Anesthesia For Thoracic Surgery Second Edition
Jonathan L. Benumof, M.D. Professor of Anesthesiology University of California, San Diego, Medical Center Department of Anesthesiology San Diego, California
W.B. SAUNDERS COMPANY A Division of Harcourt Brace & Company Philadelphia London Toronto Montreal Sydney
Tokyo
W.B. SAUNDERS COMPANY A Division of Harcourt Brace ά Company The Curtis Center Independence Square West Philadelphia, Pennsylvania 19106
Library of Congress Cataloging-in-Publication Data Benumof, Jonathan Anesthesia for thoracic surgery / Jonathan L. Benumof.—2nd ed. p.
cm.
Includes bibliographical references and index. ISBN
0-7216-4467-8
1. Chest—Surgery. 2. Anesthesia. I. Title. [DNLM: 1. Anesthesia. 2. Thoracic Surgery. 1994]
WF 980 B478a
RD536.B46 1995 617.9'6754—dc20 DNLM/DLC
93-^2770
ANESTHESIA FOR THORACIC SURGERY, SECOND EDITION
ISBN 0-7216-4467-8
Copyright © 1995, 1987 by W.B. Saunders Company All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Printed in the United States of America. Last digit i s the print number:
9
8
7
6
5
4
3
2
1
Contents
CHAPTER 1 History
of
Anesthesia
for
Thoracic
Surgery
/
Introduction, 2 Where Are We Today? Brief Summary of Modern Anesthetic Practice for Thoracic Surgery, 2 How Did We Get There? Historical Evolution of Modern Anesthetic Practice for Thoracic Surgery, 4 The Future, 12 Summary, 13
CHAPTER 2 Thoracic Anatomy
15
Introduction, 16 The Thoracic Wall, 16 The Airway and Lung, 21 The Mediastinum, 38
CHAPTER 3 General Respiratory Physiology and Respiratory Function During Anesthesia
43
Respiratory Physiology, 44 Respiratory Function During Anesthesia, 94
CHAPTER 4 Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
123
Introduction, 124 Physiology of Spontaneous Ventilation With an Open Chest, 124 Physiology of the Lateral Decubitus Position and the Open Chest During Controlled Two-Lung Ventilation: Distribution of Perfusion (Q) and Ventilation (V), 125 Physiology of One-Lung Ventilation, 131
CHAPTER 5 Preoperative Cardiopulmonary Evaluation
152
Introduction, 153 Tumors and Other Masses of the Lung and Bronchi, 153 Mediastinal Masses, 198 Pleural Disease/Effusion, 201 Pericardial Disease/Effusion, 203
CHAPTER 6 Preoperative Respiratory Preparation
211
Introduction, 212 Correlation of Respiratory Complications with Degree of Pre-Existing Lung Disease, 213 Correlation of Respiratory Complications With Site of Operation, 213 Proof That Preoperative Pulmonary Preparation Decreases Incidence of Postoperative Respiratory Complications, 213 Preoperative Respiratory Preparation Maneuvers, 214 Mechanism of Preoperative Respiratory Preparation Benefit, 227 Premedication, 227
CHAPTER 7 Monitoring
232
Introduction, 233 Tier I: Essential Monitoring System, 234 Tier II: Special Intermittent and/or Continuous Monitoring, 252 Tier III: Advanced Monitoring Techniques, 256
xi
CHAPTER 8
Choice
of
Anesthetic
Drugs
and
Techniques
300
Introduction, 301 Most Common and Important Cardiopulmonary Considerations for Patients Undergoing Thoracic Surgery, 301 Choice of Anesthesia and Arterial Oxygenation During One-Lung Ventilation, 308 Recommended Anesthesia Induction and Maintenance Drugs and Techniques, 320
CHAPTER 9
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
330
Introduction, 331 Indications for Separation of the Two Lungs, 331 Double-Lumen Tube Intubation, 334 Bronchial Blockers (With Single-Lumen Endotracheal Tubes), 370 Endobronchial Intubation With Single-Lumen Tubes, 383
CHAPTER 10
Routine Surgical Considerations that have Anesthetic Implications
390
Introduction, 391 Positioning the Patient, 391 Thoracic Incisions, 394 Common Major Elective Thoracic Operations, 402
CHAPTER 1 1
Conventional and Differential Lung Management of One-Lung Ventilation
406
Introduction, 407 One-Lung Ventilation Situation, 407 Conventional Management of One-Lung Ventilation, 408 Differential Lung Management of One-Lung Ventilation, 413 Recommended Combined Conventional and Differential Lung Management of One-Lung Ventilation, 426
CHAPTER 12
High-Frequency and High-Flow Apneic Ventilation During Thoracic Surgery ...
432
Introduction, 433 High-Frequency Ventilation, 433 Low- and High-Flow Apneic Ventilation, 445
CHAPTER 13
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
453
Introduction, 454 Management of Bronchospasm, 454 Management of Blood Loss, 459 Treatment of Nonblood Loss, Deleterious Hemodynamic Changes, 472 Re-Expansion of Collapsed Lung During and at the End of Thoracic Surgery, 479 Transport of Patient, 480
CHAPTER 14
Anesthesia for Special Elective Diagnostic Procedures
491
Introduction, 492 Bronchoscopy, 492 Mediastinoscopy, 504 Thoracoscopy, 508
CHAPTER 15
Anesthesia for Special Elective Therapeutic Procedures Introduction, 514 Laser Resection of Major Airway Obstructing Tumors, 514 Tracheal Resection, 530 Broncholithiasis, 541 Giant Bullous Emphysema and Air Cysts, 542 Pulmonary Resection in Patients After Pneumonectomy, 548 Unilateral Bronchopulmonary Lavage, 548 Pulmonary Arteriovenous Malformations, 554 Lung Transplantation, 556 Tumors at the Confluence of the Superior, Anterior, and Middle Mediastina, 567 Repair of Thoracic Aortic Aneurysms, 575
513
Contents Thymectomy for Myasthenia Gravis, 575 One-Lung Anesthesia in Morbidly Obese Patients, 579 Thoracic Outlet Syndromes, 580
CHAPTER 16 Anesthesia
for
Esophageal
Surgery
591
Preoperative Considerations, 592 Intraoperative Anesthetic Considerations, 604 Postoperative Considerations, 609
CHAPTER 17 Anesthesia
for
Emergency
Thoracic
Surgery
612
Surgery
657
Introduction, 613 Massive Hemoptysis, 613 Thoracic Aortic Aneurysms and Dissections/Disruptions, 619 Bronchopleural Fistula, 626 Lung Abscesses and Empyema, 631 Chest Trauma, 633 Transvenous Pulmonary Embolectomy, 645 Emergency Room Thoracotomy in the Management of Trauma, 646 Removal of Tracheobronchial Tree Foreign Bodies, 648
CHAPTER 18 Anesthesia
for
Pediatric
Thoracic
Introduction, 659 Special Problems Related to Premature and Newborn Infants, 659 Congenital Diaphragmatic Hernia, 669 Esophageal Atresia and Tracheoesophageal Fistula, 674 Ligation of Patent Ductus Arteriosus in Premature Infants, 681 Vascular Rings, 682 Congenital Parenchymal Lesions: Lobar Emphysema, Cysts, Sequestrations, and Cystic Adenomatoid Malformations, 683 Thoracic Surgical Procedures Requiring One-Lung Ventilation, 686 Diagnostic Bronchoscopy, 688 Bronchography, 690 Asphyxiating Thoracic Dystrophy (Jeune's Syndrome), 691
CHAPTER 19 Early Serious Complications Specifically Related to Thoracic Surgery
696
Introduction, 697 Herniation of the Heart, 697 Pulmonary Torsion, 699 Major Hemorrhage, 699 Bronchial Disruption, 700 Respiratory Failure, 703 Unilateral Re-Expansion (Intraoperative)Pulmonary Edema, 707 Right-Sided Heart Failure, 708 Myocardial Ischemia/Infarction, 710 Arrhythmias, 710 Right-to-Left Shunting Across a Patent Foramen Ovale, 712 Massive Postpneumonectomy Malignant Pleural Effusion and Chylothorax, 713 Systemic Tumor Embolism, 714 Neural Injuries, 714 Complications of Intrathoracic Intercostal Nerve Blocks, 715
CHAPTER 20 Mechanical Ventilation and Weaning
720
Introduction, 721 Initial Ventilator Settings: Intermittent Mandatory Ventilation, 721 Goal # 1 : F,02 5 L/min) minimize the problem with almost all systems for almost all patients.
c. INCREASED CARBON DIOXIDE PRODUCTION
All of the causes of increased 02 consumption will also increase C0 2 production, namely, hyperthermia, shivering, catecholamine release (light anesthesia), hypertension, and thyroid storm. If minute ventilation, total dead space, and ventilation-perfusion relationships are constant, an increase in C0 2 production will result in hypercapnia.
2. Hypocapnia The mechanisms of hypocapnia are the reverse of those that produce hypercapnia. Thus, with all other factors being equal, hyperventilation (spontaneous or controlled ventilation), decreased deadspace ventilation (change from mask airway to endotracheal tube airway, decreased PEEP, increased pulmonary artery pressure, or decreased
110
General Respiratory Physiology and Respiratory Function During Anesthesia
rebreathing), and decreased C0 2 production (hypothermia, deep anesthesia, hypotension) will lead to hypocapnia. By far, the most common mechanism of hypocapnia is passive hyperventilation by mechanical means. H. Physiologic Effects of Abnormalities in the Respiratory Gases 1. Hypoxia
The end-products of aerobic metabolism (oxidative phosphorylation) are carbon dioxide and water, both of which are easily diffusible and lost from the body. The essential feature of hypoxia is the cessation of oxidative phosphorylation when mitochondrial Po2 falls below a critical level. Anaerobic pathways, which produce energy (ATP) inefficiently, are then utilized. The main anaerobic metabolites are hydrogen and lactate ions, which are not easily excreted. They accumulate in the circulation where they may be quantified in terms of the base deficit and the lactate-pyruvate ratio. Since the various organs have different blood flow and oxygen consumption rates, the presentation and clinical diagnosis of hypoxia are usually related to symptoms arising from the most vulnerable organ. This is usually the brain in an awake patient and the heart in the anesthetized patient (see the following), but in special circumstances it may be the spinal cord (aortic surgery), kidney (acute tubular necrosis), liver (hepatitis), or limb (claudication, gangrene). The cardiovascular response to acute hypoxemia is a product of both reflex (neural and humoral) and direct effects (Table 3-9).336-338 The reflex effects occur first and are excitatory and vasoconstrictive. The neuroreflex effects result from aortic and carotid chemoreceptor, baroreceptor, and central cerebral stimulation, and the humoral reflex effects result from catecholamine and renin-angiotensin release. The direct local vascular effects of hypoxia are inhibitory and vasodilatory and occur late. The net response to hypoxia in a subject de-
pends on the severity of the hypoxia; the severity of hypoxia determines the magnitude of and balance between the inhibitory and excitatory components; the balance may vary according to the type and depth of anesthesia and the degree of preexisting cardiovascular disease. Mild arterial hypoxemia (arterial saturation less than normal but still 80 per cent or higher) causes a general activation of the sympathetic nervous system and release of catecholamines. Consequently, heart rate, mean circulatory pressure (and index of venous tone and cardiac preload), systemic blood pressure, stroke volume, cardiac output, and myocardial contractility (as measured by a shortened pre-ejection period [PEP], left ventricular ejection time [LVET], and a decreased PEP/LVET ratio) are increased (Fig. 3-52).339 In healthy, awake humans, most of the increase in cardiac output with mild isocapnic hypoxemia is due to the increase in heart rate.340 Changes in systemic vascular resistance are usually slight.336·340 However, in patients under anesthesia with beta blockers, hypoxia (and hypercapnia when present) may cause circulating catecholamines to have only an alpha-receptor effect, and the heart may be unstimulated (even depressed by a local hypoxic effect), and systemic vascular resistance may be increased. Consequently, cardiac output may be decreased in these patients. With moderate hypoxemia (arterial oxygen saturation 60 to 80 per cent), local vasodilatation begins to predominate, and systemic vascular resistance and blood pressure decrease, but heart rate may continue to be increased because of a systemic hypertension-induced stimulation of baroreceptors. However, even in halothane-anesthetized sheep, moderate hypoxemia does not cause left ventricular contractility dysfunction.341 Finally, with severe hypoxemia (arterial saturation less than 60 per cent), local depressant effects dominate, and blood pressure falls rapidly, the pulse slows, shock develops, and the heart either fibrillates or becomes asystolic. It should be remembered that significant pre-existing hypotension will convert a mild hy-
General Respiratory Physiology and Respiratory Function During Anesthesia
111
gen demand. Fourth, late systemic hypotension may decrease myocardial oxygen supply owing to decreased diastolic perfusion pressure. Fifth, coro nary blood flow reserve may be exhausted by a late maximally increased coronary blood flow (as a result of maximal coronary vasodilation).341 The level of hypoxemia that will cause cardiac arrhyth mias cannot be predicted with certainty because the myocardial oxygen supply and demand rela tionship in a given patient is not known (i.e., the degree of coronary artery atherosclerosis may not be known). However, if a myocardial area (or areas) becomes hypoxic and/or ischemic, unifocal or multifocal premature ventricular contractions, ventricular tachycardia, and ventricular fibrillation may occur.
Figure 3-52 Changes in the minute ventilation and circula tion of normal awake humans during progressive isocapnic hypoxia and hyperoxic hypercapnia. VF = expired minute ven tilation; Q = minute cardiac output; Ρ,.,Ο, = end-tidal Po 2 ; P ET CO : = end-tidal Pco 2 ; S, = slope during the first phase of slowly increasing ventilation and/or circulation; S: = slope during the second phase of sharply increasing ventilation and/or circulation. (From Serebrouskaya TV: Comparison of respiratory and circulatory human responses to progressive hy poxia and hypercapnia. Respiration 59:35-41, 1992. Used with permission.)
poxemia-hemodynamic profile into a moderate hy poxemia-hemodynamic profile, and a moderate hypoxemia-hemodynamic profile will convert into a severe hypoxemia-hemodynamic profile. Simi larly, in well-anesthetized and/or well-sedated pa tients, early sympathetic nervous system reactivity 342 to hypoxemia may be reduced, and the effects of hypoxemia may be expressed only as bradycar dia with severe hypotension and, ultimately, cir culatory collapse. Hypoxemia may also cause cardiac arrhythmias, and the cardiac arrhythmias may in turn potentiate the already mentioned deleterious cardiovascular effects. Hypoxemia-induced arrhythmias may be caused by multiple mechanisms; the mechanisms are inter-related by virtue of the fact that they all cause a decrease in the myocardial oxygen supplydemand ratio, which in turn increases myocardial irritability. First, arterial hypoxemia may directly decrease myocardial oxygen supply. Second, early tachycardia may cause an increased myocardial oxygen consumption, and a decreased diastolic filling time may cause a decreased myocardial ox ygen supply. Third, early increased systemic blood pressure may cause an increased afterload on the left ventricle, which increases left ventricular oxy
The cardiovascular response to hypoxia includes a number of other important effects. Cerebral blood flow increases (even if hypocapnic hyper ventilation is present). Ventilation will be stimu lated no matter why hypoxia exists. Not surpris ingly, acute progressive isocapnic hypoxemia causes bronchodilation in both normal subjects and in patients with chronic lung disease (includ ing asthma). 343 The pulmonary distribution of blood flow is more homogeneous owing to an in creased pulmonary pressure. The most common causes of chronic hypoxia in humans (excluding existence at high attitude) are pulmonary and/or cardiac problems that lead to insidious reduction of arterial blood and tissue ox ygen tension, such as chronic obstructive pulmo nary disease and cardiac failure. The most impor tant effects of chronic hypoxemia are hematologic (increased RBC mass) and increased minute ven tilation.344 The higher hemoglobin concentration and hematocrit found after prolonged exposure to hypoxia are believed to be caused by a sustained increase in renal release of erythropoietin, the gly coprotein hormone that stimulates the formation of RBCs. Chronic hypoxia does not affect the stan dard oxygen consumption of large mammals. To maintain a normal level of oxygen consumption, a certain degree of hyperventilation must be main tained. As hypoxemia increases, the ventilation re quired to maintain the same amount of oxygen consumed per unit of time must increase.344 The bulk of evidence shows that, at rest, cardiac output in chronic hypoxia is normal. Systemic oxygen transport (i.e., the product of 02 content times car diac output) is maintained within normal levels because the absolute increase in hemoglobin sus tains an unchanged arterial oxygen content. Al though chronic hypoxia will cause a right-shifted oxygen-hemoglobin dissociation curve (because of either an increase in 2,3-DPG or acidosis), which is an effect that increases tissue Po 2 , oxygen up take by tissue may still be impaired, possibly be cause of increased blood viscosity.
112
General Respiratory Physiology and Respiratory Function During Anesthesia
2. Hyperoxia (Oxygen Toxicity) The dangers associated with the inhalation of excessive oxygen are multiple. Exposure to a high 02 tension clearly causes pulmonary damage in healthy individuals.345·346 Because pulmonary tissue Po2 is directly determined by PA02, arterial hypoxemia does not delay the onset of oxygen toxicity at 1 atmosphere of pressure.347 A dosetime toxicity curve for humans is available from a number of studies.345 Because the lungs of normal human volunteers cannot be directly examined to determine the rate of onset and the course of toxicity, indirect measures such as the onset of symptoms have been used to construct the dose-time toxicity curves. Examination of the curve indicates that 200 to 500 per cent oxygen (2 to 5 atmospheres) should not be administered for more than 6 hours, 100 per cent oxygen should not be administered for more than 12 hours, 80 per cent oxygen should not be administered for more than 24 hours, and 60 per cent oxygen should not be administered for more than 36 hours.348 No measurable changes in pulmonary function or blood-gas exchange occur in humans during exposures to less than 50 per cent oxygen or less even for long periods.348 Nevertheless, it is important to note that in the clinical setting, these dose-time toxicity relationships are often generally obscured347 because of the complex multivariable nature of the clinical setting. The dominant first symptom of oxygen toxicity in human volunteers is substernal distress, which begins as a mild irritation in the area of the carina and may be accompanied by occasional coughing.349 As exposure continues, pain becomes more intense, and the urge to cough and to deep breathe also becomes more intense. These symptoms will progress to severe dyspnea, paroxysmal coughing, and decreased vital capacity when the F,02 has been 1.0 for greater than 12 hours. At this point, recovery of mechanical lung function usually occurs within 12 to 24 hours but may require more than 24 hours in some individuals.348 No individual measure of pulmonary function has been found to be uniquely satisfactory for monitoring rates of development or reversal of pulmonary oxygen poisoning. As toxicity progresses, pulmonary functions studies such as compliance and arterial blood gases deteriorate. Pathologically, in animals the lesion progresses from a tracheobronchitis (exposure for 12 hours to a few days) to involvement of the alveolar septa with pulmonary interstitial edema (exposure for a few days to a week) to pulmonary fibrosis of the edema (exposure greater than a week).350 Ventilatory depression may occur in those patients who, by reason of drugs or disease, have been ventilating in response to a hypoxic drive. By
definition, ventilatory depression resulting from removal of a hypoxic drive by increasing the inspired oxygen concentration will cause hypercapnia but does necessarily produce hypoxia (because of the increase in F,02). Absorption atelectasis has been previously discussed (see High Inspired Oxygen Concentration and Absorption Atelectasis) and is not covered here. Retrolental fibroplasia, an abnormal proliferation of the immature retinal vasculature of the prematurely born infant, can occur following exposure to hyperoxia. Very premature infants are most susceptible to retrolental fibroplasia (i.e., those of less than 1.0 kg birth weight and 28 weeks gestation). The risk of retrolental fibroplasia exists whenever an F,02 causes a Pa02 of > 80 mm Hg for more than 3 hours in an infant whose gestational age plus life age is less than 44 weeks. If the ductus arteriosus is patent, arterial blood samples should be drawn from the right radial artery (umbilical or lower extremity Pa02 is lower than the Pa02 to which the eyes are exposed owing to ductal shunting of unoxygenated blood [see chapter 18]). The mode of action of toxicity of oxygen in tissues is complex, but interference with metabolism seems to be widespread. Most important, there is oxygen free radical (partially reduced oxygen) inactivation of many enzymes, particularly those with sulfhydryl groups.347 Neutrophil recruitment and release of mediators of inflammation occur next and greatly accelerate the extent of endothelial and epithelial damage and impairment of the surfactant systems.347 The most acute toxic enzyme effect of oxygen in humans is a convulsive effect that occurs during exposure to pressures in excess of 2 atmospheres absolute. The free radical theory has provided the principles on which to evaluate and possibly augment tolerance to elevated oxygen pressures.351 First, various regimens of intermittent or pre-exposure to oxygen or adaptation to hypoxia have been shown, at least in some models, to augment oxygen tolerance. The apparent basis for increased tolerance is repair of toxic effects and/or augmentation of antioxidant defenses. A second approach is the manipulation of nutritional status, because dietary deficiency of antioxidant factors can decrease tolerance. These dietary components include protein to provide essential amino acids for synthesis of glutathione, selenium as a component of glutathione peroxidase, and alpha-tocopherol (vitamin E) as a radical scavenger. Vitamins C and A may also have important antioxidant roles. A third strategy for increasing oxygen tolerance is to augment the cellular antioxidant defenses. This very active area of investigation has seen several major developments. The administration of antioxidant en-
General Respiratory Physiology and Respiratory Function During Anesthesia
zymes has been evaluated. Liposomes and RBCs have been tried as delivery systems to promote entry of exogenous enzymes into cellular compartments. The administration of sulfhydryl antioxidants has also shown promise with development of newer agents that are internalized by cells. For the future, genetic engineering may offer a more suitable method for altering antioxidant defenses. A final strategy for extending tolerance is to treat the toxic manifestations of oxygen to prevent death or disability. Although this approach is clearly less appealing than the methods discussed previously, it may have the most immediate practicality for extending oxygen tolerance. Possible therapies include administration of exogenous surfactant or use of anti-inflammatory and antifibrotic drugs. High inspired oxygen concentrations can be of use therapeutically. Clearance of gas loculi in the body may be greatly accelerated by inhalation of 100 per cent oxygen. Inhalation of 100 per cent oxygen creates a large N2 gradient from the gas space to the perfusing blood. As a result, N2 leaves the gas space, and the space diminishes in size. The technique of using oxygen to remove gas may be used to ease intestinal gas pressure in patients with intestinal obstruction, to hasten recovery from pneumoencephalography, to decrease the size of an air embolus, and to aid in absorption of pneumoperitoneum and pneumothorax.
3. Hypercapnia The effects of carbon dioxide on the cardiovascular system are as complex as they are for hypoxia. As with hypoxemia, hypercapnia appears to cause direct depression of both the cardiac muscle and the vascular smooth muscle. At the same time, however, it causes reflex stimulation of the sympathoadrenal system, which compensates to a
113
greater or lesser extent for the primary cardiovascular depression.336·341 With moderate hypercapnia, a hyperkinetic circulation results, with increased cardiac output and systemic blood pressure (Fig. 3-52). 339 · 352 Even in patients under halothane anesthesia, plasma catecholamine levels increase in response to increased C0 2 levels in much the same way as they do in conscious subjects. Thus, hypercapnia, like hypoxemia, may cause increased myocardial oxygen demand (tachycardia, early hypertension) and decreased myocardial oxygen supply (tachycardia, late hypotension). Table 3-10 summarizes the interaction of anesthesia with hypercapnia in humans: The increased cardiac output is maintained during anesthesia, but the systemic blood pressure and vascular resistance are decreased.352·353 The increase in cardiac output is most marked during anesthesia with drugs and enhances sympathetic activity and is least marked with halothane and nitrous oxide. The decrease in systemic vascular resistance is most marked during anesthesia with enflurane and accompanying hypercapnia. Arrhythmias have been reported in unanesthetized humans during acute hypercapnia but have seldom been of serious importance. A high P a CO : level, however, is more dangerous during general anesthesia, and with halothane anesthesia arrhythmias will frequently occur above a P a C0 2 arrhythmic threshold, which is often constant for a particular patient. The maximal stimulant respiratory effect is attained by a P a C0 2 of about 100 mm Hg. With a higher P a C0 2 , stimulation is reduced, and at very high levels respiration is depressed and later ceases altogether. The Pco2-ventilation response curve is generally displaced to the right, and its slope is reduced by anesthetics and other depressant drugs.354 With profound anesthesia the response curve may be flat, or even sloping downward, and
114
General Respiratory Physiology and Respiratory Function During Anesthesia
carbon dioxide then acts as a respiratory depressant. In patients with ventilatory failure, carbon dioxide narcosis occurs when the P a C0 2 rises above 90 to 120 mm Hg. Thirty per cent carbon dioxide is sufficient for the production of anesthesia, and this concentration causes total but reversible flattening of the electroencephalogram.355 As expected, hypercapnia causes bronchodilation in normal and lung disease patients.341 Quite apart from the effect of carbon dioxide on ventilation, hypercapnia exerts three other important effects that influence the oxygenation of the blood and tissue. First, if the concentration of nitrogen (or other "inert" gas) remains constant, the concentration of C0 2 in the alveolar gas can only increase at the expense of 0 2 , which must be displaced. Thus, P A 0 2 and P a 0 2 may decrease. Second, hypercapnia shifts the oxygen-hemoglobin curve to the right, facilitating tissue oxygenation. Third, hypercapnia may depress spontaneous diaphragmatic function.356 Chronic hypercapnia results in increased resorption of bicarbonate by the kidneys, further raising the plasma bicarbonate level and constituting a secondary or compensatory "metabolic alkalosis." Chronic hypocapnia decreases renal bicarbonate resorption, resulting in further fall of plasma bicarbonate and producing a secondary or compensatory "metabolic acidosis." In each case, arterial pH returns toward the normal value, but the bicarbonate ion concentration departs even further from normal. Hypercapnia is accompanied by a leakage of potassium from the cells into the plasma. A good deal of the potassium comes from the liver, probably from glucose release and mobilization, which occurs in response to the rise in plasma catecholamine levels.357 Since the plasma potassium level takes an appreciable time to return to normal, repeated bouts of hypercapnia at short intervals result in a stepwise rise in plasma potassium.
4. Hypocapnia In this section, hypocapnia is considered to be produced by passive hyperventilation (by the anesthesiologist or ventilator). Hypocapnia may cause a decrease in the cardiac output by three separate mechanisms. First, if present, an increase in intrathoracic pressure will decrease the cardiac output. Second, hypocapnia is associated with a withdrawal of sympathetic nervous system activity, and this can decrease the ionotropic state of the heart. Third, hypocapnia can increase pH, which can in turn decrease ionized Ca + + , which may in turn decrease the ionotropic state of the heart. Hypocapnia with an alkalosis will also shift the oxygen-hemoglobin curve to the
left, which increases the hemoglobin affinity for 0 2 , which will impair O, unloading at the tissue level. The decrease in peripheral flow and impaired ability to unload oxygen to the tissues is compounded by an increase in whole body oxygen consumption caused by an increased pH-mediated uncoupling of oxidation from phosphorylation357; P a C0 2 of 20 mm Hg will increase tissue 02 consumption by 30 per cent. Consequently, hypocapnia may simultaneously increase tissue 02 demand and decrease tissue 02 supply. Thus, in order to have the same amount of 02 delivery to the tissues, cardiac output or tissue perfusion has to increase at a time when it may not be possible to do so. It has been suggested that the cerebral effects of hypocapnia may be related to a state of cerebral acidosis and hypoxia, since hypocapnia may cause a selective reduction in the cerebral blood flow and also shifts the oxygen-hemoglobin curve to the left. Hypocapnia may cause VA/Q abnormalities by inhibiting HPV or by causing bronchoconstriction and a decreased lung compliance. Finally, passive hypocapnia will produce apnea.
REFERENCES 1. West JB, Dollery CT, Naimark A: Distribution of blood flow in isolated lung: Relation to vascular and alveolar pressures. J Appl Physiol 19:713, 1964. 2. Puri GD, Venkataranan RK, Singh H, Jindal SK: Physiological dead space and arterial to end-tidal CO\ difference under controlled normocapnic ventilation in young anaesthetized subjects. Indian J Med Res 94:41-46, 1991. 3. Permutt S, Bramberger-Barnea B, Bane HN: Alveolar pressure, pulmonary venous pressure and the vascular waterfall. Med Thorac 19:239, 1962. 4. West JB, Dollery CT, Heard BE: Increased pulmonary vascular resistance in the dependent zone of the isolated dog lung caused by perivascular edema. Circ Res 17:191206, 1965. 5. West JB (ed): Regional Differences in the Lung. New York, Academic Press, 1977. 6. Hughes JMB, Glazier JB, Maloney JE, et al: Effect of lung volume on the distribution of pulmonary blood flow in man. Respir Physiol 4:58-72, 1968. 7. Hughes JM, Glazier JB, Maloney JE, et al: Effect of extra-alveolar vessels on the distribution of pulmonary blood flow in the dog. J Appl Physiol 25:701-709, 1968. 8. Permutt S, Caldini P, Maseri A, et al: Recruitment versus distensibility in the pulmonary vascular bed. In Fishman AP, Hecht H (eds): The Pulmonary Circulation and Interstitial Space. Chicago, University of Chicago Press, 1969, pp 375-387. 9. Maseri A, Caldini P, Harward P, et al: Determinants of pulmonary vascular volume. Recruitment versus distensibility. Circ Res 31:218-228, 1972. 10. Hakim TS, Lisbona R, Dean GW: Gravity-independent inequality in pulmonary blood flow in humans. J Appl Physiol 63:1114-1121, 1987. 11. Hoppin FG Jr, Green ID, Mead J: Distribution of pleural surface pressure. J Appl Physiol 27:863, 1969. 12. Milic-Emili J, Henderson JAM, Dolovich MB, et al: Regional distribution of inspired gas in the lung. J Appl Physiol 21:749, 1966.
τ CHAPTER
4
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation I. Introduction II. Physiology of Spontaneous Ventilation With an Open Chest A. Mediastinal Shift B. Paradoxical Respiration III. Physiology of the Lateral Decubitus Position and the Open Chest During Controlled Two-Lung Ventilation: Distribution of Perfusion (Q) and Ventilation (V) A. Distribution of Q, V, V/Q—Lateral Decubitus Position, Awake, Closed Chest B. Distribution of Q, V—Lateral Decubitus Position, Anesthetized, Closed Chest C. Distribution of Q, V—Lateral Decubitus Position, Anesthetized, Open Chest D. Distribution of Q, V—Lateral Decubitus Position, Anesthetized, Open Chest, Paralyzed E. Summary of Physiology of Lateral Decubitus Position and the Open Chest IV. Physiology of One-Lung Ventilation A. Comparison of Arterial Oxygenation and C0 2 Elimination During TwoLung Versus One-Lung Ventilation B. Blood Flow Distribution During OneLung Ventilation -.
Blood Flow to the Nondependent, Nonventilated Lung 1. Distribution of Hypoxia 2. Low V/Q Versus Atelectasis 3. Vasodilator Drugs 4. Anesthetic Drugs 5. Cardiac Output and Pulmonary Vascular Pressure
6. P A 7. F,02 in the Normoxic Compartment 8. Vasoconstrictor Drugs 9. Drugs That Increase HPV (a) Almitrine (b) Products of Arachidonic Acid Metabolism (Vasoconstrictor Leukotrienes) 10. P A C0 2 and pH Changes 11. PEEP and Alveolar Pressure 12. Other Important Systemic Factors (Hypertension, Age, Sex, Infection) Blood Flow to the Dependent, Ventilated Lung Miscellaneous Causes of Hypoxemia During One-Lung Ventilation
I
123
124
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
I. INTRODUCTION It is not possible to have adequate gas exchange during spontaneous ventilation with an open chest because of the occurrence of mediastinal shift and paradoxical respiration (see the "pneumothorax problem" discussed in chapter 1). Consequently, intrathoracic surgery cannot be performed with spontaneous ventilation. The first part of this chapter briefly reviews this physiology. Patients undergoing thoracic surgery are usually in the lateral decubitus position under general anesthesia, have an open chest wall (nondependent hemithorax), are pharmacologically paralyzed, and, of course, have ventilation controlled. Even if both lungs are being ventilated, each of these anesthesia and surgical requirements can cause major alterations in the distribution of perfusion (Q), and/or ventilation (V), and ventilation-perfusion relationships (V/Q) (compared with the awake state in the upright position). The second part of this chapter discusses the physiologic effects of each one of these anesthetic and surgical events on the distribution of Q, V, and V/Q during two-lung ventilation. In addition, a good deal of thoracic surgery must be performed in the lateral decubitus position with the nondependent lung nonventilated and the dependent lung ventilated (one-lung ventilation). One-lung ventilation imposes a new host of determinants on the distribution of blood flow (and, of course, ventilation). The third, and by far the largest, part of this chapter considers the physiology of one-lung ventilation.
lung, contributing to the pleural pressure gradient. With the nondependent hemithorax open, atmospheric pressure in that cavity exceeds the negative pleural pressure in the dependent hemithorax; this imbalance of pressure on the two sides of the mediastinum causes a further downward displacement of the mediastinum into the dependent thorax. During inspiration, the caudad movement of the dependent-lung diaphragm increases the negative pressure in the dependent lung and causes a still further displacement of the mediastinum into the dependent hemithorax. During expiration, as the dependent-lung diaphragm moves cephalad, the pressure in the dependent hemithorax becomes relatively positive, and the mediastinum is pushed upward out of the dependent hemithorax (Fig. 4l). 2 Thus, the tidal volume in the dependent lung is decreased by an amount equal to the inspiratory displacement caused by mediastinal movement. This phenomenon is called mediastinal shift and is one mechanism that results in impaired ventilation in the open-chest spontaneously breathing patient in the lateral decubitus position. The mediastinal shift can also cause circulatory changes (decreased venous return) and reflexes (sympathetic activation) that result in a clinical picture similar to shock: The patient is hypotensive, pale, and cold, with dilated pupils. Local anesthetic infiltration of the pulmonary plexus at the hilum and the vagus nerve can diminish these reflexes. More practically, controlled positive-pressure ventilation abolishes these ventilatory and circulatory changes associated with mediastinal shift.
B. Paradoxical Respiration II. PHYSIOLOGY OF SPONTANEOUS VENTILATION WITH AN OPEN CHEST A. Mediastinal Shift An examination of the physiology of the open chest during spontaneous ventilation reveals why controlled positive-pressure ventilation is the only practical way to provide adequate gas exchange during thoracotomy. (However, it should be noted that spontaneous ventilation may be indicated in patients with a large mediastinal mass.1) In the spontaneously breathing, closed-chest patient in the lateral decubitus position, gravity causes the pleural pressure in the dependent hemithorax to be less negative than that in the nondependent hemithorax (see Fig. 3-5), but there is still negative pressure in each hemithorax on each side of the mediastinum. In addition, the weight of the mediastinum causes some compression of the lower
When a pleural cavity is exposed to atmospheric pressure, the lung is no longer held open by negative intrapleural pressure, and it tends to collapse because of unopposed elastic recoil. Thus, the lung in an open chest is at least partially collapsed. It has long been observed during spontaneous ventilation with an open hemithorax that lung collapse is accentuated during inspiration, and, conversely, the lung expands during expiration. This reversal of lung movement with an open chest during respiration has been termed paradoxical respiration. The mechanism of paradoxical respiration is similar to that of mediastinal shift. During inspiration, the descent of the diaphragm on the side of the open hemithorax causes air from the environment to enter the pleural cavity on that side through the thoracotomy opening and fill the space around the exposed lung. The descent of the hemidiaphragm on the closed-chest side causes gas to enter the closed-chest lung in the normal manner. However,
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
125
Spontaneous Ventilation with Open Chest Mediastinal Shift
Paradoxical Respiration Inspiration
/
Figure 4-1 Schematic representation of mediastinal shift and paradoxical respiration in the spontaneously ventilating patient with an open chest and placed in the lateral decubitus position. The open chest is always exposed to atmospheric pressure (Θ). During inspiration, negative pressure ( θ ) in the intact hemithorax causes the mediastinum to move vertically downward (medias tinal shift). In addition, during inspiration movement of gas from the nondependent lung in the open hemithorax into the dependent lung in the closed hemithorax and movement of air from the environment into the open hemithorax causes the lung in the open hemithorax to collapse (paradoxical respiration). During expiration, relative positive pressure (Θ) in the closed hemithorax causes the mediastinum to move vertically upward (mediastinal shift). In addition, during expiration the gas moves from the dependent lung to the nondependent lung and from the open hemithorax to the environment; consequently, the nondependent lung expands during expiration (paradoxical respiration).
gas also enters the closed-chest lung (which has a relatively negative pressure) from the open-chest lung (which remains at atmospheric pressure); this results in further reduction in the size of the openchest lung during inspiration. During expiration the reverse occurs, with the collapsed, open-chest lung filling from the intact lung and air moving back out of the exposed hemithorax through the thoracotomy. The phenomenon of paradoxical res piration is illustrated in Figure 4-1. 2 Paradoxical breathing is increased by a large thoracotomy and by increased airway resistance in the intact lung. Paradoxical respiration may be prevented by either manual collapse of the open-chest lung or, more commonly, controlled positive-pressure ventila tion. In summary, in a spontaneously breathing patient in the lateral decubitus position with an open, nondependent chest, the mediastinum will move away from the observer, the nondependent lung will decrease in size during inhalation, the mediastinum will move toward the observer, and the nondependent lung will increase in size during exhalation.
III. PHYSIOLOGY OF THE LATERAL DECUBITUS POSITION AND THE OPEN CHEST DURING CONTROLLED TWO-LUNG VENTILATION: DISTRIBUTION OF PERFUSION (Q) AND VENTILATION (V) A. Distribution of Q, V, V/Q—Lateral Decubitus Position, Awake, Closed Chest Gravity causes a vertical gradient in the distri bution of pulmonary blood flow in the lateral de cubitus position for the same reason that it does in the upright position (see Fig. 3-5). Since the ver tical hydrostatic gradient is less in the lateral de cubitus position than in the upright position, there is ordinarily less zone 1 blood flow (in the nonde pendent lung) in the former position compared with the latter position (Fig. 4-2). Nevertheless, blood flow to the dependent lung is still signifi cantly greater than blood flow to the nondependent
126
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
Distribution of Blood Flow Lateral Decubitus Position
Figure 4-2 Schematic representation of the effects of gravity on the distribution of pulmonary blood flow in the lateral decubitus position. The vertical gradient in the lateral decubitus position is less than that in the upright position (see Fig. 3-1). Consequently, there is less zone 1 and more zone 2 and zone 3 blood flow in the lateral decubitus position compared with the upright position. Nevertheless, pulmonary blood flow increases with lung dependency and is greater in the dependent lung compared with the nondependent lung. (PA = alveolar pressure; Ppa = pulmonary artery pressure; Ppv = pulmonary venous pressure.) (Modified with permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Livingstone Inc., 1983, chapter 8.)
lung (Fig. 4-2). Thus, when the right lung is nondependent, it should receive approximately 45 per cent of total blood flow as opposed to the 55 per cent of the total blood flow that it receives in the upright and supine positions.3,4 When the left lung is nondependent, it should receive approximately 35 per cent of total blood flow as opposed to the 45 per cent of the total blood flow that it receives in the upright and supine positions.3,4 Because gravity also causes a vertical gradient in pleural pressure (Ppl) in the lateral decubitus position (as it does in the upright position, see Figs. 3-4 and 3-5), ventilation is relatively increased in the dependent lung compared with the nondependent lung (Fig. 4-3). In addition, in the lateral decubitus position, the dome of the lower diaphragm is pushed higher into the chest than the dome of the upper diaphragm, and therefore the lower diaphragm is more sharply curved than the upper diaphragm. As a result, the lower diaphragm is able to contract more efficiently during spontaneous respiration. Thus, in the lateral decubitus position in the awake patient, the lower lung is
normally better ventilated than the upper lung, regardless of the side on which the patient is lying, although there remains a tendency toward greater ventilation of the larger right lung.5 Since there is greater perfusion to the lower lung, the preferential ventilation to the lower lung is matched by its increased perfusion, so that the distribution of the ventilation-perfusion ratios of the two lungs is not greatly altered when the awake subject assumes the lateral decubitus position. Since the increase in ventilation is less than the increase in perfusion with lung dependency, the V/Q ratio decreases from dependent to nondependent lung (just as it does in upright and supine lungs). B. Distribution of Q, V—Lateral Decubitus Position, Anesthetized, Closed Chest In comparing the awake patient with the anesthetized patient in the lateral decubitus position, there is no difference in the distribution of pulmo-
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
127
Awake, Closed Chest Distribution of Ventilation Upright Position
Lateral Decubitus Position
Figure 4-3 Pleural pressure (Ppl) in the awake upright patient (A) is most positive in the dependent portion of the lung, and alveoli in this region are therefore most compressed and have the least volume (see Fig. 3-5). Pleural pressure is least positive (most negative) at the apex of the lung, and alveoli in this region are therefore least compressed and have the largest volume. When these regional differences in alveolar volume are translated over to a regional transpulmonary pressure-alveolar volume curve, the small, dependent alveoli are on a steep (large-slope) portion of the curve, and the large, nondependent alveoli are on a flat (smallslope) portion of the curve (see also Fig. 3-6). In this diagram regional slope equals regional compliance. Thus, for a given and equal change in transpulmonary pressure, the dependent part of the lung receives a much larger share of the tidal volume than the nondependent part of the lung. In the awake patient in the lateral decubitus position (B) gravity also causes pleural pressure gradients and therefore similarly affects the distribution of ventilation. The dependent lung lies on a relatively steep portion, and the upper lung lies on a relatively flat portion of the pressure-volume curve. Thus, in the lateral decubitus position the dependent lung receives the majority of the tidal ventilation. (V — alveolar volume; Ρ = transpulmonary pressure.) (Modified with permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York. Churchill Livingstone Inc., 1983, chapter 8.)
nary blood flow between the dependent and nondependent lungs. Thus, in the anesthetized patient, the dependent lung continues to receive relatively more perfusion than the nondependent lung. The induction of general anesthesia, however, does cause significant changes in the distribution of ventilation between the two lungs. In the lateral decubitus position, the majority of ventilation is switched from the dependent lung in the awake subject to the nondependent lung in the 6 7 anesthetized patient (Fig. 4-4). · In fact, in the lateral decubitus position, the nondependent lung will receive approximately 55 per cent of each 8 tidal volume. There are several interrelated rea sons for this change in the relative distribution of ventilation between the nondependent and depen dent lung. First, the induction of general anesthesia usually causes a decrease in functional residual capacity (FRC), and both lungs share in the loss of lung volume. Since each lung occupies a differ ent initial position on the pulmonary pressure-vol ume curve while the subject is awake, a general anesthesia-induced reduction in the FRC of each lung causes each lung to move to a lower but still different portion of the pressure-volume curve (Fig. 4-4). The dependent lung moves from an initially steep part of the curve (with the subject
awake) to a lower and flatter part of the curve (after anesthesia is induced), while the nondepen dent lung moves from an initially flat portion of the pressure-volume curve (with the subject awake) to a lower and steeper part of the curve (after anesthesia is induced). In fact, in the lateral decubitus position, the ratio of nondependent to dependent lung FRC is approximately 1.5 (average values are 1400 ml/900 ml) in the adult patient.* Similar findings may be expected in children.4 Thus, with the induction of general anesthesia, the lower lung moves to a less favorable (flat, noncompliant) portion and the upper lung to a more favorable (steep, compliant) portion of the pres sure-volume curve. In fact, in the lateral decubitus position, the compliance of the nondependent and dependent lung is 30 and 23 cm H 2 0, respec tively.8 Second, if the anesthetized patient in the lateral decubitus position is also paralyzed and ar tificially ventilated, the high curved diaphragm of the lower lung no longer confers any advantage in ventilation (as it does in the awake state), since it 10 is no longer actively contracting. Third, the me diastinum rests on the lower lung and physically impedes lower lung expansion as well as selec tively decreases lower lung FRC. Fourth, the weight of the abdominal contents pushing cepha-
128
Special Respiratory Physiology of the Uiteral Decubitus Position, the Open Chest, and One-Lung Ventilation
Closed Chest, Lateral Decubitus Position Distribution of Ventilation Awake
Anesthetized
Figure 4-4 This schematic diagram shows the distribution of ventilation in the awake patient in the lateral decubitus position (A) and the distribution of ventilation in the anesthetized patient in the lateral decubitus position (B). The induction of anesthesia has caused a loss in lung volume in both lungs, with the nondependent lung moving from a flat, noncompliant portion to a steep, compliant portion of the pressure-volume curve and the dependent lung moving from a steep, compliant part to a flat, noncompliant part of the pressure-volume curve. Thus, the anesthetized patient in a lateral decubitus position has the majority of the tidal ventilation in the nondependent lung (where there is the least perfusion) and the minority of the tidal ventilation in the dependent lung (where there is the most perfusion). (V = alveolar volume; Ρ = transpulmonary pressure.) (Modified with permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Livingstone Inc., 1983, chapter 8.)
lad against a passive, flaccid, paralyzed diaphragm is greatest in the dependent lung, which physically impedes lower lung expansion the most and disproportionately decreases lower lung FRC. Fi nally, suboptimal positioning, which fails to pro vide room for lower lung expansion, may consid erably compress the dependent lung. Opening the nondependent hemithorax further disproportion ately increases ventilation to the nondependent lung (see the following discussion). Thus, in com paring the supine position to the lateral decubitus position in both spontaneously breathing and me chanically ventilated children and adults, the lung that is made the nondependent lung has an in crease in FRC, compliance, and relative ventila tion, and the lung that is made the dependent lung has a decrease in FRC, compliance, and relative ventilation.8·9· " The change in each lung ventila tion is proportional to the change in each lung volume.8,9· " In summary, the anesthetized patient, with or without paralysis, in the lateral decubitus position and with a closed chest has a nondependent lung that is well ventilated but poorly perfused and has a dependent lung that is well perfused but poorly ventilated, which results in an increased degree of mismatching of ventilation and perfusion. The ap plication of positive end-expiratory pressure (PEEP) to both lungs restores the majority of ven
tilation to the lower lung.4 I2 Presumably, the lower lung returns to a steeper, more favorable part of the pressure-volume curve, and the upper lung resumes its original position on a flat, unfa vorable portion of the curve.
C. Distribution of Q, V—Lateral Decubitus Position, Anesthetized, Open Chest Compared with the condition of the anesthe tized, closed-chest patient in the lateral decubitus position, opening the chest wall and pleural space alone does not ordinarily cause any significant al teration in the partitioning of pulmonary blood flow between the dependent and nondependent lungs; thus, the dependent lung continues to re ceive relatively more perfusion than the nondepen dent lung. However, if the compliance of the nondependent lung increases so much (see discussion immediately following) that nondependent lung vascular resistance decreases significantly, then nondependent-lung blood flow may increase rela tive to dependent-lung blood flow. In addition, the vertical distance between the heart and the nonde pendent lung may be decreased, which in the face of a constant pulmonary artery pressure might, in theory, result in an increased perfusion of the non-
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
dependent lung.13 Opening the chest wall causes only very minor alterations in pulmonary and sys temic vascular pressures and cardiac output. 13 Opening the chest wall and pleural space, how ever, does have a significant impact on the distri bution of ventilation (which must now be deliv ered by positive pressure). The change in the distribution of ventilation may result in a further mismatching of ventilation with perfusion (Fig. 4-5). 14 If the upper lung is no longer restricted by a chest wall and the total effective compliance of that lung is equal to that of the lung parenchyma alone, it will be relatively free to expand and will consequently be overventilated (and remain underperfused). Conversely, the dependent lung may continue to be relatively noncompliant and poorly ventilated and overperfused.3 From a practical point of view, it is necessary to mention that sur gical retraction and compression of the exposed upper lung can provide a partial, although nonphysiologic, solution to this problem in that if ex pansion of the exposed lung is mechanically or externally restricted, ventilation will be diverted to the dependent, better-perfused lung.13
D. Distribution of Q, V—Lateral Decubitus Position, Anesthetized Open Chest, Paralyzed In the open-chest anesthetized patient in the lat eral decubitus position, the induction of paralysis alone does not cause any significant alteration in
129
the partitioning of pulmonary blood flow between the dependent and nondependent lungs. Thus, the dependent lung continues to receive relatively more perfusion than the nondependent lung. There are, however, strong theoretical and experimental considerations that indicate that paralysis might cause significant changes in the distribution of ventilation between the two lungs under these con ditions. In the supine and lateral decubitus positions, the weight of the abdominal contents pressing against the diaphragm is greatest on the dependent part of the diaphragm (posterior lung and lower lung, re spectively) and least on the nondependent part of the diaphragm (anterior lung and upper lung, re spectively) (Fig. 4-5). In the awake, sponta neously breathing patient, the normally present ac tive tension in the diaphragm overcomes the weight of the abdominal contents, and the dia phragm moves the most (largest excursion) in the dependent portion and least in the nondependent portion. This is a healthy circumstance because this is another factor that maintains the greatest amount of ventilation where there is the most per fusion (dependent lung) and the least amount of ventilation where there is the least perfusion (nondependent lung). During paralysis and positivepressure breathing, the passive and flaccid dia phragm is displaced preferentially in the nonde pendent area, where the resistance to passive dia phragmatic movement by the abdominal contents is least; conversely, the diaphragm is displaced minimally in the dependent portion where the re-
Anesthetized, Lateral Decubitus Position Distribution of Ventilation Closed Chest Open Chest
Figure 4-5 This schematic of a patient in the lateral decubitus position compares the closed-chest anesthetized condition with the open-chest anesthetized and paralyzed condition. Opening the chest increases nondependent lung compliance and reinforces or maintains the larger part of the tidal ventilation going to the nondependent lung. Paralysis also reinforces or maintains the larger part of tidal ventilation going to the nondependent lung because the pressure of the abdominal contents (P AB ) pressing against the upper diaphragm is minimal (smaller arrow), and it is therefore easier for positive-pressure ventilation to displace this lesser resisting dome of the diaphragm. (V = alveolar volume; Ρ = transpulmonary pressure.) (Modified with permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Livingstone Inc., 1983, chapter 8.)
130
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
sistance to passive diaphragmatic movement by the abdominal contents is greatest.15 This is an unhealthy circumstance because the greatest amount of ventilation may occur where there is the least perfusion (nondependent lung), and the least amount of ventilation may occur where there is the most perfusion (dependent lung).15
E. Summary of Physiology of Lateral Decubitus Position and the Open Chest In summary (Fig. 4-6), the preceding section has developed the concept that the anesthetized,
paralyzed patient in the lateral decubitus position with an open chest may have a considerable ven tilation-perfusion mismatch, consisting of greater ventilation but less perfusion to the nondependent lung and less ventilation but more perfusion to the dependent lung. The blood flow distribution is mainly and simply determined by gravitational ef fects. The relatively good ventilation of the upper lung is caused, in part, by the open chest and paralysis. The relatively poor ventilation of the dependent lung is caused, in part, by the loss of dependent-lung volume with general anesthesia and by compression of the dependent lung by the
Lateral Decubitus Position Anesthetized Open Chest
Ν / Nondependent ^ ^ w ^ < Lung Ζ
Mediastinum
1 1i
Paralysis (Flaccid Diaphragm) /
AB
v Suboptimal y ^Positioning' Effects Figure 4-6 Schematic summary of ventilation-perfusion relationships in the anesthetized patient in the lateral decubitus position who has an open chest and is paralyzed and suboptimally positioned. The nondependent lung is well ventilated (as indicated by the large dashed lines) but poorly perfused (small perfusion vessel), and the dependent lung is poorly ventilated (small dashed lines) but well perfused (large perfusion vessel). In addition, the dependent lung may also develop an atelectatic shunt compartment (indicated on the left side of the lower lung) because of the circumferential compression of this lung. (P A B = pressure of the abdominal contents.) (Modified with permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Livingstone Inc., 1983, chapter 8.)
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
pendent lung in group 1 patients resulted in an adequate Pa02 with a lower inspired 02 concentration during surgery and a smaller P(A-a)02 at the end of surgery than when both lungs were ventilated with ZEEP. Thus, even if the selective PEEP to the dependent lung increased dependent-lung pulmonary vascular resistance and diverted some blood flow to the nondependent lung, the diverted blood flow could still participate in gas exchange with the ZEEP-ventilated nondependent lung.19 However, it should be noted that this technique requires that the nondependent (and operative) lung be ventilated, and this may impede the performance of surgery.
mediastinum, abdominal contents, and suboptimal positioning effects. In addition, poor mucociliary clearance and absorption atelectasis with an increased F,02 may cause further dependent-lung volume loss. Indeed, on rare occasion, the dependent lung may be massively atelectatic and edematous.16 Consequently, two-lung ventilation under these circumstances may result in an increased alveolar-arterial oxygen tension difference [P(A-a)02] and less than optimal oxygenation. A physiologic solution to the adverse effects of anesthesia and surgery in the lateral decubitus position on the distribution of ventilation and perfusion during two-lung ventilation would be the selective application of PEEP to the dependent lung (via a double-lumen endotracheal tube).12 Selective PEEP to the lower lung should increase the ventilation to this lung by moving it up to a steeper, more favorable portion of the lung pressure-volume curve. Indeed, this has been done with reasonably good success,12·17 although all the results have not been entirely uniform.18 A series of 22 mechanically ventilated patients (both lungs) undergoing thoracotomy in the lateral decubitus position was divided into two groups.17 Group 1 patients had 10 cm H20 of PEEP applied to the dependent lung while zero end-expiratory pressure (ZEEP) was applied to the nondependent lung. Group 2 (control) patients were intubated with a standard endotracheal tube, and both lungs were ventilated with ZEEP. Selective PEEP to the de-
Two Lung Ventilation
131
IV. PHYSIOLOGY OF ONE-LUNG VENTILATION A. Comparison of Arterial Oxygenation and C0 2 Elimination During Two-Lung Versus One-Lung Ventilation As discussed previously, the matching of ventilation and perfusion is impaired during two-lung ventilation in an anesthetized, paralyzed, openchest patient in the lateral decubitus position. The reason for the mismatching of ventilation and perfusion is relatively good ventilation but poor perfusion of the nondependent lung and poor ventilation and good perfusion of the dependent lung (Figs. 4-6 and 4-7A). The blood flow distribution
vs
One Lung Ventilation
Figure 4-7 Schematic representation of two-lung ventilation versus one-lung ventilation. Typical values for fractional blood flow to the nondependent and dependent lungs as well as P a 0 2 and Qs/Q, for the two conditions are shown. The Qs/Q, during two-lung ventilation is assumed to be distributed equally between the two lungs (5 per cent to each lung). The essential difference between two-lung and one-lung ventilation is that during one-lung ventilation the nonventilated lung has some blood flow and, therefore, an obligatory shunt, which is not present during two-lung ventilation. The 35 per cent of total flow perfusing the nondependent lung, which was not shunt flow, was assumed to be able to reduce its blood flow by 50 per cent by hypoxic pulmonary vasconstriction.'2 The increase in Qs/Q, from two-lung to one-lung ventilation is assumed to be solely due to the increase in blood flow through the nonventilated, nondependent lung during one-lung ventilation.
was seen to be mainly and simply determined by gravitational effects. The relatively good ventila tion of the nondependent lung was seen to be caused, in part, by the open chest and paralysis. The relatively poor ventilation of the dependent lung was seen to be caused, in part, by the loss of dependent-lung volume with general anesthesia and by circumferential compression of the depen dent lung by the mediastinum, abdominal contents, and suboptimal positioning effect. The compres sion of the dependent lung may cause the devel opment of a shunt compartment in the dependent lung (Figs. 4-6 and 4-7Λ). Consequently, twolung ventilation under these circumstances may result in an increased P(A-a)0 2 and impaired oxy genation. However, if the nondependent lung is nonventilated, as during one-lung ventilation, then any blood flow to the nonventilated lung becomes shunt flow, in addition to whatever shunt flow might exist in the dependent lung (Fig. 4-ΊΒ). (Table 4-1 and discussion on page 135 show a quantitative analysis of the two-lung ventilation to the one-lung ventilation conversion process with respect to arterial oxygenation.) Thus, one-lung ventilation creates an obligatory right-to-left transpulmonary shunt through the nonventilated, nondependent lung, which is not present during twolung ventilation. Consequently, it is not surprising to find that, given the same inspired oxygen con centration (F,0 2 ) and hemodynamic and metabolic status, one-lung ventilation results in a much larger alveolar-to-arterial oxygen tension differ ence P(A-a)0 2 and lower P a 0 2 than does two-lung ventilation. This contention is best supported by studies that compare arterial oxygenation during two-lung ventilation and one-lung ventilation, wherein each patient serves as their own control. 20 One-lung ventilation has much less of a steadystate effect on P a C0 2 in comparison with its effect on P a 0 2 . Blood passing through underventilated alveoli retains more than a normal amount of C 0 2 and does not take up a normal amount of 0 2 ; blood traversing overventilated alveoli gives off more than a normal amount of C 0 2 but cannot take up a proportionately increased amount of 02 owing to the flatness of the top end of the oxygen-hemoglo bin dissociation curve (see Fig. 3-32). Thus, dur ing one-lung ventilation (the one-lung minute ven tilation = the two-lung ventilation), the ventilated lung can eliminate almost enough C 0 2 to compen sate for the nonventilated lung, and P A C 0 2 and P E T C0 2 to P a C0 2 gradients are small; however, the ventilated lung cannot take up enough 02 to com pensate for the nonventilated lung, and P A 0 2 to P a 0 2 gradients are usually large. With a constant minute ventilation (two-lung ventilation compared with one-lung ventilation), the retention of C 0 2 by
blood traversing the nonventilated lung usually slightly exceeds the increased elimination of C 0 2 from blood traversing the ventilated lung, and the P a C0 2 will usually slowly increase (along with the end-tidal C 0 2 ; see later discussion) over time un less the respiratory rate is increased (see chapter 11 and following paragraph). The initiation of one-lung ventilation has much more of an acute effect (first 5 min) on P E T C0 2 than it does on P a C0 2 . When one-lung ventilation is begun (keeping total tidal volume and respira tory rate constant), the ventilated lung is immedi ately hyperventilated in relation to its perfusion (i.e., has an increased V/Q ratio) and P E T C0 2 from this lung decreases in the first minute (e.g., by 5 mm Hg). 21 Over the next 5 min, hypoxic pulmo nary vasoconstriction (HPV) in the nonventilated lung shifts blood flow over to the ventilated lung, increases ventilated lung perfusion, decreases ven tilated lung V/Q ratio and increases the P E T C0 2 back to the baseline two-lung ventilation value.21 Thereafter, and as discussed previously, P E T C0 2 will slowly increase (along with P a C0 2 ) because the same total minute ventilation to one lung is not as effective as when it is delivered to both lungs (i.e., there is an increased alveolar dead space within the one ventilated lung).
B. Blood Flow Distribution During One-Lung Ventilation 7. Blood Flow to the Nondependent, Nonventilated Lung Fortunately, there are both passive mechanical and active vasoconstrictor mechanisms that are usually operant during one-lung ventilation that minimize the blood flow to the nondependent, nonventilated lung and thereby prevent the P a 0 2 from decreasing as much as might be expected on the basis of the distribution of blood flow during two-lung ventilation. The passive mechanical mechanisms that decrease blood flow to the nondependent lung consist of gravity, surgical inter ference with blood flow, and perhaps the amount of pre-existing disease in the nondependent lung (Fig. 4-8). Gravity causes a vertical gradient in the distribution of pulmonary blood flow in the lateral decubitus position for the same reason that it does in the upright position (see Figs. 3-4 and 4-2). Consequently, blood flow to the nondependent lung is less than blood flow to the dependent lung. The gravity component of blood flow reduction to the nondependent lung should be constant with respect to both time and magnitude. Severe surgical compression (directly compress ing lung vessels) and retraction (causing kinking
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
133
One Lung Ventilation: Determinants Of Blood Flow Distribution Surgica Interference
Nondependent Lung Gravity
Dependent Lung Gravity
Hypoxic Pulmonary Vasoconstriction and/or Lung Disease
Lung Disease and/or Hypoxic Pulmonary Vasoconstriction
Figure 4-8 Schematic diagram of the determinants of blood flow distribution during one-lung ventilation. The major determinants of blood flow to the nondependent lung are gravity, surgical interference with blood flow, the amount of nondependent lung disease, and the magnitude of nondependent lung hypoxic pulmonary vasconstriction. The determinants of dependent lung blood flow are gravity, amount of dependent lung disease, and dependent lung hypoxic pulmonary vasoconstriction. (RV — right ventricle.)
and tortuosity of lung vessels) of the nondependent lung may further passively reduce nondependentlung blood flow. In addition, ligation of pulmonary vessels for pulmonary resection will greatly decrease nondependent-lung blood flow. The surgical interference component of blood flow reduction to the nondependent lung should be variable with respect to both time and magnitude. However, it should be noted that there is evidence indicating that some physical stimuli, such as stroking of pulmonary tissue, may cause local release of vasodilator prostaglandins,22 and other stimuli such as lung compression with a moist gauze may cause release of a quick-acting and terminating substance such as endothelium-derived relaxing factor.23 In view of these findings, it is not surprising that one study has shown that the shunt fraction increases significantly when the nonventilated lung is exposed to mild-to-moderate degrees of surgical manipulation.24 Thus, it appears that the effect of surgical manipulation on arterial oxygenation may depend on the exact nature and force/strength of the physical stimulus. The amount of disease in the nondependent lung should also be a significant determinant of the amount of blood flow to the nondependent lung. If the nondependent lung is severely diseased, then there may be a fixed reduction in blood flow to
this lung preoperatively, and collapse of such a diseased lung may not cause much of an increase in shunt (see Table 4-1). The notion that a diseased pulmonary vasculature might be incapable of HPV is supported by the observation that administration of sodium nitroprusside and nitroglycerin (which should abolish any pre-existing HPV) to chronic obstructive pulmonary disease patients (who have a fixed reduction in the cross-sectional area of their pulmonary vascular bed) does not cause an increase in shunt,25 whereas these drugs do increase shunt in patients with acute regional lung disease who have an otherwise normal pulmonary vascular bed.26 If the nondependent lung is normal and has a normal amount of blood flow, then collapse of such a normal lung may be associated with a higher nonventilated-, nondependentlung blood flow and shunt. A higher one-lung ventilation shunt through the nondependent lung is therefore theoretically more likely to occur in patients who require thoracotomy for nonpulmonary disease.27 There have been only two studies that have systematically validated the inverse correlation between the amount of nondependent-lung disease and shunt during one-lung ventilation.2* 2i) Figure 4-9 shows the correlation between the percentage of the cardiac output perfusing the operative lung, as measured by preoperative perfusion
134
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
Χ Flow to Operative Lung
Figure 4-10 Effect of unilateral hypoxia in humans. The pulmonary vascular resistance of the hypoxic lung is plotted against its P A 0 2 . (From Harris P, Heath D: The Hu man Pulmonary Circulation. 2nd ed. Edinburgh, Churchill Livingstone, 1977, p. 456. Used with permission.)
Figure 4—9 The relative preopera tive perfusion to the operative lung correlated with the P a 0 2 after 10 min of one-lung anesthesia. Relative per fusion to the operative lungs was measured by scintigraphy performed after the intravenous injection of technetium 99m macroaggregated human albumin. (From Hurford WE, Kolker AC, Strauss HW: The use of ventilation/perfusion lung scans to predict oxygenation during one-lung anesthesia. Anesthesiology 67:841844, 1987. Used with permission.)
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
scans, and the intraoperative one-lung ventilation P,02.28 The most significant reduction in blood flow to the nondependent lung is caused by an active vasoconstrictor mechanism. The normal response of the pulmonary vasculature to atelectasis is an increase in pulmonary vascular resistance (in just the atelectatic lung), and the increase in atelectatic lung pulmonary vascular resistance is thought to be due almost entirely to HPV. 303I The effect of unilateral hypoxia on the hypoxic lung pulmonary vascular resistance in humans is shown in Figure 4-10.32 The selective increase in atelectatic lung pulmonary vascular resistance diverts blood flow away from the atelectatic lung toward the remaining normoxic or hyperoxic ventilated lung. The diversion of blood flow minimizes the amount of shunt flow that occurs through hypoxic lung. Figure 4-11 shows the theoretically expected effect of HPV on arterial oxygen tension (P a 0 2 ) as the amount of lung that becomes hypoxic increases.33 When very little of the lung is hypoxic (near 0 per cent), it does not matter, in terms of P a 0 2 , whether the small amount of lung has HPV or not because in either case the shunt will be small. When most of the lung is hypoxic (near 100 per cent), there is no significant normoxic region to which the hypoxic region can divert flow, and, again, it does not matter, in terms of P a 0 2 , whether the hypoxic region has HPV or not. When the percentage of lung that is hypoxic is intermediate (between 30 and 70 per cent), which is the amount of lung that is hypoxic during the one-lung ventilation/anesthesia condition, there is a large difference between the P a 0 2 expected with a normal amount of HPV (which is a 50 per cent blood flow reduction for a single lung)33 compared with when there is no HPV. In fact, in this range of hypoxic lung, HPV can increase P a 0 2 from hypoxemic levels to much higher and safer values. It is not surprising, then, that numerous clinical studies on one-lung ventilation20·27· 34^2 found that the shunt through the nonventilated lung is usually 20 to 30 per cent of the cardiac output as opposed to the 40 to 50 per cent shunt that might be expected if there was no HPV in the nonventilated lung.41 Thus, HPV is an autoregulatory mechanism that protects the P a 0 2 by decreasing hypoxic lung. It is possible to model the two-lung ventilation conversion process for various initial two-lung ventilation shunts; the model shown in Table 4-1 makes several assumptions. First, the initial twolung ventilation shunt flow is equally distributed between the nondependent and dependent lungs. Second, the remaining normal blood flow to the nondependent lung can decrease its blood flow by 50 per cent owing to HPV,33 and any initial shunt flow (i.e., during two-lung ventilation) does not
135
Figure 4-11 The graph is a model of the effect of hypoxic pulmonary vasoconstriction (HPV) on P a 0 2 as a function of the per cent of lung that is hypoxic. The model assumes an F,0 2 of l.O, a normal hemoglobin, cardiac output, and oxygen consumption. In the range of 30 to 70 per cent of the lung being hypoxic, the normal expected amount of HPV can increase P a 0 2 significantly.
participate in an HPV response. Third, the total one-lung ventilation shunt flow is a sum of half the normal flow to the nondependent lung when it was ventilated plus the original nondependent and dependent lung shunt flows. Table 4-1 shows that as the two-lung ventilation shunt increases, the amount of nondependent lung flow that is able to participate in an HPV response decreases; therefore, the amount of blood flow diversion due to nondependent lung HPV decreases, and, thus, the total one-lung ventilation shunt increases. Figure 4-12 outlines the major determinants of the amount of atelectatic lung HPV that might occur during anesthesia. In the following discussion, the HPV issues or considerations are numbered as they appear in Figure 4-12.
1. Distribution of Hypoxia The distribution of the alveolar hypoxia is probably not a determinant of the amount of HPV; all regions of the lung (either the basilar or dependent parts of the lungs [supine or upright] or discrete anatomic units such as a lobe or single lung) respond to alveolar hypoxia with vasoconstriction.43
136
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
Table 4-1
MODEL OF CONVERTING TWO-LUNG TO ONE-LUNG VENTILATION Two-Lung Ventilation Fractional Normal Flow
Lung ND
::
Total Shunt Flow
0.400
One-Lung Ventilation Fractional Shunt Flow
Fractional Normal Flow
Total Shunt Flow*
Fractional Shunt Flow 0.2001 (0.200 + 0)
0.200
D
0.600
0.800
ND
0.350
0.050
0.225t (0.175 + 0.050)
D
0.550
0.050
0.725
0.050
ND
0.300
0.100
0.250t (0.150 + 0.100)
D
0.500
0.100
0.650
0.100
ND
0.200
0.200
0.300t (0.100 + 0.200)
D
0.400
0.200
0.500
0.200
ND
0.100
0.300
0.350t (0.050 + 0.300)
D
0.300
0.300
0.350
0.100
0.275
0.200
0.350
0.400
0.500
0.600
0.650 0.300
*Sum of ND and D lung fractional shunt flows. tHalf of two-lung ventilation fractional normal flow (due to hypoxic pulmonary vasoconstriction) plus all of two-lung ventilation fractional shunt flow.
Abbreviations: ND = nondependent lung; D = dependent lung.
Anesthetic Experience and Regional HPV NORMOXIC COMPARTMENT
HYPOXIC COMPARTMENT
Figure 4-12 This figure lists many of the components of the anesthetic experience that might determine the amount of regional hypoxic pulmonary vasoconstriction (HPV). The clockwise numbering of considerations corresponds to the order in which these considerations are discussed in the text. (PVP = pulmonary vascular pressure; PEEP = positive end-expiratory pressure; Q, = pulmonary blood flow; V/Q = ventilation-perfusion ratio.)
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
However, recent evidence suggests that, on a sublobar level, collateral ventilation may be the first line and HPV the second line of defense against the development of arterial hypoxemia (Fig. 413).44 In species with extensive collateral ventilation, such as canines, the development of sublobar atelectasis or low V/Q areas does not cause as much sublobar HPV because collateral ventilation prevents the sublobar area in question from becoming very hypoxic. This protective phenomenon seems reasonable when one considers that air (0 2 ) is less dense than blood and therefore easier to redistribute. Collateral ventilation is most efficient when very small airways (1.6 mm) are occluded, because many channels are available for by-passing the obstruction, whereas collateral ventilation is least efficient when large airways (4.8 mm) are occluded, because fewer by-pass channels are available (see Figs. 2-19 and 2-20). 45 On the other hand, in consolidated lesions and in species with no collateral ventilation,46 such as the coatimundi, ferret, cattle, and swine, the development of sublobar atelectasis and low V/Q areas does elicit a great deal of sublobar HPV, which minimizes the decrease in P a 0 2 .
2. Low V/Q Versus Atelectasis As with low V/Q and nitrogen-ventilated lungs, it appears that the vast majority of blood flow reduction in the acutely atelectatic lung is due to HPV, and none of the blood flow reduction is due to passive mechanical factors (such as vessel
137
tortuosity).30· 31 This conclusion is based on the observation that re-expansion and ventilation of a collapsed lung with nitrogen (removing any mechanical factor) does not increase the blood flow to the lung, whereas ventilation with oxygen restores all of the blood flow back to precollapse values. This conclusion applies whether ventilation is spontaneous or with positive pressure and whether the chest is open or closed.47 In canines, a slight amount of further subacute (greater than 30 min) decrease in blood flow to the atelectatic lung may have been due to some mechanical effect of the atelectasis on lung blood vessels.48 However, in humans, a prolonged unilateral hypoxic challenge during anesthesia results in an immediate vasoconstrictor response with no further potentiation or diminution of the response.49
3. Vasodilator Drugs A major concern with administering a drug that can cause pulmonary vasodilatation is that the drug will inhibit pre-existing HPV and thereby increase Qs/Qt and decrease P a 0 2 . However, the in vivo physiology that follows vasodilator therapy (usually both systemic and pulmonary circulations vasodilate) is complex, and there are three major reasons why, in a given experiment, administration of a vasodilator drug may not decrease HPV and. if it does decrease HPV, does not result in a decrease in P a 0 2 . First, the pre-existing hypoxia may involve all of the lung (e.g., as a result of decreased F,0 2 ); by definition, P a 0 2 cannot change
Sublobar Ventilation-Perfusion Regulation Collateral Ventilation
HPV (No Collateral Ventilation)
First Line of Defense
Second Line of Defense
Hypoxic
Normoxic
Hypoxic
Normoxic
(Air is less dense than blood and therefore easier to redistribute) Figure 4-13 Because gas is much less dense than blood and, therefore, easier to redistribute, collateral ventilation may be the first line of defense and HPV the second line of defense against the development of arterial hypoxemia caused by sublobar atelectasis (ATEL) or low V/Q (ventilation-perfusion ratio) areas. (HPV = hypoxic pulmonary vasoconstriction.)
138
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
significantly, yet hypoxic pulmonary vasoconstriction may be greatly decreased. Second, the amount of pre-existing HPV may be slight (e.g., healthy awake or anesthetized patients, presence of high concentrations of halogenated anesthetics or other inhibiting factors); by definition, inhibition of a slight amount of HPV has no effect on arterial oxygenation. Third, most pulmonary vasodilators are also systemic vasodilators and therefore cause an increase in cardiac output. The increase in cardiac output may cause an increase in P 9 0 2 (see section 6) and the increase in P ç 0 2 may offset any decrease in HPV and result in no change or even increase in P a 0 2 (see Table 3-5). However, and adding to the complexity of the situation, changes in cardiac output in either direction may increase transpulmonary shunt (see section 5). Virtually every drug known to cause pulmonary vasodilatation has been studied with regard to ef-
fects on either HPV and/or arterial oxygenation. Because most vasodilator drugs have been so extensively studied, Tables 4-2 to 4-6 list only those studies performed after 1986 (since the first edition); these latest studies (and their references) and the references in this text will allow the reader to construct most of if not the entire literature tree for each drug. Almost all of the studies with the very potent systemic and pulmonary vasodilator drugs (Table 4-2) have shown inhibition of HPV or have a clinical effect (increased shunt, decreased P a 0 2 ) that is consistent with decreased HPV. The potent vasodilating drugs that have been shown to decrease HPV or have a clinical effect consistent with decreased HPV are nitroprusside,26·50-60 nitroglycerin,26·6,_72 many β 2 ^οηΪ8ΐ8 (isoproterenol,73-77 terbutaline,78), ritodrine,79 orciprenaline,80 salbutamol,81 adenosine,82 and nitric oxide. 83-87 Interest ingly, nitroglycerin is metabolized intracellularly
Table 4-2 EFFECT OF THE VERY POTENT VASODILATORS ON HPV* AND/OR ARTERIAL OXYGENATION
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
139
Table 4-3 EFFECT OF VASODILATOR PROSTAGLANDINS ON HPV* AND/OR ARTERIAL OXYGENATION
into nitric oxide.88 The vasodilator prostaglandins (E and I2) (Table 4-3) 57 · 68 · 85 · 89 " 91 and the calcium channel blockers (Table 4-4) 92 -' 03 inhibit HPV to a much lesser extent and often have no effect on arterial oxygenation.57·90· ·'· l02· ,03 Although most animal studies with dobutamine show significant inhibition of HPV,73· 74· 77· ,04 - , ° 7 human one-lung ventilation studies72· lo severity.18 |y For one example, the incidence of lung cancer increases with age, but the lung cancer in the third, fourth, and fifth decades of life is much more aggressive (brief duration of symptoms, advanced stage, decreased time of survival) than lung cancer in older patients.20-22 Conversely, heavy smokers do not necessarily develop lung cancer. For another example, dietary factors (beta carotene, vitamins A and E, selenium) may play a role in the causation/prevention of lung cancer.23-25 Lung carcinoma has a high incidence of occurrence in workers in some chemical industries (asbestos.26· 27 arsenic, chromâtes, coal gas, and nickel) than in the general population. Uranium miners have a greatly increased risk for the development of lung carcinoma, especially if they smoke. Radon gas (sixth daughter of uranium 238) is ubiquitous and is now recognized as perhaps the second most important cause of lung cancer.28 Five per cent of the patients with lung carcinoma are completely asymptomatic at the time of discovery,5 and in this group the tumor is discovered only on routine roentgenographic examina-
Table 5-1
STANDARDIZED MORTALITY RATIOS FOR LUNG CANCER IN WOMEN IN ACS-CPS II BY NUMBER OF CIGARETTES CURRENTLY SMOKED AND DURATION OF SMOKING't
Duration of Smoking (Years)
1-10
11-19
20
21-30
31 +
21-30 31-40 41-70
2.9 7.9 10.0
6.7 19.2 17.0
13.6 19.2 25.1
18.4 26.5 34.3
18.9 25.3 38.3
Cigarettes Per Day
*Adapted with permission from Garfinkel L, Stellman SD: Smoking and lung cancer in women: Findings in a prospective study. Cancer Res 48:6951-6955, 1988. tin comparison with nonsmoking women. Abbreviation: ACS-CPS II = American Cancer Society Cancer Prevention Study II.
154
Preoperative Cardiopulmonary Evaluation
Figure 5-1 The preoperative evaluation of masses of the lung and bronchi involves three basic steps. Step 1 consists of determining whether lung carcinoma is present and, if so, the cell type. Step 2 consists of determining whether the carcinoma has spread beyond its local confines. Step 3 involves physiologic assessment of the patient for the planned surgical procedure. This preoperative evaluation diagram displays the logic necessary for a patient to arrive at thoracotomy .
Death due to lung cancer in relation to cigarette smoking
Figure 5-2 Number of lung cancer deaths/100,000 smokers and nonsmokers. (From Spiro S: Lung cancer: Presentation and treatment. Medicine International 3798-3805, 1991. Used with permission.)
Preoperative Cardiopulmonary Evaluation
tion of the chest. The vast majority of patients, however, have one or more symptoms related to the presence of the tumor. The symptoms may be designated as broncho-pulmonary, extrapulmonary intrathoracic, extrathoracic metastatic, extrathoracic nonmetastatic, and nonspecific (Table 5-2).'~5 On the average, symptoms have been present for 6 to 7 months prior to the time the patient seeks medical advice; since the first chest x-ray findings frequently antedate the first symptoms by several months, lung carcinoma will be at least a year old (and perhaps 2 to 5 years old) by the time of clinical presentation.
a.
Bronchopulmonary symptoms arising from involvement of the lung are due to bronchial irritation, ulceration, obstruction, infection distal to the obstruction, or a combination of these processes.
Table 5-2
COUGH
In a large series of patients with carcinoma of the lung, 75 per cent had cough as one of the major symptoms, and this symptom was severe in 40 per cent of the patients. However, cough is possibly the most common manifestation of respiratory disease in general. It is so common among cigarette smokers that many of them regard a morning cough as "normal." The most common stimulus to cough is the formation of sputum in the respiratory tract (see the following), and the cough process is an essential element in keeping the tract clear. b.
1. Bronchopulmonary Symptoms
155
SPUTUM
The normal adult produces about 100 ml of mucus from the respiratory tract in a day. When excess mucus is formed, it may accumulate, stimulate the mucous membrane, and be coughed up as sputum. Sputum in patients with bronchogenic
FREQUENCY OF SYMPTOM OCCURRENCE AT PRESENTATION IN BRONCHOGENIC CARCINOMA*
Symptom
*Based on data from Shields,' Spiro,2 Le Roux,3 Jones,4 and Ferguson.5
Frequency of Occurrence (Per Cent)
156
Preoperative Cardiopulmonary Evaluation
carcinoma may be formed in response to physical, chemical, or infective insult to the mucous mem brane of the airways. Mucoid sputum is clear or white. Black sputum is due to the detritus of cigarette or atmospheric smoke. Purulent sputum contains pus mixed with mucus. Purulent sputum is usually yellow, but if it has been stagnant it may be green, owing to the action of verdoperoxidase, derived from neutro phils. Failure to clear a recent change in the quality and quantity of sputum within a few days of initi ating antibiotic therapy should raise suspicion of a neoplasm. Blood-stained sputum can vary from small streaks to gross hemoptysis (see the follow ing) and always warrants investigation for carci noma. Hemoptysis, generally episodic blood streaking of the sputum, is present in 57 per cent of patients with bronchogenic carcinoma and is the first symptom in many. C. HEMOPTYSIS
Hemoptysis affects 50 to 70 per cent of patients with lung cancer at some point in their clinical course. It is usually not severe, rarely is life-threat ening, and may be occasional and/or a one-time event. d. CHEST PAIN
Chest pain is present in 40 per cent of patients presenting with a new carcinoma. It is usually a mild, constant dull ache on the side of the tumor and is often due to erosion of a rib (disease within the lung is usually painless). Another important form of chest pain with lung carcinoma is pleuritic pain. It is due to direct tumor extension to the parietal pleura and is characteristically sharp, is worse on breathing and coughing, and can usually be accurately localized by the patient. Mediastinal tumors can cause pain that is usually aching and retrosternal but poorly localized. Θ. DYSPNEA
Dyspnea is a common complaint in patients with chronic lung disease and lung carcinoma (39 per cent). In chronic diseases, it is common to find that patients begin to complain of dyspnea only after the respiratory reserve is quite severely im paired, whereas in patients with lung carcinoma dyspnea occurs more abruptly and with less objec tive functional impairment. The degree of dyspnea should be approximately quantified (i.e., how far can the patient walk, how many steps can be climbed, and so on). The time period over which dyspnea has developed is important in the diagno sis; dyspnea due to lung carcinoma develops over weeks to months.
f. WHEEZE
This is described by 10 per cent of patients with lung cancer and is frequently localized to one side. It is due to airway obstruction by the tumor and, if located in the trachea, severe dyspnea and stridor (i.e., inspiratory wheeze) may develop.
2. Extrapulmonary Intrathoracic Symptoms Other symptoms of chest disease occur as a result of growth of the tumor beyond the confines of the lung. These symptoms are due to involve ment of the pleura, chest wall, diaphragm, medias tinal structures, and contiguous nerves. Approxi mately 15 per cent of patients with carcinoma of the lung have these kinds of extrapulmonary intra thoracic symptoms. Pleural effusion is due to either metastatic in volvement of the pleura (blood stained) or obstruc tion of peripheral lymphatic drainage (clear color) or central (thoracic duct) lymphatic drainage (chy lous effusion). Chest wall pain is due to direct involvement of the chest wall by tumor. Dys phagia is due to partial obstruction of the esopha gus by the tumor in the paraesophageal lymph nodes. The superior vena cava syndrome (dyspnea, dysphagia, stridor, blackouts, and severe headache on coughing) is due to obstruction of the superior vena cava by right paratracheal lymphadenopathy. Pain down the arm is due to involvement of the branches of the brachial plexus from tumors lo cated in the superior sulcus. Horner's syndrome (small pupil, partial ptosis, enophthalmos, lack of thermal sweating on the ipsilateral half of the face) is due to involvement of the cervical sympathetic chain and will often be present in patients who have pain down the arm. Hoarseness is due to paralysis of the vocal cord as the result of involve ment of the left recurrent laryngeal nerve (at the left hilum) or rarely, of the right recurrent laryn geal nerve. Paralysis of the recurrent laryngeal nerve(s) may also cause difficulty with expectora tion and chronic aspiration (caused by failure of the vocal cords to adduct). Brachial plexus neuritis can occur with an apical tumor (usually T, distri bution), and pericarditis is caused by direct in volvement of the pericardium.
3. Extrathoracic Metastatic Symptoms Symptoms resulting from metastatic spread of the tumor outside the thorax account for a small percentage of the presenting or major complaints of patients with carcinoma of the lung (although as many as 50 per cent of patients with lung car cinoma actually have asymptomatic extrathoracic
Preoperative Cardiopulmonary1 Evaluation
metastases).5 These extrathoracic metastatic symptoms can be referable, in order of general decreasing frequency, to brain, skeleton, liver, adrenals, gastrointestinal tract, kidneys, and pancreas. In these cases, the history is extremely important because any positive history referable to these organs requires specific organ workup for metastatic disease (see staging later) and, if found, precludes surgery. Extrapulmonary small-cell carcinoma (without lung involvement) is a distinct clinicopathologic entity, is associated with smoking, and can occur in virtually every organ.17
4. Extrathoracic Nonmetastatic Symptoms The extrathoracic nonmetastatic symptoms (may occur in up to 10 per cent of patients with lung carcinoma)5 are usually due to paraneoplastic syndrome caused by secretion of endocrine or endocrine-like substances by the tumor (see chapter 3, Table 3-7). The endocrine-like manifestations include Cushing's syndrome, excessive antidiuretic hormone secretions (low serum sodium and plasma osmolarity, high urine osmolarity), carcinoid syndrome, hypercalcemia, ectopic gonadotropin secretion, and hypoglycemia. The neuromuscular manifestations consist of carcinomatous myopathies (Eaton-Lambert syndrome) and various myopathies related to brain dysfunction. Other manifestations can be skeletal (clubbing, pulmonary hypertrophic osteoarthropathy), dermatologie (scleroderma, acanthosis nigricans), vascular (thrombophlebitis), and hematologic.
157
lymph nodes because lung carcinoma is most often in an advanced, inoperable stage at the time of presentation. Clubbing of the fingers should always raise the possibility of intrathoracic neoplasm. Because it is mandatory subsequently to use much more sensitive radiologic means of determining resectability, further discussion of physical examination methods for making these determinations is not continued here. Similarly, the questions of overall severity of chronic lung disease and whether the patient can tolerate the planned procedure (operability) can be much more quantitatively answered by pulmonary function testing than by the findings of the physical examination (see Physiologic Assessment of the Patient for Surgery).
C. Common Laboratory Tests Some of the routine laboratory tests that are performed on all patients are especially relevant to the preoperative evaluation of the patient with a lung or bronchial mass. These tests can be divided into those that help establish the diagnosis of lung cancer (e.g., chest x-ray and sputum for cytology), those that help establish the diagnosis of metastatic lung cancer (e.g., liver and bone enzymes, blood urea nitrogen and creatinine, and the urinalysis), and those that help physiologic assessment (e.g., hemoglobin concentration). Each of these three diagnostic areas is fully discussed in the following pages (along with the role of the common laboratory test in the workup logic).
5. Nonspecific Symptoms Weight loss, anemia, protein-energy malnutrition,29 weakness, anorexia, lethargy, and malaise occur in a large number of patients. Vague febrile respiratory syndromes (cold-like) may be present in 22 per cent of these patients. In 10 to 15 per cent, these symptoms instigate the initial visit to the physician.
B. Physical Examination The basic tools of observation, inspection, palpation, and percussion should allow the physician to assess, in a gross way, the overall severity of chronic lung disease, whether major consolidation, atelectasis, or pleural effusion is present, and whether there is any obvious extrathoracic complication of thoracic carcinoma (Table 5-3). However, it should be noted that the most common manifestation of lung cancer on physical examination is the presence of palpable supraclavicular
D. Diagnosis of Lung Cancer 1. The Carcinomas of the Lung Lung cancer1·2-30·31 is the most common malignant disease and cause of cancer death in both sexes. On the basis of ordinary light microscopy histopathologic findings, lung cancer was formerly divided into four major types: squamous (epidermal), adeno, large cell, and small cell. However, there is growing acceptance of the hypothesis that all lung cancer starts by activation of an oncogene (e.g., by smoke, radiation) (an activated oncogene is called a proto-oncogene) or by loss of an antioncogene in a cell capable of differentiating into the various pathologic forms (Fig. 5-3).32~36 The cancer cell generation scheme in Figure 5-3 suggests that there is a common stem cell for all types of lung cancer, but that there is a basic genetic difference between small-cell lung carcinoma (SCLC) and non-small-cell lung carcinoma (nonSCLC) (i.e., the adeno, large, squamous type).
Table 5-3
PHYSICAL FINDINGS THAT OCCUR WITH PULMONARY PATHOLOGY Inspection and Palpation
Condition Normal Consolidation Major atelectasis Pleural effusion or empyema Cavitation Diffuse pulmonary fibrosis; interstitial lung disease
Equal rib and diaphragm movement Slight restriction of motion on side affected Slightly small and restricted on side affected Reduced movement on side affected Normal Symmetrically diminished
Percussion
Fremitus
Adventitious Sounds
Resonant
Present
Vesicular
None
Dull
Increased
Bronchial
Rales
Dull
Normal
Diminished
Dull or flat
Absent
Diminished or absent
Rales after deep breath or cough Friction rub early
Usually normal Normal
Usually present Normal
Amphoric Harsh vesicular with prolonged expiration Diminished with prolonged expiratory phase Vesicular with prolonged expiration Diminished with prolonged expiratory phase Bronchial if interstitial; vesicular if alveolar Diminished or absent
Emphysema
Enlarged and restricted bilaterally
Hyper-resonant
Normal or reduced
Bronchitis
Normal
Normal
Normal
Bronchial asthma
Normal or enlarged
Hyper-resonant
Reduced
Pulmonary edema Pneumothorax
Normal
Normal
Normal
Slightly enlarged and restricted movement Small and very restricted movement
Hyper-resonant
Absent
Dull
Present
Fibrothorax
Breath Sounds
Reduced to absent
Coarse rales Coarse rales uninfluenced by coughing Occasional rhonchi fine rales late in inspiration Rhonchi with coarse rales Wheezes
Other "Bronchophony" in normal spoken voice "Egophony" (E to A) and whispered pectoriloquy Tracheal and mediastinal shift toward Mediastinal shift away Coin sign
Hoover's sign; high clavicle; muscular wasting (pink puffer) Cyanotic (blue bloater) Distress
Moist rales
Distress
None
Mediastinal shift away
None
Mediastinal shift toward
Figure 5-3
Hypothesis for the development of histologic types of lung cancer.12^
160
Preoperative Cardiopulmonary Evaluation
This is supported by the occurrence of mixed pathologic types within the non-small-cell group such as adenosquamous carcinoma (by light microscopy). Indeed, lung carcinoma can be seen to differentiate spontaneously into various histologies in tissue culture, and the presence of adenosquamous histologies may simply represent a phase in this ongoing differentiation from a common stem cell.37 This hypothesis is even more dramatically supported by the finding that, when large-cell carcinoma is examined by electron microscopy and immunoperoxidase studies, it is obvious that this tumor may be further divided into five groups as follows: squamous, adenomatous, adenosquamous, neuroendocrine, and undifferentiated.38 Finally, lung cancers of both small-cell and non-small-cell variety share a number of common antigens and neuroendocrine markers32· 39; this overlap in the expression of biomarkers between small-cell and non-small-cell lung cancer, and within the various subtypes of non-small-cell lung cancer, also suggests a common stem cell for all types of lung cancer. The common ancestor of the epithelial lung tumor is thought to be the simple cuboidal endodermal cell of the primitive lung bud.39 The most common carcinomas of the lung (epidermoid, small cell, adeno, large cell), their relative incidence of occurrence, and most usual growth rate characteristics are listed in Table 5-4. However, it should be realized that all of these carcinomas are capable of a wide range of growth characteristics, and all cell type combinations are possible (combined adenosquamous carcinoma is the most common mixed tumor; this mixed carcinoma is responsible for 2 per cent of all lung
Table 5-4
carcinoma37· 40 ). Other very rare, malignant pulmonary neoplasms not listed in Table 5-4 are alveolar cell carcinoma, carcinoid, bronchial gland tumors, papillary tumors, sarcomas, and melanomas. Clinically, the most relevant differential diagnosis is that between SCLC on the one hand and the other lung carcinomas (non-SCLC) on the other. This is so because SCLC differs from nonSCLC by showing a higher growth rate and a propensity to earlier and more extensive metastastic spread (see Table 5-4). Therefore, SCLC has a much worse prognosis. In non-SCLC, surgery, if possible, is the treatment of choice. The early occurrence of metastatic lesions in most SCLC cases excludes surgical intervention as an effective treatment for this type of lung cancer. However, SCLC shows a remarkable initial sensitivity to chemoand radiotherapy.42 In SCLC, a clinically complete response is often observed after induction chemotherapy. Only few patients turn out to be cured, however, and at relapse a tumor may emerge whose cells are biologically different from those present in the pretreatment tumor (i.e., they are refractory to further treatment). In order to understand fully the diagnosis and the staging of lung carcinoma, it is necessary to have some appreciation of the natural history of each of the lung carcinoma cell types. Carcinoma of the lung may spread by direct extension, by lymphatic metastasis to lymph nodes, and by hematogenous metastasis to lymph nodes, and by hematogenous metastasis to distant organs (Fig. 54). Direct extension can be in any direction and can involve any structure within the chest (pleura, chest wall, diaphragm, all mediastinal structures).
CARCINOMAS OF THE LUNG
*Taken in part from Bains41 and Humphrey et al.4" tThe incidence of squamous carcinomas is on the decrease, and the incidence of adenocarcinoma is on the increase. ^Central = Inner two thirds of lung or proximal to third to fourth bronchial generation and usually visualized with a fiberoptic bronchoscope. ^Peripheral = Outer one third of lung or distal to third to fourth bronchial generation and not usually visualized with a fiberoptic bronchoscope.
Preoperative Cardiopulmonary Evaluation
161
Spread of Carcinoma of the Lung
2. Lymphatic Metastasis
1. Direct Extension
3. Hematogenous Metastasis Figure 5-4 Carcinoma of the lung spreads in three ways: first, by direct extension to the mediastinum, pleura and chest wall, diaphragm, and bronchi; second, by lymphatic metastasis to hilar, paratracheal and supraclavicular, and cervical nodes (proceeding from the distal to the proximal nodes, the incidence of involvement of the nodes decreases); third, by hematogenous metastasis to brain, liver, bone, adrenals, and kidney.
Blockage of a bronchus can cause distal atelectasis and infection. Cavitation may be due to either necrosis within the tumor mass or abscess formation distal to an obstructed bronchus. Lymphatic metastasis follows the lymph sump pathway of hilar and mediastinal nodal stations to the venous outlets (see chapter 2). All mediastinal structures can be potentially affected by lymph node enlargement and erosion. Hematogenous spread is due to invasion of the pulmonary veins by tumor cells. The blood-transported tumor cells are most frequently deposited and grow in brain, bone, liver, adrenal glands, and kidney. 2. Diagnosis of the Presence of Lung Cancer (Is Lung Carcinoma Present? Cell Type?)
The diagnosis of the presence of lung cancer'"* · is made most often by use of the chest roentgenogram, bronchoscopy, sputum cytology, and percutaneous needle biopsy (Fig. 5-5). Although clearly positive results from any of these four tests establishes the diagnosis of lung carcinoma, the diagnosis is most often established by a combination of chest roentgenogram consistent with the diagnosis along with a positive result from one of the other three tests. The latter three tests permit diagnosis of cell type in 75 per cent of cases. i0 31
a. CHEST ROENTGENOGRAM
It is useful to think of the radiograph as consisting broadly of four optical densities: black (air). dark gray (fat), light gray (soft tissue/fluid), and white or colorless (bone/calcification). To be seen as a separate structure on a chest radiograph, an object needs to have a radiograph density that contrasts with its environment and borders that are tangential to the X-ray beam. Lesions of soft tissue density in the air-density lung often fulfill these conditions, making the high-voltage (penetrating) 130- to 140-kV chest radiograph the most important preliminary investigative modality once lung cancer is suspected. It has been estimated that when a tumor of the lung is first detected on a chest roentgenogram, it has completed three fourths of its natural history,4^ and this first roentgenographic abnormality frequently antedates the first symptoms or signs of the disease by 7 or more months.44 By the time bronchial carcinoma becomes symptomatic, the chest roentgenogram is abnormal in 98 per cent, and the abnormality is most suggestive of tumor in more than four fifths of all these patients. The roentgenographic findings present due to carcinoma of the lung (Fig. 5-6) may be the result of the presence of the tumor itself within the lung (70 per cent are centrally located), of changes in the pulmonary parenchyma distal to a bronchus
162
Preoperative Cardiopulmonary Evaluation
Preoperative Evaluation of Masses of the Lung and Bronchi ŞţepJ: Is Lung Carcinoma Present? Cell Type? History, Physical Examination, Chest Roentgenogram
Diagnosis of Lung Carcinoma and Cell Type Confirmed Figure 5-5 Preoperative evaluation logic of step 1 for determining the presence of lung carcinoma type. See text for full explanation.
obstructed by the tumor (atelectasis, infection, and cavitation), and of spread of the tumor to extrapul monary intrathoracic sites (hilar and mediastinal lymph nodes, pleura, chest wall, and diaphragm). Because the primary route of lymphatic spread of lung carcinoma is to the hilum and mediastinum, familiarity with the mediastinal silhouette is an important diagnostic concern (Fig. 5-7). The early roentgenographic features are, unfor tunately, subtle in nature and often are appreciated
only in retrospect.44 The earliest signs visible in the roentgenogram of the chest are locally pro duced by the tumor itself. These signs may include any abnormal density within the lung parenchyma (most common), lobulation and cavitation of a mass (most specific for lung carcinoma),45 seg mental atelectasis, a simple cavitation within the lung, and a mediastinal mass (uncommon). A newly appreciated pulmonary density must be compared with old films to establish how long it
Chest Roentgenographic Findings Due to Lung Carcinoma
Parenchymal Mass
Hilar and Mediastinal Masses
Distal Atelectasis and Infection
rect Extension \f_^y ( Pathology
Figure 5-6 In patients with lung carcinoma the chest roentgenographic findings result from the presence of the tumor itself within the lung (parenchymal mass), changes in the pulmonary parenchyma distal to a bronchus obstructed by the tumor (atelectasis and infection), and spread of the tumor to extrapulmonary intrathoracic sites (hilar and mediastinal masses and other direct extension pathology).
Preoperative Cardiopulmonary Evaluation
163
Mediastinal Silhouette and Major Mediastinal Lines and Interfaces
Figure 5-7 The mediastinal silhouette is determined by the borders of the great vessels (aorta, great veins, pulmonary artery) and the heart.
has been present. Malignant pulmonary lesions usually have doubling times of less than 1 year. A lesion that remains the same size for at least 2 years can be presumed to be benign (see Table 55).46 The usual roentgenographic manifestations of lung carcinoma more frequently include the hilar Table 5-5
RADIOLOGIC CRITERIA FOR DIFFERENTIATING MALIGNANT FROM BENIGN PULMONARY OPACITIES*
I. More Likely Malignantf 1. 2. 3. 4.
Opacity larger than 3 cm in diameter Spicular margins Noncalcified Increasing size (doubling time 30-490 days)
II. More Likely Benignf 1. Stable size for 2 yearst or doubling time less than 30 days (probably infectious) or more than 490 days (probably benign) 2. Benign pattern of calcification 3. Well-circumscribed margins 4. Small (< 2 cm) size 5. Nearby satellite lesions 6. Cavitated with thin walls or with air-fluid level
III. Indeterminate or Noncontributory Factors 1. Age of lesion is unknown (no prior radiographs) 2. Noncalcified or eccentric calcification 3. 2-3 cm in size, with smooth margins *From Batra P, Brown K, Aberle DR, et al: Imaging techniques in the evaluation of pulmonary neoplasms. Chest 101:239-244, 1992. Used with permission. tThe criteria in each section are "additive" (e.g., the presence of two or three criteria has greater impact than one alone). ÎMajor criteria of benignancy.
and extrapulmonary intrathoracic manifestations in addition to the pulmonary parenchymal manifestations (see Table 5-5).46 In a review of the chest roentgenograms of 600 patients with carcinoma of the lung,35 the average lung cancer mass at radiologic presentation was 3 to 4 cm in diameter. A larger parenchymal mass was present in 22 per cent and a smaller mass in 20 per cent; multiple masses were present in only 1 per cent. Obstructive pneumonitis, collapse, or consolidation was present in 41 per cent. A hilar abnormality, either alone or associated with other abnormalities, was present in 41 per cent of the patients. The various extrapulmonary intrathoracic manifestations, with mediastinal widening, pleural effusion, and raised hemidiaphragm being the most common of these, were present in 11 per cent. Certain roentgenographic patterns are characteristic of the various cell types.47 Calcification is the most reliable sign of benign disease, especially if the lesion is a peripheral nodule. Since squamous cell carcinoma has a propensity to obstruct a bronchus, squamous cell carcinoma most often presents the picture of obstructive pneumonitis, collapse, consolidation, or cavitation. A hilar abnormality is also usually present with a squamous cell carcinoma. Adenocarcinomas are most often peripheral masses, and two thirds of these are larger than 4 cm. Hilar abnormalities, obstructive parenchymal lesions (atelectasis, abscess), and cavitation are infrequent to rare. Large-cell undifferentiated carcinomas are most likely to be peripheral lesions and are larger than 4 cm. Cavitation is infrequent, and
Preoperative Cardiopulmonary Evaluation
hilar abnormalities and parenchymal changes are present in approximately 30 per cent of cases. Small-cell undifferentiated tumors appear primarily as hilar abnormalities (80 per cent), and a parenchymal obstructive lesion occurs in approximately 40 per cent. b. BRONCHOSCOPY The examination of the tracheobronchial tree with either flexible fiberoptic (by far most common) or rigid bronchoscope should be done in almost all patients suspected of having a tumor of the lung. Indeed, 44 per cent of lesions that completely obstruct a bronchus (segmental to lobar), that are observed with a flexible fiberoptic bronchoscope, will have no radiographic sign of the obstruction.36 An exception to performing bronchoscopy may be made in patients with a very small peripheral lesion with no evidence of hilar or mediastinal lymph adenopathy. Direct visualization of the tumor, positive biopsy findings, positive bronchial brushing or trap suction specimens, or some combination of these three bronchoscopic findings is obtained in a high percentage of the patients. The technique of transbronchial needle aspiration has clearly made mediastinal and hilar nodes accessible to biopsy, thereby obviating the need for further invasive and costly surgical staging (e.g., mediastinoscopy) in some cases.37 Transbronchial needle aspiration of peripheral masses and subtle endobronchial lesions has greatly increased the diagnostic yield of bronchoscopy.37 Cell type influences the rate of positive finding; small-cell tumors are identified proportionately more often than are squamous cell or large-cell, undifferentiated tumors, and adenocarcinomas are identified least frequently of all. In addition to actual assessment of the tumor, other valuable information may be obtained at bronchoscopy. For example, the length of normal bronchus proximal to the tumor and the status of the carina (subcarinal nodes) may be determined, both of which are determinants of the exact surgical procedure to be performed (lobectomy, pneumonectomy, sleeve resection, or inoperable). In addition, bronchoscopy may reveal the presence of a second central tumor that was not visible on chest X-ray. C. SPUTUM CYTOLOGY
Cytologic examination of sputum has been found to be positive in approximately half of patients suspected of having carcinoma of the lung. With appropriate cytologic study of several sputum specimens, tumor cells may be found in a higher percentage of patients. In one study, one or two sputum samples yielded a 59 per cent positive result, three sputa a 69 per cent, and four sputa a 90 per cent positive result.50 A false-positive inci-
dence of only 1 per cent was found, although in most laboratories this incidence is reported to be in the range of 2 to 3 per cent. Cell type as determined by cytologic study agrees with that of the final histologic diagnosis in approximately 85 per cent of the patients. Well-differentiated epidermoid carcinomas, undifferentiated small-cell carcinomas, and adenocarcinomas can all be effectively typed by cytology. The undifferentiated large-cell carcinomas, the poorly differentiated epidermoid carcinomas, and combined carcinomas are more difficult to type correctly. Cytologic studies are most often positive in patients with large central tumors that communicate with the main bronchi (e.g., squamous cell). Peripheral parenchymal lesions frequently do not communicate with a large bronchus, and cytologic studies in patients with such lesions are less rewarding (e.g., adeno cell). d. NEEDLE BIOPSY
Percutaneous, transthoracic needle biopsy (with either fluoroscopic or computed tomographic [CT] guidance) has been suggested as a routine procedure for indeterminate (and noncommunicating) solitary peripheral lesions.48-50 With these lesions, this procedure is very useful when sputum and bronchoscopic methods fail to establish a definitive histologic diagnosis.51 The procedure can be performed on central lung, mediastinal, and pleural lesions and abscesses as well. Although percutaneous fine-needle aspiration is very accurate when positive (97 per cent), the false-negative rate may be as high as 30 per cent.52·53 Usually fine needles (18-23 gauge) are used but with CT scan guidance, larger needles (providing more accurate diagnosis) can be used.54 Percutaneous needle biopsy causes a significant incidence (7-14 per cent) of a small pneumothorax, which unfortunately requires chest tube drainage in 25 to 50 per cent of the patients with this complication.51·55·56 The anesthesiologist must be aware of the history of a recent needle biopsy because of the possibility of the development of a tension pneumothorax with the commencement of positivepressure ventilation. Finally, when all else fails to establish a diagnosis, open-lung biopsy may rarely be necessary.57 e. SUMMARY OF DIAGNOSIS OF LUNG CARCINOMA
The diagnosis of lung carcinoma is almost always certain, but in 20 to 25 per cent of patients a histologic cell type diagnosis is not known preoperative^. For these patients, the findings at surgery (which often begins with an open-lung biopsy) complete the first step in the workup.
Preoperative Cardiopulmonary Evaluation
E. Staging of Lung Cancer (Has the Carcinoma Spread?) Simply making the diagnosis of lung cancer alone is a grossly inadequate preoperative evalua tion. In order to devise a rational approach to treat ment, it is absolutely essential to determine the nature and extent of the disease in terms of any direct extension of the tumor to adjacent struc tures, metastasis of the tumor to the thoracic lymph node system, and extension of the tumor to extrathoracic structures. Any of these three forms of tumor extension may render the tumor inopera ble, and the preoperative diagnosis of such exten sion will greatly decrease the incidence of unnec essary thoracotomy and surgical morbidity and mortality. The need for precise classification of the
Table 5-6
165
anatomic extent or stage of lung cancers led to the application of the size of the tumor-nodal involve ment-metastasis staging system'· 2,58-66 to the dis ease; the international staging system for lung can cer64 provides a reference standard that has consistent meaning worldwide (Table 5-6). Staging is the quantitative assessment of malig nant disease and allows logical grouping of pa tients with a similar extent of disease of prog nostic, therapeutic, and analytic purposes. In bronchogenic carcinoma, a stage is assigned based on size, location, and the extent of invasion of the primary tumor as well as the presence of any re gional or metastatic disease. Selecting the most appropriate treatment for a patient with broncho genic carcinoma depends on precise staging. In brief, the Τ (for tumor) classification de-
NEW INTERNATIONAL NON-SMALL-CELL LUNG CARCINOMA PRIMARY SIZE OF TUMOR (T)-NODAL INVOLVEMENT (N)- DISTANT METASTASIS (M) STAGING SYSTEM
Factor
Classification
Description
Primary tumor size (T)
Tumors proven by the presence of malignant cells in bronchopulmonary secretions but not visualized roentgenographically or bronchoscopically, or any tumor that cannot be assessed, as in a retreatment staging No evidence of primary tumor Carcinoma in situ A tumor that is 3.0 cm or less in greatest dimension, surrounded by lung or visceral pleura, and without evidence of invasion proximal to a lobar bronchus at bronchoscopy A tumor more than 3.0 cm in greatest dimension, or a tumor of any size that either invades the visceral pleura or has associated atelectasis or obstructive pneumonitis extending to the hilar region; at bronchoscopy, the proximal extent of demonstrable tumor must be within a lobar bronchus or at least 2.0 cm distal to the carina; any associated atelectasis or obstructive pneumonitis must involve less than an entire lung A tumor of any size with direct extension into the chest wall (including superior sulcus tumors), diaphragm, or the mediastinal pleura or pericardium without involving the heart, great vessels, trachea, esophagus, or vertebral body, or a tumor in the main bronchus within 2 cm of the carina without involving the carina A tumor of any size with invasion of the mediastinum or involving heart, great vessels, trachea, esophagus, vertebral body, or carina, or presence of malignant pleural effusion.
Nodal involvement (N)
No demonstrable metastasis to regional lymph nodes Metastasis to lymph nodes in the peribronchial or the ipsilateral hilar region, or both, including direct extension Metastasis to ipsilateral mediastinal lymph nodes and subcarinal lymph nodes Metastasis to contralateral mediastinal lymph nodes, contralateral hilar lymph nodes, ipsilateral or contralateral scalene, or supraclavicular lymph nodes
Distant metastasis (M)
No (known) distant metastasis Distant metastasis present—specify site(s)
Stage grouping
166
Preoperative Cardiopulmonary Evaluation
scribes the size of the tumor and any direct exten sion of the tumor into surrounding tissues. T0 in dicates no evidence of a primary tumor; Tx indicates malignant cytology, but tumor is not seen roentgenographically or bronchoscopically. The remaining descriptors for the primary tumor sub divide the category into four levels. T, tumors are less than 3 cm in diameter and are completely surrounded by lung tissue. In contrast, T2 tumors are greater than 3 cm or involve the visceral pleura or cause lobar atelectasis. A tumor of any size that extends into the chest wall, diaphragm, or medias tinal pleura, which grows to within 2 cm of the carina or cause atelectasis of an entire lung, is labeled T 3 . A new class, T4 tumors, includes those involving structures that are typically unresectable (heart, great vessels, trachea, esophagus) or are associated with malignant pleural effusions. Regional lymph nodes (N) include N, nodes, those totally enclosed by visceral pleura, or N2 nodes, located in the ipsilateral mediastinum or subcarinal region. A new category, N-,, is assigned to nodes in the contralateral mediastinum, contra lateral hilum, or either supraclavicular or scalene region. The presence or absence of distant metastases (M) is normally specified according to site. The various T, N, and M categories are organ ized into various stage groupings (Table 5-6 and Figures 5-8 through 5-11). Stage I (Fig. 5-δ) 6 4 includes patients with T, or T2 tumors without
evidence for nodal (N0) or distant metastatic spread (M 0 ). Stage II cancers (Fig. 5-9) M are sim ilar to those in stage I except that N, nodes (peri bronchial, lobar, hilar nodes) are involved. Stage III has been subdivided into stage Ilia (Fig. 510)64 patients (=s T 3 , ^ N 2 ), who are usually can didates for definitive surgical treatment and stage IHb (Fig. 5-11 )M patients (T4 or N 3 ), who are normally not candidates for operative intervention. Stage IV includes patients with distant metastatic disease (M,). With this new, revised TNM staging system, both the designation of stage (I, II, Ilia, IHb, IV) and the T, N, and M status are all significant indi cators of prognosis. Each successive T, N, and M descriptor is associated with a worse prognosis, and additionally each successive stage also carries a poorer prognosis. 65 · 66 The breakdown of stages into three groups serves three purposes. First, the stage of the dis ease closely correlates with survival rate (see Prog nosis and Survival as a Function of the Staging of Lung Cancer) except for small-cell carcinoma (see section II.E.5. Staging in Small-Cell Lung Cancer, and section H.E.7., Prognosis and Survival as a Function of the Staging of Lung Cancer). Smallcell carcinoma is thought to have metastatic spread by the time of diagnosis; therefore, its natural his tory and behavior are independent of (and much more lethal than) the TNM system predictions. The diagnosis of small-cell carcinoma should be
Preoperative Cardiopulmonary Evaluation
167
168
Preoperative Cardiopulmonary Evaluation
T4 N3 MO Involvement of mediastinum, (ipsilateral and) contralateral mediastinal lymph nodes, contralateral hilar nodes, supraclavicular lymph nodes
Figure 5-11 Stage Illb disease. (From Moutain CF: A new international staging system for lung cancer. Chest 89:225S, 1986. Used with permission.)
made from the bronchoscopic procedures (brush ings, biopsy) or sputum cytology or needle biopsy (see preceding discussion). Second, the stage of the disease (I and II, approximately 30 to 35 per cent of all patients) usually dictates the surgical procedure of choice (see section II.E.6., Surgical Procedures as Dictated by Staging). Third, the di agnosis of stage III (approximately 60 to 65 per cent of all patients) may preclude surgery; there fore, staging tests are most vitally concerned with determining T 3 , N2, or M, disease. However, it should be noted that outstanding 5-year survival results (30 per cent) have been obtained with some stage III subsets (T3, N 0 -N 2 , M 0 ). 6 7 In summary, the introduction of the TNM stag ing system has encouraged an orderly assessment for selecting those cases most suitable for surgery and the type of surgery that should be performed. As a corollary to this appropriate staging should reduce the incidence of unnecessary thoracotomy to less than 20 per cent. 6 8 · 6 9 As a consequence of these improvements, the staging should result in an overall improvement in the present postsurgery 5- and 10-year survival rates of approximately 30 and 16 to 18 per cent, respectively. Although T, N, and M staging are actually done in parallel, for the sake of clarity they are discussed separately in the following sections (Fig. 5-12).
1. Τ Staging (Tumor Size and Direct Extension) Τ staging is primarily done by CT scanning and bronchoscopy. The advent of CT scans of the thorax (lung parenchyma, pleura, and medias tinum) has had a profound and simplifying effect on the Τ staging of lung cancer (see Fig. 5-12). CT scans of the thorax have been shown to pro vide a clear delineation of the tumor mass and can suggest direct tumor extension (particularly direct spread to the pleura with or without an accompa nying effusion and direct spread to the mediastinal structures) in many patients in whom the more conventional diagnostic radiologic methods fail to do so. For example, CT scanning in non-small-cell cancer when compared with conventional chest ra diology, tomography, and bronchoscopy increases the Τ stage in 40 per cent of cases 70 and in smallcell cancer, it increases the Τ stage from I or II to III in anywhere from 30 to 84 per cent of cases. 71 · 72 These Τ stage changes were due to direct tumor extension into the mediastinum, pleura, or dia phragm. Visualization of the mediastinum and hi lar structures by CT scanning may be markedly enhanced by intravenous contrast; this permits in tense opacification of the vessels and heart during the scan, allowing improved visualization of nonvessel structures.72
Preoperative Cardiopulmonary Evaluation
169
Figure 5-12 Preoperative evaluation logic of step II to determine whether lung carcinoma has spread beyond its local confines. Definition of T, N, and M staging and the various T, N, and M subsets are described in Table 5-6. The dashed lines indicate that surgery has recently been performed with an encouraging degree of success. 76 · 78 · 79, M S5 Step III is physiologic assessment of the patients for the planned surgical procedure. (CT = computed tomography.)
The results of and extra information gained from Τ staging with CT have to be interpreted carefully with regard to pulmonary nodules. CT detects 50 per cent more pulmonary nodules than 73 whole-lung tomograms. Most of these nodules are small, less than 6 mm, and serial follow-up shows that the majority (60 per cent) are benign. Thus, an additional intrapulmonary nodule (other than the primary) seen only with CT does not necessarily represent an intrapulmonary metastasis and is not necessarily a contraindication to sur gery. The proximal extent of a tumor in the airway is determined by flexible or rigid bronchoscopy. Bronchoscopy should be carried out to assess the lesion fully, to exclude synchronous tumors, and to plan the operative approach. Tumors within a major lobar or main bronchus but greater than 2 cm from the carina are classified as T 2 . Tumors less than 2 cm from the carina but not involving the carina are T3 lesions. Tumors that invade the
trachea or carina are T4 tumors. The geographic distribution of lung carcinoma is shown in Table 5-7. 40 Several other roentgenographic studies may demonstrate direct extension involvement of tho racic structures, which may preclude resection. Cardiac angiocardiography may demonstrate main-stem pulmonary artery invasion (at least l to 2 cm of a main-stem pulmonary artery that is free of tumor is required for pneumonectomy). Pulmo nary artery angiography may be necessary to dem onstrate pulmonary arteriovenous malformations and aortography to demonstrate pulmonary se questrations. A barium swallow may demonstrate fixation and/or distortion of the esophagus. Bron chography may be necessary for defining bron chopleural and bronchoesophageal fistulas. Azy gography may reveal mediastinal lymph node involvement as well as involvement of the vena cava by tumor. In a patient with suspected or diagnosed lung
170
Preoperative Cardiopulmonary Evaluation
Table 5-7 LOCATION OF LUNG CARCINOMA* Side of Lung Right lung Left lung Both lungs
Main Bronchus (%)
Upper Lobe (%)
Middle Lobe (%)
Lower Lobe (%)
4 4 2
33(19.4) 27(22.4)
5(8.3)
14(25.2) 12(24.7)
*From Humphrey WS, Smart CR, Winchester DP, et al: National Survey of the pattern of care for carcinoma of the lung. J Thorac Cardiovasc Surg 100:837-843, 1990. Used with permission. The data are given as a percentage of the total number of patients. The numbers in parentheses are the tissue mass of the given lobe as a percentage of the whole-lung mass.
cancer, a pleural effusion may be evidence for extension of disease beyond the primary tumor. Thoracentesis with cytologic analysis of the fluid provides specific staging and in some cases diag nostic information. If cytologic analysis of the fluid obtained by thoracentesis is not diagnostic, thoracoscopy should be considered. This can facil itate collection of greater amounts of pleural fluid for cytologic evaluation and allows a directed bi opsy of pleural based masses and peripherally lo cated primary tumors.5
2. Ν Staging (Metastasis to Lymph Nodes) The goal of preoperative evaluation of the medastinum is to identify those patients with N3 dis ease, exclude them from surgical consideration, and select those patients with operable N2 or less disease. N, nodes are segmental, lobar, interlobar, and hilar. N2 nodes are inferior mediastinal, aortic, and superior mediastinal. N-, nodes are any contra lateral nodes and ipsilateral scalene or supraclavic ular lymph nodes.74 The final precise enumeration of nodal involvement often requires staging at the time of operation (see section II.E.4., Intraopera tive and Postsurgical Staging). The preoperative assignment of Ν classification involves the use of CT of the thorax (lung paren chyma, hilum, and mediastinum) and mediastinos copy (Fig. 5-12). As with Τ staging, use of CT has also had a profound and simplifying effect on tumor staging. Although paratracheal and hilar lymph node adenopathy can usually be adequately identified by conventional radiology, CT scans are clearly superior in detecting subcarinal node en largement. Lymph nodes larger than 15 mm should be regarded as CT positive (but do not necessarily contraindicate surgery).72 CT com pared with conventional radiology causes an up staging of Ν status in 32 per cent of cases of lung carcinoma.70 The results of and extra information gained from CT scanning for nodes also have to be inter preted carefully. The demonstration of enlarged mediastinal glands should not automatically lead
to the conclusion that they are infiltrated by tumor. Glands are frequently enlarged and fleshy and may be subsequently found to be free of disease (e.g., centrally located squamous cell carcinoma can cause postobstructive pneumonia with reactive re gional lymph node involvement), and the falsepositive rate of CT mediastinal lymphadenopathy is about 25 per cent of cases.75 Consequently, a positive CT scan result still requires mediastinos copy. However, because CT tends to "overstage" tumors, thoracic surgical units without CT scan ners will achieve equally good Ν staging results (and perhaps better) 76 with routine preoperative mediastinal exploration. On the other hand, the predictive value of a negative CT scan is on the order of 90 to 95 per cent, and in such cases mediastinoscopy (see the following) is omitted by some surgeons before thoracotomy.75 However, it should be noted that, for the remaining 5 to 10 per cent, a normal-sized mediastinal gland does not mean that it is tumor free, and the results of studies comparing CT findings in the mediastinum with histologic evaluation show that there is a definite small false-negative rate. 7 5 · 7 7 Indeed, two authors found a significant rate of metastatic disease in nodes that were small on CT scan (< 1 cm) at surgery (i.e., the nodes were completely replaced by tumor but yet of normal size), especially with adenocarcinoma. 76,78 · 79 It is clear from this discussion that Ν staging with CT involves a moderate false-positive rate and a small false-negative rate. Many recent stud ies (N = 18; these studies are conveniently summarized in Daly et al.'s 80 Table 3) carefully quantitated the sensitivity, specificity, accuracy, positive predictive index, and negative predictive index (see legend of Table 5-8 for the definition of these indices) of CT scanning. Table 5-8 shows the accuracy of each Ν staging modality from one typical study and concludes (as do most studies) that mediastinoscopy is the most accurate staging investigation and recommends its continued rou tine use.81 Table 5-9 shows one author's actual guidelines for the use of mediastinoscopy (see chapter 14).82 In recent years, magnetic resonance imaging
Preoperative Cardiopulmonary Evaluation
171
Table 5-8 ACCURACY OF EACH STAGING MODALITY' Statistical Parameter!
MR
CT
CXR
Mediastinoscopy
Sensitivity Specificity Positive predictive value Megative predictive value accuracy
71.0 90.6 81.5 84.2 83.3
71.0 88.7 78.6 83.9 82.1
80.7 43.4 45.5 79.3 57.1
87.1 100.0 100.0 93.0 95.2
*From Patterson GA, Ginsberg RJ. Poon PY, et al: A prospective evaluation of magnetic resonance imaging, computed tomography, and mediastinoscopy in the preoperative assessment of mediastinal node status in bronchogenic carcinoma. J Thorac Cardiovasc Surg 94:679-684, 1987. Used with permission. tThe sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were calculated by the following formulas: Number of true positives Sensitivity X 100 Number of true positives + Number of false negatives Positive Number true positives predictive value = x 100 Number of true positives + Number of true negatives X 100 Specificity = Number of false positives Number of true negatives + Number of false positives Negative Number of true negatives predictive value X 100 Number of true positives + Number of true negatives + Number of true negatives Number of false negatives Accuracy X 100 Total number of instances Abbreviations: CXR = chest roentgenography; MR = magnetic resonance; CT = computed tomography.
(MRI) has been compared with CT scanning in its usefulness for staging lung cancer patients. MRI has advantages in that it is able to distinguish vascular from solid structures without the use of contrast materials (which CT scanning requires). MRI is still in its infancy from a technologic standpoint, requiring a long scanning time that re sults in motion degradation of images and poor spatial resolution. For these reasons, comparative studies between CT and MRI show no difference between the two or a slight to moderate advantage for CT in the assessment of both primary cancer and nodal involvement. 5 · 81 - 83 The present role of MRI should be confined to resolving specific quesTable 5-9 ACTUAL GUIDELINES FOR THE USE OF MEDIASTINOSCOPY* So mediastinoscopy Peripheral squamous cell carcinoma Undiagnosed peripheral nodule < 3 cm Cervical mediastinoscopy Right-sided centrally located squamous cell carcinoma Every undifferentiated carcinoma and adenocarcinoma (except for left upper lobe and left central tumors) \nterior mediastinoscopy Left upper lobe tumors Left-sided centrally located tumors (when anterior mediastinoscopy is negative, a cervical approach should be performed) *From Van Schil PEY, Van Hee RH GG, Schoots ELG: The ralue of mediastinoscopy in preoperative staging of broncho genic carcinoma. J Thorac Cardiovasc Surg 97:240-244, 1989. Jsed with permission.
tions in patients with possible cord involvement by the primary tumor or in whom the use of iodinated contrast material is contraindicated.5 Mediastinoscopy or mediastinotomy (anterior resection of a costosternal cartilage) should be car ried out as the final procedure to assess Ν staging when CT results are positive or when CT is not available. 84 · 85 Mediastinoscopy or mediastinotomy should be performed even if the mediastinum looks normal with conventional radiologic tech niques, since almost 50 per cent of patients with mediastinal nodal involvement do not have me diastinal widening on chest X-ray.86 Mediastinot omy is particularly important for left upper lobe lesions because it can provide better access to the left anterior mediastinal lymph nodes. If glands resected at mediastinoscopy or mediastinotomy are found to contain tumor (N 2 ), most would classify the case as inoperable.
3. M Staging (Metastasis) M staging begins with a complete history and physical examination that covers the function of all organ systems, with particular emphasis on brain, bone, and liver (Fig. 5-12). If the history or physical examination is positive for any given or gan system, then that particular organ system should be further investigated for metastases. The investigation should first involve an organ scan of some type (radioisotopic, CT). Organ scanning should not be done in the absence of symptoms;
172
Preoperative Cardiopulmonary Evaluation
5. Staging in Small-Cell Lung Cancer The very high incidence of mediastinal lymph node disease in SCLC and coexisting evidence of extrathoracic dissemination at the time of presentation mean that the T, N, and M classification has less bearing on SCLC prognosis. In fact, the situation at one time was thought to be so dismal that patients were classified only into limited (involvement of only one hemithorax) versus extensive disease, surgical resection had no place in the treatment in these patients, and chemotherapy was considered the main treatment modality.95·96 However, concern arose that these understandings and dictums needed revision when it was noted that some patients with ''limited" disease had an apparently complete response to chemotherapy and/ or radiation (i.e., SCLC is often initially exquisitely responsive to chemotherapy.97-|0° These reports suggested that surgery might be of value as a component of combined-modality therapy, under carefully defined conditions, and, indeed, subsequent studies have shown improved survival in stage I and II disease (for both stages the reported 5-year range is as great (and as good) as 10 to 70 per cent; see also Table 5-11).1()|-"° The dominant opinion in 1992 appears to be that the treatment of choice for stage I and II SCLC is curative surgical resection combined with pre- and postoperative chemotherapy and radiation treatment. Thus, TNM staging of clinically localized small-cell carcinoma is necessary to define the lesions for which surgical resection might be of benefit. Nevertheless, one must still keep in mind that, of all patients seen with limited SCLC, only about 10 per cent will have sufficiently early disease to warrant consideration for surgical treatment.1"
6. Surgical Procedure as Dictated by Staging All agree that, in the absence of contraindications, surgical excision of lung carcinoma offers the best likelihood of survival and quality of life and that the appropriate surgical procedure is one that excises all the carcinoma and yet, at the same time, preserves as much normal lung as possible."2 In this regard, the neodymium:yttrium-aluminumgarnet laser has been shown to be an excellent tool for removing parenchymal lesions while sparing surrounding normal lung tissue." 3 It is possible that socioeconomic factors may impact on the exact choice of lung cancer treatment."4, "5 Segmental resection is indicated for small (< 3 cm in diameter) stage I (T,, N0, M0) peripheral lesions and for patients with severely compromised cardiopulmonary status. Relative contraindications to segmental resection are T2 lesions or a
Preoperative Cardiopulmonary Evaluation
Lobectomy
173
Pneumonectomy or Bronchoplastic Procedure
Figure 5-13 Algorithm of operating-room decision for stage I and stage II disease based on the size of the primary tumor, its location, and the involvement of the hilar nodes. (From DeMeester TR, Albertucci M: Stage I non-small cell lung cancer: Surgical therapy. In Bitran JD, Golomb HM, Little AG. et al (eds): Lung Cancer: A Comprehensive Treatise. Orlando, Grune & Stratton. 1988, ρ I4l. Used with permission.)
lesion that crosses an intersegmental plane. Mor tality for segmental resection is less than 0.5 per cent." 6 At present, lobectomy is the operation of choice for carcinoma of the lung. It is indicated when technically feasible for all stage I and II disease and some stage III lesions. Lobectomy is not indi cated when the tumor involves structures in the pulmonary hilum, crosses an interlobar fissure, or involves the main bronchus. Lobectomy is associ ated with a mortality rate of l to 3 per cent." 6 ""* Indications for bilobectomy are tumor extending across the fissure, absent fissure, endobronchial tu mor, extrinsic tumor or nodal invasion of the bron chus intermedius, and vascular invasion." 9 The mortality rate associated with bilobectomy is 4 per cent." 9 Pneumonectomy is indicated with more exten sive disease, generally T 2 , N,„ M() or T ; , N,, M0 with hilar involvement. Figure 5-13 shows an ex ample of how the intraoperative decision/choice between pneumonectomy versus lobectomy is made based on intraoperative findings. The mor tality for pneumonectomy is about 5 per cent." 6 "" 8 T3 lesions (stage ΠΙΑ) will require additional re section of involved structures (chest wall, dia
phragm, pericardium, etc.). N2 lesions (ipsilateral mediastinal nodes, stage ΠΙΑ) may benefit from pneumonectomy combined with neoadjuvant ra diation and/or chemotherapy. 120-122 With few exceptions, patients with stage IIIB disease are not candidates for definitive surgical resection.67 The exceptions may include certain pa tients with central lesions and only hilar node in volvement, Pancoast's tumors, or certain lesions with chest wall involvement. Specially selected patients with limited carinal involvement (T4) have undergone resection for cure by sleeve pneumo nectomy or carinal resection (see Fig. 5-12). 8 4 ' 8 5 In the rare case of resectable solitary brain metas tasis, M, disease is potentially surgically curable (see Fig. 5_i2). 7 K · 7 9 ·^· 1 2 4 Patients who are not candidates for surgical re section may receive radiation therapy, chemother apy, and/or immunotherapy. Unfortunately, none of these latter therapeutic modalities have signifi cantly or dramatically impacted on survival rates for bronchogenic carcinoma. The use of chemo therapy for non-SCLC is toxic and at present very controversial. 125-129 Surgery, when it can be per formed, is still a patient's best and only hope for cure of bronchogenic carcinoma.
174
Preoperative Cardiopulmonc ι
Evaluation
I g 135.136 Approximately 65 per cent of the patients would be considered inoperable at presentation be cause of intrathoracic spread, the detection of extrathoracic metastases, or extremely poor lung function. The remaining 35 per cent would undergo a thoracotomy, although the figure may be nearer 20 to 30 per cent if mediastinoscopy was routinely performed prior to definitive surgery. At thoracotomy, 15 per cent of the original 100 cases would be found to be inoperable because of pre viously unrecognized tumor spread. The remaining 20 per cent would undergo "curative" resection. The 5- and 10-year survival rates for all persons undergoing resection would be 40 per cent and 16 to 18 per cent, respectively (eight and three to four patients, respectively, of the original 100 patients). "Curative" resection may fail so often in the lower stage because of the inability to detect small asymptomatic metastases before operation.
•o 61.4%
37.4% 25.8%
* · - · • » .
8.4% 48 η
60
Figure 5-14 Survival rates for 1479 patients in M0 group after re section of lung cancer, according to Τ classification (includes one patient with TX and five with TO disease). Differences between groups: Tl ver sus T2, ρ < .01; T2 versus T3, ρ < .01; T3 versus T4, ρ < .01. (From Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new in ternational staging system. J Thorac Cardiovasc Surg 96:440-447, 1988. Used with permission.)
Preoperative Cardiopulmonary Evaluation
175
Figure 5-15 Survival rates for 1479 patients in MO group after re section of lung cancer, according to postoperative Ν classification. Dif ferences between groups: NO versus Nl, ρ < .01; NI versus N2, ρ < .01; N2 versus N3, ρ < .01. (From Na ruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new in ternational staging system. J Thorac Cardiovasc Surg 96:440-447, 1988. Used with permission.)
Figure 5-16 Survival rates for 1737 patients after resection of lung cancer according to postoperative M classification. Differences between groups; M0 versus Ml. ρ < .01. (From Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg 96:440-447. 1988. Used with permission.)
176
Preoperative Cardiopulmonary Evaluation
Figure 5-17 Survival rates for 1737 patients after resection of lung cancer, classified by postoperative stage (four patients in stage 0 in cluded). Differences between groups: stage I versus stage Π, ρ < .01 ; stage II versus stage IIIA, ρ < .01; stage III A versus stage IIIB, ρ < .01; stage IIIB versus stage IV, NS. (From Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new in ternational staging system. J Thorac Cardiovasc Surg 96:440-447, 1988. Used with permission.)
tors being equal, e.g., cell type); possibly hor monal factor(s) may contribute to the prognosis.137 Although most studies have found that age greater than 80 years does not impose a higher risk of complications after lung resection for carci noma,138-140 two studies found that age greater than 60 years does.141·142 Preoperative protein depletion has been found to be as important as site of inci sion as a risk factor for postoperative pneumo nia.143 The available data suggest that postopera tive atelectasis is much more common in obese patients, but the risk of death is, in general, not 144 significantly increased. Aside from smoking/stopping smoking, there are three other aspects of poor lung function that can be easily diagnosed by history and/or physical
Table 5-10
POSTOPERATIVE 5-YEAR SURVIVAL IN PATIENTS WITH NON-SMALL-CELL LUNG CANCER ACCORDING TO THE NEW (1986) INTERNATIONAL STAGING SYSTEM
examination and can be treated or improved preoperatively; therefore, the patient should always be questioned about these areas. If the patient re ports a significant departure from his or her usual state of health with regard to any of these three aspects, then a preoperative pulmonary preparation regimen should be instituted (see chapter 6). First, all patients should be questioned about bronchospasm. Most patients who have bronchospasm (or wheezing) are very aware of varying degrees of airway resistance and chest tightness and of which medications relieve these symptoms. Auscultation of the chest for bronchospasm should also always be performed. If the amount of bronchospasm as reported by the patient, or heard by auscultation, is more than usual, then therapeutic levels of bronchodilator drugs should be administered (see chap ter 6). Second, all patients should be questioned about the amount of their secretions. If secretions are much more copious than usual, a few days of secretion removal (see chapter 6) may be of great benefit. Third, if the color of the sputum has re cently changed from mucoid to yellow or green, a sputum culture and sensitivity test should be per formed and appropriate antibiotic therapy insti tuted. 2. Pulmonary Function Tests If only studies that permit determination of pre operative and postoperative probabilities of mor bidity, mortality, sensitivity, and specificity are
Preoperative Cardiopulmonary Evaluation
177
Table 5-11 SURVIVAL RATES ACCORDING TO HISTOLOGIC TYPE AND STAGE OF DISEASES
*From Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg 96:440-447. 1988. Used with permission. tSixty patients with adenosquamous carcinoma and 27 patients with unclassified carcinoma are excluded.
considered, then it can be concluded that preoperative pulmonary function testing (spirometer) has a measurable benefit in predicting outcome in lung resection candidates.145 In selected patients, split perfusion lung scanning and pulmonary exercise testing are also useful.145 Therefore, it is thought
that pulmonary function testing is more sensitive than history or physical examination for detecting preoperative lung disease and the post-test probability of pulmonary complications and that patients with abnormal studies would benefit from preoperative respiratory therapy designed to decrease
Fate of 100 "Typical" Patients with Non-Small Cell Lung Carcinoma
Figure 5-18
The fate of 100 "typical" patients with non-small cell lung carcinoma.
178
Preoperative Cardiopulmonary· Evaluation
their risk for complication. In addition, in recent years, an effort has been made to use preoperative pulmonary function testing to predict which patients will develop intraoperative hypoxemia during one-lung ventilation (see chapter 11). However, it should be noted that it is only common sense to realize that not all the risk of thoracotomy for lung cancer is predictable and that some of it is of a random nature.140 In this section, the various individual whole-lung and regional lung function tests are described first, and then the sequence in which these tests should be performed is detailed. a. WHOLE-LUNG FUNCTION TESTS
All patients undergoing resectional thoracic surgery should have a timed expiratory spirogram (spirometry), a maximum breathing capacity, and an arterial blood gas analysis. Pulse oximetry may prove to be a useful preoperative screening test.146 Increasing consideration is being given to the routine performance of a flow-volume loop and some form of exercise tolerance test (see the following appropriate sections). The minimal pulmonary function criteria indicating increased risk to performing various-sized pulmonary resectional surgery are shown in Table 5-12.30· l47 Values below the ones listed in Table 5-12 are associated with a greatly increased incidence of postoperative respiratory complications.30·I47 The preoperative values listed in Table 5-12
Table 5-12
should be interpreted with three very important cautions in mind. First, technical and/or iatrogenic factors, such as bronchopleural fistula, overhydration, dehydration, bleeding, and transfusion, can certainly modify the postoperative course. Second, pulmonary function tests (spirometry) only test one essential component of the gas exchange system, namely, the ventilatory pump (lung mechanics) and ignore the other two essential components, namely, the transport system (the heart, vasculature and blood), and the gas-exchange systems (the V/Q matching in the lungs).I4H~|S7 Third, the predicted postoperative pulmonary function values (based on preoperative V/Q scanning) (see radioisotope regional lung functions later) are more predictive of postoperative function than preoperative whole-lung function tests14*'157 (see Fig. 5-23 to 5-25). (1) SPIROMETRY. The simplest spirometry test is measuring vital capacity. To measure vital capacity, the patient inspires maximally to total lung capacity (TLC) and then exhales either slowly (slow vital capacity) or forcefully (forced vital capacity, FVC) and completely, and the exhaled volume is recorded. With regard to the lung volumes shown in Figure 3-24, inspiratory capacity, inspiratory reserve volume, and expiratory reserve volume can all be measured directly from the vital capacity spirogram. The volume exhaled rapidly from one breath has
MINIMAL PULMONARY FUNCTION TEST CRITERIA FOR VARIOUSSIZED PULMONARY RESECTIONS*
*Data are based on Miller,30 Gass and Olsen,147 Bechard and Wetstein,14* Wahi et al.,144 Putman et al.,1-" Markos et al.,151 Nakahara et al.,153 Miller and Hatcher,1" Ferguson et al.,154 Eugene et al.,,5î Smith et al.,156 and Batton et al.1" Abbreviations: MBC = maximum breathing capacity; FVC = forced vital capacity; FEV, = forced expiratory volume in 1 sec; FEV,V,__7„ = forced expiratory volume between 25% and 75% of FVC; Vo:m.lx = maximum oxygen consumption.
Preoperative Cardiopulmonary Evaluation
been quantified in many ways; the most commonly used value is the volume of gas exhaled in the first second and is usually expressed as a percentage of the FVC (forced expired volume in 1 sec/FVC or FEV,%) (see Fig. 5-19). Normally, an individual can exhale 70 to 80 per cent of the VC in 1 sec; the remainder may take an additional 2 sec (FEV3). Patients with significant airway obstruction are able to exhale much less volume in the first second (the FEV,/FVC ratio is decreased), and they require a much longer time to deliver the entire volume, whereas in patients with lung parenchymal disease (i.e., restriction) the FEV,/FVC ratio is normal (but the FVC is decreased). The measurement of the maximum instantaneous "peak" expiratory flow rate (PEFR) is measured with an automated device such as an electronic flowmeter. As a patient forcefully exhales, expiratory flow rate increases rapidly to the PEFR (see Fig. 5-20, the flow-volume loop), and then the flow rate gradually decreases in parallel with the decrease in lung volume. The PEFR is defined as the highest expiratory flow rate sustained for at least 10 msec. Because flow rate at any given lung volume or moment is equal to the slope of the tangent line to the timed expiratory spirogram curve (such as the one shown in Fig. 519), and it may be difficult to accurately pick out the steepest slope, the PEFR cannot be measured easily or with precision by hand from a recorded spirogram. However, a comparable, albeit slightly lower value, the maximum expiratory flow rate (MEFR), can be derived by taking the average rate of flow (slope) over a 1-L segment of the early portion of the spirogram trace (Fig. 5-19). Usually, the segment between 200 and 1200 ml of expired volume is chosen; hence, the measurement is often referred to as FEF200_,200 the subscript designating the volume segment used. FEV,, PEFR, and MEFR are easily measured (electronically or with a dry-wedge spirometer) indices of large airway obstruction (normally 80 per cent of the airway resistance is in the large airways) and correlate well with gross functional impairment. Nevertheless, they are relatively insensitive to increases in the frictional resistance in small airways, which normally are responsible for 20 per cent of the airway resistance and therefore could be very diseased, but the disease would not be reflected in the PEFR, FEV,, or MEFR. The small airways are increasingly important at low lung volumes near the terminal portion of the FVC maneuver. Rather, FEV,, PEFR, and MEFR represent measurements of flow occurring within the first second of the FVC maneuver and at relatively high lung volumes. These measurements are highly dependent on patient effort (and are therefore variable) as well as on the resistance of the
0 1 2 TIME (seconds)
179
3
Figure 5-19 Spirogram of a patient taking two small tidal volume breaths, then an inspiration to total lung capacity, and then a forced exhaled breath to residual volume cycle. (MEFR = maximum expiratory flow rate—the average rate of flow [slope] over a l-L segment of the early portion of the expiratory spirogram trace. FEV, 0 = forced expired volume in l sec. FVC = forced vital capacity, which equals 100 per cent of the y-axis. MMEF = maximum midexpiratory flow rate, which is measured on the timed expiratory spirogram between 25 and 75 per cent of the FVC curve.)
larger airways. The value of FEV,, PEFR, and MEFR as screening tests is thus limited by the extent to which disease involves the larger airways. This helps to explain the relative lack of sensitivity of these measurements in detecting early mild (small airway) obstructive disease. However, these effort-dependent large-airway pulmonary function tests are the ones most responsive to and capable of detecting bronchodilator drug effects; they are, therefore, the most useful tests in the assessment of the presence of bronchospasm and the therapeutic benefit of these drugs. Indeed, in view of recent demonstrations that cigarette smoking causes an acute decrease in airway caliber and that cigarette smokers have a much increased airway responsiveness, pre- and postbronchodilator spirometry may be or should be more frequently performed.158·I59 The assessment of the reversibility of increased airway resistance employs beta-adrenergic drugs and spirometry. The maximum midexpiratory flow rate (MMEF) is measured on the timed expiratory spirogram during the middle half of the FVC curve between 25 and 75 per cent (see Fig. 5-19). However, it is not truly the "maximum" flow the (PEFR) because the latter occurs at a point much closer to the TLC. For this reason, use of "forced midexpiratory flow" has been advocated, with the more precise symbol of FEV25_75% The MMEF, by considering only the middle segment of the FVC. eliminates the initial highly effort-dependent and more variable segment of the trace. The flow rate during this segment has been shown to be largely
180
Preoperative Cardiopulmonary Evaluation
independent of patient effort and is slowed by obstruction of smaller airways. Because of this, the MMEF is accepted as an indirect measure of small-airway resistance and has been advocated as a sensitive test for the early detection of smallairway disease.160 The fact that successful treatment of left-sided cardiac failure results in marked improvement of FEV5(MK FEV73% FEV25%_75% but not in PEFR and FEV25% is strong support for the effort-independent small-airway hypothesis.161 (2) FLOW-VOLUME CURVES. The recording, during spirometry, of the expiratory flow plotted against volume, instead of time, produces a triangle-shaped envelope (Fig. 5-20). The virtually instantaneous peak is the peak expiratory flow (PEF), and the gradually decreasing flow rates follow the progressive airway narrowing down to zero flow at residual volume (RV). An inspiratory loop can also be obtained by asking the patient to take a maximal breath in from RV back to TLC (see Fig. 5-20). The semicircular shape of the normal inspiratory flow curve reflects the time taken for the respiratory muscles to generate maximal inspiratory force and flow; the maximal rate has usually been achieved by the time 50 per cent of the vital capacity (VC) has been inspired. See chapter 15 (Fig. 15-1) for the effect of relief of
major obstruction of a main-stem bronchus on the flow-volume curve. (3) CARBON MONOXIDE DIFFUSING CAPACITY.
The magnitude of carbon monoxide gas transfer (DL co ) depends on the volume and flow of blood through the pulmonary capillary bed (Q,) and the matching between ventilation and perfusion (V/Q) within the lungs. In the single-breath method, the patient breaths out to RV and then inspires a gas mixture of helium and 0.03 per cent carbon monoxide in air, to TLC. This is held in the lungs for 10 sec, and then the patient slowly exhales. An initial volume of 750 ml is discarded, and then a similarly sized sample of expired gas is collected and analyzed for the alveolar concentrations of helium and carbon monoxide. The change in concentration of the gases from the original inspired mixture enables two measurements to be obtained. First, helium is inert and is not absorbed; the change in helium concentration as a result of dilution within the lungs, therefore, provides a measurement of the volume of alveolar gas into which the gas mixture was inspired (VA). Second, the rate of uptake of carbon monoxide, a highly soluble gas that combines rapidly with hemoglobin in the capillaries, gives the DLC() as the uptake of carbon monoxide/minute/unit of partial pressure of the gas
Figure 5-20 Flow-volume loops demonstrating normal pattern. Diagram includes tidal volume (inner loop) and vital capacity (outer loop) loops. The vital capacity loop shows the V 50% and V 25% points. (PEF = peak expiratory flow; RV = residual volume.
Preoperative Cardiopulmonary Evaluation
(mmol/min/KpJ. D L c o is reduced when the alveo lar capillaries are reduced to number, there is in creased ventilation-perfusion mismatch, the hemo globin concentration is decreased, and the patient is a heavy smoker with a very high level of carboxyhemoglobin. D L c o is difficult to assess in pa tients in -whom line VC is iess than \ L οτ in children younger than 7 years. The uptake of carbon monoxide per unit of lung can be determined by dividing the D L c o by the alveolar volume, thereby determining the ¥^0 The Kco is useful in distinguishing a reduced transfer coefficient caused simply by a reduction in lung volume (reduced D L c o and normal Kço), as, for example, after pneumonectomy, from situations when the lung and its volume are both abnormal (reduced DL c o and reduced K ^ ) , as in emphysema. In one study, of 38 preoperative and operative risk factors, the DL c o was the most important predictor of morality (p < .01) and was the sole predictor of postoperative pulmonary complications (p < .005).I54 The authors concluded that the DL c o can reveal structural and functional abnormalities of emphysema even when spirometric values are acceptable. A DL c o less than 60 per cent of that predicted154 or a predicted postoperative percentage of less than 40 per cent151 is considered to be indicative of very high risk (Table 5-12). (4) MAXIMUM BREATHING CAPACITY. T h i s IS
the maximum amount of air that can be breathed in 1 min. It is expiratory effort dependent and reflects the total function of the entire cardiorespiratory apparatus. However, the maximum breathing capacity (MBC) may have unique value in that it also depends on the intangible variables of cooperation, motivation, and stamina in addition to cardiorespiratory function. Thus, the predictive value of MBC should be similar to an exercise test, such as bicycle ergometry or treadmill walking (see next subsection).162 (5) EXERCISE TESTING. Exercise increases utilization of oxygen peripherally and requires the entire interlocking lung/heart/vascular oxygen transport system to react. Thus, the potential exists to evaluate much of the cardiopulmonary system with just one test.163 Although exercise may be as simple as walking or stair climbing,157 exercise is usually undertaken on a cycle ergometer or treadmill. Initially, the patient cycles at rest with no resistance to the pedals or walks at a slow pace on the treadmill. At regular intervals of 1 to 3 min (according to protocol), the resistance to the cycle ergometer pedals or the speed and/or slope of the treadmill is increased by standard increments. At any increment, a steady state may be achieved for a submaximal end point, or the increments may be increased to a
181
maximal (symptom-limited) end point (onset of chest pain, breathlessness, or other discomfort or maximal allowable heart rate). During exercise, the following measurements should be made: minute ventilation, heart rate from an electrocardiogram (EKG), and mixed expired carbon aioxiûe and mixed expired oxygen (F E C0 2 and F É 0 2 ) concentrations. These measurements are used to calculate minute ventilation, oxygen uptake (Vo2), carbon dioxide production (Vco 2 ) and hence the respiratory exchange ratio R(Vco 2 /Vo 2 ). For each patient, heart rate and gas exchange are documented minute by minute and can be plotted against normal values (Fig. 5-21). The normal response to exercise consists of a linear increase in heart rate when plotted against oxygen consumption (Vo2) until maximum heart rate is reached. As workload is progressively increased, Vo 2 increases until a plateau is reached where further work produces no further increase in Vo 2 . This point is called a Vo 2max Typical values for a healthy 70-kg adult man are a resting Vo 2 = 245 ml/min (see chapter 3, Fick Principle) and a Vo 2 maximum of 2800 ml/min (cardiac output can increase fourfold and the arteriovenous 02 content can increase threefold).164 Minute ventilation increases progressively during exercise. The tidal volume increases until it reaches 60 per cent of the VC, and then ventilation increases solely by an increased breathing rate. Maximal exercise ventilation in the normal subject is 60 to 70 per cent of the MBC. In patients with lung disease, there is an increase in ventilation-perfusion mismatch and increased dead space. The increase in dead space requires a high minute ventilation to achieve an alveolar ventilation sufficient to maintain normal arterial blood gas tensions. Many patients with lung disease can maintain normal arterial gas tensions only at the expense of excessive minute ventilation. When lung disease is severe, arterial Po2 drops, and this is immediately detected by the oximeter if severe. Physiologic abnormalities capable of limiting Vo 2 at maximal exercise are anemia/ carboxyhemoglobinemia, heart disease, metabolic disease, muscular disease, peripheral vascular disease, pulmonary disease, and reduced effort. At some point, the utilization of oxygen by the muscles may exceed the oxygen availability provided by the transport axis. At this point, anaerobic mechanisms are used by the muscles, and lactate is produced. Information concerning this point may be obtained by direct measurement of the level of lactate in the blood.165 Considering these physiologic responses to exercise in patients with lung disease and cardiac disease, it is apparent there are similarities and dissimilarities (Table 5-13).166 Similar responses
182
Preoperative Cardiopulmonary Evaluation
Normal Response to Progressive Exercise Test
Diagnosis: normal control Figure 5 - 2 1 The V E and heart rate data for a 62-year-old man from a progressive exercise test plotted against the range of values for normal controls. (Modified from Spiro SG, Roberts CM. Lung function tests. Medicine International 3661, I 9 9 l . Used with permission.)
Table 5-13
CARDIOPULMONARY EXERCISE TEST RESULTS
Type of Test
Test
Pulmonary Limitations
Cardiac Limitations
Tests that do not differentiate between pulmonary and cardiac limitation (i.e., nonspecific)
Vo 2max Maximal heart rate Ventilation/work Pattern of breathing
Reduced Reduced Increased Rapid, shallow
Reduced Reduced Increased Rapid, shallow
Tests that do differentiate between pulmonary and cardiac limitations (i.e., specific)
Breathing reserve Blood gas
Reduced Hypoxemia, hypercapnia, respiratory acidosis
Normal
Anaerobic threshold Blood pressure response Electrocardiogram
Indeterminate or normal Normal Normal
Indeterminate or reduced Blunted Ischemia, arrhythmia
Preoperative Cardiopulmonary Evaluation
include reduced Vo2max and maximal heart rate, increased ventilation relative to workload, and rapid, shallow breathing. Dissimilar responses include reduced breathing reserve, hypoxemia, and hypercapnia in lung disease and low anaerobic threshold (expressed in ml 0 2 /kg/min), blunted blood pressure response, and abnormal EKG in cardiac disease. Patients may have both cardiac and pulmonary disease, and exercise testing may be helpful in establishing the disease process that is contributing the most to limitation. However, the precise percentages of contributions of cardiac and pulmonary limitations are difficult to determine. In a practical sense, it is most important to be aware that both diseases are present. Specific therapeutic trials can then be undertaken and the effect on exercise response measured. Several studies used exercise tolerance as a predictor of postpulmonary resection morbidity and mortality.148·,51·155, l56 The results to date have been very encouraging; if the preoperative value is below that shown in Table 5-12, then there is a 75 to 100 per cent chance of significant morbidity or mortality. (6) LUNG VOLUME. The functional residual capacity (FRC) is defined as the volume of gas in the lung that exists at the end of a normal expiration when there is no airflow and alveolar pressure equals the ambient pressure. The FRC is approximately 45 per cent of TLC. The expiratory reserve volume is additional gas below FRC that can be consciously exhaled and results in the minimum volume of lung possible, known as the RV. The RV is approximately 30 per cent of the TLC. Thus, the FRC equals the RV plus the expiratory reserve volume (Fig. 3-24). Total lung volume, FRC, and residual volume all contain a fraction (the residual volume) that cannot be measured by simple spirometry. However, since FRC is usually measured (by N2 washout, by He dilution, or plethysmographically; see chapter 3), TLC and RV can be easily derived by using the other lung volumes that are measured by simple spirometry. The size of RV is determined by the force of the expiratory muscles opposed by the tendency of the thorax to recoil outward at low lung volumes. Beyond the third decade of life, the RV increases as a result of dynamic compression or closure of airways; as a fraction of TLC, RV increases from approximately 25 per cent at the age of 20 years to 40 per cent at 70 years. Airway disease accelerates the increase in RV. Any airway narrowing or loss of elastic recoil, allowing dynamic compression, facilitates the trapping of gas within the lungs, and an increasing RV is a characteristic feature of early obstructive dis-
183
ease. The RV/TLC per cent is a sensitive measure of air trapping subsequent to airflow obstruction. To improve the efficiency of breathing, the obstructed patient is obliged to breathe at a volume above normal FRC, causing overinflation. These patients' lungs are overinflated in terms of the lung volume at which they pursue their tidal breathing. Recently, abnormal permanent enlargement of airspaces (emphysema) has been noninvasively and successfully assessed using CT scans taken at full expiration.167 b.
REGIONAL LUNG FUNCTION TESTS
When a tumor completely occludes a mainstem bronchus, then tests of whole-lung function evaluate the function of the nonaffected side; in other words, physiologically, the patient has already had a pneumonectomy. However, in the large majority of cases, tests of whole-lung function do not separate out the relative contribution made by the lung tissue to be resected from the contribution made by the lung tissue that would remain after resection to the total preoperative ventilation and gas exchange. In fact, ventilation-perfusion scans have shown that, apparently, small tumors may greatly distort the distribution of ventilation or perfusion or both, and this may result in a misleading interpretation of whole-lung function, especially if there is coexistent underlying generalized disease such as chronic bronchitis.168 Thus, tests of wholelung function may fail to resolve whether the patient would survive a resection without being left unduly dyspneic in cor pulmonale. The regional lung function tests consist of radioisotope studies (radiospirometry, which is the most important one) and somewhat less important nonradioisotope studies (the lateral position test, bronchial blockade, main-stem pulmonary artery blockade). (1) RADIOISOTOPE REGIONAL PERFUSION, VENTILATION-PERFUSION STUDIES (RADIOSPIROMETRY).169"171 In the past, regional lung function was measured by differential bronchospirometry. The procedure involved considerable patient discomfort due to double-lumen endotracheal tube insertion (usually into an awake patient) and technical and physiologic uncertainties related to the required use of the Fick principle. Presently, rightleft split pulmonary functions are obtained with easily performed noninvasive '"Xe radiospirometry and '"Xe and macroaggregate ("Te) perfusion scanning. Usually, three studies are performed to determine right-left split pulmonary function (regional ventilation, regional perfusion, regional lung volume [and regional ventilation/perfusion, ventilation/volume, perfusion/volume relationships by appropriate division of the primary three num-
184
Preoperative Cardiopulmonary Evaluaţii
bers]; Fig. 5-22). First, l33 Xe (or w T e or l 3 l I· MAA)' 47 is infused as a bolus intravenously (Fig 5-22A). Because l33 Xe is a very insoluble gas almost all of the l33 Xe (95 per cent) evolves out o: the blood into the alveoli, and this is respired ou of the body into an open system. Regional externa chest scintillation counters record these events ai a time (x-axis) versus activity (y-axis) curve. The time versus activity curve shows a rapid vascular to-alveolar wash-in peak and then an exponentia alveolar-to-environment washout. Since l33 Xe ap pears in the airspace only in proportion to the perfusion of the region, the regional perfusion wash-in peak count divided by the total counl equals the regional fractional perfusion. Since "Te and l3l I-MAA simply lodge in the pulmonary mi crocirculation, the regional activity count divided by the total chest count with these isotopes more directly yields the regional fractional perfusion.
Figure 5-22 Three studies are usually performed to deter mine regional lung function. Regional perfusion (A) is deter mined by intravenous injection of an insoluble radioisotope, which distributes to the two lungs according to the blood flow to each lung. Peak radioactivity over each lung (peak counts with time) is proportional to each lung blood flow. Regional ventilation (B) is determined by having the patient inhale an insoluble gaseous radioisotope. Peak radioactivity over each lung (peak counts with time) is proportional to each lung ven tilation. Regional volume (C) is determined by equilibrating radioactive material within the lungs with a connected closed space. The plateau radioactivity of each lung (plateau count with time) is then proportional to the volume of each lung. See text for more detailed explanation.
In the second study, a vital capacity breath of a bolus of l33 Xe results in a ventilation wash-in peak (environment-to-alveolar) followed by an expo nential alveolar-to-environment washout of l33Xe (Fig. 5-22Z?). Regional ventilation wash-in peak counts divided by the total chest count yield the regional fractional ventilation. In the third study, l33 Xe is again administered as an intravenous bolus, but after the l33 Xe evolves into the alveoli from the vascular space, it is res pired into a closed system (Fig. 5-22C). After 10 to 15 min of equilibration, the l33 Xe should have reached all gas spaces within the lungs (slow and fast time constants), and the concentration of 133Xe should be uniform throughout the lungs and the attached closed system. The regional radioactive counts/unit time then is proportional to the re gional volume of lung. Radioactive-scanning radiospirometry therefore yields regional perfusion, ventilation, and lung volume. Dividing regional perfusion and ventila tion by regional lung volume results in regional perfusion and ventilation per unit lung volume. Finally, dividing regional ventilation by regional perfusion results in the regional ventilation-perfu sion ratio. Conventional whole-lung function tests (VC, FEV,, maximal breathing capacity) cannot predict postpneumonectomy lung function because the amount of lung function to be removed is other wise unknown. However, multiplication of preop erative whole-lung pulmonary function tests by the percentage of lung to be removed (or that percent age of lung that will remain), as determined by radioactive-scanning radiospirometry, should the oretically predict postoperative pulmonary func 147 tion testing. Specifically,
Preoperative Cardiopulmonary Evaluation
Predicted postoperative FEV, = Preoperative FEV, X [% perfusion to "remaining" lung/100] For example, if the preoperative VC and FEV, was 2.00 L and 1.40 L, respectively, and the perfusion and/or ventilation of the lung to be removed was 40 per cent of the total ventilation, then the predicted postoperative VC and FEV, would be 1.20 L and 0.84 L, respectively. In the absence of adequate resolution of the V/Q scans (especially at the lobar level) prediction of postlobectomy function and the loss in function can be calculated as:172 Loss of function = preoperative FEV, X [# of functional segments in lobe to be resected/total # of segments in both lungs] (where the total number of segments in both lungs = 42). Combining radiospirometry with conventional pulmonary function tests has, in fact, resulted in a fair to good correlation between predicted and measured postpneumonectomy pulmonary function testing, including FVC, MBC, FEV,, FRC, DL co and exercise capacity (Vo2max).173-178 Figures 5-23, 5-24, and 5-25 show the results of one study that compared per cent preoperative perfusion to the lung to be resected to the change in
185
several different indices of pulmonary function (FEV,, DL co %, K co %, VEmax, V o 2 m J ; obviously, the correlation for all indices was very good.178 On the basis of this type of pulmonary function analysis, the literature suggests that a predicted postoperative FEV, of 0.8 or greater yields an "acceptable" incidence of postoperative morbidity and mortality.147,169·l73, l75 However, in view of the fact that a short, elderly, thin women probably does not need as large a postpneumonectomy FEV, as a tall, young, muscular man, the predicted postoperative percentage of predicted FEV, should be used for determinations of operability rather than the absolute value of FEV, (800 ml in Table 5-12). Specifically, Postpneumonectomy FEV, as % normal = [predicted postpneumonectomy FEV,/normalFEV,] X 100 The lowest proposed cutoff for FEV, is 30 per cent (see Table 5-12); for other acceptable predicted postpneumonectomy values as a per cent of normal, see Table 5-12. Preoperative ventilation-perfusion scans appear to identify those patients at greatest risk of hypoxemia during one-lung anesthesia (as a result of continued perfusion of the nonventilated, operative
·/· preoperative perfusion to resected lung 10 20 30 40 50
60
Figure 5-23 Relationship between observed changes in FEV, (y-axis) and percentage of preoperative perfusion to resected lung (x-axis) in 28 patients (r = -0.86, y = -0.788x + 0.14). The broken lines indicate 95% confidence limits. (From Corris PA. Ellis DA, Hawkins T, Gibson GJ: Use of radionuclide scanning in the preoperative estimation of pulmonary function after pneumonectomy. Thorax 42:285-291, 1987. Used with permission.)
186
Preoperative Cardiopulmonary Evaluation
lung). Figure 4-9 provides good evidence that hypoxemia during one-lung ventilation results, in part, from persistent blood flow to the operative lung.179 Although the correlation between the per cent of flow to the operative lung and the one-lung ventilation P a 0 2 is good, the preoperative scan cannot take into account the influence of posture, anesthesia, and surgery on pulmonary shunt and ventilation-perfusion matching (see chapter 8). Taken altogether, the studies discussed in the section on regional lung scanning and their high degree of accuracy in predicting intraoperative and postoperative function, combined with the availability of lung scanning in most hospitals and the ready acceptance of the test by patients, make the quantitative lung scan a most useful preoperative regional lung function test.
(2) LATERAL POSITION TEST. This has had some success in approximating individual lung ventilation.180 When a patient with normal lungs turns from the supine position to the lateral decubitus position, there is an increase in total lung FRC because the increase in nondependent lung FRC (due to less exposure of the nondependent lung to the weight of the mediastinum and pressure of the abdominal contents) exceeds the decrease in dependent lung FRC (due to greater exposure of the dependent lung to the weight of the mediastinum and the pressure of the abdominal contents).181-183 Thus, the procedure is thought to be a test of the expansile and ventilatory capabilities of the nondependent lung. The test is performed by having the patient breathe continuously into a spirometer while se-
Preoperative Cardiopulmonary Evaluation
187
Figure 5-25 (a) Relationship between observed changes in maximum ventilation (VF max (y-axis) and percentage of preoperative perfusion to resected lung (x-axis) in 14 patients (r = -0.89, y = -0.56x 0.03). The broken lines indicate 95% confidence limits, (b) Relationship between observed changes in Vo2 max (y-axis) and percentage preoperative perfusion to resected lung (x) in 14 patients (r = -0.90, y = -9.7x -0.09). The broken lines indicate 95% confidence limits. (From Corns PA, Ellis DA, Hawkins T, Gibson GJ: Use of radionuclide scanning in the preoperative estimation of pulmonary function after pneumonectomy. Thorax 42:285-291, 1987. Used with permission.)
quentially turning from the supine position to one of the lateral decubitus positions, back to the supine position, then into the other lateral decubitus position, and finally back into the supine position. The increase in total lung FRC is easily and noninvasive^ determined by noting the rise in the end-expiratory level of the continuously recorded spirogram; that is, the sine wave tidal ventilation on the spirogram first shows a gradual slope upward and then a plateau at a new stable level. The results are expressed as the increase in FRC in one lateral decubitus position divided by the increase in FRC in both lateral decubitus positions. The relative increase in FRC in each lung has correlated well with oxygen consumption and minute ventilation determined bronchospirometrically184 in each lung and with radionuclide ventilation-perfusion studies when the lung disease was symmetrical,185 and moderately well when the lung disease was asymmetrical,186 but predicts postpneumonectomy FEV, only fairly well.187 The lat-
eral position test is most likely to give false information whenever the disease process has little effect on the compliance of the lung in question. It should be stressed that since radionuclide ventilation-perfusion studies are well tolerated, safe, and simple to perform and provide exquisitely detailed information about both regional ventilation and perfusion, they must be considered the tests of choice of regional lung function. However, radionuclide studies require the use of sophisticated equipment, which may not be available in all hospitals, and it would seem that the lateral position test might be a useful alternative in this situation. (3) REGIONAL BRONCHIAL BALLOON OCCLUSION. Postpneumonectomy (or after any resection) ventilatory function can also be simulated preoperatively by passing, with the aid of a fiberoptic bronchoscope, a balloon occlusion catheter, which can occlude either lung (or any lobe), and then measuring spirometry of the remaining lung tissue (after careful withdrawal of the bronchoscope).
188
Preoperative Cardiopulmonary Evaluation
Supplemental oxygen must be administered during bronchial blockade because the blocked segment would still be perfused, and all of this perfusion would be right-to-left shunt flow, which would create a risk of hypoxemia. The regional bronchial balloon occlusion method of predicting postpneumonectomy ventila tory function has been studied during cycling ex ercise with a steady-state load equivalent to walk 188 ing at a brisk pace for that patient. The effects on minute ventilation and oxygen uptake were ob served during occlusion of the bronchus to the diseased lobe. If the patient was able to continue cycling and maintain the same workload during occlusion, this was regarded as evidence that they would withstand resection of the occluded lung tissue. All patients who could maintain the work load during preoperative bronchial occlusion were able to do so postoperatively. (4) REGIONAL PULMONARY ARTERY BALLOON
Postpneumonectomy pulmonary cir culatory and right ventricular function can be sim ulated preoperatively by passing into the main pul monary artery on the side to be resected a pulmonary artery catheter (using fluoroscopy) that has a 5-ml balloon at the tip of the catheter and a port for measuring pressure that is proximal to the balloon. Inflation of the distal balloon functionally resects the vasculature distal to the balloon; the pulmonary artery balloon inflation can be done with and without exercise. Under these conditions, all the pulmonary blood flow is diverted to the lung that will remain after the pneumonectomy, and the distensibility and compliance of the re maining pulmonary vascular bed are therefore tested. However, the test is unphysiologic in that the blocked lung would still be ventilated, and all the ventilation to this lung would be dead-space ventilation, which would not be present postpneu monectomy. If the mean pulmonary artery pres sure increases above 40 mm Hg, the P u C0 2 in creases above 60 mm Hg, or the P a 0 2 decreases below 45 mm Hg, it is likely that the patient will not be able to tolerate resection of that amount of pulmonary vascular bed without development of right ventricular failure and cor pulmonale.147· I74 189 l y 3 Because this test is highly invasive, requires special equipment, and is technically difficult, it is rarely clinically used today. OCCLUSION.
Simultaneous balloon occlusion of both a mainstem pulmonary artery and bronchus should com pletely simulate the physiologic effects of pneu monectomy and provide the most realistic assess ment of total postpneumonectomy lung function; there would be no acute increase in shunt and/or dead space. However, since this potentially very accurate simulation of the postresection condition is so invasive and complicated, it can only be
regarded as a research tool; the combined blockade test has not yet been reported in humans. C. SEQUENCE OF TESTS FOR LUNG RESECTION SURGERY With special reference to the performance of a pneumonectomy, there is a consensus that pulmo nary function testing should proceed in three I47 l74 l89 m 196 phases (Fig. 5-26). · · · ~ The first phase evaluates total lung (both, bilateral lung) function and consists of arterial blood gas measurements, simple spirometry, and perhaps carbon monoxide diffusion capacity and exercise capability. In creased risk is present when hypercapnia is found on a room air blood gas sample, the FEV, and/or the MBC is less than 50 per cent of that predicted, the RV/TLC is greater than 50 per cent, and/or D L c o per cent or exercise tolerance is markedly reduced. If any of these whole-lung pulmonary function values are worse than these stated limits, testing should proceed to the second phase, which evaluates the function of each lung separately (sin gle, unilateral lung function) and consists of meas urement of the ventilation and perfusion of each individual lung (as a fraction of the total) by ra dioisotopic ('"Xe, "Te) scanning. Combining right-left fractional lung function tests with con ventional spirometry and/or D L c o or Vo 2max should yield the appropriate predicted postoperative pul monary function. If this second level criterion can not be met and surgery is still contemplated or desired, the postoperative condition of the patient can be simulated (the third phase of testing; see also next section on pulmonary vascular function testing) by functionally resecting the vascular bed of the lung to be taken out by temporary balloon occlusion of the major pulmonary artery on that side with and without exercise. An increase in mean pulmonary artery pressure above 40 mm Hg (or an increase in P a C0 2 above 60 mm Hg, or a decrease in P a 0 2 below 45 mm Hg) indicates an inability to tolerate the removal of this amount of lung. 147 · I74 · ι 8 9 - | 9 λ This pulmonary function test cas cade is logical because it starts out with simple, inexpensive, and noninvasive tests and only in creases the degree of difficulty, expense, and in vasiveness as necessary; thus, in practice, the third phase of testing is rarely performed. In interpreting the results of this preoperative pulmonary function test cascade, physicians should always ask them selves what is an acceptable surgical mortality risk in a disease that has close to 100 per cent natural history 5-year mortality rate (even if the patient receives radiotherapy).197 Although less restrictive pulmonary function test criteria for operability for pulmonary resec tions less than pneumonectomy have been pub lished (see Table 5-12), 198 there are several rea-
Preoperative Cardiopulmonary Evaluation
189
Preoperative Evaluation of Masses of the Lung and Bronchi Step III: Respiratory Function
Whole Lung Function
Post-Operative Function Figure 5-26 Sequence of tests to determine pulmonary function for lung resection surgery (step III). The first group of tests determines whole lung function, the second test determines regional lung function, and the third test mimics postoperative lung function. See text for full explanation and definition of abbreviations.
190
Preoperative Cardiopulmonary Evaluation
sons why in some patients it may be prudent to think of a lobectomy (and lesser procedures) as a functional pneumonectomy.199 First, in the immediate postoperative period, the function of the lung tissue remaining on the operative side may be significantly impaired by atelectasis and perhaps infection; consequently, these patients may experience significant transient postoperative functional impairment.200 Patients who are most likely to have a stormy postoperative course with minor resections are those in whom the surgeon had intraoperative exposure problems that required severe and prolonged lung handling, retraction, compression, and packing. Intraoperative exposure problems are more likely to occur when the lung under operation is large and moving (large tidal volume positive-pressure ventilation). Second, at the time of thoracotomy, more accurate staging of the disease is possible, and it may become apparent that it is necessary to do a pneumonectomy.194 Third, the function of the lung on the nonoperated side may be impaired preoperative^ 200 and may acutely deteriorate further intraoperatively (aspiration and/or spillage of blood and/or pus from the operated to the nonoperated lung, inability of the nonoperated lung to tolerate dependency and compression in the lateral decubitus position). Finally, postoperative studies have shown that, although the ventilation and perfusion of the lung remaining on the operated side increase significantly during the long-term interval (3 to 51 months) after lobectomy, the volume of the remaining lung increases even more, so that the ventilation and perfusion per unit volume of the remaining lung decrease; this is equivalent to hyperinflation.176 The compensatory hyperinflation represents dilatation of the pre-existing respiratory units (as opposed to hyperplasia [alveolar multiplication] or hypertrophy [increase in existing mass by active growth])201 without disruption or fragmentation of the elastic tissue as seen in pathologic emphysema; however, the pulmonary hyperinflation decreases the compliance and, therefore, the ventilation per unit volume of the ipsilateral remaining pulmonary tissue. In addition, the hyperinflated lung stretches and thins out the capillaries in the alveolar walls, which decreases the perfusion per unit volume of the remaining pulmonary tissue on the ipsilateral side.
3. Pulmonary Vascular and Right Ventricular Function and Testing Assuming that there are no intracardiac shunts, right ventricular output is equivalent to that of the left ventricle, but because the pulmonary artery pressure is much lower than systemic blood pressure, the design and performance of the two ventricles are quite different. The right ventricle has a
thin lateral free wall (it pumps against a low pressure) and is a highly compliant chamber that better accommodates increases in filling pressure. Systolic pressures in the right ventricle and pulmonary artery are usually less than 30 mm Hg. At rest, right ventricular end-diastolic pressure is less than 6 mm Hg and mean PAP ranges from 10 to 18 mm Hg. During exercise, recruitment of small arterioles and capillaries of the pulmonary vascular bed compensates for an increase in cardiac output, and mean PAP rarely exceeds 20 to 30 mm Hg. Thus, in the healthy individual, vigorous demands are seldom made of the right ventricle (if the demands are excessive, acute dilation may occur, but hypertrophy will not).202 Most patients with pulmonary tumors have had a long history of smoking; consequently, they have varying degrees of chronic obstructive pulmonary disease (COPD). The cardiovascular response to the chronic pathologic alveolar and airway changes in COPD consists of the development of pulmonary hypertension and increased pulmonary vascular resistance (secondary to hypoxic pulmonary vasodilation, the destruction of alveolar septae, increased blood viscosity resulting from polycythemia and thromboembolism, and vascular compression caused by air trapping)202 followed by right ventricular hypertrophy, dilation, and dysfunction (cor pulmonale).203 Right ventricular dysfunction can lead to a shift of the interventricular septum within the pericardium and impair left ventricular filling and function. The prognosis of cor pulmonale in patients with COPD is poor; 3-year mortality is estimated at 60 per cent.204 Increased PVR has very important implications for patients undergoing pulmonary resection. Whereas a normal pulmonary vasculature is distensible and capable of accommodating large increases in pulmonary blood flow (to approximately 2 to 2.5 times greater than normal, as would occur through the remaining lung following a pneumonectomy) with only minor increases in PAP (Fig. 5-27), the relatively rigid and restricted pulmonary vascular bed of patients with chronic lung disease cannot accommodate even small increases in pulmonary blood flow without concomitant increases in pulmonary vascular pressure.205 The inability to tolerate increases in blood flow occurs over the entire range of physiologic cardiac output and may be an important contributing factor to the development of postpneumonectomy pulmonary edema when it occurs.206 The preoperative pulmonary function testing cascade as outlined in phases one and two in the section Sequence of Tests for Lung Resection Surgery (which, in reality, is the extent to which the large majority of patients are studied preoperatively) does not allow for specific diagnosis of
Preoperative Cardiopulmonary Evaluation
Cardiac Output (in multiples of resting output) Figure 5-27 Mean pulmonary artery pressure (y-axis) does not increase until cardiac output (x-axis) has been increased to 2 to 2.5 times when the pulmonary vascular bed is normal, whereas mean pulmonary artery pressure increases linearly when cardiac output is increased when the pulmonary vascular 204 bed is restricted.
increased PVR and right ventricular disease. In creased PVR may be noninvasively suspected pre operative^ by the presence of the auscultatory and radiographic signs of pulmonary hypertension and by electrocardiographic evidence of right atrial and ventricular hypertrophy (Table 5-14). In COPD patients without waking hypoxemia, cor pulmo
nale can be detected twice as sensitively and fre quently by echocardiography (definition criteria are pulmonary hypertension and right ventricular enlargement and/or hypertrophy) as by the clinical methods just discussed.207 Indeed, invasive preop erative monitoring (pulmonary artery catheteriza tion) in a consecutive series of 148 patients older than 65 years awaiting all types of major surgery showed a 50 per cent incidence of elevated mean PAP. 208 Clinically, the onset of cor pulmonale is most often indicated by the development of a pos itive hepatojuglar reflex, ascites, and peripheral edema. Not surprisingly, the incidence of death and total cardiac complications is increased in pa tients with pulmonary hypertension in the periop erative period.209 Measurements of PVR have been directly made by determining mean PAP and pulmonary wedge pressure at various levels of cardiac output pro duced by varying treadmill exercises (Fig. 5-28). Thus, using the patient's own cardiac output, pul monary vascular compliance can be determined. PVR measurements made in this way have been good indicators of risk for pneumonectomy.2I(V~:" Operative risk was considered to be increased if PVR was greater than 190 dynes/sec/cm~5. How ever, if the risk, expense, and time have been taken to insert a pulmonary artery catheter, then it is logical to take one further step and measure pul monary vascular pressures during temporary uni lateral pulmonary artery balloon occlusion in states of rest and exercise. This specifically tests the
Table 5-14 NONINVASIVE DIAGNOSIS OF PULMONARY HYPERTENSION ( ţ PAP), INCREASED PULMONARY VASCULAR RESISTANCE ( f PVR), RIGHT ATRIAL AND VENTRICULAR HYPERTROPHY ( f RA AND ţ RV), AND COR PULMONALE (CP)
Abbreviations: P-A = posterior-anterior.
191
Figure 5-28 Preoperative evaluation logic to determine cardiovascular function in patients with lung and bronchial carcinoma (step III). See text for full explanation. (CABG = coronary artery bypass grafting; CAD = coronary artery disease; EKG = electrocardiogram.)
Preoperative Cardiopulmonary Evaluation
compliance of just the pulmonary vascular bed that will remain after pneumonectomy. Temporary unilateral pulmonary artery balloon occlusion simulates the pulmonary vascular conditions to be expected after pneumonectomy (see Fig. 5-28). If significant pulmonary hypertension (P- > 40 mm Hg) or arterial hypoxemia ensues, pneumonectomy will likely not be tolerated because of the high risk of causing cor pulmonale, pulmonary edema, and low ventilation-perfusion relationships. Performing this procedure during exercise is the most realistic preoperative approximation of pulmonary vascular and right ventricular function that can be expected in the ambulatory postpneumonectomy patient.174189 nl93
4. Left Ventricular and Coronary Artery Function and Testing Ischemic heart disease is the number one cause of morbidity and mortality in the United States.212· 213 Cigarette smoking is thought to be responsible for an estimated 80 per cent of all heart attacks214,215 and 30 per cent of the fatalities caused by this disease.216 The lethality of smoking is understandable in view of the fact that patients with coronary artery diseases who smoke have significantly and substantially more active myocardial ischemia during daily life than patients who do not.213 The 1979 report of the U.S. Surgeon General on smoking and health217 leaves no doubt that smoking is a major independent cause of coronary artery disease and myocardial infarction. The adverse cardiovascular effects of cigarette smoking generally demonstrate a dose-response relationship in the sense that the risk increases with increased duration of smoking, increased quantity of cigarettes smoked, and the increased depth of inhalation of smoke.218 For example, the risk of a fatal or nonfatal coronary event in smokers of 25 or more cigarettes a day was more than 500 per cent the risk in nonsmokers; the risk in even the lightest smokers (one to four cigarettes a day) was more than doubled.214· 215 Thus, considering the usual age, the long and heavy smoking history, and the frequently sedentary lifestyle of patients undergoing thoracic surgery, it is not surprising that most of these patients have coronary artery disease and ischemic heart disease to some extent and that coronary artery disease is by far the most likely independent cause of left ventricular dysfunction. A number of mechanisms might explain the relationship between smoking and coronary disease. Carbon monoxide, one of the most poisonous byproducts of cigarette smoking, makes up approximately 2.7 to 6 per cent of the smoke219 and may cause damage by injuring vascular endothelium and enhancing the deposition of cholesterol in the
193
development of atherosclerosis. Nicotine, a key ingredient of cigarettes, acts to release catecholamines, which alter blood pressure and heart rate, thereby increasing myocardial oxygen demand. In combination with carbon monoxide, nicotine may be the predisposing factor in the development of coronary artery disease and myocardial infarction. In addition to increasing blood pressure and heart rate, nicotine also causes constriction of muscular arteries, which may interfere with regional tissue flow220 and has been shown to lower thresholds of ventricular fibrillation.221 Additionally, cigarette smoking increases the adhesiveness of platelets222· 223 and lowers high-density lipoprotein cholesterol levels.224,225 Myocardial ischemia leading to infarction may occur throughout the perioperative period, although peaks of incidence occur during operation226 and on the third day after operation.227 The first peak is caused by intraoperative changes in hemodynamics, and the second peak is caused by episodes of hypoxia, uneven administration of pain medication and withdrawal, or alteration of drug therapy.228 There are only two proven preoperative clinical predictors of perioperative cardiac morbidity (defined as occurrence of myocardial infarction, unstable angina, congestive heart failure, serious dysrhythmia or cardiac death during the intraoperative or in-hospital postoperative periods), and they were recent (less than 6 months) myocardial infarction and current congestive heart failure.229 The value of other historic predictors, such as previous (old) myocardial infarction, angina, previous congestive heart failure, hypertension, diabetes mellitus, and age is still unresolved. The classic historic intraoperative predictors of emergency surgery, prolonged (more than 3 hours) surgery, and thoracic or upper abdominal surgery also appear to be independent predictors of perioperative morbidity, whereas choice of anesthetic is not. The dynamic intraoperative predictors of perioperative cardiac morbidity are intraoperative hypotension and tachycardia. Hypertension remains a controversial predictor. Angina pectoris is a constellation of symptoms that reflect transient inadequacy of myocardial blood flow and can be classified as follows230: ( 1 ) classic angina, a transient precordial or substernal discomfort typically provoked by exertion and promptly relieved by rest or nitrates; (2) atypical angina, similar symptoms but with the absence of one or more of the criteria for classic angina (e.g.. the pain may not be consistently related to exertion or relieved by rest); (3) angina equivalent, usually the sensation of dyspnea as the sole or major manifestation (some physicians also label isolated pain referral, e.g., localized left arm or shoulder pain,
194
Preoperative Cardiopulmonary Evaluation
as an angina equivalent); and (4) variant or Prinz metal's angina, angina pectoris that occurs at rest and may manifest in stereotyped patterns such as nocturnal symptoms or symptoms only after exer cise. Variant angina is thought to be caused by coronary artery spasm with or without underlying coronary artery disease. The daily activities of the patient can be used to classify the frequency of angina (Table 5-15). Limitations of these classifi cations include the inability to categorize patients who are not accurate observers, overlap between classes, and inapplicability to the patient who does not have exertion-related angina. The EKG and the chest X-ray may help to con firm the presence of coronary artery disease and ischemic heart disease. The EKG may show Q waves (previous infarction), left bundle branch block, ST-segment elevation (transmural ische mia), ST-segment depression (subendocardial is chemia), T-wave inversions, and positive U wave (left main coronary artery disease). On the chest X-ray, cardiomegaly may be found in as many as 15 per cent of patients with coronary artery dis ease; the absence of cardiomegaly does not ex clude left ventricular dysfunction or even dila tion.231 If the history, physical examination, and EKG raise suspicion of coronary artery disease, then further perioperative evaluation of coronary artery function is necessary (see Fig. 5-29). a. EXERCISE-EKG TESTING The first step should be noninvasive exercise testing.232 Electrocardiography and thallium scans in that order appear to be the best such exercise tests at this time. Because exercise raises heart rate and blood pressure and increases myocardial oxy gen consumption, it provides an excellent assess ment for the probability of developing ischemia during the perioperative period. Exercise is most commonly performed on a treadmill in the United States. There are three basic treadmill protocols; the treadmill gradient may be increased progres sively with a constant speed (e.g., the Balke and
Naughton protocols), the speed is increased at a constant gradient (e.g., the Ellestead protocol), or both are increased (e.g., the Bruce protocol). Ex ercise EKG testing is highly predictive of subse quent cardiac events when the ST-segment changes are characteristic, large (> 2.5 mm), im mediate (first 1-3 min), sustained into the recovery period, or associated with the decrease in blood pressure.229 There are two major causes of false-negative results. First, intraoperative ischemia may not be associated with any hemodynamic changes.229 233. 234 s e c o n ( j ţ false-negative results most fre quently are due to inadequate exercise stress (gen eral debility, orthopedic limitation, or claudica tion). If an adequate exercise EKG stress testing workload is achieved, the sensitivity of exercise of EKG stress testing for the diagnosis of coronary artery disease is 62 to 80 per cent (mean, about 65 per cent) and the specificity is 83 to 96 per cent (mean, about 85 per cent). 229 The end points in clude workload and duration of exercise, ST-seg ment and T-wave changes, blood pressure and heart rate responses, and symptoms (chest pain or dyspnea). If the exercise EKG is normal, then sur gery should proceed: If the exercise EKG indicates ischemia, then coronary angiography should be considered. If the exercise EKG cannot be prop erly interpreted/performed (pre-existing EKG ab normalities, cannot exercise), then radionuclide angiography or exercise-thallium testing should be done. b. EXERCISE-THALLIUM TESTING Exercise-thallium imaging is another preopera tive test that can be used in patients who have abnormal EKGs suggestive of ischemic heart dis ease (and which may preclude interpretation of the EKG) (see Fig. 5-28).2-" In addition, thallium im aging can better identify the location and extent of myocardium at risk compared with conventional exercise electrocardiography.232 After intravenous injection of thallium-201, myocardial uptake of the
Table 5-15 NEW YORK HEART ASSOCIATION FUNCTIONAL CLASSIFICATION OF ANGINA PECTORIS Functional Class I II III IV
Occurrence of Symptoms With unusual activity With prolonged or slightly more than usual activity With usual activity of daily living At rest
Exercise Tolerance (Met) 7-8 (or more) 5-6 3-4 1-2
Functional Impairment Minimal or none Mild (can do light and general industrial work) Moderate (may be able to have desk job) Severe (incapacitated)
From Shub C: Angina pectoris and coronary heart disease. In (eds): Cardiology: Fundamentals and Practice. Vol. 2. Brandenburg RO\ Fuster V, Giuliani ER, McGoon DC. Chicago, Year Book Medical Publishers, 1987, pp 1073-1104. Used with permission.
Preoperative Cardiopulmonary Evaluation
radioactive tracer is dependent on regional myo cardial blood flow. In normal persons, the myocar dial uptake is homogeneous, whereas in areas of markedly reduced perfusion (> 90 per cent steno sis), the uptake after 30 to 60 min is markedly reduced, and a perfusion defect appears (cold-spot imaging) (as opposed to hot-spot imaging with technetium, which indicates increased uptake in an ischemia area). The appearance of thallium perfu sion defects in two or three separate left ventricu lar segments identifies approximately 75 per cent of patients with multivessel coronary artery dis ease, and approximately 95 per cent of patients with three-vessel coronary artery disease have ab normal scans.231 It should be remembered that pos itive results indicate areas of reduced blood flow (at a time when the rest of the coronary circulation is fully dilated secondary to exercise), but that does not necessarily mean that these areas are con currently ischemic. A period of thallium redistri bution occurs at 2 to 4 hours. A repeat image is performed at 4 hours. Defects in the 1-hour image and 4-hour image indicate infarcted tissue, whereas a defect in the 1 -hour tests alone indicates reversible or collateralized reduced perfusion areas. If the thallium exercise test is negative, then the planned pulmonary resection should proceed. If the thallium exercise scan is positive for ische mia, the coronary angiography should be done (see Fig. 5-28). C. DIPYRIDAMOLE-THALLIUM TESTING
For patients who cannot exercise adequately (which eliminates exercise EKG and thallium test ing) and for patients with already abnormal EKG (eliminates exercise EKG testing), dipyridamolethallium testing is an attractive alternative (see Fig. 5-28). Dipyridamole, like strenuous exercise, causes marked coronary vasodilation and a general increase in coronary blood flow. Because highly stenotic coronary vessels cannot dilate normally, areas of myocardium supplied by them will take up less thallium during scanning than areas sup plied by normal vessels. Later, after the vasodila tion has resolved, these underperfused areas will "fill in" as the thallium redistributes. Thus, a "cold" area on the scan that fills in during the later (redistribution) phase of the procedure im plies the presence of a highly stenotic coronary artery.236 The sensitivity and specificity of dipyridamolethallium imaging for detecting coronary artery disease are both approximately 90 per cent, ap proaching the sensitivity and specificity of thal lium-exercise scintigraphy.231 Like the exercisethallium scan, the dipyridamole-thallium scan reflects characteristics of coronary flow rather than ischemia, and therefore the test is limited to deter
195
mining the quantity of myocardium at risk from reduced perfusion distal to a coronary artery ste nosis and not the risk of developing myocardial ischemia at a critical heart rate. d. RADIONUCLIDE ANGIOGRAPHY
Radionuclide angiography provides useful in formation about global and regional left ventricu lar systolic function. The ejection fraction values obtained are reproducible and compare favorably with the results of contrast ventriculography. The development of exercise-induced regional wall motion abnormalities and an abnormal ejection fraction response during exercise radionuclide an giography provide sensitivity of 85 to 90 per cent for the diagnosis of coronary artery disease, which is higher than that with treadmill testing alone.231 Θ. ECHOCARDIOGRAPHY
The major use of two-dimensional echocardi ography in patients with coronary artery disease is to evaluate resting regional and global left ventric ular systolic function, although additional infor mation, such as assessment of cardiac valves, peri cardium, ventricular wall thickness, and the presence of left ventricular aneurysm and throm bus, can be obtained concurrently. Exercise twodimensional echocardiography has been shown to be useful in selected patients with coronary artery disease and compares favorably with radionuclide techniques. f. CORONARY ANGIOGRAPHY
Patients with a positive preoperative nuclear cardiology imaging test (see sections N.F.4. b., c, d.) that demonstrates a large reversible defect should undergo coronary catheterization if surgical intervention is considered a possibility. If, for any reason, there is a strong suspicion that the patient is indeed having significant angina, even though exercise testing is negative or equivocal, coronary angiography is indicated. Consideration should al ways be given to coronary angiography in the pa tient with proven previous myocardial infarction, especially if the patient currently has angina. If significant coronary artery disease is present, the patient needs coronary artery by-pass grafting before or at the time of pulmonary resection. For lesser degrees of coronary artery disease, pulmo nary resection for carcinoma of the lung should be done after appropriate medical therapy for coro nary insufficiency has been initiated. If the patient needs coronary artery by-pass grafting, and a lim ited resection can encompass the cancer, both pro cedures can be done under the same anesthetic, but the coronary artery by-pass grafting should be 237 2 3 8 done before pulmonary resection. After by pass, if the patient is stable with good myocardial
Preoperative Cardiopulmonary Evaluation
function and not bleeding, a pulmonary wedge resection can be done. For patients who require coronary artery by-pass grafting and have pulmonary lesions that require segmentectomy, lobectomy, or pneumonectomy, there is a good possibility that the prolonged nature of the pulmonary procedure will increase the operative mortality (and therefore should not be done), although a small number of successful combined procedures have been reported.238· 23y In one series of 21 patients, the pulmonary mass was discovered on the preoperative chest X-ray (for cardiac surgery), and therefore before the occurrence of any symptoms, and by definition constituted a fortuitously early diagnosis. Not surprisingly, resection at the time of cardiac surgery (17 wedge resections, four lobectomies) resulted in a 95 per cent 5-year survival.240 In cases that require large resections in very comprised patients, coronary artery by-pass grafting should be done first, and pulmonary resection should be delayed until the patient has gained weight and muscle mass (usually 4 to 6 weeks). The risk of general anesthesia for a noncardiac operation in the patient with previous coronary artery by-pass grafting is similar to that in patients without proven coronary artery disease.241·242 Although it is not possible to estimate the true effects of delay in pulmonary resection, in terms of tumor spread in a possibly immunocompromised patient (especially after general anesthesia243), it seems reasonable that in the latter group (those requiring by-pass grafting and major pulmonary resection) the operative risk of combined procedures probably exceeds the risk of tumor spread. 5. Cardiovascular Sequelae of Therapeutic Thoracic Radiation Heart disease resulting from therapeutic radiation is now well recognized. Once thought to be radioresistant, the heart, as well as the spinal cord, appears to be the dose-limiting organ for thoracic irradiation.244 The pericardium, myocardium, endocardium, valves, conduction system, and coronary arteries all may be affected by radiation. The common pathogenetic factor involves microcirculatory injury. How frequently radiation damages the heart depends significantly on the total dose, how it is given, how long the patient is observed, and the intensity with which cardiac abnormalities are sought (Table 5-16). Clinically significant radiation heart disease most frequently takes the form of pericarditis (including tamponade and constriction). Radiationrelated myocardial dysfunction is typically mild or subclinical but can include cardiomyopathic heart failure, particularly of the right ventricule. Valvular thickening and deformity are common findings
Table 5-16.
FACTORS INFLUENCING INCIDENCE OF RADIATION HEART DISEASE
Total number rad delivered (must be > 3000 rads) Number rad per treatment (inversely proportional) Time span over which radiation is given (inversely proportional) Techniques used Anterior port versus anterior posterior ports Use of apical and subcarinal shields (?) Presence and bulk of mediastinal tumor (?) Age of patient (older than 40 years) Length of follow-up
in irradiated hearts that come to postmortem examination. The aortic valve is involved most often, with the mitral valve next. Multiple causes for coronary artery disease in cancer patients exist, including age-related atherosclerosis (most common), compression or infiltration of an artery by tumor, embolization of tumor fragments into an artery, a hypercoagulable state, and chemotherapy (principally fluorouracil), as well as radiotherapy. Therefore, it is understandable that coronary events occur in young patients previously exposed to mediastinal radiation who have no other risk factor. The most devasting noncardiac but rare sequela or mediastinal radiation is myelitis. Characteristically beginning 6 months to 2 years after radiation, the process is generally irreversible, and the end result is functional transection of the cord. Mediastinal radiation may cause acute pneumonitis as well as a more chronic form of pulmonary fibrosis. Significant mediastinal fibrosis must also be considered in the patient with dyspnea after radiation. 6. Anesthetic Implications of Cancer Chemotherapy Cancer chemotherapeutic agents act through a variety of mechanisms. In general, these agents are more effective against small rather than large tumor burdens; kill a fixed percentage of tumor cells, not a fixed number; affect rapidly dividing normal as well as neoplastic cells, especially those of gastrointestinal or surface epithelium and bone marrow; and are more effective against dividing rather than testing cells. After repeated exposure to antineoplastic agents, tumor cells may become resistant, and combinations of drugs with different mechanisms of action must be devised. The antineoplastic drugs are classified into several groups on the basis of the point in the cell cycle at which they act, their chemical nature, and their biologic source. Table 5-17 summarizes the major agents included in each category (common
Preoperative Cardiopulmonary Evaluation
197
Table 5-17. COMMON CHEMOTHERAPEUTIC AGENTS Type of Drug
Generic Name
Trade Name
Chemical Name
Alkylating agents*
Nitrogen mustard* Melphalan (L-PAM) Cytoxan* Chlorambucil Myleran Thiotepa Amsacrine
Mustargen Alkeran Cytoxan Leukeran Myleran
Mechlorethamine 1-phenylalanine mustard Cyclophosphamide Chlorambucil Busulfan Triethylenethiophosphoramide M-amsa
Antimetabolites
Methotrexate* 5-FU 5-FUDR Ara-C 6-TG 6-MP
Methotrexate
Amethopteran 5-fluorouracil 5-floxuridine Cytosine arabinoside 6-thioguanine 6-mercaptopurine
Cytosar
Mitotic inhibitors (vinca alkyloids)
Vincristine* Vinblastine Vindesine
Oncovin Velban Eldisine
Antibiotics
Adriamycin* Daunomycin Bleomycin* Mithramycin Mitomycin C Actinomycin D Streptozocin
Adriamycin
Nitrosoureas and miscellaneous agents
Carmustine Lomustine* Methyl-CeeNU Dacarbazine Procarbazine* L-asparaginase C/i-platinum* Hydroxyurea Etoposide*
Blenoxane Mithracin Mutamycin Cosmegen
Vinca leukoblastine Doxorubicin hydrochloride Daunomycin Bleomycin Mithramycin Mitomycin C Dactinomycin Streptozocin
BiCNU CeeNu
Bis-chloroethylnitrosurea Cyclohexylchloroethylnitrosourea
DTIC-Dome Matulane
Dimethyltrianzeno-imidazole carboxdamide
Cisplatin Hydrea VePesid
Cis-dichlorodiamminoplatinum Hydroxyurea
''Drugs used against cancers relevant to thoracic surgery.
name, trade name, chemical name). Table 5-18 shows the commonly used dosages of chemotherapeutic agents against SCLC.245 The alkylating agents covalently cross-link strands of deoxyribonucleic acid (DNA), thus preventing replication and subsequent transcription into ribonucleic acid (RNA). In contrast, the nitrosoureas with alkylating properties inhibit both DNA and RNA synthesis; they also demonstrate significant lipid solubility. The antibiotics, doxorubicin and daunorubicin, bind to DNA by intercalation, inhibit both DNA and RNA synthesis, and cause chromosome breaks. The antibiotics cause some of the most diverse cell toxicities, making them particularly notable for anesthesiologists. The mitotic inhibitors are plant alkaloids that cause metaphase arrest by binding to microtubular protein involved in the formation of the mitotic spindle. The antimetabolites act by interfering with the synthesis of new nucleic acids. Because this synthesis occurs only in actively dividing cells, antimetabolites are considered to be cycle specific.
The enzyme L-asparaginase is unique in that it depletes tumor cells of the amino acid asparagine. which is essential for their propagation. The classic chemotherapy complications are bleomycin pulmonary toxicity, doxorubicin cardiac toxicity, vincristine neurotoxicity, and the effects of the alkylating agents on neuromuscular function. Many different antineoplastic drugs have been associated with pulmonary toxicity. They are summarized in Table 5-19. No matter the particular agent, antineoplastic drug-induced pulmonary toxicity presents in a similar manner. In general, most of the patients develop a pneumonitis that can ultimately lead to pulmonary fibrosis. Bleomycin's action is similar to the effect of radiation, and bleomycin and radiation may act synergistically during simultaneous therapy.246 Pulmonary toxicity is the most life-threatening, drug-limiting effect, reported in 15 to 25 per cent of patients.247248 The predisposing factors for bleomycin pulmonary toxicity appear to be patient age greater than 70 years, a total dose greater than 450
198 Preoperative Cardiopulmonary Evaluation Table 5-18.
SMALL-CELL LUNG CANCER: COMMONLY USED REGIMENTS OF COMBINATION CHEMOTHERAPY* Regimen
Drug Cyclophosphamide Lomustine Vincristine •
Etoposide Cyclophosphamide Doxorubicin Vincristine Etoposide
"
' • · •
Doxorubicin Cyclophosphamide
1 g/m2 intravenously every 4 weeks, day 1 70 mg/m2 by mouth every 4 weeks, day 1 1.3 mg/m2 intravenously (maximum = 2.0 mg) every 4 weeks, day 1, except first 4 weeks, when administered weekly 70 mg/m2 by mouth, on days 3,4, 5 and 6, every 4 weeks 1500 mg/m2 intravenously every 3 weeks 40 mg/m2 intravenously every 3 weeks 2 mg/m2 intravenously every 3 weeks 50 mg/m2 intravenously 5 days every 3 weeks 45 mg/m2 intravenously every 3 weeks 1 g/m2 intravenously every 3 weeks
•From Hanson HH, Kristijansen PEG: Changing concepts in the management of patients with lung cancer. Med J Aus 149:77-84,1988. Used with permission.
units, radiation therapy to the chest, and pre-existing lung disease.249 Signs and symptoms of pulmonary toxicity are cough, dyspnea, and basal rales. The disease may manifest itself with minimal radiologic changes and normal resting P a 0 2 , or it may progress to severe hypoxemia at rest, with radiologic changes similar to severe adult respiratory distress syndrome. A controversial factor that may predispose patients to pulmonary toxicity is the administration of oxygen in high concentrations (see chapter ll). 2 5 0 Mitomycin C is also highly toxic and can cause pulmonary fibrosis and nephrotoxicity.251 Thus, patients receiving mitomycin C should have their pulmonary and renal status fully evaluated in the preoperative period. Drugs associated with cardiac toxicity are listed in Table 5-20. The most severe cardiac toxicity occurs as a cardiomyopathy, which can be lethal. Toxicity secondary to doxorubicin includes severe cardiomyopathy, seen in 1.8 per cent of patients treated. When cardiomyopathy develops, it has been shown to be irreversible in 60 per cent of the Table 5-19.
patients, with death occurring within 3 weeks of the onset of the symptoms.252 The left ventricular failure that occurs with doxorubicin is refractory to inotropic drugs. EKGabnormalities are also part of doxorubicin toxicity, but they resolve 1 to 2 months after cessation of therapy.253 Once doxorubicin administration is discontinued, left ventricular function generally continues to show deterioration in patients older than 40 years, whereas it may improve in younger patients.254 Table 5-21 lists the drugs that have been associated with neurotoxicity. The mitotic inhibitors, vinblastine and especially vincristine, pose the greatest challenge to the anesthetic management of patients treated with them. Essential nervous system effects are rare (decreased consciousness), whereas peripheral neuropathy, such as induced by vincristine, is frequent. Antineoplastic drugs associated with reduced plasma cholinesterase activity include the alkylating agents, cyclophosphamide, nitrogen mustard, and thiotepa, listed in decreasing order of effect on cholinesterase. Cisplatin, methotrexate, 6-mercaptopurine, mithramycin, and streptozocin have been associated with major alterations of renal function. The hypomagnesemia induced by cisplatin and mithramycin is of particular interest. It may be associated with tetany and increased susceptibility to digoxin toxicity.
III. MEDIASTINAL MASSES •
A. Types of Masses The anatomy of the mediastinum is complex and contains many structures (see chapter 2) and cell types (see Fig. 2-23, 2-24, and 2-25). The most common tumors by location are as follows (Fig. 5-29): In the anterior mediastinum are thymoma, mesenchymal tumors, dermoid cysts, thyroid and parathyroid tumors, and lymphoma; in the middle mediastinum are pericardial cysts, bronchogenic cysts, and lymphomas; and in the posterior mediastinum are neurogenic and enterogenous tumors and cysts, aortic aneurysms, and paravertebral abscess.30 The most common tumors by cell type are neurogenic (14 to 24 per cent),1 cysts (pericardial, bronchogenic, enteric, and non-
..· • DRUGS ASSOCIATED WITH PULMONARY TOXICITY BY CATEGORY
Alkylating Agents
Antimetabolites
Antibiotics
Nitrosoureas
Miscellaneous
Busulfan Chlorambucil Cyclophosphamide Melphalan
Azathioprine Cytarabine 6-mercaptopurine Methotrexate
Bleomycin Mitomycin C
BiCNU Methyl-CeeNU
Procarbazine
Preoperative Cardiopulmonary Evaluation
Table 5-20 DRUGS ASSOCIATED WITH CARDIAC TOXICITY
Table 5-21
Cisplatinum* Cyclophosphamide Daunorubicin Doxorubicin*
B. Signs and Symptoms Signs and symptoms of mediastinal masses can be referable to any of several of the many organs within the mediastinum (heart, great vessels, the airway, esophagus, vertebral column). The lifethreatening complications are superior vena cava
DRUGS ASSOCIATED WITH NEUROTOXICITY L-Asparaginase Cisplatinum 5-Fluorouracil Methotrexate Nitrogen mustard Procarbazine Vinblastine Vincristine
*Drugs used against cancer relevant to thoracic surgery.
specific, totaling 19 to 25 per cent), and teratodermoids, thymomas, lymphomas, and others (each 10 to 20 per cent).253 2 5 5 · 2 5 6 Approximately 30 to 40 per cent of the nonvascular masses are malignant.256·257
199
obstruction (dyspnea, cough, orthopnea, facial swelling), pulmonary artery obstruction (postural hypotension), carinal obstruction (respiratory distress), and cardiac compression. Major, but nonlife-threatening, complications include esophageal compression, which can cause dysphagia, pressure on a nerve and erosion of bone, which can cause severe pain, and compression of the spinal cord, which may cause paralysis. More minor complications include hoarseness resulting from entrapment of the recurrent laryngeal nerve, Horner's syndrome caused by involvement of the high pos-
Mediastinal Masses
ANTERIOR MEDIASTINUM
MIDDLE MEDIASTINUM
POSTERIOR MEDIASTINUM
Aortic Aneurysm
Thyroid Adenoma Parathyroid Adenoma
Esophageal Lesions
Thymoma
Enterogenous Cysts
Mesenchymal Tumors Neurogenic Tumors: Pheochromocytoma Ganglioneuroma Neurofibroma Neuroblastoma
Terato Dermoid Cyst Lymphoma
Hiatus Hernia Meningocele
Lymphoma
Paravertebral Abscess
Bronchial Cysts Figure 5-29 The location of commonly occurring mediastinal tumors. See Figure 2-23 for the division of the mediastinum into conventional compartments.
200
Preoperative Cardiopulmonary Evaluation
terior mediastinal sympathetic nerves, lethargy, and weight loss. In one large series, the symptoms consisted of cough (40 per cent), pain (40 per cent), dyspnea (20 per cent), dysphagia (20 per cent), and hoarseness (3 per cent). The signs con sisted of weight loss (24 per cent), fever (24 per cent), superior vena cava obstruction (16 per cent), tracheal deviation (12 per cent), Horner's syn drome (7 per cent), spinal cord compression (5 per cent), cyanosis (3 per cent), and mediastinal wid ening (3 per cent). Symptoms were absent in 22 per cent, signs were absent in 20 per cent, and both signs and symptoms were absent in 20 per cent.257 Usually, mediastinal masses are suggested by a combination of clinical signs and symptoms and standard and high-voltage anteroposterior and lateral chest roentgenograms. Contour distortions of the normal interfaces between the lung and the mediastinum may occur even with small medias tinal masses, which, with enlargement, may pro ject into the adjacent hemithorax. Obvious points of diagnostic differentiation include vascular ver sus nonvascular mass, mass that can be surgically cured (thymomas, teratomas, neurogenic tumors) versus those that cannot be surgically cured (e.g., lymphoma), metastatic versus nonmetastatic, and those that require specific tissue typing for correct choice of chemotherapy and irradiation alone (lymphoma) versus those that require chemother apy, irradiation, plus resection (germ cell tu mor).2583
C. Diagnostic Workup Logic for Mediastinal Masses Once a mediastinal mass is suggested, CT is the single best test to perform in a stable patient to identify the nature and the location of the mass (Fig. 5-30). In fact, CT has been shown to be more useful in evaluating the mediastinum than any other region in the thorax. Mediastinal disease may easily be detected by CT even in the presence of a normal chest radiograph. CT can determine more accurately than other diagnostic procedures whether the mass is primarily vascular, fatty, cys tic, or soft tissue in nature. For example, fluoros copy, conventional tomography, barium esophagogram, radionuclide angiography and thyroid imaging, angiography, bronchoscopy, scalene node biopsy, and mediastinoscopy will all be nor mal in patients with mediastinal lipomatosis. 259,260 Echocardiography may be helpful in the recogni tion of the anatomic and functional aspects of me diastinal masses.261 Once a soft tissue mass is identified in the me diastinum, the next logical question is whether the process is benign or malignant. Approximately 30 to 40 per cent of mediastinal masses are malig nant. 257 Unfortunately, CT cannot distinguish be nign from malignant masses based on any intrinsic appearance of the mass itself. However, CT may demonstrate invasion of the pulmonary arteries, airway, pericardium, or myocardium or may dem-
Preoperative Evaluation of Mediastinal Masses
ι
History, Physical Examination, Chest X-rays
I I
Diagnosis Suspected
Figure 5-30 Preoperative evalua tion logic of mediastinal masses. (CT = computed tomography.) See text for full explanation.
Preoperative Cardiopulmonary Evaluation
onstrate pleural or parenchymal metastases. In each of these instances, the malignant nature of the mediastinal mass is more apparent. More commonly, the diagnosis of benign versus malignant will depend on a diagnosis of cell type, and this must come from more invasive procedures such as mediastinoscopy, mediastinotomy, needle biopsy, bronchoscopy, or esophagoscopy. Biopsy may be guided by fluoroscope, CT, or ultrasonography.262 Lymphomas or metastases require only diagnostic biopsy because irradiation or chemotherapy (or both) is the treatment of choice. The accurate diagnosis of thoracic lymphoma or new thoracic lesions in patients with lymphomas usually requires that enough tissue be taken for immunophenotyping. Providing adequate tissue samples and treating new lesions that are not lymphomas often requires major thoracomediastinotomies for immunophenotyping and more than one operation.263 Many germ cell tumors, even though they display elevated levels of «-fetoprotein or beta-human chorionic gonadotropin, require a confirming biopsy because they are treated initially with multimodality therapy followed by resection.25H If the mass is not suspected of being a metastasis, lymphoma, or germ cell tumor, then plans to resect the lesion primarily can proceed without preoperative biopsy if the patient is otherwise in good health. Thymomas, benign teratomas, cysts, and neurogenic tumors are types of lesions that can be resected without a preoperative biopsy. Resection not only resolves any symptoms caused by a space-occupying lesion but also determines the exact histologic diagnosis.258 If the lesion may potentially involve the aorta or any of its branches, this is best demonstrated by intra-arterial digital or conventional angiographic examinations.264 The evaluation of acute traumatic injuries of the thorax is still best accomplished using conventional arteriography and angiography (especially if the diagnosis of dissection of the aorta or its major branches is suspected). The assessment of physiologic function of the patient
Table 5-22
with a mediastinal mass is similar to that for a patient with lung cancer.
IV. PLEURAL DISEASE/EFFUSION A. Anatomy and Pathophysiology The pleura is the double serous membrane that separates the lung from the chest wall and mediastinal structures. The surface of each pleural membrane consists of a uniform layer of mesothelial cells supported on a connective tissue framework, which is well supplied with capillaries and lymph vessels. The parietal pleura also contains pain-sensitive nerve fibers supplied by the intercostal and phrenic nerves. In health, there is a continuous movement of water, electrolytes, and a little protein into the pleural space from the parietal surface; the hypotonic pleural fluid thus formed is reabsorbed across the visceral membrane, under the influence of hydrostatic and osmotic gradients (Starling's law). The parietal pleura is perfused by systemic arteries, whereas the visceral layer receives blood from the low-pressure pulmonary circulation. In a normal individual, the net volume of pleural fluid may be as little as 1.0 ml. The volume of pleural fluid will increase when there is a rise in intravascular hydrostatic pressure (e.g., in cardiac failure) or a fall in plasma oncotic pressure (e.g., in hypoproteinemic states). An increase in the permeability of the pleural capillaries, caused by inflammatory or neoplastic disease, leads to accumulation of fluid in the pleural space. Lymphatic obstruction can also contribute to the accumulation of pleural fluid. The causes of pleural effusion are listed in Table 5-22.
B. Symptoms, Signs and Diagnostic Workup Logic The normally negative intrapleural pressure is due to the elastic recoil of the lung pulling inward
CAUSES OF PLEURAL EFFUSION
Transudates A. Increased hydrostatic pressure 1. Cardiac failure 2. Constrictive pericarditis 3. Obstruction of superior vena cava B. Decreased oncotic pressure I. Hypoalbuminemia (liver and renal disease) C. Miscellaneous 1. Peritoneal dialysis 2. Acute atelectasis 3. Subclavian catheter misplacement
201
Exudates A. B. C. D. E.
Infections Neoplasms Collagen Intra-abdominal disease Miscellaneous 1. Drug-induced pleural disease 2. Pulmonary infarction 3. Hemothorax 4. Chylothorax or pseudochylothorax
ΓΟ
ο
ΙΟ
Preoperative Cardiopulmonary Evaluation
in opposition to the thoracic wall. At the relaxed, end-expiratory position of the breathing cycle, the intrapleural pressure is about 5 cm H 2 0 below atmospheric pressure. Accumulation of liquid or leakage of air into the pleural space is accompa nied by an increase in pleural pressure, a reduction in lung volume, and a slight expansion of the ipsilateral hemithorax. This results in a restrictive ventilatory defect, which is in proportion to the volume of liquid or gas in the pleural space. VC, FRC, and TLC are all reduced, and the transfer factor is also slightly reduced. The most common symptoms of pleurisy are localized inspiratory pain, dry cough, slowly in creasing breathlessness in larger effusions, and fe ver in infectious disease. On physical examination, there may be a pleural friction rub at the onset, restriction of chest wall movement, diminished tactile vocal fremitus, dullness on percussion, and decreased breath sounds with pleural fluid accu mulations of at least 300 ml. A scheme for diag nosis and management is outlined in Figure 5-31. V. PERICARDIAL DISEASE/ EFFUSION A. Anatomy and Pathophysiology The pericardium has three functions: mechani cal, membranous, and ligamentous. Mechanically, it protects against acute dilation of the ventricles and mediates diastolic coupling between ventri cles, and it is involved with the pressure-volume relationship between the ventricles. Membranous functions include lubrication and protection against infection; ligamentous function consists of limiting heart displacement.265 There are numerous causes of cardiac tampon ade (see Table 5-23). In medical patients, neo plasms, primarily of the breast and lung, are most Table 5-23 CAUSES OF CARDIAC TAMPONADE 1. Malignant disease (primarily breast and lung) 2. Trauma 3. Iatrogenic causes, including radiation therapy, central venous pressure placement, pacemaker placement, and cardiac surgery 4. Acute cardiac infarct with anticoagulation 5. Infections 6. Dissecting aortic aneurysm 7. Postpericardiotomy syndrome 8. Chylous effusions 9. Connective tissue diseases, primarily systemic lupus erythematosus and rheumatoid arthritis 10. Amyloidosis 11. Myxedema 12. Uremia
203
Table 5-24 HEMODYNAMIC ALTERATIONS IN CARDIAC TAMPONADE 1. 2. 3. 4. 5. 6. 7. 8.
Limited atrial and ventricular diastolic filling Increase and equalization of diastolic pressures Absent or positive y descent and prominent χ descent Decreased stroke volume Increased heart rate Decreased pulse pressure Increased systemic vascular resistance Decreased cardiac output
commonly responsible for cardiac tamponade.266 Traumatic causes include direct, penetrating, or blunt insults to the heart. Hemopericardium also may be seen after dissecting ascending aortic aneurysms, myocardial infarction, or anticoagula tion.267 Table 5-24 summarizes the hemodynamic alter ations during cardiac tamponade. Limited diastolic ventricular filling resulting from increased intrapericardial pressure causes a fall in cardiac output and stroke volume. The fall in stroke volume and subsequently hypotension elicit a generalized sym pathetic response, which, in turn, causes tachy cardia, increased systemic vascular resistance, decreased pulse pressure, increased preload, de creased diastolic filling time, and increased con tractility with subsequent decrease in cardiac per fusion.268 B. Diagnosis and Treatment Symptoms seen with cardiac tamponade are usually nonspecific. The patient may complain of dyspnea, fullness of the head and neck, abdominal pain, nausea, or a vague oppressive feeling in the chest. The dyspnea is believed to be secondary to decreased cardiac output and to restriction of lung volumes by pericardial and pleural effusions. Signs of cardiac tamponade are summarized in Table 5-25. Acute tamponade is well described by Beck's triad: a small, quiet heart; increased venous pres sures, and hypotension. The triad is caused by pericardial fluid softening heart sounds; resistance
Table 5-25 SIGNS OF CARDIAC TAMPONADE 1. Beck's triad: small, quiet heart; increased venous pressures; and hypotension 2. Pulsus paradoxus 3. Equalization of diastolic pressures 4. Large and globular cardiac shadow radiographically 5. Nonspecific electrocardiogram changes 6. Right-heart compression during diastole by echocardiography
2U4
Preoperative Cardiopulmonary Evaluation
to right-heart filling, resulting in increased jugular venous distention and neck vein dilation; and a reduction in cardiac output with subsequent hypo tension, decreased pulse pressure, pulsus para doxus, and tachycardia. Increased central venous pressure (CVP) may not be seen in the patient with extreme volume depletion. Cardiac catheterization shows equalization of CVP, diastolic right ventricular pressure, pulmo nary capillary wedge pressure, left atrial pressure, and LVEDP. Radiographically, one sees an en larged and globular cardiac shadow with a convex or straight left-heart border. Definitive treatment of cardiac tamponade is drainage. It may be accomplished through pericar diocentesis or surgical decompression. The former is the treatment of choice for nontraumatic cardiac tamponade. Surgical drainage is usually chosen for traumatic tamponade, after cardiac surgery, or if fluid reaccumulates after pericardiocentesis. The surgical approaches include the subxiphoid ap proach, which is easier to perform and more easily tolerated with local anesthesia but affords a limited exposure. A left thoracotomy incision provides ex cellent exposure and is indicated if a larger field is required, such as after penetrating chest injury. Initial noninvasive management of cardiac tam ponade includes fluid administration and pharma cologic support. It is important to maintain ade quate preload to optimize stroke volume so that administration of a fluid load is necessary; ideally colloid, plasma, or blood is used. The volume should be titrated to the response of the initial load. CVP should be maintained above 15 to 20 cm H 2 0. The increases in filling pressures will oppose pericardial pressure.269 Positive inotropes may be considered. Isoproter enol may be ideal because it sustains a tachycardia (because stroke volume is usually limited, heart rate must remain high to preserve cardiac output), increases stroke volume, and decreases systemic vascular resistance. It is important to select anesthetic agents that do not depress the myocardium, severely decrease afterload, or cause bradycardia. With these limita tions in mind, low-dose ketamine is an attractive choice. Controlled ventilation significantly de creases preload and decreases cardiac output from the right ventricle. Positive end-expiratory pres sure further decreases preload and cardiac output. If it is not possible to perform pericardiotomy without general anesthesia, it is suggested that the patient be allowed to breathe spontaneously until the chest is opened and drainage of the pericardial space is imminent. If spontaneous ventilation is not possible, ventilation with high rates and low tidal volumes should be considered. Acute pulmonary edema has been reported after
relief of pericardial tamponade. 270 The acute in crease in venous return may overload the left ven tricle in the face of increased systemic vascular resistance secondary to elevated catecholamine levels. This mismatch of preload and afterload is exacerbated in patients with decreased left ventric ular compliance. Slow drainage of the effusion is recommended. Right ventricular filling pressures should be monitored.
REFERENCES 1. Shields TW: Carcinoma of the lung. In General Thoracic Surgery. Philadelphia, Lea & Febiger, 1983, chapter 54, pp 729-769. 2. Spiro SG: The diagnosis and staging of lung cancer. In Smyth JF (ed): The Management of Lung Cancer. Balti more, Edward Arnold, Ltd., 1984, chapter 3, pp 36-52. 3. Le Roux BT: Bronchial Carcinoma. London, E. & F. Livingstone, Ltd., 1968. 4. Jones DP: Diagnostic work-up of chest disease. Sympo sium on noncardiac thoracic surgery. Surg Clin North Am 60:743-755, 1980. 5. Ferguson MK: Diagnosing and staging of non-small cell lung cancer. Hemalol/Oncol Clin North Am 4(6): 10531068. 1990. 6. Garfinkel L, Stellman SD: Smoking and lung cancer in women: Findings in a prospective study. Cancer Res 48:6951-6955, 1988. 7. Spiro S: Lung cancer: Presentation and treatment. Med Int 3798-3805, 1991. 8. Petty TL: Pulmonary medicine. JAMA 263:2677-2678. 1990. 9. Tockman MS. Antonisen NR. Wright EC. Donathan MG: Airways obstruction and the risk of lung cancer. Ann Intern Med 106:512-518, 1987. 10. Fielding JE, Phenow KJ: Health effects of involuntary smoking. Ν Engl J Med 319:1450-1460, 1988. 1 I. Capewell S, Sankaran R, Lamb D, Mclntyre M, Sudlow MF: Lung cancer in lifelong non-smokers. Thorax 46:565-568, 1991. 12. Janerich DT, Thompson WD, Varela LR, Greenwald P, Chorost S, Tucci C, Zaman MB, Melamed MR, Kiely M, McKneally MF: Lung cancer and exposure to tobacco smoke in the household. Ν Engl J Med 323:632-636. 1990. 13. U.S. Office on Smoking and Health: Smoking and health: A report to the Surgeon General of the Public Health Service. Washington, DC, U.S. Department of Health and Human Services, 1979. 14. U.S. Department of Health and Human Services: The Health Consequences o\' Smoking for Women: A Report of the United States Surgeon General. Washington DC. U.S. Department of Health and Human Services, 1980. 15. Peto R, Doll R: The control of lung cancer. Ν Scientist 24:26-30, 1985. 16. Mathe G, Reizenstein P: Extra-pulmonary tumors caused by smoking. Biomed Pharmacother 42:87-88. 1988. 17. Remick SC, Hafez GR, Carbone PP: Extrapulmonary small-cell carcinoma: A review of the literature with em phasis on therapy and outcome. Medicine 66:457-471, 1987. 18. Hill GB: Smoking and lung cancer. Arch Intern Med 148:2538-2539, 1988. 19. Petruzzelli S. Hietanen E, Bartsch H, Camus AM, Mussi A, Angeletti CA, Săracei R, Giuntini C: Pulmonary lipid peroxidation in cigarette smokers and lung cancer pa tients. Chest 98:93-935. 1990.
CHAPTER
6
Preoperative Respiratory Preparation I. Introduction II. Correlation of Respiratory Complications with Degree of PreExisting Lung Disease III. Correlation of Respiratory Complications With Site of Operation IV. Proof That Preoperative Pulmonary Preparation Decreases Incidence of Postoperative Respiratory Complications V. Preoperative Respiratory Preparation Maneuvers A. Discontinue Smoking B. Dilating the Airways 1. Overall Bronchodilating Plan 2. Ş2-Agonists 3. Anticholinergic Drugs
4. Inhaled Steroids 5. Inhaled Sodium Cromoglycate 6. Methylxanthines 7. Oral and Parenteral Steroids C. Loosening the Secretions D. Removing the Secretions E. Ancillary Measures/Issues 1. General Measures (Treatment of Systemic Disease) 2. Treatment of Cor Pulmonale 3. Preoperative Digitalization F. Measures to Increase Motivation and Education and to Facilitate Postoperative Respiratory Care VI. Mechanism of Preoperative Respiratory Preparation Benefit VII. Premedication
212
Preoperative Respiratory Preparation
I. INTRODUCTION Thoracic surgical patients are at high risk for (he development of postoperative pulmonary complications. In most of the literature, "postoperative complications" refers to the development of atelectasis and/or pneumonia.' The incidence of pneumonia usually parallels the incidence of atelectasis, and the onset of pneumonia lags behind the onset of atelectasis because atelectasis provides the ventilatory and mucociliary stasis condition necessary for the development and culture of pneumonia.2,3 There are three major reasons why thoracic surgery promotes postoperative pulmonary complications; these reasons originate in the preoperative, intraoperative, and postoperative period (Fig. 6l ). First, the incidence of postoperative respiratory complications following any surgery is positively correlated with the degree of preoperative respiratory dysfunction, and most thoracic surgical patients come to surgery with some degree of preoperative lung dysfunction. Preoperative pulmonary function testing will identify the patients at high risk due to poor preoperative lung function. Second, the performance of thoracic surgery can impair lung function in any patient. During surgery, nondependent-lung function may be impaired by resection of functional lung and/or by trauma to the remaining nondependent lung (as a
result of various nondependent-lung manipulations) and/or by overexpansion of the remaining nondependent lung, and dependent-lung function may be impaired as a result of the development of atelectasis and edema formation. Third, thoracotomy incisions are painful and cause patients to resist deep breathing and coughing in the postoperative period, leading to retained secretions, atelectasis, and pneumonia. The second factor (impaired lung function due to the performance of thoracic surgery) can be minimized by appropriate intraoperative management (such as one-lung ventilation, positive end-expiratory pressure, continuous positive airway pressure) (see chapter 8). The third factor can be minimized by appropriate postoperative pain management (e.g., epidural narcotics) (see chapter 20). The impact of the first factor (presence of preoperative respiratory dysfunction) can be significantly reduced by preoperative prophylactic respiratory preparation measures. This chapter first documents the correlation between degree of preexisting respiratory disease and incidence of postoperative respiratory complications. Next, the correlation between thoracic surgery and the increased incidence of postoperative respiratory complications is discussed. Proof is then offered that preoperative respiratory preparation decreases the incidence of postoperative respiratory complications, and the main body of the chapter details a full preoperative respiratory prep-
Thoracic Surgery Impairs Postoperative Lung Function
Patients Who Have Thoracic Surgery Usually Have Pre-existing Lung Disease
Figure 6-1 function.
Nondependent Lung: Functional Tissue Resected or Traumatized
Painful Incision
Dependent Lung: Compressed, Edematous
Fail to Deep Breathe and Cough
There are preoperative, intraoperative, and postoperative reasons why thoracic surgery impairs postoperative lung
Preoperative Respiratory Preparation
aration regimen. The mechanism by which preoperative respiratory preparation decreases the incidence of postoperative respiratory complications is considered. Finally, premedication, as part of a preoperative respiratory preparation plan, is discussed.
II. CORRELATION OF RESPIRATORY COMPLICATIONS WITH DEGREE OF PRE-EXISTING LUNG DISEASE The relationship between lack of preoperative pulmonary reserve and postoperative respiratory morbidity and mortality is very well recognized and may be very dramatic. Compared with nonsmokers, smokers have decreased pulse oximetry values, increased post anesthesia care unit (PACU) stay, and a sixfold increase in the incidence of postoperative pulmonary complications after major operative procedures.4-8 Three major mechanisms appear to be responsible for the adverse smoking/surgery interaction, and they are small airway disease (spasm, collapse), hypersecretion of mucus, and impairment of tracheobronchial tree clearance. In patients with chronic lung disease, compared with normal healthy patients, there is a 20-fold increase in the incidence of postoperative pulmonary complications.9 Thus, it is not surprising to find widespread agreement that the risk of postoperative pulmonary complications progressively increases as preoperative pulmonary function progressively decreases. Preoperative pulmonary function is best quantitated by preoperative pulmonary function tests, and patients with a vital capacity, maximum breathing capacity, FEV, or FEF257f_75C/i of less than 50 per cent of predicted capacity and/or grossly hypercapnic patients are at very high risk (see chapter 5 references and Handlin and Baker7 for extensive substantiation of this contention). However, even if pulmonary function tests are not available, it should be remembered that gross observations such as the production of a great deal of sputum by the patient (more than 2 ounces per day),10 minimal exercise tolerance capability, severe cardiac disease, obesity, sepsis, and very advanced age also indicate great risk for developing postoperative pulmonary complications." 12
III. CORRELATION OF RESPIRATORY COMPLICATIONS WITH SITE OF OPERATION The correlation between postoperative pulmonary complications in patients with and without respiratory disease and the site and type of opera-
213
tion has long been known, with the highest incidence of complications following major thoracic and upper abdominal procedures.10-14 In a series of 1500 surgical patients with a wide variety of respiratory diseases treated over a 30-year period, the incidence of respiratory complications averaged 63 per cent following thoracic and gastric operations, 15 to 19 per cent following midabdominal operations, and 9 per cent following lower abdominal procedures.15 In a group of 464 patients with chronic respiratory disease who did not have any preoperative respiratory preparation, the highest risk of pulmonary complications involved patients with thoracotomy or abdominal operations (compared with surgery on other parts of the body).16 Even when the patient population was specifically defined as those with chronic obstructive pulmonary disease (COPD) who underwent preoperative respiratory preparation, the incidence of respiratory complications was still highest for thoracic and abdominal operations compared with surgery on other parts of the body.10 Similarly, other patients with severe chronic obstructive pulmonary disease who had operations on the thorax or upper abdomen were found to have twice the mortality compared with those with a similar amount of respiratory impairment and who were subjected to operations on other body regions.17 The mechanisms by which thoracic surgery (and abdominal surgery) especially predisposes patients to postoperative respiratory complications include resection of, or trauma to, functional lung and the degree to which the bellows function of the lung is affected (Fig. 6-1). It is painful for patients who have incisions in these areas to deep breathe and to stretch either the chest or abdominal wall (i.e.. the incision); consequently, they fail to cough (which requires a deep breath), and they retain secretions (which promotes the development of atelectasis and infection). When the lateral thoracotomy incision is limited (preservation of the latissimus dorsi, splitting of the serratus anterior, and cutting of only the intercostal muscles without rib resection) and the bellows function is therefore better preserved, postoperative pulmonary reserve may be increased.18 In addition, it has been claimed (but not proven) that operative time, blood loss and postoperative pain, intensive care unit stay, and morbidity are decreased.19
IV. PROOF THAT PREOPERATIVE PULMONARY PREPARATION DECREASES INCIDENCE OF POSTOPERATIVE RESPIRATORY COMPLICATIONS Considerable evidence has accumulated over the last three decades to demonstrate that vigorous
214
Preoperative Respiratory Preparation
preoperative pulmonary preparation can significantly reduce postoperative pulmonary complications.20 It is useful to monitor pulmonary function during the preoperative "tune-up" because those who improve their pulmonary function have a better prognosis than those who do not.10 Several decades ago, it was demonstrated that simple preoperative physical therapy instruction, as opposed to the same instruction given postoperatively or not at all, decreased the incidence of atelectasis from 42 per cent (no instruction), to 27 per cent (postoperative instruction), to 12 per cent (preoperative instruction).21 About the same time, it was demonstrated that the use of nebulized isoproterenol three times daily before and after surgery, in conjunction with postural drainage and chest physiotherapy, reduced the incidence of postoperative atelectasis from 43 to 9 per cent.22 In a much more recent study, normal patients were prospectively randomized into a supervised preoperative breathing exercise treatment group (diaphragmatic breathing, deep breathing, forced expiration, coughing, and dead-space rebreathing using a tube) and a nontreatment group. The incidence of postoperative (upper abdominal surgery) complications (criteria derived from chest roentgenograms, arterial blood gas samples, and temperature registration) in the treatment group and in the control group were 19 per cent and 60 per cent, respectively.23 Patients with chronic obstructive pulmonary disease are the ones most likely to benefit from preoperative respiratory preparation. In 1959 it was impressively shown in a group of 250 surgical patients with obstructive pulmonary disease that when an intensive pulmonary preparation was utilized, consisting of oral therapy with a bronchodilator drug, intermittent positive-pressure ventilation with nebulized bronchodilator therapy, use of an expectorant, postural drainage, cough training, and antibiotic treatment for purulent sputum, only two patients (1 per cent) developed atelectasis.24 In 1970 randomly selected "poor-risk" patients were treated preoperatively and postoperatively with a very comprehensive respiratory care regimen (bronchodilators, antibiotics, inhalation of humidified gas, segmental postural drainage, and chest physical therapy). When the treated patients were compared with the nontreated patients, the treated patients had a pulmonary complication rate of only 22 per cent (all mild), whereas the untreated patients had a pulmonary complication rate of 60 per :ent (of which 60 per cent were severe).25 Similarly, in a retrospective study of patients with ;hronic obstructive pulmonary disease who underwent various surgical procedures under inhalation inesthesia, the incidence of postoperative respiraory complications was only 24 per cent in those >iven preoperative pulmonary preparation com-
pared with 43 per cent in a control, nonprepared group.16 Since preoperative pulmonary preparation can significantly reduce postoperative pulmonary complications, especially in patients with lung disease, it is not surprising to find that a high-risk group of patients with moderate to severe respiratory dysfunction who required thoracic surgery also greatly benefited from preoperative pulmonary preparation.26 This high-risk group was managed with vigorous prophylactic measures and suffered less pulmonary complications than did a more normal respiratory function group that had not been exposed to the same preoperative regimen. Similarly, in another study, the differences in pulmonary complication rate were insignificant between untreated young "good-risk" and treated "poorrisk" patients.25 Thus, the frequency and severity of pulmonary complications may not increase in going from nontreated young "good-risk" patients to treated "poor-risk" patients (implying that the preoperative treatments corrected, at least in part, the preoperative respiratory dysfunction), but the respiratory complication rate definitely increases in going from treated "poor-risk" patients to untreated "poor-risk" patients.16·24"26 More recently (1979), a very careful, meticulous prospective report described the effect of a standardized preoperative pulmonary regimen on preoperative pulmonary function tests in a series of 157 patients with chronic obstructive pulmonary disease.10 The pulmonary regimen consisted of 48 to 72 hours of aerosolized isoproterenol in saline given four times a day, oral therapy with theophylline, guaiacol glyceryl ether, hydration, and chest physiotherapy. In addition, most of the patients discontinued smoking. The preoperative respiratory preparation regimen improved preoperative pulmonary function tests but not in a predictable or consistent way. The rate of complications was highest in those patients who had upper or long abdominal incisions or thoracotomy, although the incidence of respiratory complications was significantly reduced as a result of the preoperative preparation. The authors were unable to determine which patients would develop significant pulmonary complications not requiring mechanical ventilation, but those requiring respiratory support were predictable on the basis of the severity of their pulmonary functional impairment and minimal response to the pulmonary preparation used (i.e., those who had irreversible disease).
V. PREOPERATIVE RESPIRATORY PREPARATION MANEUVERS The preceding data indicate that patients undergoing thoracic surgery are particularly sus-
Preoperative Respiratory Preparation
ceptible to postoperative respiratory complications 4 · 5 · 9-17 and that prophylactic measures do decrease postoperative respiratory complications.10·I7, 21, 22. 24-26 Consequently, preoperative evaluation should be followed by preoperative preparation efforts directed toward optimally managing any preexisting pulmonary disease.27 In general, a full preoperative respiratory preparation regimen involves a five-pronged attack on airway disease. The five elements of the preoperative regimen are stopping smoking, dilating airways, loosening and removing secretions, and taking measures to increase motivation and education and to facilitate postoperative care (Table 6-1 and Fig. 6-2). The five treatment modalities are instituted and proceed in parallel fashion. Before discussing each element separately, it is important to point out that the desirable results of these maneuvers should be, for the purpose of understanding their interaction, achieved in a sequential manner (Fig. 6-2). The logic behind this concept is as follows. First, stopping smoking eliminates the stimulation for the production of airway secretions and bronchoconstriction. Next, the airways should be dilated to facilitate secretion removal. Similarly, thick, tenacious, and adherent secretions must be loosened in order to be removed. Once the airways are dilated and the secretions loosened, it makes sense to use physical maneuvers to remove the secretions. Finally, the patients should assist, as much as possible, in their preoperative preparation and postoperative respiratory care. The studies cited later indicate that using the maneuvers in this sequence (dilating the airways, loosening the secretions, removing the secretions) allows the maneuvers to complement one another in improving mucociliary
Table 6-1
215
PREOPERATIVE RESPIRATORY CARE REGIMEN
1. Stop smoking. 2. Dilate the airways. a. 32-agonists b. Theophylline c. Steroids d. Cromolyn sodium 3. Loosen the secretions. a. Airway hydration (humidifier/nebulizer) b. Systemic hydration c. ? Mucolytic and expectorant drugs d. Antibiotics 4. Remove the secretions. a. Postural drainage b. Coughing c. Chest physiotherapy (percussion and vibration) 5. Perform ancillary measures. a. General measures (treatment of systemic disease) b. Treatment of cor pulmonale c. Preoperative digitalization 6. Provide increased education, motivation, and facilitation of postoperative care. a. Psychologic preparation b. Incentive spirometry c. Exposure to secretion removal maneuvers d. Exercise e. Weight loss/gain f. Stabilization of other medical problems
transport function. The following section discusses each of these five preoperative preparation maneuvers in the order and context of this conceptual approach. Of course, it is recognized that patients who do not have bronchospasm will not benefit from bronchodilator treatment and that patients who do not have secretions will not benefit from measures to enhance secretion removal.28
Preoperative Respiratory Preparation Regimen
1. Terminate Stimulus for Bronchoconstriction and Secretions Stop Smoking Figure 6-2 A full, aggressive, preoperative respiratory preparation regimen consists of a five-pronged attack: (1) Require the patient to stop smoking, (2) dilate the airways, (3) loosen secretions, (4) remove secretions, and (5) increase patient participation. Using these five maneuvers in the numbered sequence allows them to complement one another in improving secretion removal.
216
Preoperative Respiratory Preparation
A. Discontinue Smoking An improvement in mucociliary transport and small-airway function and a decrease in airway secretions, reactivity, and markers of injury occur over several weeks after cessation of smoking. 29-32 Indeed, in addition to the usual clinical syndrome of chronic bronchitis (chronic cough and phlegm), smoking may cause wheezing by increasing the level of nonspecific airways responsiveness32 " and/or by narrowing of airways secondary to in flammation and structural changes.34 Numerous past studies have clearly shown that a period of at least 6 to 8 weeks of abstinence is necessary to effect substantial improvement in smoking-induced small-airways disease, 3 5 3 6 hypersecretion of mucus," reduced closing volume,38 and impaired tracheobronchial clearance.39 Consequently, it is not surprising to find that preoperative cessation of smoking for more than 4 to 8 weeks is associated with a decrease in the incidence of postoperative respiratory complications.29· 30 Thus, despite the fact that as few as 20 per cent of smokers will quit when advised to do so by their physician well in advance of an elective operation,40 it is a worth while effort. Although stopping smoking for only 24 hours will do nothing to actually decrease the amount of secretions (at least l to 2 weeks are required), 30 · 41 airway irritability, and incidence of postoperative respiratory complications, there are still a number of other important, perhaps intraoperative, benefits that accrue in the first 1 to 2 days of abstinence.29· 30 4I The most important benefit results from a decrease in carboxyhemoglobin levels. Carbon monoxide avidly combines with hemoglobin, even at low tensions of carbon monoxide, because of the extraordinary chemical affinity of carbon mon oxide for the iron atoms in hemoglobin. The affin ity of hemoglobin for carbon monoxide is more than 200 times that for oxygen. Carbon monoxide readily displaces oxygen from hemoglobin and re duces the volume of oxygen carried by the hemo globin in the circulation. In addition, the presence of carboxyhemoglobin in the blood causes a left ward shift in the oxygen-hemoglobin dissociation curve. Thus, oxygen dissociates from hemoglobin at a lower tension, which, in turn, results in a lower driving pressure and reduced availability of oxygen in the tissues and metabolizing cells. 42 · 43 Cessation of smoking for as short a time as 1242 to 48 44 hours has been shown to decrease carboxy hemoglobin levels (e.g., after 9 and 12 hours of abstinence from smoking one to two packs per day, carboxyhemoglobin concentrations decrease from 6 and 7 per cent to 1 and 4 per cent, respectively),42·45 (Fig. 6-3). The decrease in car boxyhemoglobin increases oxyhemoglobin and
Figure 6-3 Mean ( ± standard error of the mean) blood nicotine and carboxyhemoglobin (COHb) concentrations in cigarette smokers. Subjects smoked cigarettes every half hour from 8:30 A.M. to 11:00 p.M for a total of 30 cigarettes per day. (Adapted from Benowitz NL: Pharmacologic aspects of cigarette smoking and nicotine addiction. Ν Engl J Med 319:1318-1330, 1980. Used with the permission of the pub lisher.)
right shifts the oxygen hemoglobin curve (in creases availability of oxygen tissues).42 4 4 4 5 Numerous studies show that the smoking-induced increase in carboxyhemoglobin may signif icantly impair cardiovascular performance in pa tients with coronary artery disease. In a doubleblind crossover study involving 30 patients with ischemic heart disease who had carboxyhemoglo bin levels of 6 per cent, there was an early onset of angina during exercise and abnormal left ven tricular function as detected by radionuclide ven triculography.46 Similarly, and perhaps of even more concern, in patients with coronary artery dis ease, cigarette smoking can produce an absolute decrease in regional myocardial blood flow, usu ally without angina pain, comparable to exerciseinduced ischemia.47 Indeed, there is a 12-fold Hoi-
Preoperative Respiratory Preparation
ter monitor documented increase in the duration of ischemic episodes in smokers (nearly all of which were silent), even though the clinical and angio graphic characteristics of the smokers and nonsmokers were similar.48 With respect to exposure to environmental car bon monoxide, it has been observed that there is a reduction in exercise time to angina along with a decrease in the product of systolic blood pressure and heart rate at the time of the onset of angina in 10 patients with coronary artery disease who were driven on a freeway in Los Angeles, in whom carboxyhemoglobin levels reached 4.0 per cent.49 Even low levels of carboxyhemoglobin (2 to 3 per cent) exacerbate myocardial ischemia (as reflected by shortening of the length of time to angina pain and ST-segment depression) during graded exer cise in subjects with coronary artery disease.50 The levels of carboxyhemoglobin in this study (2 to 3 per cent) occur in smokers of one half to one pack of cigarettes per day, passive smokers in an unventilated room, firefighters actively fighting a fire, and vehicular tunnel works.51 At low doses of carboxyhemoglobin, similar to those seen during cigarette smoking, the cardio vascular effects appear to be mediated by the cen tral nervous system, either through the activation of chemoreceptor afferent pathways or by direct effects on the brain stem. The net result is sympa thetic neural discharge, with an increase in blood pressure and heart rate. 45 Smoking may also have an arrhythmogenic effect, perhaps related to the increase in sympathoadrenal activation.52 In a manner consistent with a half-life of 2 hours, nic otine accumulates over 6 to 8 hours of regular smoking and persists overnight, even as the smoker sleeps (see Fig. 6-3). 45 Nevertheless, absti nence from smoking for 9 hours results in mark edly lower level of nicotine in the blood (see Fig. 6-3) and can be expected to decrease nicotine30 4I induced tachycardia and hypertension · and lower the risk of arrhythmias. Thus, the acute ef fects of decreasing both carboxyhemoglobin and nicotine levels from stopping smoking for several hours may confer a critical benefit to the marginal patient. There a few select patients in whom the risks of stopping smoking for a day or two may be thought to outweigh the benefits. First, although stopping smoking will cause some anxiety in many patients, cessation of smoking may induce a great deal of anxiety in some patients (due in part to acute nic 45 53 otine withdrawal in nicotine-addicted patients). · In patients with significant coronary artery disease the large increase in anxiety preoperatively may lead to a critical ischemic event. Second, sudden cessation in smoking can occasionally induce a hypersecretory and bronchospastic state. In pa
217
tients who have difficulty with secretions, a pre operative increase in secretions may lead to air ways that are more obstructed preoperatively.29 Third, there are some experimental data to indicate that continued smoking may result in a decreased incidence of postoperative deep vein thrombo sis.30, 4I Nevertheless, a good argument can be made that these potential benefits of continuing smoking during the day or two prior to surgery can just as easily be accomplished with anxioly tics, bronchodilators, and anticoagulants. Thus, the potential benefits do not provide sufficient reason to continue smoking.
B. Dilating The Airways The next step should be to dilate the airways. Patients who have increased airway responsive ness and therefore are candidates for preoperative bronchodilation are smokers,54 atopic individuals,54 patients with airway symptoms of allergies,54 pa tients with C O P D 5 5 5 8 and asthmatics.59 60 Perhaps the least obvious candidates for preoperative bron chodilation among these groups of patients are those with COPD. When FEV, is measured before and after the one-time administration of an inhaled bronchodilator (three puffs of metaproterenol metered-dose inhaler [MDI]), in order to assess the reversibility of chronic airway obstruction in nonasthmatic COPD patients, 38 per cent had a signif icant response to the bronchodilator (15 per cent increase in FEV,). 55 In another group of stable COPD patients taking oral prednisolone (30 mg every day), the one-time administration of salbutamol, 200 μg, from an MDI or 5.0 mg from a nebulizer caused at least a 15 per cent increase in FEV, in 56 per cent of patients.56 The most important bronchodilator drugs (β : agonist, steroids, and anticholinergics) are admin istered by inhalation of an aerosol. Therapeutic aerosol can be generated by several means: MDIs with fluorocarbon propellant, jet nebulizers pow ered by compressed air oxygen or an electric pump, jet nebulizers coupled with an intermittent positive-pressure breathing (IPPB) device, aerosol generated by an ultrasonic nebulizer, or inhaled fine powder. When correctly administered, all these forms of delivery are probably equally effec tive. Because MDIs are small, compact, and rela tively cheap and they administer a small dose of medication with each puff, they are the ideal form of therapy. Therefore, β-agonists are most widely administered from an MDI. Optimal operation of the inhalation aids requires actuation of the MDI followed by a slow, deep inhalation followed by breath holding for several seconds.
218
Preoperative Respiratory Preparation
MDIs have two major disadvantages. First, the MDI requires the patient to learn the inhalation technique. Numerous publications attest to the fact that many patients with asthma have incorrect in halation technique. Even after repeated instruc tions, 10 to 20 per cent, particularly the very young and the elderly, cannot learn the technique. The most frequent problem is inability to coordi nate the actuation of the canister and inhalation by mouth. The more fundamental problem is that the patient is unable to control inhalation: when to start inhalation and how fast to inhale through the open mouth. Second, a sufficient amount of medi cation will not reach the airways when breathing is rapid and shallow. Efficiency of penetration into the lung can be improved by interposing spacers and expansion chambers between the device and the mouth (Fig. 6-4). Large-volume spacer cham bers, which contain the aerosol cloud delivered by the MDI, are also useful in patients who cannot coordinate the ordinary MDI, because they can inhale the drug via the one-way valve after trigger ing the aerosol. Spacers ensure some increase in delivery to peripheral lung and reduce oropharyn geal deposition. This is very important in highdose inhaled steroids (when spacers should always be used to reduce oropharyngeal complications [candidiasis]). Nebulized bronchodilator solutions are valuable for treating acute episodes of severe obstruction when breathing is rapid and shallow and for pa tients who cannot master the coordination needed to use MDIs.
1. Overall Bronchodilating Plan Before presenting the individual types of bron chodilator drugs, it is important to consider the order in which the drugs should be administered (i.e., what is the pharmacologic bronchodilating plan?). At present, there are two basic different approaches to the use of bronchodilators in asthma and COPD patients, and the essential difference between the two approaches is whether the inhaled β-,-agonists and/or anticholinergics are considered the front-line drugs, with inhaled steroids consid ered the second-line drugs or vice versa (Table 62). In the first approach, espoused by Bone,61 phase I consists of inhaled bronchodilators (β 2 agonists for asthmatics, anticholinergics for COPD patients [step 1, Table 6-2] and as adjunctive ther apy in both conditions [step 2, Table 6-2]).61 Phase II for asthmatics consists of prophylaxis and/or treatment of inflammation with agents that have few side effects (inhaled steroids, cromolyn) (step 3, Table 6-2). Theophylline can be used if asthma is not adequately controlled with inhaled β,-agonists and either cromolyn and/or inhaled steroids (step 4, Table 6-2). Phase III for asthmat ics is treatment of inflammation with agents that have greater toxicity (oral or intravenous steroids) (step 5, Table 6-2). For COPD patients, theophyl line and steroids follow the acute bronchodilators (steps 1-5, Table 6-2). In the second approach, espoused by Lam and Newhouse62 and Newhouse,63 there is an increased appreciation of the importance of inflammation in the pathogenesis of
Figure 6-4 spacer.
Child using metered aerosol inhaler with
Preoperative Respiratory Preparation
219
Table 6-2 TWO BASIC DIFFERENT APPROACHES TO THE USE OF BRONCHODILATORS IN ASTHMA AND COPD Approach to Bronchodilation in Asthma and COPD Bone'
asthma, and inhaled bronchodilators have been rel egated to second-line therapy for controlling asthma by the increasingly potent inhaled steroids and cromolyn, both of which have an excellent therapeutic ratio. Thus, in this approach, first-line therapy for moderate chronic asthma should be dose-optimized inhaled steroids and possibly in haled cromolyn sodium (cromoglycate sodium), especially in children (step 1, Table 6-2). Secondline therapy is P2-agonists (step 2, Table 6-2). Theophylline is considered rarely of additional value except, perhaps, as a steroid-sparing strategy in the maintenance therapy of patients with asthma requiring systemic steroids for control despite maximum tolerated doses of inhaled steroid, ad renoceptor agonists, and (rarely) anticholinergic agents (step 4, Table 6-2). For COPD patients, after the reversal of priority for the first- and second-line drugs (steps 1 and 2, Table 6-2), the approach is similar (steps 3-5, Table 6-2). All the drugs (and their dosages) that are contained within all approaches to the preven tion and treatment of bronchospasm in patients with asthma and COPD are listed in Table 6-3.
2.
fi2-Agonists
β-Receptors are divided into β, and β2 sub types; β,-receptor agonists subserve chronotropic and inotropic cardiac effects, and β2-Γεΰερΐ0Γ8 re lax the airway and vascular smooth muscle. At the molecular level, sympathomimetics (so-called first messengers) cause β-receptor activation, which leads to the activation of adenylate cyclase, which converts adenosine triphosphate to cyclic adeno sine monophosphate (cAMP), which acts as a se cond messenger for the production of protein ki
6263
Lam, Newhouse
nases, leading to smooth-muscle relaxation (Fig. 6-5). 64 When used by inhalation, β 2 ^οηΐ8ί8 cause a peak bronchodilator effect within 15 min, which is maintained for approximately 4 to 6 hours, de pending on the dose inhaled. β 2 ^οηΪ8ί8 may be absorbed from the bronchial mucosa and distrib uted to more peripheral airways. Only about 10 per cent of inhaled drug reaches the lung; the rest is swallowed. Substantial bronchodilation can be achieved without detectable circulating levels of β 2 ^οηΐ5ί5. Therefore, side effects common to in travenous administration and overdoses, such as muscle tremor, tachycardia, and hypokalemia, are rarely observed with inhaled therapy. In addition, adrenergic compounds (epinephrine, ephedrine, isopropyl, norepinephrine) can also increase ciliary activity, which may help in removing secretions. 65 · 66 β 2 ^π^ίηοιτπΓηείΐΰ drugs, such as albuterol, terbutaline, and metaproterenol, are administered to patients who have a demonstrable reversible bronchospastic airway component to their respira tory disease. 61 - 67 · 68 Their use in asthma is obvious. Studies in stable outpatients with COPD demon strate that, with adequate dosing and delivery, an inhaled bronchodilator (either a β 2 ^ ο η ΐ 5 ΐ or an antimuscarinic) can result in complete57 5* or partial 55 · 56 bronchodilation (as measured by FEV,). In these studies, addition of a second agent did not result in further spirometric improvement. In ad dition, if a therapeutic blood level of a p2-agonist is achieved, then the ventricular performance of COPD patients may also be improved because of a decrease in afterload (particularly on the right ventricle, which may be pumping against chronic pulmonary hypertension) and by a positive ino tropic effect.69
ro Ν) ο
Preoperative Respiratory· Preparation
221
Figure 6-5 Occupation of the B-receptor by agonist activates adenylate cyclase (AC) via coupling protein, guanine nucleotide regulatory protein (N). Cyclic adenosine monophosphate (AMP) is broken down by phosphodiesterase (PDE). (ATP = adenosine triphosphate.) From Chung KF, Barnes PJ: Drugs for respiratory disease. Medicine International 2469-2473, 1988. Used with permission.)
3. Anticholinergic Drugs The main anticholinergic agent of clinical use as a bronchodilator is ipratropium bromide, a qua ternary ammonium derivative of atropine (N-isopropylnoratropine). Ipratropium bromide can be prescribed only as an aerosol, either from an MDI inhaler or from a nebulizer. It is also available combined with a p2-agonist. Ipratropium bromide is a competitive antagonist of acetylcholine at the parasympathetic muscarinic receptor. It causes bronchodilatation by relieving the increase in bronchomotor tone as a result of tonic vagal nerve impulses releasing acetylcholine at cholinergic nerve terminals in airways. Muscarinic receptors are more densely distributed in the larger, proxi mal airways, consistent with physiologic studies that demonstrate a greater effect on larger airways than on small airways. Compared to β 2 ^οηΪ8ί5, ipratropium bromide is slower in achieving a max imal bronchodilator effect, with approximately 75 per cent of the effect occurring at 15 min and the complete effect by 1 hour. The duration of signif icant bronchodilation is about 6 to 8 hours. Maxi mal responses are usually achieved with 80 to 120 μg of ipratropium bromide. At the recommended doses, ipratropium bromide aerosol causes rela tively few adverse reactions. Anticholinergic side effects, such as blurred vision, dry mouth, hesi
tancy of micturition, are uncommon (< 1 per cent), because of its poor absorption. Most asthmatics respond better to a p2-agonist than to ipratropium bromide, in contrast to patients with chronic bronchitis and COPD in whom ipra tropium may be/should be used as an alternative to a β-agonist as first-line treatment. 6 1 · 7 0 Ipratro pium bromide works as well in COPD patients because they have an increased amount of cholin ergic tone compared with normal patients.70 How ever, in asthma, ipratropium may be tried after failure to respond to a β-agonist,60·71 and ipratro pium may also be useful as additional therapy in patients with chronic bronchitis or asthma who are already on therapeutic doses of β 2 ^οηΐ5ΐ. 6 0 7: Ipratropium (10 puffs or 200 μg nebulized) may also be useful in patients with nonallergic bron 73 chial asthma and in COPD patients being me chanically ventilated after cardiac operations.74
4. Inhaled Steroids Inflammation of the airways is the main patho physiologic process in asthma, and it can be found even in patients with newly diagnosed, mild, or 75 76 asymptomatic asthma. · Thus, the disease is now often described as chronic desquamating eosino philic bronchitis.77 Airway hyperresponsiveness in asthma is now viewed as a secondary consequence
222
Preoperative
Respiratory
Preparation
of the inflammation that narrows the lumen ana tomically and increases the responsiveness of neu romuscular control of the airways through the ac tion of the mediators of inflammation. Inhaled corticosteroids therapy (beclomethasone dipropionate, flunisolide acetate, budesonide, and tri amcinolone acetate) is the best anti-inflammatory treatment available for asthma.62 It decreases airway inflammation and nonspecific bronchial hyperresponsiveness, potentiates response to sympa thomimetics and aminophylline, improves pulmo nary function, and decreases the frequency and severity of asthma symptoms. An expert panel, convened by the National Institutes of Health and a number of interested professional societies, has recommended that the care of patients with daily (or nightly) asthma include one of the anti-inflam matory aerosols—cromolyn or a glucocorticoid— along with bronchodilators for symptomatic re lief.7" A large fraction of the dose delivered by an MDI held in the mouth—up to 90 per cent—is deposited in the pharynx and larynx, where ste roids can cause dysphonia and predispose the pa tient to oropharyngeal candidiasis. After it is swal lowed, the fraction of drug deposited in the pharynx also contributes to systemic steroidal ac tion. Inhaling slowly through a spacer device re duces deposition in the mouth and pharynx and reduces both local and systemic adverse effects. Training the patient in aerosol-inhalation tech nique with a spacer adds to the complexity of prescribing steroid aerosols, but it is essential for success. Recent studies showed that the effectiveness of these medications increases if they are used with an aerochamber (spacer). This device also de creases the incidence of oropharyngeal candidi asis.7V Inhaled corticosteroids have been underused because patients who are accustomed to rapid re lief from an inhaled bronchodilator find that they experienced no immediate relief from inhaled cor ticosteroids. In addition, the suppressant effect of inhaled steroids (budesonide) on airway inflam mation, edema, and reactivity is dissipated within 2 weeks of dosing if it is not repeated.80 The lack of dramatic onset and dissipation of effect may constitute an apparent lack of response to the pa tient and lead to noncompliance. Thus, patient ed ucation regarding therapeutic expectations for these drugs is of paramount importance.
5. Inhaled Sodium Cromoglycate Sodium cromoglycate is a synthetic derivative of the herbal plant, khellin, and is used for the prophylaxis and prevention of asthma attacks. An other closely related compound, nedocromil so
dium, has recently become available. Sodium cromoglycate is highly ionized and water soluble and is therefore poorly absorbed when taken orally. For the treatment of asthma, it is inhaled either as a fine powder from a spinhaler or as an aerosol from an MDI. A nebulizer solution is also available. Prevention of mast cell degranulation by aller gens was thought to be the main mechanism of action of sodium cromoglycate in the prophylaxis of asthma. However, other compounds with greater potency as mast cell stabilizers have little value in the prophylaxis of asthma. Other possible modes of action of cromoglycate include an inhib itory effect of eosinophil activation to release me diators and inhibition of neurogenic reflex bronchoconstriction perhaps by an action on sensory nerves. It is not a bronchodilator. Only a small proportion of inhaled sodium cromoglycate can be detected in the circulation, the amount depending on inspiratory flow rate. Peak plasma levels are achieved within 15 to 20 min of inhalation, and sodium cromoglycate is ex creted unchanged in bile and urine. The time course of action of sodium cromoglycate is diffi cult to assess, but some studies suggest that its prophylactic effect may take a few weeks to be come optimum. Sodium cromoglycate is remarka bly devoid of side effects. Patients may complain of irritation of the throat and of cough when in haled as a dry powder, probably because of the irritant effect of the powder. Sodium cromoglycate improves asthmatic symptoms and leads to a re duction in the use of bronchodilators in asthmatic patients. However, in general, sodium cromogly cate is less effective as a prophylactic treatment than inhaled steroids. Clinically, it is only effective in a proportion of asthmatics, particularly in the younger, atopic group. Sodium cromoglycate should be considered in asthmatics whose symp toms are not well controlled by a β-agonist inhaler alone, particularly in children. If cromoglycate is not effective, then inhaled steroids should be used. However, in adults, there is a preference for in haled steroids as first-line prophylaxis. Sodium cromoglycate should be used regularly over a pe riod of 2 to 3 weeks before assessing the response. Improvement may be dose dependent. Sodium cromoglycate should not be used for treatment of acute asthma and in patients with COPD.
6. Methylxanthines Theophylline is a methylxanthine found in var ious plants, including tea, and is closely related to caffeine and theobromine. Theophylline can be taken orally or by suppository. Aminophylline is a theophylline salt, is more soluble than theophyl-
Preoperative Respiratory· Preparation
line, and is therefore used intravenously. Both methylxanthines are difficult drugs to use because of the large variation in metabolism between indi viduals and because of the narrow therapeutic ra tio. c'AMP is broken down by a cytoplasmic en zyme, phosphodiesterase, whose activity can be inhibited by methylxanthines, such as theophylline and aminophylline. Thus, the methylxanthines also increase c'AMP but by a mechanism different from that of the p2-agonists (see Fig. 6-5). Be cause the methylxanthines and β 2 ^οηΐ8ί8 act by different mechanisms, theophylline is often added to the regimen of patients with bronchospasm al ready receiving beta-adrenergics, and thus they work in synergy to increase intracellular concen trations of c'AMP. 8 1 , 8 2 In addition, aminophylline improves diaphragmatic contractility and renders it less susceptible to fatigue.83 In fact, in COPD patients with severe "fixed" obstruction (FEV, < 30 per cent and unresponsive to β 2 ^ ο η ΐ 8 θ , the ophylline increases respiratory function and de creases dyspnea as a result of increased strength of the respiratory muscles. 84 · 85 Thus, it is no surprise that the methylxanthines cause subjective im provement in patients with chronic airflow ob struction.86 Recently, however, there has been considerable concern regarding the short-term toxicity of in haled beta-adrenergic agents when used in the presence of methylxanthines; specifically, myocar dial ischemia might occur as a result of the drugs' combined effect on the heart with resultant (and possibly fatal) ventricular arrhythmias.87 88 In one study of patients in status asthmaticus, with a mean age of 39 ± 6 years, 17 per cent of the patients exhibited severe ventricular and atrial ar rhythmias during combined therapy with amino phylline and the p2-agonist terbutaline.89 Some studies have not supported exercising great con cern with respect to arrhythmias with concurrent use of aminophylline and β 2 ^οηΪ8ί, 9 0 · 9| but it should be remembered that the dose required to produce therapeutic/toxic levels can be quite vari able because of variation in hepatic metabolism. Thus, appropriate caution should be exercised when combining inhaled β 2 ^οηΐ8ί8 with meth 88 92 ylxanthines. Optimal therapeutic serum levels of theophylline (10 to 20 mg/ml) can be safely approached and toxicity avoided if an intravenous loading dose (5 to 7 mg/kg) is given, followed by continuous infusion (0.2 to 0.8 mg/kg/hr: dose de 93 creases with increasing age). Although the pa tient's subjective feeling of relief is an important end point (see chapter 5), the effect of bronchodilator drug treatment should be quantitated by pul monary function tests.
223
7. Oral and Parenteral Steroids Oral or intravenous corticosteroids should be used as a last-line measure. Because corticoste roids are rapidly and almost completely absorbed from the gastrointestinal tract, and both routes of administration have somewhat of a delayed effect (but still less than a day),94 it is not surprising that oral corticosteroids are nearly as effective as intra venous corticosteroids for status asthmaticus.61 Corticosteroids should be used in patients whose attacks are not controlled with the combi nation of bronchodilators and inhaled anti-inflam matory medications. Oral dosage (for outpatients) should be as high as 0.75 to 1.0 mg/kg/day, and intravenous dosages (for inpatients) of corticoste roids should be in the range of 0.5 mg/kg every 6 hours, for 1 to 2 weeks, with daily monitoring of peak flow rates (if possible), and the oral or intra venous dosage should be tapered over 1 to 2 weeks. If the patient is not taking inhaled steroids or cromolyn, one of these should be added as the systemic corticosteroid dosage is tapered. If oral corticosteroids cannot be replaced by the inhaled anti-inflammatory agents, alternate-day steroids in conjunction with the anti-inflammatory agents should be considered. The anti-inflammatory agents constitute the ideal regimen in the patient with severe asthma, with oral corticosteroids re served for severe exacerbation. The oral cortico steroid dosage is then tapered as soon as control is achieved.
C. Loosening the Secretions The next step should be to thin and to loosen thick adherent secretions. The most efficacious method is hydration. When tracheal mucus trans port velocity is quantitatively measured by radio active tracer methods, it can clearly be shown that dehydration decreases and rehydration increases, respectively, tracheal transport velocity.95 The most common method of hydrating secretions is by use of a jet humidifier or ultrasonic nebulizer to produce a heated, sterile water aerosol that is delivered by a close-fitting mask for 20 min to a deeply spontaneously breathing patient. Concur rently, continuous systemic hydration must be en sured orally or intravenously. Very occasionally, administration of mucolytic agents (such as acetylcysteine [Mucomyst]) by a nebulizer and/or oral expectorants (such as guai fenesin, potassium iodide, iodinated glycerol) may be of limited benefit in patients with very viscous secretions, but both of these treatments (mucolytic agents, oral expectorants) have side effects that
224
Preoperative Respirator}· Preparation
would obviate their use in most patients. The ben efit of mucolytic agents is that they decrease the viscosity of secretions (by depolymerizing muco polysaccharides),96 but at the same time they may induce irritability of the airways and broncho spasm. The benefit of oral expectorants is that they increase the amount of secretions removed, but at the same time they increase the absolute amount of secretions (and may cause gastrointestinal upset and skin lesions).96 Pulmonary infection, if present, is treated ac cording to the results of culture and sensitivity tests; broad-spectrum antibiotics such as ampicillin or a cephalosporin frequently have the required specificity and potency. If the antibiotic treatment clears the infection to any extent, it may also de crease the tenacity, viscosity, and volume of secre tions.
D. Removing the Secretions The next preoperative respiratory step is the ac tual removal of secretions, and this is accom plished by a combination of postural drainage (several different positions may be required), coughing or forced expiration (see later discus sion), and perhaps chest percussion and vibration (common methods include tapping with cupped hands and electric vibrators) for a period 15 to 20 min several times day. 9 7 · 9 8 Consequently, therapy to remove secretions uses gravity, patient-gener ated expiratory airflow, and mechanical measures to dislodge and to propel the secretions proximally. The additional use of inhaled β2-agonists 99 may increase secretion removal. In a patient who has been in bed for a long period of time, getting the patient out of the bed is one of the most im portant lung-expansion maneuvers that can be per formed, often generating a 10 per cent to 20 per 20 cent increase in functional residual capacity. When removal of tracheobronchial secretions is quantitatively measured with radioactive tracer methods in patients with chronic obstructive lung disease, chest physiotherapy (chest percussion and vibration along with postural drainage) with cough is effective in increasing both central and periph eral airway clearance and sputum yield, whereas cough alone is effective in increasing only central airway clearance and sputum yield.100 Further evi dence that physiotherapy with coughing is most effective at clearing secretions from the main bronchi has come from a study of patients with acute lobar collapse.101 The success rate in regain ing lung volume was strongly related to whether there was an air bronchogram: The resolution rate was only 25 per cent when a bronchogram was present compared with almost 90 per cent when
no bronchogram was present. The explanation wa that, with a bronchogram, there was distal alveola collapse but no bronchial sputum plug, whereas n< bronchogram meant that sputum had blocked ; bronchus and caused secondary collapse. In patients with chronic bronchitis, coughing ex ercises have been investigated with radioaeroso techniques and have produced as much centra lung clearance as physiotherapy102 and greater tota clearance than no treatment.103 Thus, chest physio therapy moves peripheral bronchial secretions t( more central airways for expectoration by cough ing. Cough alone cannot clear peripheral airway: because an effective cough must attain a higr enough airflow rate so as to shear secretions awa) from the airway wall. In patients with chronic lun£ disease, flow rates are low (especially peripher ally), and the shearing of secretions by cough ma> well be limited to the trachea and perhaps the firsi two airway generations.104 Obviously, with eithei chest physiotherapy, forced expiration technique, or cough, it will be much easier to expel the secre tions if the airways have already been dilated and the secretions loosened. Chest physical therapy is relatively contraindicated in patients with lung abscesses, bone metas tases, and a history of significant hemoptysis and inability to tolerate the postural drainage positions One review pointed out the possibility of shortterm hypoxemia lasting up to 30 min after physio therapy.10^ Most patients tolerate this hypoxia, but postural drainage, percussion, and vibration must be used with particular care in critically ill patients and in those with low blood concentrations of ox ygen (e.g., monitor with pulse oximetry). Thus, because of the risk of causing either bronchospasm or short-term hypoxemia, postural drainage, per cussion, and vibration should be used only in pa tients in whom its value has been proved (e.g., a patient with copious secretions and a weak cough). The forced expiration technique combined with postural drainage should be strongly considered as an alternative (see later discussion). It is generally agreed that IPPB regimens are not sufficiently ef ficacious to warrant the excessive cost ($60 per treatment) of routine use. 1 0 6 - 1 0 9 The forced expiration technique (FET) is in creasingly regarded as more effective in removing secretions than a cough. 110 The FET comprises a forced expiration starting from midlung volume (50 per cent of inspiratory reserve lung volume) to a low lung volume, usually the residual volume (Fig. 6-6), followed by a period of relaxation andi diaphragmatic breathing. This forceful expiration maneuver differs from a cough because it is per formed without closure of the glottis and the ac companying compressive phase that characterizes cough. The transpulmonary pressure is less during
Preoperative Respiratory Preparation
225
THE FORCED EXPIRATION TECHNIQUE
Figure 6-6 The forced expiration technique begins at 50% of the inspiratory reserve volume and ends at residual volume.
the FET than during cough, resulting in less airway compression and permitting improved proximal and distal clearance of bronchial secretions compared with a conventional cough (as documented by radioaerosol techniques)."0 The FET is now being increasingly used as an alternative to cough during chest physiotherapy."0
E. Ancillary Measures/Issues There are a number of ancillary preoperative therapeutic measures that, in selected patients, may be very important and contribute to the function of the respiratory system. These therapies consist of general measures to improve systemic and major organ well-being and fitness for surgery and specific measures to treat cor pulmonale, including the controversial issue of preoperative digitalization.
1. General Measures (Treatment of Systemic Disease) Several general measures can significantly reduce respiratory morbidity and mortality. Graded activity programs may provide considerable preoperative subjective improvement. If obese patients will cooperate, preoperative weight loss should be strived for. Malnutrition, sometimes present in patients with carcinoma or advanced pulmonary disease, may require preoperative treatment (by nasogastric or intravenous feeding).'" Any other concurrent medical problems (e.g., diabetes, angina [see later discussion]) should be stabilized. Oxygen supplementation by nasal cannula at flows of l to 2 L/min may be helpful for patients
with severe hypoxemia and therefore have a risk of arrhythmia. Most people with angina require drug therapy. Coronary thrombosis is prevented by aspirin. The remaining pharmacologic goals are reduction of myocardial oxygen demand and coronary vasodilation. Nitrates produce venous dilatation and therefore a reduction in venous return and cardiac output. They also cause dilatation of the coronary vasculature. The beta blockers and some calcium antagonists (e.g., verapamil) reduce myocardial oxygen demand by inhibiting exercise-induced tachycardia. Other calcium antagonists (e.g., nifedipine) primarily have a vasodilator effect, reducing peripheral arteriolar resistance and thereby reducing left ventricular work. They may also have a direct vasodilator effect on the coronary arteries. Patients with hypertension are frequently treated with diuretics. If the patient has a hypochloremic alkalosis secondary to diuretic therapy, this should be corrected to reduce any component of hypoventilation in response to the metabolic abnormality. The treatment of choice is KG supplementation. Hypercalcemia caused by lytic bone metastases should be treated with saline hydration, furosemide, possibly drugs to prevent bone resorption (the bisphosphonates etidronate or pamidronate), plicamycin (Mithracin), calcitonin, gallium nitrate, and glucocorticoid."2 Antibiotics are an important component of therapy in COPD patients with acute pulmonary infections that may increase pulmonary vascular resistance. The most common infecting organisms are Haemophilus influenzae and pneumococci, both of which usually can be treated effectively with ampicillin or a cephalosporin."3 Certainly, most clinicians would treat a purulent bronchitis (without
226
Preoperative Respiratory Preparation
evidence of a pneumonia) with these antibiotics. Many surgeons use preoperative (1 week before thoracotomy) antibiotic prophylaxis (deoxytetracycline or cefuroxine) against postoperative infection."4
2. Treatment of Cor Pulmonale The prevalence of cor pulmonale increases with the severity of airway obstruction, occurring in 40 per cent of patients with an FEV, of less than 1.0 L and in 70 per cent of patients with an FEV, of less than 0.6 L." 5 It is also higher in patients with hypoxemia, C0 2 retention, and polycythemia, which likely explains why cor pulmonale is seen more frequently in patients with chronic bronchitis than in those with emphysema. The mainstays of therapy for cor pulmonale are a vigorous pulmonary toilet regimen, oxygen supplementation to relieve hypoxic pulmonary vasoconstriction, diuretics, and possibly digitalis (see the following discussion). Diuretics have been shown to offer significant benefit to patients with right ventricular failure from cor pulmonale." 6 Excess lung water that interferes with gas exchange and increases pulmonary vascular resistance can be reduced by systemic diuresis; careful monitoring of serum electrolytes must be performed. Phlebotomy should be considered whenever the hematocrit reaches 55 per cent or greater. Secondary erythocytosis increases blood viscosity as well as intravascular volume and thereby increases both right ventricular preload and afterload. Hypoxic patients with COPD and polycythemia show improvements in right ventricular ejection fraction, cerebral blood flow, exercise tolerance, and neuropsychologic function after phlebotomy that cannot be attributed to volume reduction alone." 5 The hemodynamic effects of many vasodilators have been studied in patients with COPD but, despite their theoretic potential benefit, vasodilator agents have failed to consistently sustain favorable effects in patients with COPD beyond those achieved by oxygen and bronchodilators. Furthermore, some deleterious effects (e.g., hypoxemia) have been observed with the use of these agents in patients with COPD and secondary pulmonary hypertension (see chapter 4). At present, vasodilator agents are not recommended for routine use in patients with COPD and cor pulmonale." 5
3. Preoperative Digitalization The preoperative use of digitalis in the thoracic surgical patient deserves special comment. Resection of pulmonary tissue reduces the available pulmonary vascular bed for perfusion and can cause postoperative right ventricular, and right atrial en-
largement. Thus, it is not surprising that the incidence of postoperative arrhythmias (due to atrial stretching) increases progressively with age and amount of lung resected. In addition, there is a higher incidence of atrial arrhythmias following left pneumonectomy than right pneumonectomy because a greater degree of manipulation of the atrium occurs during the former operation. Although the postoperative incidence of arrhythmias has provided the basis for the prophylactic use of digitalis in thoracic surgical patients without evidence of congestive heart failure, this practice is still controversial."7-122 Widely accepted indications for preoperative digitalization in patients without cor pulmonale undergoing thoracic surgery include congestive (left-sided) heart failure and supraventricular arrhythmias with a rapid ventricular response.123 I24 The indications for preoperative digitalization are more straightforward in patients with cor pulmonale. However, it is important to note that these patients have a propensity to develop hypoxemia, hypercarbia, and acidosis, and they are therefore at an increased risk of developing digitalis toxicity; the drug should therefore be used with caution.125· I26 If these patients do receive digitalis preoperatively, it is important to normalize the serum potassium in order to decrease the risk of arrhythmias. Digitalis should probably be withheld on the day of surgery to help avoid confusion with digitalis intoxication if arrhythmias occur postoperatively.127
F. Measures to Increase Motivation and Education and to Facilitate Postoperative Respiratory Care The last step consists of general measures designed to increase motivation and education and to facilitate postoperative respiratory care. Preoperative psychologic preparation (i.e., positive suggestion and encouragement), including orientation to the intensive care environment, can reduce fear and improve patient outlook and cooperation. Still, patients should be given realistic expectations about postoperative pain and how it will be handled. With this type of positive but realistic approach, patients will have a diminished postoperative narcotic analgesic requirement and a shorter period of hospitalization.128 Preoperative patient education in the procedures that will be used for respiratory care postoperatively and explaining why these procedures are going to be used will greatly help to ensure optimal postoperative compliance with and performance of the respiratory care maneuvers. There is no question that a critical factor in treating patients
Preoperative Respiratory Preparation
is to preserve and to enhance an individual's ability to do things for him- or herself.129 In this regard, preoperative practice with an incentive spirometer prepares the patient to participate in restoring lung volume and coughing postoperatively. Controlled studies have shown that incentive spirometry is superior to IPPB in terms of both effectiveness and cost.130, ,31 Several incentive spirometers are available, with the more popular ones being Triflo II (which is a relatively inexpensive, small, and easily carried flow-sensitive device), the original volume-sensitive BartlettEdwards Incentive Spirometer, and Spirocare, a relatively expensive, nondisposable, volume-sensitive device that must be plugged into an electric outlet.132 There does not appear to be any significant difference in effectiveness among these three different devices.12 However, it has become eminently clear that the most effective use of these devices depends very heavily upon an effective bedside coach (nurse, respiratory therapist, family member, or friend) who provides frequent verbal encouragement designed to maximize patient compliance with the incentive of the device12 (the devices use either visual or auditory cues to signal achievement of either the volume or flow end point). Other deep-breathing maneuvers, such as blow bottles and blow gloves, actually depend on the inspiratory maneuver that must precede the emphasized expiratory maneuver and, therefore, must be regarded as mild, short-duration forms of incentive spirometry. Carbon dioxide breathing devices had a short period of popularity until it was shown that the resulting increase in minute ventilation was due to an increase in respiratory rate rather than depth of breathing. Preoperative exposure to chest physiotherapy (percussion and vibration), postural drainage, and deep-breathing exercises facilitates use of these maneuvers postoperatively and thereby reduces the incidence of atelectasis.21
VI. MECHANISM OF PREOPERATIVE RESPIRATORY PREPARATION BENEFIT There are several possible mechanisms to explain why preoperative respiratory care preparation maneuvers benefit the patient and result in a decreased incidence of postoperative respiratory complications. Preoperative removal of secretions from the airways is probably the main beneficial result of such intensive preparations for surgery. The removal of secretions is accomplished by the therapeutic cascade of dilating the airways, loosening the secretions, and then removing the secre-
227
tions. Second, the patients may respond nonspecifically to the attention given them, and because of an improved psychologic and motivational status, they may have improved compliance with postoperative respiratory care maneuvers and may ambulate earlier. Third, preoperative instructions in deep-breathing exercises, incentive spirometry and coughing maneuvers, and exposure to chest physiotherapy probably improve the efficaciousness of these maneuvers in the postoperative period. Fourth, there is evidence that many of the techniques used (aminophylline, the various exercises) improve respiratory muscle strength and endurance; however, it is not clear whether the usual 1- to 2-day preoperative regimen is capable of producing significant muscular conditioning as compared with the 5-week period allowed for in more carefully controlled situations.10 Finally, the preoperative respiratory care regimen may actually improve pulmonary function and pulmonary function testing, although this does not occur in every prepared patient even though the incidence of respiratory complications decreases in prepared patients versus nonprepared patients.10 It is likely that some or all of these mechanisms contribute in part to the ability of preoperative respiratory care preparatory maneuvers to decrease the incidence of postoperative respiratory complications.
VII. PREMEDICATION Premedication is individualized according to the psychologic needs of the patient, the severity of pre-existing pulmonary disease, and the anticipated operation. An explanation of the need for various vascular catheters, specific monitoring devices, and use of the face mask for oxygenation (and possibly for inhalation induction of anesthesia) helps alleviate anxiety and promotes patient cooperation in the operating room. For most patients with reasonably good preoperative pulmonary function, a combination of a narcotic analgesic with a benzodiazepine (diazepam, lorazepam) in moderate dosage provides sedation, perioperative analgesia, reduction of anesthetic requirements, and amnesia without concern of predisposing the patient to preoperative respiratory depression. Depression of spontaneous ventilation during anesthesia caused by this type of premedication is rarely a concern because the vast majority of these patients will have their ventilation controlled intraoperatively. Excessively long-acting drugs or very heavy sedation is to be avoided if the operative procedure is short and early postoperative mobilization is desired. The use of anticholinergic drugs in most normal
228
Preoperative Respiratory Preparation
patients and in patients with mild to moderate COPD often causes a fairly uncomfortable feeling as a result of excessive drying; the drying may theoretically cause difficulty in removing secre tions. In addition, modern inhalation anesthetics are much less sialorrheic than their predecessors. Consequently, anticholinergic drugs are not ordi narily used preoperatively. However, because there is good evidence that atropine does not in crease the viscosity of secretions but merely de creases the volume of secretions,133 these drugs should be considered for patients with copious and troublesome amounts of secretions. Although anti cholinergic drugs do increase the dead space to tidal volume ratio, the increase is small. For patients in whom histamine release might be a problem (patients with chronic obstructive lung disease and reactive airways, asthmatics), premedication with the Η,-blocker diphenhydra mine hydrochloride (Benadryl) would seem to be a logical choice; Benadryl provides both sedation and Η,-receptor blockade of histamine-induced bronchospasm. The use of H2-receptor blockers (cimetidine, rantidine) poses the risk of provoking bronchospasm. Histamine mediates bronchoconstriction via Η,-receptors, whereas bronchodilation is mediated by H2-receptors; hence, selective blockade of H2-receptors could unmask unopposed bronchoconstriction. For patients who are thought to have, or may develop, predominance of vagal tone that could induce bronchospasm, atropine is relatively indicated. For patients with a predomi nance of alpha-adrenergic activity that might result in bronchospasm and hypertension, use of droperidol for premedication is relatively indicated. Both hydroxyzine and droperidol decrease airway resis tance. The histamine-releasing opioids, such as morphine and meperidine, also should theoreti cally be avoided in patients with bronchospastic disease. Should opioid premedication be required, fentanyl is suggested. Patients who are hypoxemic on room air (P a 0 2 < 60 mm Hg) or hypercarbic (P a C0 2 > 45 mm Hg) are given little or no premedicants that might further depress gas exchange. Patients who are already on supplemental oxygen preoperatively must be transported to the operating room with the same oxygen flow as previously administered, along with appropriate personnel in constant attendance. Patients who experience or thopnea need to be transported in a semiupright position. Manual or mechanical ventilation during intrahospital transport of critically ill, mechani cally ventilated patients is safe provided the person performing manual ventilation knows the inspired oxygen fraction and minute ventilation required before transport and is trained to approximate them during transport.134
REFERENCES 1. Ford GT, Guenta CA: Toward prevention of postopera tive pulmonary complications. Am Rev Respir Dis 130:4-5, 1984. 2. Lansing AM, Jamilson WG: Mechanisms of fever in ate lectasis. Arch Surg 87:168, 1963. 3. Harris JD, Johanson WG Jr, Pierce AK: Bacterial lung clearance in hypoxic mice. Am Rev Respir Dis 1 11:910, 1975. 4. Latimer G, Dickman M, Clinton DW, Gunn MI, DuWayne SC: Ventilatory patterns and pulmonary compli cations after upper abdominal surgery determined by pre operative and postoperative computerized spirometry and blood gas analysis. Am J Surg 122:622, 1971. 5. Morton HJV, Camb DA: Tobacco smoking and pulmo nary complications after operation. Lancet 1:368, 1944. 6. Tait AR, Kyff JV, Crider B, Santibhavank V, Learned D: Post-operative arterial oxygen saturation—Up in a puff of smoke? Anesth Analg 68:S284, 1989. 7. Handlin DS, Baker T: The effect of heavy and light smoking on patient duration in the post anesthesia care unit. Anesth Analg 72:S99, 1991. 8. Pearce AC, Jones RM: Smoking and anesthesia: Preop erative abstinence and perioperative morbidity. Anesthe siology 61:576-584, 1984. 9. Stein M, Koota GM, Simon M, et al: Pulmonary evalua tion of surgical patients. JAMA 181:765-770, 1962. 10. Gracey DR, Divertie MB, Didier EP: Preoperative pul monary preparation of patients with chronic obstructive pulmonary disease. Chest 76:123-129, 1979. 11. Tisi GM: Preoperative evaluation of pulmonary function. Am Rev Respir Dis 119:293-310, 1979. 12. Van De Water JM: Preoperative and postoperative tech niques in the prevention of pulmonary complications. Surg Clin North Am 60:1339-1348, 1980. 13. Johnson WC: Postoperative ventilatory performance. De pendence upon surgical incision. Am J Surg 41:615-619. 1967. 14. Ali J, Weisel RD, Layng AB, Kripke BJ, Hechtman HB: Consequences of postoperative alterations and respiratory mechanics. Am J Surg 128:376-382, 1974. 15. Anderson WH, Dosett BE Jr, Hamilton GE: Prevention of postoperative pulmonary complications. JAMA 186:763-766, 1963. 16. Tarhan S, Moffitt EA, Sessler AD, et al: Risk of anesthe sia and surgery in patients with chronic bronchitis and chronic obstructive pulmonary disease. Surgery 74:720726, 1973. 17. Harmon E, Lillington G: Pulmonary risk factors in sur gery. Med Clin North Am 63:1289-1298, 1979. 18. Lemmer JH, Gomez MN, Symreng T, Ross AF, Rossi NP: Limited lateral thoracotomy: Improved postoperative pulmonary function. Arch Surg 125:873-877, 1990. 19. Mitchell RL: The lateral limited thoracotomy incision: Standard for pulmonary operations. J Thorac Cardiovasc Surg 99:590-596, 1990. 20. Jackson CV: Preoperative pulmonary function. Arch In tern Med 148:2120-2127, 1988. 21. Thoren L: Postoperative pulmonary complications: Ob servations on their prevention by means of physiotherapy. Acta Chir Scand 107:193-205, 1954. 22. Palmer K.N, Sellick BA: The prevention of postoperative pulmonary atelectasis. Lancet 1:164-168, 1953. 23. Roukema JA, Carol EJ, Prins JG: The prevention of pul monary complications after upper abdominal surgery in patients with noncompromised pulmonary status. Arch Surg 123:30-34, 1988. 24. Veith FJ, Rocco AG: Evaluation of respiratory function
CHAPTER
7
Monitoring I. Introduction A. Special Intraoperative Conditions B. Pre-Existing Lung Disease C. Tiered Monitoring System II. Tier I: Essential Monitoring System A. Checking the Anesthesia Machine B. Continuous Monitoring of the Oxygen-Delivery System C. Continuous Monitoring of Apnea D. Minute Ventilation E. Gas Exchange 1. Cyanosis 2. Pulse Oximetry (Sp02) a. Basic Principles and Clinical Use b. Limitations of Pulse Oximetry c. Use of S p 0 2 From the Digit of the Dependent Hand in the Lateral Decubitus Position During One-Lung Ventilation 3. Capnometry and Capnography a. Basic Principles b. Clinical Use (1) Is Exhaled C0 2 (Waveform) Present? (2) Monitoring Cardiac Output and Cardiopulmonary
232
Resuscitation by EndTidal Carbon Dioxide Concentration (3) Phase I: Inspiratory Baseline (4) Phase II: Expiratory Upstroke (5) Phase III: Expiratory (Alveolar) Plateau (6) Phase IV: Inspiratory Downstroke 4. Approximation of Arterial BloodGas Tensions Using Various Other Body and Equipment Compartments F. Airway Mechanics G. Cardiovascular Parameters H. Muscle Relaxation I. Temperature III. Tier II: Special Intermittent and/or Continuous Monitoring C. Continuous Monitoring of Apnea D. Minute Ventilation E. Gas Exchange 1. Arterial Blood-Gas Analysis 2. Venous Blood-Gas Analysis 3. Transcutaneous 0 2 and C0 2 Tensions
a. Basic Principles b. Clinical Use F. Airway Mechanics G. Cardiovascular Parameters IV. Tier III: Advanced Monitoring Techniques E. Gas Exchange F. Airway Mechanics G. Cardiovascular Parameters 1. Left Ventricular Preload and Left Ventricular Function 2. Various Indices of Left Ventricular End-Diastolic Pressure a. Left Atrial Pressure b. Pulmonary Artery Occluded Pressure (Ppao, Large Pulmonary Vein Pressure) (1) Definition of Ppao (2) Ppaoi PEEP, and Zone III Location (3) Ppao) PEEP, and Zone I Location (4) Ppao and Determination of Zone III Location (5) Special Ppao Considerations Relevant to Preoperative and Postoperative Thoracic Surgery Patients (Excluding the OneLung Ventilation Situation) c. Pulmonary Arterial (True) Wedge Pressure (Ppdw, Pulmonary Distal Wedge Pressure, Small Pulmonary Vein Pressure) d. Pulmonary Capillary Pressure V cap/
e. Pulmonary Artery Diastolic Pressure (Ppad) 3. Various Indices of Right Ventricular End-Diastolic Pressure a. Right Atrial Pressure b. Central Venous Pressure (1) Normal Heart (2) Abnormal Heart 4. Clinical Value of the Pulmonary Artery Catheter 0) Ppao a n c l t n e Morphology of Ppao (Giant a and cv Waves) (2) Cardiac Output (3) Systemic Vascular Resistance
Monitoring (4) Monitoring Mixed Venous Oxygen Saturation (5) Total Benefit 5. Special Pulmonary Vascular Monitoring Considerations Related to Thoracotomy in the Lateral Decubitus Position 6. Risks and Complications of Pulmonary Vascular Pressure Monitoring a. Complications of Gaining Central Venous Access
I. INTRODUCTION Although thoracic surgery may affect both pul monary and cardiovascular function, it more com monly threatens respiratory function much more so than cardiovascular function. Consequently, this chapter emphasizes the monitoring of respiratory function more than cardiovascular function for the routine thoracic surgery patient. In addition, for the very ill patient or for very physiologically in trusive and demanding thoracic surgery cases, ad vanced cardiovascular function monitoring tech niques are equally and fully discussed. Patients undergoing thoracic surgery are most prone to impaired gas exchange because with either one- or two-lung ventilation both nondependent- and dependent-lung function (with reference to lateral decubitus position) will be impaired (see chapter 4). The consequences of both inadequate oxygenation and carbon dioxide elimination are serious. Even slight decreases in the content of oxygen in arterial blood from a marginally normal level may alter a delicate balance of supply and demand to certain tissues and also have wide spread effects due to activation of the sympathetic nervous system. Similarly, changes in blood car bon dioxide content can alter sympathetic nervous system activity, and, in addition, concomitant changes in pH may impair the function of several vital organs. The heart, especially in patients with coronary artery disease, may be the first organ to dramatically show the occurrence of hypoxemia and hypercapnia with the development of ischemia and arrhythmias (see chapter 3). Thus, the need to monitor respiratory function during thoracic anes thesia is a critical and continuous responsibility. The monitoring responsibility can be fulfilled in various ways with various degrees of sophistica tion depending on a particular patient's preopera tive condition and intraoperative requirements. The philosophy espoused by this approach is based upon the concept that there are two considerations that dictate the type of monitoring used. First, pa tients undergoing thoracic operations have varying
233
b. Complications of Actually Passing and Floating the Pulmonary Artery Catheter c. Complications of Maintaining Central Venous Access 7. Transesophageal Echocardiography 8. Lung Water Measurements 9. Computed Tomographic Pulmonary Scan
degrees of pre-existing cardiorespiratory disease (see chapter 5). Second, the very nature of thoracic procedures causes further derangements in cardio respiratory function during the perioperative pe riod (see chapter 4). Thus, on the basis of these two considerations and their interactions, individ ual patients can and should be categorized into a progressively sophisticated and complex tier sys tem with regard to what monitoring is necessary to make possible accurate and rapid diagnosis and therapy during anesthesia (Table 7-1).
A. Special Intraoperative Conditions Patients undergoing thoracic surgery may expe rience several hazardous intraoperative conditions. First, most thoracic surgical procedures are per formed in the lateral decubitus position, which may compromise gas exchange (with either oneor two-lung ventilation, but especially with onelung ventilation; see chapter 4). Second, the sur gical procedure may adversely affect the function of mediastinal organs (e.g., irritate the heart, ob struct the vena cava). Third, some thoracic proce dures may require massive transfusion (such as excision of some arteriovenous malformations); hypotension and massive transfusion are associ ated with the development of the adult respiratory distress syndrome. Fourth, operations necessitating deliberate hypotension increase monitoring re quirements because of changes in pulmonary vas cular autoregulation (hypoxic pulmonary vasocon striction), in VD/VŢ, and perhaps in oxygen transport. Fifth, operations of excessive duration will promote transudation of fluid into dependent regions of the lung and thereby cause progressive oxygenation difficulties. Last, operations involving the airway, such as laryngoscopy, which fre quently necessitates apnea, or bronchoscopy, which frequently imposes restrictions on positivepressure ventilation, require increased respiratory function monitoring.
234
Monitoring
Table 7-1 DERIVATION OF A TIERED MONITORING SYSTEM FOR ANESTHESIA FOR THORACIC SURGERY Pre-Existing Lung Disease
And/Or
Special Intraoperative Conditions
Risk of Respiratory Morbidity and Mortality
Tiered Monitoring System
None
None
Very low
None Moderate Moderate
Moderate None Moderate
Moderate Moderate High moderate
II. Special intermittent and/or continuous monitoring
None Severe Severe
Severe None Severe
High High Very high
III. Advanced monitoring
B. Pre-Existing Lung Disease Patients undergoing thoracic surgery often have significant pre-existing cardiopulmonary disease (see chapter 5). Most of these patients have a long history of heavy smoking and therefore have chronic lung disease with reactive airways and excessive secretions. Due to the lifestyle associated with smoking, many may have coronary artery disease and some may be very obese (with a propensity for decreases in functional residual capacity [FRC] during anesthesia; see chapter 3). Some patients presenting for emergency thoracic surgery will have acute chest disease (pulmonary infection, hemorrhage, contusion, infarction) or acute systemic diseases (sepsis; renal, cardiac, or liver failure; or multiple trauma). Some patients requiring thoracic surgery may be very old.
I. Essential monitoring
interstitial lung disease who requires an open-lung biopsy. Anyone with some lung disease undergoing one-lung ventilation for a major thoracic procedure must be considered to be in a high tier II category, which actually represents a transition category between the tier II and tier III continuum. Finally, a third tier (tier III) of monitoring requirements is constructed for patients with significant pre-existing cardiopulmonary disease and/or for those who will experience major compromising intraoperative conditions. An example of such a patient is one with cor pulmonale undergoing lobectomy or pneumonectomy. Thus, it should be apparent from this monitoring approach that an individual with severe pulmonary disease who is to undergo a minor surgical procedure may well require as extensive a monitoring system as a patient with normal lungs who is to have extensive thoracic surgery.
C. Tiered Monitoring System On the bases of the presence of special intraoperative conditions and the degree of pre-existing lung disease, and the interaction between these two factors, a progressively sophisticated three-tiered monitoring system should be used (see Table 71 ). The first tier (tier I) includes healthy, young patients without special intraoperative conditions, such as a young patient undergoing pleurodesis. This tier contains the minimal, yet essential, monitoring that is required for any patient undergoing a thoracic procedure. The second tier (tier II) represents an increase in risk, caused by either the presence of special unfavorable intraoperative conditions for relatively healthy patients or the presence of significant pre-existing cardiopulmonary disease in patients who will not experience special unfavorable intraoperative conditions. An example of the former circumstance is a patient with mild lung disease having a lobectomy. An example of the latter circumstance is a patient with moderate
II. TIER I: ESSENTIAL MONITORING SYSTEM An essential monitoring system is used for healthy patients undergoing simple, physiologically nonintrusive thoracic procedures (Table 72). Patients belonging exclusively to this monitoring tier make up relatively few of the patients undergoing thoracic surgery. However, since the system represents the minimum amount of monitoring, it is a component of all levels of monitoring for all patients and should allow one to anticipate incipient ventilatory failure as well as to recognize ventilatory failure when it does occur. Although superficially this system may seem unsophisticated, an alert anesthesiologist uses the senses of sight, sound, and touch to gather automatically and reflexly a great deal of information about the wellbeing of the patient. In view of the many possible mechanical mishaps described in chapter 3 that can impair respiratory function, monitoring respi-
Table 7-2
TIERED MONITORING SYSTEM BASED ON AMOUNT OF PRE-EXISTING LUNG DISEASE AND PRESENCE OF SPECIAL INTRAOPERATIVE CONDITIONS
Abbreviations: PIP = peak inspiratory pressure; EKG = electrocardiogram.
236
Monitoring
ratory function during anesthesia begins prior to the induction of anesthesia with checking the anesthesia machine.
A. Checking the Anesthesia Machine Failure to check equipment properly before the induction of anesthesia is responsible for 22 per cent of critical incidents that occur during anesthesia.' Accordingly, preanesthesia check lists have been written to help fulfill this responsibility completely.2 In August, 1986, the U.S. Food and Drug Administration (FDA) published its "Anesthesia Apparatus Checkout Recommendations," shown in Table 7-3. This check applies as a general guideline only, and the FDA encourages users to modify this checkout "to accommodate differences in equipment design and variations in local clinical practice. . . . Users should refer to the operator's manual for special procedures or precautions." In this respect, the user must understand the basic arrangement and functions of the anesthesia machine components so as to apply the correct checkout. (Check lists based on the 1986 FDA Anesthesia Apparatus Checkout Recommendations are available from Organon Inc., West Orange, NJ 07052.) The following mentions only some important anesthesia machine checkout procedures. Inspect the machine to ensure that component parts are connected in the proper order. Vaporizers should be filled and the caps and drains closed. The central gas supply lines should be properly seated in the appropriate pin-indexed wall or ceiling outlets. The tanks should have a wrench, be color coded, be on the proper yokes, and be pin indexed. The tanks should be opened serially and adequately pressurized and the flow of gas observed through the appropriate flowmeter. The carbon dioxide absorber cannister should be full, functional as indicated by color, and fastened securely. The breathing circuit should be tested for leaks under pressure, but the circuit should only be partially filled, and the pressure should be released slowly to avoid depositing soda lime in the inspiratory limb.3 Competence of unidirectional valves should be tested by breathing into the circuit and by noting any undue resistance, presence of irritating gas, and motion (sticking) of the directional valves. The ventilator, scavenger and suction systems, the monitors, and their alarm settings must be checked and tested. A head strap should be available. For the vast majority of thoracic surgery cases, the function of the anesthesia machine ventilator needs to be assessed preoperatively. This can be
done easily for both pressure- and volume-limited, ventilators by serially attaching a rebreathing bag to the patient connection, closing the anesthesia machine overflow valve, turning on the ventilator to either a high pressure or a volume limit, and; observing the movements of the bag. The bag, serving as a test lung, should expand smoothly and easily and deflate in a similar way through the ventilator overflow valve. B. Continuous Monitoring of the Oxygen-Delivery System Since some of the causes of failure of the oxygen delivery system listed in chapter 3 can occur even when the anesthesia machine appeared to function properly preoperatively (e.g., wrong gas in central storage tanks, crossed pipe lines, failure of fail-safe mechanism), the oxygen concentration in the inspired gas should be continuously monitored.4 Numerous paramagnetic, polarographic, and fuel-cell analyzers are available. Ideally, the oxygen analyzer should have a fast response time and high and low alarm setting capabilities, which, when exceeded, trigger audio and visual alarm signals. A model for the study of commercially available oxygen monitors has recently been published.5 It should be noted that a mass spectrometer can function as a most elegant in-line oxygen analyzer. If the oxygen sensor is placed on the expiratory side of the anesthesia circle system (between the corrugated tubing and the expiratory unidirectional valve), then the oxygen sensor will detect the minimum oxygen concentration in the circuit (usually the fraction of oxygen in expired gas is 0.05 less than the fraction in the inspired gas). However, in this position, the oxygen monitor can also double as a circuit disconnection alarm (in addition to continuous chest auscultation, chest observation, and low-pressure alarms) if a falling bellows ventilator (which is the most common type) is in use. The falling bellows of the ventilator draws room air past the oxygen probe, and within one or two breaths following circuit disconnection the alarm will sound if the low limit is set above 20 per cent oxygen/'
C. Continuous Monitoring of Apnea (Fig. 7-1) Precordial and esophageal stethoscopes enable essential and continuous monitoring of breath sounds. Bilateral breath sounds may be easily monitored by joining two taped-on precordial stethoscopes at a stopcock, which is attached to an
Monitoring
237
Table 7-3 ANESTHESIA APPARATUS CHECKOUT RECOMMENDATIONS' This checkout, or a reasonable equivalent, should be conducted before administering anesthesia. This is a guideline which users are encouraged to modify to accommodate differences in equipment design and variations in local clinical practice. Such local modifications should have appropriate peer review. Users should refer to the operators manual for special procedures or precautions. * 1 . INSPECT ANESTHESIA MACHINE FOR: machine identification number valid inspection sticker undamaged flowmeters, vaporizers, gauges, supply hoses complete, undamaged breathing system with adequate CO, absorbent correct mounting of cylinders in yokes presence of cylinder wrench •2. INSPECT AND TURN ON: electrical equipment warm-up (ECG/pressure monitor, oxygen monitor, etc.) *3. CONNECT WASTE GAS SCAVENGING SYSTEM: adjust vacuum as required •4. CHECK THAT: flow-control valves are off vaporizers are off vaporizers are filled (not overfilled) filler caps are sealed tightly CO, absorber by-pass (if any) is off *5. CHECK OXYGEN (0 2 ) CYLINDER SUPPLIES: a. Disconnect pipeline supply (if connected) and return cylinder and pipeline pressure gauges to zero with O, flush valve. b. Open O, cylinder; check pressure; close cylinder and observe gauge for evidence of high pressure leak. c. With the O, flush valve, flush to empty piping. d. Repeat as in b. and c. above for second O, cylinder, if present. e. Replace any cylinder less than about 600 psig. At least one should be nearly full. f. Open less full cylinder. *6. TURN ON MASTER SWITCH (if present) •7. CHECK NITROUS OXIDE (N 2 0) AND OTHER GAS CYLINDER SUPPLIES: Use same procedure as described in 5a. & b. above, but open and CLOSE flow-control valve to empty piping. Note: N,0 pressure below 745 psig. indicates that the cylinder is less than VA full. *8. TEST FLOWMETERS: a. Check that float is at bottom of tube with flow-control valves closed (or at min. O, flow if so equipped). b. Adjust flow of all gases through their full range and check for erratic movements of floats. •9. TEST RATIO PROTECTION/WARNING SYSTEM (if present): Attempt to create hypoxic 0,/N,0 mixture, and verify correct change in gas flows and/or alarm. *10. TEST O, PRESSURE FAILURE SYSTEM: a. Set O, and other gas flows to mid-range. b. Close O, cylinder and flush to release O, pressure. c. Verify that all flows fall to zero. Open O, cylinder. d. Close all other cylinders and bleed piping pressures. e. Close O, cylinder and bleed piping pressure. f. CLOSE FLOW-CONTROL VALVES. •II. TEST CENTRAL PIPELINE GAS SUPPLIES: a. Inspect supply hoses (should not be cracked or worn). b. Connect supply hoses, verifying correct color coding. c. Adjust all flows to at least mid-range. d. Verify that supply pressures hold (45-55 psig.). e. Shut off flow-control valves. *If an anesthetist uses the same machine in successive cases, these steps need not be repeated or may be abbreviated after the initial checkout.
* 12. ADD ANY ACCESSORY EQUIPMENT TO THE BREATHING SYSTEM: Add PEEP valve, humidifier, etc., if they might be used (if necessary remove after step 18 until needed). •13. CALIBRATE O, MONITOR: *a. Calibrate O, monitor to read 21% in room air: *b. Test low alarm. c. Occlude breathing system at patient end; fill and empty system several times with 100% O,. d. Check that monitor reading is nearly 100%. 14. SNIFF INSPIRATORY GAS: There should be no odor. •15. CHECK UNIDIRECTIONAL VALVES: a. Inhale and exhale through a surgical mask into the breathing system (each limb individually, if possible). b. Verify unidirectional flow in each limb. c. Reconnect tubing firmly. •*I6. TEST FOR LEAKS IN MACHINE AND BREATHING SYSTEM: a. Close APL (pop-off) valve and occlude system at patient end. b. Fill system via O, flush until bag just full, but negligible pressure in system. Set O, flow to 5 L/min. c. Slowly decrease 02 flow until pressure no longer rises above about 20 cm Η,Ο. This approximates total leak rate, which should be no greater than a few hundred ml/min. (less for closed circuit techniques). CAUTION: Check valves in some machines make it imperative to measure flow in step c. above when pressure just stops rising. d. Squeeze bag to pressure of about 50 cm H,0 and verify that system is tight. 17. EXHAUST VALVE AND SCAVENGER SYSTEM: a. Open APL valve and observe release pressure. b. Occlude breathing system at patient end and verify that negligible positive or negative pressure appears with either zero or 5 L/min. flow and exhaust relief valve (if present) opens with flush flow. 18. TEST VENTILATOR: a. If switching valve is present, test function in both bag and ventilator mode. b. Close APL valve if necessary and occlude system at patient end. c. Test for leak and pressure relief by appropriate cycling (exact procedure will vary with type of ventilator). d. Attach reservoir bag at mask fitting, fill system and cycle ventilator. Assure filling/empyting of bag. 19. CHECK FOR APPROPRIATE LEVEL OF PATIENT SUCTION. 20. CHECK. CONNECT, AND CALIBRATE OTHER ELECTRONIC MONITORS. 21 CHECK FINAL POSITION OF ALL CONTROLS. 22. TURN ON AND SET OTHER APPROPRIATE ALARMS FOR EQUIPMENT TO BE USED. (Perform next two steps as soon as is practical) 23. SET O, MONITOR ALARM LIMITS. 24. SET AIRWAY PRESSURE AND/OR VOLUME MONITOR ALARM LIMITS (if adjustable). **A vaporizer leak can only be detected if the vaporizer is turned on during this test. Even then, a relatively small but clinically significant leak may still be obscured.
*As developed by the (U.S) Food and Drug Administration, August 1986. Abbreviations: ECG = electrocardiogram; PEEP = positive end-expiratory pressure; APL = airway pressure line.
238
Monitoring
Monitoring of Only Apnea and Minute Ventilation
Figure 7-1 The number of things that an anesthesiologist does to monitor minute ventilation and to detect continuously the onset of apnea is considerable. Most of the monitoring utilizes the basic senses of sight, sound, and touch and is done reflexively and automatically by an experienced anesthesiologist. These automatic monitoring efforts, for the most part, make up the tier I monitoring level (solid arrows). Additional tier I monitoring modalities include use of an in-line F,0 : monitor and low and high airway pressure alarm systems. Tier II monitoring efforts include movement of in-line spirometer and observation of respiratory variation in vascular pressures (dashed arrows). Both tier I and tier II monitoring efforts should be continuously conducted in a systematic and frequently recurring circular pattern (cockpit analogy).
ear piece. This simple device allows one to listen to both hemithoraces by simply turning the stopcock and is useful in a variety of surgical cases with difficult access to the chest and when lung isolation is required. 7 · s The system is particularly useful in neonates and small infants because of their short trachea and consequent high risk for endobronchial intubation. Additionally, the anesthesiologist should frequently (every minute) scan the chest, breathing bag, ventilator bellows, and pressure manometers on the anesthesia machine for appropriate movement. The anesthesia machine-mechanical ventilator system should have a low and high positive-pressure audiovisual alarm system. This latter monitoring aid must now be viewed as essential because some accidental anesthesia-caused or -related deaths (brain and/or whole body) have been, and presently are, due to ventilator disconnections in otherwise healthy patients. As just discussed, oxygen sensors can serve as disconnection alarms if placed on the expiratory limb of an anesthesia circle system and if a falling bellows ventilator is in use.
D. Minute Ventilation For the healthy patient having a physiologically nonintrusive thoracic procedure, only a crude assessment of minute ventilation (see Fig. 7-1) is necessary. Minute ventilation is the product of respiratory rate per minute and tidal volume. Counting respiratory rate with either spontaneous, assisted, or mechanically controlled respiration presents no difficulties. However, the limitation in measuring minute ventilation precisely without instrumentation is in accurately measuring the tidal volume. When a tier I patient is mechanically ventilated, the setting of ventilator bellows (10 to 15 ml/kg) and patient chest movement usually suffices to gauge the tidal volume. However, it should be realized that the movement of ventilator bellows may be in error by 50 per cent, depending on respiratory rate, inspiratory-to-expiratory ratio, gas flow rates, and patient compliance and resistance. With most anesthesia machine ventilators, the higher the flowmeter settings (0 2 , N 2 0, air) during inspiration, the greater will be the tidal volume (by
Monitoring
an amount = flow rate X inspiratory time), and the longer the inspiratory time (lower I:E) for any given flow rate, the greater will be the tidal volume (by an amount = flow rate X the inspiratory time).9 Thus, a high gas flow rate with a low I:E ratio can increase minute ventilation by 50 per cent (with correspondingly lower P a C0 2 ) above what is set by the ventilator bellows and respiratory rate.9 In addition, it should also be realized that, with spontaneous ventilation, visual or "educated" hand assessments of chest or anesthesia reservoir bag movements in questionable or marginal cases may be misleading,10 especially with inexperi enced observers." Conversely, with respect to con trolled ventilation, experienced anesthetists can achieve a higher minute ventilation with lower peak and end-expiratory alveolar pressures with manual compared with mechanical ventilation with two commonly used anesthesia ventilators (Ohmeda 7000 and Drăger AV-E).12 In any pa tient, an unusually strong respiratory drive in the presence of high doses of inhaled or intravenously administered anesthetics may be the consequence of hypercapnia and hypoxia.
E. Gas Exchange 1. Cyanosis The simplest and perhaps most common signs of severe hypoxemia are purple or dark shed blood' 3 u and/or the appearance of cyanosis.'5· '6 Most of the blood in skin and mucous membranes is venous and is related to the C a 0 2 as follows (rearranged Fick equation; see equation ll in chapter 3): C v 0 2 = C a 0 2 - C(a-v)0 2 = C a 0 2 -
tissue Vo-,1 -r tissue Q
The oxygen consumption by skin (Vo 2 ) is very low in relation to its circulation (Q) so that the quantity in the far right term is generally small. Therefore, skin C v 0 2 is close to C a 0 2 . Cyanosis can be detected by experienced observers in 95 per cent of awake patients with an arterial saturation (S ;1 0 2 ) of 89 per cent with the aid of fluorescent lighting17 (corresponds to a P a 0 2 = 55 mm Hg), whereas cyanosis does not always become evident in awake patients in ambient lighting until S a 0 2 is H 72 per cent (corresponds to P a 0 2 = 36 mm Hg).' It has been observed that cyanosis is not at all consistently detected in anesthetized patients even at S p 0 2 values as low as 72 per cent. 18 This sug gests that, during inhalation anesthesia, cyanosis is 15, ,6 l8 a less reliable sign than in awake patients. · The greater difficulty in detecting cyanosis during
239
halogenated hydrocarbon anesthesia may be re lated to relative hypoperfusion of the skin and mucous membranes.14 It has also been noted that, during clinical epi sodes of hypoxemia (S p 0 2 < 85 per cent), the heart rate electrocardiogram (EKG), blood pres sure, and respiration hardly change.18 This obser vation is explained by the results of studies of volunteers that showed that both cardiovascular and ventilatory responses to induced moderate hy poxemia (i.e., P E T 0 2 40-45 mm Hg) are severely impaired or abolished by commonly used halogen ated anesthetics. 19~21 Clearly, cardiorespiratory monitoring, including the EKG, has little or no value as an indicator of moderate hypoxemia dur ing anesthesia with these agents.14 In addition, in cases of anemia in which it is not possible to obtain the level of 5 g/100 ml of reduced hemoglo bin, which is generally accepted to be required for the appearance of cyanosis, cyanosis may not oc cur in the presence of abnormal C a 0 2 ("central cyanosis"). 22 Clearly, cyanosis could never occur if the hemoglobin concentration was only 5 g/100 ml. Alternatively, when skin circulation is reduced in relation to skin oxygen consumption, as may occur in hypovolemia (vasoconstriction) and in the Trendelenburg position (stagnant flow), cyanosis may occur in the presence of normal C a 0 2 (periph eral cyanosis). Thus, cyanosis should always be looked for, but since it can be falsely positive, it should be regarded as a warning sign requiring further investigation. Since the absence of cy anosis can be falsely negative, the anesthesiologist should not be lulled into a sense of complacency. In addition, recognition of cyanosis depends upon the quality of lighting, reflection from drapes, and the observers themselves.
2. Pulse Oximetry (SpOJ Pulse oximetry is the technology whereby the pulse oximeter provides a noninvasive estimate of arterial hemoglobin saturation with oxygen. The American Society of Anesthesiologists Standards for Basic Intraoperative Monitoring (last amended October 23, 1990, to become effective January 1, 1991) state (standard II, oxygenation, methods 2) that "during all anesthesia a quantitative method of assessing oxygenation, such as pulse oximetry, shall be employed." a. BASIC PRINCIPLES AND CLINICAL USE
Pulse oximetry is a combination of two technol ogies, namely, spectrophotometry (which meas ures hemoglobin saturation by the amount of light absorbed by the blood) and optical or photoplethysmography (which compares light absorbance in the absence of a pulse [low finger blood vol-
240 M on it or ÎHÎ; urne], its zero, and in the presence of the arterial pulsation [high finger blood volume]). Pulse oximeters provide instantaneous, in vivo measurements of arterial oxygenation by determining the color of the blood between a light source and a photodetector. To be able to distinguish between two species of hemoglobin (i.e., oxyhemoglobin [HbO:] and deoxyhemoglobin [Hb]), it is necessary to measure absorption at two different wavelengths. This is accomplished by using a light source that consists of two different light-emitting diodes, one emitting red (^όόΟ nm) and the other infrared (^940 nm) light. Oxyhemoglobin absorbs less red light than deoxyhemoglobin, accounting for its red color; at infrared wavelengths, the op posite is true. During each cardiac cycle, light absorption by tissue beds varies cyclically (Fig. 7-2): During "diastole," absorption is caused by venous blood, tissue, bone, and pigments (melanin, nail polish, etc.); during "systole," there is an increase in light absorption that pulse oximeters assume to be cre ated by the influx of arterialized blood into the tissue bed. The oximeter then determines the dif ference between background absorption during dias tole and peak absorption during systole at both red and infrared wavelengths; these changes correspond to the absorption caused by arterialized blood.23 Ox ygen saturation determines the red:infrared absorp tion ratio; thus, the red:infrared ratio of these pul satile differences can be used to compute the pulse oximeter reading (S p 0 2 ), which is an estimate of arterial oxygen saturation (S ;1 0 2 ). 23 The pulse oximeter will function anywhere it can reliably detect a pulse (e.g., the digits, ear, tongue, 2 4 · 2 S nose,26 buccal mucosa27). The further the sensing site is from the lung, the longer it will take for the pulse oximeter to express the change in oxygenation (i.e., acute changes in S a 0 2 may not be reflected by finger vs. ear S p 0 2 for as long as 30 sec vs. 5-10 sec, respectively 28 · 29 ). The re sponse time at low S p 0 2 is not nearly so dependent on sampling site as are changes at high S p 0 2 . 3()
Response time in change from S p 0 2 = 100 pi cent will also be a function of F,0 2 , whereas η sponse time for change from a low S p 0 2 will n< be.31 A pulse waveform display enables the open tor to determine whether the oximeter is detectir valid pulses or interfering signals. By analogy, a arterial catheter without a waveform display migl give accurate pressure readings under most cond tions, but erroneous readings resulting from art fact would be difficult to recognize. S p 0 2 valut may be expressed as a compressed spectral arra (i.e., as a function time and the amount of tim spent at a particular S p 0 2 ) (Fig. 7-3). λ2 Most mar ufacturers specify that their oximeter readings ca be expected to have a 2 to 3 per cent standar deviation in the 70 to 100 per cent saturatio range. At these relatively high saturations, the 9 per cent prediction limits of a single pulse oxime ter reading are ± 6 per cent (i.e., there is a 95 pe cent probability that an oximeter reading of 90 pe cent corresponds to an arterial saturation betwee 84 and 96 per cent). In the range of saturatio equal to 40 to 70 per cent, the variation may be a great as 10 per cent. 29 The lack of reliability at lo\ S p 0 2 is confirmed by studies in infants with SaC less than 60 per cent.13 The definition of saturation is the cause of muc confusion. Laboratory co-oximeters separatei measure deoxygenated hemoglobin (RHb), ox> genated hemoglobin (Hb0 2 ), and the dyshemoglc bins methemoglobin (MetHb) and carboxyhemc globin (COHb). From these can be calculated the
Figure 7-2 Components of lighi absorption by a tissue bed as a func tion of time. (From Alexander CM Teller LE, Gross JB: Principles ol pulse oximetry: Theoretical anc practical considerations. Anesth An· alg 68:368-376, 1989. Used with permission.)
Monitoring
241
Figure 7-3 Sp02 CSA of a 58-yearold male after laparotomy/adhesiolysis. Mild hypoxemia was noted in the first hour, but then patient was instructed in the use of 02 nasal cannula (2 L/min), which he wore all night. O-, was discontinued by physician or nurse (pt unsure) after 0800, after which patient developed increasing hypoxemia with S p 0 2 in the low 80s by midafternoon. (From Cohen MS, Eichhorn JH, Darvish AH: Postanesthetic hypoxemia due to removal of supplemental oxygen therapy. Anesth Analg 74:545, I992.)
2 per cent, the S a 0 2 would be 100 per cent (96/0 + 96 = 96/96), whereas the Hb0 2 per cent would be 96 per cent (96/0 + 96 + 2 + 2 = 96/100). However, because pulse oximeters respond to the dyshemoglobins in a very complex way (see later discussion), the reading from a pulse oximeter is a unique value and must be given its own designation S p 0 2 . When the dyshemoglobinemia is mild/moderate, the S p 0 2 is a close approximation of both Hb0 2 per cent and S a 0 2 (see later discussion). The waveform that a pulse oximeter displays is essentially a photoplethysmogram. Acute hypovolemia may cause cyclic beat-to-beat variation in both systolic blood pressure and the pulse oximeter waveform that is synchronous with respiration. Even though the display has an automatic gain and hence cannot be considered quantitative, variation in the pulse wave seems to correlate with systolic pressure variation and was found to be a useful guide to fluid therapy.34 When the pulse oximeter fails to detect the usual pulse during positive-pres-
Figure 7 - 4 Top panel, Waveform during positive-pressure ventilation. Displayed rate is equal to respiratory rate: Displayed saturation is low. Bottom panel, Waveform during expiratory pause in ventilation. Displayed rate is equal to heart rate; displayed saturation is normal. (From Scheller J, Loeb P: Respiratory artifact during pulse oximetry in critically ill patients. Anesthesiology 69:602-603, 1988. Used with permission.)
sure ventilation (Fig. 7-4),35 it may mean that the patient is hypovolemic and the extremity blood flow is further reduced by a positive-pressure ventilation-induced decrease in venous return. In this situation, instead of readjusting the pulse oximeter or changing the pulse oximeter to a new site, positive-pressure ventilation should be briefly interrupted to obtain an accurate reading (see Fig. 7-1 ) and fluid boluses administered to restore perfusion and pulse oximeter function during positive-pressure ventilation. Certainly, in oximeters that display the pulsatile waveform and that can fix gain (signal strength), the amplitude of the waveform may be helpful as an absolute monitor of the circulatory volume and as an evaluation of the hemodynamic significance of cardiac arrhythmias. b. LIMITATIONS OF PULSE OXIMETRY
A pulse oximeter will not function well under a variety of conditions (Table 7-4). First, to measure pulse-added absorbance, a pulse oximeter requires a pulsatile arterial bed to be present. Readings may
242
Monitoring
Table 7-4
LIMITATIONS OF PULSE OXIMETRY
1. 2. 3. 4. 5. 6.
Will not function without a pulse Motion artifact Interference from large venous pulsation Interference from dyes, nail polish, dark skin, jaundice Interference by dyshemoglobins (MetHb and COHb) Interference by electrocautery, ambient light, and infrared radiation 7. Insensitive with high P^CX
not be obtainable when pulsatile flow is lost, such as during intense vasoconstriction, hypothermia, severe peripheral vascular disease, hypovolemia, or cardiopulmonary bypass. Hypothermia-induced loss of pulse oximeter function, presumably resulting from vasoconstriction, may be reversed by volar digital nerve blocks.36 However, from a rheometric standpoint, it should always be remembered that the presence of a functioning pulse oximeter should not be construed as evidence of adequate tissue oxygenation or oxygen delivery to vital organs." Second, patient or probe movement may interfere with oximeter function because these motions cause changes in optical path length. The oximeter may then fail to recognize actual pulsatile flow. Flexible, adhesive probes, which are taped in place, tend to be less susceptible to motion artifacts than clip-on probes (i.e., the orientation of the probe to the finger is more stable). Third, pulse oximetry requires a nonpulsatile venous bed. Venous pulsations (e.g., caused by tricuspid regurgitation) are not distinguishable from arterial. Venous pulsations will contribute to the pulse-added absorbance signal, causing a spurious desaturation reading. Fourth, pulse oximetry requires hemoglobin as the dominant color species present. Intravascular dyes, if present, may interfere with function. Methylene blue and indocyanine green cause spurious desaturation, whereas indigo carmine has little or no effect on the S p 0 2 reading. Similarly, certain colors of nail polish, skin pigmentation, and jaundice may cause spurious readings.38 Fifth, the saturation reading is based on absence of significant amounts of dyshemoglobins (e.g., MetHb, COHb). Their presence may result in spurious readings. Oximeters respond to COHb as if it were HbO,, and the pulse oximeter reading (SpO;) is the sum of COHb and Hb0 2 (i.e., if a patient had 90 per cent Hb0 2 , 7 per cent COHb, and 3 per cent Hb, a pulse oximeter would read S 02 = 97 per cent). MetHb corresponds to an S p 0 2 reading of 85 per cent. Therefore, in the presence of MetHb, the pulse oximeter reading will be a weighted average of the 02 saturation of the hemoglobin available for transport plus 85 per
cent. With fetal hemoglobin and the hemoglobinopathies, the heme moiety itself is unaffected, and such changes do not produce clinically significant alterations in S p 0 2 readings. Sixth, pulse oximeters are sensitive to interference by electrocautery, intense ambient light, and infrared radiation. Seventh, pulse oximeters are insensitive to changes in oxygenation at the high end of the oxyhemoglobin dissociation curve. At arterial oxygen tensions above about 120 mm Hg, saturation is 100 per cent. The pulse oximeter is thus a poor predictor of P a 0 2 until the saturation falls below 100 per cent. A patient receiving an F,0 2 of 1.0 may have a large shunt and yet still show an Sp02 of 100 per cent. For this reason, the pulse oximeter is a poor indicator of endobronchial intubation or a poor early indicator of esophageal intubation, especially if the F,0 2 is greater than O.5.31 c. USE OF S p 0 2 FROM THE DIGIT OF THE DEPENDENT HAND IN THE LATERAL DECUBITUS POSITION DURING ONE-LUNG VENTILATION
It is clear that pulse oximetry performs well, with a predictable error range of 2 per cent in normal patients who do not have any of the limitations just discussed. However, in a patient in the lateral decubitus position undergoing one-lung ventilation, there are many potential (and perhaps unavoidable) sources of error. These potential sources of error are rapidly changing oxygenation levels with lung surgery (e.g., clamping, compression of vessels, and lung parenchyma), which are reflected relatively slowly at a distal sensing site, venous congestion in a dependent extremity, hypovolemia, decreased cardiac output with impaired venous return, arrhythmias, hypothermia with the chest open, large tidal volume positive-pressure ventilation of one lung, and electrocautery. One study very clearly highlights this problem by showing that there was excellent agreement between S p 0 2 and S a 0 2 while a series of thoracic surgery patients were awake, but once anesthesia had been induced, the patients were in the lateral decubitus position, and surgery had begun, the agreement between S p 0 2 and S;,02 was much worse.39 These data support the conclusion that the accuracy of pulse oximeters is altered during anesthesia for thoracic surgery and should be relied on only as a warning of impending hypoxemia when the S p 0 2 falls. This recommendation is similar to that of Chung et al.,40 who also found that transcutaneous monitors of oxygen were not accurate during thoracic surgery. It was appropriately concluded that, although the information derived from pulse oximetry is useful in identifying impending hemoglobin desaturation caused by hypoxemia, S p 0 2 alone should not be used as a sub-
Monitoring
stitute for frequent arterial blood gas determinations during thoracic surgery.
3. Capnometry and Capnography a. BASIC PRINCIPLES
Airway C0 2 can be monitored continuously by either small, freestanding operating room infrared spectrophotometers or large, centrally housed respiratory mass spectrometers. Freestanding infrared units are by far the most commonly used because there is no delay in read-out and there are two types: in-line and side-stream monitors. The inline C0 2 sensor is connected directly to the endotracheal tube and receives the total alveolar gas flow, whereas the side-stream type receives a portion of the alveolar flow via a side-port connector. The in-line type has the intrinsic advantage of maximal frequency response and minimal delay and is therefore most suitable for infants. Disadvantages of the in-line type include calibration inaccuracy while in use and the encumbrance of a sensing unit at the patient interface. The weight of the transducer and its associated connections may restrict patient movement and may necessitate sedation of restless patients. The side-stream monitor adds essentially no encumbrance to the patient interface and has the additional advantage of in-use calibration capability. However, to achieve adequate frequency response, side-stream flows between 125 and 500 ml/min are required. These rates are too high for accurate end-tidal measurements in neonates and infants. The best measure of P E T C 0 2 is obtained when (1) tidal volumes are large enough to displace dead space; (2) fresh gas flow rates are low enough to prevent dilution or washing out of C0 2 ; (3) sample aspiration rates are low enough that they do not interfere with patient ventilation or entrain air that may dilute the C0 2 ; (4) the sampling site is close to the patient, minimizing the dead space; and (5) the waveform is displayed for end-tidal alveolar plateau analysis.41 Despite the use of elegant mucus traps and filters, occlusion of sampling lines and sensor by aspirated mucus and debris is a common clinical problem. In infants and small children (weighing 12 kg or less) ventilated with a partial rebreathing circuit, the sampling site becomes a significant dependent variable. Capnographic waveforms from distal endotracheal tube sampling sites must and do show constant plateau phases during expiration, whereas those from the proximal endotracheal tube sites fail to achieve a plateau and underestimate the PaC02.42 This is consistent with the hypothesis that gas sampled at the proximal site is a mixture of expired alveolar gas and fresh gas removed from
243
the circuit by sampling. In any situation in which P E T C 0 2 does not correlate with what the clinician expects (e.g., P E T C 0 2 monitoring in infants), P a C0 2 should be measured as a guide for P E T C O : tracking and interpretation. A change in the P a C0 2 -PETC0 2 gradient in itself may indicate an important pathophysiologic change (e.g., change in dead space; see later discussion). b. CLINICAL USE
The anesthesiologist will obtain the most information out of a capnograph if it is examined systematically. First, the anesthesiologist must determine whether exhaled C0 2 (i.e., a waveform) is present. Second, the shape of the waveform must be analyzed systematically by looking at, and in sequence, phase I (inspiratory baseline), phase II (expiratory upstroke), phase III (expiratory plateau), and phase IV (inspiratory downstroke) (Fig. 7-5). This discussion will follow this systematic approach but will emphasize diagnoses that can be obtained from the phase III alveolar plateau. In the future, capnography may display carbon dioxide concentration as a function of the volume of exhaled gas from each breath (the single-breath test, or SBT) because the SBT allows on-line determination of Vco 2 (C0 2 concentration X tidal volume) and is a more sensitive test of, and reveals more information about, gas elimination in the late phase of each breath.43 ( 1 ) . IS EXHALED CO z (WAVEFORM) PRESENT? If
the presence or persistence of C0 2 is not detected by the capnometer, or capnogram, failure to ventilate the patient's lungs must be assumed. The differential diagnosis of absent C0 2 includes esophageal intubation, accidental tracheal extubation, disconnection of the breathing circuit, complete obstruction of the endotracheal tube, conducting system (kink, inspissated blood or secretions, extremely severe bronchospasm), or breathing circuit, apnea, and cardiac arrest (Table 7-5).44 With respect to esophageal intubation, capnometry/capnography can make the diagnosis in one breath and is therefore far superior to pulse oximetry, which usually requires some time for desaturation to occur.31 If the stomach contains exhaled gas from previous mask ventilation attempts or carbonated beverages, a few tidal ventilations through an esophageal tube may contain minimal and progressively diminishing concentrations of carbon dioxide. It should be noted that very severe bronchospasm (enough to cause complete airway obstruction) can prevent carbon dioxide from being registered by the capnogram even though the trachea has been properly intubated.45 Only after all of these life-threatening possibilities have been quickly ruled out and ventilation of the patient's lungs confirmed by clinical examination
244 M ont tarins
EXHALED C02 CONCENTRATION
Figure 7-5
The four phases of the capnogram.
should failure of the capnometer or capnogram be considered in the differential diagnosis. A rapid qualitative check of the capnogram consists of simply removing the C 0 2 sensing or sampling port and exhaling into it. Of all the entities in the differential diagnosis listed previously, monitoring cardiac output and cardiopulmonary resuscitation during cardiac ar rest is a new application of capnography and there fore is the only entity further discussed. ( 2 ) . MONITORING CARDIAC OUTPUT AND CAR DIOPULMONARY RESUSCITATION BY END-TIDAL CARBON
DIOXIDE
CONCENTRATION.
During
steady-state gas exchange equilibrium, the alveolar Pco : (Ρ Λ 0 2 ), tissue C 0 2 production (Vco 2 ), and
alveolar ventilation (VA) are uniquely related as given by P A C 0 2 = (K)Vco/V A . During constant minute ventilation and Vco 2 , an abrupt reduction in cardiac output (Q,) reduces P E T C O , by two mechanisms.46 47 First, a reduction in venous re turn causes a decrease in C 0 2 delivered to the alveolar compartment, resulting in decreased P A C0 2 . Second, the increase in alveolar dead space, which results from the reduced pulmonary vascular pressures, will dilute the C 0 2 from nor mally perfused alveolar spaces to decrease P E T C O below P A C0 2 (see later discussion). During a sus tained reduction in Q,, increasing C 0 2 accumula tion in the peripheral tissues and in venous blood will begin, after 10 to 20 min, to restore C 0 2 delivery to the lung and P E T C 0 toward baseline levels. Reciprocal changes in P E T C O will occur during acute increases in Q,. These Q, versus P E T C 0 observations have been made quantita tively and with very good correlation in both con trolled experiments in animals 46 and in patients undergoing major vascular and cardiac surgery.47 With cardiac arrest, there is no pulmonary blood flow and therefore no delivery of carbon dioxide to the lungs. Consequently, P E T C 0 exponentially decreases over a dozen breaths, and there is no steady-state exhaled C 0 2 (waveform).48 However, with external cardiac compression, pulmonary blood flow will begin again and the amount of :
2
:
Table 7-5
DIFFERENTIAL DIAGNOSIS OF ABSENT EXHALED C0 2
I. Exhaled CO : usually present, then absent 1. Accidental tracheal extubation 2. Disconnection of breathing circuit 3. Monitor failure 4. Complete obstruction of endotracheal tube 5. Cardiac arrest 6. Patient becomes apneic II. Exhaled C 0 3 usually absent/minimal initially, remains ab sent 1. Esophageal intubation
2
2
Monitoring
C0 2 excreted by the lungs (i.e., P E T C 0 2 ) will be proportional to the amount of pulmonary blood flow (see previous discussion). Indeed, the efficacy of external cardiac compression can be continuously and quantitatively followed by the amount of C0 2 excreted (Fig. 7-6, top panel).49·50 Exhaled C0 2 during cardiopulmonary resuscitation can also be used prognostically.5153 In one study the initial P E T C 0 2 was 19 mm Hg in those who eventually regained spontaneous pulses but only 5 mm Hg in those who did not (p < .0001 ).54 Furthermore, a sharp increase in ETco 2 often heralded, and was the first indicator of, the resumption of spontaneous circulation.52 54 Almost identical data were found in yet another study (Fig. 76, bottom panel).49 Figure 7-7 shows all of this, in addition to the effects of sodium bicarbonate infusion, in a patient undergoing ventricular fibrillation, cardiopulmonary resuscitation, and resumption of a spontaneous circulation.51 Thus, capnography provides an instantaneous and continuous guide to the efficacy of external chest compression and the resumption of spontaneous pulmonary perfusion. (3). PHASE i: INSPIRATORY BASELINE. The in-
spiratory baseline is traced as fresh gas moves over the C0 2 sensing or sampling site. The C0 2 level during this phase should be zero; if it is not, C0 2 is being rebreathed. This may be intentional and/or a characteristic (desirable or undesirable) of the
245
equipment being used. The inspiratory baseline becomes elevated if C0 2 is added to the fresh inspired gas, if a C0 2 rebreathing or by-pass valve (which is present on older anesthesia machines) is open, if the C0 2 absorbent is partially exhausted or gas is channeling through the absorbent, if the expiratory valve is missing or incompetent (exhaled gas in the exhalation limb goes back into the patient's lungs during inhalation, thereby pulling C0 2 containing gas by the sampling site during inhalation), or if a Bain circuit is being used. ( 4 ) . PHASE II: EXPIRATORY UPSTROKE. S o o n
after exhalation begins, C0 2 containing gas arrives at the C 0 2 sampling site, and it quickly washes away the fresh gas from the previous inspiration. Thus, the expiratory upstroke is steep. When the expiratory upstroke phase of the capnogram becomes prolonged (i.e., the upstroke becomes less steep), delivery of C0 2 from the lungs to the CO : sampling site is delayed. Possible causes include mechanical obstruction in the equipment, such as a kinked endotracheal tube, or slow emptying of the lungs, such as with chronic obstructive pulmonary disease (COPD) or bronchospasm. The expiratory upstroke also becomes prolonged when a side-stream capnogram samples gas too slowly or when the capnogram has a slow response time and the respiratory rate is fast. ( 5 ) . PHASE III: EXPIRATORY (ALVEOLAR) PLA-
Figure 7-6 Top panel, Infrared analyzer trace showing end-tidal carbon dioxide concentration with two different physicians giving external chest compression. Bottom panel, infrared analyzer trace showing changes in end-tidal carbon dioxide concentration with successful defibrillation and recurrence of ventricular fibrillation. (From Nielson MS, Fitchet A, Saunders DA: Monitoring cardiopulmonary resuscitation by end-tidal carbon dioxide concentration. Br Med J 300:1012-1013, 1990. Used with permission.)
246
Monitoring
Figure 7-7 Serial changes in the end-tidal carbon dioxide concentration (ETco : ) and arterial (A) and mixed venous (PA) bio gases in a representative patient before and immediately after cardiac arrest, during precordial compression, and after detibrillati (DF) and resuscitation. The transient increase in the ETco : after the administration of sodium bicarbonate (NaHCO,) is a demonstrated. The original tracing has been modihed because of space limitations. (From Falk JL, Rackow EC, Weil ΜΗ: Ει tidal carbon dioxide concentration during cardiopulmonary resuscitation. Ν Engl J Med 318:607-61 1. 1988. Used with permissio
TEAU. The expiratory plateau should detect mixed alveolar gas at the CO-, sampling site. During ini tial exhalation, elimination of gas from the ana tomic dead space (infinite V/Q, zero CO ; concen tration) is followed by elimination of gas from the well-ventilated, low-resistance regions of the lung (relatively high V/Q, low CO, concentration). Later, gas from poorly ventilated, high-resistance regions of the lung (relatively low V/Q, high CO-, concentration) is eliminated. The continuum of V/Q ratios between high and low V/Q areas cre ates a positive slope (upward to the right) to the alveolar plateau part of the C 0 2 elimination wave form, with the result that the end-tidal CO, con centration is the last and highest (the peak) con centration on the alveolar or the expiratory plateau. 55,56 In addition, continued production and evolution of C 0 2 into the alveolar space during exhalation contributes to the rise in C 0 : concentra tion during exhalation. Theoretically, the slope is such that the end-tidal value should be about 2 per cent higher than the time-weighted mean in normal resting subjects and about 4 per cent higher in exercising subjects, assuming tidal volumes are large enough to displace dead space.41 Because the alveolar plateau expresses the V/Q continuum in the lung, analysis of the alveolar
plateau may result in a wealth of diagnostic infc mation. First, the steepness of the alveolar plate is directly related to the degree of airway res: tance. Second, biphasic waveforms may reveal t presence of a two-compartment lung. Third, lea in the sampling system may alter the alveolar pi teau in a characteristic and, at first glance, peculi way. Fourth, the P . , C O - P E T C 0 gradient is c rectly related to alveolar dead space. Steepness of the Slope of the Alveolar Plates Is a Function of Expiratory Resistance. Whi the lungs of a patient without lung disease a being mechanically ventilated, the V/Q units in ti lung are relatively uniform and homogeneo (have the same C 0 2 concentration), and the exţ ratory plateau is smooth and nearly horizont; However, when there is significant lung disea and a large spread in V/Q ratios within the lun£ very well-ventilated, high-V/Q, low-CO : conce tration areas empty first, causing the alveolar pi teau to be relatively low. Following this, ve poorly ventilated, low-V/Q, high-CO : concentr tion areas empty, causing the alveolar plateau be relatively high. Thus, with a large spread in tl V/Q ratios, the upward positive slope of the alve lar plateau will be very steep to the right (it possible, but unusual, for there to be simultaneoi :
2
Monitoring
247
Figure 7-8 The phase III alveolar plateau slopes upward to the right because of a spread in ventilation-perfusion (V/Q) ratios from high (early phase III) to low (late phase III). Bronchospasm spreads the distribution of V/Q ratios and increases the slope of phase III. (Rairvi = airway resistance.)
emptying of alveoli with very different V/Q ratios and therefore a minimal slope to the expiratory plateau). Thus, it is not surprising that the increase in airway resistance that is associated with bronchospasm, which causes a large spread in the V/Q ratios and emptying times, to be tightly correlated with an increase in the slope of the alveolar plateau (Fig. 7-8).57 Biphasic Waveform in Patients Who Have Markedly Different Individual Lungs. Theoretically, if the continuum of the V/Q ratios is broken into two distinctly different lung regions (a lowresistance, high-V/Q, low-CO : concentration region and a high-resistance, low-V/Q, high-C0 2 concentration region), a biphasic C0 2 excretion waveform might be expected. Such a biphasic C0 2 waveform has been described in the patient who is in the lateral decubitus position, in which the nondependent lung has relatively low airway resistance, high V/Q ratio, and low C0 2 concentration compared with the dependent lung; in a patient with severe rotary kyphoscoliosis causing severe
Figure 7-9 Biphasic CO, excretion waveform during manual intermittent positive-pressure breathing. (From Nichols K, Benumof JL, Clausen J, Ozaki G: Expiratory CO : plateau slope predicts airway resistance. Anesthesiology 71:A1072, 1989. Used with permission.)
compression of one lung (Fig. 7-9) 58 ; and in a major main-stem bronchial intubation (Fig. 710).39 Some patients with COPD may also display a slight biphasic expiratory plateau if they have, throughout both lungs, two distinct populations of alveoli with very different time constants. In this situation, rapidly exchanging alveolar spaces are overinflated during inspiration (their compliance is high) so that their C0 2 concentration is low. whereas slower exchanging alveoli empty only during the latter part of exhalation, releasing a higher C 0 2 content.60 In patients with active expiratory efforts, a similar pattern may also be precipitated by airway closure because of increased intrapleural pressure during expiration.60 Finally, spontaneous breathing effort during a mechanical positive-pressure breath will create a cleft in the expiratory plateau and therefore a biphasic appearance. Sampling Line Leak Equals Very Unusual Waveform. When there is a leak in the sampling
248
Monitoring
Figure 7-10 Left panel (read right to left), Biphasic capnogram during right main-stem bronchial intubation. Each horizonta line indicates 10 mm Hg C0 2 concentration. The first C0 2 concentration peak ranged from 21 to 23 mm Hg, and the second CO) concentration peak ranged from 26 to 29 mm Hg. No spontaneous ventilation was apparent during this period of controlled ventilation. Right panel (read right to left), Normal appearing capnogram after the tip of the endotracheal tube was pulled baclj above the tracheal carina. Each horizontal line indicates 10 mm Hg C0 2 concentration: The end-tidal C0 2 concentration at the en 5 mm Hg produced a progressive positive discrepancy between P p a o and P la (i.e., P p a o > Pla) when the pulmonary arte rial catheter was wedged above the left atrium. In fact, when PEEP = 10 and 15 mm Hg, P pao equaled 10 and 15 mm Hg, respectively, whereas the Pu equaled 6 and 8 mm Hg, respectively.136 When the pulmonary artery catheter was wedged below the left atrium, it was not possible to create a P p a o -P| a gradient with even high levels of PEEP. A zone I catheter position may be suspected by observing the effects of sudden increases or de creases of PEEP on the measured P pao ; if the P p a o increases by greater than 50 per cent of the change in the applied PEEP (or CL suddenly increases), a non-zone III catheter tip is likely.137 (4). P p a o AND DETERMINATION OF ZONE III LO CATION. If the tip of the pulmonary artery catheter is vertically below or at the level of the left atrium, zone III conditions usually exist, unless extreme hypovolemia or high levels of PEEP are applied.137 Anteroposterior chest radiographs may not reliably locate the position of the tip of the pulmonary
With PEEP
Static Column of Fluid Catheter
Factors Causing Preservation of Ppao = Pla During PEEP 1. Location of catheter in dependent lung (Zone 3) 2. Maintenance of some spontaneous ventilation (venous return) 3. Non-compliant lung (decreased transmission of PEEP) Figure 7-21 In the clinical situation, zone l conditions are created by the application of positive end-expiratory pressure (PEEP). PEEP may collapse the pulmonary capillaries, interrupting the static column of fluid normally created by inflation of the pulmonary artery catheter balloon. Consequently, the pulmonary artery wedge pressure (P p a J will sense alveolar pressure (PEEP) rather than a downstream of vascular pressure (left atrial pressure [PJ). Several commonly present clinical conditions tend to preserve the Ρ ao as a reflection of Pla even when PEEP is being used; these consist of dependent-lung catheterization, presence of spontaneous ventilation, and noncompliance of the lung.
266
Monitoring
Monitoring
267
Figure 7-22 Schematic representation of pulmonary artery pressure tracings in the absence of respiratory effort (muscle paralysis), during spontaneous breathing and during controlled ventilation. The arrow indicates inflation of the balloon to obtain the pulmonary artery wedge pressure, and the asterisks represent the points of end expiration. (Reproduced with permission from Tobin MJ. Essentials of Critical Care Medicine. New York, Churchill Livingstone, 1989, ρ 37.)
c. PULMONARY ARTERIAL (TRUE) WEDGE PRESSURE (Ppdw, PULMONARY DISTAL WEDGE PRESSURE, SMALL PULMONARY VEIN PRESSURE) When the balloon on a pulmonary catheter is inflated, the segment of the pulmonary circulation that is occluded is the relatively large area sub tended by a large pulmonary artery, and flow will first begin in a relatively large pulmonary vein (Fig. 7-23, upper panel).151 Therefore, the pressure measured when the balloon on the pulmonary ar tery catheter is inflated is the pressure in a large vein (P p a o ; see prior discussion). When the tip of a pulmonary artery catheter is wedged in a small pulmonary artery, the segment of the pulmonary circulation that is occluded is subtended by a rela tively small pulmonary artery, and flow will first begin again in a relatively small pulmonary vein (see Fig. 7-23, bottom panel).151 Therefore, the pressure measured by the distal wedging of the pulmonary artery catheter tip is the pressure in a
small pulmonary vein (denoted by P p d w ). Ob viously, the pressure in a small pulmonary vein (Ppdw) will t>e slightly higher than the pressure in a large pulmonary vein (P pao ), and the P p d w -P p ; 1 0 gra dient will increase as venous resistance increases. The P p d w -P p a o gradient was 1.1 ± 0.5 mm Hg in nine patients with normal lungs and was signifi cantly higher in the 13 patients with chronic conges tive heart failure (3.8 ± 0.8 mm Hg, ρ < .01) and 22 patients with adult respiratory distress syn drome (3.8 ± 0.8 mm Hg; ρ < .01), but not in 20 patients with COPD (1.8 ± 0.7 mm Hg). 151 The distribution of the pulmonary vascular re sistance was, therefore, clearly different among the four groups.151 The fraction of the total pulmonary vascular resistance attributable to large and me dium pulmonary veins was significantly increased (p < .01) in adult respiratory distress syndrome (27.5 ± 12 per cent) and cardiac patients (27.5 ± 9 per cent) compared with patients with COPD ( 13 ± 5 per cent) and normal lungs (13.5 ± 6 per cent). The P p d w has been used clinically to assess
268
Monitoring
Monitoring
ual decline in pressure until the classic occluded pressure (P pao ) is attained (see Fig. 7-24). The fast decline is due to the diastolic pressure run-off in the pulmonary artery (to the P cap ), and the slow component of the pressure decay curve represents the discharging of the pressure/volume in the large capillary bed to the pressure/volume in the large pulmonary veins and left atrium. The inflection point or transition from fast to slow decline is thought to reflect the P cap , and the P c a p -P p a o gradient is thought to reflect postcapillary resistance (Fig. 7-24). Large gradients between estimated P c a p and P pa0 have been shown in patients with increased pulmonary venous resistance after mitral valve re placement.154 The P c a p may be suspected of being greater than the P p a o by simply observing an in creased difference between the P p a d and the P p a o (e.g., greater than 3 mm Hg), whereas a difference of 2 to 3 mm Hg means the P p a o more closely approximates the P cap . Although numerous mathematical/computer methods for determining the inflection point have been described, the visual technique has been thought to be adequate (see Fig. 7-24).'" The vis ual technique is best accomplished by placing a straight edge directly on the tracing, adjusted for the best fit, and marking the inflection point (i.e., when the curve clearly shows the slow component as being separate from the rapid component). The easiest and best determinations are made in me chanically ventilated, paralyzed patients (which eliminates respiratory artifact) with a pulmonary artery catheter in the proximal one third of the lung. Interobserver variability in determining Ρ using Electronic for Medicine recorder varied ± 0.2 mm Hg standard error of the mean. 1 " How ever, it should be noted that it still may be difficult to identify an unequivocal break in the pressure158 decay curve. e. PULMONARY ARTERY DIASTOLIC PRESSURE (Ppad)
Normally, pulmonary artery diastolic pressure equals left ventricular end-diastolic pressure. In Figure 7-25, pulmonary arterial (heavy line) and
Figure 7-24 Estimation of P c a p using the balloon pressure profile after inflation of pulmonary artery catheter (PAC) balloon (open arrow).
269
left ventricular (thin line) blood pressures are su perimposed on one another during three condi tions: (1) normal circumstance (A), (2) in the pres ence of left ventricular failure (B), and (3) in the presence of increased pulmonary vascular resis tance (C).m In the first two instances, both dia stolic blood pressures fall to the same level at the end of diastole (single arrow). The pulmonary hy pertension shown in Figure 1-25B is a passive consequence of the increased left ventricular enddiastolic pressure. The vessel walls remain thin and distensible under these conditions and do not impose an appreciable increase in resistance to blood flow. In contrast, the panel on the right illus trates that when pulmonary hypertension stems from abnormal structure or function of the pulmo nary vessels, pressure in the pulmonary artery will exceed the left ventricular pressure at the end of diastole, as indicated by the two arrows. Disease has rendered the pulmonary vessels sufficiently obstructed so that resistance to blood flow is greatly increased. The vertical distance between the two arrows, which represents the end-diastolic pressure gradient between the pulmonary artery and the left ventricle, is proportional to the in crease in pulmonary vascular resistance. In sum mary, when pulmonary vascular resistance is nor mal, pressure in the pulmonary artery during diastole can run off and decrease to left ventricular end-diastolic pressure (Fig. 7-26, left panel) (and blood flow will momentarily cease), whereas when pulmonary vascular resistance is increased, pres sure in the pulmonary artery during diastole cannot run off and decrease to left ventricular end-dia stolic pressure (see Fig. 7-26, right panel). Increased resistance to flow can be caused by both passive mechanical and active vasoconstrictor mechanisms. The passive mechanical mechanisms consist of capillary and venous compression by interstitial fluid and/or blood and/or fibrosis, en dothelial cell edema, capillary compression by PEEP, arterial obstruction by thrombosis and/or microembolism, and medial hypertrophy. The ac tive vasoconstriction mediators can be alveolar hypoxia, systemically released vasoconstrictor
270
Monitoring
Monitoring
amines and peptides, decreased mixed venous oxygen tension, and acidosis (Fig. 7-27).159 These changes result in an increase in pulmonary vascular resistance, which is "a universal feature of acute respiratory failure."160 Tachycardia (heart rate greater than 120/min) can also independently cause an end-diastolic pressure gradient (as much as 10 mm Hg) between the pulmonary artery and left ventricle because of insufficient diastolic run-off time.161 In another study, the left atrium was paced from 74/min to 124/min.162 As heart rate increased, Ppad increased but PLVED decreased, thus creating a significant pressure gradient (11 mm Hg). Therefore, this study also demonstrates that, with less diastolic filling time, another effect of tachycardia is that less blood is transferred from the pulmonary vasculature to the left side of the heart. Finally, another study showed that when the heart rate was above 115/min, Ppad was always greater than
271
Ppao·163 Thus, in patients with acute respiratory failure who have these pathologic changes and increased pulmonary vascular resistance, and often an associated tachycardia, it'is highly likely that pulmonary artery diastolic pressure will be higher than left ventricular end-diastolic pressure. Several studies showed that an increased PpadPpao gradient carries a very poor prognosis. In one study of a large series of critically ill patients, the Ppad w a s 6.0 mm Hg or more higher than the Ppao in 30 per cent of the patients.140 The pulmonary vascular resistance in these patients averaged (± standard deviation) 257 ± 145 dyne/sec/cm-5 (normal = 80 to 160 dyne/sec/cm~5). The mean pulmonary vascular resistance in the other 74 patients was slightly lower (158 ± 72 dyne/sec/cm-5). The mortality rate in the patients with the increased Ppad-Ppao gradients was 59 per cent. This was significantly higher than the mortality rate of 34 per cent seen with lower Ppad-Ppao gradients. In
Causes of Increased Pulmonary Vascular Resistance in Acute and Chronic Lung Disease
Figure 7-27 The causes of increased pulmonary vascular resistance during acute and chronic respiratory failure are multiple. In an anatomic progression from the arterial to the venous side of the pulmonary circulation, the causes of increased pulmonary vascular resistance consist of medial hypertrophy, endothelial cell edema, pulmonary thromboembolism, arteriolar constriction by vasoactive amines, peptides, decreased mixed venous oxygen tension ( [ P«02), acidosis, and alveolar hypoxia ( [ P A 0 2 ). Positive end-expiratory pressure (PEEP) can compress the pulmonary capillaries. Increased interstitial hydrostatic pressure resulting from transudated fluid and blood, which can later fibrose, can compress the venous side of the pulmonary capillaries. All these causes of increased pulmonary vascular resistance will create a pulmonary artery diastolic (Ppad) to left ventricular end-diastolic pressure (PLVED) gradient (see inset).
272
Monitoring
another study, 22 of 37 patients with sepsis had an increased P p a d -P p a o gradient. Patients in whom the P p a d exceeded the P p a o by more than 5.0 mm Hg had an average pulmonary vascular resistance ex ceeding 300 dyne/sec/cm -5 and had increased mortality rate. 164 Finally, another study noted that a persistent or increasing P p a d -P p a o gradient exceed ing 5 mm Hg was associated with a 91 per cent mortality rate. 165 Thus, if the P p a d exceeds the P p a o by 6.0 mm Hg or more, the patient has probably developed significant pulmonary hypertension and has a much poorer prognosis. A right bundle branch block (RBBB) also cre ates pressure gradients in the pulmonary vascula ture. 166 In the presence of RBBB (Fig. 7-28), 130 right ventricular systole is delayed, allowing the pulmonary pressure to continue to fall during the χ descent of the left atrial pressure trace. 130 The P p a d is then actually lower (may be up to 7 mm Hg lower166) than the mean P la . This will occur only when there is normal pulmonary vascular resis tance and thus rapid equilibrium between P p a d and Pu.
A pulmonary artery catheter should be floate to the most proximal position from which both tlj wedge pressure and phasic pulmonary artery prej sure can be obtained by simple sequential inflaţie and deflation of the flotation balloon. If the cathl ter floats out too far in the peripheral lung parei chyma, it may permanently wedge and greatly ii crease the risk of pulmonary infarction. If tl catheter is floated to a too proximal position in tl main pulmonary artery, the catheter tip may inte mittently whip or dip into and out of the rigl ventricle. Under these circumstances, the "dii stolic" pressure may be intermittently too low b cause several of the recorded beats will have a tually been right ventricular (diastolic pressure the right ventricle is close to zero). ι
3. Various Indices of Right Ventricular End-Diastolic Pressure If the right and left ventricles have parallel fun tion, then right ventricular end-diastolic pressu should follow left ventricular end-diastolic pre sure. The difference in absolute value between tl two end-diastolic pressures would be caused on by the difference in compliance between the tv ventricles; the right ventricle, being more cor pliant than the left ventricle, has a lower en diastolic pressure (see Normal Heart, which fc lows). With this reservation in mind, various inc ces of right ventricular end-diastolic pressure cl be used to assess preload clinically. a. RIGHT ATRIAL PRESSURE
Tricuspid and pulmonic valve disease will ci ate a right atrial to right ventricular end-diastoi pressure gradient. Tricuspid stenosis causes rig atrial pressure to be higher than right ventricul end-diastolic pressure, and pulmonic regurgitatii causes right atrial pressure to be lower than rig ventricular end-diastolic pressure.
b. CENTRAL VENOUS PRESSURE
J
Figure 7-28 A normal left atrial (LA) and pulmonary artery (PA) pressure tracing (A) is compared with tracings with right bundle branch block (RBBB) (B). The a wave corresponds to left atrial contraction, the c wave to the bulging of the mitral valve in ventricular systole, and the ν wave to the passive filling of the atrium. Mean LA pressure is read between a and c waves and should correspond to left ventricular end-diastolic pressure. The χ descent represents relaxation of the left atrium, the y descent the opening of the mitral valve. Note that the delayed right ventricular systole in RBBB is associated with a low P p a d (arrow). (Modified from Nadeau S, Noble WH: Mis interpretation of pressure measurements from the pulmonary artery catheter. Can Anaesth Soc J 33:352, 1986. Used with permission.)
To clearly answer the question of whether when central venous, as opposed to pulmona vascular, pressure monitoring can be used (P cv j P p a d and P pao ) to assess preload, it is necessary consider the answer separately for a normal he versus an abnormal heart. (1). NORMAL HEART. The left ventricle is thickly muscled (walled) cavity and therefore relatively noncompliant (Fig. 7-29, right-ha panel). When fluid is systemically infused (p load increased), the increase in left ventricu end-diastolic volume causes a large increase in 1 ventricular end-diastolic pressure (as measured Ppad or Ppao)· The right ventricle is a thinly muse (walled) cavity and therefore is relatively co
Monitoring
273
Pressure jure 7-29 The relationship between central venous pressure (P cv ) and pulmonary artery occluded pressure (Ppao) and pulmonary ery diastolic pressure (Ppad) is normally based on the differences in compliance between the right ventricle (RV) and the left ntricle (LV). The right ventricle is relatively compliant because it is thinly muscled, and the left ventricle is relatively noncomant because it is thickly muscled. Thus, when preload is changed in a normal heart (such as with fluid infusion), the central nous pressure (as a reflection of right ventricular end-diastolic pressure) increases only a small amount for a given increase in fdiac output, whereas the pulmonary artery occluded pressure and pulmonary artery diastolic pressure (as reflections of left ntricular end-diastolic pressure) increase a large amount for the same increase in cardiac output. Thus, because of the differences compliance between the two ventricles, the absolute central venous pressure is always less than the pulmonary artery occluded d pulmonary artery diastolic pressures, and the change in the central venous pressure is always less than the change in pulmonary ery occluded and pulmonary artery diastolic pressures.
iant (Fig. 7-29, left-hand panel). The infusion of lids systemically causes the increase in right intricular end-diastolic volume to be accompaed by only a relatively small increase in right intricular end-diastolic pressure (as measured by mtral venous pressure [Pcv]). Thus, when intraiscular volume is acutely increased, the Pcv ineases only a small amount compared with a relively large increase in the wedge pressure.167 deed, when preload is manipulated in normal itients (fluid infusion or diuresis168 or by position langes169) and simultaneous measurements of the L and P pao are made, the initial and final absolute μ is always approximately 100 per cent greater [in Pcv (Figs. 7-29 and 7-30). Figure 7-31, upper nel, shows the good predictability and high deee of correlation (lowest individual patient r > 72) in patients with ejection fractions greater an 0.5 undergoing position changes.169 In other
words, initial and final P pao is twice as great as the initial and final Pcv, and the change in P pao is al ways approximately twice the change in the Pcv (see Figs. 7-29 and 7-30). Thus, in a normal heart, there is an orderly and predictable relationship be tween the filling pressures of the right and left sides of the heart, and the Pcv can be used to follow left-heart function. However, because the regression line relating Pcv (x-axis) to Ppad and P pao (y-axis) in relatively normal patients has an average slope of 2 (which means P pao changes twice as much as the Pcv) (see Fig. 7-30),167-170 the P pao may be easier, and per haps more reliable, to use than the Pcv to follow intravascular volume status. In other words, it may be easier to distinguish physiologically meaningful changes above and beyond the usual background noise and variation in the P pao compared with the Pcv. In addition, it is not known how other changes
274
Monitoring
(e.g., afterload and changes in myocardial contractility) in normal patients will affect simultaneously measured Pcv and Ppad and Ppao.171 In summary, because the right and left sides of the heart have an orderly and predictable functional relationship to one another in the normal heart, the Pcv may be used as a substitute for pulmonary vascular pressure monitoring in normal patients, but the changes in Pcv will be less marked than those that might be observed in Ppao. (2). ABNORMAL HEART. In an abnormal heart, left- and right-heart ventricular function curves (filling pressure vs. output) may be markedly different from one another. For example, it is possible to have a failing right ventricle and a normal left ventricle, resulting in a high Pcv and low Ppao (which is an uncommon combination in the general surgical population but very common in the multiple-trauma patient,172 whereas with an ischemic left ventricle and a normal right ventricle (which is a common combination found in many cardiac patients), Ppa0 will be high while Pcv may be low (Fig. 7-32). Consequently, the discrepancy between the Pcv and the Ppad and Ppao increases a great deal when intravascular volume is augmented in patients with heart disease,169-171 and it is usually misleading to use the Pcv as a meaningful guide to the filling pressure of the left side of the heart. Figure 7-31, bottom panel, shows the unpredictability and extremely poor correlation (highest individual patient r < 0.53) between Pcv and Ppao in patients with ejection fraction less than 0.5 undergoing position changes.169 Thus, candidates for pulmonary vascular^ pressure monitoring include all patients who have any significant myocardial compromise who are undergoing significant perioperative stress.
Monitoring
275
Figure 7—31 Upper panel, Regression lines and correlation coefficients for the 15 patients in group I (ejection fraction > 0.5). For each patient, the correlation between P c v and Ρ was high; little dispersion of data is found around each individual regression line. Lower panel. Regression lines and correlation coefficients for the 15 patients in group II (ejection fraction < 0.5). The P c v and P p a o were uncorrected or poorly correlated for each group II patient, and the dispersion of data around individual regression lines was large. (From Mangano DT: Monitoring pulmonary arterial pressure in coronary artery disease. Anesthesiology 53:364370, 1980. Used with permission.)
276
Monitoring
Figure 7-32 Patients with abnormal hearts may have either a nom ι (LV) function curve (most common) (solid line) or an abnormal RV fun (dashed line).
ventricular (RV) and an abnormal left ventricular ve and a normal LV function curve (less common)
Monitoring
277
Clinical Value of Pulmonary Artery Catheter
Figure 7-33 The clinical value of the pulmonary artery catheter is great. Measurement of pulmonary artery diastolic (P pad ) and pulmonary artery wedge (Ppil()) pressures can be early indicators of left ventricular ischemia (increase in mean pressure and the appearance of "giant" a and cv waves). These two pressures allow estimation of left ventricular end-diastolic pressure (P L V ED)' which allows for assessment of intravascular volume, and based on this assessment decisions regarding preload changes (position, fluid infusion, or diuresis) can be made objectively. Measurement of the cardiac output (CO.) along with P p a o allows for estimation of myocardiac contractility and for an objective decision to be made as to whether inotropic or suppressant drugs should be used. Measurement of cardiac output along with systemic pressure allows for determination of systemic vascular resistance (SVR) and for an objective decision to be made as to whether vasodilator or vasoconstrictor drugs should be used. Measurement of SvO, may be used as an index of global well-being because it is a function of cardiac output, oxygen consumption (Vo : ), Hb concentration, and Ρ.Ο,.
278
Monitoring
ECG
CVP
Monitoring
279
GIANT cv WAVE
NORMAL
GIANT a WAVE Figure 7-35 Giant a waves are caused by atrial contraction into a noncompliant stiff ventricle (as may be caused by ventricular ischemia). Giant cv waves are actually giant c waves and are due to atrioventricular valve regurgitation when the supporting papillary muscles dysfunction (as may be caused by ventricular ischemia).
280
Monitoring
for assessment of myocardial contractility. Knowl edge of cardiac output along with systemic and pulmonary vascular pressures allows for calcula tion of systemic and pulmonary vascular resis tances, respectively. Knowledge of cardiac output along with heart rate allows for calculation of stroke volume. Measurement of P-02 (pulmonary artery blood) allows for calculation of Qs/Q„ and P-02 alone can be used as an indirect gross meas ure of cardiac output and oxygen consumption.174 (2). CARDIAC OUTPUT. The ability to obtain rap idly accurate measurements of cardiac output (CO), and, in conjunction with Ppad and Ppao, assess cardiac contractility in critically ill patients is one of the principal advantages of the pulmonary ar tery catheter. The temperature of pulmonary artery blood is instantly measured by a thermistor located near the distal tip of the pulmonary artery catheter. The injection of cold solution into the proximal (right atrial) lumen of the catheter allows the CO to be determined by the indicator-dilution tech nique. As right-sided CO increases, the indicator (cold solution) becomes more and more diluted by the warm venous blood, and less temperature drop (from body temperature) is detected by the pul monary artery thermistor. CO is usually calculated by a computer that measures the area under the thermodilution curve (obtained by plotting the de cline in pulmonary artery temperature from body temperature vs. time). Representative curves from patients with low, normal, and high COs are illus trated in Figure 7-36Λ, B, and C, respectively. ( 3 ) . SYSTEMIC
VASCULAR
RESISTANCE.
The
physiologic parameter known as systemic vascular resistance (SVR) is commonly calculated from the following relationship:
where P s a equals mean arterial pressure and Ρ equals right atrial pressure. The logical basis fc this expression arises by analogy from Poiseuille1 law for fluid flow within rigid pipes, which state that flow is proportional to the difference betwee the upstream and downstream pressures. Althoug CO, stroke volume, and stroke work are usuall indexed to body size, it is not yet common practic to do so for vascular resistance. Because the pi ripheral arterial tree branches in parallel and to a extent that is proportional to body size, large indi viduals have lower resistance (more vessels in paj allel) than smaller individuals. For example, a 1Κ lb, 62-inch female with P s a of 90 mm Hg, P c v of mm Hg, and CO of 4.2 L/min has a calculate SVR of 1617 dynes-sec-cm-5, whereas a 210-lt 72-inch male with the same P s a and P c v and a C( of 6.16 L/min (same cardiac index) has a calcu lated SVR of 1103 dynes-sec-cm-5. Calculatin SVR index using cardiac index instead of CO i the denominator normalizes the parameter to bod surface area and yields an SVR index of 242 dynes-sec-cm~5/m2 for both individuals, suggestin that their vascular trees were similarly matched t their CO. It seems that a method of indexing SV1 should be universally adopted into clinical prac tice." 9 ( 4 ) . MONITORING MIXED VENOUS OXYGEN SAT
In specific capillary beds, factors detei mined by local tissue metabolism induce releas of oxygen from hemoglobin (Hb) and produc unique, tissue-specific differences in arteriovenou oxygen content or saturation (the Fick principle^ For example, the arteriovenous oxygen conter difference in the normal working heart is quit large, usually 11.4 cc of oxygen per 100 ml α blood; in the skin, it is significantly smaller: 1 c of oxygen per 100 ml of blood. Because of wid URATION.
Monitoring
Variations in regional venous oxygen saturation, )lood from the pulmonary artery, where total mixng has occurred, represents the best average value or SA. For total body oxygen uptake, or (Vo2), the Fick )rinciple can be expressed mathematically as follows: Vo2 = Qt(Ca02 - C A ) The equation for the determination of S-O2 c a n De derived from the Fick equation as follows (the derivation involves substitution of S-O2 a n d Sa02 X [Hb][1.36] for C-02 and Ca02, respectively, factoring and transposing [Hb][1.36] and then transposing Sa02): SA = Sa02 - [Vo2/(Hb)(1.36)(Qt)l X 10 From this equation, we see that a decrease in SA can result from ( 1 ) a decrease in oxygen saturation of arterial blood, (2) an increase in oxygen consumption, (3) a decrease in cardiac output (Qt), and (4) a decrease in hemoglobin. When respiratory function, hemoglobin concentration, and oxygen consumption are stable, it follows from the Fick equation that changes in SA2 may accurately reflect parallel changes in cardiac output.175 This situation, a most common occurrence in the operating room, provides the basis for rapid correlation of SA2 with therapeutic pharmacologic interventions when cardiac output is low. Indeed, in patients who undergo coronary artery surgery, it has been shown that the probability of a decrease in SA2 being due to a reduction in cardiac output is 86 per cent.175 Even if the clinical situation is slightly more complex, the change in SA2 c a n usually be correctly interpreted. The correct interpretation of changes in SA2 m u s t be correlated with information about inspired oxygen concentration, changes in respiratory function, hemoglobin concentration, oxygen consumption, and cardiac index. In most clinical circumstances, the cause of SA2 changes can usually be inferred from other available laboratory data or astute clinical observation. For example, in hypovolemic shock, reductions in both hemoglobin and cardiac output may be additive factors in decreasing SA2· During myocardial ischemia, a reduction in cardiac index is most likely the cause of reduced SA2· However, reductions in arterial saturation secondary to pulmonary congestion may also be contributory. An arterial blood-gas sample or continuous monitor of arterial oxygen saturation, together with measurement of cardiac output, will usually establish the cause of reduced SA· Finally, the changes in SA2 during initiation of one-lung ventilation usually cause a large de«,
281
crease in Pa02 (a factor that decreases SA2) an d a moderate increase in Qt (a factor that increases SAX which results in a final small decrease in SA2·176·'77
If three of the four primary variables that determine SA2 (Qt» Vo2, Hb, Pa02) are not constant and/or secondary compensatory mechanisms kick in (e.g., interorgan and intraorgan redistribution of blood flow, right-shifted oxygen-hemoglobin dissociation curve), then changes in SA2 m a v °e uninterpretable. In fact, large changes in oxygen supply (Q„ Hb, Pa02) and demand (Vo2) may occur without any change in SA2·175 In t m s context, a decrease in SA2» which is probably not a desirable change, can only serve as an early warning device. For example, a large decrease in Q, may be compensated for by increase in F,02 and a decrease in Vo2, which is a circumstance that commonly occurs during general anesthesia. On emergence, the increased oxygen consumption and decreased arterial oxygenation resulting from the increased ventilation-perfusion mismatch associated with anesthesia and surgery may act in combination to decrease SA2 despite a notable increase in cardiac output. A relatively common experience after cardiopulmonary by-pass is that transfusion to a normal hemoglobin and hematocrit will increase SA2 considerably even as cardiac output decreases, sometimes to worrisome levels. Clinical situations in which changes in SA2 have been found to be uninterpretable involve critically ill patients in an ICU,178 postoperative cardiac surgical patients,179 patients in circulatory shock,180 patients undergoing aortic surgery with cross-clamping,181 and patients with acute myocardial infarction.182 There are two other situations/factors that limit the usefulness of SA2 monitoring. First, only pooled oxygen uptake and saturation values can be obtained, although the important information is whether oxygen supply is adequate to meet the needs of all organ systems, including those of the heart and brain. Therefore, SA2 monitoring provides little information on specific organ perfusion.182 In that sense, and in the case of complicated patients with many variables changing at the same time, decreases in SA2 (which are probably never a good sign) can serve at best only as a general early warning indicator. Second, diseases that uncouple total oxygen transport from oxygen consumption either by creating anatomic or physiologic peripheral shunts invalidate the role of SA2 m following the net oxygen supply and demand relationship. Renal patients with surgical arteriovenous fistulas commonly have elevated central and mixed venous oxygen saturations despite anemia and marginal arterial content. These patients clearly have
282
Monitoring
anatomic arteriovenous shunting. Patients with hypophosphatemia, nitroprusside toxicity, carboxyhemoglobinemia, cyanide poisoning, carbon monoxide poisoning, and methemoglobinemia have lesions that limit hemoglobin unloading at the tissue level. The combined anatomic and metabolic lesions of hepatitis, sepsis, major burns, pancreatitis, some drug-intoxicated states, and perhaps adult respiratory distress syndrome are poorly understood but widely known to be associated with elevated venous oxygen saturation despite clinical evidence for anaerobic metabolism (lactic acidosis).175 (5). TOTAL BENEFIT. The use of advanced technology such as pulmonary artery catheterization is based on the assumption that prompt recognition and therapy of hemodynamic abnormalities should improve outcome (i.e., the benefits outweigh the complications [see later discussion], and there is net benefit or the risk/benefit ratio is small); however, the correctness of this assumption has been debated.183-188 Nevertheless, pulmonary artery catheterization has been shown, by careful analysis
of the literature, to certainly be more accurate than clinical assessment alone in critically ill patients for determining the cause of shock (hypovolemic, cardiogenic, or septic) (Fig. 7-37, upper panel) and for assessing the cause of severe pulmonary edema (cardiogenic or noncardiogenic) (see Fig. 7-37; compare upper with lower panel).183·184· 189~191 In addition, it has been shown that the diagnosis of cardiac failure in medical or surgical patients with invasive hemodynamic monitoring provides physiologic data that guide pharmacologic treatment, which may favorably influence preload and afterload in the failing or ischemic heart.183·184·189_191 Figure 7-37 illustrates some of the obvious hemodynamic insights that pulmonary artery catheter monitoring can provide.183 The upper panel of the figure shows examples of typical and uncomplicated hemodynamic data for two common pathologic conditions (hypovolemia vs. cardiac failure caused by myocardial infarction). In both hemodynamic syndromes, or constellations, systemic pressure (Psa), central venous pressure (Pcv), heart rate (HR), CO, and SVR are identical or at least
DECISION ANALYSIS · HEMODYNAMIC DATA
DECISION ANALYSIS • HEMODYNAMIC DATA
Figure 7-37 Data from a pulmonary artery catheter can be the differentiating factor among clinical syndromes that otherw might appear hemodynamically similar. The shaded boxes provide the key hemodynamic data for diagnostic analysis and therapei decision making. See text for more complete explanation and definition of abbreviations.
Monitoring
uite similar, and the differential diagnosis beween the two conditions is not possible without neasurement of pulmonary vascular pressure. Vith simple hypovolemia, the heart is normal, and he P cv and P p a o correlate, whereas with myocardial nfarction the heart is abnormal and the P c v and 5 do not correlate. Consequently, the pulmonary pao — /ascular pressures are shaded in and are the key iifferentiating factors. Note that the correct diaglosis dictates therapy strategies that are quite different. Similarly, the lower panel of Figure 7-37 shows that measurement of P sa , P c v , and HR may not be dramatically different during sepsis and cardiac failure (because of very increased systemic afterload). However, measurement of CO and SVR clearly differentiates between the two conditions and would result in very different treatment strat egies.
— 5. Special Pulmonary Vascular Monitoring Considerations Related to Thoracotomy in the Lateral Decubitus _ Position __
—
_
~
-7-
_
I
Pulmonary arterial catheters usually (greater than 90 per cent) float to and locate in the right lung142; it is possible to increase the incidence of left-lung catheterization to 50 per cent using the right lateral decubitus position (which implies that balloon flotation vertically upward is as important as the pathway and curvature of blood flow.192 Consequently, during a right thoracotomy (left lat eral decubitus position), the pulmonary artery catheter will usually be in the nondependent right lung (if no special effort was made to place it in the left lung) and, therefore, either in a collapsed lung if one-lung ventilation is used or possibly in a zone 1 or 2 region of the lung if large tidal volume two-lung ventilation is used. Conversely, when a left thoracotomy is performed (patient in the right lateral decubitus position), the pulmonary artery catheter will be in the dependent lung and will probably be in a zone 3 region. Thus, it is theoretically possible that the pulmonary artery catheter might function differently or yield differ ent pulmonary vascular pressure and cardiac out put data during right versus left thoracotomies and during two-lung versus one-lung ventilation. Indeed, with the pulmonary artery catheter tip located in the right lung, the cardiac output is lower during right thoracotomy with one-lung ven tilation (right lung collapsed) than during left thoracotomy with one-lung ventilation (left lung col lapsed) in patients who were otherwise similar (Fig. 7-38Λ).193 Consequently, it is possible that when the pulmonary artery catheter is located in
283
the collapsed lung, where blood flow patterns may be distorted or the function of the thermistor inter fered with (not free in the lumen of the vessel), the measured output is indeed lower. This hypoth esis is supported by the concurrent finding that continuously measured mixed venous oxygen sat uration is also decreased during right thoracoto mies compared with left thoracotomies when the pulmonary artery catheter is located in the col lapsed nondependent lung. The decrease in mixed venous oxygen saturation may have been caused by stagnant blood flow and, therefore, not truly representative of whole patient mixed venous ox ygen saturation.193 When the nondependent lung is ventilated with varying levels of PEEP (in contrast to nondependent-lung collapse), there is no difference in the cardiac output measured simultaneously from thermistors located in the nondependent and de pendent lungs.194 This finding implies that, when the pulmonary artery catheter tip is in the nonde pendent lung and the nondependent lung is venti lated, blood flow to the nondependent lung is undistorted and/or there is no interference with the function of the thermistor. Indeed, in a conven tional two-lung ventilation unilateral lung injury model (hydrochloric acid aspiration) in sheep, all hemodynamic and respiratory measurements ob tained with a pulmonary artery catheter are accu rate and identical whether the catheter is located ipsilateral or contralateral to the injury.195 When the pulmonary artery catheter is in the nondependent lung and the nondependent lung is ventilated with large tidal volume, PEEP, or con tinuous positive airway pressure (CPAP), the wedge pressure may not reflect left atrial pressure (Fig. 7-38Z?).145 When the pulmonary artery cath eter is in the dependent lung and presumably in the zone 3 region, wedge pressure should accu rately reflect left atrial pressure even when PEEP 145 is applied to the dependent lung. In summary, the lateral decubitus position is important with regard to pulmonary artery catheter monitoring in three situations. First, when the nondependent lung is collapsed and the catheter is in the nondependent lung, the measured cardiac out put and P^02 may be decreased compared with more normal conditions or the "real" value. Sec ond, when the nondependent lung is ventilated with PEEP and the catheter is in the nondependent lung, P p a o may not equal P ]a . Third, when the cath eter is in the dependent lung, P p a o will be a faithful index of P la even if PEEP is used. In spontaneously ventilating critically ill patients, there is no signif icant difference in P p a o or cardiac output in the supine versus either lateral decubitus position.196
284
Monitoring
Conditions During Thoracotomy in Lateral Decubitus Position When Pulmonary Artery Catheter Data May be Inaccurate
Figure 7-38 Conditions during thoracotomy in the lateral decubitus position when pulmonary artery catheter data may be inaccurate. A, During right thoracotomy with a pulmonary artery (PA) catheter located in the collapsed right lung (one-lung ventilation [1LV]), the cardiac output (CO) may be lower than when the right lung is ventilated. The thermistor in the collapsed lung may be exposed to abnormal flow patterns or vascular wall interference. B, When the pulmonary artery catheter is in the nondependent lung and the nondependent lung is exposed to continuous positive airway pressure (CPAP) or positive end-expiratory pressure (PEEP), the pulmonary artery occluded pressure (Ppao) may be inaccurate. Nondependent lung CPAP or PEEP may cause zone 1 conditions in the nondependent lung. The Ppao is probably always reasonably accurate when the pulmonary artery catheter is in the dependent lung, even if the dependent lung is exposed to PEEP.
Monitoring
6. Risks and Complications of Pulmonary Vascular Pressure Monitoring Pulmonary vascular pressure monitoring is an invasive and sophisticated procedure and consequently involves numerous potential physical risks and requires intelligent and informed interpretation of the results. The physical risks must be known to make an objective risk/benefit ratio decision to insert a pulmonary artery catheter. Failure to understand the rationale for each index of PLVED (Pia)» PPao< PCap> Ppa> Pcv> as previously discussed, may cause the unwary and unknowing therapist to use the wrong pressures to arrive at ill-advised, if not contraindicated, therapeutic decisions. The physical risks are related to complications of gaining and maintaining central venous access and actually passing and floating the pulmonary artery catheter into the pulmonary circulation. This section briefly discusses the physical risks. a. COMPLICATIONS OF GAINING CENTRAL VENOUS ACCESS
The most obvious and often dramatic complication associated with pulmonary vascular pressure monitoring has to do with the fact that it is necessary to gain controlled access to the central venous circulation (Table 7-7). Frequently, the internal jugular, external jugular, and subclavian veins are used for this purpose. Because the needle that finds and identifies these neck/central veins (which have various advantages and disadvantages; Table 7-8) can inadvertently strike nearby vital organs, it is important to understand these potential complications.' 97 · ,98 First, and probably most common, the carotid and subclavian arteries lie next to and parallel to the corresponding internal jugular and subclavian veins, and the incidence of arterial puncture is site dependent (see Table 7-8) but overall is approximately 2.0 per cent.149 Because the risk of arterial puncture is least with the antecubital veins and it is difficult to tamponade the subclavian artery externally, the antecubital vein should be used and the subclavian vein avoided in patients with bleeding diatheses. The potential for vascular injury is greatly minimized if the finder needle is of small bore (approximately 22 gauge). Nerves can be damaged at multiple sites, but brachial plexus, stellate ganglion, and phrenic nerve injuries, in particular, have been reported. I have observed the withdrawal of cerebrospinal fluid on aspiration of an internal jugular vein finder needle. The pleural cavity can be invaded by the probing needle, causing either pneumo-, hemo-, or chylothorax. The incidence of pneumothorax has been
285
reported to range from 1 to 6 per cent with the subclavian vein approach to catheterization, but use of the internal jugular vein approach for insertion of the catheter has dramatically reduced this complication. The pericardium and vessels in the mediastinum can be invaded by the probing needle, causing cardiac tamponade. If the syringe used for venous blood aspiration is disengaged from the finder needle during a spontaneous inspiration, it is possible for environmental air to be sucked into the central venous circulation and result in air embolism.199 Air embolism has also been reported with the use of a pulmonary artery catheter-introducer kit that did not provide a self-sealing introducer port200 as well with the luminal dilators that pass through the selfsealing valves.201 It is estimated that, during 1 sec, 100 ml of air can flow through a 14-gauge needle with a 5-cm H 2 0 pressure drop across it,144 and 100 to 300 ml of air can be fatal. If this occurs, the source of air entry must be occluded, the F,0 : increased to 1.0 (to decrease the size of the embolus by causing resorption of nitrogen in the air). and the patient turned on the left side and to the Trendelenburg position, which facilitates moving an air bubble to the apex of the right ventricle (away from the right ventricular outlet) and increases CVP. Aspiration of air through a catheter in the right ventricle can be effective. Pulmonary artery catheters have been fractured in half202 and have been cut in two at the time of surgery.203 Catheter fragmentation is particularly likely to occur with a catheter through-needle device. In this case, retraction of the catheter may cause it to shear off at the tip of the needle. Thus, if a pulmonary artery catheter needs to be removed (e.g., ventricular arrhythmia has occurred) and the introducer needle is still in the vein, it is imperative to withdraw the catheter and the needle simultaneously. Application of a plastic drape over the patient's face for pulmonary artery catheterization while the patient receives oxygen, 3 L/min via nasal cannulas, is associated with a mild amount of C0 2 rebreathing (P,C0 2 = 6 ± 2 mm Hg; P E T C O , increased from 37 to 39 mm Hg and PaC(X increased from 41 to 43 mm Hg).204 b. COMPLICATIONS OF ACTUALLY PASSING AND FLOATING THE PULMONARY ARTERY CATHETER
Rupture of an inflated pulmonary artery catheter balloon has been reported (see Table 7-7), but no serious embolic sequelae from 1.5 ml of air have been noted.205 Rupture of the balloon of a pulmonary artery catheter is partly related to the duration of catheterization, because the balloon loses elasticity with exposure to blood. Rupture of the bal-
286
loon is detected by noting a decreased resistance when inflation is attempted. Also, blood may be aspirated from the balloon port. Arrhythmias are not uncommon during passage of the catheter through the right side of the heart (approximately 15 per cent of pulmonary artery catheter passages cause premature ventricular contractions), although persistent atrial arrhythmias,206 transient RBBB, complete heart block,207 and ven-
tricular fibrillation have occurred.208 In patients with pre-existing left bundle branch block, transient interruption of conduction via the right bundle branch may occur during insertion of a balloon flotation catheter, which may result in complete atrioventricular block and asystole. However, previous insertion of a pacing catheter or the use of a pacing-port pulmonary artery catheter can prevent this complication.183 The major risk factors for the
Table 7-7 CLINICALLY SIGNIFICANT COMPLICATIONS OF GAINING CENTRAL VENOUS ACCESS, INSERTING AND USING A PULMONARY ARTERY CATHETER, AND MAINTAINING CENTRAL VENOUS ACCESS* Pulmonary Catheterization Time/Event
Approximate Incidence Complication
(%)
Prevention
Treatment
Monitoring
287
Table 7-7 CLINICALLY SIGNIFICANT COMPLICATIONS OF GAINING CENTRAL VENOUS ACCESS, INSERTING AND USING A PULMONARY ARTERY CATHETER, AND MAINTAINING CENTRAL VENOUS ACCESS* Continued Pulmonary Catheterization Time/Event
Complication Pulmonary artery rupture
Maintaining central venous access
Thrombosis
Sepsis secondary to catheter
Abbreviations: PVC airway pressure.
Approximate Incidence (%) 0.2
>50% by venography
0-1
Prevention
Treatment
Use pulmonary artery Pull back catheter 1-2 cm. catheter position inflate balloon until Ppa monitoring catheter (right trace becomes damped, ventricular port), avoid single-lumen tube/positivedistal migration (pull pressure ventilation, pulmonary artery catheter double-lumen tube/high tip back to main CPAP to bleeding lung, pulmonary after each Ppao lobectomy, transfusion, determination, especially if systemic support during cardiopulmonary by-pass), minimize inflation time, avoid hyperinflation of balloon, inflate slowly with continuous waveform monitoring Increase cardiac output and blood pressure, use heparin-bonded catheters, minimize indwelling/ catheterization time Sterile technique, change catheters, minimize catheterization time
None
See chapters 18, 19; remove catheter; administer antibiotics, systemic support
premature ventricular contractions; RBBB = right bundle branch block; CPAP = continuous positive
development of arrhythmias include myocardial ischemia, hypoxemia, electrolyte disorders, and acidosis.'49 Prolonged catheterization time is another risk factor. Although this may indicate a less skillful operator, longer catheterization time may also reflect increased difficulty of catheterization in patients with severe shock, large right ventricles, dilated pulmonary arteries, and marked pulmonary hypertension.200 Occasionally, it may be difficult to pass the pulmonary artery catheter out of the right ventricle, especially if the right ventricle is dilated (a circumstance that also promotes premature ventricle contractions). The head-up and right lateral tilt position appears superior to Trendelenburg's position
for passage of pulmonary artery catheter in the awake patient209 and the anesthetized patient210; this may be due to the fact that balloon flotation (tendency of the balloon to seek a nondependent position) is more important than blood flow.192 Very occasionally, a catheter may not pass out of the right atrium; tricuspid regurgitation creates a whirlpool type of blood flow in the right atrium, denying entry of the air-inflated balloon into the right ventricle.2" Catheters have knotted on themselves,212· 2I3 with other indwelling catheters,214 and around papillary muscles.215-216 A number of nonsurgical techniques for removing knotted catheters have been described.212-218 Passage of pulmonary artery catheters has
Table 7-8 COMPLICATIONS OF CENTRAL VENOUS CATHETERIZATION IN RELATION TO SITE*
*From Tobin M: Pulmonary artery catheter problems. Appl Cardiopulm Pathophysiol 3:279-285, 1990. Used with permission.
288
Monitoring
caused damage to the tricuspid219 and pulmonary valves220 as well as endothelial trauma along the entire catheter path.200 Catheters have permanently wedged in distal pulmonary vessels with subsequent pulmonary infarction in the distribution of the vessel containing | the catheter.221·222 Infarction is most likely to occur within 12 to 24 hours of insertion. The risk of infarction can be decreased by careful monitoring of the vascular waveform and probably also by the use of continuous flush with heparin solution. Pulmonary artery catheters have also perforated pulmonary vessels,223 causing massive pulmonary hemorrhage224,225 and hemoptysis,226 and have perforated the right ventricle, causing pericardial tamponade.197 The mortality rate after rupture of a pulmonary artery by a pulmonary catheter is greater than 50 per cent and is still higher in heparinized patients. Associated factors include pulmonary artery hypertension, advanced age, fragility of tissues, peripheral catheter-tip locations, anticoagulation (which is unlikely to predispose to rupture but probably contributes to morbidity once it occurs), and cardiopulmonary by-pass. Retraction of the heart, especially to expose distal circumflex vessels, may push the catheter peripherally, and hypothermia stiffens the catheter, increasing the risk of perforation. Although the exact mechanisms of catheter-associated pulmonary artery rupture have not been identified, the common factor in most reported cases appears to be the location of the catheter tip in distal pulmonary artery branches outside the mediastinal shadow on the chest X-ray. Smaller branches may be lacerated by spearing of the vessel wall by the catheter tip propelled by contractions of the heart or by overdistention of the vessel at the time of balloon inflation. Patients typically present with hemoptysis, which may occur up to 20 hours after inserting the catheter. In about one third of patients, rupture and hemoptysis occur almost immediately after catheter introduction, and typically there is a chest Xray infiltrate near the tip of the catheter. Several important preventive procedures can be used. First and foremost, with respect to distal migration of the catheter tip, phasic Ppa waveform should always be displayed. There is now available a pulmonary artery catheter position-monitoring catheter (PA Watch Catheter; Baxter Healthcare, Santa Ana, CA), which has a right ventricular port 10 cm from the tip of the pulmonary artery catheter.227·228 This catheter is considered to be in proper position when the middle-lumen port, which is located 10 cm from the tip, transmits a right ventricular pressure waveform. Using this pulmonary artery catheter, it is very clear that pulmonary artery catheters migrate distally (right ven-
tricular port enters pulmonary artery) very frequently (i.e., 50 227 -100 per cent228 of patient^ Withdrawal distances (to get the right ventricul • port out of the pulmonary artery) ranged from 1 io 6 cm.227 Thus, this new catheter may add a margin of safety to pulmonary artery monitoring a lower its overall cost by eliminating the need L·* chest radiographs ordered solely to confirm catheter-tip location. Second, balloon hyperinflation should I avoided. The balloon should always be inflatec slowly, with continuous waveform monitori· (Ppao position should not be detected by palpatij of the pilot balloon).229 Third, relocating the catheter tip in the proxirr** pulmonary artery (especially during cardiopulm nary by-pass) and readvancing it for each wedgT pressure determination significantly reduces Û* risk of catheter-induced pulmonary hemorrhage. Management of a ruptured pulmonary arter should be progressively aggressive as necessan The pulmonary artery catheter should be pull» back 1 to 2 cm and the balloon inflated until tkphasic Ppa trace dampens; this partial unilater; occlusion will greatly decrease the blood flow the bleeding area; infusion of epinephrine throu[__ the distal port may also be helpful. AnticoaguL tion should be reversed if possible. To avoid co tamination of the nonbleeding lung, the patie_ may need to be placed in a decubitus position wi the affected side down. Tracheal toilet and fibe optic bronchoscopy (perhaps with placement of bronchial blocker) should follow. If respirato~ distress transpires, endotracheal tube intubatl· and positive-pressure ventilation are indicate Double-lumen tube intubation is indicated if t bleeding is massive or if bilateral contamination a real threat. With double-lumen tube intubatk differential lung ventilation is possible, and 1 bleeding lung may be tamponaded and held s from the airway side by high (40-50 cm H2 CPAP (but not ventilated). With all these man< vers, systemic support should be provided as n< essary (transfusion, vasopressure, and so on), nally, if all else fails to control the bleedi surgical resection may be necessary. c. COMPLICATIONS OF MAINTAINING CENTRAL VENOUS ACCESS
Thrombophlebitis of the veins in question is ways possible (see Table 7-7), and indwelling ν cular catheters and sheath introducers can alw serve as a source of infection and sepsis.205·231 Thrombi commonly form on the surface of \ monary artery catheters soon after their insert as a result of an interaction between blood and physiochemical properties of the catheter's suri as well as secondary to catheter-induced endoi
Monitoring
Hal damage (including heart valve damage, predominantly the pulmonary valve). In addition to thrombus formation on the catheter, venous thrombosis has been found by venography in two thirds of patients with internal jugular vein catheters.217 Such catheter-related thrombosis has been shown to be associated with an increased risk of proximal pulmonary artery thromboembolic38 Thrombotic complications appear to have been reduced by heparin bonding of catheters and by minimizing the time the catheter is indwelling.239·240 Catheter-related infection is thought to arise in the fibrin sheath that forms around the catheter within 24 to 48 hours of insertion.236 Most cannula-related infections arise from organisms that migrate from the patient's skin into the cannula tract. These originate from the patient's own skin flora or from the hands of medical personnel. The tip may also be colonized hematogenously by organisms from other sites. Prevention and treatment of sepsis are standard and are discussed to some extent in chapter 19. These complications of central venous access and catheter flotation are admittedly rare events; therefore, many of the references to these complications are case reports (which involve only one or a few patients) rather than studies of series of patients. Not surprisingly, therefore, one large series of 1400 patients concluded that, taking everything into consideration, pulmonary artery catheterization is associated with a very acceptable low incidence of morbidity and mortality.198 Nevertheless, to make an intelligent judgment about whether a pulmonary artery catheter should be used in a given patient, it is necessary to enter this information into the risk-benefit equation or balance.
7. Transesophageal Echocardiography Although transesophageal echocardiography (TEE) can reveal information about atrial and ventricular septal, valvular, and aortic function, intracardiac masses, and ventricular volumes, the main purpose of intraoperative TEE for noncardiac thoracic surgery is to monitor ventricular ischemia. The main contraindications to the use of TEE involve esophageal diseases such as strictures, varices, scleroderma, esophagitis, upper gastrointestinal bleeding, dysphagia, history of esophageal surgery, and chest wall radiation therapy.241 Complications from the insertion and use of the TEE probe are not a common occurrence (~1 per cent) and consist of arrhythmias, bronchospasm, unsuccessful insertion, and esophageal perforation (0.02 per cent).241 At present, the standard TEE images are all transverse (cross sectional). There are four major
289
transverse views: the basal short-axis view (Fig. 7-39A), frontal long-axis or four-chamber view (Fig. 7-395), the short-axis view of the left ventricle at midpapillary region (Fig. 7-39C), and the long-axis view of the descending thoracic aorta (Fig. 7-39D).242 Semiquantitative and quantitative tools like color-flow mapping (CFM) and Doppler echocardiography have further enhanced the utility of TEE. CFM simultaneously presents real-time images of intracardiac blood flow and structure in two dimensions. The blood flow going toward the transducer is depicted in various hues of red and away from the transducer in various hues of blue. CFM facilitates evaluation of valvular and congenital cardiac lesions. Doppler echocardiography uses the Doppler principle; in valvular heart disease, by accurately measuring the flow velocity by Doppler shift, the valve cross-sectional area and the regurgitant flow can be reliably estimated.24. The identification of regional wall abnormalities and their association with coronary artery ischemia has played a major role in the development of TEE. Ischemia may be supply ischemia, demand ischemia, or mixed ischemia. The earliest sign of ischemia is impaired ventricular relaxation. This sign is followed (in order of decreasing sensitivity ) by regional wall-motion abnormalities, impaired global systolic function, ventricular pressure-volume (compliance) changes, electrocardiographic changes of the ST segment, and only then by chest pain. In one study of 98 patients anesthetized for coronary artery bypass surgery, myocardial ischemia was identified by TEE (wall-motion abnormalities) in 14 patients; in 10 of these, it was associated with concomitant ST-segment depression of at least 1 mm. The onset of ischemia, as defined by TEE, was accompanied by a mean increase in pulmonary capillary wedge pressure of 3.5 ± 4.8 mm Hg compared with a mean change of 0 ± 2.2 mm Hg between observations not associated with the onset of ischemia (p < .01). An increase in Ppao of at least 3 mm Hg tested as an indicator of ischemia, but the sensitivity of this indicator was only 33 per cent and its positive predictive value was only 16 per cent.121 The left ventricle short-axis view at the level oï the papillary muscles is used to determine ischemia (this area represents perfusion by all three coronary vessels), and the ventricular wall may be divided into four segments (Fig. 7-40). I21 Each myocardial segment (anterior, posterior, septal, lateral) may be examined separately. Normal segmental contraction of the heart may be defined as shortening of the radius of the left ventricle by more than 30 per cent with wall thickening. One of the representative scoring schemes for regional wall-motion abnormalities defines wall-motion ab-
290
Monitoring
Basal short axis
Basal 4-chamber
Figure 7-39 Schematic representa tion of different transesophageal echo cardiography (TEE) scan planes and the corresponding images. The Ρ at the apex of the sector indicates that the transducer lies posteriorly to the imaged structure. A, Basal short-axis views: (1) main pul monary artery, (2) pulmonic valve, (3) aortic root and coronanes, (4) aortic valve. B, Basal four-chamber views: (5) left ventricle (LV) outflow tract, (6) clas sic four chamber, (7) coronary sinus.
Monitoring
LV short axis
Figure 7-39 Continued C, LV short-axis views: (8) mitral valve, (9) midpapillary muscle, (10) apical. D, LV long-axis views: (11) apical, (12) obtuse angle, (13) short-axis view of the descending thoracic aorta. A = anterior; Ao = aorta; AL = anterolateral papillary muscle; CS = coronary sinus: FO = fossa ovalis; IVC = inferior vena cava; LA = left atrium; LAA = left atrial appendage; LCA = left coronary artery; MPA = main pulmonary artery; PM = posteromedial papillary muscle; PV = pulmonic valve or pulmonary vein; RA = right atrium; RAA = right atrial appendage; RCA = right coronary artery; RLPV = right lower pulmonary vein; RPA = right pulmonary artery; RUPV = right upper pulmonary vein; RV = right ventricle; SVC = superior vena cava. (From Thys DM, Hillel Z: Echocardiography. In Benumof JL (ed): Clinical Procedures in Anesthesia and Intensive Care. Philadelphia, J. B. Lippincott, 1992, pp 459^196. Used with permission.)
LV long axis and descending aorta
291
292
Monitoring
Figure 7—40 Schematic representation of echocardiographic short-axis view of left ventricle at the level of papillary mus cles, indicating division into four segments of myocardium used for grading wall motion per segment (see text). A = anterior; S = septal; L = lateral; Ρ = posterior; LV = left ventricle; RV = right ventricle. (From Van Daele MERM, Sutherland GR, Mitchell MM, et al: Do changes in pulmonary capillary wedge pressure adequately reflect myocardial ische mia during anesthesia? Circulation 81:865-871, 1990. Used with permission.)
normalities as follows: mild hypokinesis, evi denced by shortening of the radius of the ventricle by less than 30 per cent but with more than 10 per cent wall thickening; severe hypokinesis, evi denced by less than 10 per cent radial shortening and minimal wall thickening; akinetic segment, evidenced by no wall thickening during systole; and, finally, a dyskinetic segment, evidenced by left ventricular wall bulging and thinning during systole. Regional wall-motion abnormalities can occur with ischemia, but it is important to remem ber that they can also occur when there is prior myocardial infarction with tissue fibrosis, an intra ventricular conduction abnormality caused by bun dle branch block, premature ventricular contrac tions, and ventricular pacing or during thoracotomy because of shifting of cardiac structures.241 Thus, onset of new wall-motion abnormality in patients with baseline impairment of systolic left ventricu lar function may be difficult to detect.243 In human subjects undergoing coronary angio plasty, visible wall-motion abnormality may occur within 10 to 15 sec of coronary occlusion.244 Onset of hypokinesia occurred within 15 sec of balloon occlusion.244 Hypokinesia develops at 50 per cent reduction of coronary blood flow, whereas, with 90 per cent reduction in flow, dyskinesia takes place. 243 New wall-motion abnormalities observed intraoperatively"may be reversible when associated with a short period of ischemia or may be irrevers ible when associated with infarction.243 TEE is valuable in the diagnosis of aortic dis
section; in one study, the sensitivity was 99 per cent and the specificity 98 per cent. 245 TEE is very sensitive in detecting as little as 0.01 ml/kg of air (microbubbles) and is therefore extremely sensi tive in detecting a patent foramen ovale and intra cardiac air. Valsalva and cough maneuvers are used to augment a right-to-left shunt in patients to rule out a patent foramen ovale with contrast and Doppler echocardiography. TEE is also helpful in patients suspected of having an intracardiac mass (thrombus, tumor, and/or vegetations), although the sensitivity and specificity are not nearly as good as for aortic dissections and intracardiac air. TEE is very beneficial in visualizing cardiac val vular function. TEE can determine whether left ventricular end-diastolic volume is very high or very low but is moderately imprecise for follow ing, on-line, changes in left ventricular end-dia stolic volume of intermediate values.
8. Lung Water Measurements The preceding pulmonary artery catheter meas urements greatly increase the understanding of how the respiratory and cardiovascular systems work together. Recently, with a combination of these techniques and methodology, extravascular lung water measurements have become available for clinical use (American Edwards Laboratories 9310 Computer). Extravascular lung water is measured using a thermal change-green dye con centration change double-indicator dilution tech nique. The two indicators are injected simultane ously into the central venous circulation and are detected by a thermistor-tipped arterial catheter. One indicator, indocyanine green dye, binds to serum albumin and remains intravascular as it passes through the lung. The other indicator, a cold bolus of dextrose solution, diffuses through out the extravascular space at the same time. Time-related dilution curves are determined and analyzed by the computer. Cardiac output and mean transient time are automatically calculated for each indicator. The product of cardiac output and mean transit time for a given indicator yields the volume distribution for that indicator. The dif ference between the volume distribution of the two indicators represents the volume of extravascular water in the lungs. The correlation between the double-indicator dilution technique and gravimétries, the accepted standard, has been good (r = 0.96),246-249 and it has been possible to perform multiple determinations rapidly and reproducibly. However, it is important to note that the correlation between extravascular lung water measurements and other observable parameters of respiratory function and the effect of various therapeutic modalities have not yet been established.
Monitoring
9. Computed Tomographic Pulmonary Scan Computed tomographic (CT) scanning may prove to be a new useful demonstration/monitor of lung disease.250 One study of adult respiratory distress syndrome patients showed that pulmonary pressure increases in concert with lung weight and that, as the mass of noninflated lung tissue increases, so do venous admixture and V[:/VT, while P a 0 2 decreases.251 Without the CT scanner, no such correlations would have been possible, because there are no other means for measuring reliably at the bedside either lung weight or noninflated lung tissue mass. Earlier inhalation and indicator dilution techniques that measured extravascular lung water (see section IV.G.8.) were restricted to assessing those areas within the reach of inhaled or perfused agents, whereas X-rays are not. The correlation in early adult respiratory distress syndrome of increased pulmonary artery pressure with increased lung weight as estimated by CT corroborates our current understanding of fluid and solute transport in the injured lung. It is expected that increasing Pp;i (caused by hypoxic and mediator-included vasoconstriction, thrombosis, and obliteration of pulmonary microvasculature; see Fig. 7-27) in the presence of lung injury will cause the lung to gain weight by directly forcing plasma into the interstitial and alveolar spaces. As in other studies, there was a marked posterior distribution of regions of high density probably resulting from the effect of gravity, causing airway and interstitial fluid to migrate to dependent lung regions, thereby filling airspaces with fluid and promoting consolidation. PEEP augmented the recruitment of lung volume during adult respiratory distress syndrome by increasing low-density, presumably ventilated lung regions at the expense of high-density, presumably nonventilated (shunted) regions.
REFERENCES 1. Cooper JB, Newbower BS, Kitz RJ: An analysis of major errors. Anesthesiology 60:34-42, 1984. 2. Paulus DA, Basta JW, Klie H, Radson EA: Preanesthetic checklist. Anesth Analg 64:264, 1985. 3. Debban DG, Bedford RF: Overdistension of the rebreathing bag: A hazardous test for circle system integrity. Anesthesiology 42:365-366, 1975. 4. Ward CS: The prevention of accidents associated with anesthetic apparatus. Br J Anaesth 40:692-701, 1968. 5. Westenskow DR, Jordan WS, Jordan R, Gillmore ST: Evaluation of oxygen monitors for use during anesthesia. Anesth Analg 60:53-56, 1981. 6. Meyer RM: A case for monitoring oxygen in the expiratory limb of the circle. Anesthesiology 61:347, 1984. 7. Lee E: Of stethoscopes and stopcocks. Anesth Analg 73:98-99, 1991.
293
8. Schwartz N: Monitoring bilateral breath sounds. Anesthesiology 66:711-712, 1987. 9. Scheller M, Jones B, Benumof JL: Effect of changes in I:E ratio and flow rate with constant tidal volume on minute ventilation and PaCOo. J Cardiothorac Anesthesiol 3:564-567, 1989. 10. Saklad M, Paliotta J, Weyerhauser A: On line monitoring of ventilatory parameters. Clin Anesth 9:335-362. 1973. 11. Egbert LD, Biano D: The educated hand of the anesthesiologist. Anesth Analg 46:195-200, 1967. 12. Silverman MS, Bishop MJ: Manual vs mechanical ventilation in a model of bronchospasm: Does the "educated hand" exist? Anesthesiology 71:A441. 1989. 13. Knill RL: Secondary prevention of hypoxemia. Anesthesiology 69:438-439, 1988. 14. Knill RL: Evaluation of arterial oxygenation during anaesthesia. Can Anaesth Soc J 32:S16-S19, 1985. 15. Comroe JH Jr, Botelho S: The unreliability of cyanosis in the recognition of arterial anoxemia. Am J Med Sci 214:1-6, 1947. 16. Medd WE, French EB, Wylie VM: Cyanosis as a guide to arterial oxygen desaturation. Thorax 14:247-250. 1959. 17. Kelman GR, Nunn JF: Clinical recognition of hypoxaemia under fluorescent lamps. Lancet 1:1400, 1966. 18. Coté C, Goldstein E. Coté M, Hoaglin DC, Ryan J: A single blind study of pulse oximetry in children. Anesthesiology 68:184-188, 1988. 19. Manninen PH, Knill RL: Cardiovascular signs of acute hypoxaemia and hypercarbia during enflurane and halothane anaesthesia in man. Can Anaesth Soc J 26:282287, 1979. 20. Knill RL, Gelb AW: Ventilatory responses to hypoxia and hypercapnia during halothane sedation and anesthesia in man. Anesthesiology 40:244-251, 1978. 21. Knill RL, Kieraszewicz HT, Dodgson BG: Chemical regulation of ventilation during isoflurane sedation and anaesthesia in humans. Can Anaesth Soc J 30:607-614. 1983. 22. Landsgaard C, Van Slyke DD: Cyanosis. Baltimore, Williams & WilkinsCo, 1923. 23. Alexander CM. Teller LE, Gross JB: Principles of pulse oximetry: Theoretical and practical considerations. Anesth Analg 68:368-376, 1989. 24. Jobes DR, Nicolson SC: Monitoring of arterial hemoglobin oxygen saturation using a tongue sensor. Anesth Analg 67:186-188, 1988. 25. Coté CJ, Daniels AL, Connolly M. Szyfelbein SK, Wickens CD: Tongue oximetry in children with extensive thermal injury: Comparison with peripheral oximetry. Can J Anaesth 39:454^157, 1992. 26. Ezri T, Lurie S. Konichezky S, Soroker D: Pulse oximetry from the nasal septum. J Clin Anesthesiol 3:447-450. 1991. 27. O'Leary RL, Landon M, Benumof JL: Buccal S P 0 : is more accurate than finger S P 0,. Anesth Analg 75:495498, 1992. 28. Kagle DM, Alexander CM, Berko RS, Guiffre M, Gross JB: Evaluation of the Ohmeda Biox 3700 pulse oximeter: Steady state and transient response characteristics. Anesthesiology 66:376-380, 1987. 29. Severinghaus JW, Naifeh KH: Accuracy of response of six pulse oximeters to profound hypoxia. Anesthesiology 67:551-558, 1987. 30. Broome IJ, Harris RW, Reilly CS: The response times during anaesthesia of pulse oximeters measuring oxygen saturations during hypoxaemic events. Anaesthesia 47:17-19, 1992. 31. Guggenberger H, Lenz G, Federle R: Early detection of inadvertent oesophageal intubation: Pulse oximetry vs
CHAPTER
8
Choice of Anesthetic Drugs and Techniques I. Introduction II. Most Common and Important Cardiopulmonary Considerations for Patients Undergoing Thoracic Surgery A. Pulmonary Considerations (TwoLung Ventilation) 1. Reactive Airways a. Effect of Anesthetic Drugs on Airway Reactivity 2. Gas-Exchange Impairment B. Cardiac Considerations 1. Coronary Artery Disease a. Effect of Anesthetic Drugs on Cardiovascular Function III. Choice of Anesthesia and Arterial Oxygenation During One-Lung Ventilation A. Effect of Anesthetics on Hypoxic Pulmonary Vasoconstriction B. Effect of Anesthetics on Arterial Oxygenation During One-Lung Ventilation 1. Two-Lung Ventilation: Blood Flow Distribution
300
2. One-Lung Ventilation: Blood Flow Distribution, Shunt Flow, and Arterial Oxygen Tension 3. Effect of Isoflurane on the OneLung Ventilation Blood Flow Distribution, Shunt Flow, and Arterial Oxygen Tension IV. Recommended Anesthesia Induction and Maintenance Drugs and Techniques A. Summary of Advantages of Anesthetic Drugs 1. Inhalational Anesthetics 2. Intravenous Anesthetics B. Recommended Anesthetic Drugs and Technique 1. Insertion of an Epidural Catheter Before Induction of Anesthesia 2. Induction of Anesthesia 3. Maintenance of Anesthesia 4. Total Intravenous Anesthesia Technique
Choice of Anesthetic Drugs and Techniques
I. INTRODUCTION The choice of anesthetic drug and technique in the vast majority of thoracic surgery cases is based on the preoperative cardiopulmonary evaluation. Some drugs and techniques may favor the overall perioperative function of one organ at the expense of another. For example, the halogenated drugs may prevent or minimize bronchospasm but at the same time may decrease myocardial contractility, whereas the narcotics may preserve myocardial contractility but may not prevent bronchospasm in patients with reactive airways and may cause postoperative respiratory depression. Because different patients will have varying degrees of dysfunction of different organs and anesthetic drugs differentially affect the various organs, the most appropriate anesthetic will depend on the patient. This chapter first briefly considers the usual and major pulmonary and cardiac problems that patients undergoing thoracic surgery may have (with both lungs ventilated) as well as the most important effect anesthetic drugs might have on these problems. Next, the specific effect of anesthetic drug and technique on gas exchange for the special one-lung ventilation situation (in particular, on hypoxic pulmonary vasoconstriction [HPV]) is covered. Finally, the chapter recommends anesthetic drugs and techniques that should simultaneously minimize the pulmonary and cardiac problems but yet incorporate enough flexibility to emphasize the function of one organ over another, should that be considered necessary. As with most of the other chapters in this book, respiratory considerations are emphasized more than cardiac considerations.
II. MOST COMMON AND IMPORTANT CARDIOPULMONARY CONSIDERATIONS FOR PATIENTS UNDERGOING THORACIC SURGERY A. Pulmonary Considerations (TwoLung Ventilation) 1. Reactive Airways The mechanisms by which increases in airway resistance may be stimulated have been summarized' and include mechanical and chemical mucosal stimulation (causing various neural reflexes that are mediated by medullary centers, local arcs, and the autonomic nervous system), anaphylactoid bronchoconstriction, histamine-induced bronchoconstriction, alpha-adrenergic predominance, vagal (cholinergic) predominance, and exercise-induced bronchoconstriction. In this discussion, the
301
term "reactive airways" applies to patients who come to surgery with pre-existing bronchospasm or will likely react to the mechanical stimulation of endotracheal tube insertion and/or tracheobronchial tree manipulation with bronchospasm. Patients undergoing thoracic surgery are likely to have an increased incidence of reactive airways for two reasons. First, the vast majority of patients have a long and significant smoking history and therefore have varying degrees of chronic obstructive pulmonary disease, excess secretions, and increased reversible airway resistance (bronchoconstriction). The presence of the increased airway resistance can often be readily demonstrated by improvement in expiratory airflow rates following the preoperative administration of bronchodilating drugs. Indeed, there is a strong dose-response relation between the amount of secretions, cough, bronchoconstriction, severity of chronic obstructive pulmonary disease, and the risk of mortality from chronic obstructive pulmonary disease and the number of cigarettes smoked per day, the number of years of smoking, and the depth of smoke inhalation.2 Second, thoracic surgery often demands direct manipulation of the tracheobronchial tree. Although much of the direct surgical contact is external (on the adventitia), the manipulations (clamping, compression between the fingers) usually cause mucosal stimulation, as does endotracheal intubation, and can therefore cause bronchospasm. Two studies suggest that, just as endothelial cells are important modulators of vascular tone, the epithelium of the airway is a metabolically active tissue that can modulate airway smooth-muscle tone/function by the production and destruction of inflammatory mediators such as prostaglandins and perhaps relaxing and constricting factors.3·4 Epithelial damage may result in loss of an epithelium-derived relaxant factor and loss of enzymes that degrade constrictor neuropeptides.5 In addition, damage to the epithelium increases access of luminal substances to muscle and to afferent nerve fibers that elicit the irritant bronchoconstrictor responses seen so often with airway instrumentation during anesthesia.5 In fact, even normal, healthy patients undergoing surgical stimulation of the lung parenchyma and airways can experience bronchospasm,6 especially if they are too lightly anesthetized. a. EFFECT OF ANESTHETIC DRUGS ON AIRWAY REACTIVITY (Table 8-1)
The following discussion of the pharmacology of anesthetic drugs is limited to those aspects that suggest their use or avoidance in patients with reactive airways. Some of the drugs used in anesthetic practice (especially the halogenated drugs) decrease the reactivity of airways. However, there
302
Choice of Anesthetic Drugs and Techniques
Table 8-1 EFFECT OF ANESTHETIC DRUGS ON AIRWAY REACTIVITY Anesthetic Drug Class Inhalation
Narcotics
Induction
Relaxants
Adjuncts
Drug
Bronchodilation
Bronchoconstriction
Comment
Isofiurane Halothane Enflurane N20 Fentanyl Meperidine Morphine Ketamine
+ + + + + + 0 0 0 0 +
0 0 0 0 0 + + + 0
Thiopental Vecuronium Pancuronium Succinylcholine Atracurium
0 0 0 0 0
+? 0 0 0 +?
Metocurine
+
i/-tubocurarine
+ +
Lidocaine
+
Neostigmine
+ + +
Atropine
+
Drug of choice See text See text —» Light anesthesia Drug of choice Releases histamine Releases histamine Drug of choice for asthmatic Light anesthesia? Drug of choice Drug of choice Drug of choice Releases histamine, high doses Releases histamine, high—moderate doses Releases histamine, normal doses Useful pre- and intraoperatively Must use atropine concomitantly See text
Key: + , + • * - , + + + = mild, moderate, severe effect, respectively.
must be some initial smooth-muscle constriction in order for a bronchodilating anesthetic to have any effect on bronchomotor tone. Thus, measurements of airway resistances in normal men have generally failed to show a bronchodilatory effect of halothane.7 Similarly, in patients with normal bronchomotor tone on cardiopulmonary by-pass, the administration of halothane both by the airway and systemically (via the pump circuit) does not alter the resistive work of breathing (as it does in patients with increased bronchomotor tone; see the following).8 When bronchoconstriction is provoked in patients by either hypocapnia9 or inhalation of ultrasonic aerosols,10 halothane and enflurane reliably decrease bronchomotor tone. There are two important mechanisms of halogenated drug inhibition of bronchomotor tone. The first, and most important, the mechanism of bronchodilation by the halogentated drugs is by direct action on the airway musculature and/or local reflex arcs rather than, or at least in addition to, an effect on centrally mediated reflex pathways. This contention is best supported by the fact that systemic (intravenous) administration of halothane via cardiopulmonary by-pass pump does not decrease hypocapnia-induced increased airway resistance, whereas inhaled halothane (during cardiopulmonary by-pass) does.8, " Because halothane has a direct relaxant effect on bronchial smooth muscle, it is not surprising that in animals it can
block acetylcholine-, histamine-, alpha-adrenergic-, and antigen-induced increases in bronchial muscle tension.12 Second, halothane also blocks central reflex pathways. Clinically relevant concentrations of halothane reduced the amount of acetylcholine released from nerve terminals in response to nerve stimulation and reduced the size of the muscle contraction. Thus, inhalation anesthetics such as halothane suppress the final step of the reflex pathway, particularly vagal reflexes.5 The fact that the direct dilating effects are more important than the effects on reflexes is highlighted by one study that showed that at 1.2 minimum alveolar concentration (MAC), 60 per cent of the decrease in response to vagal nerve stimulation caused by halothane could be attributed to a direct effect on smooth muscle.13 Isoflurane is probably just as efficacious as halothane or enflurane in decreasing elevated bronchomotor tone. 5 · I2 · l4 Because the halogenated drugs are the most potent bronchodilating anesthetic drugs used today, they must be considered the anesthetic drug of choice for patients with reactive airways. If the halogenated drug is used as a primary induction agent for a patient with reactive airways, then halothane might be considered the induction drug of choice because it is less pungent than isoflurane. However, isoflurane is a better choice for the maintenance of anesthesia for four reasons. First, isoflurane has a high arrhythmogenic
Choice of Anesthetic Drugs and Techniques
threshold (in contradistinction to halothane, which sensitizes the myocardium to catecholamines and is commonly associated with ventricular arrhyth mias). This consideration has increased importance in patients with reactive airways because they may more likely receive aminophylline and β 2 ^οηΪ8ί8 and may more likely become acidotic (all of which may cause arrhythmias). Second, isoflurane is not metabolized as much as halothane (it has less or no hepatic toxicity). Third, isoflurane provides much more cardio vascular stability and potency than enflurane. Fourth, isoflurane is efficacious in treating very severe bronchospasm (status asthmaticus). 15-18 Consequently, isoflurane should be regarded as the halogenated drug of choice for maintenance anes thesia in patients with reactive airways. Fentanyl does not have any effect on bronchomotor tone, which is consistent with the fact that, in humans, fentanyl does not change plasma his tamine concentrations.19 In contrast, morphine is known to release histamine and to increase central vagal tone. In dogs, bronchoconstriction caused by morphine has been shown to be reduced either by administration of an antihistamine intravenously or by bilateral vagotomy.20 Similarly, meperidine has been shown in dogs to have a bronchoconstricting effect similar to that of morphine.20 Consequently, in patients with reactive airways for whom a nar cotic supplement to nitrous oxide (which has no effect on bronchomotor tone) or halogenated drug anesthesia is desired, or to use a narcotic alone in high doses, fentanyl is the preferred choice. How ever, it should be remembered that an N 2 0 nar cotic-relaxant anesthetic, as ordinarily adminis tered with low to moderate doses of narcotic, produces a light anesthesia and will not prevent bronchospasm in patients with reactive airways. Ketamine and thiopental are intravenous anes thetic drugs that are mainly associated with the induction of anesthesia. For the patient with reac tive airways, ketamine is believed to be more ad vantageous than most other anesthetic drugs used for induction. In dogs, ketamine protects against antigen-induced bronchospasm, while thiopental does not; the protective effect can be blocked by propanolol, suggesting that the mechanism of ac tion of ketamine is, perhaps, due to beta-adrener 21 gic stimulation (although the drug may have a direct bronchodilating effect on airway smooth muscle).22 Thus, ketamine has been used very suc cessfully in patients with a history of asthma23 and in treating patients with bronchospasm while on mechanical ventilation (who were refractory to 24 maximum broncholytic therapy), whereas thio pental has clinically been associated with more bronchospasm than any other anesthetic drug (per haps because of light levels of anesthesia that
303
leave airway reflexes relatively intact).25 Conse quently, ketamine (1 to 2 mg/kg) may be the drug of choice for the intravenous induction of anesthe sia in patients with bronchospastic disease requir ing a rapid induction of anesthesia. Ketamine has been successfully used as the sole anesthetic for thoracic surgery.26 With the exception of d-tubocurarine, all of the muscle relaxants may be used in patients with re active airways. In humans, i/-tubocurarine releases histamine and increases airway resistance.27 Metocurine can also release histamine in ordinary clin ical doses but not nearly as much as d-tubocurarine. Atracurium may release histamine but only in very high doses. Succinylcholine is structurally related to acetylcholine and could theoretically re lease histamine, but clinically it is unassociated with bronchospasm, even in asthmatics. Pancuro nium and vecuronium do not release histamine. Lidocaine, given immediately before intubation (1-2 mg/kg intravenously), is a useful drug in the prevention of reflex bronchoconstriction and laryngospasm provoked by instrumentation of the airway. Similarly, lidocaine by infusion (1-3 mg/kg/hour) may be useful in diminishing the reactivity of airways throughout surgery in patients with limited cardiac reserve who would not hemodynamically tolerate the usual doses of the usual anesthetics. Intravenous administration of lidocaine has also been successfully used to treat bronchospasm dur ing anesthesia.28 Intravenously administered lido caine also significantly lowers the incidence of induced laryngospasm (laryngospasm quantita tively defined/measured as an increase in intraglottic pressure of 45 ± 15 mm Hg for 61 ± 19 sec); a serum lidocaine level of 0.4 μg/ml appears to be the lower limit of the therapeutic range, although further study is necessary to construct a dose-re sponse curve.29 Lidocaine, inhaled in an ultrasonic aerosol, has been shown to produce mild bronchodilation in healthy subjects, whereas normal saline induced 30 mild bronchoconstriction. In a similar study, li docaine administered by aerosol both prevented and reversed increases in airway resistance pro voked by ultrasonically nebulized water.31 The lidocaine was thought to be acting directly on bronchial smooth muscle. Thus, it appears that ad ministration of lidocaine either via the airway or intravenously has a role in preventing and protect ing against the development of bronchospasm in patients with reactive airways. Neostigmine, physostigmine, and pyridostig mine are cholinesterase-blocking drugs that can be expected to produce an increase in airway resis tance by increasing cholinergic activity. In clinical practice, of course, atropine is used to block this
304 Choice of Anesthetic Drugs and Techniques
effect when anticholinesterases are administered to reverse the effects of neuromuscular blocking drugs. Atropine, the classic cholinergic blocker, not only reverses effects of anticholinesterase drugs but can also directly dilate the airways. A decrease in airway resistance was found following intravenous administration of atropine (0.84 mg/70 kg) to patients anesthetized with 75 per cent nitrous oxide and oxygen.32 In addition to its potential effect on airway resistance, atropine has been shown to increase respiratory dead space in both dogs and humans, presumably by dilating larger bronchi.33 Table 8-1 summarizes the effect of inhalation and intravenous anesthetic drugs on bronchomotor tone and, in view of the preceding discussion, ranks the various drugs of each class in terms of their efficacy for the patient with reactive airways. Ketamine appears to be the intravenous induction drug of choice, with an adequate dose of thiopental as a close second. Intravenous lidocaine is probably a useful adjunct in the peri-induction period. Isoflurane appears to be the halogenated drug of choice for maintenance anesthesia. Fentanyl is the narcotic of choice. Relaxation can be facilitated by succinylcholine, but if a rapid intubation is not necessary, vecuronium or pancuronium may be the relaxants of choice, with atracurium a reasonably close second choice. Nitrous oxide is a benign drug for the patient with reactive airways (if anesthesia caused by other drugs is adequate), but in view of the fact that a large number of patients undergoing thoracic surgery will have one-lung ventilation and therefore require a high F,02, use of nitrous oxide is limited in these patients. 2. Gas-Exchange Impairment Chapter 4 describes in great detail the pathophysiology of two-lung ventilation in the openchest paralyzed patient in the lateral decubitus position. In summary, the nondependent lung may be well ventilated owing to an increase in compliance but may be poorly perfused owing to gravitational effects. The dependent lung may be poorly ventilated owing to a decrease in compliance but may be well perfused. Consequently, there may be mismatching of ventilation and perfusion in the openchest, paralyzed patient in the lateral decubitus position. The low ventilation-perfusion ratio in the dependent lung can be improved with selective dependent-lung positive end-expiratory pressure (PEEP) while the nondependent lung is ventilated with zero end-expiratory pressure (ZEEP). The differential lung ventilation combination of dependent-lung PEEP and nondependent-lung ZEEP provides better arterial oxygenation compared with ventilation of both the nondependent and dependent lungs with ZEEP.
Chapter 4 also describes in great detail the pathophysiology of one-lung ventilation in the lateral decubitus position. The one-lung ventilation condition creates a large obligatory right-to-left shunt that is not present during the two-lung ventilation condition. This right-to-left shunt during one-lung ventilation, however, is minimized by the presence of HPV in the nondependent, nonventilated lung. Section III of this chapter considers in great detail the effects of anesthetic drugs on HPV in the nonventilated, nondependent lung. B. Cardiac Considerations 1. Coronary Artery Disease Patients undergoing thoracic surgery, especially those beyond their fourth decade, should be approached with an increased index of suspicion that coronary artery disease may be present. Ten major cohort studies, accounting for more than 20 million person years of observation in several countries, support the statement in the 1983 Surgeon General's report that "cigarette smoking should be considered the most important of the known modifiable risk factors for coronary heart disease in the United States."34 These studies showed that men 40 to 59 years of age who were smoking a pack or more per day at the time of initial examination had a risk for a first major coronary event that was 2.5 times as great as that of nonsmokers, with a strong dose-response relation.35 Studies both in the United States and abroad have demonstrated consistently that women whose smoking patterns are similar to those of men have a similar increased risk of death from coronary heart disease and for common morbidity from the disease, such as angina pectoris, compared with nonsmokers.34·36 The risk of death from coronary heart disease among both male and female smokers is increased by early initiation of smoking, long total exposure to smoking, and deep smoke inhalation. Although smoking, hypertension, and hypercholesterolemia confer approximately the same average increase in the risk of coronary heart disease in populations, smoking in the presence of other risk factors for coronary heart disease appears to create a synergistic effect on mortality from the disease.34 The lifestyle of many smokers also includes caffeine and nicotine addiction and obsessive-compulsive type A behavior and results in an increased incidence of systemic hypertension; the combination of smoking and systemic hypertension increases the risk of coronary artery disease. Pipe and cigar smokers have a risk of experiencing a major coronary event and subsequent morbidity from chronic heart disease that is intermediate between that for nonsmokers and cigarette smokers.34
Choice of Anesthetic Drugs and Techniques
Smoking cessation results in a decreased risk of mortality from coronary heart disease, and the degree of risk reduction is determined by the length of time after cessation, the amount smoked, and the duration of smoking before cessation. Although the risk of coronary heart disease attributable to smoking declines by approximately 50 per cent 1 year after cessation, only after a decade or more does it approach that of a person who has never smoked.34 The anesthetic management and choice of anesthesia for patients with coronary artery disease are based on the factors that determine myocardial oxygen supply and demand (Fig. 8-1). Fortunately, the reserve on the supply side is large enough to make ischemia impossible in normal hearts even with the most vigorous level of exercise/stress. However, in the presence of coronary
305
artery disease or abnormal left ventricular hypertrophy, an imbalance between supply and demand may occur with minimal stress/exercise or even at rest and during anesthesia.37 Myocardial oxygen supply is determined by the product of coronary artery blood flow (most important, by far, in most clinical situations) and the oxygen content of coronary artery blood. For a given coronary vascular resistance, coronary artery blood flow is increased by an increased diastolic filling time (slow heart rate, most important oxygen-supply factor; Q in Fig. 8-1), and the coronary artery perfusion pressure (the second most important oxygen-supply factor; (2) in Fig. 8-1). The coronary perfusion pressure is equal to coronary artery diastolic pressure (Psad) (there is no perfusion during systole) minus the preload (which is ventricular end-diastolic pressure or the outlet
DETERMINANTS OF MYOCARDIAL OXYGEN SUPPLY AND DEMAND: THE BALANCE
Figure 8—1 The determinants of myocardial oxygen supply and demand. See text for full explanation. (SVR = systemic vascular resistance; EDRF = endothelial relaxing factor; PLVBD = left ventricular end-diastolic pressure; PSAD = systemic arterial diastolic pressure.
306
Choice of Anesthetic Drugs and Techniques
or back pressure). Coronary vascular resistance is (1) inversely proportional to myocardial metabolic rate (this coupling mechanism acts within one cardiac cycle, and maximal coronary dilation or constriction can be elicited within 15 to 20 sec), (2) directly proportional to perfusion pressure so that blood flow is kept constant between 60 and 160 mm Hg (at less than 60 mm Hg, blood flow becomes pressure dependent), (3) directly proportional to systolic compression (the magnitude of this effect decreases from endocardium to epicardium, and, in fact, left ventricular subendocardial blood flow ceases during systole), (4) decreased in response to humoral factors (angiotensin II, serotonin, thromboxane, prostacyclin, and bradykinin) that can promote or inhibit release of endothelium-derived relaxing factor, and (5) under neural influence (alpha-adrenergic stimulation induces coronary constriction; beta-adrenergic and parasympathetic stimulation induces coronary vasodilation).37 The oxygen content of arterial blood is a function of the amount of hemoglobin (most important), position of the oxygen-hemoglobin dissociation curve (P50), and the ventilation-perfusion relationships within the lung (which determines P a 0 2 ). Myocardial oxygen demand is raised by tachycardia (the most oxygen-expensive factor; ® in Fig. 8-1), increased diastolic and systolic myocardial wall tension, and increased contractile state of the heart (as caused degree of sympathetic nervous system activity). Diastolic myocardial wall tension is determined by preload (left ventricular end-diastolic [LVED] pressure; the second most oxygenexpensive factor; (5) in Fig. 8-1), and systolic myocardial wall tension is determined by afterload (systemic systolic blood pressure and vascular resistance). Coronary obstruction causes a decrease in perfusion pressure. The pressure decrease across a plaque is proportional to the radius (fourth power), the length of stenosis, and the magnitude of flow. As a stenosis increases, a larger pressure decrease develops. This reduction in perfusion pressure (Psad — PLVED) i s compensated for by a distal bed vasodilation until the critical lower threshold of coronary perfusion pressure (Psad — P LVE D) *S reached (—60 mm Hg). At this point, vasodilation is maximal, further autoregulation is not possible, and flow becomes dependent on perfusion pressure in a linear manner. Therefore, the significance of coronary artery disease becomes clinically apparent when the perfusion pressure of coronary vasculature falls below this threshold. The area of myocardium that is most vulnerable to the effects of coronary artery disease and reduced perfusion pressure is subendocardium (inner one third or one fourth of the myocardium). The
subendocardium is most susceptible because it has the highest metabolic rate (resulting from an increased systolic and diastolic wall tension) compared with epicardium; consequently, the ratio of endocardial to epicardial blood flow is approximately 1.25:1.0 in conscious dogs.38 From the foregoing discussion of myocardial oxygen supply and demand, one can see that an increase in the diastolic phase (decreased heart rate, decreased P LVE D) m a v have a dramatic impact on subendocardial perfusion. On the demand side, the importance of controlling heart rate is again evident as is a reduction in wall tension through decreasing preload and afterload. The medicopharmacologic armamentarium to achieve these end points is made up of three classes of agents (excluding antiplatelet agents): (1) nitrates, (2) betaadrenergic receptor blockers, and (3) calcium channel blockers. Nitrates relax the vascular smooth muscle of the venous and coronary arterial circulation. The vasodilatory effect of nitrates on coronary arteries yields increased supply of oxygen to the myocardium. Through venous and arterial vasodilation, nitrates decrease preload and afterload (diastolic and systolic wall tension, respectively) and therefore myocardial oxygen consumption. The increased supply of and decreased demand for oxygen caused by nitrates favorably affect the subendocardium; the net result is a preferential redistribution of blood to the subendocardium as a result of treatment with nitrates.39 By controlling the heart rate and contractility, beta-adrenoreceptor blocking agents decrease the myocardial oxygen requirement at times of increased sympathetic output. Slower heart rate not only reduces myocardial oxygen consumption but also leads to a prolonged diastolic phase, which has a favorable effect on subendocardial blood flow. The three commonly used calcium channelblocking agents in the United States are nifedipine, diltiazem, and verapamil. They all inhibit calcium influx into smooth-muscle cells and cardiac cells, causing vasodilation and negative inotropy, thereby reducing myocardial oxygen consumption as well as enhancing myocardial oxygen delivery. In summary, common precipitating factors for myocardial ischemia are tachycardia (increases oxygen demand and decreases oxygen supply), increased preload (increases oxygen demand and decreases oxygen supply), hypertension (increases systolic wall tension more than increasing oxygen supply from the increase in perfusion pressure), and hypotension (decreases oxygen supply more than oxygen consumption). It follows that anesthetic management should minimize myocardial oxygen demand by continuing preoperative beta
Choice of Anesthetic Drugs and Techniques
blockers until the time of surgery, minimize preinduction anxiety, keep heart rate low, maintain ad equate levels of anesthesia, and use myocardial depressant drugs when indicated. Anesthetic man agement should also maximize myocardial oxygen supply by keeping diastolic blood pressure normal or increased, keeping heart rate low, ensuring ad equate arterial oxygenation, and using venous and arterial vasodilators to reduce preload and afterload, respectively, and to relieve coronary artery spasm. a. EFFECT OF ANESTHETIC DRUGS ON CARDIOVASCULAR FUNCTION (Table 8-2) This discussion is not intended as a review of the general pharmacology of anesthetic drugs; rather, it discusses those features of anesthetic drugs that suggest their use or avoidance in pa tients with coronary artery disease. Depending on the drug and the amount used, anesthetic technique may alter all of the determinants of myocardial oxygen supply and demand. The narcotics, especially fentanyl, have minimal primary hemodynamic effects if adequate doses are used (or are supplemented by other drugs). High doses of fentanyl (greater than 60 μg/kg) provide remarkable hemodynamic stability for cor onary artery by-pass surgery and, in this group of patients, low to moderate doses of fentanyl (15 μg/kg) will also provide stable anesthesia for in
307
duction and intubation, provided the fentanyl is administered rapidly (within 12 sec). 40 The de crease in heart rate with fentanyl is a desirable characteristic in patients with coronary artery dis ease. Meperidine has a negative inotropic effect and positive chronotropic effect. Propofol appears to affect the cardiovascular system in a biphasic manner.41 First, there is a marked reduction in systemic vascular resistance with an associated tachycardia, an increase in car diac output, and a decrease in arterial pressure. Second, as the systemic vascular resistance in creases toward normal, there is a decrease in heart rate and cardiac output, and a further slight de crease in arterial pressure. At 1 to 2 MAC, the halogenated drugs have a number of unfavorable cardiovascular effects. A major cardiovascular disadvantage of the halogen ated drugs is a 20 to 40 per cent decrease in sys temic blood pressure. Since cardiac output is de creased 20 to 40 per cent by halothane and enfiurane, and not at all by isoflurane, and in the context of a 20 to 40 per cent decrease in systemic blood pressure caused by all drugs, systemic vas cular resistance is only minimally affected by hal othane and enfiurane but is decreased by isoflur ane. Since filling pressures are increased by halothane and enfiurane, and not at all by isoflur ane, and in the context of a decrease in cardiac output caused by halothane and enfiurane, but not
Table 8-2 EFFECT OF ANESTHETIC DRUGS ON CARDIOVASCULAR FUNCTION
*Ketamine has a direct myocardial depressant action that becomes evident when its sympathetic stimulating effects are blocked or when the sympathetic nerves are maximally stimulated, as in severe hypotension. Key: 0 = no change; 1 and 11 = 10 to 20 per cent and 20 to 40 per cent decrease, respectively; f , ţ \ , ţ ţ ] = progressively greater increases; ± = small changes that depend on circumstances and reflex activity.
308 Choice of Anesthetic Drugs and Techniques
by isoflurane, cardiac contractility is decreased by halothane and enflurane and not at all by isoflurane. Halothane greatly sensitizes the myocardium to catecholamines, and arrhythmias may be prominent in patients with irritable ventricular foci (which may be due to ischemia caused by coronary artery disease). Enflurane causes a 20 to 40 per cent and isoflurane a 10 to 20 per cent increase in heart rate, which may be very and moderately disadvantageous, respectively, to the patient with coronary artery disease. Isoflurane may act as a nonspecific coronary vasodilator, which can theoretically result in a "coronary steal"; that is, vasodilation and an increase in blood flow in normal areas can occur at the expense of blood flow to ischemic areas that have a fixed vascular resistance.42 Isoflurane may also induce myocardial ischemia in patients with coronary artery disease by causing systemic hypotension and tachycardia (in addition to a coronary steal). One study of anesthetized patients just before undergoing coronary artery by-pass grafting surgery determined the effect of equipotent doses of halothane and isoflurane on coronary blood flow and myocardial ischemia and thereby provided an objective basis for analyzing the contention that isoflurane predisposes to the development of ischemia by "stealing" blood flow.43 Global and regional myocardial blood flow and metabolism were examined in 20 patients with coronary artery disease before surgical stimulation. Half were anesthetized with halothane (0.8 per cent) and half with isoflurane (1.2 per cent) (these concentrations are equipotent). Coronary perfusion pressure decreased similarly in both groups. During halothane anesthesia, coronary sinus blood flow, an index of global perfusion, decreased from an awake value of 129 ± 7 to 97 ± 7 ml/min (p < .05), and great cardiac vein blood flow, an index of regional perfusion (drains the area supplied by the left anterior descending coronary artery), decreased from 60 ± 8 to 44 ± 5 ml/min (p < .05). In contrast, during isoflurane anesthesia, global coronary blood flow increased from 131 ± 13 to 153 ± 16 ml/min (p < .05), whereas regional blood flow decreased from 68 ± 7 to 56 ± 6 ml/ min (p < .05). Thus, the ratio of great cardiac vein blood flow to coronary sinus blood flow was unchanged during halothane anesthesia but decreased significantly during isoflurane. Neither global nor regional coronary vascular resistance was altered by halothane, whereas isoflurane decreased global coronary vascular resistance without affecting regional coronary vascular resistance. All patients receiving halothane had net myocardial lactate extraction. In the isoflurane group, four patients showed global lactate production and three regional lactate production. All patients demonstrat-
ing lactate production also developed electrocardiographic evidence of myocardial ischemia, which was not present before induction. The authors concluded that halothane is a preferable anesthetic to isoflurane in patients with coronary artery disease because the latter has the propensity to induce maldistribution of the coronary circulation (steal) and myocardial ischemia. The maldistribution of coronary artery blood flow (steal) caused by isoflurane is attributed to vasodilation in normal coronary arteries while stenotic ones must remain constant in size. Indeed, one case report documented the development of ischemia in an isoflurane-anesthetized patient that was successfully treated by switching to halothane.44 The neuromuscular blocking drugs have hemodynamic effects, but they are not generally of large magnitude. Pancuronium has the most undesirable effects for the patient with coronary artery disease because it can stimulate the sympathetic nervous system, causing tachycardia and increased blood pressure and cardiac output. i/-tubocurarine also has undesirable effects for the patient with coronary artery disease because of the development of hypotension and tachycardia. Succinylcholine may mimic acetylcholine at nicotinic and muscurinic receptors; consequently, a dose-related tachycardia followed by bradycardia may be seen owing to sequential stimulation of these receptors, respectively. The other relaxants have only minimal effects on cardiovascular function. Ketamine can cause a significant stimulation of the sympathetic nervous system that results in an increase in contractility, tachycardia, and hypertension, whereas thiopental can cause significant depression of myocardial contractility and hypotension. Consequently, both of these intravenous anesthesia-inducing drugs have major disadvantages in patients with coronary artery disease. However, the sympathomimetic effect of ketamine may be used to good advantage in hypovolemic patients.
III. CHOICE OF ANESTHESIA AND ARTERIAL OXYGENATION DURING ONE-LUNG VENTILATION A. Effect of Anesthetics on Hypoxic Pulmonary Vasoconstriction As discussed in chapters 3 and 4, an undesirable property of general anesthesia is inhibition of HPV in the nonventilated, nondependent lung by the anesthetic drug. All of the inhalation anesthetics and many of the injectable anesthetics have been studied with regard to their effect on HPV. Halothane has been the most extensively studied agent
Choice of Anesthetic Drugs and Techniques
309
Table 8-3 EFFECT OF HALOTHANE ON HYPOXIC PULMONARY VASOCONSTRICTION (HPV) IN VARIOUS EXPERIMENTAL PREPARATIONS
Key: j = decrease; f = increase; J /Dr-to % = HPV was progressively decreased to the maximum shown in column 5 over the concentration range shown in column 4.
(Table 8-3).45-*7 The experimental preparation used may be divided into four basic categories: (1) in vitro, (2) in vivo-not intact (pumped perfused lungs, no systemic circulation or neural function), (3) in vivo-intact (normally perfused lungs, normal systemic circulation), and (4) humans (volunteers or patients). It appears, according to this breakdown of experimental preparation, that inhibition of HPV by halothane is a universal finding in the in vitro and in vivo-not intact preparations. The site of inhibition is precisely where one would expect it to be; namely, at the arteriolar level (middle segment of an inflow/outflow occlusion model), which is the same site of action of hypoxia.55· 56· 68 However, in the more normal or physiologic in vivo-intact and human studies, halothane has caused no or only a very slight decrease in HPV response. Thus, it appears that a fundamental property of halothane is its inhibition of HPV in experimental preparations, which can be controlled for other physiologic influences (e.g., pulmonary vascular pressure, cardiac output, mixed venous oxygen tension, C0 2 level, and temperature) that can have an effect on the HPV response. In the more biologically complex in vivo models, other factors seem to be involved that greatly
diminish the inhibitory effect of halothane on HPV. Important méthodologie differences between the in vitro and in vivo-not intact preparations and the in vivo-intact and human models that could account for the observed differences in halothane effect on HPV are presence (or absence) of perfusion pulsations, perfusion fluid composition, size of perfusion circuit61; baroreceptor influences, absence of bronchial blood flow (which abolishes all central and autonomic nervous activity in the lung)69, chemical influences (i.e., pH, Po,). humoral influences (i.e., histamine and prostaglandin release from body tissues), lymph flow influences, and, very importantly, unaccounted for or uncontrolled changes in physiologic variables, such as cardiac output, mixed venous oxygen tension, and pulmonary vascular pressures, which might have directionally opposite effects on HPV, and the use of different species.70-72 Ether has been the next most studied drug, and it appears that the quantitative effect of ether on HPV is also dependent on the type of experimental preparation used. Thus, the in vitro and in vivonot intact models show much more inhibition of HPV by anesthetic drug (ether) than the in vivointact and human models (Table 8-4). 46 · 47 · 5() - 5I · 54 · 64 "• 74 Although the number of studies involving
Table 8-4
EFFECT OF ETHER, ISOFLURANE, ENFLURANE, METHOXYFLURANE, FLUROXENE, TRICHLOROETHYLENE, NITROUS OXIDE, AND INJECTABLE ANESTHETICS ON HYPOXIC PULMONARY VASOCONSTRICTION (HPV)
Anesthetic Drug Ether
Isoflurane
Experimental Preparation
Species
Regional (R) vs. Whole (W)Lung Hypoxia
In vitro: Heart-lung Heart-lung Heart-lung Lung Lung In vivo: Not intact, pump perfused In vivo: Intact, normally perfused Human
Rat Rat Rat Cat Cat Cat
w w w w w w
In vitro: Heart-lung In vivo: Not intact, pump perfused In vivo: Intact, normally perfused
Human
Enflurane
In vitro: Heart-lung Heart-lung Heart-lung Human
Effect on HPV/Magnitude of Change
Dose (Converted to MAC) 0.5-1.0 1-2 4-6 0.5-5.0 1 2.5-5.0
1/Dr-to 100% 1/Dr-to 100% 160 + 70% I 90-95% 185% 1 /Dr-to 95%
Bredesen et al.41 Bjertnaes47 Bjertnaes et al.5 Sykes et al.51 Hurtig et al." Loh et al.54
Dog
1.5-3.0
1 /55%
Sykes et al.74
Human
1-2
1 /33%
Pavlin et al.64
0.2 0-2
1 /Dr-to 90% 0
Marshall et al.49 Gardaz et al.76
1 /Dr-to 60% 1 /Dr-to 50% 0 1 /Dr-to 50% 0 0 0 0 0 0
Benumof & Wahrenbrock62 Mathers et al.63 Saidman & Trousdale75 Domino et al.77 Chuda et al.78 Naeije et al.79 Groh et al.8" Jolin-Carlsson et al.81 Benumof et al.67 Carlsson et al.82
1 /Dr-to 90% 1 /60% 1/Dr-to 100% 0
Marshall et al.49 Bjertnaes et al.M Bjertnaes & Mundal8' Carlsson et al.84
Rat Dog
w
Dog Dog Dog Dog Dog Dog Rabbit Human Human Human
R R R R R W R R R R
1-3 1-2 1-2 0-2.4 1.3 1.0 1.2 1.0-1.5 1.0 1.0
Rat Rat Rat Human
W W W R
0-2 1-3 1-3 1-2
R
Study
Key: \ = decrease; \ = increase; Dr-to % = HPV was progressively decreased to the maximum shown in column 6 over the concentration range shown in column 5. *Drugs used in these experiments were fentanyl, propofol, meperidine, morphine, thiopental, pentobarbital, hexobarbital, droperidol, diazepam, chlorpromazine, ketamine, pentazocine, lidocaine, buprenorphine. For doses and blood levels, see the references in the far right column.
312 Choice of Anesthetic Drugs and Techniques
halogenated drugs other than halothane—namely, isoflurane,49· 6 2 ^· 67· 75"82 enflurane,49· 5a 8 3 · 84 methoxyflurane,45^7·85·86 fluroxene,62·63 and trichloroethylene51—have been too small to permit recognition of a clear experimental preparation result pattern, most of these anesthetics have demonstrated inhibition of HPV in the in vitro models (see Table 8-4). However, both isoflurane and enflurane show no inhibition of HPV in the only studies performed in humans 81 · 84 ; because of the importance of these studies, they are discussed at some length presently. Nitrous oxide seems to cause a small, somewhat consistent inhibition of HPV (see Table 8-4). 47 · 58 · 62 · 63 · 87 - 89 All injectable anesthetics studied to date have no effect on HPV (see Table 8-4).46·47·59·62·90-96 The influence of isoflurane on HPV was studied in eight subjects before undergoing elective surgery.82 The lungs were ventilated separately with a double-lumen endobronchial catheter. After oxygen ventilation of both lungs for 30 min during intravenous barbiturate anesthesia, the test lung was rendered hypoxic by ventilation with 8 per cent 02 in nitrogen. The control lung was ventilated continuously with 100 per cent 0 2 . Isoflurane was added to the inspired gas, so that end-tidal concentrations of 1 per cent and 1.5 per cent were obtained. Cardiac output (Qt) was determined by thermodilution, and the distribution of blood flow between the lungs was assessed from the excretion of a continuously infused, poorly soluble gas (SF6). The hypoxic challenge during intravenous anesthesia resulted in a reduction in the fractional
perfusion of the test lung from 54 per cent to 41 per cent of Q, (Fig. 8-2). Mean pulmonary arterial pressure increased by 46 per cent, and pulmonary vascular resistance (PVR) of the test lung more than doubled. Arterial oxygen tension fell from 375 mm Hg (50 kPa) to 101 mm Hg (13.5 kPa). Adding isoflurane to the inhalation gas, first at a concentration of 1 per cent and then 1.5 per cent, caused no further significant change in the distribution of pulmonary blood flow, although six of the eight subjects showed a small increase in test lung blood flow at isoflurane 1.5 per cent (see Fig. 8-2). There was no change in PVR or in any other circulatory variable. Arterial blood gases remained essentially unaltered. When the hypoxic challenge was discontinued, all variables returned to control values. It is possible that higher isoflurane concentrations would have caused a clear change in the blood flow distribution, but, at clinical concentrations, the effect of HPV in the human is all but unmeasurable. Similarly, the degree of HPV was also studied in eight subjects during enflurane anesthesia and was compared with that during intravenous pentobarbital anesthesia in the same subjects.84 The lungs were ventilated separately with the aid of a double-lumen endobronchial catheter. After preoxygenation of both lungs for 30 min, during intravenous anesthesia, the right lung (test lung) was rendered hypoxic by ventilation with 6 per cent 02 in nitrogen. The left lung (control lung) was ventilated continuously with 100 per cent 0 2 . Cardiac output and the distribution of blood flow between the two lungs was assessed as previously described
Figure 8-2 The perfusion of the test lung (Q^) and the control lung (QCL) in the control period and during hypoxia without and with isoflurane. Note the redistribution of pulmonary blood flow during hypoxia, indicating functioning hypoxic vasoconstriction before and during isoflurane anesthesia. (Reproduced with permission from Carlsson AJ, Bindslev L, Hedenstierna G: Hypoxia-induced pulmonary vasoconstriction in the human lung. The effect of isoflurane anesthesia. Anesthesiology 66:312-316, 1987.)
Choice of Anesthetic Drugs and Techniques
in the isoflurane study. The hypoxic challenge resulted in a reduction of the fractional perfusion to the test lung from 57 per cent to 36 per cent of Qt. Mean pulmonary arterial pressure increased by 37 per cent, and PVR of the test lung doubled. Arterial oxygen tension decreased from 45.9 to 9.5 kPa. Inhalation of enflurane to an end-tidal concentration of 2 per cent to both lungs caused no significant change in the distribution of the pulmonary blood flow, PVR, or any other circulatory variables (Fig. 8-3). The arterial blood gases remained unaltered. When the hypoxic challenge was discontinued, all variables returned toward control values. The findings suggest that the inhalation anesthetic enflurane does not reduce the hypoxic vasoconstrictor response in the human lung. To summarize previous animal studies, it appears that a fundamental property of inhalational anesthetics is to decrease HPV. However, in intact animal preparations, some biologic or physiologic property seems to remove or greatly lessen the inhibitory effect of anesthetic drugs on HPV. It may be that the cause(s) of the difference in effect of anesthetic drugs on regional HPV from prepa-
313
ration to preparation, anesthetic to anesthetic, and species to species97 (see the following) is closely related to the mechanism of HPV, which is still unknown. B. Effect of Anesthetics on Arterial Oxygenation During One-Lung Ventilation An often-made extrapolation of the much more numerous in vitro and in vivo-not intact HPV studies is that anesthetic drugs might impair arterial oxygenation during one-lung anesthesia by inhibiting HPV in the nonventilated lung. One of the previously mentioned studies on the effect of isoflurane on regional canine HPV was especially well controlled and showed that, when all nonanesthetic drug variables that might change regional HPV are kept constant, isoflurane inhibits single-lung HPV in a dose-dependent manner.77 Additionally, the study is valuable because the authors offer the reader an easily comprehensible quantitative summary of the relationship between dose of isoflurane administered and degree of inhibition of the single-lung canine HPV response.
Figure 8-3 The fractional perfusion of the test (right) lung and control (left) lung during the control situations and during hypoxia without and with enflurane (Enfl.)· A functioning hypoxic vasoconstriction was evident from the diversion of pulmonary blood flow from the test lung to the control lung during hypoxia and enflurane inhalation. (Reproduced with permission from Carlsson AJ, Hedenstierna G, Bindslev L: Hypoxia-induced vasoconstriction in human lung exposed to enflurane anaesthesia. Acta Anaesthesiol Scand 31:57-62, 1987.)
314
Choice of Anesthetic Drugs and Techniques
If the summary can be extrapolated or applied to the clinical one-lung ventilation situation (at least as an approximation), insights can be gained into what might be expected with regard to arterial oxygenation when such patients are anesthetized with isoflurane. In order to put these insights into sharp clinical focus, it is necessary to first understand what should happen to blood flow, shunt flow, and arterial oxygenation, as a function of a normal amount of HPV, when two-lung ventilation is changed to one-lung ventilation in the lateral decubitus position. Once the stable one-lung ventilation condition has been described, it is then possible, using the data from the previously mentioned study, to see how isoflurane administration affects the one-lung ventilation blood flow distribution, shunt flow, and arterial oxygen tension.
1. Two-Lung Ventilation: Blood Flow Distribution Gravity causes a vertical gradient in the distribution of pulmonary blood flow in the lateral decubitus position for the same reason that it does in the upright position. Consequently, blood flow to the dependent lung is significantly greater than blood flow to the nondependent lung. When the right lung is nondependent, it should receive approximately 45 per cent of total blood flow as opposed to the 55 per cent that it received in the upright and supine positions. When the left lung is nondependent, it should receive approximately 35 per cent of total blood flow as opposed to the 45 per cent that it received in the upright and supine positions (closed-chest data with normal pulmo-
nary artery pressure) (Fig. 8^l·).98·99 If these blood flow distributions are combined (both the right and left lungs being nondependent an equal number of times), average two-lung ventilation blood flow distribution in the lateral decubitus position would consist of 40 per cent of total blood flow perfusing the nondependent lung and 60 per cent of total blood flow perfusing the dependent lung (Fig. 84, right-hand panel and Figs. 8-3 and 8-4, lefthand panels). It is possible that nondependent-lung blood flow may increase slightly when the nondependent hemithorax is open for two reasons.100 First, if the compliance of the nondependent lung increases so much that nondependent-lung alveolar pressure decreases significantly, nondependent-lung blood flow may increase relative to dependent-lung blood flow. Second, if the nondependent lung falls away from the open chest wall, the vertical distance between the heart and the nondependent lung may decrease, which in the face of a constant pulmonary artery pressure might result in an increased perfusion of the nondependent lung. Consequently, the 40/60 per cent nondependent-/dependent-lung blood flow ratio during closed-chest two-lung ventilation may be a slight underestimation of the ratio during open-chest two-lung ventilation.
2. One-Lung Ventilation: Blood Flow Distribution, Shunt Flow, and Arterial Oxygen Tension When the nondependent lung is nonventilated (made atelectatic), HPV in the nondependent lung
Blood Flow Distribution: Two Lung Ventilation Left Lung Nondependent
Right Lung Nondependent
Average of Both Lungs Being Nondependent
Figure 8—4 This schematic diagram shows that when the left lung is the nondependent lung, the distribution of blood flow between the nondependent and dependent lungs is 35/65 per cent. When the right lung is the nondependent lung, the blood flow distribution between the nondependent and dependent lungs is 45/55 per cent. When the left and right lungs are nondependent an equal number of times, then the average one-lung ventilation blood flow distribution would consist of a nondependent and dependent lung blood flow ratio of 40/60 per cent.
Choice of Anesthetic Drugs and Techniques
will increase nondependent-lung PVR and decrease nondependent lung blood flow. In the absence of any confounding or inhibiting factors to the HPV response, a single-lung HPV response should decrease the blood flow to that lung by 50 per cent (Figs. 8-5, 8-6, and 8-7).101 Consequently, the nondependent lung should be able to reduce its blood flow from 40 to 20 per cent of total blood flow and the nondependent-/dependentlung blood flow ratio during one-lung ventilation should be 20/80 per cent (see Fig. 8-7, middle panel). All the blood flow to the nonventilated nondependent lung is shunt flow, and therefore one-lung ventilation creates an obligatory right-to-left transpulmonary shunt flow that was not present during two-lung ventilation. If no shunt existed during two-lung ventilation conditions (ignoring the normal 1 to 3 per cent shunt flow due to the bronchial, pleural, and thebesian circulations), we would expect the ideal total shunt flow during one-lung ventilation to be a minimal 20 per cent of total blood flow. With a normal hemodynamic and metabolic state, the arterial oxygen tension should be approximately 280 mm Hg (Fig. 8-8).",2 Table 4-1 shows a quantitative example of a model of blood flow to each lung during two-lung
315
and one-lung ventilation, with an increasing initial two-lung ventilation shunt through both lungs. As the initial nondependent-lung shunt flow increases (i.e., during two-lung ventilation), the amount of nondependent-lung blood flow able to participate in nondependent-lung HPV and the amount of nondependent-lung HPV blood flow diversion decrease, and the one-lung ventilation shunt increases. In addition, as the fraction of the Q, that normally perfuses the operative lung increases (e.g., in a patient who has little or no lung disease, i.e., undergoing esophageal surgery), the shunt flow through the operative lung during one-lung ventilation increases and the P a O : decreases.103 As the initial dependent-lung shunt flow increases (i.e., during two-lung ventilation), the amount of one-lung ventilation shunt also increases irrespective of nondependent-lung HPV. 3. Effect of Isoflurane on the One-Lung Ventilation Blood Flow Distribution, Shunt Flow, and Arterial Oxygen Tension Domino et al.77 found that per cent inhibition of regional HPV response equals 22.8 (per cent alveolar isoflurane) minus 5.3. As previously de-
Amount of Lung That is Hypoxic, % Figure 8-5 The x-axis shows the amount of lung that is hypoxic. The left-hand y-axis shows the expected amount of blood flow reduction to the hypoxic lung as a result of hypoxic pulmonary vasoconstriction (HPV). The right-hand y-axis shows the amount of perfusion pressure increase expected because of HPV. When 40 per cent of the lung is hypoxic, the blood flow reduction to the hypoxic lung should be very near 50 per cent. (Redrawn with permission from Marshall BE, Marshall C: Continuity of response to hypoxic pulmonary vasoconstriction. J Appl Physiol 49:189-196, 1980.)
316
Choice of Anesthetic Drugs and Techniques
Conversion of Two-Lung to One-Lung Ventilation: Blood Flow Distributions
Figure 8—6 This schematic diagram shows that the two-lung ventilation nondependent/dependent lung blood flow ratio is 40/60 per cent (left-hand side). When two-lung ventilation is converted to one-lung ventilation (l LV) (as indicated by atelectasis of the nondependent lung), the hypoxic pulmonary vasoconstriction (HPV) response decreases the blood flow to the nondependent lung by 50 per cent so that the nondependent/dependent lung blood flow ratio is now 20/80 per cent (right-hand side).
Effect of 1 MAC Isoflurane Anesthesia on Shunt During One Lung Ventilation (1LV) of Normal Lungs
Figure 8-7 This schematic diagram shows that the two-lung ventilation nondependent/dependent lung blood flow ratio is 40/60 per cent (left-hand side). When two-lung ventilation is converted to one-lung ventilation (1 LV) (as indicated by atelectasis of the nondependent lung), the hypoxic pulmonary ventilation (HPV) response decreases the blood flow to the nondependent lung by 50 per cent, so that the nondependent/dependent lung blood flow ratio is now 20/80 per cent (middle). According to the data of Domino et al," administration of 1 minimum alveolar concentration (MAC) isoflurane anesthesia should cause a 21 per cent decrease in the HPV response, which would decrease the 50 per cent blood flow reduction to a 40 per cent blood flow reduction HPV response. Consequently, the nondependent/dependent lung blood flow ratio would now become 24/76 per cent, representing a 4 per cent increase in the total shunt across the lungs (right-hand side). (Reproduced with permission from Benumof JL: Isoflurane anesthesia and arterial oxygenation during one-lung ventilation. Anesthesiology 64:419-422, 1986.)
Choice of Anesthetic Drugs and Techniques
317
Figure 8-8 The x-axis shows the inspired oxygen concentration. The left-hand y-axis presents the expected arterial Po, for a family of intrapulmonary shunts. The tick intervals on the right-hand y-axis are the same as those for the left-hand y-axis. As shunt increases, the family of isoshunt lines becomes flatter and closer together, and for a given F,0 2 an increase in shunt has a decreasing effect on decreasing arterial Po : . The model assumes relatively normal hemoglobin, P a C0 2 , and a-v 02 content difference (upper left-hand corner). (Redrawn with permission from Lawler PGP. Nunn JF: A reassessment of the validity of the isoshunt graph. Br I Anaesth 56:1325-1335, 1984.)
318
Choice of Anesthetic Drugs and Techniques
scribed, under normal conditions, collapse of the nondependent lung in the lateral decubitus position causes a nondependent-lung HPV response to de crease nondependent-lung blood flow by 50 per cent; that is, from 40 to 20 per cent of total flow (see Figs. 8-5, 8-6, and 8-7). Using these values as a model of the normal two-lung to one-lung ventilation conversion process, we can construct a table (Table 8-5) that sequentially relates per cent alveolar isoflurane to per cent inhibition of the nondependent-lung HPV response to the resultant nondependent-lung HPV response (expressed as a per cent decrease in nondependent-lung blood flow), to the resultant increase in atelectatic non dependent-lung blood flow (which is the shunt during one-lung ventilation), to an absolute in crease in shunt, to a decrease in arterial oxygena tion during one-lung ventilation (from 280 mm Hg [F,0 2 = 1.0] to some lower value). Table 8-5 and the right-hand panel of Figure 87 show that 1 MAC isoflurane anesthesia would inhibit the nondependent-lung HPV response by approximately 21 per cent, which would decrease the nondependent-lung HPV response from a 50 to 40 per cent nondependent-lung blood flow re duction, which would increase nondependent-lung blood flow from 20 to 24 per cent of total blood flow, causing shunt to increase by 4 per cent of the cardiac output and P a 0 2 to decrease a moderate amount to 205 mm Hg (F,0 2 = 1.0) (see Fig. 88). Table 8-5 shows that one-half MAC isoflurane anesthesia would cause a very small increase in the total one-lung ventilation shunt and a small decrease in P a 0 2 , whereas 2 MAC isoflurane an esthesia would cause a moderate increase in the total one-lung ventilation shunt and a large de crease in P a 0 2 . Since isoflurane causes undesirable hemodynamic effects at high doses (greater than 1 MAC), and moderate doses of fentanyl (20 μg/kg) have a relative absence of hemodynamic effect, isoflurane anesthesia is usually administered in a 1
MAC or less concentration and is often supple mented with moderate doses of narcotics (or vice versa) (see Recommended Anesthesia Induction and Maintenance Drugs and Techniques). A number of important nonanesthetic drug fac tors might make the administration of isoflurane anesthesia have less of an effect on shunting and arterial oxygenation during one-lung ventilation than the preceding analysis suggests. First, and most important, the absolute level of shunt is al most always higher in surgical patients than the minimal 20 per cent used in the preceding analysis of one-lung ventilation (see Table 4-1). The effect of a given increase in shunt on P a 0 2 depends on the absolute level of the initial shunt and the in spired oxygen concentration (see Fig. 8-5). '°2 With an F,0 2 of 1.0, an increase in shunt from 20 to 24 per cent of the cardiac output decreases the P a 0 2 a moderate amount. However, if the two-lung and one-lung ventilation shunt is increased, per haps owing to pre-existing or anesthesia-induced lung disease, the same isoflurane-induced increase in shunt will cause much less of a decrease in P a O ; (the larger isoshunt lines of Fig. 8-8 are much flatter and closer together). For example, if the one-lung ventilation shunt is 30 per cent without isoflurane and 34 per cent with isoflurane, the de crease in P a 0 2 will be very small and perhaps not detectable, given the usual accuracy of clinical methodology. In fact, in clinical one-lung ventilation studies involving intravenously anesthetized patients with this level of shunting, administration of 1 MAC isoflurane (and halothane) anesthesia during stable one-lung ventilation conditions causes no detect able decrease in P a 0 2 (Figs. 8-9 and 8-10). 6 5 · 6 7 In one of these clinical studies,65 stable one-lung ven tilation conditions in the lateral decubitus position were established in patients who were anesthetized with only intravenous drugs (see Fig. 8-9). While stable one-lung ventilation was maintained, inha-
Abbreviations: HPV = hypoxic pulmonary vasoconstriction; MAC = minimum alveolar concentration.
Choice of Anesthetic Drugs and Techniques
HALOTHANE GROUP
319
ISOFLURANE GROUP
Figure 8-9 Changes in P a O : throughout the entire experimental sequence (experimental steps 1-5). Patients are divided according to the inhalational anesthetic drug they received (halothane or isoflurane). (2-LV = two-lung ventilation; 1-LV = one-lung ventilation; IV = intravenous anesthesia; IH = inhalational anesthesia. Open circles = individual patient data; closed circles = mean ± standard deviations for each group.) (From Rogers SN, Benumof JL: Halothane and isoflurane do not decrease PaO: during one-lung ventilation in intravenously anesthetized patients. Anesth Analg 64:946-954, 1985. Used with permission.)
lational anesthetics were administered (halothane and isoflurane end-tidal concentrations were greater than 1 MAC for at least 15 min) and then discontinued (halothane and isoflurane end-tidal concentrations decreased to near zero) (see Fig. 8-9). In the other clinical study,67 steady-state onelung ventilation conditions in the lateral decubitus position were established in patients who were anesthetized with only inhalational drugs (halothane and isoflurane end-tidal concentrations were constantly greater than 1 MAC for more than 40 min) (see Fig. 8-10). While one-lung ventilation was continued, an intravenous anesthesia was administered (halothane and isoflurane end-tidal concentrations decreased to near zero) (see Fig. 8-10). There was no significant difference in P a O : during inhalation anesthesia with either halothane or isoflurane compared with intravenous anesthesia during one-lung ventilation in either of the two experimental sequences (see Figs. 8-9 and 8-10). In addition, there were no significant changes in physiologic variables, such as cardiac output, pul-
monary vascular pressure, and mixed venous oxygen tension, that might secondarily alter nondependent-lung HPV. Thus, irrespective of whether inhalational anesthesia is administered before (see Fig. 8-10) or after (see Fig. 8-9) intravenous anesthesia during one-lung ventilation, inhalation anesthesia does not further impair arterial oxygenation. These findings are consistent with the interpretation that 1 MAC halothane and isoflurane do not inhibit HPV in patients with a moderate level of shunting enough to cause a significant decrease in P a 0 2 during one-lung ventilation in the lateral decubitus position. Further considering the implications of Figure 8-8, some anesthesiologists use an F,0 2 less than 1.0 during one-lung ventilation (which I do not recommend). As can be seen from Figure 8-8, the family of isoshunt lines is much closer together at F,0 2 = 0.5 than at F,0 2 = 1.0, and the decrease in P a 0 2 with a given increase in shunt is much less (but the absolute level of P a 0 2 is uncomfortably low). Second, as pointed out by Domino et al.,77 the
320 Choice of Anesthetic Drugs and Techniques
Figure 8-10 P a 0 2 and Qs/Q( values during the four experimental sequence steps. Conversion from two-lung ventilation (2-LV) to one-lung ventilation (1-LV) during inhalation anesthesia (IH) in both groups (step 1 to step 2) caused a very large and significant decrease in PaO, and increase in Qs/Q,. Conversion from IH to intravenous anesthesia (IV) during 1-LV (step 2 to step 3) caused a slight but significant increase in P^CX and decrease in Qs/Q, in the halothane group and a very slight and nonsignificant increase in P a 0 2 and decrease in Qs/Q, in the isoflurane group. Conversion from 1-LV to 2-LV during IV anesthesia (step 3 to step 4) caused a very large and significant increase in P a 0 2 and decrease Qs/Q, in both the halothane and isoflurane groups. The small numbers to the left of the first data point refer to the patient number. (From Benumof JL, Augustine SD, Gibbons JA: Halothane and isoflurane only slightly impair arterial oxygenation during one lung ventilation in patients undergoing thoracotomy. Anesthesiology 67:910— 915, 1987. Used with permission.)
secondary effects of anesthesia with isoflurane may counteract the direct HPV-inhibitory effect of the drug. Thus, a decrease in cardiac output, mixed venous oxygen tension, and pulmonary artery pressure, all of which may accompany isoflurane anesthesia, would intensify nondependent-lung HPV at the same time isoflurane was decreasing it. Third, the presence of chronic irreversible disease in the vessels of the nondependent lung may render these vessels incapable of an HPV response.104· '°5 Fourth, the presence of disease in the dependent
.
lung (either pre-existing or anesthesia-induced), which increases dependent-lung vascular resistance, will make the dependent lung less able to accept redistributed blood flow and thereby decrease the nondependent-lung HPV response.101, 106-108 Y^g s m a u e r me j-jVP response, the less of an effect isoflurane anesthesia can have on the HPV response. Fifth, surgical interference with blood flow to the nondependent lung will also decrease the effect that isoflurane anesthesia can have on the one-lung ventilation shunt. Sixth, species differences70-72·97 and differences in the study and clinical one-lung ventilation methodology (nitrogen ventilation vs. atelectasis, administration of isoflurane to the hypoxic lung vs. the normoxic or hyperoxic lung, and large vs. small alveolar-to-mixed venous isoflurane tension gradients, respectively) may alter the precise relationship between per cent inhibition of single-lung HPV and the alveolar concentration of isoflurane. In summary, as demonstrated by Domino et al., isoflurane anesthesia has a direct inhibiting effect on regional HPV in dogs.77 In the simple case, in which physiologic variables (cardiac output, mixed venous oxygen tension, pulmonary vascular pressures, and carbon dioxide tension) are normal and the amount of lung disease is minimal, the effect of isoflurane on shunting during one-lung ventilation is reasonably predictable and moderately small. In the complex case, in which physiologic variables are abnormal and/or the amount of lung disease is extensive, the effect of isoflurane on shunting during one-lung ventilation is much less predictable but almost certainly still small. Nevertheless, it should be remembered that it is the compromised patient who will be most intolerant of any further anesthesia-induced inhibition of HPV. For this kind of patient, the effect of isoflurane anesthesia on shunting must be carefully considered, arterial oxygenation must be closely monitored, and therapeutic measures to decrease shunting, such as nondependent-lung, continuous positive airway pressure (CPAP) and return to two-lung ventilation, should be quickly instituted, if necessary.
IV. RECOMMENDED ANESTHESIA INDUCTION AND MAINTENANCE DRUGS AND TECHNIQUES A. Summary of Advantages of Anesthetic Drugs 1. Inhalational Anesthetics General anesthesia with controlled ventilation is the safest method of anesthetizing patients for the
Choice of Anesthetic Drugs and Techniques
321
vast majority of elective thoracic procedures. Al though a variety of general anesthesia techniques can be used, the volatile halogenated anesthetic drugs are good choices for several reasons. First, the halogenated drugs have a salutary effect on airway irritability. The mechanism of this action is controversial, but, as previously discussed, there is evidence that these drugs can block specific forms of bronchoconstriction 12 · l09 as well as have a non specific bronchodilating effect that is related to the depth of anesthesia.8 Obtundation of airway re flexes in patients who have reactive airways (i.e., smokers)" 0 - " 2 and who may have their airways directly manipulated by the surgeon is a highly desirable property of the general anesthesia pro duced by these drugs. Second, the use of volatile halogenated drugs allows for delivery of a high inspired oxygen con centration without loss of anesthesia. Although a nitrous oxide-oxygen-narcotic-relaxant anesthesia technique can be used, nitrous oxide necessitates a significant decrease in the inspired oxygen concen tration and increases the chance of developing hy poxemia (especially if one-lung ventilation is em ployed)." 1 Unless very high doses of narcotics are used, airway reflexes and reactivity may remain at a high level. Third, since the volatile halogenated drugs can be rapidly eliminated, concern over postoperative hypoventilation in extubated patients may be di minished. Doses of intravenous anesthetics, such as the narcotics, ketamine, and the barbiturates, which render the patient areflexic to surgical stim ulation, may cause the patient to require a period of postoperative ventilation. Fourth, in the usual clinical doses (near 1 MAC), the halogenated anesthetic drugs provide a reasonable degree of cardiovascular stability. This may be of particular importance in patients with coronary artery disease and systemic hypertension. Fifth, the halogenated drugs do not appear to decrease P a 0 2 any more than intravenous anes thetics during one-lung ventilation (see the following).65·67
ronium (0.1 mg/kg), and the anesthetic state main tained with titratable infusions of propofol (3-4 mg/kg/hour), alfentanil (1.0 μg/kg/min), and ve curonium." 4 The lungs may be ventilated with ox ygen or an air/oxygen mixture to avoid all inhalational anesthetic agents. At the termination of surgery, all infusions are discontinued, residual muscle relaxation is reversed, and arousal occurs promptly, usually within 10 min." 5 Total intravenous anesthesia has several acute intraoperative benefits, and most of these benefits have to do with the elimination of nitrous oxide. Intraoperative risks attendant to nitrous oxide use include the limitation of inspired oxygen concen tration, exacerbation of air embolus or pneumo thorax, diffusion into closed body cavities such as the inner ear, gastrointestinal tract, or emphyse matous bullae, elevation of PVR and pulmonary artery pressure in predisposed patients, bone mar row suppression after prolonged administration, and elevation of intracranial pressure in patients with pre-existing intracranial hypertension. Total intravenous anesthesia also has several disadvantages. Whenever fixed doses of intrave nous agents are used, their pharmacokinetic and pharmacodynamic effects vary according to the patient's physiology. A given dose of an anesthetic will affect a patient in hemorrhagic shock differ ently than a hemodynamically stable patient. The hemodynamic state may vary greatly during any surgical procedure (as in thoracic surgery), espe cially those in which large fluid shifts may occur. Unlike a volatile agent, once an intravenous drug is given, it cannot be recovered. With volatile agents, their partial pressure in the blood can be estimated by monitoring end-tidal concentrations; no equivalent real-time measures of intravenous drug blood concentrations exist. Should hidden blood loss occur, profound circulatory depression may occur as a result of deepening of anesthesia. Because the anesthetic state depends on the intra venous infusion of drugs, the technique seems to require a second dedicated intravenous catheter. Finally, intraoperative awareness is a very impor tant problem.
2. Intravenous Anesthetics
New short-acting hypnotics and analgesics are available with a wide therapeutic ratio that show little accumulation and that are suitable for contin uous infusion. Propofol and alfentanil are both suitable to be given by infusion because their phar macokinetic profiles predict little cumulative ef fects within therapeutic dosage ranges. Propofol has a rapid onset of action; permits rapid awaken ing, resulting in clear-headed, alert, oriented pa tients; provides smooth maintenance anesthesia that is easy to control, provides stable hemody namics, and does not inhibit HPV96; therefore, it provides excellent P a 0 2 during one-lung ventila-
The concept of balanced intravenous anesthesia is not new. Specific anesthetic drugs are used to produce specific effects that make up the anes thetic state, namely hypnosis, obtundation of au tonomic reflexes, and muscular relaxation. For one example of total intravenous anesthesia, hypnosis can be induced with a loading dose of propofol (2 mg/kg), the pressor response to intubation atten uated with a loading dose of alfentanil (10-15 μg/kg or bolus 10 μg/kg/min for 10 min), paraly sis and tracheal intubation facilitated with vecu
322 Choice of Anesthetic Drugs and Techniques
tion."6 An example of an infusion strategy for general surgery was given above previously, and slightly different strategies for thoracic surgery are given next. The narcotics, especially alfentanil and fentanyl, have a number of desirable properties that could be used to advantage for patients undergoing tho racic surgery. First, fentanyl and alfentanil have no adverse hemodynamic effects and, therefore, are useful in patients who have significant coronary artery disease. Second, if significant blood levels exist at the end of surgery, the narcotics can allow an intu bated patient to have a smooth transition from surgery into the postoperative period. Third, the narcotics, if used in moderate dosage, greatly diminish the amount of volatile halogenated drug anesthesia required to achieve surgical levels of anesthesia."7 Fourth, high doses of narcotics or moderate doses in conjunction with halogenated drugs allow for the use of a high inspired oxygen concentration without loss of anesthesia. Fifth, the narcotics are thought not to diminish regional HPV and, therefore, should permit opti mal oxygenation during one-lung ventilation. Ketamine, in combination with nitrous oxide and a muscle relaxant, has also been used for an esthesia in thoracic surgery."8 Although we do not ordinarily use ketamine for elective thoracic pro cedures, the drug is very useful for the induction of general anesthesia in critically ill patients undergoing emergency thoracic surgery for several reasons. First, ketamine has sympathomimetic properties"9 that are highly desirable because many emergency thoracic procedures are associ ated with hypovolemia (gunshot and stab wounds of the chest, blunt trauma, and massive hemopty sis). However, it should be remembered that keta mine will depress cardiovascular function (sys temic blood pressure, contractility) if the degree of hypovolemia is severe and the patients are sympa thetically exhausted. Second, ketamine has a rapid onset of action and can be used safely, along with cricoid pres sure, to induce anesthesia in patients with full stomachs. Third, ketamine may reduce bronchospasm in asthmatic patients21-24·26· l2°; in view of the number of reports describing decreased airway resistance in patients with bronchospasm with ketamine in fusion, it seems reasonable to extrapolate these findings to thoracic surgery patients. Fourth, ketamine does not impair arterial oxy genation during one-lung ventilation (perhaps ow 95 ing to lack of effect on HPV). Rapid sequence induction is used to secure the airway in emergency surgical procedures to avoid
aspiration of gastric contents. Succinylcholine is ordinarily the relaxant of choice because of its rapid onset of action (1 min to full relaxation). However, succinylcholine may produce muscle fasciculations, myalgia, and increases in intragas tric, intraocular, and intracranial pressures. Succi nylcholine should also be avoided in patients with burns, crush injuries, and renal failure, in patients who have chronic lack of use of muscles, and in patients with a history of malignant hyperthermia or pseudocholinesterase deficiency. Fortunately, there have been several new devel opments that have facilitated the speed of onset and safety of inducing and maintaining neuromus cular blockade with nondepolarizing muscle relax ants. First, but least important because of unpre dictability, is use of the priming principle. For example, a priming dose of 15 μg/kg of vecuron ium with 100 μg/kg total dose, produces excellent intubating conditions in a mean onset time (75 ± 10 sec) not significantly different from that of suc cinylcholine, 1.5 mg/kg (58 ± 4 sec).121 However, it should be remembered that occasionally the priming dose of a nondepolarizing muscle relaxant may produce an undesirable degree of blockade and attendant problems (inadequate respiration, as piration). Second, the administration of a small dose of alfentanil (30 μg/kg) decreases the incidence of unacceptable intubating conditions to 4 per cent when intubation is performed 1 min after admin istration of alfentanil and an intubating dose of vecuronium (priming dose, priming interval, and intubating dose of vecuronium were 0.01 mg/kg, 4 min, and 0.1 mg/kg, respectively).122 Preintubation administration of lidocaine may also be useful in 123 this regard. Third, use of triple to quadruple the usual rec ommended intubating dose of a nondepolarizing relaxant can increase the speed of onset of com plete blockade. For example, the ED95 (i.e., the effective dose for 95 per cent of patients) dose of mivacurium is 0.7 mg/kg, which yields good to excellent intubating conditions in 3.3 min, whereas 0.15 mg/kg and 0.2 to 0.25 mg/kg yields good to excellent intubating conditions in 2.5 and 2.0 min, respectively.124-126 Fourth, the new short- and intermediate-acting nondepolarizing muscle relaxants (mivacurium, vecuronium, and atracurium) can be easily admin istered by continuous infusion, which results in easy and complete reversal. Average initial bolus and continuous infusion doses for mivacurium, ve curonium, and atracurium are 0.15 mg/kg and 6 to 7 μg/kg/min, 80 μg/kg and 1.0 μg/kg/min, and 0.5 mg/kg and 6 μg/kg/min, respectively; in all cases with all drugs, the continuous infusion rate should be adjusted according to the findings with a neuromuscular blockade monitor.
Choice of Anesthetic Drugs and Techniques 323
B. Recommended Anesthetic Drugs and Technique It should be obvious from the foregoing discus sions that there are advantages and disadvantages to both the inhalation and intravenous anesthetic drugs. One study made this point very well by examining the respiratory and cardiac differences between various anesthetic regimens (intravenous based, inhalation based, epidural based) before and during one-lung ventilation.127 The patients (n = 36) were randomly allocated to one of the follow ing three groups: group A: propofol, 10 mg/ kg/hour, fentanyl; group B: 1 MAC enflurane, fentanyl; group C: thoracic epidural anesthesia, 0.4 per cent enflurane. Before induction of anesthesia, significant differences were not found. During two-lung ventilation, cardiac index was signifi cantly decreased in group Β patients in comparison with those in group C (p < .01). During one-lung ventilation, significant differences were not found except for an increased shunt fraction (Qs/Qt) in group Β patients (groups A-B: ρ < .05; groups B-C: ρ < .05; groups A-C: not significant). Be cause Qs/Q, was significantly increased and hypox emia occurred in group Β patients, regimen A or C might be preferred in patients at high risk for hypoxia or when the application of CPAP to the nondependent lung is not possible (see chapter 11). Cardiac index was best maintained in group C patients. Regimen C might be of value in pa tients at high risk for poor perfusion, taking into account both the possible complications of the ep idural block and the reduced need for postopera tive analgesic agents. The following recommended anesthetic tech nique takes advantage of the desirable properties and minimizes the undesirable properties of these drugs. Thus, the halogenated drugs are used for their effect on bronchomotor tone, to administer 100 per cent oxygen, and to allow for early extubation while not decreasing hemodynamic func tion and arterial oxygenation, whereas fentanyl is used to ensure hemodynamic stability while not jeopardizing early extubation if desired. If it is thought that the patient will not be extubated early or if greater hemodynamic stability is desired, an esthesia consisting of more fentanyl and less halo genated drug can be used.
1. Insertion of an Epidural Catheter Before Induction of Anesthesia Epidural anesthesia has been reported to exert beneficial effects in various surgical procedures, including thoracic operations.128 The definite ad vantages (major and consistently reported) are a great reduction in the intraoperative requirements for general anesthetics; a great decrease, if not
complete elimination, in the postoperative require ment for systemic analgesics and improved post operative ventilatory function (if the catheter is used postoperatively); and improved cardiac per formance if local anesthetic is used (causes de creased afterload because of sympathetic block ade). Other possible benefits include decreased incidence of venous thromboembolism as well as blood loss, better suppression of stress responses, and a positive effect on postoperative nitrogen bal ance. Potential disadvantages include the time re quired to establish the epidural anesthesia, intra vascular volume needed to avoid a decrease in blood pressure, and technical complications such as epidural hematoma. Reported levels of insertion and preoperative drugs and dosages of drugs suc cessfully used are TI2_M and 12 to 15 ml of 0.5 per cent bupivacaine,128 L2_3 and fentanyl 4 μg/kg in 20 ml saline 40 min before incision,129 T6_7 and fentanyl 250 μg, l 3 0 T4_5 and 50 μg sufentanil in 7 ml saline or 8 ml 0.5 per cent bupivacaine.131
2. Induction of Anesthesia (Fig. 8-11) The patient is preoxygenated by spontaneously breathing 100 per cent oxygen through a black rubber anesthesia mask that is connected to an anesthesia circle system. The preoxygenation or denitrogenation process with tidal breathing (typi cally 3-5 min) may be greatly speeded up by hav ing the patient take several (four to eight) deep or maximal breaths. The exact comparison/result of a few deep breaths with several minutes of tidal breathing with respect to preinduction P a 0 2 or S a 0 2 and time to various desaturation levels will obviously depend on how many deep breaths were taken and the duration of the tidal breathing. 132 " 136 Fentanyl is administered intravenously until the respiratory rate is approximately 8 to 10 breaths/min. This usually corresponds to a dose of 10 to 15 μg/kg and is usually administered over 3 min. When the respiratory rate is relatively slow and deep, and response to commands are becom ing sluggish, a small dose of thiopental (Pentothal; 2 to 3 mg/kg) or ketamine (1 to 2 mg/kg) (if the patient is thought to have an especially reactive airway or to be minimally to moderately hypovo lemic) is administered, which renders the patient unconscious and usually apneic. The airway is then established, and ventilation is controlled with intermittent positive-pressure oxygen via the black rubber mask. While the patient is being ventilated with positive pressure, concentrations of 2.5 to 0.5 per cent isoflurane are administered. The higher isoflurane concentration is used initially for a short period of time (overpressure, 1 to 2 min), and as the patient demonstrates signs of deepening anes thesia, the inspired isoflurane concentration is de creased. In view of the fact that general anesthetics
324
Choice of Anesthetic Drugs and Techniques
Anesthetic Technique for Typical One Lung Ventilation Thoracic Surgery Cases ACTION Airway
ENDPOINT Drug
Laryngoscopy Lidocaine I n t r a t r a c h e a l ^ Double Lumen Tube Intubation
Administer Maintenance Isoflurane, Fentanyl, Paralysis Figure 8-11 This action-end point flow diagram describes the anesthetic technique used for a typical one-lung ventilation thoracic surgery case. The actions are divided into those that are primarily concerned with airway management versus those that involve administration of a" drug. All drug administrations are associated with an end point. See text for full explanation. (IPPB = intermittent positive-pressure breathing.)
Choice of Anesthetic Drugs and Techniques
significantly decrease the ventilatory response to C 0 2 (to a much greater degree in patients with mechanical ventilatory impairment compared with normal patients), patients are not allowed to breathe spontaneously until the end of the proce dure; alarming degrees of hypercapnia have been observed in similar circumstances when sponta neous ventilation was allowed.137 Early during the period of positive-pressure ventilation with isoflurane, paralysis is induced with either pancuronium 0.02 mg/kg and metocurine 0.08 mg/kg, or vecuronium 0.1 mg/kg, or atracurium 0.5 mg/kg. The development of full paralysis is monitored with a neuromuscular blockade monitor. During the period of deepening isoflurane anesthesia and paralysis, blood pressure is supported with an infusion of approximately 10 ml/kg crystalloid. If more cardiovascular support is required, the first drugs used (while fluid is being infused) are ephedrine, 0.05 to 0.1 mg/kg, atropine, 0.02 mg/kg, and calcium, 5 mg/kg. In patients with stable angina undergoing operations of short duration, the use of nitroglycerin infusion (0.9 μg/kg) and low-dose fentanyl (3 μg/kg) be fore laryngoscopy and intubation significantly de creases the incidence of myocardial ischemia as sociated with induction of anesthesia and tracheal intubation (compared with fentanyl, 8 μg/kg alone).138 When the patient has been judged to be ade quately anesthetized (surgical stage as judged by changes in blood pressure, heart rate, and eye signs [the eyes should be central, conjugate, fixed, star ing, without tears and with nondilated pupils]) and paralyzed in the previously described manner, lidocaine, 1.5 mg/kg,139· , 4 0 is administered intrave nously (optimal time of injection is 3 min before tracheal intubation),141 laryngoscopy is performed, the tracheobronchial tree sprayed with a laryngotracheobronchial spray system, and the trachea in tubated with a double-lumen tube (see chapter 9). The intravenous and intratracheal lidocaine should diminish both the airway and cardiovascular re sponse to endotracheal intubation. 139-142 If the pa tients have received a higher dose of fentanyl (25 μg/kg), then intratracheal lidocaine alone is suffi cient and preferable.143 Although lidocaine and fentanyl protect against increases in blood pressure, intravenous esmolol (5-10 mg) will also protect against increases in heart rate in addition to increases in blood pressure (half-life = 9 min). 1 4 4 - 1 4 7 This is an important con sideration because tachycardia may be more stren uous on the heart than an increase in afterload.148 Esmolol also decreases the incidence of arrhyth 147 mias attributable to tracheal intubation. Conse quently, rapid intravenous injection of esmolol before rapid sequence induction is a simple, con
325
venient, and effective technique that should be considered for blunting hemodynamic responses to intubation. After intubation, the patient is venti lated with maintenance doses of isoflurane and administered maintenance doses of narcotics and relaxants. Use of maintenance paralysis decreases isoflurane requirements, possibly allowing for a more rapid emergence from anesthesia.149
3. Maintenance of Anesthesia Anesthesia is maintained with both isoflurane (concentration approximately 0.5 to 1.0 MAC) and narcotics. Isoflurane is primarily used if the patient is thought to stand a reasonable chance of being extubated within the first couple of hours postop eratively. Narcotics (fentanyl) are primarily used if the patient is thought not to have a reasonable chance of being extubated in the immediate post operative period and will require a significant pe riod of postoperative ventilation. Relaxants (pri marily pancuronium) are administered in small doses to keep the level of neuromuscular blockade, as judged by a neuromuscular blockade monitor, near the 90 per cent paralysis level. If the patient is thought to have a reasonable chance of being extubated in the first postoperative hour, the patient is turned supine, the double-lu men tube changed to a single-lumen tube, paraly sis reversed, and spontaneous ventilation allowed to recur. Fentanyl is administered in extremely small increments (0.3 μg/kg) while the patient is breathing spontaneously. The goal of the fentanyl administration is to have the patient breathing relatively slowly (approximately 10 to 12 breaths/min) and deeply when surgery is com pleted. The presence of a moderate narcotic base allows the patient to be returned to the recovery room for a short period of mechanical ventilatory support (if needed) and weaned and extubated in a relatively smooth manner.
4. Total Intravenous Anesthesia Technique Propofol's well-known pharmacokinetic prop erties make this drug most suitable for infusion techniques. The same is true of alfentanil; these two drugs together form an ideal combination for total intravenous anesthesia inasmuch as one pro vides anesthesia and the other hypnosis. There have been two studies using total intravenous an esthesia for thoracic surgery. In one study," 6 an alfentanil bolus of 1 mg plus lidocaine, 1 mg/kg, was given intravenously. This was followed by propofol, 2 to 2.5 mg/kg, and atracurium, 0.5 mg/kg. After intubation, two infusions, one with propofol and one with alfentanil, were started im-
326 Choice of Anesthetic Drugs and Techniques
mediately thereafter by way of two separate infu sion pumps. Alfentanil was given at a rate of 3 μg/kg/min for 10 min followed by a constant in fusion of 1 μg/kg/min. The infusion was discon tinued approximately 20 min before the end of the operation. The propofol infusion was started at 0.2 mg/kg/min (12 mg/kg/hour). Thereafter, it was varied to adjust the depth of anesthesia. Between 0.1 and 0.15 mg/kg/min (6-9 mg/kg/hour) was required, and the infusion was discontinued at the time of skin closure. Extra boluses of 20 mg of propofol were given in addition when blood pres sure and pulse rate increases made them necessary, together with repeat doses of the muscle relaxant when slight movements occurred. Not surpris ingly, the other important total intravenous anes thesia for thoracic surgery study (propofol with or without nitrous oxide) showed that the total amount of propofol infused could be decreased from 7.2 ± 2.7 mg/kg/hour with air oxygen to 5.7 ± 2.0 mg/kg/hour with nitrous oxide oxygen.150 REFERENCES 1. Aviado DM: Regulation of bronchomotor tone during anesthesia. Anesthesiology 42:68, 1975. 2. Fielding JE: Smoking: Health effects and control. Ν Engl J Med 313:491^198, 1985. 3. Furchgott RF, Vanhoutte PM: Endothelium-derived re laxing and contracting factors. Fed Am Soc Exp Biol J 3:2007-2018, 1988. 4. Munakata M, Huang I, Mitzer W, Menkes H: Protective role of epithelium in the guinea pig airway. J Appl Phys iol 66:1547-1552, 1989. 5. Hirschmann CA, Bergman NA: Factors influencing intrapulmonary airway calibre during anaesthesia. Br J Anaesth 65:30-42, 1990. 6. Bennett DJ, Torda TA, Horton DA, et al: Severe bronchospasm complicating thoracotomy. Arch Surg 101:555, 1970. 7. Brakensiek AL, Bergman NA: The effects of halothane and atropine on total respiratory resistance in anesthetized man. Anesthesiology 53:341, 1970. 8. Patterson RW, Sullivan SF, Malm JR, et al: The effects of halothane on human airway mechanics. Anesthesiol ogy 29:900, 1968. 9. McAslan C, Mima M, Norden I, et al: Effect of halothane and methoxyflurane on pulmonary resistance to gas flow during lung bypass. Scand J Thorac Cardiovasc Surg 5:193, 1971. 10. Waltemath CL, Bergman NA: Effect of ketamine and halothane on increased respiratory resistance provoked by ultrasonic aerosols. Anesthesiology 41:473, 1974. 11. Meloche R, Norlander O, Norden I, Herzog P: Effects of carbon dioxide and halothane on compliance in pulmo nary resistance during cardiopulmonary bypass. Scand J Thorac Cardiovasc Surg 3:69, 1969. 12. Hirshman CA, Edelstein G, Peetz S, Wayne R, Downes H: Mechanism of action of inhalational anesthesia on airways. Anesthesiology 56:107, 1982. 13. Vettermann J, Warner DO, Brichant JF, Rehder K: Halo thane directly relaxes airway smooth muscle. Anesthe siology 71 :A1070, 1989. 14. Vettermann J, Beck KC, Lindahl SHE, Brichant JF, Reh der K: Actions of enflurane, isoflurane, vecuronium, atra-
curium and pancuronium on pulmonary resistance in dogs. Anesthesiology 69:688-695, 1988. 15. Revell S, Greenhalgh D, Absalom SR, Soni N: Isoflurane in the treatment of asthma. Anaesthesia 43:477^179, 1988. 16. Parnass SM, Feld JM, Chamberlin WH, Segil LJ: Status asthmaticus treated with isoflurane and enflurane. Anesth Analg 66:193-195, 1987. 17. Bierman MI, Brown M, Muren O, Keenan RL, Glauser FL: Prolonged isoflurane anesthesia in status asthmaticus. Crit Care Med 14:832-833, 1986. 18. Johnston RG, Noseworthy TW, Friessem EG, Yule HA, Shustock A: Isoflurane therapy for status asthmaticus in children and adults. Chest 97:698-701, 1990. 19. Moss J, Rosow CE, Savarese JJ, et al: Role of histamine in the hypotensive action of d-tubocurarine in humans. Anesthesiology 55:19-25, 1981. 20. Shemano I, Wendel H: Effects of meperidine hydrochlo ride and morphine sulfate on the lung capacity of intact dogs. J Pharmacol Exp Ther 149:379-384, 1965. 21. Hirshman CA, Downes H, Farbood A, et al: Ketamine block of bronchospasm in experimental canine asthma. BrJ Anaesth 51:713-718, 1979. 22. Ben-Harari RR, Bansinath M: Reinvestigation of the mechanism of bronchodilation by ketamine. Anesthesiol ogy 75: A627, 1991. 23. Corssen G, Gutierrez J, Reeves JG, et al: Ketamine in the anesthetic management of asthmatic patients. Anesth An alg 51:588-596, 1972. 24. Hemmingsen C, Nielsen PK, Odorico J: Ketamine in the treatment of bronchospasm during mechanical ventila tion. Anesth Analg 74:S135, 1992. 25. Clarke RSJ, Dundee JW, Garrett RT, McArdle GK, Sut ton JA: Adverse reactions to intravenous anesthesia. Br J Anaesth 47:575, 1975. 26. Rees DI, Howell ML: Ketamine-atracurium by continu ous infusion as the sole anesthetic for pulmonary surgery. Anesth Analg 65:860-864, 1988. 27. Crago RR, Bryan AC, Laws AIC, Winestock AE: Respi ratory flow resistance after curare and pancuronium meas ured by forced oscillations. Can Anaesth Soc J 19:607614, 1972. 28. Brandus V, Joffe S, Benoit CV, Wolff WI: Bronchial spasm during general anesthesia. Can Anaesth Soc J 17:269-274, 1970. 29. Jarnberg PO, Andersen P, Cohen J, Jamond M, Smith J: Intravenous lidocaine inhibits induced laryngospasm in dogs during halothane anesthesia. Anesthesiology 75:A971, 1991. 30. Gal TJ: Airway responses in normal subjects following topical anesthesia with ultrasonic aerosols with 4 per cent lidocaine. Anesth Analg 59:123-129, 1980. 31. Loehning RW, Waltemath CL, Bergman NA: Lidocaine and increased respiratory resistance produced by ultra sonic aerosols. Anesthesiology 44:306-310, 1976. 32. Don HF, Robson JG: The mechanics of the respiratory system during anesthesia: The effects of atropine and carbon dioxide. Anesthesiology 26:168-178, 1965. 33. Severinghaus JW, Stupfel M: Respiratory dead space in crease following atropine in man and atropine, vagal or ganglionic blockade and hypothermia in dogs. J Appl Physiol 8:81-87, 1955. 34. U.S. Department of Health and Human Services: The health consequences of smoking: Cardiovascular disease. A report of the Surgeon General. Rockville, MD, U.S. Department of Health and Human Services, 1983. 35. The Pooling Project Research Group: Relationship of blood pressure, serum cholesterol, smoking habit, relative weight, and ECG abnormalities to incidence of major coronary events: Final report of the Pooling Project. J Chronic Dis 31:201-306, 1978.
CHAPTER
9
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
I. Introduction II. Indications for Separation of the Two Lungs A. Absolute Indications B. Relative Indications III. Double-Lumen Tube Intubation A. Various Double-Lumen Endotracheal Tubes 1. Robertshaw Double-Lumen Endotracheal Tube a. Original Red Rubber Robertshaw Double-Lumen Endotracheal Tube b. Modern Plastic Disposable Robertshaw Double-Lumen Endotracheal Tube B. Conventional (Nonfiberoptic) Double-Lumen Tube Intubation Procedure 1. Choice of Left- Versus RightSided Double-Lumen Tube 2. Choice of Double-Lumen Tube Size 3. Double-Lumen Tube Intubation Procedure a. Preintubation Procedures b. Intubation c. Routine Checking of DoubleLumen Tube Position d. Auscultation and Unilateral Clamping Maneuvers to Diagnose Double-Lumen Tube Malposition e. Nonfiberoptic Method for Intubation of Left Main-Stem Bronchus When a Left-Sided Tube Will Not Pass Routinely C. Use of Fiberoptic Bronchoscope to Insert the Bronchial Lumen of a Double-Lumen Tube Into a MainStem Bronchus D. Use of Fiberoptic Bronchoscope to Determine Precise Double-Lumen Tube Position 1. Margin of Safety in Positioning Double-Lumen Tubes a. Left-Sided Double-Lumen Tubes (1) Definition of Margin of
330
Safety in Positioning Left-Sided DoubleLumen Tubes (2) Measurement of Length of Left Main-Stem Bronchus (3) Measurement of Margin of Safety for Left-Sided Double-Lumen Tubes (4) Clinical Examples/ Implications of Exceeding Margin of Safety for Left-Sided Double-Lumen Tubes b. Right-Sided Double-Lumen Tubes (1) Definition of Margin of Safety in Positioning Right-Sided DoubleLumen Tubes (2) Measurement of Length of Right Main-Stem Bronchus and Right Upper Lobe Bronchial Orifice (3) Measurement of Margin of Safety for Right-Sided Double-Lumen Tubes (4) Clinical Examples/ Implications of Exceeding Margin of Safety for Right-Sided Double-Lumen Tubes 2. Relationship of Fiberoptic Bronchoscope Size to DoubleLumen Tube Size Use of Chest X-Ray to Determine Double-Lumen Tube Position Other Methods to Determine Double-Lumen Tube Position Securing the Double-Lumen Tube Quantitative Determination of CuffSeal Pressure Hold Complications of Double-Lumen Endotracheal Tubes Relative Contraindications to the Use of Double-Lumen Endotracheal Tubes
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
IV. Bronchial Blockers (With SingleLumen Endotracheal Tubes) A. Univent Bronchial Blocker Tube 1. Description of Univent Bronchial Blocker Tube 2. Insertion of Univent Tube and Positioning of Bronchial Blocker 3. Advantages/Noteworthy Positive Attributes of Univent Bronchial Blocker Tube System 4. Potential Limitations of Univent Bronchial Blocker Tube System and Solutions
331
5. Methods to Obtain a Just-Seal Volume in Bronchial Blocker Cuff 6. Firm Clinical Indications for Use of Univent Bronchial Blocker System B. Bronchial Blockers That Are Independent of a Single-Lumen Tube V. Endobronchial Intubation With Single-Lumen Tubes
I. INTRODUCTION
A. Absolute Indications
The complete functional separation of the two lungs is often the most important anesthetic consideration for patients undergoing thoracic surgery. The procedure can occasionally be life-saving, and very frequently it greatly facilitates the conduct of surgery. Newly introduced disposable plastic double-lumen tubes (DLTs), which are relatively nontraumatic and easy to insert, and the advent of fiberoptic bronchoscopy (FOB), which makes location of a DLT under direct vision possible and therefore a precise, repeatable, low-risk maneuver, have greatly increased the efficacy and use of DLTs. In addition, the new Univent bronchial blocker tube (Fuji Corp., Tokyo, Japan) makes the use of a bronchial blocker much easier and simpler by coupling the insertion of the main single-lumen tube with the insertion of the bronchial blocker; because the single-lumen tube and bronchial blocker are inserted together, the only new skill requirement is to push, under fiberoptic guidance and during continuous ventilation, the bronchial blocker into the desired main-stem bronchus. This chapter sequentially discusses the indications for separation of the two lungs, conventional techniques of DLT insertion, and determination of precise DLT position by FOB. The chapter then considers how to position properly the bronchial blocker of the Univent tube. Next, endobronchial intubation with a single-lumen tube is briefly considered, as are a variety of other ways to separate the two lungs (coaxial tubes, the Wilson tube, Ttube ventilating system).
Separation of the two lungs for any of the absolute indications discussed here should be considered a life-saving maneuver because failure to separate the two lungs under any of these conditions could result in a life-threatening complication or situation. There are three absolute indications for separating the two lungs (Table 9-1 and Fig. 9-1 ). First, separation of one lung from the other is
II. INDICATIONS FOR SEPARATION OF THE TWO LUNGS There are several absolute and relative indications for separation of the two lungs during thoracic operations or procedures (Table 9-1).
Table 9-1
INDICATIONS FOR SEPARATION OF THE TWO LUNGS (DOUBLELUMEN TUBE INTUBATION) AND/ OR ONE-LUNG VENTILATION
Absolute 1. Isolation of one lung from the other to avoid spillage or contamination A. Infection B. Massive hemorrhage 2. Control of the distribution of ventilation A. Bronchopleural fistula B. Bronchopleural cutaneous fistula C. Surgical opening of a major conducting airway D. Giant unilateral lung cyst or bulla E. Tracheobronchial tree disruption F. Life-threatenng hypoxemia caused by unilateral lung disease 3. Unilateral bronchopulmonary lavage A. Pulmonary alveolar proteinosis
Relative 1. Surgical exposure—high priority A. Thoracic aortic aneurysm B. Pneumonectomy C. Thoracoscopy D. Pulmonary resection via median sternotomy E. Upper lobectomy F. Mediastinal exposure 2. Surgical exposure—medium (lower) priority A. Middle and lower lobectomies and subsegmental resections B. Esophageal resection C. Procedures on the thoracic spine 3. Postcardiopulmonary by-pass status after removal of totally occluding chronic unilateral pulmonary emboli 4. Severe hpyoxemia caused by unilateral lung disease
332
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Absolute Indications for Lung Separation/One Lung Ventilation
Figure 9-1 This schematic diagram shows the absolute indications for lung separation and/or one-lung ventilation. Hemorrhage and infection in one lung can contaminate and soil the other lung. A low-resistance ventilation pathway, such as a bronchopleural fistula (BPF) or a surgically opened major airway, can make positive-pressure ventilation of the other lung impossible. A giant unilateral cyst or bulla can rupture if exposed to positive-pressure ventilation and can result in a tension pneumothorax. Unilateral lung lavage requires the instillation of large amounts of saline into one lung while the other lung is being ventilated. Major tracheobronchial disruption can lead to mediastinal and pulmonary interstitial emphysema if exposed to positive pressure.
absolutely necessary to prevent spillage of pus or blood from an infected (abscessed) lung or bleeding lung, respectively, to a noninvolved lung. Acute contamination of a lung with either blood or pus from the other lung usually results in severe massive (bilateral) atelectasis, pneumonia, and sepsis. Second, a number of unilateral lung problems can prevent adequate ventilation of the other noninvolved side. A large bronchopleural or bronchopleural cutaneous fistula or a surgically opened conducting airway has such a low resistance to gas flow that a tidal inspiration delivered by positive pressure will exit via the low-resistance pathway, and it will become impossible to ventilate the other, more normal lung adequately. A giant unilateral bulla or cyst may rupture if exposed to positive-pressure ventilation and result in a tension pneumothorax or pneumomediastinum. Very severe or life-threatening hypoxemia caused by unilateral lung disease may require differential lung ventilation and positive end-expiratory pressure (PEEP).1 Positive-pressure ventilation of a lung with a tracheobronchial tree disruption can result in dissection of gas into the pulmonary interstitial
space or mediastinum, resulting in a tension pneumomediastinum. Third, separation of the two lungs is absolutely necessary to perform unilateral bronchopulmonary lavage in patients with pulmonary alveolar proteinosis (and, rarely, asthma and cystic fibrosis).
B. Relative Indications There are a large number of relative indications for separation of the two lungs, and they are all for the purpose of facilitating surgical exposure by collapsing the lung in the operative hemithorax. These relative indications can be divided into high-priority and medium-, or lower, priority categories (Table 9-1 and Fig. 9-2). Of the relative indications, repair of a thoracic aortic aneurysm is usually the highest priority because it requires exposure of the thoracic aorta as it runs the entire length of the left hemithorax. A pneumonectomy, especially if performed through a median sternotomy,2 is greatly aided by the wide exposure of the lung hilum, which is afforded by collapse of the
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
operative lung. Examination of the pleural space (thoracoscopy) and pulmonary resections through a thoracoscope are considerably aided by collapse of the ipsilateral lung. Similarly, an upper lobectomy, which is technically the most difficult lobectomy, and many mediastinal exposures may be made much easier by eliminating ventilation to the lung on the side of the procedure. The surgical items in the medium-priority category do not routinely require collapse of the lung on the operative side but still significantly aid surgical exposure and eliminate the need for the surgeon to handle (retract, compress, pack away) the operative lung. Severe intraoperative retraction of the lung on the operated side can traumatize the operative lung and impair gas exchange both intra-3 4 and post-
333
operatively.5· 6 The lower priority items consist of middle and lower lobectomies, less extensive pulmonary resections, thoracic spinal procedures that are approached anteriorly through the chest,7 and esophageal surgery. However, even relatively small operations such as wedge and segmental resections benefit by DLT insertion because of the ability to alternate easily and quickly between lung collapse and inflation, which is sometimes required to visualize lung morphology better and to facilitate identification and separation of important planes and fissures. Finally, the separation of the lungs after removal of totally occluding and predominantly unilateral chronic pulmonary emboli (postcardiopulmonary by-pass) can be very helpful if significant transudation of hemorrhagic fluid
Relative Indications for Lung Separation/ One Lung Ventilation
Significant Hypoxemia due to Unilateral Lung Disease Figure 9—2 This schematic diagram depicts the relative indications for lung separation and/or one-lung ventilation. The indications are all for facilitating surgical exposure and can be divided into high-priority and low-priority categories. The high-priority items for lung collapse are thoracic aortic artery aneurysm repair (requires exposure of the entire thoracic aorta, which is greatly facilitated by collapsing the left lung), pulmonary resection via median sternotomy pneumonectomy (requires wide exposure of the lung hilum), thoracoscopy (especially when therapeutic and video-assisted) and upper lobe lobectomy (technically the most difficult lobe to expose). The lower priority items for lung collapse are middle and lower lobe lobectomy and esophageal surgery. Procedures on the thoracic spine that are approached anteriorly via the chest are facilitated by collapse of the lung on the operative side. Lung separation is useful if unilateral pulmonary edema occurs following removal of a chronic, totally occluding, unilateral pulmonary embolus (postcardiopulmonary by-pass).
334
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
across the alveolar capillary membrane in the region of the lung supplied by the previously occluded vessel occurs after cardiopulmonary bypass (reperfusion of a previously and chronically nonperfused vascular bed). Should significant and predominantly unilateral postthromboembolectomy postcardiopulmonary by-pass pulmonary edema occur, the patient should be returned to cardiopulmonary by-pass, and a double-lumen endotracheal tube should be inserted so that differential lung ventilation may be used (see chapters 11 and 20). Finally, significant hypoxemia resulting from unilateral lung disease may be treated more easily by differential lung ventilation and PEEP.1
intraoperative/postoperative risk factor for making laryngoscopy and intubation difficult is the infusion of massive amounts of blood and fluids (the upper airway becomes edematous). There are two relatively minor in situ disadvantages to DLTs related to the fact that the lumens of a DLT may be narrow. First, suctioning may be more difficult down a narrow lumen, but this is usually not a problem with the new disposable Robertshaw-type DLTs, which have nonadhering suction catheters that slide easily down the lumens of the DLT. Second, although airway resistance may be increased with a narrow lumen, the increased airway resistance can be easily overcome by positive-pressure ventilation.9
III. DOUBLE-LUMEN TUBE INTUBATION
A. Various Double-Lumen Endotracheal Tubes
Double-lumen endotracheal tubes have evolved to become considered the lung-separation technique of choice for the majority of thoracic surgery cases and are discussed here in great detail. Bronchial blockade with the Univent tube in adults is greatly increasing and is also described at length. Endobronchial tubes are not often used today and are only briefly described at the end of the chapter. DLTs are favored over bronchial blockers and endobronchial tubes for lung separation primarily because they are more versatile than bronchial blockers or endobronchial tubes. The most important double-lumen function not available with a bronchial blocker is independent bilateral suctioning. In addition, it is easier to apply continuous positive airway pressure (CPAP) to the nonventilated operative lung with a DLT compared with a bronchial blocker. Endobronchial tubes are very limited in function and allow only one-lung ventilation. There are two firm disadvantages/contraindications to the use of a DLT compared with a bronchial blocker. First, very distorted tracheobronchial tree anatomy, including exophytic and stenotic lesions, as well as tortuosity may preclude successful correct placement/positioning of a DLT. Second, changing from a DLT to a single-lumen tube during or at the end of a case can be expected to be a difficult/risky procedure on occasion. Examples of situations in which a change is required from DLT to a single-lumen tube during the operative period, and the need for this change is known to exist preoperatively, are (1) turning the patient from the supine to the prone position (45 per cent of cases involving an anterior approach to surgery on the vertebral bodies7) and (2) for postoperative ventilation in the intensive care unit (ICU). There are many known anatomic causes of difficult laryngoscopy and intubation,8 but the most important
DLTs are essentially two catheters bonded together, and each lumen is intended to ventilate one of the two lungs. DLTs are made as left- and rightsided tubes. A left-sided tube means that the left lung catheter is placed into the left main-stem bronchus, whereas the right lung catheter ends in the trachea; therefore, for a left-sided tube, the left lung catheter is longer than the right lung catheter (Fig. 9-3). A right-sided tube means that the right lung catheter is placed into the right main-stem bronchus, whereas the left lung catheter ends in the trachea; therefore, for a right-sided tube, the right lung catheter is longer than the left lung catheter (Fig. 9-3). All the DLTs have a proximal cuff for the trachea and a distal cuff for a mainstem bronchus; the endobronchial cuff causes separation and sealing off of the two lungs from each other, and the tracheal cuff causes separation and sealing off of the lungs from the environment. The part of the right lung catheter of the right-sided DLT that is in the right main-stem bronchus must be slotted to allow for ventilation of the right upper lobe (Fig. 9-3) because the right main-stem bronchus is too short to accommodate both the right lumen tip and the right endo-bronchial cuff. All the double-lumen endotracheal tubes have two curves that lie in planes approximately 90 degrees apart from one another. The distal curve is designed to facilitate placement of the distal catheter tip into the appropriate main-stem bronchus, and the proximal curve is designed to approximate the oropharyngolaryngeal curve. The DLTs that have been used for lung separation and one-lung ventilation include the Carlens, White, Bryce-Smith, and Robertshaw types. The Robertshaw double-lumen endotracheal tube is by far the most commonly used, and the disposable polyvinylchloride Robertshaw tube has almost completely replaced the red rubber Robertshaw
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker intubation)
335
Double-Lumen Tubes
tube (the former is easier to pass and suction through, is positioned more quickly, has a lower airflow resistance, and causes less mucosal damage).10"12 Consequently, the modern polyvinylchloride tube is described in great detail. The other DLTs (the first three mentioned) are seldom used; however, some anesthesiologists still routinely use the Carlens tube, and some prefer the right-sided red rubber Robertshaw type (when a right-sided tube is used) because the right upper lobe ventilation slot is longer than it is in the polyvinylchloride tube13 and may better locate opposite/ventilate the right upper lobe (see Margin of Safety in Positioning Double Lumen Tubes). The remaining DLTs are largely of historical interest and are only briefly described here. The left-sided Carlens tube (Fig. 9-4) was the first double-lumen endotracheal tube utilized for one-lung ventilation.14 The tube has a carinal hook to aid in its proper placement and to minimize tube movement after placement. Potential problems with carinal hooks include increased difficulty (more rotations) and laryngeal trauma during intubation, amputation of the hook during passage, malpositioning of the tube due to the hook (see Margin of Safety in Positioning Double-Lumen Tubes), and physical interference when performing a pneumonectomy.15 Therefore, some anesthesiologists prefer to use the tube with the hook cut off. The tube is available in four sizes: 41, 39, 37, and 35 French (which correspond to an internal diameter for each lumen of approximately 7.0, 6.5, 6.0, and 5.5 mm, respectively). The cross-sectional shape of each lumen is oval, and this accounts for the occasional difficulty in passing a suction catheter down the lumen (see Table 9-2). The White tube was essentially a modified rightsided Carlens tube and was used for right main-
stem bronchus intubation.16 The right main-stem bronchial cuff is slotted to provide for ventilation of the right upper lobe. As with the Carlens tube. suctioning may occasionally prove to be difficult. and the carinal hook can cause a variety of physical problems. The Bryce-Smith tube (Fig. 9-5) represents another modification of the Carlens tube and was
Placement at the Carina
A. Carlens Tube Figure 9-4 A, Sketch of the red rubber Carlens doublelumen tube. B. Close-up of the placement of the red rubber Carlens double-lumen tube at the carina. Note that the left endobronchial lumen and carinal hook straddle the carina.
336
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Table 9-2 PHYSICAL CHARACTERISTICS OF VARIOUS DOUBLE-LUMEN TUBES*
*Based in part on Brodsky.'°t The 37, 39, and 41 F correspond to small, medium, and large, respectively, in the red rubber tubes. tAverage, each lumen.
intended to reduce trauma to the larynx and tracheobronchial tree.17 This tube was originally designed for placement in the left main-stem bronchus, although a right-sided tube was soon developed as well.18 The cuff for the right mainstem bronchus is slotted to allow for ventilation of the right upper lobe. These tubes do not have a carinal hook, and both lumens (arranged anteriorly and posteriorly) are round, allowing for greater ease in passing a suction catheter. Bryce-Smith double-lumen endotracheal tubes are available in
B. Placement at the Carina
A. Left Bryce-Smith Tube
three sizes according to the internal diameter of the lumen: 7, 6.5, and 6 mm. A new coaxial double-tube system for lung separation has been described.19 Separate ventilation is achieved by means of an appropriate cuffed bronchial tube whose tip is positioned in a main bronchus by passing it through a standard orotracheal tube (Fig. 9-6). The tubes assembled in this way make up two independent channels, each in communication with one of the main bronchi. Ventilation is then carried out separately through
D. Placement at the Carina
C. Right Bryce-Smith Tube
Figure 9-5 A, Sketch of the left Bryce-Smith double-lumen tube. B, Close-up of the placement of the left Bryce-Smith doublelumen tube at the carina. C, Sketch of the right Bryce-Smith tube. D, Close-up of the placement of the right Bryce-Smith doublelumen tube at the carina.
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
To the respirotor
Figure 9-6 Schematic representation of the coaxial double tube with the bronchial tube positioned within the left mainstem bronchus. (From Nazeri S, Trazzi R, Moncalvo F, et al: Selective bronchial intubation for one lung anesthesia in thoracic surgery. Anaesthesia 41:519-526, 1986. Used with permission.)
the bronchial tube on one side and the residual tracheal tube lumen on the other side. The bronchial tube can move freely in the main lumen, and the seal between the two channels is airtight. The bronchial tubes for the right and left sides are curved differently, and the left bronchial tube is fitted with a tracheal carinal hook, whereas the right bronchial tube has a hook for the first mainstem carina (separating the right upper lobe from the right middle and right lower lobes). The sizes of the tubes are such that they supply two channels with similar airflow resistance: This allows the
Figure 9-7 Modified left-sided polyvinyl chloride (PVC) tracheostomy double-lumen endobronchial tube (DLT) next to a standard leftsided PVC DLT (Broncho-Trach, Sheridan, Argyle, NY). (From Brodsky JB, Tobler HG, Mark JBD: A double-lumen endobronchial tube for tracheostomies. Anesthesiology 74: 387-388, 1991. Used with permission.)
337
anesthestic gases to be delivered through a Yshaped connector to both channels (see Fig. 9-6), supplying the same flow to the two lungs when they are both being ventilated. Because this new coaxial double-tube system is not in widespread use (since its introduction in 1986) and it is not apparent at present that it will be in widespread use in the foreseeable future, the reader is referred to the original articles for the details of insertion/operation. Finally, passing a conventional DLT through a tracheostomy presents some special problems that can be solved by using a modified DLT. Conventional DLTs made of polyvinylchloride can be used with tracheostomies, but, because of the shortened length of the upper airway, special attention is required to prevent kinking and subsequent luminal obstruction. Fixation of the tube to prevent movement and displacement may also be a problem with the standard DLT. These problems are greatly magnified if the tube is required for extended periods, as may occur when selective splitlung ventilation in the ICU is indicated. A DLT has been designed (left-sided 41 French Broncho-Trach DLT [Sheridan, Argyle, NY]) that has been used successfully in patients with tracheostomies (Fig. 9-7).20 The distance between the distal tip of the endobronchial lumen and the bifurcation of the tracheal and endobronchial lumens is shortened to 18.5 cm from 32.0 cm to reflect the markedly reduced length of the upper airway. The proximal length of both lumens after they bifurcate is also shortened to 3.0 cm from 7.5 cm to reduce the chance of kinking. A 90-degree bend is placed approximately 2.5 cm proximal to the proximal edge of the tracheal cuff to allow the tube to exit from the neck at a less awkward angle. The tubings to the pilot balloons are shortened to 11.0 cm from 23.0 cm for convenience. In all other aspects.
338
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
the tube is identical to a standard 41 French Sheridan polyvinylchloride DLT. If necessary, other (37F and 39F) polyvinylchloride DLTs may be similarly modified for smaller patients.
1. Robertshaw Double-Lumen Endotracheal Tube a. ORIGINAL RED RUBBER ROBERTSHAW DOUBLE-LUMEN ENDOTRACHEAL TUBE
The original Robertshaw DLT, introduced in 1962, was made as a reusable red rubber tube (Fig. 9-8).21 This tube was designed to provide the largest possible lumen in order to decrease airway resistance and to facilitate removal of secretions. The lumens are D-shaped and lie side by side, like those of the Carlens tube, but are larger in size. As with the other double-lumen endotracheal tubes, it has two curves (in planes 90 degrees apart) that facilitate intubation and proper endobronchial placement. Both right- and left-sided tubes are available, and the absence of a carinal hook allows for easier tracheal intubation and perhaps correct positioning. Because of these features, the original Robertshaw DLT rapidly gained popularity.22 The right-sided tube has a relatively long, slotted endobronchial cuff (22 mm vs. 11 mm for the polyvinylchloride tube) to permit ventilation of the
B. Placement at the Carina
A. Left Robertshaw Tube
right upper lobe. The long right upper lobe ventilation slot results in a much lower incidence of right upper lobe obstruction after "blind" insertion compared with the polyvinylchloride tube (10 per cent vs. 89 per cent).13 The endobronchial cuff has an additional area of inflation on the nonslotted side above the slot to effect a more reliable seal (in contrast to the endobronchial cuffs of the other right-sided tubes that do not have this inflation area). On the slotted side, inflation of the endobronchial cuff is restricted. However, the right endobronchial cuff design forces the right upper lobe slot to lie flat against the right upper lobe orifice, and if the right upper lobe slot is not perfectly aligned with the right upper lobe orifice, the right upper lobe ventilation slot will be perfectly aligned with the right upper lobe orifice, and the right upper lobe ventilation slot will be blocked (obstructed) by the right main-stem bronchial wall (and vice versa). Unfortunately, the bronchial cuff on the red rubber cuff is a low-volume-high-pressure cuff; the mean just-seal volume and intracuff pressure (± standard deviation) are 2.2 ± 1.1 ml and 130 ± 41 mm Hg, respectively;23,24 although only 10 to 20 per cent of intracuff pressure is transmitted/applied to the bronchial mucosal wall,25 mucosal perfusion is probably impaired to some degree by the sealing of the bronchial cuff of the red rubber tube. In addition, high-pressure
D. Placement at the Carina
C. Right Robertshaw Tube
Figure 9-8 A, Sketch of the left-sided red rubber Robertshaw double-lumen tube. B, Close-up of the placement of the left-sided red rubber Robertshaw double-lumen tube at the carina. C, Sketch of the right-sided red rubber Robertshaw double-lumen tube. D, Close-up of the placement of the right-sided red rubber Robertshaw double-lumen tube at the carina.
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
cuffs such as the bronchial cuff of the red rubber tube have a tendency to dilate asymmetrically, so greater volumes may be needed to effect a seal. b. MODERN PLASTIC DISPOSABLE ROBERTSHAW DOUBLE-LUMEN ENDOTRACHEAL TUBE
The Robertshaw-type tube is now made of a clear, nontoxic, tissue-implantable plastic (denoted by the marking Z-79) and is disposable (Fig. 9-9). The tubes are made in sizes 41, 39, 37, 35 and 28 French (internal diameter of each lumen is approx imately 7.4, 7.0, 6.5, 6.0, and 4.5 mm, respec tively) (see Table 9-2). These tubes are relatively easy to insert and have appropriate end-of-lumen and cuff arrangements that minimize lobar ob struction. Each lumen is color coded (bronchial is blue and tracheal is clear colorless) and labeled. The endobronchial cuff is colored brilliant blue, which is a very important feature for recognition when using a fiberoptic bronchoscope. The ends of both lumens have a black radiopaque line, which is an essential recognition marker when viewing a chest X-ray. The tubes have high-vol ume-low-pressure tracheal and endobronchial cuffs; the mean just-seal volume and intracuff pressure (± standard deviation) are 3.0 ± 2 ml and 56 ± 21 mm Hg, respectively. Because the actual pressure transmitted to the bronchial wall is only approximately 10 to 20 per cent of the meas ured intracuff pressure (the difference between measured intracuff pressure and the pressure ap plied to the bronchial wall is the pressure required simply to distend and maintain the inflation of the cuff itself), mucosal perfusion is probably main tained. The slanted doughnut-shaped endobron chial cuff on the Mallinkrodt right-sided DLT al lows the right upper lobe ventilation slot to ride off of (away from) the right upper lobe orifice, which minimizes the chance of right upper lobe obstruction by the tube. The clear color-coded tubing is helpful because it permits continuous observation of the tidal movement of respiratory moisture as well as ob servation of secretions from each lung. When the polyvinylchloride material heats to body tempera ture, the tube shape changes to approximate the shape of the airway more closely. The tubes are packaged with malleable stylets and are relatively easy to insert and position. These tubes have large internal to external diameter ratios and gentle cur vatures, thereby allowing relatively easy section ing, and have a low resistance to airflow. They are packaged with their own nonadhering suction catheters. For these reasons, the Robertshaw-type tubes are now considered by far the double-lumen endotracheal tube of choice by most anesthesiolo gists. As expected, several manufacturers make
339
these DLTs (Mallinkrodt, Sheridan, Rusch, Portex).
Β. Conventional (Nonfiberoptic) Double-Lumen Tube Intubation Procedure
1. Choice of Left- Versus Right-Sided Double-Lumen Tube It has been recommended that the nonoperated main-stem bronchus be routinely intubated be cause intubation of the operative main-stem bron chus may interfere with the performance of sur gery. Consequently, there is no controversy regarding using a left-sided double-lumen endotra cheal tube for right thoracotomies requiring col lapse of the right lung and ventilation of the left lung (Fig. 9-10). However, there is controversy regarding using a right-sided DLT for left lung surgery (see later discussion), and for this reason, either a left- or right-sided tube may be used for left thoracotomies requiring collapse of the left lung and ventilation of the right lung (see Fig. 910). However, because the right upper lobe venti lation slot of a right-sided tube has to be closely apposed to the right upper lobe orifice to ensure unobstructed right upper lobe ventilation, and be cause there is considerable anatomic variation in the exact position of the right upper lobe orifice and, therefore, length of the right main-stem bron chus (in fact, it is well known that an anomalous right upper lobe can take off from the trachea; the incidence is 1:250 in normal patients and 1:50 in patients who have some other congenital defect [see later discussion]), use of a right-sided tube for left lung collapse introduces the risk of inadequate right upper lobe ventilation. For this reason, a leftsided tube is preferable for most cases requiring one-lung ventilation. If clamping of the left mainstem bronchus is necessary, the tube can be with drawn at that time into the trachea and then used in the same manner as a single-lumen endotracheal tube (ventilate the right lung with both lumens with the endobronchial cuff deflated) (see Fig. 910). If after withdrawal the tracheal cuff is supraglottic, then the bronchial cuff should be inflated and the patient ventilated through the bronchial lumen. Contraindications to the use of a left-sided DLT are proximal left main-stem bronchial lesions that could be traumatized by the passage of a left-sided tube. These lesions include strictures, endoluminal tumors, tracheobronchial disruptions, and com pression of the airway by an external mass (includ ing a thoracic aortic aneurysm).26 In addition, left lower lobe and left upper lobe tumors may push
The Advantages of the Modern Plastic Disposable Robertshaw Double-Lumen Endobronchial Tubes
Β Figure 9-9 A, Left- and right-sided disposable Robertshaw double-lumen tubes. B, Schematic diagram depicts the advantages of left-sided and right-sided modern plastic disposable Robertshaw double-lumen endobronchial tubes. Both lumens of the left-sided double-lumen tube are shown, whereas only the distal endobronchial lumen of the right-sided double-lumen tube is shown.
340
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker intubation)
Left Lung Surgery and Right-Sided Double-Lumen Tube
Right Lung Surgery and Left-Sided Double Lumen Tube
341
Left Lung Surgery and Left-Sided Double-Lumen Tube Pulled Back
Β Figure 9-10 Use of left-sided and right-sided double-lumen tubes for left- and right-lung surgery (as indicated by the clamp). When surgery is going to be performed on the right lung, a left-sided double-lumen tube should be used (A). When surgery is going to be performed on the left lung, a right-sided double-lumen tube can be used (B). However, because of uncertainty as to the alignment of the right upper lobe ventilation slot to the right upper lobe orifice, a left-sided double-lumen tube can also be used for left lung surgery (C). If the left-lung surgery requires a clamp to be placed high on the left main-stem bronchus, the left endobronchial cuff should be deflated, the left-sided double-lumen tube pulled back into the trachea, and the right lung ventilated through both of the lumens (use the double-lumen tube as a single-lumen tube).
and pull, respectively, the left main-stem bronchus off the trachea at a very acute angle. This distor tion of the tracheobronchial tree may make it im possible to cannulate the left main-stem bronchus, even over a fiberoptic bronchoscope. Finally, sleeve resection of the proximal left main-stem bronchus contraindicates the presence of an endo bronchial catheter in the surgical field. The largest size tube that can comfortably pass the glottis should be used because a relatively small DLT may require excessive cuff volume for an endobronchial cuff seal to be obtained (see the discussion of endobronchial cuff problems in sec tion III. D. and Fig. 9-20), may be inadvertently inserted too deeply, may not protect against transbronchial spread of pus, blood, or necrotic tumor if the endobronchial cuff is deflated during twolung ventilation (larger space around small lu men),27 and may cause difficulty with suctioning secretions. In summary, the new plastic disposable Robertshaw-type DLTs are by far the most commonly used DLTs. Because a right-sided tube incurs the risk of inadequate right upper lobe ventilation, leftsided tubes are used more commonly than rightsided tubes. Consequently, the rest of this chapter emphasizes the insertion and location of the leftsided Robertshaw-type DLT.
2. Choice of Double-Lumen Tube Size The appropriate size of DLT is one that results in a just-seal volume for the endobronchial cuff that is greater than l ml but less than 3 ml. If the endobronchial just-seal cuff volume is less than l ml, then the outside diameter of the bronchial lu men is very near the internal diameter of the bron chial lumen, and the risk of bronchial wall damage is increased, especially if the DLT is inserted too deeply. If the endobronchial just-seal cuff volume is greater than 3 ml, then the outside diameter of the endobronchial tube is too small in relation to the bronchial lumen, the intracuff pressures may be excessive (the bronchial cuff will behave as a high-pressure cuff),28 and the risk of bronchial wall damage will be increased, especially if nitrous ox ide is used. In general, as height and weight increase, the appropriate size of the DLT (as defined previ ously) increases, although height is much more important than weight.29 Because women are usu ally shorter than men, sizes 35 French and 37 French are ordinarily appropriate for women, whereas sizes 39 French and 41 French are appro priate for men. However, choice of DLT by height, irrespective of gender, makes more sense; Figure 9-11 shows the frequency of choice of DLT size
342
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
3. Double-Lumen Tube Intubation Procedure a. PREINTUBATION PROCEDURES
Figure 9—11 Larger double-lumen tubes were chosen as patient height increased. (From Brodsky JB, Benumof JL, Ehrenwerth J, Ozaki GT: Depth of placement of left double-lumen endobronchial tubes. Anesth Analg 73:570-572, 1991. Used with permission.)
by grouped interval of height (panels A, B, and C) in a large series of patients from three institutions when it was known that the DLT was in proper position and cuff volume was appropriate.30 Although there are strong tendencies for short and tall individuals to receive small and large DLTs, respectively, the choice of size was still moderately variable.
Before intubation with a double-lumen endotracheal tube, both cuffs and lumen connections are checked. A 3-ml syringe should be placed on the end of the bronchial cuff pilot tube because proper bronchial cuff inflation rarely requires more than 1 to 2 ml of air; a 5- or 10-ml syringe should be placed on the tracheal cuff pilot tube. Both cuffs should be tested for leaks. Because the highvolume-low-pressure cuffs can be easily torn by teeth, the distal tube is coated with a lubricating ointment (preferably containing a local anesthetic) to minimize this possibility. If a less than optimal view of the larynx is anticipated, the stylet that comes packaged with the tube is lubricated, inserted into the endobronchial (in this case, the left) lumen, and appropriately curved. The patient is then anesthetized and paralyzed as described in chapter 8. With the induction of anesthesia, patients who are apneic for more than the normal amount of time are at risk for hypoxia before intubation and/or ventilation is achieved; this may occur despite "preoxygenation." During this apneic period, oxygen can be drawn from the pharynx into the trachea and lungs, provided that the airway is patent; the transfer of oxygen in this manner should delay the onset of hypoxia. One randomized, double-blind study proved that insufflation of oxygen into the nasopharynx of denitrogenated patients could effectively prolong the safe apneic period before intubation.31 Patients were between 35 and 65 years old (52 ± 4 years, mean ± standard error) and had significant histories of tobacco use (20 ± 6 pack-years). During pharyngeal oxygen insufflation, Sa02 never fell below 97 per cent during the entire 10 min of apnea in any subject; the mean minimum saturation achieved ( ± standard deviation) was 98 ± 1 per cent. Conversely, in the absence of oxygen insufflation, the duration of apnea was 6.8 ± 0.6 min (p < .001), and the mean minimum saturation observed ( ± standard deviation) was 91 ± 1 per cent (p < .0001). These results were independent of whether preoxygenation was by spontaneous or controlled ventilation. Thus, insufflation of oxygen via nasopharyngeal cannula provides at least 10 min of adequate oxygenation in unintubated, denitrogenated, apneic patients whose airways are unobstructed. By significantly delaying the onset of hypoxia, this technique may be life-saving: Extra time is allowed to obtain control of the airway in critical situations. On rare occasions in which tracheal intubation is known to be very difficult or respiration is already compromised, it may be ap-
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
propriate to institute prophylactic transtracheal jet ventilation before inducing general anesthesia or attempting conventional or fiberoptic tracheal intubation with the patient either awake or anesthetized.8· 32 b. INTUBATION
A curved open-phalange blade (Macintosh) is usually preferred for laryngoscopy because it approximates the curvature of the tube and therefore provides the largest possible area through which to pass the tube. However, a straight (Miller) blade may be a better choice in patients with overriding teeth or an excessively anterior larynx. Double-lumen endotracheal tubes with carinal hooks are first inserted through the vocal cords with the distal curve concave anteriorly (just like a single-lumen tube) and the hook facing posteriorly. When the tip of the tube has passed the
343
vocal cords, the tube is rotated 180 degrees, so that the hook passes anteriorly through the glottis. After the tube tip and the hook pass the larynx, the tube is rotated 90 degrees so that the tube tip is curved toward and enters the appropriate bronchus and the hook engages the carina. The Robertshaw DLT is passed with the distal curvature initially concave anteriorly (Fig. 9-12A) (just like a single-lumen tube). After the tube tip passes the larynx, and while anterior force on the laryngoscope is continued, the stylette (if used) is removed and the tube is carefully rotated 90 degrees (so that the distal curve is now concave toward the appropriate side and the proximal curve is concave anteriorly) to allow endobronchial intubation on the appropriate side (Fig. 9-125). Continued anterior force by the laryngoscope during tube rotation prevents hypopharyngeal structures from falling in around the tube and interfer-
Passage of Left-Sided Double-Lumen Tube
Figure 9-12 This schematic diagram depicts the passage of the left-sided double-lumen tube in a supine patient. A, The tube is held with the distal curvature concave anteriorly and the proximal curve concave to the right and in a plane parallel to the floor. The tube is then inserted through the vocal cords until the left cuff passes the vocal cords. The stylet is then removed. B, The tube is rotated 90° counterclockwise so that the distal curvature is concave anteriorly and the proximal curvature is concave to the left and in a plane parallel to the floor. C. The tube is inserted until either a moderate resistance to further passage is encountered or the end of the common molding of the two lumens is at the teeth. Both cuffs are then inflated, and both lungs are ventilated. Finally. one side is clamped while the other side is ventilated and vice versa (see text for further explanation).
344
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
ing with a free 90-degree distal tube tip rotation. Failure to obtain close to a 90-degree rotation of the distal tube tip, while the proximal end does rotate 90 degrees, will cause either a kink or twist in the shaft of the tube and/or prevent the distal end of the lumen from lying free in the main-stem bronchus (i.e., not up against the bronchial wall). After rotation, the tube is advanced until most of it is inserted.30 When the proper depth of insertion is defined as the cephalad surface of the bronchial cuff being immediately below the carinal bifurcation, the average depth of insertion for both male and female patients 170 cm tall is 29 cm, and for each 10-cm increase or decrease in height, average placement depth is increased or decreased 1 cm (Fig. 9-13).30 The correlation between depth of insertion and height is highly significant (p < .0001) for both male and female patients. Nevertheless, it should be understood that the depth of DLT insertion at any given height is still normally distributed (see Fig. 9-13), and correct DLT position should always be confirmed fiberoptically after initial placement.
If the proximal end of the common or twolumen binding mold is near or at the level of the teeth (i.e., at a depth of 32-33 cm) in a normalsize person and/or moderate resistance to further passage is encountered, it usually means that the tube has been pushed in too far and the tube has been firmly seated in a distal bronchus (Fig. 912C). Double-lumen endotracheal tubes may also be passed successfully via tracheostomy, although it should be remembered that the tracheal cuff may be at the tracheal stoma or lie partly outside the trachea in this situation.33· 34 A special tracheostomy DLT has been designed for use with tracheostomies (see Figure 9-7).20 C. ROUTINE CHECKING OF DOUBLE-LUMEN T U B E POSITION
Once the tube tip is thought to be in an endobronchial position, the following steps are carried out to ensure proper functioning of the tube. First, inflate the tracheal and endobronchial cuffs until moderate tension is palpated in the external pilot balloons. The endobronchial cuff should not require more than 2 to 3 ml of air; if it does, the cuff
Figure 9-13 The depth of insertion for left double-lumen tubes for all patients and for three grouped intervals based on patient height. At each grouped interval, the depth of insertion was normally distributed. (From Brodsky JB, Benumof JL, Ehrenwerth J, Ozaki GT: Depth of placement of left double-lumen endobronchial tubes. Anesth Analg 73:570-572, 1991. Used with permission.)
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
may be herniating out of the main-stem bronchus (which is, by far, the most likely possibility) or the main-stem bronchus is malotic, or the DLT size is inappropriately small. Later, after one has completed fiberoptic confirmation of proper DLT position (see later discussion), the tracheal lumen can be clamped and the endobronchial cuff deflated. Then, while ventilating through only the endobronchial lumen, its cuff can be slowly inflated until the minimum amount of air needed to prevent an air leak is determined (which is the just-seal volume of air). Other more quantitative methods of determining the just-seal volume of cuff air are the positive-pressure ventilation/air bubble underwater technique (see section III.H.), negative circle system pressure/collapse of reservoir bag, and use of capnography (see Figs. 9-39 to 9-41 and Hannallah,35 Hannallah and Benumof,36 and Essig and Freeman37). Determination of an air leak when the cuff is deflated is important because it rules out the possibility that the DLT is too tightly impacted in the bronchus. Several positive-pressure ventilations should be delivered and the chest auscultated and observed bilaterally to determine that the trachea, rather than the esophagus, has been intubated and that both lungs are being ventilated (Fig. 9-12Q. In addition to seeing the tube go through the vocal cords, correct intubation position is checked by feeling and observing the anesthesia reservoir bag to make sure it has the appropriate compliance and movement, maintaining normal pulse oximetry and end-tidal C0 2 values, and perhaps palpating the tracheal cuff in the neck. If only unilateral breath sounds or chest movement are present, it is likely that both of the lumens of the tube have entered a main-stem bronchus (if both of the lumens enter the left main-stem bronchus, the findings may mimic an esophageal intubation and vice versa). In this situation, quickly deflate the cuffs, withdraw the tube 1 to 2 cm at a time, inflate the cuffs, and reassess ventilation until bilateral breath sounds are heard. If bilateral breath sounds are not heard, and the tube has been withdrawn a significant amount, the entire procedure must be repeated, beginning with establishing the airway and oxygen ventilation via mask, laryngoscopy, and reinsertion of the DLT through the vocal cords. If bilateral breath sounds are present, then one side is clamped, and breath sounds and chest movement should disappear on the ipsilateral side and remain on the contralateral side. Next, the clamped side should be undamped, and the breath sounds and chest movement should reappear on that side. During unilateral clamping, the breath sounds on the ventilated side should be compared with and calibrated against unilateral chest wall movements and the inspiratory disappearance and expiratory
345
appearance of respiratory gas moisture in the clear tubing of the ventilated side (Fig. 9-14). In addition, the compliance of the lung should be gauged using hand ventilation. The unilateral clamping and unclamping should then be repeated on the opposite side to ensure adequate lung separation and cuff seal. DLT adaptors have been described that permit each lumen independently to be either open to the mechanical ventilator or the atmosphere or completely blocked (as though it was clamped) by simply turning a dial/stopcock to the desired setting without the need for airway disconnection and/or external clamping maneuvers.38·39 With one adaptor, the two lungs may receive any combination of PEEP, CPAP, FOB, and/or suctioning.17 These features greatly facilitate the testing for lung separation as well as use of all of the other onelung ventilation/anesthesia maneuvers that are demanded by modern anesthetic practice (see chapter 11). In summary, when double-lumen endotracheal tube position is correct, the breath sounds are normal and follow the expected unilateral pattern with unilateral clamping, the chest rises and falls in accordance with the breath sounds, the ventilated lung feels reasonably compliant, no leaks are present, and respiratory gas moisture appears and disappears with each tidal ventilation (see Fig. 9-14). When comparing both lumens during two-lung ventilation and when changing from two-lung ventilation to one-lung ventilation with a known constant tidal volume, the exhaled tidal volume should not decrease by more than 15 per cent, expiratory flow rate from either lung should not slow markedly, peak airway pressure should not increase by more than 60 per cent, and there should not be an obvious difference in the rate of appearance of the fog in the two lumens during exhalation. However, it should be realized that, in the presence of advanced lung disease, loss of lung tissue, or atelectasis, more exaggerated changes in the just-mentioned variables are expected when switching from two-lung ventilation to diseased-lung ventilation. The auscultation findings with a DLT, whose endobronchial cuff has gone past an upper lobe orifice while still allowing ventilation of the upper lobe by the tracheal lumen (see Fig. 9-27), may closely mimic the findings expected for a properly positioned DLT during two-lung ventilation.40 4I Conversely, when the double-lumen endotracheal tube is malpositioned, any or all of the following may occur: The breath sounds are poor and correlate poorly with unilateral clamping, the chest movements do not follow the expected pattern, the ventilated lung feels noncompliant, leaks are present, or the respiratory gas moisture in the clear tubing is relatively stationary. It is very important
346
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Correct Position of Double-Lumen Tube and Unilateral Clamping
Ipsilateral Breath Sounds Disappear
Contralateral Breath Sounds Remain and Have the Expected Quality
Contralateral Hemithorax Rises and Falls
Ipsilateral Hemithorax Does Not Move
Ipsilateral Respiratory Gas Moisture is Stationary
Contralateral Respiratory Gas Moisture Disappears on Inhalation and Reappears on Exhalation Breathing Bag Has the Expected Compliance for One-Lung Ventilation
to realize, however, that even if the DLT is thought to be properly positioned based on clinical signs, subsequent FOB will reveal a 40 to 48 per cent incidence of malpositioning.42,43 (see section III.D.). Obviously, the auscultation and fiberoptic (see later discussion) findings can always be supplemented by direct observation of the operative lung and mediastinum when the chest is open. It is controversial as to whether the endobronchial cuff should be left inflated or deflated after confirmation of proper position of the DLT (see section III.D.)· Deflation of the cuff during twolung ventilation in the initial stages of surgery (preparing the chest, draping, opening the chest wall, etc.) minimizes the chance of pressure damage to the bronchial mucosa. Alternatively, leaving the cuff inflated (after readjustment to a justseal volume) prevents trans-main-stem bronchial spread of secretions (pus, blood) and necrotic tumor.22 In cases in which migration of material from the nondependent lung to the dependent lung is a possibility, adjustment of the cuff to a just-
Figure 9-14 This schematic diagram shows the results of unilateral clamping when the double-lumen tube is in the correct position.
seal volume seems to minimize risk and maximize benefit. d. AUSCULTATION AND UNILATERAL CLAMPING MANEUVERS TO DIAGNOSE DOUBLE-LUMEN TUBE MALPOSITION
When it is felt that the double-lumen endotracheal tube is malpositioned based on clinical signs, it is theoretically possible to diagnose the malposition of the tube more precisely by a combination of several unilateral clampings, chest auscultation, and left endobronchial cuff inflation/deflation maneuvers (Fig. 9-15). With reference to a left-sided double-lumen endotracheal tube, there are three possible gross malpositions: in too far on the left (both lumens in the left main-stem bronchus), out too far (both lumens in the trachea), and in or down the right main-stem bronchus (at least the left lumen is in the right main-stem bronchus). When the right (tracheal) side is clamped (breath sounds should be heard only over the left lung) and the tube is in too far on the left side, breath
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
347
Double-Lumen Tube Malpositions
Procedure Clamp Right Lumen Both Cuffs Inflated Clamp Left Lumen Both Cuffs Inflated Clamp Left Lumen Deflate Left Cuff
Breath Sounds Heard Left
Left and Right
Right
None or Very 11 None or Very 11 None or Very 11 Left
Left and Right
Right
Figure 9-15 There are three major (involving a whole lung) malpositions of a left-sided double-lumen endotracheal tube. The tube can be in too far on the left (both lumens are in the left main-stem bronchus), out too far (both lumens are in the trachea), or down the right main-stem bronchus (at least the left lumen is in the right main-stem bronchus). In each of these three malpositions, the left cuff, when fully inflated, can completely block the right lumen. Inflation and deflation of the left cuff while the left lumen is clamped creates a breath sound differential diagnosis of tube malposition. (See text for full explanation.) (L = left; R = right: 1 = decreased.)
sounds will be heard only on the left side. When the tube is out too far and the right side is clamped, breath sounds will be heard bilaterally (the tube needs to be advanced further). When the tube is in or down the right side and the right lumen is clamped, breath sounds will be heard only on the right side (the tube needs to be pulled back, rerotated, readvanced). When the left side is clamped and the left endobronchial cuff is inflated, the right lumen is blocked by the left cuff in all three malpositions. Consequently, with the left side clamped and the left cuff inflated, no, or very diminished, breath sounds will be heard bilaterally in all three malpositions, and there will be marked resistance to airflow (the right lumen opens be tween two inflated cuffs). When the left side is clamped and the left cuff is deflated, so that the right lumen is no longer blocked by the left cuff, breath sounds will be heard only on the left side when the tube is in too far on the left (the tube needs to be pulled back), breath sounds will be heard bilaterally if the tube is out too far (tube
needs to be advanced), and breath sounds will be heard only on the right side when the tube is in the right side (tube needs to be pulled back, rero tated, and readvanced). The left-cuff infla tion/deflation findings provide the key diagnostic data because they essentially define the position of the right tracheal lumen by blocking and unblock ing it with the left cuff. Another possible diagnos tic sequence is presented in Figure 9-1 ό. 44 There are, however, several situations in which unilateral clamping, auscultation, and cuff inflation and deflation maneuvers for determining the integ rity of lung separation are either unreliable or im possible. First, and most importantly, when the patient is in the lateral decubitus position, has had a skin preparation, and is draped, access to the chest wall is impossible, and the anesthesiologist cannot listen to the chest. Second, the presence of unilateral or bilateral lung disease, which either existed before anesthesia and surgery or was in duced by anesthesia, may markedly obscure the crispness of the chest auscultation end points.
348
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Pulmonary Surgery
Figure 9-16 Decision tree outlining a sequence of cuff inflation and positive-pressure ventilation through a left-sided double-lumen tube that will ensure proper functioning of the tube. (From Katz J A, Fairley HB: Pulmonary Surgery. In Marshall BE, Longnecker DF, Fairley HB (eds): Anesthesia for Thoracic Procedures. Boston, Blackwell Scientific, 1988, pp 363^413. Used with permission.)
Third, the diagnosis of exactly where the doublelumen endotracheal tube is located may be confused when the tube is just slightly malpositioned. Fourth, the tube may have moved because of some event such as coughing, turning into the lateral decubitus position, and tracheal manipulation and hilar retraction by the surgeon. Finally, some combination of these circumstances may culminate in uncertainty as to where the DLT is located. The solution to any uncertainty as to the exact position of the DLT is to determine the position by use of FOB (see Use of Fiberoptic Bronchoscope to Determine Precise Double-Lumen Tube Position). e. NONFIBEROPTIC METHOD FOR INTUBATION OF LEFT MAIN-STEM BRONCHUS W H E N A LEFTSIDED TUBE WILL NOT PASS ROUTINELY
If a left-sided tube locates in the right mainstem bronchus and repeated attempts at correct
placement are unsuccessful, rotating the patient's head and neck to the right before rotating and advancing the tube may result in proper lateralization of the left-sided double-lumen endobronchial tube.45 Bronchoscopists have long recognized the increased difficulty of inserting a rigid bronchoscope into the left main-stem bronchus because of the angle it makes with the trachea and because, in 74 per cent of patients, its orifice is partly covered with the tracheal carina.46 The technique recommended for passage of a rigid bronchoscope into the left main-stem bronchus is that the patient's head, face, and neck are turned to the right. Kubota et al.47 reported their experience of selective blind left endobronchial intubations using a single-lumen endotracheal tube in 300 adults. The highest success rate (275/300, or 92 per cent) was achieved when the tube was rotated 180 degrees (so that the bevel faced toward
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
the right) and the patient's head was turned to the right. When the head was not rotated to the right, the success rate was only 61 per cent (182/300). The difference in success rates was statistically significant (p < .01). Turning the head and neck to the right improves the success of tube placement most likely because it shifts the larynx to the right in relation to the carina. This tends to bring the axis of the left main-stem bronchus more into line with that of the trachea (i.e., the bronchus would arise at a smaller angle), and the endobronchial tube would have a straighter line to the left mainstem bronchus.45 It is also possible that turning the head stretches the trachea and left main-stem bronchus, thereby altering the anatomy of the origin of the left main-stem bronchus to make it wider or less slit-like, either way rendering it more receptive to the passage of the bronchial catheter of a left-sided double-lumen endobronchial tube.
C. Use of Fiberoptic Bronchoscope to Insert the Bronchial Lumen of a DoubleLumen Tube Into a Main-Stem Bronchus The insertion of the bronchial lumen of a DLT into the appropriate main-stem bronchus may be aided by FOB (Fig. 9-17). The procedure is indicated whenever the endobronchial lumen will not locate in the appropriate main-stem bronchus by blind insertion (e.g., ventilation through only the endobronchial lumen results in breath sounds over the wrong lung field or side), when the endobronchial lumen needs to be maneuvered past some sort of disease (e.g., stenosis in the left main-stem bronchus caused by thoracic aortic aneurysm),26 or when intubation is known to be very difficult and fiberoptic DLT intubation while the patient is awake (or anesthetized) is indicated48· 49 (as it would be for a single-lumen tube; of course, the awake patient requires topical and nerve block anesthesia).48 To use a fiberoptic bronchoscope to insert and precisely position DLTs and bronchial blockers successfully, one must be familiar with the normal anatomy of the tracheobronchial tree.50 The trachea of an adult is 10 to 14 cm long and is composed of 18 to 24 incomplete, horseshoe-shaped cartilages that appear as ridges from the four o'clock position to the eight o'clock position anteriorly. These are connected to one another posteriorly by the trachealis muscle (from the four o'clock position to the eight o'clock position posteriorly). The right side is therefore at three o'clock and the left side is at nine o'clock. The distal end of the trachea is displaced to the right by the aortic arch. A prominent aortic arch and
349
enlarged paratracheal and bronchial lymph nodes can cause compression, narrowing, and displacement of the trachea and main-stem bronchi. When the patient is supine, the carina lies at the level of the fourth vertebra. The right main-stem bronchus has a very variable length with an average of 15 to 18 cm (female vs. male) before the right upper lobe bronchus branches off. The right upper lobe bronchus arises from the lateral aspect (three-o'clock position) of the main-stem bronchus and divides into three segmental bronchi: the apical, posterior, and anterior branches. The right middle-lobe bronchus arises from the anterior wall of the intermediate bronchus (12 o'clock position), about 2.4 cm beyond the orifice of the upper lobe bronchus. It divides into two segmental bronchi: the lateral and the medial branches. The right lower lobe bronchus is a continuation of the right main-stem bronchus and branches into five segmental bronchi. The left main-stem bronchus is an average of 44 to 49 mm long (female vs. male). It is slightly more narrow and has a more acute origin from the carina than does the right main-stem bronchus. The "secondary carina" lies at the division of the left upper and left lower lobe bronchi. Although it may occupy any position between the horizontal and vertical axes, it usually lies obliquely between the eight-o'clock and the two-o'clock positions. The left upper lobe bronchus divides into an upper division and the lingular bronchus. The upper division constitutes the upper lobe bronchus and divides into two branches: the apical posterior and anterior segments. The lingular bronchus divides into the superior and inferior lingular segments. The left lower lobe bronchus is a continuation of the left main-stem bronchus and branches into four segmental bronchi. The DLT is usually first placed in the trachea in a conventional manner (laryngoscopy, manual tube insertion) (unless the patient is awake and the fiberscope is used from the outset) until the tracheal cuff just passes the vocal cords. The tracheal cuff is then inflated, and both lungs are ventilated with both lumens (use the DLT as if it were a single-lumen tube). A pediatric fiberoptic bronchoscope can then be inserted into the bronchial lumen through a self-sealing diaphragm in the elbow connector to the bronchial lumen (which permits continued positive-pressure ventilation through that lumen around the fiberoptic bronchoscope) and passed into the appropriate main-stem bronchus.48· 50 The tracheal cuff is then deflated, and the bronchial lumen is passed over the fiberoptic bronchoscope stylet into the appropriate main-stem bronchus. With respect to placing a right-sided DLT correctly (i.e., getting the right upper lobe ventilation
350
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Use of Fiberoptic Bronchoscope to Insert Left-Sided Double-Lumen Tube
Insert Double-Lumen Tube Into Trachea in Conventional Manner and Ventilate Both Lungs
Pass Fiberoptic Bronchoscope Down Left Lumen Into Left Main Stem Bronchus
Push Double-Lumen Tube in Over Fiberoptic Bronchoscope Until Left Lumen is in Left Main Stem Bronchus
Figure 9-17 This schematic diagram portrays use of the fiberoptic bronchoscope to insert a left-sided double-lumen tube. The double-lumen tube can be put into the trachea in a conventional manner, and both lungs can be ventilated by both lumens (A). The fiberoptic bronchoscope may be inserted into the left lumen of the double-lumen tube through a self-sealing diaphragm in the elbow connector to the left lumen; this allows continued positive-pressure ventilation of both lungs through the right lumen without creating a leak. After the fiberoptic bronchoscope has been passed into the left main-stem bronchus (B), it is used as a stylet for the aftercoming left lumen (C). The fiberoptic bronchoscope is then withdrawn. Final precise positioning of the double-lumen tube is performed with the fiberoptic bronchoscope in the right lumen (see Fig. 9-18).
slot opposite the right upper lobe bronchial orifice), the right upper lobe bronchial orifice may first be identified fiberoptically and then, while holding the fiberoptic bronchoscope steady, the endobronchial lumen may be passed over the fiberoptic bronchoscope into the right main-stem bronchus (the tip of the fiberoptic bronchoscope must be returned to a neutral position during passage of the DLT over the fiberoptic bronchoscope) until the tip of the fiberoptic bronchoscope can lookthrough the right upper lobe ventilation slot into the right upper lobe bronchial orifice (Fig. 9-215). The fiberoptic bronchoscope is then withdrawn from the bronchial lumen to determine the precise DLT position (see the following section). Alternatively, once the DLT is in the trachea, the fiberoptic bronchoscope can be inserted into the tracheal lumen through a self-sealing diaphragm in the elbow connector to the tracheal lumen (which permits continued positive-pressure ventilation through that lumen around the fiberoptic bronchoscope) and passed just proximal to the tracheal carina.51 While the carina and the two main-stem bronchial orifices are in view, the DLT
can be advanced and the degree of lateral rotation adjusted so that the left lumen enters the left mainstem bronchus. Final precise positioning (see the following section) can be done with the fiberoptic bronchoscope remaining in the tracheal lumen. D. Use of Fiberoptic Bronchoscope to Determine Precise Double-Lumen Tube Position The deleterious consequences of a malpositioned DLT can be great, even life-threatening. With almost all DLT malpositions, gas exchange can be significantly/profoundly impaired, the nonoperative lung can be difficult/impossible to ventilate, and the operative lung may not collapse on the initiation of one-lung ventilation. In addition, failure to separate the lungs in some specific situations may result in additional catastrophes such as flooding both lungs with blood, pus, or lavage fluid and tension pneumothorax. There is a very high incidence of malpositioned DLTs, as determined by FOB, when the DLT is
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
inserted blindly (i.e., without a fiberoptic bronchoscope). In one study, 48 per cent of all DLTs were found to be malpositioned.52 In another study, 83 per cent of disposable right-sided DLTs were malpositioned.53 In a third study, FOB after auscultation (which led the physician to believe the DLT was in optimal position) resulted in repositioning of 78 per cent of left-sided DLTs and 83 per cent of right-sided DLTs.54 In a fourth study, 38 per cent of all DLTs were malpositioned.55 In a fifth study, 44 per cent of disposable tubes required readjustment using the fiberoptic bronchoscope during the initial intubation, and 30 per cent of disposable tubes required readjustment using the fiberoptic bronchoscope during the operation.56 The authors of the last three studies concluded that "auscultation is an unreliable method of confirming the position of DLTs and should be followed by fiberoptic bronchoscopy";54 "because auscultation for tube position is unreliable, bronchoscopic assessment of final position should be performed in every instance"; 55 and "in certain situations the fiberoptic bronchoscope may have been life-saving."56 Given the high incidence of malpositioned DLTs when DLT position is determined by only auscultation (i.e., "blindly") and the potentially serious consequences associated with a malpositioned DLT, it is only a matter of simple common sense to use an FOB routinely to determine easily, quickly, and precisely the position of the DLT. In the studies just quoted, it is open to conjecture and speculation as to what would have happened if the position of the DLT was not corrected with the aid of FOB. However, one study has shown that, when the position of the DLT is checked only by clinical signs, there will be intraoperative problems 25 per cent of the time with either deflating the nondependent lung, ventilating the dependent lung, or completely separating the two lungs.57 If one accepts the premise that the position of a DLT should be confirmed by FOB, then I also contend that it should be done first in the supine position, again after the patient has been turned into the lateral decubitus position, and again whenever there is a question about the DLT position (in conjunction with other maneuvers such as palpation of the lumen tips by the surgeon, observation of the operative lung, mediastinum, etc.). There are several reasons for repeating FOB. First, when the patient is supine, access to the patient's head and DLT is optimal, the orientation of the patient and FOB is most certain, and correlation of the auscultatory findings is easiest (by whatever unilateral clamping and auscultation technique one prefers)44·58-70 (Fig. 9-15) with the FOB findings. The main value of FOB at this time (supine position), given that all auscultation algorithms can
351
rule in or rule out cannulation of the wrong mainstem bronchus or failure to cannulate either mainstem bronchus, is to rule in or out upper lobe obstruction (by determining endobronchial cuff position or by visualizing the upper lobe bronchial orifice). I do not believe one can accurately diagnose upper lobe obstruction by auscultation alone because breath sounds are transmitted from the ipsilateral lower lobe and across the mediastinum from the contralateral lung. Second, if the DLT is properly positioned in the supine position with certainty (i.e., with FOB confirmation), then the second and more important confirmation of DLT position after the patient is in the lateral position is greatly facilitated by the first look in the supine position. In addition, events that take place between the supine and lateral decubitus position may affect the functioning of the DLT. For example, placement of the axillary roll has caused almost complete obstruction of the right main-stem bronchus in a 10-year-old child.71 Similarly, third and further intraoperative views are facilitated by the second view in the lateral position. This series/cascade of FOB views of DLT position interact/overlap to give the anesthesiologist much tighter control of vital respiratory functions throughout the anesthetic with very little risk. The exact position of a left-sided double-lumen endotracheal tube can be ascertained at any time, in less than a minute, by simply passing a pediatric fiberoptic bronchoscope through the tracheal lumen of the DLT. It is rarely necessary also to have to pass the fiberoptic bronchoscope down the left endobronchial lumen. With reference to a leftsided DLT, looking down the right tracheal lumen. the endoscopist should see a clear, straight-ahead view of the tracheal carina (the tracheal lumen should be approximately 1-2 cm above the carina), the left lumen going off to the left, and the upper surface of the left endobronchial balloon just below the tracheal carina (Figs. 9-18 and 9-19). (The importance of seeing the upper surface of the left endobronchial cuff below the tracheal carina is emphasized in the following section.) It is important that the volume of air used to fill the left endobronchial left cuff does not cause the endobronchial cuff to herniate over the tracheal carina or cause the tracheal carina to deviate to the right (Figs. 9-19 and 9-20); both cuff herniation and carinal deviation can be readily appreciated looking down the tracheal lumen. Looking down the left lumen (which is sometimes done when inserting a left-sided DLT with a fiberoptic bronchoscope [see section III.C] and in all cases of bronchopulmonary lavage in which perfect tube position and tight cuff seal are extremely critical), the endoscopist should see a very slight narrowing of the left lumen (due to endobronchial cuff près-
352
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Use of Fiberoptic Bronchoscope Down the Right Lumen to Determine Precise Left-Sided Double-Lumen Tube Position
Figure 9—18 This schematic diagram portrays use of the fiberoptic bronchoscope down the right lumen to determine precise left-sided double-lumen tube position. The endoscopist should see a clear straight-ahead view of the tracheal carina, the left lumen going off into the left main-stem bronchus, and, most importantly (in bold print), the upper surface of the blue left endobronchial cuff just below the tracheal carina.
sure) as well as the bronchial carina distal to the end of the tube (see Fig. 9-19). The endoscopist should not see excessive left luminal narrowing (due to excessive left cuff pres
sure) (see Fig. 9-20). Thus, aside from gross mal position, important undesirable findings on endos copy are related to excessive left cuff inflation and pressure and consist of cuff herniation over the tracheal carina, carinal deviation to the right (both of which may block the right main-stem bronchial orifice and impair right lung ventilation), and ex cessive left lumen constriction (invagination), which may impair left lung ventilation (Fig. 920).61 In addition, when an inappropriately under sized tube is used, the large endobronchial cuff volume required for endobronchial cuff seal tends to force the entire DLT cephalad, making a func tional bronchial seal more difficult.62 With reference to a right-sided DLT, looking down the left (tracheal) lumen the endoscopist should see a clear straight-ahead view of the tra cheal carina, with the right lumen going off to the right (Fig. 9-21Λ). The upper surface of the right endobronchial balloon may not be visible below the tracheal carina. Looking down the right lumen, the endoscopist should see the right middle-lower lobe bronchial carina distal to the end of the tube. Most importantly, the endoscopist should locate the right upper lobe ventilation slot and be able to look directly into the right upper lobe orifice through the right upper lobe ventilation slot by simply flexing the tip of the fiberoptic broncho scope laterally and superiorly (Fig. 9-21B). There
Use of Fiberoptic Bronchoscope to Determine Precise Left-Sided Double-Lumen Tube Position
View Down Left (Bronchial) Lumen
View Down Right (Tracheal) Lumen
Figure 9-19 This schematic diagram depicts the complete fiberoptic bronchoscopy picture of left-sided double-lumen tubes (both the desired view and the view to be avoided from both of the lumens). When the bronchoscope is passed down the right lumen of the left-sided tube, the endoscopist should see a clear straight-ahead view of the tracheal carina and the upper surface of the blue left endobronchial cuff just below the tracheal carina. Excessive pressure in the endobronchial cuff, as manifested by tracheal carinal deviation to the right and herniation of the endobronchial cuff over the carina, should be avoided. When the bronchoscope is passed down the left lumen of the left-sided tube, the endoscopist should see a very slight left luminal narrowing and a clear straight-ahead view of the bronchial carina off in the distance. Excessive left luminal narrowing should be avoided.
.
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
353
Excessive Left Cuff Inflation: Problems
Figure 9-20 Excessive inflation of the left cuff of a left-sided double-lumen tube can cause impaired ventilation of both the right and left lungs. Right-lung ventilation may be impaired by left-cuff herniation over the tracheal carina (as a result of excessive left-cuff volume) and by tracheal carinal deviation to the right (because of excessive left-cuff pressure). Left-lung ventilation may be impaired by invagination of the left lumen caused by excessive left-cuff pressure.
Use of Fiberoptic Bronchoscope to Determine Precise Right-Sided Double-Lumen Tube Position
Right Lumen Going off to the Right
View Down Left (Tracheal) Lumen
Right Upper Lobe Bronchial Orifice
View Down Right (Bronchial) Lumen
Figure 9-21 This schematic diagram portrays use of a fiberoptic bronchoscope to determine precise right-sided double-lumen tube position. A, When the fiberoptic bronchoscope is passed down the left (tracheal) lumen, the endoscopist should see a clear straight-ahead view of the tracheal carina and the right lumen going off into the right main-stem bronchus. B, When the fiberoptic bronchoscope is passed down the right (bronchial) lumen, the endoscopist should see the bronchial carina off in the distance; when the fiberoptic bronchoscope is flexed laterally and cephalad and passed through the right upper lobe ventilation slot, the right upper lobe bronchial orifice should be visualized.
354
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
should be no overriding of the right upper lobe ventilation slot on the bronchial mucosa, and the bronchial mucosa should not be covering any of the right upper lobe ventilation slot. The fact that there is little room for error in aligning the right upper lobe ventilation slot with the right upper lobe orifice is emphasized in the following section. The right upper lobe ventilation slot may be easily located by first finding the bronchial lumen radiopaque marker. The radiopaque marker appears as either a white or black line on the inside of the bronchial lumen, which ends at the proximal end of the right upper lobe ventilation slot. If one simply follows the radiopaque line, it will lead the endoscopist to the right upper lobe ventilation slot. In my experience, in eight of 10 cases, the clinical signs (breath sounds, chest movements, compliance of the lung[s], movement of respiratory gas moisture) indicate that the lungs are apparently clearly and without doubt completely separated when the DLT is first inserted with the patient in the supine position. However, in view of the finding by many authors52-56 that between 40 and 50 per cent of DLTs are malpositioned to some extent in the supine position, even though the clinical signs may indicate there is no problem, it is strongly advisable to check the position of the tube with a fiberoptic bronchoscope in the supine position (especially considering that the procedure takes less than a minute). Even if no problem is identified, the procedure still allows the endoscopist to become familiar with the patient's anatomy and facilitates the more important endoscopy performed after turning the patient into the lateral decubitus position. In approximately two of 10 cases, there is doubt about tube location in the supine position, and in these patients the fiberoptic bronchoscope is always used to correct the DLT malposition. The fiberoptic bronchoscope is always used to determine DLT position after the patient has been turned into the lateral decubitus position. Of course, a determined effort is made to prevent dislodgment of the tube during turning by holding onto the tube at the level of the incisors and by keeping the head absolutely immobile in a neutral or slightly flexed position. Head extension can cause movement of the tube in a cephalad direction, which may result in bronchial decannulation; head flexion can cause movement of the tube in a caudad direction, which may result in an upper lobe obstruction or in both lumens being in a main-stem bronchus (see the next section).63·64 Finally, the fiberoptic bronchoscope is used anytime during the procedure when there is a question about DLT position. This is not an infrequent occurrence and is usually caused by surgical manipulation and traction on the hilum, carina, or trachea.
There are some risks to the routine use of FOB to determine DLT position. First, depending on the relative sizes of the FOB and the DLT, the use of a self-sealing diaphragm in the elbow connector, and the position of DLT, ventilation may (need to) be interrupted during use of the FOB. Second, if the FOB is inserted through a self-sealing diaphragm in the elbow connector to the side being visualized, one needs to be cognizant of the continuous presence and status of the self-sealing diaphragm (i.e., a torn fragment is not carried/pushed into the tracheobronchial tree). Third, poor FOB technique may injure the mucosa of the tracheobronchial tree. Fourth, with poor cleaning technique, cross infection or direct mucosal injury (e.g., by Cidex) is possible. Finally, erroneous interpretation of the view is always possible. However, all of these complications are avoidable with proper education, forethought, protocols, and experience. Use of FOB, especially while the patient is being ventilated or has been hyperventilated on 100 per cent oxygen, should be no more distracting than inserting a central venous catheter, nasogastric tube, and so on, and should also be no more time consuming than the multiple, and often confusing, unilateral clamping and declamping/auscultation periods that are usually required when the DLT is not in the right position. 1. Margin of Safety in Positioning Double-Lumen Tubes A correctly positioned DLT does not cause obstruction of any conducting airway. This section discusses the length of adult male and female tracheobronchial trees, over which differently sized and manufactured DLTs can be moved and still be correctly positioned: This length is defined as the margin of safety.65 Because left- and right-sided DLTs are constructed differently and the left and right main-stem bronchi have different anatomy, the definitions of positioning left and right DLTs are unique. a. LEFT-SIDED DOUBLE-LUMEN TUBES (1) DEFINITION OF MARGIN OF SAFETY IN POSITIONING LEFT-SIDED DOUBLE-LUMEN TUBES. The
left-sided tubes made by different manufacturers have the same basic structural design and, therefore, can be considered together. The most proximal acceptable position of a left-sided tube is when the left endobronchial cuff is just below the tracheal carina (Fig. 9-22, top left panel). If the endobronchial cuff is placed in a progressively more proximal position, then the endobronchial cuff would progressively fill the space above the carina and obstruct the trachea and contralateral (right) main-stem bronchus. In addition, the posi-
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
355
The Margin of Safety (MS) in Positioning Double-Lumen Endotracheal Tubes
Figure 9-22 This schematic shows the definitions of most proximal and most distal acceptable positions of left- and rightsided double-lumen tubes and the margin of safety in positioning these double-lumen tubes. Top panel = all left-sided double-lumen tubes; middle panel = Mallinkrodt right-sided double-lumen tube; bottom panel = Rusch right-sided double-lumen tube; LMS = length left main-stem bronchus; RMS = length right main-stem bronchus; MS = margin of safety in positioning double-lumen tube; LUL = left upper lobe; RUL = right upper lobe. (From Benumof JL, Partridge BL, Salvatierra C, Keating J: Margin of safety in positioning modern double-lumen endotracheal tubes. Anesthesiology 67:729-738, 1987. Used with permission.)
tive-pressure gas and fluid seal between the two lungs would be lost. The most distal acceptable position of a leftsided tube is when the tip of the left lumen is at the proximal edge of the left upper lobe bronchial orifice (see Fig. 9-22, top right panel). If the tip of the left lumen is placed in a progressively more distal position, then the tip of the left lumen would progressively obstruct the left upper lobe bronchial orifice. This definition of the most distal acceptable position for a left-sided tube is valid, provided that two assumptions inherent in this definition are correct. First, the distance between the right and left lumen tips (see Table 9-5) must be greater than the length of the left main-stem bronchus (which it is); this means that the tip of the right lumen will not enter the left main-stem bronchus before the tip of the left lumen by-passes the left upper lobe. Second, the outside diameter of the left lumen tip of a left-sided tube must be nearly equal to the internal diameter of the second-gen-
eration bronchus to the left lower lobe (which it is); this means that there is no, or minimal, space between the tip of the left lumen and the left lower lobe bronchus for retrograde gas exchange between the left lumen and left upper lobe. The length of tracheobronchial tree between the most distal and most proximal acceptable position is the margin of safety in positioning a left-sided tube. This margin of safety is thus the length of the left main-stem bronchus minus the distance between the proximal margin of the left endobronchial cuff and the tip of the left lumen (see Fig. 922, top panel). (2) MEASUREMENT OF LENGTH OF LEFT MAIN STEM BRONCHUS. Table 9-3 shows the averages
and ranges of left main-stem bronchial lengths for three different measurement techniques (in vivo FOB, autopsy study, lung cast study).65 From study method to study method, and for the separate sex subgroups and the combined sex groups, the results are strikingly similar. In all subgroups and
356
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Table 9-3
MAIN-STEM BRONCHIAL LENGTHS FOR MALES, FEMALES, AND SEXES COMBINED IN THE IN VIVO FIBEROPTIC BRONCHOSCOPY, AUTOPSY AND CAST STUDY GROUPS*
*From Benumof JL, Partridge BL, Salvatierra C, Keating J: Margin of safety in positioning modern double-lumen endotracheal tubes. Anesthesiology 67:729-738. tValues are mean ± standard deviation. Values in parentheses are ranges. +One right upper lobe takeoff was above tracheal carina. Abbreviations: RMSB = right main-stem bronchus; LMSB = left main-stem bronchus; — = sex of the casts was unknown.
combined groups, the main-stem bronchial lengths were approximately normally distributed. The in vivo bronchoscopically determined lengths of left main-stem bronchus for males and females (±
standard deviation) are 50 ± 8, and 45 ± 7 mm, respectively, with a slight clinically and statistically significant positive correlation with height (Fig. 9-23).65 Considering all groups, the lengths of the left main-stem bronchus in males is 48 to 49 mm and in females 44 mm. (3) MEASUREMENT OF MARGIN OF SAFETY FOR LEFT-SIDED DOUBLE-LUMEN TUBES. The margin
of safety in positioning a left-sided DLT is the difference between the length of the left main-stem bronchus and the length of tube between the proximal margin of the left endobronchial cuff and the left lumen tip (Table 9-4). This average margin of safety in positioning left-sided DLTs ranged from 16 to 19 mm. DLTs with a carinal hook have the hook placed approximately 10 mm proximal to the proximal surface of the endobronchial cuff. Consequently, the tip of the left lumen of a left-sided DLT with a carinal hook enters the left main-stem bronchus 10 mm deeper than a tube without a carinal hook,66 and the margin of safety is reduced by approximately 10 mm. (4) CLINICAL EXAMPLES/IMPLICATIONS OF EXCEEDING MARGIN OF SAFETY FOR LEFT-SIDED DOUBLE-LUMEN TUBES. Figure 9-24, left panel,
Figure 9-23 In vivo fiberoptic bronchoscopy measured length of the left main-stem bronchus (LMS) plotted as a function of the height of the patients. A weak but statistically significant correlation was found. (From Benumof JL, Partridge BL, Salvatierra C, Keating J: Margin of safety in positioning modern double-lumen endotracheal tubes. Anesthesiology 67:729-738, 1987. Used with permission.)
shows an example of a common situation using a 37 French clear plastic left-sided DLT in an average-sized female. With the left endobronchial cuff placed just below the tracheal carina, and using average values for DLT and left main-stem bronchus lengths, the margin of safety is 21 mm. However, because the distance between the right and left lumen tips for the clear plastic tube (69 mm; see Table 9-5) is longer than the length of the left main-stem bronchus (50 mm) in this typical example, it is possible for the right lumen to be above the tracheal carina while the left lumen tip obstructs the left upper lobe (see Fig. 9-24, right panel). Figure 9-25 shows a postoperative chest
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
357
Table 9-4 AVERAGE MARGIN OF SAFETY IN POSITIONING A LEFT-SIDED DOUBLE-LUMEN TUBE IS THE LENGTH OF THE LEFT MAIN-STEM BRONCHUS MINUS THE PROXIMAL MARGIN OF LEFT CUFF TO LUMEN TIP LENGTH* Average Length of Left Main-Stem Bronchus, mm Proximal Margin of Average Margin of Safety, mm Left Cuff to Left Size Lumen Tip, mm Female Combined Male Female Combined Manufacturer 35-41 -Fr Male Mallinkrodt Rusch Sheridan
All AH All
49 49 49
44 44 44
48 48 48
29t 32 30
20 17 19
15 12 14
19 16 18
*From Benumof JL, Partridge BL, Salvatierra C, Keating J: Margin of safety in positioning modern double-lumen endotracheal tubes. Anesthesiology 67:729-738, 1987. Used with permission. tManufacturer estimates a ± 2 mm (variation) tolerance for this value.
roentgenogram of a patient who underwent a right upper lobectomy and demonstrates this problem: A left-sided DLT tube is in situ, the right lumen tip is above the tracheal carina, and the remaining right lung is well aerated, while the left upper lobe is partially collapsed. The same situation may occur when a left-sided DLT with a carinal hook is used (because the hook is proximal to the proximal surface of the left cuff) (i.e., the distance between
the left tip and hook exceeds the length of the left main-stem bronchus).66 In addition to significant variability in length of the left main-stem bronchus, there is also fairly significant variability (up to 20 per cent) between individual left-sided DLTs of different manufacturers and same-size and different-size DLTs of the same manufacturer with respect to right and left lumen tip and proximal endobronchial cuff to
37 French Left-Sided Double-Lumen Tube and Adult Female Left Main Stem Bronchial Dimensions and Relationships
Figure 9-24 This schematic diagram shows the relationship of the lumen tips and cuffs of a correctly and an incorrectly positioned 37 French left-sided double-lumen tube to the tracheobronchial tree of an average-sized adult female. When the left endobronchial cuff is just below the tracheal carina (correct position), the margin of safety (MS) is 21 mm because the length of the left main-stem bronchus (LMS) exceeds the distance between the cephalad surface of the left cuff and the tip of the left lumen (distance A). However, because the distance between the right and left lumens (distance B) exceeds the length of the left main-stem bronchus, it is possible to still have the right lumen above the tracheal carina while the left lumen tip and the left endobronchial cuff obstruct the left upper lobe (LUL). Thus, as a left-sided tube is inserted further (so that the left cuff cannot be visualized below the carina), the first airway likely to be obstructed will be the left upper lobe.
358
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Figure 9-25 This x-ray shows an example of a patient with the problem schematically depicted in Figure 9-22. The right lumen is above the tracheal carina, and the right lung is well ventilated, while the left lumen tip is deeply inserted into the left lung, and the left upper lobe is opacified.
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Table 9-5 LEFT TO RIGHT LUMEN TIP DISTANCE FOR VARIOUS SIZED AND MANUFACTURED DOUBLELUMEN TUBES Left-Sided Double-Lumen Tube Manufacturer Mallinkrodt broncho catheter
Carlens
Leyland
French
Left to Right Tip Distance (mm)
41 39 37 35 41 39 37 35 44 38 32
70 70 69 66 67 65 64 60 90 82 72
left lumen tip lengths. The 20 per cent variation in these distances between manufactures and sizes in the clear plastic DLTs is due to slightly different designs by the manufacturers and the fact that cutting of the proximal lumen and placement of the endobronchial cuff is done by hand at the end of the manufacturing process. The variation in DLT lengths (lumen to lumen and left tip to proximal endobronchial cuff) and left main-stem bronchial length greatly reinforces the need to determine DLT location with a fiberoptic bronchoscope. The variation in these lengths has three more important implications. First, a patient with a small left main-stem bronchus will have a smaller margin of safety. In fact, if a patient with a small left main-stem bronchus (two standard deviations less than the mean presented in Table 9-3) received a 37 French plastic tube that had a large left tip to proximal left cuff distance and a large lumen to lumen distance (20 per cent greater than the mean value in Table 9-5), there would be no margin of safety, even if the left endobronchial cuff is positioned just below the tracheal carina (Fig. 9-26); this situation has the smallest margin of safety realistically possible. Indeed, with a short left main-stem bronchus, it is possible to have the left endobronchial cuff beyond the left upper lobe so that the left lower lobe is ventilated by the left lumen while the left upper lobe and right lung are ventilated by the right lumen (Fig. 9-27). 40 · 41 Second, and conversely, if a patient with a long left main-stem bronchus (two standard deviations greater than the mean presented in Table 9-3) received a 37 French plastic DLT with a short lumen to lumen distance (20 per cent less than the mean value in Table 9-5), the right lumen could be in the left main-stem bronchus (right main-stem
359
bronchus obstructed) while the left lumen tip is still above the left upper lobe (Fig. 9-28). However, it must be emphasized that no matter which manufactured tube or sized tube is used and no matter how long or short the left main-stem bronchus is (within the range of extremes observed in the cadaver studies), when the upper surface of the left endobronchial balloon is just below the tracheal carina, it is not possible for the left lumen tip to obstruct the left upper lobe or the right (tracheal) lumen to be near a main-stem bronchus. Third, in view of the fact that head flexion and extension can move the tip of a left-sided DLT in or out 27 mm,64 these considerations show that it is easily possible to cause left upper lobe obstruction with head flexion and bronchial decannulation with head extension (Fig. 9-29). Finally, larger, rather than smaller, sized leftsided DLTs should be used. As tube size increases, proximal margin of left cuff to left lumen tip length (all manufacturers) remains constant. Although the margin of safety remains constant with increasing DLT size, airway resistance and difficulty in secretion removal decrease. In view of the positive correlation between patient height and
Example of No Margin of Safety (MS) When Left Main Stem Bronchus (LMS) is Short and Length A is Large
MS=0
Left Cuff Just Below Tracheal Carina
Figure 9-26 This schematic diagram shows that there will be no margin of safety (MS) when a double-lumen tube is used that has a large length A in a patient who has a short left mainstem bronchus (LMS), even though the left endobronchial cuff is positioned just below the tracheal carina. Nevertheless, with the left endobronchial cuff just below the tracheal carina, the left upper lobe is unobstructed.
360
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Figure 9-27 Using a left double-lumen tube, after clamping the left (endobronchial) lumen, the left upper lobe and entire right lung continued to be ventilated. Withdrawing the tube back 2 cm corrected the problem. (From Brodsky JB, Shulman MS, Mark JBD: Malposition of left-sided double-lumen endobronchial tubes. Anesthesiology 62:667-669, 1985. Used with permission.)
main-stem bronchial length, and with the provision that an intubation can be atraumatically performed, use of 39 and 41 French tubes in patients who are 68 inches (170 cm) or taller and 35 to 39 French tubes in patients who are shorter appears indicated. b. RIGHT-SIDED DOUBLE-LUMEN TUBES (1) DEFINITION OF MARGIN OF SAFETY IN POSITIONING RIGHT-SIDED DOUBLE-LUMEN TUBES.
The Mallinkrodt, Sheridan, Rusch, and Leyland right-sided DLTs are designed differently (the shapes of the right endobronchial cuffs are very different); therefore, they must be considered separately. Although the Sheridan double-balloon configuration of the right-sided tube is unique, the positioning margin of safety is similar to that of the Mallinkrodt right-sided DLT. The most proximal acceptable position of a Mallinkrodt rightsided tube is when the endobronchial cuff is just below the tracheal carina (Fig. 9-22, middle left panel). If the endobronchial cuff is placed in a progressively more proximal position, then the right endobronchial cuff would progressively fill
the space above the carina and obstruct the trachea and contralateral (left) main-stem bronchus. In addition, the positive-pressure gas and fluid seal between the two lungs would be lost. The most distal acceptable position of the Mallinkrodt right-sided tube is when the distal margin of the right endobronchial cuff is at the proximal margin of the right upper lobe bronchial orifice (see Fig. 9-22, middle right panel). If the Mallinkrodt (or Sheridan) right endobronchial cuff is placed in a progressively more distal position, then the Mallinkrodt right endobronchial cuff would progressively obstruct the right upper lobe bronchial orifice. This definition of the most distal acceptable position for a Mallinkrodt (or Sheridan) right-sided tube is valid because of the unique shape of the Mallinkrodt (or Sheridan) right endobronchial cuff (which forces the right upper lobe ventilation slot to ride off the right main-stem bronchial wall), which allows gas exchange between the right upper lobe ventilation slot, right lumen tip (for the Mallinkrodt tube), and the right upper lobe (even though the right upper lobe ven-
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker intubation)
Example of Right Lumen Tip Being in Left Main Stem Bronchus when Left Lumen Tip is Above Left Upper Lobe When Left Main Stem Bronchus (LMS) is Long and Length Β is Short
361
bronchial orifice. If the Rusch or Leyland rightsided tube is placed in a progressively more distal position, then the proximal margin of the right endobronchial cuff and outside lateral wall of the right endobronchial lumen would progressively obstruct the right upper lobe bronchial orifice. These definitions of the most proximal and most distal acceptable position for a Rusch and Leyland right-sided tube are valid because the shape of the Rusch and Leyland right endobronchial cuff forces the right upper lobe ventilation slot and outside lateral wall of the right endobronchial tube to be in apposition to the right upper lobe bronchial ori fice and lateral bronchial mucosa. The margin of safety in positioning the Rusch and Leyland rightsided tube is, therefore, the distance between the distal and proximal margins of the right endobron chial cuff (i.e., the length of the right upper lobe ventilation slot) minus the diameter of the right upper lobe bronchial orifice (see Fig. 9-22, bottom panel). ( 2 ) MEASUREMENT
OF
LENGTH
OF
RIGHT
MAIN-STEM BRONCHUS AND RIGHT UPPER LOBE BRONCHIAL ORIFICE. Table 9-3 shows the aver
Figure 9-28 This schematic diagram shows that it is possi ble for the right lumen to be in the left main-stem (LMS) bronchus when the left lumen tip is still above the left upper lobe if the left main-stem bronchus is long and length Β is short.
tilation slot may not be properly aligned with the right upper lobe bronchial orifice). The margin of safety in positioning the Mallinkrodt right-sided tube is, therefore, the length of the right main-stem bronchus minus the distance between the proximal and distal margins of the right endobronchial cuff (i.e., the length of the right endobronchial cuff) (see Fig. 9-22, middle panel). The most proximal acceptable position of the Rusch and Leyland right-sided tube is when the distal margin of the right endobronchial cuff is at the distal margin of the right upper lobe bronchial orifice (see Fig. 9-22, bottom panel). If the Rusch or Leyland right-sided tube is placed in a progres sively more proximal position, then the distal margin of the endobronchial cuff and the outside lateral wall of the right lumen tip would progres sively obstruct the right upper lobe bronchial ori fice. The most distal acceptable position of a Rusch or Leyland right-sided tube is when the proximal margin of the right endobronchial cuff is at the proximal margin of the right upper lobe
ages and ranges of right main-stem bronchial lengths for three different measurement techniques (in vivo FOB, autopsy study, cast study).65 From study method to study method and for the separate sex subgroups and the combined sex groups, the results are very similar. In all subgroups and com bined groups, the main-stem bronchial lengths are approximately normally distributed. Considering all groups, the length of the right main-stem bron chus is 18 to 19 mm in males and 13 to 14 mm in females. The average diameters (± standard de viation) of the right and left upper lobe bronchial orifices in the lung casts were 10.7 ± 2.1 mm and 10.3 ± 1.9 mm, respectively, with a range of values of 6 to 18 mm. The distribution of values was approximately normal. These findings are nearly identical to those of a previous and more comprehensive, but not as quantitative, study (Fig. 67 9-30). (3) MEASUREMENT OF MARGIN OF SAFETY FOR
The margin of safety in positioning right-sided tubes is shown in Table 9-6. The margin of safety ranged from 1 to 10 mm and differed between manufacturers. The Leyland right-sided DLT has a greater margin of safety than the Rusch (or Mallinkrodt) rightsided DLT because it has a long (21 mm) right upper lobe ventilation slot. However, it should be remembered that the unique slanted doughnut shape of the Mallinkrodt right endobronchial cuff of the right-sided DLT allows the right upper lobe ventilation slot to ride off the right upper lobe orifice, thereby increasing (from 1 mm) to an un known extent the margin of safety in positioning RIGHT-SIDED DOUBLE-LUMEN TUBES.
362
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker intubation)
Left Double Lumen Tube Position and Head Flexion and Extension
Figure 9—29 Head flexion moves an endotracheal tube inward, and head extension moves an endotracheal tube outward. A, Correct position of a left-sided double-lumen tube along with average values (in mm) for left main-stem bronchus, right to left lumen tip, and left lumen tip to left upper lobe lengths; the latter length is the margin of safety. When the left cuff is just below the tracheal carina, the margin of safety is 25 mm. B, Extreme head flexion can cause left upper lobe obstruction. C, Extreme head extension can cause left main-stem bronchial decannulation. Data from Saito, Dohi, and Naito.62
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
363
upper lobe takes off from the trachea (1:250 in normal patients68 and 1:50 patients with congenital heart disease69). ( 4 ) CLINICAL EXAMPLES/IMPLICATIONS OF EXCEEDING MARGIN OF SAFETY FOR RIGHT-SIDED
This analysis of rightsided DLT margin of safety has five clinical implications. First, with a right main-stem bronchus less than 10 mm long or when the right upper lobe has a tracheal takeoff, it will be impossible to position a right-sided DLT without obstructing the right upper lobe. Indeed, if the right upper lobe has a low tracheal origin and a left-sided DLT is inserted too deeply, the tracheal cuff may obstruct the right upper lobe and the left lumen/cuff may obstruct the left upper lobe (Fig. 9-31).70 Second, a clinical prospective study confirms that the Leyland rightsided DLT has a greater positioning margin of safety than the Mallinkrodt tube.53 In 18 of 20 patients intubated with a Leyland right-sided DLT, compared with only one of nine patients intubated with a Mallinkrodt right-sided DLT, the right upper lobe was bronchoscopically patent when the patient was supine. Third, the small margin of safety for right-sided DLTs, along with the large variation in the length of the right main-stem bronchus, indicates that whenever all other considerations are equal, a left-sided DLT is preferable to a right-sided DLT. Fourth, tolerance of head movement with a right-sided DLT will be obviously much less than for a left-sided DLT. Fifth, if a right-sided DLT has to be used, the clear plastic disposable ones may be best because of the slanted doughnut shape of the right endobronchial cuff DOUBLE-LUMEN TUBES.
Figure 9-30 Mean tracheobronchial tree diameter (in cm) Note that the diameter of the right upper lobe bronchial orifice is 1 cm. (From Merendino KA, Keriluk LB: Human measurements involved in tracheobronchial resection and reconstruction procedures. Surgery 35:590-597, 1954. Used with permission.)
this particular right-sided tube. The margin of safety in positioning right-sided DLTs may be negative if the right main-stem bronchus is less than 10 mm (1:6 normal patients)68 or if the right
Table 9-6 AVERAGE MARGIN OF SAFETY IN POSITIONING A MALLINKRODT RIGHT-SIDED DOUBLE-LUMEN TUBE IS THE LENGTH OF THE RIGHT MAIN-STEM BRONCHUS MINUS THE WIDTH OF THE RIGHT ENDOBRONCHIAL CUFF, AND THE MARGIN OF SAFETY IN POSITIONING A RUSCH RIGHT-SIDED DOUBLE-LUMEN TUBE IS THE LENGTH OF THE RIGHT UPPER LOBE VENTILATION SLOT MINUS THE DIAMETER OF THE RIGHT UPPER LOBE BRONCHIAL ORIFICE*
364
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
French tubes in patients who are shorter appears indicated. In summary, the analyses of margin of safety in positioning left- and right-sided DLTs indicate that if a question arises concerning DLT position (using conventional unilateral clamping and auscultation methods), blind attempts to adjust the position of the DLT have a very good chance of being unsuccessful and/or uncertain. Instead, the DLT position should be checked by FOB, and for a left-sided DLT the endobronchial cuff should be positioned just below the tracheal carina (and, therefore, left upper lobe obstruction will not be possible). When a DLT is used for clinical research purposes, a fiberoptic bronchoscope must be used to confirm the proper DLT position to prevent otherwise unrecognizable upper lobe obstruction and to eliminate this possibility of gathering uninterpretable and nonrepresentative data.
2. Relationship of Fiberoptic Bronchoscope Size to Double-Lumen Tube Size
Figure 9-31 The left double-lumen tube was advanced down the trachea until moderate resistance to further passage was encountered. With the tube in this position while ventilating the patient through both lumens of the double-lumen tube, the left upper lobe bronchus was obstructed by the inflated bronchial cuff while the right upper lobe bronchus, which originated in the trachea, was simultaneously obstructed by the inflated tracheal cuff. (From Brodsky JB, Mark JBD: Bilateral upper lobe obstruction from a single double-lumen tube. Anesthesiology 74:1163-1164, 1991. Used with permission.)
(Mallinkrodt), which decreases the likelihood of right upper lobe obstruction by causing the right upper lobe ventilation to be pushed away from the lateral wall of the main-stem bronchus by the endobronchial cuff (see Fig. 9-9). Finally, a larger, rather than smaller, right-sided DLT should be used. As tube size increases, length of Mallinkrodt right cuff remains constant, and length of Rusch right upper lobe ventilation slot increases; thus, the margin of safety either remains constant (Mallinkrodt right-sided tubes) or increases (Rusch rightsided tubes) with increasing tube size. Although the margin of safety remains constant or increases with increasing DLT size, airway resistance and difficulty in secretion removal decrease. In view of the positive correlation between patient height and main-stem bronchial length, and with the provision that an intubation can be atraumatically performed, the use of 39 and 41 French tubes in patients who are 68 inches (170 cm) or taller, and 35 to 39
The clear plastic disposable right- and left-sided double-lumen endotracheal tubes are manufactured in five sizes: 28, 35, 37, 39, and 41 French. A 5.6mm outside-diameter diagnostic fiberoptic bronchoscope will not pass down the lumens of any sized DLT. A 4.9-mm external diameter fiberoptic bronchoscope passes easily through the lumens of the 41 French tube, passes moderately easily with lubrication through the 39 French tube, causes a tight fit that needs a liberal amount of lubrication and a strong pushing force to pass through the 37 French tube, and does not pass through the lumen of the 35 French tube. A silicon-based fluid (such as that made by the American Cystoscope Co.) is the best lubricant for a fiberoptic bronchoscope because it does not dry out or crust and does not interfere with the view even if it coats the tip of the bronchoscope. Fortunately, from the point of view of using a 4.9-mm outside-diameter fiberoptic bronchoscope, a 37 French tube or larger can be used in almost all adult females and a 39 French tube or larger can be used in almost all adult males. A 3.6- to 4.2-mm outside-diameter (pediatric) fiberoptic bronchoscope passes easily through the lumens of all sized double-lumen endotracheal tubes and, because the bronchoscope has an increased amount of space, the maneuverability of the tip of the bronchoscope is greatly increased. Therefore, the 3.6- to 4.2-mm outside-diameter bronchoscope is obviously the bronchoscope of choice for DLTs. Table 9-7 summarizes these fiberoptic bronchoscope DLT relationships. Several companies (Olympus, Machida, Pentax) presently
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
365
Table 9-7 RELATIONSHIP OF FIBEROPTIC BRONCHOSCOPE SIZE TO DOUBLE-LUMEN TUBE SIZE Fiberoptic Bronchoscope Size Outside Diameter (mm)
Double-Lumen Tube Size (French)
Fit of Fiberoptic Bronchoscope Inside Double-Lumen Tube
5.6
All sizes
Does not fit
4.9
41 39 37 55
Easy passage Moderately easy passage Tight fit, needs lubricant,* hard push Does not fit
3.6-4.2
All sizes
Easy passage
*Lubricant recommended is a silicon-based fluid made by the American Cystoscope Company.
manufacture 4.9- and 3.6- to 4.2-mm outside-diameter fiberoptic bronchoscopes that are of adequate length and have a suction channel. E. Use of Chest X-Ray to Determine Double-Lumen Tube Position The chest roentgenogram can be used to determine DLT position. The usefulness of a chest roentgenogram may be greater than conventional unilateral auscultation and clamping in some patients, but it is always less precise than FOB. To use the chest roentgenogram, the DLT must have radiopaque markers at the end of the right and left lumens. The key to discerning DLT position on the chest roentgenogram is seeing where the marker at the end of the tracheal lumen is in relation to the tracheal carina and whether the endobronchial lumen located in the correct main-stem bronchus. The end of the tracheal lumen marker must be above the tracheal carina; however, this does not guarantee correct position because this technique may not reveal a subtle obstruction of an upper lobe (see Fig. 9-25 and Margin of Safety in Positioning Double-Lumen Tubes). If the tracheal carina cannot be seen (which is often the case with a portable anterior-posterior film), then the chest roentgenogram method of determining DLT position is not usable. Furthermore, the chest roentgenographic method is time consuming (for film transport, film development), costly, and awkward to perform and may dislodge the tube (the cassettes are often difficult to place under the operating room table and may require movement of the patient). Instillation of 1 ml of a radiopaque marker into the endobronchial cuff (diatrizoate meglumine and diatrizoate sodium solution [Renografin 60] is water soluble and is easy to inject but can be irritating to the bronchial mucosa if it were to leak from a cuff) could enhance the utility of using the chest X-ray to determine DLT position.
F. Other Methods to Determine DoubleLumen Tube Position Three other methods may help to determine the position of a DLT. First, comparison of capnography (waveform and P E T C 0 2 value) from each lumen may reveal a marked discrepancy. For example, and all other conditions being equal, one lung may be very poorly ventilated in relation to the other lung (high P E T C 0 2 ) , indicating obstruction to that lung; one lung may be extremely overventilated in relation to the other lung (low P E T C 0 2 ) , indicating, perhaps, just ventilation of a lobe of that lung; or the capnogram from a given lung may have a much steeper slope to the alveolar plateau, indicating exhalatory obstruction.72·73 Second, continuous spirometric data (Datex Capnomac Ultima) from both lungs and each individual lung, such as pressure-volume or flow-volume loops. may be displayed and compared to a control loop that is stored in memory.74 Third, the surgeon may be able to palpate the position of the DLT from within the chest and may be able to redirect or assist in changing the position of the DLT (by deflecting the DLT away from the wrong lung. etc.).75
G. Securing the Double-Lumen Tube After the correct placement of a double-lumen endotracheal tube, it is customary to secure firmlv the DLT at the level of the lips using adhesive or umbilical tape. A simple and practical method for securing the DLT is to use a simple single tie around the tube at the level of the lips, followed by a bow-tie knot at the bifurcation of the DLT connector (Fig. 9-32). 76 Using this technique, stabilization of the tube is not dependent on tight knots around the tube itself but is instead accomplished by the bow-tie knot around the connector bifurcation, preventing the tube from pulling out.
366
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Figwe 9—32 TYiç. dwtate-Vavwtn votat \% %ţab\lV/fcd w\iK & t\«. around it at the ievei of hps and a bow tie at the bifurcation of the lumens of the double-lumen tube. (From Cohen E, Koorn R: An easy way to safely tie a double-lumen tube. J Cardiothorac Anesth 5:194-195, 1991. Used by permission.)
If readjustments need to be made, the DLT can be freed by simply pulling on either end of the bowtie knot. The method is simple, avoids outward displacement, and most importantly allows the tube to be freed (untied) in a single maneuver for proper positioning and manipulation. The timeconsuming struggle of untying multiple knots in a contaminated, wet, and slippery environment is avoided.
H. Quantitative Determination of CuffSeal Pressure Hold The use of FOB to determine DLT position does not provide evidence or a guarantee that the two lungs are functionally separated (i.e., against a fluid and/or air pressure gradient). There are times,
such as during the performance of unilateral pul monary lavage, when the anesthesiologist must be absolutely certain that functional separation has been achieved. Complete separation of the two lungs by the left endobronchial cuff can be dem onstrated in a left-sided tube by clamping the con necting tube to the right lung proximal to the right suction port and attaching a small tube (i.e., intra venous extension tubing) to the open right suction port (by appropriate adaptors) (Fig. 9-33). The free end of this tube is submerged in a beaker of water. When the left lung is statically inflated to any pressure considered necessary, and the left endobronchial cuff is not sealed, air will enter the left lung as well as escape out from around the unsealed left cuff, up the right lumen to the small connecting tube, and bubble through the beaker of water (Fig. 9—335). If the left endobronchial cuff is sealed, no bubbles should be observed passing through the beaker of water (Fig. 9-33Λ). Following demonstration of functional lung sep aration, the right connecting tube is undamped, the right suction port is closed, and ventilation to both lungs is resumed. To test for lung separation with the pressure gradient across the endobronchiâJ baboon reversed, îhe tefî airway connecting tube is clamped proximai to the feft suction port, the left suction port is opened to the beaker of water via the small tube, the right lung is statically inflated to any desired pressure, and the absence or presence of air bubbles in the beaker of water Yi *?& B,
+
10 to 25
25
340 fig/puff
+++
Tachyphylaxis
1 PA
Abbreviation: IV = intravenously.
1 to 2
Poor choice
458
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
bronchomotor and peripheral vascular tone. Iso proterenol is a very potent bronchodilator, but it is not β2 selective and results in tachycardia and ar rhythmias. The propensity toward tachycardia and arrhythmias is exaggerated by halothane and min imized by isoflurane. In addition, isoproterenol is not as effective by the tracheal route. Conse quently, it is fortunate that the following new drugs, with substitutions in various parts of the catecholamine structure, have been developed; they differ from isoproterenol in that they have higher β2 selectivity, may be administered by the tracheal route, and have a longer action of dura tion.12 Albuterol (salbutamol) can be administered as a metered aerosol and, by this route, is as effective a bronchodilator as isoproterenol is intravenously but has fewer cardiovascular effects. Nevertheless, salbutamol may still increase cardiac performance by decreasing systemic vascular resistance and in crease cardiac contractility.14 Its duration of action is approximately 6 hours. Terbutaline is now avail able as a metered aerosol, and a water-soluble preparation is available for nebulizer use. Metaproterenol and isoetharine are slightly less effec tive, have slightly less β2 selectivity, and are shorter acting than albuterol. They offer no advan tage. Second, although no clinical studies have been done with anesthetized patients, combining the anticholinergic inhalant ipratropium (Atrovent) to P2-agonists appears to be a reasonable option to consider.15·16 Three studies17-19 have indicated that the combination of an anticholinergic agent, ipra tropium bromide, to β 2 ^οηΪ8ΐ8 is more effective in severe acute asthma than either agent alone. This additive effect is not surprising because the anticholinergic agents differ in their mode of ac tion (blockage of efferent limb of irritant neural reflex) from β2^οηΐ8ί8 (blockage/antagonism of bronchoconstrictor mediator). Figure 13-3 shows typical results from one study of awake asthmatics and patients with chronic bronchitis.17 No increase in side effects has been observed when ipratro pium has been used in combination with other bronchodilating drugs. Because parasympatholytic drugs have little or no effect on mediator-induced airway constriction, they are often more effective in patients with chronic bronchitis than in those with asthma20 perhaps because, generally, they de crease the volume of secretions in and, specifi cally, in response to irritation/instrumentation of the airway; therefore, they reduce bronchoconstricting reflexes. Thev bronchodilating effect of anticholinergic inhalants peak at 1 hour. Corticosteroids may be extremely useful in treating the inflammatory basis of bronchospasm, but it should be realized that the onset of therapeu
tic action may require 1 to 3 hours,21 and the mechanism of action in the modulation of airway tone is incompletely understood. Proposed mecha nisms include stabilization of membranes, de creased bronchial mucosal edema and inflamma tory cells, and potentiation of catecholamine action. Despite the 1 to 3 hours of onset of action, the initiation of intraoperative treatment will ob viously have postoperative benefits.1·22 The advent of an aerosol steroid offers a major advantage in the treatment of chronic asthma. Side effects are greatly minimized, but adrenal suppres sion may still occur. Beclomethasone dipropionate (Beclovent) and triamcinolone (Azmacort) appear to be the best inhaled agents currently available. Nevertheless, in the acute intraoperative setting, intravenous steroids are logically more reliable and quick acting. Hydrocortisone is considered the ste roid of choice for parental administration in both the preoperative preparation of patients with reac tive airway disease (1-2 mg/kg) and the treatment of intraoperative bronchospasm (4 mg/kg). Methylprednisolone, 2 mg/kg, may also be used. Finally, theophylline compounds may be bene ficial prophylactically in asthmatic patients and therapeutic in bronchospastic patients. Phosphodi esterase inhibition is the major mechanism of ac tion of this drug in contrast to direct β2 stimula tion; this allows aminophylline to potentially be useful in the patient who is taking beta blockers or is already using a β 2 ^οηΐ8ί. Side effects include hypotension and/or cardiac arrhythmias, especially during anesthesia with halogenated drugs. The ar rhythmias are magnified by halothane and mini mized by isoflurane. In addition, aminophylline may not cause additional bronchodilation in the presence of 1.5 MAC halothane.23 Therefore, if optimal doses of β 2 ^οηΐ8ΐ8 are being given, there appears to be no advantage and several disadvan tages to using aminophylline.1 Should a patient require/receive aminophylline during anesthesia, a loading dose of 5 to 6 mg/kg should be given intravenously and slowly during 15 to 20 min (which is a disadvantage compared with the β2agonists) followed by an infusion of 0.5 to 1.0 mg/kg/hour depending on age (inversely propor tional). Theophylline blood levels should be ob tained on all patients, and the therapeutic range is 10 to 20 μ^πιΐ. The third category of therapy involves the ad ministration of compounds that should be expected to have a minor impact on bronchospasm but may, in selected patients, be helpful, especially when administration of the major bronchodilating drugs has not completely resolved the problem (see Fig. 13-1, category III). If the increase in airway resis tance is thought to be a result of coughing or straining (reflex induced), intravenous lidocaine, 1
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
459
Figure 13-3 FEV, increased in 11 patients with asthma and 10 patients with chronic bronchitis after treatment with ipratropium (40 μg four times a day) alone, salbutamoi (200 μg four times a day) alone, and the combination of both drugs. (Reproduced with permission from Lightbody JM, Ingram CG, Legge JS, Johnston RN: Ipratropium bromide, salbutamoi and prednisolone in bronchial asthma and chronic bronchitis. Br J Dis Chest 72:181-186, 1978.)
mg/kg, should be administered.24 Additionally, 4 per cent lidocaine may be administered intracheally (down the tracheal lumen of a doublelumen tube for anesthesia of the carina and one main-stem bronchus and down the bronchial lu men for anesthesia of the other main-stem bron chus). Both forms of lidocaine administration in cur little risk and may be of significant benefit (they decrease coughing reflex and bronchospasm and protect against or treat ventricular arrhyth mias). Finally, sodium bicarbonate should be adminis tered following blood-gas determination if a met abolic acidosis is present (acidosis inhibits the action of catecholamines). Diphenhydramine (a histamine blocker) may be administered to supple ment the level of anesthesia. However, administra tion of an antihistamine should not be expected to have a large effect on the degree of bronchospasm because histamine is usually not the sole mediator involved in producing bronchoconstriction, and the antihistamines do not block reflex airway con striction, which is so important during thoracic surgery. Cromolyn is a mast cell stabilizer and is thought to inhibit release of mediators from the mast cells. The action of the drug is entirely pro phylactic, and it is not useful in the treatment of
bronchospasm intraoperatively. However, it is a remarkably benign drug in terms of interaction with anesthetics, and patients who are receiving this drug should continue to receive it right up to the time of surgery. Mucolytics have no proven value in the treatment of bronchospastic disease. Indeed, they have been shown to provoke reflex bronchoconstriction and are best avoided.
III. MANAGEMENT OF BLOOD LOSS A. Assessment of Blood Loss The anesthesiologist should be continuously aware of the amount of blood loss. The assessment of blood loss involves both direct observation and measurement of physiologic responses as indirect indicators of blood loss (Fig. 13-4).
7. Direct Observation Surgical swabs and packs should be observed for the amount of blood staining. Fully stained 4 by 4 gauze sponges and laparotomy pads contain 10 and 50 ml of blood, respectively. If there is a
460
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Assessment of Blood Loss
Figure 13—4
See legend on opposite page
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
question or concern about the exact amount of blood on the sponges and pads, the sponges and pads should be weighed; each gram greater than the known dry weight represents 1 ml of blood loss. The suction bottles should be frequently observed for blood accumulation. The operative field, the drapes around the operative field, and the floor should also be assessed for blood loss. Furthermore, if uncertainty about the amount of blood loss persists, there should be periodic attempts to look under the drapes and other concealed places for blood loss. With all these forms of direct observation, the anesthesiologist must be aware of how much saline irrigation was used (to wash out pleural cavity, test integrity of bronchial stumps) and how much the irrigation might have contributed to the observed shed volumes. Similar considerations apply to pleural fluid that was present preoperatively and then suctioned intraoperatively. With regard to the aspirated blood in the suction bottles, the hematocrit of the material in the suction bottles can always be measured and evaluated in light of the blood loss, intravenous fluid infusion, and surgical irrigation and suction history. Usually these data, when processed by the anesthesiologist, circulating nurse, and surgeon, result in a reasonably accurate educated estimate of the blood loss. Tying off of the pulmonary vessels for lung resection results in blood loss due to entrapment of blood within the specimen. If the pulmonary veins are tied off first, the specimen may be very blood-engorged (due to continued inflow) and with a pneumonectomy the resected lung may contain 500 ml of blood. If the pulmonary arteries are tied off first, then the entrapped blood loss in the specimen is much less. Postoperative drainage adds to this blood loss, and after pneumonectomy the pleural space fills with an unknown but considerable volume of blood and plasma.
2. Physiologic Response to Blood Loss The patient's physiologic response to fluid and blood loss (Fig. 13-4) is an indirect indicator of how much has been lost and how adequate replacement has been. In the face of continued blood loss and replacement with just crystalloid infusion, serial measurement of the hematocrit and hemo-
globin will reveal a progressive decrease. Hemodynamically, blood loss causes a decrease in systemic arterial, central venous, pulmonary arterial systolic, diastolic, and wedge pressure, cardiac output, and urine output and an increase in systolic pressure variation with respiration (i.e., the variation in the difference between maximal and minimal systolic blood pressure during the respiratory cycle). The heart rate and systemic vascular resistance usually increase with blood loss if the depth of anesthesia is not profound. The heart rate may not increase in patients taking preoperative beta blockers.
B. Minimal Blood Loss ( 10 L/min) and hy potension still exists despite high infusion rates of dopamine or epinephrine, norepinephrine can be tried. Norepinephrine is the most potent vasocon-
474
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Figure 13-9 This figure lists the important hemodynamic variables that may change during anesthesia and the surgical procedure (left-hand column). These hemodynamic variables may change in either direction (middle column), and there are multiple therapeutic options for treating any change in any hemodynamic variable in any direction (right-hand column). See the text for a full explanation.
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
475
Table 13-14 HEMODYNAMICALLY ACTIVE DRUGS COMMONLY ADMINISTERED BY IV INFUSION IN THE PERIOPERATIVE PERIOD
Abbreviations: VF = ventricular fibrillation; VT potassium; q = every.
ventricular tachycardia; D?W = 5 per cent dextrose in water; Κ
Tabe 13-15 TYPICAL COMMON HEMODYNAMIC SYNDROMES AND THEIR TREATMENT
476
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
strictor of both the venous and arterial vascular systems. At low infusion rates of less than 2 μg/ min (less than 0.03 μg/kg/min) there is a mixed alpha- and beta-adrenergic effect; however, above this rate, effects from stimulation of the a,-adre nergic receptor will predominate. In this situation, "pure" a,-agonists such as phenylephrine or methoxamine may also be used. Whenever a sympathomimetic agent other than dopamine is used to maintain perfusion pressure and cardiac output, renal blood flow is always reduced. Maintaining a low infusion rate of dopa mine (2-3 μg/kg/min) in conjunction with another vasoactive agent may help maintain urine output. Sympathomimetic support should always be weaned as rapidly as possible. Continuous sympa thomimetic infusions will decrease receptor sensi tivity to catecholamines and potentially produce a serious cardiomyopathy. There is also the potential of life-threatening arrhythmias and regional ische mia. Finally, sympathomimetic support usually
causes increased heart rate, which is a distinct dis advantage in the patient with coronary heart disease. Amrinone is a noncatecholamine, nonglycoside alternative in the treatment of perioperative myo cardial dysfunction. A type 3 phosphodiesterase enzyme inhibitor, amrinone combines positive ino tropic support with systemic and pulmonary vaso dilation. In patients with heart failure, the absence of drug-induced tachycardia and arrhythmias is of particular clinical benefit. In these patients, amri none's ability to augment cardiac performance without increasing heart rate or myocardial oxygen consumption offers significant clinical advantages during the perioperative period.75 B. Pulmonary Artery Hypertension and Right Ventricular Failure Under normal conditions, right ventricular per formance is governed, in order of decreasing im-
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
portance, by end-diastolic volume (preload), contractility, pulmonary vascular impedance (afterload), and coronary perfusion. However, in pathologic conditions, such as pulmonary hypertension, the order of right ventricular performance determinants is reversed, and the most important factor governing right ventricular function is status of the pulmonary vascular impedance (i.e., afterload).76 Thus, the contractility of the right ventricle is more dependent on the maintenance of a normal afterload than is the function of the left ventricle. The thin-walled right ventricle will not compensate for changes in afterload, particularly if these changes are acute. In situations in which pulmonary vascular impedance is acutely increased, right ventricular performance is dramatically affected. In situations in which this increase in afterload
477
occurs gradually (i.e., COPD), hypertrophy develops as a compensatory mechanism, and the right ventricular output remains normal at rest. However, in situations in which pulmonary impedance is increased acutely, these compensatory mechanisms are not operative/adequate and right ventricular failure ensues. In clinical situations that result in an acutely increased pulmonary vascular resistance (e.g., sepsis, adult respiratory distress syndrome, hypoxia, and pulmonary emboli), prompt and effective reductions in right ventricular afterload will result in maintenance of normal right ventricular function. A variety of intravenous agents are available as pulmonary vasodilators (Table 13-17), but the results have been mixed and a decrease in pulmonary and systemic pressure is to be expected (i.e..
478
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Table 13-17 VASODILATOR DRUGS CLINICALLY USED FOR PULMONARY HYPERTENSION* Drug Nitroglycerin Sodium nitroprusside Hydralazine Isoproterenol Nifedipine Prostaglandin E, (Alprostadil) Amrinone
Trade Name Nitrostat Nipride Apresoline Isuprel Procardia Prosti η VR
Predominant Vasodilator Effects
HR
CO
PAP
0.25-1.5 μg/kg/min 0.30-5.0 μg/kg/min 0.05-0.1 mg/kg bolus 0.02-0.08 μg/kg/min 5-10 mg po or si 0.03-0.10 μg/kg/miη
0/+1 0/+1 +1 +3 0/+1 0/+ 1
0 0 +1 + 2/ +3 0/+1 + 1/ + 2
-2 -1/-2 0/-1 0/-1 -1 -3
Systemic > pulmonary
5-10 mg/kg/5 min bolus 5-10 μg/kg/min
+2
-2
Systemic = pulmonary
Dose Range
Systemic Systemic Systemic Systemic Systemic
> pulmonary > > > pulmonary = pulmonary >> pulmonary = pulmonary
*From Prielipp R: Right-sided heart failure and pulmonary hypertension. In Cheng EY, Kay J (eds): Manual of Anesthesia and the Medically Compromised Patient. Philadelphia, JB Lippincott Co, 1990, pp 112-124. Used with permission. Abbreviations: HR = heart rate; CO = cardiac output; PAP = pulmonary artery pressure; po = orally; si = sublingual.
there is no available selective pulmonary vasodi lator except for inhaled nitric oxide; at the time of this writing nitric oxide is an investigational drug). Occasionally, the decrease in afterload will cause an increase in cardiac output, which will maintain the increase in pulmonary artery pressure or even cause a further increase. However, the use of am rinone in this context is most promising.75
C. Coronary Artery Disease
1. Monitoring for Ischemia (See Chapter 7): Further Considerations Recently, it has been demonstrated that, in pa tients in whom intraoperative ischemia develops, the risk of postoperative infarction is increased two- to threefold. These data imply that intraoper ative ischemia should be aggressively monitored and treated. Significant advances in echocardiog raphy have enabled this technique to be used in patients with ischemic heart disease. With ische mia, wall-motion abnormalities and systolic thin ning develop, followed by decreased left ventricu lar compliance (increased Ppao; for extensive discussion of this change, see chapter 7) followed by electrocardiogram changes and, in the awake patient, angina. Studies in animals have demonstrated that, very soon after coronary occlusion (in seconds), the ischemic myocardium first develops wall-motion and wall-thickening abnormalities. Both systolic shortening and thickening of the myocardial fibers become impaired. These regional effects produce abnormal contractions of the wall known as dyssynergia, which can pass through the phases of hypokinesis, akinesis and, finally, dyskinesis or aneurysm-type bulging. The effects of ischemia on wall thickening are important as well. In fact, wall thickening may be a more specific index of ischemia than wall mo
tion. In the area of acute myocardial infarction, segmental lengths of the involved tissue are in creased; paradoxical bulging is noted, and systolic thinning of the wall occurs. For those patients at significant risk for intra operative ischemia and older patients undergoing major and prolonged surgery, five electrode elec trocardiographic systems are in common use. The lead systems consist of either a bipolar arrange ment, measuring the potential between two elec trodes, or a unipolar electrode, measuring the po tential between an electrode and a combination of the remaining electrodes. The standard limb leads (I, II, IN) are bipolar, whereas the precordial leads (V,-V6) and the augmented limb leads (aVR, aVL, aVF) are unipolar. For ischemia detection, in an awake person undergoing exercise stress testing, lead V5 is associated with the largest number of ST-segment abnormalities. Furthermore, the diag nostic mode, which filters frequencies below 0.14 Hz, is more useful than the monitor mode (high pass at 4 Hz) for the detection of these ST-segment changes. The criteria for ischemic ST changes are a horizontal or down-sloping depression of 1 mm or more occurring 80 msec beyond the J point. This depression can be either new or additive and is associated with a greater incidence of compli cations. The depth of depression of the ST seg ment has been related to subsequent coronary events. Certainly, a 2-mm or more depression car ries the most significant prognosis. Recently, the comparative height of the R wave, used as a nor malizing factor, has been shown to give greater validity to the ST-segment change.
2. Coronary Artery Disease and Good Ventricular Function The first three rows of Table 13-15 represent a progression of events, beginning with inadequate anesthesia and ending with cardiac failure secon-
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
dary to myocardial ischemia, that might occur in a patient with coronary artery disease and good ventricular function. In this type of patient, angina pectoris is a primary symptom, hypertension is frequently present, the cardiac index is normal, the ejection fraction is greater than 0.5, left ventricular end-diastolic pressure is less than 12 mm Hg, there are no ventricular wall-motion abnormalities, and the cardiovascular system is capable of a hyperdynamic response to sympathetic stimulation. Also, inadequate anesthesia causes all of the determinants of myocardial oxygen consumption to increase; that is, heart rate, preload and afterload, and contractility (row 1, Table 13-15). These early changes may be aborted by providing adequate anesthesia early on, perhaps along with nitroglycerin vasodilatation and beta blockade. If the left ventricle experiences excessive afterload, it may become ischemic and begin to fail, as demonstrated by a decreased cardiac output despite an increased pulmonary artery wedge pressure (row 2 in Table 13-15). These intermediate changes may be reversed by dilating the arteriolar side of the circulation with nitroprusside. If the stress on the left ventricle remains unrelieved, the left ventricle then may become further ischemic and fail even more as evidenced by a further decrease in cardiac output, despite a further increase in pulmonary artery wedge pressure, and because the cardiac output is now so low there is also the beginning of systemic hypotension (row 3, Table 13-15). Under these late and severe circumstances, the heart must be aided with an inotropic drug. It should be remembered that whenever vasodilators are used, but particularly with the decreased preload effect of nitroglycerin infusion, concomitant volume infusion may often be required to maintain systemic and cardiac filling pressures at normal levels. This scenario emphasizes the importance of adequate anesthesia in patients with coronary artery disease and hyperdynamic ventricles.
3. Coronary Artery Disease and Poor Ventricular Function The patient with coronary artery disease and poor ventricular function may have a history of or electrocardiographic evidence of previous myocardial infarction and symptoms of congestive heart failure. There is little or no cardiac reserve, the cardiac index is less than 2 L/min/m2, the ejection fraction is less than 0.4, left ventricular end-diastolic pressure is often greater than 18 mm Hg, there may be hypokinetic or dyskinetic segments of the ventricular wall, and the heart does not tolerate loss of sympathetic input. In this patient, deep anesthesia and beta-blocking drugs should be
479
avoided and cardiac function maintained by using narcotic anesthesia, nitroglycerin preload and afterload reduction, inotropic drugs if cardiac failure supervenes, and maintenance of preinduction filling pressures. It should be remembered that, in order to maintain preinduction filling pressures in patients with poorly functioning ventricles, nitroglycerin increases the oxygen supply/demand ratio, in part, by decreasing preload (diastolic wall tension). Therefore, nitroglycerin infusion often needs to be accompanied by volume infusion to restore simultaneously toward normal both cardiac output and preload.
D. Valvular Heart Disease Patients with valvular heart disease are also a great challenge to successfully and safely anesthetize. For example, aortic stenosis is associated with a 14-fold increase in perioperative mortality.77 Table 13-18 shows the important hemodynamic goals to be achieved during anesthesia for patients with valvular heart disease.
E. Arrhythmias Thoracic surgery patients are exposed to a great deal of manipulation of the heart and mediastinum; these manipulations are associated with a high incidence of supraventricular tachyarrhythmias and ventricular ectopy.78 Neck dissections may stimulate the carotid sinus reflex, resulting in bradycardia and hypotension. Usually, informing the surgeon of the occurrence of these arrhythmias and terminating the mechanical stimulus causes the cardiac rhythm to revert back to normal. However, on occasion the arrhythmias may persist, and it may be necessary to use one or more of the treatment modalities listed in Table 13-19. Of the halogenated anesthetics, halothane is most arrhythmogenic and isoflurane the least arrhythmogenic.
V. RE-EXPANSION OF COLLAPSED LUNG DURING AND AT THE END OF THORACIC SURGERY (SEE CHAPTER 11) Pulmonary edema is well known to occur after re-expansion of lung following evacuation of a pneumothorax and a pleural effusion.79 Less well known and appreciated is that pulmonary edema may follow re-expansion of atelectatic lung after a right main-stem bronchial intubation (with the chest closed) (left lung becomes edematous) as well as re-expansion of atelectatic lung during
480
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Table 13-18 HEMODYNAMIC GOALS DURING ANESTHESIA FOR PATIENTS WITH VALVULAR HEART DISEASE Valvular Lesion
Variable Heart rate Rhythm
Preload
AS
AR
IHSS
Neither increases nor Avoid increases decreases Maintain normal Maintain normal sinus rhythm (do sinus rhythm not lose atrial kick, cardiovert atrial fibrillation) Maintain preload Maintain preload
Maintain normal sinus rhythm
MR
Avoid increased heart Avoid decreased heart rate rate Not as important as Maintain normal in other valvular sinus rhythm lesions
Maintain preload Maintain preload Maintain preload especially if (however, vasodilator is used increased heart rate may lead to increased Ppao) Systemic vasodilation May need some Systemic decrease if may lead to vasodilation may lead to decreased pulmonary vascular decreased regurgitation resistance requires regurgitation a decrease (Table 13-17). If hypertensive, use fluid, inotrope. Avoid decreases Amrinone Maintain
Afterload
Maintain blood pressure
Contractility
Neither increases nor Avoid increases decreases Full heart, maintain Adequate anesthesia, Avoid decreased normal sinus avoid heart rate rhythm hypertension and increased contractility
Global, most important
Maintain blood pressure
Avoid decreases
MS
Avoid decreased Avoid increased pulmonary vascular heart rate, avoid resistance and increased decreased right pulmonary vascular ventricular function resistance (hypoxemia, (hypoxemia, hypercapnia, hypercapnia, acidosis), avoid acidosis) increased heart rate
Abbreviations: AS = aortic stenosis; IHSS = idiopathic hypertrophic subaortic stenosis; AR = aortic regurgitation; MS = mitral stenosis; MR = mitral regurgitation.
open thoracotomy (see chapter 11). All of these situations are relevant to the condition of patients undergoing thoracic surgery and the one-lung ventilation situation. The clinical implications of this literature on re-expansion pulmonary edema (see chapter 11) are twofold. First, the collapsed lung should be expanded slowly and over several nonstaggered but progressively increasing tidal volumes (minimize the development of negative interstitial pressures). The same considerations apply to demonstrating an airtight bronchial closure just before closure of the chest after a pneumonectomy by pressurizing the lung to 35 to 40 cm H20 pressure while the surgeon fills the pleural space with irrigating fluid. Second, the anesthesiologist should be aware that the pulmonary edema may become manifest anywhere between a few minutes to several hours post re-expansion. The treatment of re-expansion pufmonary edema is mechanical ventilation with PEEP, diuretics, and other general support measures.
VI. TRANSPORT OF PATIENT80 Some thoracic surgery patients can be extubated while on the operating room table. These patients include those who had normal cardiopulmonary function preoperatively, underwent relatively short, physiologically nonintrusive surgery, and have an adequate postoperative vital capacity, peak inspiratory force, and spontaneous minute ventilation. These extubated patients should be sent to the postanesthesia recovery room for close observation for a period of time. However, in many thoracic surgery patients, it is obvious, based on either the preoperative condition of the patient or the nature of the surgical procedure, that a period of postoperative mechanical ventilation and intensive care will be required; it is most cost effective to send these intubated patients directly to the intensive care unit (ICU). In these patients, preparation for transport to the ICU should be begun toward the end of the procedure.
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery-
481
Table 13-19 ARRHYTHMIA TREATMENT SUMMARY Rhythm
First Treatment Choice
Second Treatment Choice
Ventricular fibrillation Ventricular tachycardia
Electrical defibrillation: 200 W/sec (3 W/sec/kg) with increased energy as required
Lidocaine (Xylocaine) 1.0-1.5 mg/kg initial IV bolus, infusion 1 to 4 mg/min (2 g/500 ml D5W)
Premature ventricular contractions
Lidocaine (Xylocaine) l .0-1.5 mg/kg initial bolus followed by infusion 1^4 mg/min Verapamil 5.0-10.0 mg IV
Propranolol (Inderal) 0.5 mg q 2 min up to 5 mg
Atrial fibrillation Atrial flutter
Paroxysmal supraventricular Verapamil 5-10 mg IV (atrial or junctional) Adenosine 0.1-0.2 mg/kg IV tachycardia Sinus tachycardia
Sinus bradycardia
Digoxin (Lanoxin) 0.25-0.5 mg IV Ouabain 0.1-0.2 mg IV Propranolol (Inderal) 0.5 mg q 2 min up to 5 mg Propranolol (Inderal) 0.5 mg q 2 min up to 5 mg Digoxin (Lanoxin) 0.25-0.5 mg IV Ouabain 0.1-0.2 mg IV Propranolol (Inderal) 0.5 mg q 2 min up to 5 mg Esmolol 1.0 mg/kg bolus, repeat q 10 min*
Eliminate underlying cause (e.g.. fever, pain, decreased blood volume) Usually no specific treatment required. If ţ HR—^ischemia, go to second treatment Atropine 0.3-2.0 mg Eliminate underlying cause (e.g., hypoxemia, fever, pain) Usually no specific treatment required. If hypotension or ventricular escape beats present, go to second treatment
*Esmolol is the beta blocker of choice in patients with bronchospasm. Abbreviations: IV = intravenously; D5W = 5 per cent dextrose in water; q
A. Preparation of Patient for Transport Once stability of the cardiovascular, respiratory, and hemostatic systems has been achieved, and the surgical wounds are being closed, preparations for transport to the ICU should be under way. The pretransport preparation check list is shown in Ta ble 13-20. Advance notice to the postoperative ICU should be given 30 to 45 min before leaving the operating suite. A suggested advance informa tion sheet is provided in Table 13-21.
1. Airway If the patient is going to be intubated and/or mechanically ventilated postoperatively, the dou ble-lumen tube must be changed to a single-lumen tube (except in rare situations in which differential lung ventilation is going to be used postoperatively [see chapter 19]). If the Univent tube is used, the bronchial blocker should be ^rendered nonfunc tional by fully retracting the bronchial blocker, taping it securely in the fully retracted position.
Third Treatment Choice Propranolol (Inderal) 0.5 mg q 2 min up to 5 mg Fourth choice: Refractory ventricular fibrillation. bretylium 5 mg/kg qs 50-100 ml DSW up to 30 mg/kg total dose Procainamide (Pronestyl) 100 mg q 5 min up to 1000 mg; may infuse 1-4 mg/min Cardioversion (first choice if patient is unstable)— external—synchronized flutter: 50 W/sec; fibrillation: 200 W/sec; internal: 5-30 W/sec— synchronized Vagal maneuvers (edrophonium. carotid sinus massage, NeoSynephrine)
Ephedrine 5-10 mg Fourth choice: isoproterenol 1 4 Fifth choice: atrial pacing if heart exposed, transvenous otherwise
every; HR = heart rate.
and cutting the pilot tube to the bronchial blocker balloon so that the bronchial blocker cuff cannot be mistaken for the cuff of the main lumen and inflated. As the end of the operation approaches, the patient should be given narcotics and left par alyzed. After the patient is turned into the supine position, both lumens of the double-lumen tube. the oropharynx and nasogastric tube should be suctioned, and the patient is then ventilated with 100 per cent oxygen. The double-lumen tube should be visualized entering the larynx by direct laryngoscopy, and, while continuing to directly visualize the larynx, the double-lumen tube should be removed and the single-lumen tube inserted. Following single-lumen tube insertion, confirma tion of proper single-lumen tube placement, and securement of the single-lumen tube with tape, the patient should be mechanically ventilated for sev eral minutes while other measures are being taken (see the following). It should be realized that an esophageal intubation following right pneumonec tomy or transposition of the stomach into the chest may be particularly difficult to diagnose (left-sided
482
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Table 13-20
PRETRANSPORT PREPARATION CHECK LIST
A. Respiratory system 1. Double-lumen tube changed to single-lumen tube 2. Arterial blood gases on F,0 2 = 1.0 3. Manual transport positive-pressure ventilation system ready 4. Stethoscope in place 5. Anesthesia mask present 6. Laryngoscope present B. Chest tube and drainage system 1. Collection compartment below level of patient 2. Blood and air leak not excessive 3. Mediastinum midline C. Circulatory system 1. Hemodynamically stable 2. EKG on oscilloscope 3. Mean arterial pressure on manometer or phasic waveform on oscilloscope 4. Vascular catheters untangled, labeled; injection ports identified; heparin lock and flush on noninfusing lines; adequate fluid for infusing lines; drug infusions labeled and in pump meters 5. Coagulation status satisfactory D. Anesthesia requirements 1. Narcotized 2. Paralyzed E. Miscellaneous 1. ICU bed present 2. Sufficient personnel available 3. ICU notified 4. Elevator called for 5. Records collected 6. All patient lines that have an operating room connection (intravascular, Foley, nasogastric, chest catheters, other monitoring lines, ventilation system) should be transferred to a transport system connection. Abbreviations: EKG = electrocardiogram; ICU = intensive care unit.
or midline breath sounds will be heard, respectively, and left or midline chest movements seen, respectively, regardless of whether the tube is in the trachea or esophagus); capnography should be used to make the differential diagnosis whenever the single-lumen tube cannot be seen to enter the trachea directly or there is any question about gas exchange. If the laryngeal aperture and entrance of the double-lumen tube into the larynx cannot be visualized with direct laryngoscopy, then this method of changing a double-lumen tube to a single-lumen tube is hazardous and has a high risk of failure. Inability to visualize the laryngeal aperture is particularly likely to occur in cases in which massive amounts of crystalloid fluids and blood have been infused and the supraglottic area has become edematous. There are two ways of changing a double-lumen tube to a single-lumen tube that utilize a stylet method and still retain the ability to ventilate the patient. The first method allows ventilation during all steps of the tube change, and the second method allows ventilation most of the time.
The first method involves the use of a jet stylet.81 A jet stylet is a small internal diameter (ID), hollow, semirigid catheter that is inserted into an in situ endotracheal tube (in this case, the tracheal lumen of the double-lumen tube) before extubation. The proper depth of insertion is to 38 to 39 cm82 (Fig. 13-11) to ensure that the tip of the jet stylet remains supracarinal; it is important that the tip of the jet stylet (tube changer) is supracarinal before administering any jet ventilation to avoid barotrauma to the one lung (the one that is ventilated if the jet stylet is subcarinal). When changing a double-lumen tube to a singlelumen tube, a jet stylet catheter of approximately 100 cm should have an appropriate reserve of length if the tracheal lumen is uncut82 and an 80cm catheter is long enough if the tracheal lumen is cut just proximal to the common molding that binds the two lumens together. After the double-lumen tube is withdrawn over the jet stylet, the small-ID hollow catheter may then be used as a means of ventilation (i.e., the jet function) and/or as an intratracheal guide for reintubation (i.e., the stylet function) (Fig. 13-12).81 In some instances, the jet function may safely allow additional time to assess the need for the reintubation stylet function. A small Sheridan tube exchanger (Sheridan Catheter Corp, Argyle, NY) fits down a 35 French double-lumen tube, and a medium size tube exchanger will fit down a 41 French double-lumen tube. Both are satisfactory to be used as a jet stylet in this context (the tracheal lumen should be cut at the common molding). The tube exchanger may be connected to the jet injector by inserting an appropriately sized intravenous catheter into the proximal end of the stylet (see Fig. 13-12) or by inserting a female Luer lock barbed cone adapter into the proximal end of a Table 13-21 ADVANCE INFORMATION SHEET FOR ICU A. Respiratory sytem 1. Endotracheal tube: route, size 2. Ventilator settings: F,0 2 , tidal volume and/or peak inspiratory pressure, rate, PEEP 3. T-piece: F,0 2 4. Extubated: mask F,0 2 5. Request immediate chest X-ray: yes, no B. Monitoring system 1. Arterial line: location 2. Central venous pressure: location 3. Pulmonary artery catheter: location C. Intravenous catheters 1. Site 1: location, fluid, drug 2. Site 2: location, fluid, drug D. Laboratory studies pending E. Estimated time of arrival Abbreviation: ICU = intensive care unit; PEEP = positive end-expiratory pressure.
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Figure 13-11 When a tracheal tube exchanger is inserted to a depth of 38 to 39 cm down an optimally placed double-lumen endobronchial tube, its tip will lie proximal to the carina. (From Hannallah M: Evaluation of tracheal tube exchanges for replacement of double-lumen endobronchial tubes. Anesthesiology 77:609-610, 1992. Used with permission.)
short length of appropriately sized ETT and the distal end of the ETT over the tube exchanger.83 The intratracheal location of the jet stylet may be preserved while concurrently confirming intratracheal placement of the reintubation tracheal tube (see Fig. 13-12).81·83·84 The intratracheal location of the jet stylet may be preserved while concurrently confirming intratracheal placement of the reintubation tracheal tube by the following procedure. A standard anesthesia circle system is modified by" replacing the existing elbow connector with a connector incorporating a self-sealing diaphragm such as the type
483
used for bronchoscopy. In addition, the connection of the proximal end of the lumen of an appropriately sized tracheal tube exchanger to a jet ventilator is made immediately available as described previously. When the new tracheal tube is passed over the tracheal tube changer, the new tracheal tube is connected to the anesthesia circuit via the bronchoscopy elbow by passing the proximal end of the tracheal tube changer retrograde through the self-sealing diaphragm of the bronchoscopy elbow connector (see Fig. 13-12). After conclusively establishing the intratracheal position and patency of the new tracheal tube by capnography and auscultation of bilateral breath sounds, the tracheal tube exchanger is removed by pulling it through the self-sealing diaphragm of the bronchoscopy elbow connector. The second method involves using a fiberoptic bronchoscope (FOB) as a stylet and cutting away the double-lumen tube (Fig. 13-13).85 After ensuring that the patient is completely paralyzed, both the tracheal and bronchial cuffs of the doublelumen tube are deflated, and the double-lumen tube is withdrawn until the bronchial lumen is above the level of the carina. The tracheal lumen adapter is cross clamped, and bilateral ventilation is maintained via the bronchial lumen. With the tracheal lumen now isolated, a no. 10 scalpel is used to create a 5 X 5-mm opening in the lateral wall of the tracheal lumen distal to the reinforced area connecting the bronchial and tracheal lumina. A lubricated (inside and outside) 25-cm long. 7mm outside diameter single-lumen tube is advanced over the FOB until the 15-mm adapter is abutting the handle of the FOB. The distal end of the single-lumen tube-FOB combination is then inserted into the tracheal lumen of the doublelumen tube through the newly created opening and is advanced until the tracheal mucosa can be visualized. The new opening in the double-lumen tube is extended distally with a pair of scissors. The double-lumen tube is withdrawn while the tracheal lumen is cut longitudinally (by a large pair of sharp scissors or sharp scalpel) until the opening of the tracheal lumen is out of the mouth. At this time, ventilation is maintained via the bronchial lumen of the double-lumen tube, which is resting in the oropharynx (Fig. 13-13). Once the FOB is freed from the tracheal lumen of the double-lumen tube, the single-lumen tube is then passed over the FOB and positioned in the trachea. Ventilation is then instituted via the single-lumen tube.
2. Respiration After 5 to 10 min of ventilation with 100 per cent oxygen via the single-lumen tube, arterial
484
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Figure 13-12 A method to preserve the intratracheal location of the tracheal tube exchanger during confirmation of intratracheal placement of a tracheal tube. The tracheal tube exchanger is passed through a self-sealing diaphragm in a fiberoptic elbow adapter. With this method, positive-pressure ventilation and carbon dioxide sampling may be around the tracheal tube exchanger but within the tracheal tube. (ETT = endotracheal tube.) (From Benumof JL: Management of the difficult airway: With special emphasis on awake tracheal intubation. Anesthesiology 75:1087-1110, 1991. Used with permission.)
blood gases should be obtained and the values known before departure. A simple foolproof man ual system for providing controlled positive-pres sure ventilation with an inspired oxygen concen tration of 100 per cent during transport should be standing by. The transport ventilating system should allow PEEP to be provided to the patient. Manual ventilation during intrahospital transport of critically ill, mechanically ventilated patients is safe provided the person performing the manual ventilation knows the inspired oxygen fraction and minute ventilation that is required before the trans port takes place and is trained to approximate the minute ventilation during transport.86 A sufficient supply of oxygen should be avail able, remembering that a completely full Ε tank of oxygen contains enough gas to supply 10 to 12 L/min for approximately 50 min. Monitoring of ventilation should be provided both by observation of chest movements and by hearing breath sounds via an esophageal stethoscope. The anesthesiolo gist should take along an appropriately sized an esthesia mask (to provide positive-pressure venti lation in case of inadvertent extubation), a la
ryngoscope (to reintubate), and some essential drugs for cardiovascular resuscitation.
3. Chest Tube and Drainage System87 The chest tube and drainage system must be functioning properly before transport. Chest X-ray confirmation of proper chest tube position is made by identifying the radiopaque stripe that outlines the proximal drainage hole (there are multiple holes) and by noting the absence of a visceral 88 pleural line. Two tubes are usually required to drain both air and fluid from the pleural space (an empty hemithorax after a pneumonectomy requires only one tube; see later discussion); one tube is placed at the anterior apex to favor the escape of air when the subject is sitting upright, and one tube is placed at the posterior base to drain accumulat ing fluid when the subject is supine. They are usually inserted through separate incisions just be fore the chest is closed. As the last stitch closing the pleural cavity is being tied, the lungs should be slowly inflated and then held fully inflated to
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
485
\ Bronchial lumen Tracheal lumen
w
- Double-lumen tube
New opening (for bronchoscope)
Proximal clamp
Longitudinal cut
Single-lumen tube over flexible fiberoptic bronchoscope
Figure 13-13 Diagrammatic representation of fiberoptic bronchoscope passed through tracheal lumen of the endobronchial tube and acting as guide for single lumen tube. The bronchial lumen of the endobronchial tube remains in close proximity to the glotticinlet, permitting ventilation of the lungs while the tube's tracheal lumen is freed from the bronchoscope. (From Gatell JA. Barst SM, Desiderio DP, et al: A new technique for replacing an endobronchial double-lumen tube with an endotracheal single-lumen tube. Anesthesiology 73:340-341, 1990. Used with permission.)
reverse any remaining atelectasis and to help evacuate the pleural space of air and fluid. Modern chest tube drainage systems have three compartments (Fig. 13-14). The first compartment is a simple graduated collection chamber where the amount of draining blood and fluid can be measured accurately (see Fig. 13-14, right panel). The second compartment is an underwater seal that serves as a simple, but very reliable, one-way valve, which allows air to escape from the pleural space to the drainage system but not to enter the pleural space from the drainage system during the next inspiration. If the inlet drainage limb to the underwater seal compartment is put just a short distance ( 1 to 2 cm) below the level of the water, there is minimal resistance to the escape of air, but a huge inspiratory effort would be needed to break the seal by drawing water from the bottle up the drain and into the chest (see Fig. 13-14, middle panel). All drainage systems, whether simple or complex, require a water seal (i.e., the water seal compartment is all that is necessary for the treatment of a spontaneous pneumothorax). The third compartment controls the amount of negative pres-
sure that the suction can generate. The inlet drainage limb to the suction control compartment is exposed to the outlet suction above a level of water. The level of the water in the outlet limb of the suction control compartment is vented to atmosphere. If the negative pressure above the water level exceeds the height of the water level, air will come in from the suction control atmospheric vent, bubble through the water, and neutralize the excessive negative pressure above the water level (see Fig. 13-14, left panel). Two factors determine the level of suction. First, the central movable vent tube, which can be raised or lowered to adjust the underwater depth, determines the negative pressure that the system will generate. Second, the actual suction pump selected may be a limiting factor. Low-pressure systems capable of generating between 15 and 20 cm H20 negative pressure and between 15 and 10 L/min of flow include the Stedman, Gomco, and Thermovac pumps. Emerson and Sorenson systems are high-pressure systems capable of pressures of - 6 0 cm H20 with flows of >20 L/min. When selecting a suction system, care should be
486
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
Figure 13-14 This schematic diagram of a modem chest tube drainage system shows that the chest tube drainage system is made up of three compartments. The first compartment is a collection chamber for fluids (blood, pus) from within the chest. The second compartment contains the origin of the chest tube suction and a water seal valve that prevents gas or fluid from being drawn, by a forceful spontaneous inspiration, into the chest. The third compartment controls, via an atmospheric vent, the degree of negative pressure that can be developed by chest tube drainage system suction. See the text for a further explanation.
exercised to provide a negative pressure greater than the possible positive pleural pressure seen on expiration. If these conditions are not met, a tension situation could be created despite continuous suction.88 In other words, if the flow rate of the suction system cannot exceed the rate of leakage of air into the pleural space, a tension pneumothorax will occur. If functioning lung tissue remains on the operated side (resection less than a pneumonectomy, all nonpulmonary surgery), the drains should be joined to the inlet of an underwater seal as soon as the pleural cavity has been closed to prevent the lung from collapsing again (owing to accumulation of air in pleural space) while the remaining layers of the chest wall are sutured. If separate bottles are used for the drainage system (see Fig.
13-14, upper panel), the collection compartment bottle should be below the level of the patient at all times to prevent the entry of fluid from the collection compartment and chest tube per se back into the chest. Modern disposable plastic systems have hooks on them that can be easily attached to the lower part of the transport bed. Simple clamping of the drains is dangerous owing to the possible development of tension pneumothorax. Air is expelled through the drain during expiration if the patient is breathing spontaneously, and the volume escaping increases during coughing. The drain will bubble continuously if there is a large leak, particularly if suction is used, and, if both blood and air are escaping simultaneously, a prohibitive amount of froth may appear. The froth can be controlled by adding an antifoaming agent
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
(such as a few drops of alcohol). If there is no air leak, the fluid level will move gently in the drainage tubing during quiet, spontaneous respiration but will cease to swing when the lung is fully expanded and opposed to the chest wall or if the drain is blocked. Frequent (every half hour) mechanical stripping of the chest tubes (a distal movement to create vacuum within the tubing to suck fluid and/or air from the chest) or milking of the chest tubes if blood is being drained (pushing air and fluid and clots back into the chest) is important because, once clots have begun to collect within the chest and chest tubes, chest tube function deteriorates, and the likelihood of the need for re-exploration for hemopneumothorax increases. In individuals who have significant drainage and/or bleeding, these maneuvers may be required every 5 min. Nevertheless, the value of these procedures in postoperative coronary artery by-pass graphing patients has been questioned.89 When the functional status of a tube is questioned, several simple maneuvers can be used to assess its integrity. Observation of synchronous water seal and respiratory motion suggest the tube is still functioning in the pleural space and all connections are tight. If an air leak is suspected at a particular point in the system, be it bottle, tube, or connector, sequential clamping with distal suction before and after the spot in question should be performed. Bubbling air through the water seal when the drainage system is clamped just distal to the point in question, which disappears when the clamp is placed just proximal to that point, identifies the site of leakage; the identified component then can be changed. If the tube is not functioning and occlusion of the drainage holes is suspected, the tube can be disconnected and flushed with saline solution in an effort to dislodge obstructing debris. The evacuation of clots and debris from the chest tube with the use of a sterile suction catheter introduced into the chest tube through a sterile cap has been proposed.90 After a pneumonectomy but before transport, the position (laterality) of the mediastinum should be ascertained.91 This can be done in two ways: clinical examination and measurement of pressure in the hemithorax. Clinical data consist of determining the position of the trachea and inspecting the anteroposterior chest roentgenogram. Perhaps more practical and accurate is the measurement of pressure in the empty hemithorax by manometer. If the pressure in the empty hemithorax is significantly positive, then it is likely that the mediastinum is shifted into the contralateral hemithorax. If the pressure in the empty hemithorax is significantly negative, then it is likely that the mediastinum is shifted into the ipsilateral hemithorax. To
487
get the mediastinum in the midline, air can be injected into or aspirated from the empty hemithorax, as dictated by the ongoing pressure manometer readings. Closing the chest without drainage after pneumonectomy makes it more difficult to recognize serious postoperative hemorrhage. Consequently, a single basal chest drain is usually used and is joined to an underwater seal bottle without suction (suction will cause a mediastinal shift to the ipsilateral side and, perhaps, cardiovascular collapse). It is left undamped until the patient is lying on the back and breathing spontaneously. Some air usually escapes through the drain when the patient is turned from the lateral to the supine position, or if coughing occurs, when spontaneous ventilation is being re-established. This air should be allowed to escape so that pressure does not build up within the space and cause surgical emphysema around the drain or the wound site. The mediastinum should be central at the end of the operation if this routine is followed, and once this has been confirmed, the tube is clamped and remains so while the patient is returned for postoperative supervision. Thereafter, it is undamped for 1 to 2 min every hour for the first 12 to 24 hours so that the excess fluid may drain, thereby preventing unwanted mediastinal shift.
4. Circulation For the patient to be a candidate for leaving the operating room, the cardiovascular system must be stable. Vascular volume status should be optimal for that particular patient, as judged by the urine output, systemic pressure, left- or right-sided filling pressures or both, and cardiac output, if available. All vascular catheters should be identified by label and should be free and untangled with at least one injection port conveniently placed. Adequate fluid for all vascular catheters should be available. Drugs being administered by continuous infusion should be reviewed, and the administration rates and cardiovascular responses confirmed to be appropriate. The cardiac rate, rhythm, and mean arterial pressure should be monitored during transport. Essential cardiovascular resuscitation drugs should be taken along.
5. Coagulation If massive transfusion was required intraoperatively, the coagulation status should be normal or as nearly normal as possible at the time of transport. All products to treat coagulation abnormalities should have been administered if possible. Blood specimens should be sent to the laboratory from the operating room for the determination of
488
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
coagulation status (ACT, aPTT, PT, fibrinogen), platelet count, hematocrit, serum potassium, and arterial blood gases. The results of these determinations will therefore be available to the ICU personnel within moments of the patient's arrival there. 6. Miscellaneous A fully functional ICU bed should be available for direct patient transfer. All records should be collected in order for them to be transported with the patient. Sufficient personnel to effect a smooth transport should be gathered. Prior to moving the patient, all possible connections of the patient to the operating room must be disengaged, such as pressure lines to transducers, electrocardiogram leads to the room monitor, chest tube and Foley catheters to the operating room table, and intravenous lines to operating room poles, and must be transferred to their appropriate places on the transport bed. If an elevator is to be used, it should be called for so that it is there waiting for the transport team. In the rare case of transport by air, the pleural cavity must be decompressed by a chest tube with a guaranteed one-way flow of gas (from the pleural cavity to the environment) (by a one-way valve or underwater seal mechanism) because at an altitude of 5000 to 6000 feet airspaces will expand by 20 to 35 per cent (Boyle's Law). Similarly, all other potential airspaces must be decompressed (e.g., stomach, bowel).92 B. Transport Preferably, the patient should be moved only once, and that is directly from the operating room table to the ICU bed. The move should be performed smoothly and gently to minimize vascular volume shifts or changes in cardiac function. A carefully coordinated move by a sufficient number of personnel should prevent accidents such as disconnected intravascular lines and various tubes being pulled out or malpositioned. The anesthesiologist should be responsible only for ventilation and monitoring, not moving, fetching, or adjusting parts of the bed. Monitoring during transport should include at least an esophageal stethoscope for respiratory and cardiovascular sounds and an oscillographic display of the electrocardiogram. In addition, if an arterial line is in situ, mean arterial pressure should be monitored, at a minimum, with a sterile anaeroid gauge directly connected to the arterial line by a 50-cm segment of sterile intravenous tubing. Still better, and presently practiced more com-
monly, is an oscillographic display of the electrocardiogram and arterial pressure waveform on a portable monitor for continuous observation during transport. A defibrillator should also accompany the very critically ill patient. The longer and more complex the route to the ICU, the more care must be exercised in planning and execution of this transport. If an elevator separates the operating room from the ICU, it should be equipped with a separate reserve oxygen system, electrical power outlet, phone communication system, and if possible, suction equipment (physicians responsible for transport should ask themselves what they would want available in an elevator that is jammed for 2 hours). C. Arrival In Intensive Care Unit On arrival in the ICU, the pretransport operating room take-down should be reversed and a quick ICU arrival hookup performed. Priorities should be the same as in a cardiac resuscitation: establishment of the airway, adequate gas exchange, and circulation. Ventilatory function is assessed as the patient's airway is connected to a ventilator, set initially with an F,02 of 100 per cent. Other initial ventilator settings such as tidal volume, respiratory rate, and PEEP level should duplicate those found optimal in the operating room (see chapter 20). The chest must be seen and heard to move bilaterally. Vital signs of immediate importance are cardiac rate and rhythm (electrocardiogram) and mean arterial and cardiac filling pressures. Infusion rates of potent drugs are verified, and desired cardiovascular responses are confirmed. Tasks of lesser importance can be performed as time permits. Laboratory and other diagnostic procedures should be performed as soon as practical (chest film, arterial blood gases, electrolytes, hematocrit, 12-lead electrocardiogram, and coagulation screen if indicated). Before the anesthesiologist leaves the ICU, he or she should ascertain the present status of the patient, confirm cardiovascular and respiratory stability, be assured that no problems of an acute nature exist, and confirm that the ICU nurses are satisfied with the patient's status. A brief early postoperative note as well as a copy of the anesthetic record should be placed in the patient's chart.
REFERENCES 1. Fitzgerald JM, Hargreave FE: The assessment and management of acute life-threatening asthma. Chest 95:888894, 1989. 2. Gold MI, Helrich M: Pulmonary mechanics during general anesthesia: V. Status asthmaticus. Anesthesiology 32:422428, 1970.
CHAPTER
14
Anesthesia for Special Elective Diagnostic Procedures 1. Introduction II. Bronchoscopy A. Fiberoptic Bronchoscope 1. Indications 2. Ventilatory Considerations 3. Anesthetic Technique a. Local Anesthesia b. General Anesthesia 4. Complications B. Rigid Ventilating Bronchoscope 1. Indications 2. Ventilatory Considerations a. Sidearm Connection b. Venturi Connection
3. Anesthetic Technique a. Local Anesthesia b. General Anesthesia 4. Complications III. Mediastinoscopy A. Indications and Surgical Considerations B. Anesthetic Technique C. Complications IV. Thoracoscopy A. Indications and Surgical Considerations B. Anesthetic Technique C. Complications
492
Anesthesia for Special Elective Diagnostic Procedures
I. INTRODUCTION There are a number of very common invasive diagnostic procedures used to evaluate thoracic surgery patients. These diagnostic procedures all have special anesthetic implications and problems. Bronchoscopy requires that the airway be shared between the anesthesiologist and the endoscopist, which causes special problems for ventilation and anesthetic technique. Mediastinoscopy can cause compression of important vessels and hemorrhage, and the anesthesiologist's attention must be focused on recognizing these complications. In addition, these patients must not be allowed to strain or cough. Thoracoscopy often involves double-lumen tube intubation and one-lung ventilation. This chapter considers these three common major diagnostic techniques in the order just mentioned.
II. BRONCHOSCOPY Bronchoscopy is the most common invasive procedure used for the diagnosis and treatment of chest diseases. There are three general diagnostic indications for bronchoscopy.1 First, bronchoscopy is indicated to investigate the cause of medically resistant chronic chest symptoms and signs (cough, pain, wheeze, atelectasis, pneumonia). Second, bronchoscopy is indicated to determine the site, extent, and cause of acute changes such as hemoptysis, acute inhalation injury, and regional abnormality on the chest X-ray. Third, bronchoscopy is an essential tool in the evaluation of patients for thoracic surgery (to evaluate the tracheobronchial tree for tumors, to rule out metastases, to obtain tissue and specimens [cytology, biopsies, cultures], to determine presence and exact level of tracheobronchial esophageal fistula, to palpate the carinal and subcarinal areas). The therapeutic indications are to aid in the treatment of acute respiratory failure (remove secretions, reverse atelectasis, drain abscess, position endotracheal tubes), to aid in laser resection of bronchogenic carciTable 14-1
noma, and to allow for removal of foreign bodies. This chapter discusses anesthesia for diagnostic bronchoscopy. The anesthetic management of the various therapeutic bronchoscopies are discussed elsewhere (laser resection of tumors in chapter 15, removal of foreign bodies in chapter 17, treatment of acute respiratory failure and aid in mechanical ventilation in chapter 20). There are three types of bronchoscopes in current use, and they consist of the flexible fiberoptic, rigid ventilating, and rigid Venturi (Sanders injector) types. The manner in which patients are ventilated with these bronchoscopes differs considerably, and the different ventilatory considerations have important implications for anesthetic technique (Table 14-1). Flexible fiberoptic bronchoscopy may be performed with local or general anesthesia, whereas rigid ventilating and Venturi bronchoscopy is best performed under general anesthesia. Since the inspired oxygen and inhalational anesthetic concentration is known with a rigid ventilating bronchoscope but not known with a rigid Venturi bronchoscope, rigid ventilating bronchoscopy can be performed with either inhalational (including nitrous oxide) or intravenous general anesthesia, whereas rigid Venturi bronchoscopy is best performed with intravenous general anesthesia and 100 per cent oxygen. Since minute ventilation is constant with either fiberoptic or rigid Venturi bronchoscopy (no need to interrupt ventilation) but not with rigid ventilating bronchoscopy (need to interrupt ventilation when the eyepiece is removed), the former two can be comfortably used for long procedures, whereas the latter should be used for relatively short procedures. In view of these differences in ventilatory and anesthetic technique, the different bronchoscopes are discussed separately. A. Fiberoptic Bronchoscope 1.
Indications
Use of the flexible fiberoptic bronchoscope is especially indicated for bronchoscopy whenever
VENTILATION CHARACTERISTICS AND ANESTHETIC IMPLICATIONS OF THE THREE DIFFERENT TYPES OF BRONCHOSCOPES
Type of Bronchoscope
RO,
Concentration of Inhalation Anesthesia
Constancy of Minute Ventilation
Suitable Duration of Procedure
Flexible fiberoptic
Known
Known
Constant
Long
Rigid ventilating (sidearm connection) Rigid Venturi
Knjawn
Known
May be inconstant without packing
Long (if packing used)
Unknown
Unknown
Constant
Long
»
Preferred Type of Anesthesia Local or general anesthesia Inhalation or intravenous general anesthesia Intravenous general anesthesia
Anesthesia for Special Elective Diagnostic Procedures
493
Endotracheal Tube Size Without Fiberoptic Bronchoscope, mm ID Figure 14-1 This diagram relates the size of an endotracheal tube without an indwelling fiberoptic bronchoscope (x-axis) to the resultant functional size that the endotracheal tube would have with an indwelling fiberoptic bronchoscope (left-hand y-axis) (three fiberoptic bronchoscope sizes are shown: 5.0-, 5.7-, and 6.0-mm outside diameter). The cross-sectional area of unoccupied endotracheal tube in mm2 that is responsible for the resultant functional size is shown on the right-hand y-axis. (The diagram is a modification [with permission] of Figure 7 from Lindholm CE, Oilman B, Snyder JV, et al: Cardiorespiratory effect of flexible fiberoptic bronchoscopy in critically ill patients. Chest 74:362-368, 1978.)
there are mechanical problems with the neck that would make rigid bronchoscopy especially difficult; whenever upper lobe and peripheral lung lesions need to be visualized and are beyond the reach and capability of a rigid bronchoscope (the fiberoptic bronchoscope can examine the third to possibly the fourth order of bronchi); and whenever there is a need to find the site of limited and perhaps distal hemoptysis, to aid in the treatment of respiratory failure, to obtain very localized cultures, biopsies, and cytology, and to perform selective bronchography.1-5 The many airway management and intensive care unit uses of a fiberoptic bronchoscope are beyond the scope of this chapter (see chapters 9, 13, and 20).
2. Ventilatory Considerations In a nonintubated patient, flexible fiberoptic bronchoscopes with outside diameters of 5.0, 5.7, and 6.0 mm occupy 6, 10, and 11 per cent, respectively, of the cross-sectional area of an average 70kg adult trachea (which has a diameter of 18 mm). Consequently, spontaneous ventilation and removal of carbon dioxide is not usually significantly im-
paired in a nonintubated, nonbronchospastic, spontaneously ventilating patient undergoing fiberoptic bronchoscopy under local anesthesia. However, since the suction port is capable of removing large amounts of air (14.2 L/min at 760 mm Hg through a 2-mm suction port), which may cause atelectasis (especially if the bronchoscope becomes wedged during suctioning), oxygen supplementation is highly desirable.6·7 This can be accomplished by having the patient breathe through nasal cannulas or, more preferably, through a facial mask (which provides a higher F,02) that has a hole cut in it to allow for insertion of the fiberoptic bronchoscope. In an intubated patient (which would usually be the case if general anesthesia were used), the fiberoptic bronchoscope occupies a very significant amount of the cross-sectional area of the endotracheal tube. Figure 14-1 shows the reduction in effective size of any sized endotracheal tube with a 5.0-, 5.7-, and 6.0-mm outside diameter fiberoptic bronchoscopy.8 For example, a 5.7-mm outside diameter fiberoptic bronchoscope occupies 41 and 52 per cent of the cross-sectional areas of 9- and 8-mm internal diameter endotracheal tubes, respectively, which reduces the functional internal
494
Anesthesia for Special Elective Diagnostic Procedures
diameters of these endotracheal tubes to 7.0 and 5.5 mm, respectively.8 In view of the greatly reduced surface area available for ventilation under these circumstances, it is obvious that ventilation must either be controlled or vigorously assisted with positive-pressure ventilation, This may be easily accomplished by passing the fiberoptic bronchoscope through a self-sealing rubber diaphragm in the elbow connector to the endotracheal tube; tidal ventilation then occurs around the fiberoptic bronchoscope but within the endotracheal tube (Fig. 14-2). However, it must be realized that if the fiberoptic bronchoscope occupies too much of the cross-sectional area of the endotracheal tube, tidal gas may enter the lungs under positive pressure, but the elastic recoil of the lung may not force the tidal gas out of the lungs, and distal gas trapping, high distal positive end-expiratory pressure (PEEP), and barotrauma may result. Consequently, a low inspiratory-to-expiratory time ratio is desirable. In light of these considerations, it is apparent that an endotracheal tube with an internal
diameter of 8.0 to 8.5 mm or larger should be used with any adult-sized fiberoptic bronchoscope.8 If a smaller endotracheal tube must be used (e.g., 7.0mm internal diameter), the use of helium/oxygen mixtures should be considered (see chapter 3).9,10 In addition, periods of hyperventilation with 100 per cent oxygen without the fiberoptic bronchoscope in place may need to be alternated with short periods of fiberoptic bronchoscopy. Oxygen can always be insufflated down the suction port of the fiberoptic bronchoscope, and it is my practice to routinely do so." In anesthetized children and adults and in awake adults, the fiberoptic bronchoscope may also be passed through a laryngeal mask airway (Fig. 14-3).I2~14 Table 14-2 shows the relationship between the outside diameter of the fiberoptic bronchoscope and internal diameter of the largest endotracheal tube that fits into the shaft of the various sized laryngeal mask airways.12 It may be inferred from Table 14-2 that the internal diameter of the shaft of the laryngeal mask airway is 3 to 4
Fiberoptic Bronchoscope Endotracheal Tube Ventilating System
Figure 14-2 This schematic diagram shows that a fiberoptic bronchoscope is ordinarily inserted through a self-sealing diaphragm in the elbow connector to an endotracheal tube. Inspired gas under positive pressure from the anesthesia circle machine goes around the outside of the fiberoptic bronchoscope but within the lumen of the endotracheal tube. Exhaled gas is via the same route into the anesthesia circle system. The seal of the diaphragm in the elbow connector around the fiberoptic bronchoscope ensures the ability to continue positive-pressure ventilation while the fiberoptic bronchoscope is in use.
Anesthesia for Special Elective Diagnostic Procedures
495
eral anesthesia is induced, and the patient is venti lated (either spontaneously or controlled with pos itive pressure) via a mask. The fiberoptic bronchoscope can be introduced into the airway down the central slit in the diaphragm of a bron choscopy elbow adaptor (Fig. 14—4)15 or through an anesthesia mask with diaphragm (Fig. 14-5)." In either case, the airway can thus be examined with little interference with ventilation. These methods avoid creating a small breathing space between the fiberoptic bronchoscope and the con duit for the fiberoptic bronchoscope (either an en dotracheal tube or laryngeal mask airway), avoid interfering with the view of the vocal cords, and perhaps avoid further edema in an already nar rowed airway. In the rare and unusual situation of high tracheal and laryngeal obstructions in infants, the institu tion of transtracheal jet ventilation may be lifesaving16 (irrespective of whether a fiberoptic or rigid instrument is used for the definitive proce dure). 3.
Anesthetic Technique
a. LOCAL ANESTHESIA
Figure 14-3 Fiberscope within a 6.0-mm ID endotracheal tube (ETT) that is within the shaft of a no. 4 laryngeal mask airway. The fiberscope and ETT pass through the central com partment of the grille at the end shaft of the laryngeal mask airway, and the cuff of the endotracheal tube is inflated to increase the distinctiveness of the ETT. (From Benumof JL: Use of the laryngeal mask airway to facilitate fiberoptic-endoscopy intubation. Anesth Analg 74:313-314, 1992. Used with permission.)
mm greater than the internal diameter of the cor responding endotracheal tube (e.g., the internal di ameter of the no. 4 laryngeal mask airway is ap proximately 9-10 mm). The difference between the internal diameters of the laryngeal mask air way and fiberoptic bronchoscope determines the resistance to airflow, and the clinical implications of an increase of airflow resistance are the same as passing a relatively large fiberoptic bronchoscope through a relatively small endotracheal tube. It is apparent from the previous discussion that passage of a fiberoptic bronchoscope through an endotracheal tube in generally anesthetized pediatric patients is problematic with respect to ventilation. Fortunately, in generally anesthetized children (as well as adults), the fiberoptic bronchoscope may be passed into the trachea without going through an endotracheal tube in the following way.15 Gen
Patients with reactive airways may need preop erative bronchodilation (β,-agonists, aminophylline, steroids; see chapter 13 and Fig. 13-1). In addition, atropine premedication17· IH is useful to decrease the volume of secretions and to cause some bronchodilation. Diazepam and barbiturates are useful to decrease anxiety and systemic toxic ity of local anesthetics. The fiberoptic bronchoscope is relatively easily Table 14-2 LARGEST SIZE ENDOTRACHEAL TUBE AND FIBERSCOPE THAT CAN FIT CONCENTRICALLY INTO THE LARYNGEAL MASK AIRWAY*
LMA Size Number 4 3 2 l
Largest ETT That Fits Into LMA (ID in mm) 6.0. 6.0, 4.5, 3.5,
cuffed cuffed without cuff without cuff
Largest Fiberscope That Fits Into ETT (Middle Column) (OD in mm) 5.0+ 5.0+ 3.5+ 2.7§
*From Benumof JL: Use of the laryngeal mask airway to facilitate hberoptic-endoscopy intubation. Anesth Analg 74:313-314, 1992. Used with permission. +01ympus BF-P20D. t-Pentax FB-10Hor F1-10P. §01ympus PF-27M. Abbreviations: LMA = laryngeal mask airway: ETT = endotracheal tube; ID = inside diameter; OD = outside di ameter.
496
Anesthesia for Special Elective Diagnostic Procedures
Figure 14—4 Fiberoptic bronchoscopy in an anesthetized child using the Olympus BF 3C10 scope (external diameter 3.6 mm). (From Khoo ST: Anaesthesia for fiberoptic bronchoscopy in children. Anaesthesia 42:248-249, 1990. Used with permission.)
Figure 14-5 Use of the anesthesia mask with diaphragm and oral airway intubator as aids to fiberoptic tracheal intubation in an anesthetized (and paralyzed) patient. (Reproduced with permission from Benumof JL: Management of the difficult airway: with special emphasis on the awake intubation. Anesthesiology 75:10871110, 1991.)
Anesthesia for Special Elective Diagnostic Procedures
Table 14-3 LOCAL ANESTHETIC TECHNIQUES FOR NASOTRACHEAL FIBEROPTIC BRONCHOSCOPY 1. Cocaine (10 per cent) to nasal mucosa 2. Soft nasopharyngeal airways liberally coated with lidocaine ointment 3. Local anesthetic spray to nasopharyngeal, oropharyngeal, laryngeal, and tracheal mucosal surfaces a. Tetracaine 0.5 per cent with epinephrine b. Lidocaine 4 per cent 4. Block of the lingual branch of IX 5. Superior laryngeal nerve block a. Externally by needle b. Internally by swab soaked in local anesthetic 6. Transtracheal block 7. Local anesthetic spray down suction channel of fiberoptic bronchoscope
and atraumatically passed through the nose into the trachea when the patient has had adequate topical anesthesia and sedation. The naris selected is the one that the patient can breathe through most easily (presumably, this is the naris that is the larger or less obstructed one). There are several local anesthetic techniques that clearly facilitate nasotracheal fiberoptic bronchoscopy; all techniques or combination of techniques must anesthetize the nose, pharynx, larynx, and trachea (Table 14-3). The nose may be anesthetized by spraying either 0.5 per cent tetracaine with epinephrine or 4 per cent lidocaine with epinephrine, by topically applying cocaine by cotton-tipped wooden pledgets (total dose over 15 min in a 70-kg patient not to exceed 300 to 350 mg [3 to 4 ml of 10 per cent solution or 3.3 mg/kg]), or by passing progressively larger soft nasopharyngeal airways that are liberally coated with lidocaine ointment. The first two nasal techniques both anesthetize the nose and shrink the nasal mucosa by active vasoconstriction, whereas the last technique anesthetizes the nose and dilates the nasal cavity by mechanical means. The pharynx may be anesthetized by spraying local anesthetic through the oral cavity over the tongue and pharynx, by gargling viscous lidocaine, or by blocking the lingual branch of IX bilaterally. This easily performed block is effective in eliminating the gag reflex and hemodynamic response to a laryngoscopy."· I9 The patient's tongue is gently retracted laterally (by pulling the tip of the tongue with gauze and by pushing it with a tongue blade), exposing the palatoglossal arch (also called the anterior tonsillar pillar (Fig. 14-6). The base of the palatoglossal arch forms a U- or J-shaped band of tissue or bridge starting from the soft palate, running along the lateral pharyngeal wall
497
to the lateral margin of the base of the tongue. The palatoglossal arch is pierced approximately 0.5 cm from the lateral margin of the root of the tongue at the point where it joins the floor of the mouth (at the trough of the U- or J-shaped band of tissue) using a 25-gauge spinal needle. The length of the spinal needle allows the local anesthetic syringe to be outside of the mouth and therefore not in the line of vision. The needle is inserted 0.5 cm and an aspiration test is performed. Air will enter the syringe if needle placement is too deep, causing the tip of the needle to exit from the posterior aspect of the palatoglossal arch and enter the oropharynx. An aspiration test is also helpful in reducing the possibility of an intravascular injection. Two milliliters of 2 per cent lidocaine are slowly injected. The procedure is then repeated on the opposite side. Because the injection is made into loose sublingual tissue, there should be minimal resistance to injection. Within a few minutes, the posterior third of the tongue, the pharynx, and the pharyngeal side of the epiglottis (vallecula) should be adequately anesthetized to allow direct laryngoscopy with a Macintosh blade with minimal discomfort or gagging. The larynx may be additionally anesthetized by
Figure 14-6 The lingual branch of the glossopharyngeal nerve (IX) can be blocked by instilling l to 2 ml of a local anesthetic at the trough of the glossopalatal arch (this band of tissue sweeps upward from the lateral base of the tongue to the palate; it is commonly called the anterior pillar). The trough of the glossopalatal arch may be best visualized by retracting the tongue with a tongue blade (in this example a gloved finger is being used). (From Benumof JL: Management of the difficult airway: With special emphasis on the awake intubation. Anesthesiology 75:1087-1110, 1991. Used with permission.)
498
Anesthesia for Special Elective Diagnostic Procedures
continuing the local anesthetic spray via the mouth and nose, by superior laryngeal nerve blocks with 2 ml of 2 per cent lidocaine, or by transtracheal block with 4 ml of 4 per cent lidocaine during exhalation. The superior laryngeal nerve block technique .consists of needle application of local anesthetic to the thyrohyoid membrane between the superior lateral cornu of the thyroid cartilage and the inferior lateral margin of the cornu of the hyoid bone (Fig. 14—7).20*21 An internal superior laryngeal nerve block technique consists of painting the pyriform fossae with sponges that are soaked with local anesthetic. Superior laryngeal nerve block anesthetizes the lower pharynx, laryngeal epiglottis, vallecula, vestibule, aryepiglottic fold, and posterior rima glottis. The trachea may be anesthetized by the local anesthetic spray as the patient breathes the nebulized material, by the transtracheal block, or by local anesthetic sprayed down the suction channel of the fiberoptic bronchoscope (spray as you go). In one study that compared 100 mg of transcricoid lidocaine with 240 mg of spray-as-you-go lidocaine, the transcricoid method was significantly more effective in decreasing the number of coughs per minute.22 Once the fiberoptic bronchoscope is introduced into the trachea, topicalization of the distal tracheobronchial tree can be accomplished by continuing to spray local anesthetic down the suction channel of the fiberoptic bronchoscope. Transtracheal block is the least performed maneuver be-
Figure 14-7 This schematic diagram shows the point of emergence of the superior laryngeal nerve through the thyrohyoid membrane between the superior lateral cornu of the thyroid cartilage and the inferior aspect of the hyoid bone. Local anesthetic injected at this point will cause a superior laryngeal nerve block.
cause of the impressive anterior and posterior mucosal bruise that can be caused by the block and regularly observed with the fiberoptic bronchoscope after the block (personal observation). As can be seen from the preceding list of techniques, the one that can anesthetize everything (naris, pharynx, larynx, trachea) is simply spraying local anesthetic through the nose and the oral cavity. In my experience, 0.5 per cent tetracaine with epinephrine provides a more complete block than 4 per cent lidocaine. A very simple but effective system that produces a very dense cloud or mist of local anesthetic is shown in Figure 14-8. In my experience, this system has made nebulization of local anesthetic to all the mucosal surfaces the single most effective local anesthetic maneuver and can provide adequate anesthesia by itself alone. It is extremely important for the anesthetist to be unhurried and complete (10-sec spraying periods should be alternated with 10- to 20-sec rest periods); this approach usually requires at least 15 min to result in adequate anesthesia. The total dose of tetracaine over 15 min in a 70-kg patient should not exceed 100 mg (20 ml of 0.5 per cent solution), and the total dose of lidocaine over 15 min in a 70-kg patient should not exceed 400 mg (10 ml of 4 per cent solution) or 6 mg/kg. Nebulized tetracaine appears to be more potent and to have a significantly longer duration of action than lidocaine. Table 14-4 shows the most important properties of the major agents used for topical anesthesia of the respiratory tract. It cannot be stressed enough that tracheal intubation with a fiberoptic bronchoscope, is a difficult procedure if anesthesia is inadequate (the operator will have a rapidly and widely moving field of vision), whereas the procedure is relatively easy if the patient is adequately anesthetized (the operator has a quiet field). Consequently, the author uses, in addition to the local anesthetic spray of all the mucosal surfaces (Fig. 14-8), cocaine to the nose or soft nasopharyngeal airways coated with lidocaine ointment (if the nose is going to be used), pharyngeal gargle of viscous lidocaine, and, perhaps, bilateral superior laryngeal nerve blocks. The topical agents can cause local tissue irritation, and in susceptible patients (asthmatics) attempted anesthesia of the respiratory tract may cause bronchospasm; the bronchospasm is unrelated to airway histamine responsiveness.23 Bronchospasm in these patients can be eliminated by prophylactically administering an inhaled p2-agonist.24 If too large a dose of a topical anesthetic is given too rapidly, absorption into the systemic circulation may occur to a degree sufficient to produce toxic reactions. The initial effect is usually one of central excitation, which may be manifested by agitation, yawning, laughing, looking around
Anesthesia for Special Elective Diagnostic Procedures
499
Figure 14-8 This schematic diagram shows the system for creating a fine mist of local anesthetic. Oxygen green tubing is connected to an oxygen tank. A hole is cut in the oxygen green tubing near the nebuhzation chamber. When oxygen is flowing in the green tubing and a finger is placed over the hole, a fine dense mist from the nebuhzation chamber results. The size and velocity of spread of the mist are proportional to the oxygen flow rate.
500
Anesthesia for Special Elective Diagnostic Procedures
the room, nausea, vomiting, and twitching; higher serum levels cause seizures. The risk of these toxic central nervous system effects may be minimized by premedicating the patient with small doses of a benzodiazepine and/or a barbiturate. Other possible adverse effects include cardiovascular depression, hypotension, syncope, and arrhythmias. Rarely, allergic or idiosyncratic reactions occur, including rashes, eczema, edema, bronchospasm, methemoglobinemia, and agranulocytosis. Hypersensitivity reactions seem to be more common with the ester type drugs (see Table 14-4). It is apparent from Table 14-3 that there are several other permutations of local anesthetic techniques that may be used to anesthetize all of the upper airway mucosal surfaces. However, no matter which technique or combination of techniques is used, unless the anesthetist is thorough and patient and allows for a sufficient amount of time (approximately 20 min) to achieve adequate anesthesia, any approach may result in spotty anesthesia. With all techniques, intravenous lidocaine is useful in depressing the cough reflex as well as decreasing the incidence of arrhythmias.25-27 b. GENERAL ANESTHESIA
It is important to realize at the outset that the amount of general anesthesia required for bronchoscopy can be greatly diminished if some topical local anesthesia is used. Again, the single most effective maneuver that one can do to anesthetize the entire upper respiratory passages is to nebulize local anesthetic to all the upper airway mucosal surfaces. Two types of general anesthesia, or a combination of the two, may be used with a fiberoptic bronchoscope. Either an oxygen-nitrous oxide, intravenous barbiturate/narcotic, short-acting muscle relaxant (succinylcholine drip, atracurium or vecuronium), and/or a halogenated drug anesthesia technique may be used. The factors that determine the choice between the two techniques are extensively discussed in chapter 8. It is important to emphasize that with either type of general anesthetic adequate anesthesia to prevent laryngospasm and bronchospasm is of paramount importance. Administration of intravenous lidocaine will significantly depress the cough reflex and decrease the incidence of premature ventricular contractions during either type of general anesthesia.25-27 Postoperatively, the patients should breathe an elevated F,02 for several hours (see complications).17·28 4. Complications Intraoperative laryngospasm and bronchospasm due to inadequate anesthesia are the most common intraoperative emergencies during fiberoptic bronchoscopy.29· 30 Intraoperative hypoxemia is also
common6-7·29·30 and can be a result of either poor ventilation (due to bronchospasm or using a fiberoptic bronchoscope that is too large) or atelectasis (due to suctioning from a wedged position). Major cardiac arrhythmias may develop in as many as 11 per cent of patients during fiberoptic bronchoscopy, and the occurrence of the arrhythmias is often associated with the presence of a Pa02 < 60 mm Hg.31 If hypoxemia is corrected, the arrhythmias are usually self-limiting and do not need to be treated. Deleterious hemodynamic effects of PEEP and parenchymal barotrauma may result intraoperatively if the cross-sectional area of the endotracheal tube is reduced greatly enough by the fiberoptic bronchoscope, thereby preventing the escape of expired gases. A chest X-ray should be obtained postoperatively to rule out mediastinal emphysema and pneumothorax if high airway pressures were noted during the procedure.3·8 A chest X-ray does not appear indicated if the perioperative period was completely uncomplicated.32· 33 Bleeding is an infrequent complication of fiberoptic bronchoscopy (may follow transbronchial biopsy).31 Hypoxemia develops during fiberoptic bronchoscopy in both healthy volunteers and sick patients, with an average decline of the Pa02 by 20 mm Hg, and lasts for 1 to 4 hours following the procedure.28·34-36 Hypoxemia occurring after fiberoptic bronchoscopy is mainly due to atelectasis caused by suctioning while the bronchoscope was in a wedged position. Consequently, in patients with an endotracheal tube, the anesthesiologist should sigh the patient with positive pressure at the end of the procedure. In addition, patients may develop increased airway obstruction after fiberoptic bronchoscopy,28 probably secondary to direct mechanical activation of cough and irritative reflexes in the airway and possibly by direct traumainduced mucosal edema.17 Use of helium oxygen mixtures can greatly decrease the resistance to airflow.9·I0
B. Rigid Ventilating Bronchoscope 1. Indications There are several relative indications for the use of a rigid ventilating bronchoscope. First, a ventilating bronchoscope has a hole near the end of the bronchoscope that allows ventilation of the contralateral lung when the tip of the bronchoscope is ir the other lung; therefore, the ventilating broncho scope permits bilateral lung ventilation when th< tip of the bronchoscope is inside one of the main stem bronchi. The ventilating bronchoscope ma; be attached to the anesthesia circle system vi;
Anesthesia for Special Elective Diagnostic Procedures
a sidearm (Fig. 14-9, bottom panel,37 and Fig. 14-10) or to a jet injector for Venturi ventilation (see Fig. 14-9, top panel37 and Fig. 14-11). Se cond, it is the instrument of choice for foreign body removal (even in broncholithiasis).38 Third, massive hemoptysis (500 ml per 24 hours) must be assessed with an open-tube bronchoscope (see chapter 17). Adequate suctioning, removal of blood clots with a large forceps, and packing of a major bleeding site can be much more readily ac
501
complished with a rigid rather than a fiberoptic bronchoscope. In addition, control of the airway can be maintained during all these therapeutic ma neuvers. On rare occasion, it may be necessary to position the bronchoscope in one main-stem bron chus to provide ventilation to one lung (and seal off the bleeding lung) while the patient is taken to the operating room. Fourth, constricting and/or bleeding lesions may be by-passed by an opentube bronchoscope if the airway becomes seriously
Figure 14-9 The Storz adult bronchoscope. Top panel, With Sanders jet injector attachment. The anesthesia circle system attachment is plugged. Bottom panel. With connection to anesthesia circle system. The jet injector attachment is plugged. (From Vaughan RS: Endobronchial intubation. In Latto IP, Rosen M (eds): Difficulties in Tracheal Intubation. London. Bailliere Tindall. 1985, ρ 162. Used with permission.)
502
Anesthesia for Special Elective Diagnostic Procedures
Rigid Ventilating Bronchoscope
Figure 14-10 This schematic diagram shows a rigid ventilating bronchoscope system, which consists of the anesthesia circle system attached to a flexible connector that is attached to the sidearm of the bronchoscope. With the proximal eyepiece in place, most of the inspired gas goes into the patient. However, since the bronchoscope cannot fully fill the area of the trachea, there is a variable leak around the distal end of the bronchoscope. Exhaled gases are through the anesthesia circle system. When the eyepiece is removed, there is a very large leak out the proximal end of the bronchoscope.
compromised. Fifth, the open-tube bronchoscope allows a large biopsy of a main bronchial neoplasm to be taken. Indeed, one report has described using the rigid bronchoscope to core out (by biopsy and by screwing the bronchoscope) obstructing lesions in patients with terminal malignancy.39 However, the use of the rigid bronchoscope in this manner has been condemned.40 Sixth, the rigid bronchoscope is necessary for the placement of tracheobronchial tree stents.41 Seventh, the rigid bronchoscope allows the surgeon to palpate the carina preoperatively to assess operability and to identify extension of subcarinal disease. Finally, the rigid bronchoscope is necessary for small children (see chapter 18). 2. Ventilatory Considerations a. SIDEARM CONNECTION
The rigid ventilating bronchoscope (see Fig. 14-9, bottom panel, the Storz or Negus rigid bronchoscope) may be attached directly to the anesthe-
sia machine circuit via a sidearm adapter (see Fig. 14-9, bottom panel, and Fig. 14-10). Consequently, inspired oxygen and anesthetic gas concentrations are known and can be administered with conventional or high-frequency positive-pressure ventilation. Although spontaneous ventilation may be allowed, the dangers of inadequate ventilation due to bronchospasm and a tight chest wall (too light anesthesia), or hypoventilation (too deep anesthesia) and tracheobronchial tree damage due to unexpected coughing usually cause the risk/benefit ratio of spontaneous ventilation to be high. Consequently, these patients should usually be paralyzed, and ventilation should be controlled with positive pressure. Because the usual rigid bronchoscope has an external diameter of 11 mm, there is normally a variable leak around the distal end of the bronchoscope (with the proximal eyepiece in place), but this may be compensated for by using a high gas flow rate (greater than 10 L/min) (first choice) or by packing the pharynx with saline-soaked gauze
Anesthesia for Special Elective Diagnostic Procedures
503
Rigid Venturi Bronchoscope
Figure 14-11 This schematic diagram of a rigid Venturi bronchoscope shows that the jet of gas exiting from a Venturi needle placed within the lumen and parallel to the long axis of the bronchoscope entrains gas from the environment. The jetted gas comes from a high-pressure source and an intermittent (I2 L/min) injector. The flow of gas from the tube into the patient is equal to the volume of gas through the jet plus the air entrained.
(second choice). It is possible to provide effective ventilation in the vast majority of patients with this system. However, the proximal eyepiece must be removed during suctioning, foreign body manipulations, or the taking of biopsy specimens. Since ventilation must be interrupted when the surgeon removes the occluding eyepiece (all the inspired gas including the inhalational anesthetics, which can sedate the endoscopist, go into the room out the open proximal end), a high F,0 2 should be used if the surgeon requires removal of the eyepiece for a considerable period of time. One to 2 min maximum of apnea may be allowed at any one time before ventilation must be resumed, with a shorter time allowed for obese patients and patients with lung disease.42 b. VENTURI CONNECTION
The rigid Venturi-effect bronchoscope relies on an intermittent (10-20 L/min) high-pressure oxygen jet to entrain air and ventilate the lungs with an air-oxygen mixture (see Fig. 14-9, top panel, and Fig. 14-11). The Venturi jet is delivered via a reducing valve (Sanders injector) to a 1.0- to 1.5-
inch, 18- or 16-gauge needle that is inside and parallel to the lumen of the bronchoscope. The high velocity of the Venturi jet exiting from the end of the needle creates a negative pressure just outside the open end of the needle that draws environmental air into the jet stream. Oxygen jets (from a 50-psi source) usually entrain two to three times their own volume of ambient air.43 The Venturi jet plus the entrained air creates a positive intraluminal tracheal pressure and a tidal volume. At any given reducing valve pressure, the exact amount of tracheal pressure and tidal volume depends on the driving pressure from the reducing valve, the size of the needle, and the diameter, length, and type of bronchoscope. With the usual Venturi bronchoscope and an 18-gauge needle orifice, a 50-psi source causes a flow of 160 L/min and a peak airway pressure of 27 cm H-,0.44 The jet reducing-valve pressure can be adjusted according to the tracheal pressure and observed chest movements; usually a tracheal pressure of 30 cm H 2 0 results in normocarbia. In comparison with the intermittent gas exchange often required by the rigid ventilating bronchoscope, the rigid
504
Anesthesia for Special Elective Diagnostic Procedures
Venturi bronchoscope provides more constant and adequate ventilation.45 Use of high-frequency jet ventilation at rates of 150 to 300 breaths/min (using a commercial high-frequency ventilator) through the rigid Venturi bronchoscope has resulted in adequate gas exchange (but still less effective than jet ventilation).46 3. Anesthetic Technique a. LOCAL ANESTHESIA
Passing a rigid bronchoscope into the trachea requires a considerable amount of pressure, mucosal stimulation, and significant neck extension that is not required for fiberoptic bronchoscopy. In addition, there is no guarantee in an awake patient that the patient will not suddenly move at a critical point as a result of some noxious or unpleasant stimulus. Consequently, rigid bronchoscopy is usually performed under general anesthesia. If local anesthesia must be used, the premedication and various local blocks, as previously described for fiberoptic bronchoscopy and used collectively, will usually render rigid bronchoscopy tolerable. b. GENERAL ANESTHESIA
Bronchospastic and asthmatic patients must have good pharmacologic control of bronchomotor tone before rigid bronchoscopy. General anesthesia for rigid bronchoscopy should also be preceded by some topical local anesthesia of the larynx. Thus, less general anesthesia will be required, permitting easier, more rapid awakening and return of laryngeal reflexes. Both nitrous oxide/intravenous anesthesia/short-acting muscle relaxant and halogenated drug/short-acting muscle relaxant anesthesia may be used with ventilating bronchoscopes. Intravenous and intratracheal lidocaine with either type of general anesthetic minimizes straining, coughing, and arrhythmias.25-27 Of course, the choice of anesthesia will also be based on the overall medical status of the patient and the skill and speed of the surgeon. Because the Venturi principle renders the inspired oxygen concentration and inhalational anesthetic concentration uncertain, an oxygen-narcotic-thiopentalmuscle relaxant general anesthetic is most often used for bronchoscopy with a Venturi bronchoscope. Paralysis is a helpful part of the Venturi technique because a compliant thorax with minimal resistance is sometimes necessary to ensure adequate ventilation. 4. Complications Experiencing the passing of a rigid bronchoscope can be very unpleasant for an awake patient.
The rigid bronchoscope may fracture teeth, and it is occasionally necessary to use extreme extension of the neck to insert the bronchoscope, which may cause vasovagal reactions. Massive hemorrhage from directly traumatized lesions may occur.30·33 The tip of the rigid bronchoscope can perforate the mucosa, causing pneumomediastinum and subcutaneous emphysema. Because a large leak may occur around the distal end of the bronchoscope, which may prevent adequate ventilation, and periods of apnea may be required, there is an increased risk of hypoxemia and hypercarbia.30·33 In view of the inherent intense stimulation associated with the procedure, arrhythmias (especially if either hypoxemia, hypercarbia, inadequate anesthesia, or halothane anesthesia is present) may occur.30·33 In addition, there are several potential special disadvantages in using the rigid Venturi bronchoscope. First, there is no continuous documentation of delivered oxygen and inhaled anesthetic drug concentration. Second, the high gas flow rates resulting from the jet ventilation may cause blood or tumor particles to be accidentally blown down into more peripheral bronchi. Third, care must be taken not to generate excessive carinal airway pressures in children owing to a tight glottic fit, which may prevent escape of gas (exhalation) around the bronchoscope. III. MEDIASTINOSCOPY A. Indications and Surgical Considerations Mediastinoscopy is commonly performed be fore thoracotomy to establish a diagnosis and/or t< determine resectability of a lung carcinoma47 am to establish a correct diagnosis in patients sus pected of having a lymphoma and in those with mediastinal mass.48 The importance of determinin cell type is most simply and dramatically undei scored by pointing out the example that lymphe mas require radiation, whereas thymomas requir resection. After a suprasternal notch incision (cervical mt diastinoscopy), a tunnel is created (through th pretracheal fascia) by blunt dissection along th anterior and lateral walls of the trachea into th mediastinum, behind (posterior to) the aortic arcl and down to the subcarinal area (Fig. 14-12). Th procedure allows for direct inspection and biops of the superior mediastinal lymph nodes, which 1: posterior to the aortic arch (i.e., the anterior ar lateral para-main-stem bronchial, anterior subcai nal, anterior, and lateral paratracheal lymph node (see Fig. 14-12). Left-sided central and apical 1< sions are much more easily observed via a le
Anesthesia for Special Elective Diagnostic Procedures
505
Mediastinoscopy
Mediastinoscope Right Common Carotid
Left Recurrent Laryngeal Nerve
Right Subclavian Artery
Esophagus Trachea Left Subclavian Artery
Innominate Artery (pinched)
Left Common Carotid Artery Anterior and Lateral Para-Tracheal and Para-Mainstem Bronchial Nodes
Anterior Subcarinal Lymph Nodes
Superior Vena Cava Figure 14-12 This schematic diagram shows the placement of a mediastinoscope into the superior mediastinum. The mediastinoscope passes in front of the trachea but behind the thoracic aorta. This location of the mediastinoscope allows for sampling of anterior and lateral para-mainstem bronchial lymph nodes, anterior subcarinal lymph nodes, and anterior and lateral para-tracheal lymph nodes. Anatomic structures that can be compressed by the mediastinoscope (see areas marked by an *) and that can cause major complications are the thoracic aorta (rupture, reflex bradycardia), innominate artery (decreased right carotid blood flow can cause cerebral vascular symptoms, and decreased right subclavian flow can cause loss of right radial pulse), trachea (inability to ventilate), and vena cava (risk of hemorrhage with superior vena cava syndrome).
506
Anesthesia for Special Elective Diagnostic Procedures
anterior approach (this anterior mediastinotomy procedure uses a second rib interspace incision and is ordinarily a much less complicated procedure than mediastinoscopy but not always).49 Tumors of the thymus and anterior mediastinum are also not examined by the usual diagnostic cervical mediastinoscopic approach because they are anterior to the great vessels; thus, an anterior mediastinotomy is also required to examine this area. Table 14-5 summarizes one author's approach to mediastinoscopy for lung regions.47 A parasternal mediastinoscopy can also be used for the evaluation of left upper lobe lesions.50 Contralateral positive nodes (to a lung carcinoma) are considered an absolute contraindication to thoracotomy, whereas if only ipsilateral nodes are involved, thoracotomy may be performed depending on the expected resectability of the tumor. Previous mediastinoscopy is a strong relative contraindication to a repeat procedure because scarring eliminates the plane of dissection. "Strong relative contraindication" is used rather than the classical "absolute" term because two studies/reviews of 10151 and 14052 repeat mediastinoscopies found a 0 per cent mortality rate and 23 per cent and 7 per cent complication rate, respectively, all of which were successfully treated. Relative contraindications to mediastinoscopy include superior vena cava syndrome, severe tracheal deviation, cerebrovascular disease, and thoracic aortic aneurysm.53 With the advent of computed tomography, and magnetic resonance imaging, the role of mediastinoscopy may diminish considerably in the future in the sense that negative scans may obviate the need for invasive staging.
B. Anesthetic Technique In addition to the usual preanesthetic evaluation of patients, one should specifically look for the Table 14-5 ACTUAL GUIDELINES FOR THE USE OF MEDIASTINOSCOPY FOR DIAGNOSIS OF LUNG LESIONS* No mediastinoscopy Peripheral squamous cell carcinoma Undiagnosed peripheral nodule 20 years, adolescent to young adult; recurrent pulmonary infection Systemic—from aorta; often a single, large vessel Pulmonary—-inferior pulmonary vein Uncommon
Foregut connection Bronchial communication
Very rare Present, small
Age at presentation and symptoms
Arterial supply Venous drainage
Extralobar Rare Male (80%) 90%. left Above or below diaphragm Has its own separate pleural investment-visceral pleura 60% < l year, neonate; respiratory distress Systemic—from pulmonary artery or aorta; usually small vessels Systemic—azygos or hemiazygos vein; less often portal vein Common (>50%), e.g., congenital diaphragmatic hernia (30%) More common None
544
Anesthesia for Special Elective Therapeutic Procedures
sheep- and cattle-raising regions of the world (i.e., Australia, New Zealand, South Africa, South America, and the Mediterranean countries of Europe, Asia, and Africa). Indeed, in one hospital (in Ankara, Turkey) over a 16-year period, 1115 patients with pulmonary hydatidosis were treated surgically.113 A human may become infected, usually in childhood, when his or her hands or food becomes contaminated with infected canine (primary host) fecal material containing eggs, which are ingested. The eggs hatch in the stomach and proximal small intestine, releasing the embryos (larval stage), which penetrate the proximal intestinal mucosa to enter the portal venous circulation. Most of the embryos are filtered out in the liver, but some escape the liver to be disseminated, the common sites being the lungs, brain, and bones. The hydatid fluid resembles crystal-clear water: It is odorless and colorless. It suspends the daughter cysts in solution. A number of significant complications are found with hydatid cyst of the lung. The cyst progressively increases in size at a rate varying from a few millimeters to 5 cm a year, growing more rapidly in children. Pressure effects occur owing to compression by the hydatid cyst of adjacent structures (i.e., bronchus, superior vena cava, esophagus, or neurovascular structures of the thoracic inlet). Spontaneous rupture of the hydatid cyst into a bronchus may cause any of the following: (1) Dramatic expectoration of cyst fluid and fragments of the laminated membrane (a spontaneous cure may follow the evacuation of cyst contents and collapse of the cyst wall); (2) sudden death from asphyxiation caused by flooding of the bronchial tree with hydatid fluid; (3) sudden death from anaphylactic reaction to the cyst contents; (4) secondary infection forming a chronic lung abscess or a localized area of bronchiectasis manifested by chronic cough, mucopurulent or dark bloody sputum, and repeated febrile episodes; (5) no symptoms. Diagnosis of pulmonary hydatid cyst should be suspected in any patient who lives in an endemic area, who is older than 3 years, and who presents with single or multiple opacities on the chest radiograph. The indirect hemagglutination test is the clinical test of choice, with a false-positive rate of only 1 to 2 per cent. The diagnosis of a pulmonary hydatid cyst is an indication for surgery. Complete excision of the disease process with maximum conservation of lung tissue is desirable. The principles of surgical excision are as follows. The intact cyst may be removed by simple enucleation without needle aspiration. During delivery of the cyst, it should not be touched or grasped to avoid rupture. The surgical field must
always be protected from the possibility of spillage of cyst fluid. The intact cyst may be rendered sterile by first instilling hypertonic saline solution (10 ml of 10 per cent NaCl solution) into the cyst. The intact cyst may also be aspirated before removal. After the endocyst and its laminated membrane have been removed, one should locate the patent bronchial openings, which are invariably present in the fibrous wall of the ectocyst or pericyst. These openings must be carefully closed with fine suture. At times, segmental resection or lobectomy may be necessary for removal of pulmonary hydatid cyst if the lung tissue has been destroyed by prolonged compression or infection.
4. Pneumatocele Pneumatoceles are inflammatory in origin, characteristically appearing during the healing phase of pneumonia. They are radiolucent zones in the lung fields that become more clearly visible as the pneumonia resolves and then disappear spontaneously.
5. Bullous Emphysema The same mechanisms of dilation and destruction that are responsible for the development of emphysema are responsible for the formation of the bullae. In fact, just as emphysema may occur on a familial basis, so may bullae (and repeated episodes of spontaneous pneumothorax)."4 Destruction of lung parenchyma from inflammation results in loss of its elasticity and formation of an air-filled thin-walled space within the lung or cyst."5· "6 The walls of the bullae may be formed by connective tissue septa and/or compressed lung parenchyma and/or pleura. Morphologically, bullae may be classified into three types: type I or narrow-necked bullae; type II or superficial (a pleural surface) broad-based bullae; and type III or deep (surrounded by emphysematous tissue) broad-based bullae. There are several complications of bullae. Expiratory obstruction, with progressive air trapping, causes the space-occupying bulla to enlarge slowly. Occasionally a "tension bulla" may produce acute cardiorespiratory embarrassment. The development of a spontaneous pneumothorax caused by rupture of a bulla may constitute an emergency indication for surgical treatment if the pneumothorax is recurrent, persistent, or life-endangering. Bullae sometimes develop an air-fluid level and appear to have a clinical picture of lung infection and abscess; however, surgery is not always necessary"2 because complete resolution
Anesthesia for Special Elective Therapeutic Procedures
may be obtained medically."7 Alternatively, infected bullae may be drained percutaneously."8 Bullectomy is the surgical resection of one or more bullae and is performed only in selected patients. The most important physiologic effect of bullae is to compress the surrounding lung, and the release of the lung compression after bullectomy is considered to be the most important factor in symptom relief rather than eliminating dead space ventilation."1 Indications for bullectomy in chronic obstructive pulmonary disease patients include intolerable breathlessness even after full medical therapy, rapidly enlarging bullae, or the repeated occurrence of pneumothorax (a history of which most patients have).119 The goal of surgery is to remove the bullae while preserving as much functioning lung as possible. Therefore, lobectomy is not only undesirable but often unnecessary. Patients who are considered too compromised to tolerate an open approach may tolerate a closed thoracoscopic approach.120 The compression of a large area of lung can be visualized on the chest roentgenogram (as well as on an angiogram) (e.g., Fig. 15-22)" 2 as the crowding together of pulmonary vessels. However, computed tomography is a more accurate way to localize the bulla as well as to determine its functional characteristics and the nature of the surrounding lung architecture (Fig. 15-23).121-123 In addition, an equally strong case for functional impairment can be made if radioisotope studies show that the compressed area has good perfusion and some, but reduced, ventilation124·125 (see chapter 5 and Fig. 5-22). Whole-lung pulmonary function tests are not useful for determining candidates for bullectomy.121 The long-term results of bullectomy for giant bullae in emphysema are encouraging.126 In 27 emphysematous patients who had either unilateral (10 patients) or bilateral (17 patients) bullae that occupied more than 50 per cent of the hemithorax, bullectomy significantly decreased dyspnea, they were no longer functionally disabled, and mean survival time was greater than 7 years. In fact, the larger the bullae (and presumably the amount of compressed normal lung), the better are the functional results.127 The postoperative spirographic improvement depends on the type of bullae (e.g., resection of bullae with open communication with the bronchial tree results predominantly in improvement of forced expiratory flow in 1 sec [FEV,] as a percentage of vital capacity). In addition, the more normal the compressed lung, the more dramatic, positive, and enduring are the results.128 Thoracoscopic C0 2 laser treatment of bullous emphysema in 20 of 22 patients has been success-
545
Figure 15-22 A, Chest roentgenogram demonstrates large ^biapical bullae. B, Pulmonary angiogram demonstrates intact but crowded vasculature. (From Klingman RR. Angelillo VA. DeMeester TR: Cystic and bullous lung disease. Ann Thorac Surg 52:576-580, I99l. Used with permission.)
ful in moderately increasing forced vital capacity. FEV,, and maximal exercise treadmill times. Thoracoscopic carbon dioxide laser oblation of bullae has been so successful because it maximizes preservation of normal lung better than any open thoracotomy approach. It is likely that this approach to bullectomy will be used with increasing frequency in the next few years.129·13() Intracavitary drainage by tube thoracostomy (which requires rib resection, a purse-string pleurotomy, and a large balloon catheter) is another method of preserving normal lung tissue.131"133 The long-term prognosis after any surgical treatment of bullous emphysema is, of course, diminished in any patient who has chronic purulent bronchitis preoperatively and/or who does not abandon smoking.
546
Anesthesia for Special Elective Therapeutic Procedures
INSP B. Anesthetic Considerations Anesthesia for the removal of bullae involves several specific ventilation hazards (Table 15-8). First, most of these patients have severe generalized chronic lung disease with little or no ventilatory reserve. Thus, ventilation (which needs to be controlled once the chest is opened) of one severely diseased lung (the one without the giant bulla) may be hazardous and runs the risks of hypoxemia, hypercarbia, and pneumothorax on the ventilated side. If the ventilated lung also contains a bulla, the risks are obviously even greater. In addition, since general anesthesia is necessary for the procedure, it is likely that many patients with severe lung disease will be committed to at least a short period of postoperative mechanical ventilatory support. Second, when a bulla or air cyst is in communication with a bronchus, positive-pressure ventilation may cause it to increase in size.134 If a significant portion of the tidal volume enters the bullous cavity, alveolar dead space ventilation will be greatly increased, and unless there is an equivalent increase in minute ventilation, the rest of the lung may be inadequately ventilated. This complication is most likely to occur when the chest is opened because the chest wall no longer limits the expansion of the bulla. Third, because of the rapidity with which closed airspaces take up nitrous oxide and expand in size,135 this agent is best avoided (especially in patients whose bullae are thought to have poor communication with the bronchial system).136 Fourth, if a check valve is present in the airway that communicates with the cavity, overinflation and air trapping may occur within the cavity. Fifth, and most important, positive pressure within the bulla might cause it to rupture, creating
EXP
Figure 15-23 Three-dimensional reconstruction (posterior oblique view) of the computed tomography scan of the lungs of a patient with a large bulla in the left upper lobe. The measured volume of the bulla does not change from inspiration (INSP) to expiration (EXP), but there is a change in shape as a result of distortion. This would invalidate the measurement of ventilation from the change in area of the bulla in a single slice. (From Morgan MDL, Denison DM, Strickland B: Value of computed tomography for selecting patients with bullous lung disease for surgery. Thorax 41:855-862, 1986. Used with permission.)
a pneumothorax, which would likely be under tension if the chest were closed (especially in patients whose bullae are thought to have good communication with the bronchial system).134 Tension pneumothorax in these patients is usually a catastrophic event owing to the impairment of venous return and cardiac output as well as further compromise of ventilation. Insertion of a chest tube at this point would create, in effect, a large bronchopleural cutaneous fistula that could divert much of the ventilation out through the chest tube. Recently, highfrequency ventilation with low tidal volumes and airway pressure has been used successfully to avoid positive-pressure rupture of a bulla.137 The cornerstone of the anesthetic management of patients with giant bullae or cysts, whether they are undergoing thoracoscopic or open procedures, is the insertion of a double-lumen tube to allow differential treatment of the two lungs. Thus, in patients with unilateral disease, the double-lumen tube can allow adequate ventilation of the nondiseased side while preventing rupture of the diseased side. In patients with bilateral disease, the doublelumen tube still allows differential lung treatment to maximize gas exchange as well as provides an increased capability to deal with the complications of a ruptured bulla. For example, a double-lumen tube allows all possible permutations of high-fre-
Table 15-8 VENTILATION HAZARDS DURING BULLECTOMY I. Rest of lung is diseased II. Bullae may increase in size owing to: A. Intermittent positive-pressure breathing B. N 2 0 C. One-way check-valve III. Bulla may rupture —» pneumothorax
Anesthesia for Special Elective Therapeutic Procedures
quency ventilation, CPAP, positive end-expiratory pressure (PEEP), and zero end-expiratory pressure to the two lungs, depending on the pathology in each lung (see Differential Lung Ventilation, chap ter 20). In addition, as each bulla is resected, the double-lumen tube allows ventilation to the oper ated lung to be re-established for short periods, enabling the surgeon to identify and suture any air leaks that may be present (identification may in volve filling the hemithorax with saline and searching for bubbles).129 A ruptured bleb is usu ally small and cannot be seen as a hole and is always covered with fibrin materials. A ruptured emphysematous bulla is usually large enough to be identified as a hole. 129 The versatility of the double-lumen tube is par ticularly important in thoracoscopic procedures, which may average 3 hours in duration. 120 A double-lumen endotracheal tube can be in serted with the patient either awake and the airway topically anesthetized (for those with histories of repeated pneumothorax and/or bilateral bullae) or under general anesthesia (for the majority of pa tients) in order to isolate the affected lung and provide positive-pressure ventilation to the contra lateral lung.138· 139 While the depth of general an esthesia is being increased, spontaneous ventila tion may be maintained (most indicated in patients with a history of repeated pneumothorax or bilat eral bullae), but it should be realized that sponta neously breathing patients with significant pulmo nary disease under general anesthesia probably will not be able to ventilate themselves adequately. Alternatively, and preferably in the majority of patients, the patient can be anesthetized and both lungs ventilated using a limited amount of positive airway pressure; gentle ventilation by hand is the best way to ensure low airway pressures. If a ma jor air leak develops intraoperatively while posi tive-pressure ventilation is being used with a vol ume-cycled ventilator, inadequate ventilation may occur (the machine delivers the preset volume but it does not go into the lungs). Under these circum stances, a pressure-cycled high inspiratory flow ventilator must be used (e.g., Siemens Servo 900C, Solna, Sweden). 120 If positive-pressure ventilation is used while the chest is closed (always the case for the nonoperative thorax), it is important that the anesthesiolo gist be able to diagnose and treat a pneumothorax rapidly. External stethoscopes should be attached over each hemithorax at the points where breath sounds are maximal in order to monitor for pneu mothorax on each side. 140 However, it should be realized that advanced bullous disease may com pletely prevent breath sounds from being heard externally at all. In addition to a decrease in breath sounds, a pneumothorax or check-valve mecha
547
nism in a cyst may be signaled by an increase in airway pressure, tracheal shift to the opposite side, or hypotension out of proportion to the depth of anesthesia. Equipment for chest tube placement must be immediately available for these patients. However, chest tube placement for a ruptured bulla poses two problems. First, rupture of only one of several bullae present in a lung may result in a somewhat localized pneumothorax that is decompressed only by precise localization of the chest tube. Second, as discussed previously, a large bronchopleural cu taneous fistula may be created when the chest tube is placed, making ventilation difficult (see prior discussion and Bronchopleural Fistula, chapter 17). Theoretically, after bullectomy, the patient's pulmonary status should be improved because the "healthy 1 ' lung tissue that was previously com pressed by the bullae should now be able to ex pand. However, in my and others' experience,120 the weaning and extubation process, especially the spontaneous subsidence of air leaks, may take up to several days in some patients with advanced disease. When mechanical ventilation is required postoperatively, positive airway pressure should be minimized, if possible, to decrease the chance of producing a pneumothorax from rupture suture lines and/or residual bullae.141 When large air leaks do occur, a pressure-cycled ventilator may be required. If the leak is greater than 50 per cent οΐ the tidal volume, then the patient may require re-exploration.120 When the air leak is large, the chest tube should be put to water seal only (active suction greatly increases the leak). 120 When the air leak is large and refractory to all manipulations, including surgical explora tion, percutaneous partial extracorporeal carbon dioxide removal and return to spontaneous venti lation may be tried (and has been used success fully).142 * When bilateral bullectomy is done because of extensive disease in both lungs, a sternal splitting incision with the patient supine is usually used.143 Indeed, some surgeons routinely use a mediastinotomy and bilateral approach/exploration because of the high frequency of bilateral problems (e.g.. 33-60 per cent spontaneous occurrence of unilateral pneumothorax after unilateral operations).144 I4S However, sequential posterolateral thoracotomies may be planned if it is desired to see how the patient responds to the first bullectomy. With bi lateral bullectomy, the same anesthetic principles apply as with unilateral bullectomy. A double-lu men endotracheal tube is again the cornerstone of management, allowing differential lung treatment throughout the operation. If sequential posterolat eral thoracotomies are planned, provision for gain-
548
Anesthesia for Special Elective Therapeutic Procedures
ing access to the closed chest should be made in the way the skin is prepared and the patient is draped in case suspicion of a pneumothorax in the closed chest arises.
VI. PULMONARY RESECTION IN PATIENTS AFTER PNEUMONECTOMY A. General Considerations Patients who have had a pneumonectomy may have a new lesion appear in the remaining lung; the density, if it persists and increases in size, is almost always malignant. Pneumonectomy has been generally considered in the past to be a contraindication to further pulmonary resection. However, six case reports involving eight patients146-151 and one recent series of 15 patients152 have provided evidence that limited pulmonary resection after pneumonectomy is feasible with a low operative mortality. The long-term results show that resection of these "secondary" tumors can result in a reasonably prolonged and worthwhile survival in some patients.
B. Anesthetic Considerations These cases must be done with a single-lumen endotracheal tube. Fluid balance must be carefully regulated because the size of the pulmonary vascular bed will be decreased further from an already small size and the risk of pulmonary hypertension will be increased by the resection. Thus, all aspects of postoperative care, such as maintenance of fluid balance (risk of pulmonary edema and pulmonary hypertension), positive-pressure ventilation (risk of right-sided heart failure), and adequate evacuation of the pleural space (risk of lung compression), must be undertaken with increased care because of the patient's increased sensitivity to derangements in any of these factors.
VII. UNILATERAL BRONCHOPULMONARY LAVAGE153 A. General Considerations Unilateral bronchopulmonary lavage or massive irrigation of the tracheobronchial tree of one lung has been employed with good success in patients with pulmonary alveolar proteinosis as a means of removing the enormous accumulations of alveolar lipoproteinaceous material that these patients characteristically have.154-161 The lipoproteinaceous
material is thought to be surfactant,162 and the abnormal accumulation is due to failure of clearance mechanisms rather than enhanced formation.163 The abnormal accumulation of alveolar lipoproteinaceous material is bilateral and symmetric (but can be unilateral)164 and causes the classic chest roentgenographic picture of airspace consolidation with patchy, poorly defined shadows throughout the lung.165 Very commonly, the distribution of Xray opacities is "butterfly" or "bat wing"; the roentgenographic picture parallels the course of the disease. Computed tomographic scan of the chest may show the extent of the disease more clearly than plain radiography.166 The diagnosis is made in most patients between the ages of 20 and 50 years (although the disease has been described in patients from 0.5 to 72 years).164 Most patients have had symptoms for 2 to 3 years before the definitive diagnosis is made. The male to female ratio is 4 to 1, and many patients (approximately 50 per cent) have a history of exposure to dusts and chemicals.164 The occurrence of the disease in several siblings suggests a genetic predisposition. The differential diagnosis consists of sarcoidosis, extrinsic allergic alveolitis, tuberculosis, and pulmonary fibrosis. The airspace consolidation causes progressive hypoxemia and shortness of breath (first on exertion and then finally at rest),165· 167 and the lungs have a low compliance. Cough, fever, and chest pains are other common symptoms. Pulmonary function tests reveal a restrictive pattern164 and are depressed the most in patients with severe X-ray changes. The diagnosis of the alveolar proteinosis is made by correlating these clinical, roentgenologic, and laboratory data, but the definitive diagnosis rests with the results of a lung biopsy (usually transbronchial at the time of fiberoptic bronchoscopy),168· 169 but an open thoracotomy may be required to obtain an adequate tissue specimen.164 The transbronchial specimens should be multiple and from the area of greatest change on X-ray.169 The indications for lavage consist of P a 0 2 less than 60 mm Hg at rest or hypoxemic limitation of normal activity.170·I71 Infrequently, lung lavage may be performed in patients with asthma, cystic fibrosis, and radioactive dust inhalation.154,159·l72· m Unilateral lung lavage is performed under general anesthesia with a double-lumen endotracheal tube, allowing lavage of one lung while the other lung is ventilated (Fig. 15-24). In patients with alveolar proteinosis, lavage is performed on one lung and then after a few days rest on the other lung. After lung lavage, these patients usually have marked subjective improvement, which correlates with increases in P a 0 2 during rest and exercise, vital capacity and diffusion capacity, and clearing
Anesthesia for Special Elect'v
' erapeutic Procedures
549
Normal saline 30 cm above lung
Clamp Left side conn ted to anesthesia circuit
Left side double-lumen tube
PA Catheter in Right Lung
-Drainage by gravity Figure 15-24 Technique for providing unilateral pulmonary lavage. A left clear plastic double-lumen endotracheal tube allows ventilation to the left lung during lavage of the right lung (and vice versa). Normal saline is infused and drained by gravity; clamps on the connection tubes determine direction of fluid flow. (PA = pulmonary artery.)
550
Anesthesia for Special Elective Therapeutic Procedures
of the chest roentgenogram.164 1 7 4 · l 7 5 Some patients require lavage every few months, whereas others remain in remission for several years, and the disease may even eventually completely remit.164 I68 174. 175 Prolonged spontaneous remission is also possible.164176 B. Anesthetic Considerations153 This section discusses the technique for managing bronchopulmonary lavage in patients with pulmonary alveolar proteinosis. When the patient is admitted to the hospital, ventilation-perfusion scans of the lung are obtained. Ventilation can be maximized during lung lavage by performing the first lavage on the most severely affected lung, allowing the "better" lung to provide gas exchange. If the scan indicates relatively equal involvement (as is usually the case), the left lung is lavaged first, leaving the larger right lung to support gas exchange. Unilateral lung lavage is performed in the operating room, where the appropriate amount and type of equipment and ancillary personnel are present to enhance safety. Patient safety is also enhanced if a relatively constant team, composed of members of the departments of anesthesia and pulmonary medicine, becomes familiar with the nuances and technique of unilateral lung lavage. Patients are usually cooperative and require only light premedication. Since many of these patients are hypoxemic at rest, they are given oxygen by face mask following premedication and during transport to the warmed operating room. The procedure takes several hours to complete, and lavage fluid temperature cannot always be precisely controlled. Thus, some patients require external warming (e.g., with heating blanket) to maintain normal body temperature. The monitoring system consists of a blood pressure cuff, electrocardiogram, precordial stethoscope, temperature probe, pulse oximeter, and peripheral arterial and central venous pressure catheters. Patients with compromised cardiovascular function are monitored with a pulmonary artery catheter in place of the central venous catheter. After several minutes of preoxygenation (see later discussion), general anesthesia is induced with 3 to 4 mg/kg of thiopental in divided doses and inhalation of either isoflurane or halothane in 100 per cent oxygen. Isoflurane is relatively indicated in patients in whom therapeutic levels of theophylline (and the risk of arrhythmias) are present. Neuromuscular blockade is monitored with a peripheral nerve stimulator and is induced with a nondepolarizing muscle relaxant. When a suitable level of anesthesia has been reached, the
trachea is topically anesthetized with lidocaine and intubated with the largest size left-sided doublelumen endotracheal tube that can be passed atraumatically through the glottis (see chapter 9). A clear, plastic, disposable left-sided double-lumen tube is used because of the ease and certainty with which it is correctly positioned, the reliable leftcuff seal obtained (the right endobronchial cuff is small and inflates asymmetrically), and the ability to continuously observe the tidal movements of respiratory gas moisture (ventilated lung) and the lavage drainage fluid for leaking air bubbles (lavaged lung). The largest size tube is used because the left endobronchial cuff will make contact over a greater bronchial mucosal area with less air in the left cuff (compared with a small double-lumen tube). In addition, a large tube facilitates suctioning, which is an important consideration at the end of the case when the lungs need to be made as clear as possible. Precise placement of the tube and detection of leaks are essential because of the serious hazard of spillage during the lavage procedure. The position of the double-lumen tube must be confirmed with a fiberoptic bronchoscope as described in chapter 9, and cuff seal must be demonstrated to hold against 50 cm H 2 0 pressure using the catheterunder-water technique described in chapter 9. The eyes should be protected with a lubricant and eye pads. The question of patient position during unilateral lung lavage is important for there are major advantages and disadvantages related to each position (Table 15-9). The lateral decubitus position with the lavaged lung dependent minimizes the possibility of accidental spillage of lavage fluid from the dependent, lavaged lung to the nondependent, ventilated lung. However, during periods of lavage fluid drainage, pulmonary blood flow, which is gravity dependent, would preferentially perfuse the nonventilated, dependent lung, and the right-to-left transpulmonary shunt would be maximal. The lateral decubitus position with the lavaged lung nondependent minimizes blood flow to the nonventilated lung but, on the other hand, inTable 15-9 UNILATERAL LUNG LAVAGE: PATIENT POSITION I. Lateral decubitus with lavaged lung nondependent Advantage: minimizes blood flow to the nonventilated lung Disadvantage: maximizes the possibility of spillage II. Lateral decubitus with lavaged lung dependent Advantage: minimizes the possibility of spillage Disadvantage: maximizes blood flow to the nonventilated lung III. Supine position Balances spillage and blood flow distribution problems
551 creases the possibility of accidental spillage of lavage fluid from the lavaged lung to the dependent, ventilated lung. As a compromise, the supine position is used in order to balance the risk of aspiration against the risk of hypoxemia. Following insertion and checking of the doublelumen endotracheal tube and positioning of the patient, baseline total and individual lung compliance should be measured. Airway pressure can be electronically transduced and continuously recorded on a paper write-out, and a Wright spirometer should be placed in the expiratory limb of the anesthesia circle system in order to accurately measure tidal volume. A volume ventilator that can deliver relatively high inflation pressures is required, since these patients have diseased and noncompliant lungs. Prélavage total dynamic compliance (chest wall and lung) of both lungs together (using 15 ml/kg of breath) and then of each lung separately (using 10 ml/kg of breath) is measured. Following measurement of total and individual lung compliance, and with the patient breathing 100 per cent oxygen, baseline arterial blood gases are measured. The patients are completely preoxygenated prior to the induction of anesthesia and lavage for two reasons. First, as with the induction of general anesthesia in any patient, an oxygen-filled FRC greatly minimizes the risk of hypoxemia during the apneic period required for laryngoscopy and endotracheal intubation. This consideration has increased importance for patients with alveolar proteinosis since they are already severely hypoxemic. Second, preoxygenation eliminates nitrogen from the lung that is to be lavaged. Alveolar gas is then composed only of oxygen and carbon dioxide. During fluid filling, these gases will be absorbed, which allows the lavage fluid maximal access to the alveolar space. Failure to remove nitrogen from the lung prior to filling with lavage fluid may leave peripheral nitrogen bubbles in the alveoli and thus limit the effectiveness of the lavage. Warmed isotonic saline is used as the lavage fluid and is infused by gravity from a height of 30 cm above the midaxillary line. After the lavage fluid ceases to flow (i.e., lung filling is complete), drainage is accomplished by clamping the inflow line and unclamping the drainage line, which runs to a collection bottle placed 20 cm below the midaxillary line (Fig. 15-24). The inflow and outflow fluid lines are connected to the appropriate endotracheal tube lumen by a Y-adapter. Each tidal lavage filling is accompanied by mechanical chest percussion and vibration to the lavaged hemithorax prior ^to drainage. The lavage fluid that is drained is typically light brown, and the sediment layers out at the bottom of the collection bottle
552
Anesthesia for Special Elective Therapeutic Procedures
Figure 15-26 Changes in shunt (Qs/Q,[%]) and cardiac output (CO) associated with each of the seven bronchopulmonary lavages. (From Cohen E, Eisenkraft JB: Bronchopulmonary lavage: Effects on oxygenation and hemodynamics. J Cardiothorac Anesth 4:609-615, 1990. Used with permission.)
tension and cardiac output are in opposite directions and have different phase lags, the trend in mixed venous oxygen saturation remains constant (see Fig. 15-25).177 With respect to measuring pulmonary vascular pressures and central venous pressures, the situation is analogous to measuring these intrathoracic vascular pressures during PEEP (absolute intraluminal vascular pressure relative to atmospheric pressure increases, whereas transmural pressure [relative to pleural pressure] remains constant or decreases [see chapters 7 and 20]). An adequate degree of neuromuscular blockade must be maintained because unexpected vigorous coughing during the procedure could alter double-lumen endotracheal tube position. If a small leak should occur during lavage, the following may be observed sequentially: (1) the appearance of bubbles in the lavage fluid draining from the lavaged lung, (2) rales and rhonchi in the ventilated lung, (3) a difference between lavage volumes administered and those drained from the lavaged lung (the former exceeds the latter), and (4) a fall in arterial oxygen saturation. If a small leak is suspected or detected by any of these signs and the lavaged lung has been only minimally treated, the lavaged lung should be drained of all fluid, and the position of the double-lumen endotracheal tube, the adequacy of cuff seal, and separation of the lungs should be rechecked with a fiberoptic bronchoscope. Before beginning the lavage procedure again, and no matter what the double-lumen tube malposition was, the functional separation of the two lungs and adequacy of cuff seal should be tested and found adequate by using the previously described air bubble leak-detection method. Massive spillage of fluid from the lavaged lung to the ventilated lung is not a subtle event and
results in a dramatic decrease in ventilated lung compliance and a rapid and large decrease in arterial oxygen saturation. Under these circumstances, the lavage procedure must be terminated no matter how much treatment has been accomplished. The patient should be moved quickly to the lateral decubitus position with the lavaged side dependent, and the operating room table should be placed in a head-down position in order to facilitate removal of lavage fluid. Vigorous suctioning and inflation of both lungs should be carried out. The doublelumen tube should be changed to a standard single-lumen tube, and the patient should be additionally treated with a period of mechanical ventilatory support with PEEP. Timing of further unilateral lung lavage attempts will be dictated by the patient's subsequent clinical course and gasexchange status. After the effluent lavage fluid becomes clear, the procedure is terminated. The lavaged lung is thoroughly suctioned, and ventilation is begun. Since the compliance of the lavaged side will be much less than that of the ventilated side at this time, large tidal ventilations (sighs) (15 to 20 ml/ kg) to that side alone (with the nonlavaged side temporarily nonventilated) are necessary to re-expand alveoli. Arterial blood oxygenation may decrease precipitously during this time, but this can be minimized by clamping the nonlavaged side after a large inspiration of 100 per cent oxygen. After lavage, the recovery procedure consists of repetitive periods of large tidal ventilations, suctioning and chest wall percussion to the previously lavaged side, conventional two-lung ventilation with PEEP, and bilateral suctioning and postural drainage while intermittently measuring combined (total) and individual lung dynamic compliance. As the compliance of the lavaged lung returns toward prélavage values, ventilation with an air-
oxygen mixture may help lavaged lung alveoli with low ventilation-perfusion ratios to remain open. When the compliance of the hemithorax of the lavaged side returns to its prélavage value, the neuromuscular blockade is reversed. Mechanical ventilation and extubation guidelines are the same as for any patient with pulmonary disease (see chapter 20); most patients are able to be extubated while still in the operating room. If the patient is not considered a candidate for extubation, the double-lumen tube is changed to a single-lumen tube, and the patient is mechanically ventilated with PEEP in a conventional manner. In the immediate postlavage period, deep breathing (incentive spirometry), coughing exercises, chest percussion, and postural drainage are used to remove remaining fluid and secretions and to re-expand the lavaged lung. After 3 to 5 days of recovery, the patient is returned to the operating room to have the opposite side lavaged. The anesthetic considerations for the second lavage are the same as those for the first lavage, although oxygenation is usually not nearly as severe a problem as during the first lavage because the treated and now near-normal lung will be used to support gas exchange. There are two special problems associated with pulmonary lavage that may be encountered. First, a few critically ill patients may be unable to tolerate the conventional procedure. Second, unilateral lavage through a double-lumen tube is not possible in children and small adults. There are three alternative (and more complicated) ways of accomplishing lung lavage in patients who simply cannot tolerate one-lung ventilation under any circumstance. First, extracorporeal membrane oxygenation (ECMO) has been utilized to provide support of gas exchange during standard unilateral lung lavage. Partial venoarterial cardiopulmonary by-pass has been used for a few hours during unilateral or bilateral lung lavage,179-181 but the distribution of oxygenated blood from the venoarterial by-pass can be markedly nonhomogeneous and dependent on the site of blood return182·l83 and requires major arterial cannulation. Venovenous by-pass allows uniform arterial distribution and the safety of not requiring major arterial cannulation.184· I8S In one of these latter cases,185 partial venovenous (femoral vein—inferior vena cava to internal jugular vein—superior vena cava connection) cardiopulmonary by-pass with a membrane oxygenator was especially successfully used to exchange respiratory gases during and after bilateral lung lavage in a patient with severe hypoxemia due to alveolar proteinosis (the patient had previously been unable to tolerate otherwise conventional unilateral lung lavage). During by-pass, the lungs were mechanically venti-
554
Anesthesia for Special Elective Therapeutic Procedures
as a second try in patients who had unacceptable oxygenation during lung drainage on a first try. Lavage in the conventional manner is not possible in children (or small adults) in whom doublelumen endotracheal tubes are too large to be inserted. This problem routinely occurs in persons weighing less than 25 to 30 kg, since the smallest double-lumen endotracheal tube made is 28 French, with each lumen being slightly less than 4.5 mm. In this situation, partial cardiopulmonary by-pass has been successfully used to provide oxygenation during unilateral or bilateral lavage.179· 189. 190 j n o n e 0 f m e s e reports, the technique was used in two brothers aged 4 and 2.5 years.189 Both these patients underwent whole-lung lavage, during which time blood was removed from both femoral veins, oxygenated, and then returned to the left femoral artery. Both patients were eventually discharged from the hospital, although they continued to require supplemental oxygen by face mask. In another report, the technique was used to support gas exchange during whole-lung lavage in a ventilator-dependent, 3.7-kg, 8-month-old child.190 The extracorporeal oxygenation system was again venoarterial (right internal jugular—right atrium catheter to right axillary artery catheter). Marked improvement in pulmonary function was noted after lavage (total 420 ml/kg), by-pass was able to be discontinued 3 hours after lavage, and the patient was extubated 48 hours following lavage. In the last report,189 partial venoarterial by-pass with a bubble oxygenator permitted bilateral simultaneous lung lavage in two siblings aged 3 and 4 years. By-pass in these patients was carried out with femoral vein and femoral artery cannulation. During by-pass, radial artery P a 0 2 ranged between 25 and 30 mm Hg, which may, in part, have been related to continued cardiac output of desaturated blood into the proximal aorta. No neurologic sequelae were noted. Oxygenation and functional levels were improved following lavage. These various efforts to treat pulmonary alveolar proteinosis in children with bilateral lung lavage support by extracorporeal oxygenation must be regarded as successful, considering that in children the average survival from the time of severe symptoms without this kind of treatment is less than 1 year. Finally, one report describes a 7-year-old patient who was lavaged via a single-lumen tube placed endobronchially with a fiberoptic bronchoscope, while the nonlavaged lung was ventilated from above via an anesthesia mask or nasopharyngeal airway (the mouth was sealed with a transparent dressing [Opsite]).191 In this case, a 3.5-mm flexible bronchoscope was passed through a 4.5-mm ID, cuffed endobronchial tube 50 cm long. The bronchoscope and tube were passed through a bronchoscope adapter attached to a tight-fitting an-
esthesia mask. Using the bronchoscope, the endobronchial tube was guided into the right main-stem bronchus with its tip just above the right middlelobe bronchus. The bronchoscope was then withdrawn. In theory, placement of the endobronchial catheter in the right main-stem bronchus with its balloon cuff below the right upper lobe bronchus allowed ventilation of the right upper lobe as well as the entire left lung. While the patient was ventilated through the mask, the right middle and lower lobes were lavaged with 250- to 400-ml aliquots of saline to a total of 8 L, using gravity to fill and drain the lung. During this ventilation procedure/setup, the stomach may gradually become distended, and an orogastic tube should be passed for gastric decompression. The success of this technique depends on several factors. First, the endobronchial catheter must be positioned accurately; the flexible bronchoscope greatly facilitates this. The catheter must remain fixed in the proper position; this is facilitated by neuromuscular blockade. Vigilance in monitoring the breath sounds and the position of the tube is necessary because if the tube were to slip proximally, or the balloon cuff failed, fluid could flood both lungs. A second requirement is that the endobronchial catheter must be large enough to allow effective lavage yet small enough to allow effective ventilation around it. The geometry of a cylinder within a cylinder is such that there is much more cross-sectional area available for ventilation around a single tube than would be available through two parallel tubes (Fig. 1527).I91 Figure 15-27 shows an example of the considerations involved in the selection of the size of the endobronchial lavage catheter.
VIII. PULMONARY ARTERIOVENOUS MALFORMATIONS192 A. General and Surgical Considerations Although most cases of pulmonary arteriovenous malformations (including aneurysms) are congenital, the condition is often not recognized until the second decade of life.192 It has been suggested that the aneurysm progressively enlarges with age in response to increasing flow, with eventual necrosis of the vessel wall. This may suddenly increase the magnitude of the right-to-left shunt or cause hemorrhage, which then leads to clinical detection. There are four presentations that appear to be sufficient to raise the suspicion of the diagnosis even when they are found in isolation: hemoptysis, abnormalities on chest roentgenogram that are
Anesthesia for Special Elective Therapeutic Procedures 5 5 5 Figure 15-27 Geometry of a tube within the trachea. The tracheal diameter is 8 mm, and that of the 4.5-mm internal diameter endobronchial tube is 5.5 mm. The cross-sectional area of the endobronchial tube lumen (A) is 15.9 mm2, whereas the crosssectional area of the trachea remaining for ventilation around the tube (B) is 26.3 mm2. If the patient were to be ventilated through the 4.5-mm tube and a second tube (for lavage) were passed alongside, thé largest internal cross-sectional area (Q the trachea could accommodate would be 1.76 mm2, thus wasting 21.4 mm2 of area (#') potentially available for ventilation. (From MacKenzie B, Wood RE, Bailey A: Airway management for unilateral lung lavage in children. Anesthesiology 70:550-553, 1989. Used with permission.) •Ë|tKTr.itvv
consistent with pulmonary arteriovenous malformation, evidence of a right-to-left shunt, and central nervous system phenomena attributed to infected or noninfected emboli without an obvious source.193 The association between pulmonary arteriovenous malformations and Osler-WeberRendu disease is well known and pulmonary arteriovenous formation occurs most commonly in patients with the Osier-Weber-Rendu disease. The presence of symptoms and abnormal physical aigris is correlated with the size of the pulmonary arteriovenous malformation and not with the number of lesions. Because the essential defect is right-to-left shunting from the pulmonary artery to pulmonary vein, it is not surprising that arterial oxygen tension is reduced in more than 80 per cent of cases, and shunt magnitudes of as much as 79 per cent of the cardiac output have been recorded.194 In the presence of significant shunting and arterial desaturation, a secondary polycythemia results. In general, hemodynamic variables including intracardiac, pulmonary arterial, and pulmonary artery occluded pressures and cardiac output are normal. ! Although the chest radiograph is abnormal in approximately 98 per cent of patients with pulmonary arteriovenous malformations (a single peripheral, circumscribed, noncalcified lesion that is connected by blood vessels to the hilus of the lung is the most common finding), pulmonary angiography remains the diagnostic standard for these lesions. In surgical candidates, angiography is mandatory to establish the number, extent, and location of the lesions and to delineate the arterial supply and venous drainage. I Therapeutic options include surgery, embolization, or regular follow-up without invasive treatment in those patients with small lesions that are clinically silent. However, and most important, the natural history of pulmonary arteriovenous malformations is far from benign, and treatment by embolization or surgical incision should be considered in all cases. In other words, the complications
from the disease are much greater than the complications from the treatment. When an aneurysm is distal to the main pulmonary trunk, resection becomes a therapeutic option. In the past, resection was often the only therapeutic option for these lesions.193 Some surgeons have opted for pneumonectomy as the intervention for aneurysms of a main pulmonary artery or multiple more peripheral aneurysms.193 Peripheral aneurysms have also been treated with lobectomy or with resections of lesser amounts of pulmonary parenchyma.193 Although such resections can be performed with low morbidity/mortality, surgery has limitations as a therapeutic approach in patients with limited pulmonary function or in tenuous medical condition. Over the last few years, embolotherapy (percutaneous localization and occlusion of vascular lesions) under local anesthesia has become the treatment of choice for pulmonary arteriovenous malformation not involving the pulmonary trunk or main pulmonary arteries. The dual circulation of the lungs allows interruption of pulmonary arteries with minimal necrosis. This fact, coupled with advances in interventional radiology, has dramatically altered the approach to pulmonary arteriovenous malformation located distal to the main pulmonary arteries over the past few years. Embolectomy can now be performed with minimal loss of lung tissue and minimal morbidity, and no mortality has been reported to date.193
B. Anesthetic Considerations There are three main anesthetic considerations. First, lung separation is necessary for two reasons. As with all major resections, lung exposure is facilitated by collapse of the operative lung. In addition, in these cases, the surgeon will need to gain control of the vasculature by gaining control of the hilar vessels. Thus, whether a pneumonectomy is
556
Anesthesia for Special Elective Therapeutic Procedures
being performed or not, collapse of the operative lung will facilitate exposure of the hilar structures. Second, multiple large-bore intravenous routes of access are necessary because of the potential for hemorrhage with these cases. In addition, an index of left ventricle preload is desirable (i.e., central venous pressure and/or pulmonary artery occluded pressure). Third, hypoxemia may be profound in those pa tients with a high degree of right-to-left shunting. Hypoxemia may be minimized by using an in spired oxygen concentration of 100 per cent and operative lung CPAP. However, early control of hilar vessels is the most definitive treatment of shunting through the operative lung in these cases. Blood gases may need to be measured from pe ripheral arteries as well as from various pulmonary veins to prove that the resection of the pulmonary arteriovenous malformation has been complete.195
IX. LUNG TRANSPLANTATION A. General and Surgical Considerations 1. Types of Lung Transplantation and Their Indications Four types of lung transplantation are currently being performed. Table 15-10 lists the four types, the indications for each type (by category of disease as well as by specific examples/problems),196-207 whether cardiopulmonary by-pass is required, and the major limitations to each type of transplanta tion or technique. The first operation, single-lung transplantation with bronchial anastomosis to the transplanted lung, is widely used when the remain ing native lung is not infected and cardiac function is not a problem. Chronic idiopathic pulmonary interstitial fibrosis, emphysema (especially a,-anti trypsin deficiency), and pulmonary hypertension with satisfactory right ventricular function are three specific examples of this general category of patient.200 Idiopathic pulmonary fibrosis is associ ated with a 40 to 80 per cent mortality within 5 years of diagnosis.208 Severe life-threatening em physema may occur in the fourth and fifth decades of life in individuals with a,-antitrypsin deficiency of the homozygous type Ζ variant. The prevalence of homozygous type Ζ is estimated to range from 1 in 1600 to 1 in 5000, but only a minority, almost always smokers, present with end-stage obstruc tive respiratory disease.209 Deficiency of a,-pro tease inhibitors is associated with a lack of protec tion against neutrophil elastase in the lower respiratory tract. Transplantation of a normal lung in juxtaposition to a lung with either pulmonary fibrosis206 or obstructive airways disease207 results
in the majority of ventilation and perfusion divert ing to the new lung, resulting in good V/Q match and excellent oxygenation.206,207 The natural history of pulmonary hypertension is quite variable, but the typical mean survival is 2 to 3 years from the time of diagnosis.210 The course of this disease is one of progressive rightsided heart failure that ultimately results in the patient's death. The major issue in assessing pa tients with pulmonary hypertension for transplan tation has been the variability in the natural history of the disorder. In some or many patients with pulmonary hypertension and reasonable right ven tricular function, right ventricular and pulmonary vascular function has significantly improved after the insertion of a new low-load pulmonary circuit in the form of a new single-lung transplant.198,204-206 For example, in one report, mean pulmonary artery pressure decreased from 71 to 15 mm Hg and mean right ventricular ejection fraction increased from 26 to 41 per cent.205 However, the singlelung transplant for pulmonary hypertension does not completely relieve hypoxemia because vir tually all blood flow is diverted to the new trans planted lung, whereas ventilation is equally split between the native and transplanted lung, both of which have a normal compliance.206 Patients with these three disorders are otherwise excellent lung-transplant candidates because of their relatively young age and lack of other organ system involvement. The second type of lung transplantation, bilat eral, sequential single-lung transplantation with bi lateral bronchial anastomosis to each transplanted lung, is essentially equivalent to two sequential single-lung transplantations.193,199,211,212 This oper ation is indicated when both lungs are infected and therefore must be removed (if one of the two in fected lungs remained, it would infect the newly transplanted lung; indeed, foci of infection in a remaining lung will likely be exacerbated by necessary immunotherapy). Examples of patients whose main disability is secondary to conditions that result in chronic sepsis are cystic fibrosis and end-stage emphysema/bronchitis. Indeed, more than 98 per cent of patients with cystic fibrosis die from pulmonary-related complications by the third to fourth decade of life.213 The third type of lung transplantation, double lung with tracheal anastomosis, is most often per formed on full cardiopulmonary by-pass. The in dications for this operation are rapidly changing at the time of this writing, but the operation has been done for virtually every disease that might benefit from lung transplantation; however, because of the disappointment with the healing of the tracheal anastomosis and the satisfaction with the healing of the bronchial anastomosis of the single-lung
Anesthesia for Special Elective Therapeutic Procedures
557
Table 15-10 TYPES OF LUNG TRANSPLANTATION AND THEIR INDICATIONS, WHETHER CPB IS UTILIZED, AND THE MAJOR LIMITATIONS TO THE TRANSPLANTATION TECHNIQUE* Type of Transplantation Single lung, bronchial anastomosis
Category of Problem
Chronic interstitial fibrosis, emphysema (all types but especially a,antitrypsin deficiency), pulmonary hypertension with satisfactory RV function (new application) Terminal lung function, Cystic fibrosis, endBilateral sequential bilateral infection, stage infection, and single lung, bronchial good cardiac function emphysema anastomosis All of the above All of the above Double lung, tracheal anastomosis
Double-lung plus heart
Terminal lung function, no infection, satisfactory cardiac function
Terminal lung and cardiac function
CPB
Specific Examples
Major Limitations to Transplantation Technique
No, partial femoral artery—femoral vein standby
1. Requires patient to tolerate 1LV 2. Requires patient to tolerate unilateral pulmonary artery clamping
No, partial femoral artery—femoral vein standby Yes
Same as above (single lung)
Severe pulmonary Yes hypertension with right ventricular failure, Eisenmenger's syndrome
1. High incidence (approximately 50%) of poor tracheal healing 2. Hemorrhage caused by heparinization for CPB 1. Hemorrhage resulting from heparinization for CPB 2. Separate independent risk of heart rejection 3. High incidence of advanced coronary artery sclerosis 4. High incidence of obliterative bronchiolitis (chronic rejection)
*Based on data from Raju et al.,196 Bisson & Bennette,197 Kaiser & Cooper,198 Kaiser et al.,199,20° Calhoun et al.,2 Cooper, Levine et al.,204 Starnes et al.,205 Kramer et al.,206 and Yacoub et al.207 Abbreviations: CPB = cardiopulmonary by-pass; RV = right ventricular; 1LV = one-lung ventilation.
transplant, the double-lung procedure is falling out of favor.199· 200· 202 The tracheal anastomosis does poorly because it does not have the mediastinal collaterals (particularly from the coronary arteries)214 that the double-lung plus heart transplant does. The fourth type, combined double-lung and heart transplantation, is used for patients with both very severe/terminal cardiac and lung disease. The lung disease may be primary (e.g., primary pulmonary hypertension) and cardiac disease secondary (e.g., right ventricular failure), or the cardiac disease may be primary (Eisenmenger's syndrome with septal defect) and the lung disease secondary (e.g., marked increases in pulmonary artery pressure). Unfortunately, this operation has the intraoperative complications related to cardiopulmonary by-pass as well as the postoperative complications of rejection episodes of both the heart and the lung (which may occur completely
independently of one another),215 sclerosis of the transplanted coronary arteries,203 and late obliterative bronchiolitis (resulting from a chronic state of rejection).203 The third and fourth lung transplantation techniques (double lung and double lung plus heart) are not discussed here because they are outside the scope of this textbook (i.e., they require cardiopulmonary by-pass). The overall results of lung transplantation are very good. Aside from intraoperative problems (see major limitation to transplantation techniques in Table 15-10 and anesthetic considerations later), the major immediate/subacute concern is the healing of the airway anastomosis. Lung allografts are unique among solid organ transplants in that their systemic bronchial arterial supply is not reconstructed at the time of transplantation, and blood supply to the anastomosis depends on the development of retrograde collateral flow from the transplanted pulmonary artery.216· 217 Although
558
Anesthesia for Special Elective Therapeutic Procedures
omentopexy (omental wrap) provides immediate support/augmentation of collateral blood flow to the bronchial anastomosis (and therefore good healing), omentopexy has not had a great impact on the healing of tracheal anastomosis.217 Nevertheless, the overall 1-year survival worldwide for all types of lung transplantations is 60 per cent, although individual centers report survival of 75 per cent.218 Although the lungs remain deinnervated after transplantation, there are no long-term differences in patterns of breathing in recipients either when awake or asleep compared with that of normal individuals. The absence of a peripheral lung cough reflex has not proven to be problematic (the patient can still consciously cough). Rehabilitation of patients after successful surgery is excellent with restoration of a normal lifestyle with little or no functional restriction. Although reoccurrence of disease in transplanted lungs remains a worry, there is to date no published evidence to suggest that it has caused clinical problems for all the diseases listed in Table 15-10. 2. Patient Evaluation and Selection Criteria It is obvious that candidates for lung transplantation should have end-stage disease with significant functional impairment that interferes with activities of daily living. Most of these patients, with the exception of the pulmonary hypertensive, require oxygen 24 hours a day. To support the need for transplantation further, there should be documented evidence of disease progression with an anticipated life expectancy of 12 to 18 months or less without transplantation intervention. Potential candidates need to be extremely well motivated to cope with the stresses associated with the preoperative and postoperative periods and with the lifelong care required after transplantation. The complete selection criteria of one institution are summarized in Table 15-11. 198 One of the critical physiologic insults of singlelung and bilateral sequential single-lung transplants into the recipient is the clamping of the pulmonary artery before pneumonectomy. This can lead to marked increases in pulmonary arterial pressure and may signal or indicate the onset of pulmonary edema and right-sided heart failure. The factor on which the transplant operation hinges, therefore, is the ability of the recipient's right ventricle to maintain sufficient cardiac output in spite of the acute increase in pulmonary vascular resistance. Thus\ it is necessary to obtain some idea of right ventricular function before operation (see chapter 5). Several groups have used right ventricular ejection fraction, obtained from a mul-
Table 15-11 SELECTION CRITERIA FOR LUNG-TRANSPLANT RECIPIENTS* I. End-stage lung disease with evidence of progression of disease and life expectancy 250 mm Hg on F,0 2 l.O, positive end-expiratory pressure 5 cm H-,0 for 5 minutes No evidence of purulent secretions or aspirated gastrointestinal contents seen at bronchoscopy No significant chest trauma or pulmonary contusion No previous major thoracic surgery Smoking history? *From Kaiser LR, Cooper JD: The current status of lung transplantation. Adv Surg 25:259-307, 1992. Used with per mission.
scarcity of suitable lungs for implantation. Proba bly only 5 to 10 per cent of available donors have lungs that currently are considered acceptable for transplantation, even when the heart, liver, kidney, and pancreas are suitable. Of these, it is estimated that organs are retrieved from only 20 to 30 per cent of these potential organ donors because of religious, social, or other reasons. I98 Suitable donors are young patients (younger than 40 years) diagnosed as brain dead and desig nated as potential organ donors, who have had a short period of ventilatory support, without postcatastrophe pulmonary injury or barotrauma, and who do not have a history of significant pre-exist ing lung disease. The criteria for the acceptance of donor lungs are, by design, somewhat rigid,221 and are summarized in Table 15-12. 198 The issue of the smoking history of a potential donor is a diffi cult one. Lungs from nonsmokers are preferred, but the reality of donor supply does not often al low this luxury. In general, a donor lung may be accepted if the smoking history is less than 10 198 pack-years. In general, an attempt is made to provide a prospective recipient with donor lungs that would be an appropriate size for that individual in the absence of their lung disease. For patients with pulmonary fibrosis, a larger lung than their native lung is chosen because considerable mediastinal shift occurs, the diaphragm lowers, and the thorax expands to accommodate the new normal lung. In contrast, patients with chronic obstructive pulmo nary disease have an overexpanded chest cavity that will decrease in size when air trapping is no longer a clinical problem; hence, lungs that are smaller, by as much as 5 to 10 cm in vertical height and 5 cm in horizontal chest diameter, have been used.198 As with other organs transplants, lung transplantation requires ABO compatibility. Guidelines for the management of the multipleorgan donor are summarized in Table 15-13. 198 The technical aspects of donor pneumonectomy are mainly surgical considerations. The lung is removed intact with as long a piece of main bron
559
chus and of main pulmonary artery as possible. The pulmonary veins are removed from the heart with a cuff of left atrium (Fig. 15-28).201 Techni cally, the left lung is easier to extract and reimplant (see Fig. 15-28). For double-lung trans plants, the donor operation is similar, except the pulmonary veins from both lungs share a common atrial cuff and the trachea is removed (Fig. 1529) 196 Yhe double-lung block is easily split for the bilateral sequential single-lung transplant proce dure.' 99 With current preservation techniques, total do nor ischemic time should be kept under 6 hours. This constraint currently limits flying time be tween donor and recipient hospitals to 2Vi hours. The lung may be preserved by flushing cold elec trolyte solution through the pulmonary artery to affect rapid cooling and gain a degree of cytoprotection. To allow more uniform distribution of the perfusate, the lungs are ventilated during the pul monary artery flush. The specimen is then im mersed in cold solution and packed for transport. Table 15-14 provides a complete listing of lungpreservation methods used to date. 198 It is currently believed that injury to donor lung results not only from ischemic insult but also from injury occur ring at the time of reperfusion of blood through the preserved organ. Several experimental models of acute lung injury implicate oxygen free radicals in the genesis of reperfusion injury. 222 · 223 A variety of mediators, including prostaglandin metabolites.
Table 15-13 GUIDELINES FOR MANAGEMENT OF MULITPLE-ORGAN DONORS* Maintain mean arterial blood pressure > 70 mm Hg Central venous pressure not to exceed 10 cm H : 0 and/or pulmonary capillary wedge pressure not to exceed 12 mm Hg Vasopressor support to maintain blood pressure with either dopamine 2.5-10 μg/kg/min or phenylephrine 0.06-0.18 mg/min Treat diabetes inspidus with vasopression (5-10 units every 8 hours intravenously) or DDAVP (0.3 μg/kg intravenously over 30 minutes) Fluid replacement—previous hour's urine output minus 100 ml Maintain urine output at 1-2 ml/kg/hr Replace electrolyte losses Maintain normothermia (35° C-37° C) PaG\ > 100 on lowest F,0 : possible Maintain 5 cm H.,0 positive end-expiratory pressure Frequent tracheobronchial suctioning; specimen for Gram's stain Arterial blood gases every 2 hours Current chest radiograph *From Kaiser LR, Cooper JD: The current status of lung transplantation. Adv Surg 25:259-307. 1992. Used with per mission. Abbreviation: DDAVP = l-deamino-8-D-arginine vaso pressin.
560
Anesthesia for Special Elective Therapeutic Procedures
LMB
Figure 15-28 Donor lung with the pulmonary veins (PV), pulmonary artery (PA), and the stapled left main bronchus (LMB). The lung is triple-bagged in a sterile plastic container with iced saline slush for transport. (From Calhoun JH, Groover FL, Gibbons WJ, et al: Single lung transplantation: Alternative indications and technique. J Thorac Cardiovasc Surg 101:816-824, 1991. Used with permission.)
Figure 15-29 Donor double lung with retained pleuropericardial flaps and atrial cuff around the^ pulmonary veins. (From Raju S, Heath BJ, Warren ET, Hardy JD: Single and double-lung transplantation. Ann Surg 211:681-692, 1990. Used with permission.)
Anesthesia for Special Elective Therapeutic Procedures
Table 15-14 METHODS OF LUNG PRESERVATION* I. Hypothermic immersion A. With atelectasis B. With inflation C. With ventilation D. With hyperbaric oxygen II. Normothermic perfusion A. Autoperfused heart-lung preparation III. Hypothermic perfusion A. Core cooling, with or without perfusion IV. Flushing A. Extracellular fluid solutions 1. Low potassium dextran 2. Fujimura 3. Ringer's, saline 4. Plasma, blood 5. Rheomacrodex B. Intracellular fluid solutions 1. Collins 2. Sacks V. Pharmacologic additives A. Prostaglandins 1. PGI2 B. Calcium channel blockers C. Free radical scavengers 1. Superoxide dismutase 2. Catalase 3. Glutathione 4. Dimethyl sulfoxide *From Kaiser LR. Cooper JD: The current status of lung transplantation. Adv Surg 25:259-307, 1992. Used with permission.
play a role in the vascular and bronchial smoothmuscle response to injury in the lung. It is likely that some of these mediators are also important in the response of the lung to ischemia. 4. Surgical Operation The surgical technique for single-lung transplantation, although requiring great attention to numerous details, is still relatively straightforward. With all considerations being equal, most surgeons chose the technically easier left side to transplant. However, the side to transplant may depend on several other factors. If the recipient has had a major thoracotomy or previous pleurodesis, it is far easier to transplant the opposite side. If a significant proportion of arterial perfusion, as measured by a radionuclide lung perfusion scan, goes to one side, the other side should be transplanted, which allows some additional protection during the early postoperative period when the function of the transplanted lung may be somewhat compromised. Occasionally, the side for transplantation is chosen based on donor considerations such as a unilateral infiltrate seen on the donor chest radiograph or the availability of only one lung when the other was being used by another center.
561
The recipient is subjected to three surgical procedures in sequence, one of which (the second described later) is now only occasionally performed (optional) and is done only for safety reasons. The first procedure, with the patient supine, is a small midline laparotomy through which the omentum is mobilized. With its vascular pedicle, part of the omentum is pushed up through the diaphragm into the thorax behind the xiphisterum from where it can be retrieved and wrapped around the donor-recipient bronchial or tracheal anastomosis (Fig. 15-30).' 96 The omental wrap provides early vascularization to the airway suture line. However, it should be noted that the problem of healing of the bronchial anastomosis has been solved in one institution by use of nonabsorbable prolene sutures and a telescoping anastomosis (instead of using an omental wrap and thereby avoiding an abdominal incision).201 Second, in some high-risk patients, after closure of the abdomen, a femoral vein and artery can be exposed and the skin closed to facilitate the speedy implementation of femoral artery to femoral vein by-pass should an untoward event occur. More commonly, the groin area is just simply prepared and left exposed in case rapid access is needed for the institution of partial cardiopulmonary by-pass. If the by-pass is established, it may be used for rewarming of the patient at the end of the procedure. 2 " Third, in the case of a single-lung transplant, the patient is then repositioned for a posterolateral thoracotomy and a pneumonectomy is carried out. In the case of bilateral sequential single-lung transplantation, the patient is left supine and a bilateral fourth interspace anterior thoracotomy and sternotomy are performed. The pulmonary artery and veins are divided as far distally from the hilum as feasible, and for single-lung transplants the bronchus is divided near the origin of the upper lobe.196 Connection of donor lung to recipient is in sequence: pulmonary veins, on their atrial pedicle, to a zone of recipient left atrium isolated in a vascular clamp; pulmonary artery anastomosis; and finally, the bronchial anastomosis (Figs. 15-31 and 15-32) and omental wrap (see Fig. 15-30).224 The bilateral sequential single-lung transplantation operation is essentially the same as two single-lung transplants in series.197 Double-lung and combined heart and lung transplantation with cardiopulmonary by-pass are beyond the scope of this textbook and are not discussed here. B. Anesthetic Considerations 1. Anesthetic Drugs In practice, single-lung transplantation is a pneumonectomy in a patient who, under normal
562
Anesthesia for Special Elective Therapeutic Procedures
•»"îr ι Μ ν } »
Omental Patch
Figure 15-30 Omental wrap around tracheal su ture line. (From Raju S, Heath BJ, Warren ET, Hardy JD: Single- and double-lung transplantation. Ann Surg 211:681-692, 1990. Used with permission.)
Figure 15-31 Single left-lung transplant with anastomosis of the donor pulmonary veins (PV) to the recipient left atrium (LA). (PA = pulmonary artery; LMB = left main bronchus.) (From Calhoun JH, Groover FL, Gibbons WJ, et al: Single lung transplantation: Alternative indications and technique. J Thorac Cardiovasc Surg 101:816-824, 1991. Used with permission.)
Anesthesia for Special Elective Therapeutic Procedures
563
Figure 15-32 Completion of a single left-lung transplant with a running 4-0 Prolene suture of the pulmonary artery (PA). (PV = pulmonary vein; LMB = left main bronchus.) From Calhoun JH, Groover FL, Gibbons WJ, et al: Single lung transplantation: Alternative indications and technique. J Thorac Cardiovasc Surg 101:816-824, 1991. Used with permission.)
circumstances, would be judged unfit for such an operation. Scrupulous attention must be paid to asepsis, and premedication should be restricted to those drugs that decrease airway secretions rather than those with analgesic or sedative properties (to which recipients are likely to be very sensitive and thereby hypoventilate). Drugs selected for induction, and for the provision of analgesia and neuromuscular blockade, should be neutral in action on the cardiac and respiratory systems. This means that the usual anesthetics used for patients undergoing cardiac surgery such as high-dose synthetic narcotics, perhaps combined with low doses of halogenated drugs as well as nitrous oxide, and vecuronium for neuromuscular blockade are perfectly acceptable.
2. Lung Separation As with any pulmonary operation, the procedure is made easier if the airway is divided and a clear surgical area for operation is produced. There have been many reports of anesthesia for lung transplantation; some institutions use double-lumen tubes, whereas others use bronchial blockers. Both lung-separation techniques have been used with
great success, but, in terms of independent lung suctioning and rapid independent lung inflation and deflation, use of a double-lumen tube is more versatile. If a left-sided double-lumen tube is used for a left lung transplant, then it is important to be sure, with fiberoptic bronchoscopy, that the left endobronchial cuff is as proximal as possible to allow mobilization and anastomosis of the bronchus to the transplanted lung.225·226
3. Monitoring The monitoring must be comprehensive as it must be in any major operation, but use of capnography and pulmonary artery catheter monitoring deserves special comment. Capnography is important in two respects. First, as noted in chapter 7, the P E T C 0 2 tracing may demonstrate a considerable slope to the phase III plateau in patients with airway obstruction and emphysema. If the expiratory time is relatively short (i.e., less than the time needed to reach a true P E T C 0 2 in emphysematous lungs), then a large and inconsistent P a C0 2 P E T C 0 2 gradient may be present throughout the anesthetic. Therefore, during the anesthetic, intermittent arterial blood-gas analysis may be required
564
Anesthesia for Special Elective Therapeutic Procedures
to optimize ventilation.211 Second, but related to the out-of-phase ventilation of the native lung (high airway resistance, low compliance) and the donor lung (low airway resistance, normal compliance), is the occurrence of a biphasic capnogram.227 Biphasic capnogram waveform secondary to severe kyphosis, scoliosis,228 and main-stem bronchial intubation229 has been observed previously in patients in whom the pattern of expiration, as determined by alveolar time constants, was distinctly different in each lung. Similarily, single-lung transplantations may also demonstrate a biphasic capnograph waveform that is produced by two different populations of alveoli (one in the remaining emphysematous lung and one in the newly transplanted lung) (Fig. 1533).227 The first peak represents expired C0 2 from the allograft, which has normal compliance, good perfusion, and normal ventilation-perfusion ratios (V/Q). The second peak most likely reflects expired C 0 2 from the native lung, because the slanted upstroke is characteristic of the mismatched V/Q ratios and differing alveolar time constants in emphysema. Independent of the perfusion characteristics of the two lungs, the differences in their resistances and compliance alone could account for the observed biphasic capnogram, with the low-resistance transplanted lung emptying more rapidly and the more compliant native lung emptying last. After an increase in respiratory rate, the capnogram will show only a single curve (Fig. 15-34).227 This probably represents preferential exhalation of C 0 2 from the transplanted allograft, with minimal contribution of expired C 0 2 from the native emphysematous lung.227 A pulmonary artery catheter, situated in the main pulmonary artery, is probably mandatory. During perfusion of just one lung, the wedge pressure should be interpreted with caution because there is some evidence that these results may be spurious when measured in patients who have had
a pneumonectomy (the inflated balloon can block blood flow to the one remaining lung, resulting in an erroneously low value).230 The index of particular acute interest is actually the pulmonary artery pressure because it is a measure of the load on the right ventricle. The use of intraoperative transesophageal echocardiography also permits continuous assessment of right ventricular function.199 The wedge pressure is of less relevance initially, provided that left ventricular function is known to be adequate. Later, during recovery, the pulmonary artery occluded pressure may be more important, as may the ability to measure cardiac output by thermodilution or other indicator technique.
4. Major Intraoperative Anesthetic Problems Potentially, and predictably, major problems may occur when elements of the physiology of lung ventilation and pulmonary blood flow are radically altered by operative maneuvers. In practice, and in the usual temporal sequence, the significant problem periods are the institution of onelung ventilation, the clamping of the pulmonary artery before pneumonectomy, and the time period, if allowed, during which the donor lung is perfused but not ventilated. The detectable changes for these three events in terms of monitored data are shown in Table 15-15, in addition to suggested directions for corrective therapy.219 a. INITIATION OF ONE-LUNG VENTILATION
With respect to the initiation of one-lung ventilation, severe hypoxemia and hypercapnia may occur not only because of the large shunt created through the nonventilated lung, but also because the ventilated lung may not support ventilation as well as a normal lung might. Obvious and simple therapeutic maneuvers consist of increasing the inspirated oxygen concentration to 100. per cent (if
Figure 15-33 C0 2 tracings at a respiratory rate of eight breaths per min and a volume of 700 ml. Left, The trend of ETco2 ove the previous 20 min. Right, Biphasic capnogram at an ETco 2 of 31 mm Hg and a graph speed of 0.5 cm/s. (From Williams L Jellish WS, Modica PA, et al: Capnography in a patient after single lung transplantation. Anesthesiology 74:621-622, 1991. Use with permission.)
Anesthesia for Special Elective Therapeutic Procedures
565
Figure 15-34 CO, tracings at a respiratory rate of 13 breaths per min and a tidal volume of 700 ml. Left, The 20-min trend of ETco : , which has been decreasing for approximately 10 min. Right, A normal-appearing capnogram at an ETco : of 25 mm Hg and a graph speed of 0.5 cm/s. (From Williams L, Jellish WS, Modica PA, et al: Capnography in a patient after single lung transplantation. Anesthesiology 74:621-622, 1991. Used with permission.)
F,0 2 = 1.0 was not in use before the initiation of one-lung ventilation) and increasing/changing the minute volume of ventilation (change tidal volume and/or respiratory rate). In general, these conven tional maneuvers are adequate to obtain acceptable gas exchange even during ventilation of one se verely emphysematous lung.231 If these simple conventional maneuvers fail, a search for the ap propriate nonventilated lung CPAP and ventilated lung PEEP levels should be quickly instituted (chapter 11). In the event of supervening circula tory deterioration (severe pulmonary hypertension and/or systemic hypotension) or failure of ventila tion (hypoxemia and/or hypercapnia) despite these corrective maneuvers, partial femoral artery to femoral vein cardiopulmonary by-pass may be needed. The problem of hypercapnia and hypox emia during one-lung ventilation should be de creased when the ipsilateral pulmonary artery is ligated.
Table 15--15
b. CLAMPING OF THE PULMONARY ARTERY With respect to pulmonary artery clamping, the crux of the operation is the ability of the right ventricle to withstand the acute increase in pul monary artery pressure induced by clamping the pulmonary artery on the side of the pneumonec tomy. Hence, the preoperative assessment of right ventricle function is emphasized. The increase is sudden, but not usually sustained, and the pulmo nary vascular resistance in the perfused lung usu ally decreases after several minutes so that some of the strain is taken off the right ventricle. In the event of right ventricular failure, inotropic support may be required. For excessive increases in pul monary artery pressure, drugs with pulmonary vasodilatory activity are indicated: Nitroglycerin, nitroprusside, and hydralazine are drugs that have been successfully used in this context. 2 "· 2 1 9 · 2 2 5 In addition, inotropic support may be necessary (e.g., dopamine, 5-10 μg/kg/min, dobutamine, 5-15 μg/
OPERATIVE EVENTS AND THEIR MANAGEMENT*
Event Onset of one-lung anesthesia
Parameter P a 0 2 decreases P,CO^ increases
Pulmonary artery clamped
Pulmonary artery pressure
Donor lung perfused but not ventilated
P.O, decreases
Therapy Increase inspired 0 2 (if not using F,0 2 = l.O) Increase minute volume Recruit nonventilated lung CPB Inotropes Vasodilators Diuretics CPB Pulmonary artery clamped: Partial Total Low-disturbance lung ventilation
Problem eliminated when > ipsilateral pulmonary artery is i gated
*From Conacher ID: Isolated lung transplantation: A review of problems and guide to anaesthesia. Br J Anaesth 61:468-474, 1988. Used with permission. Abbreviations: CPB = partial (femorofemoral) cardiopulmonary by-pass.
566
Anesthesia for Special Elective Therapeutic Procedures
kg/min). 219 · 225· 232 Experience is not sufficient to support a definitive opinion on this aspect of therapeutic invention, and the use of prostacyclins is also conjectural. Once again, if these therapeutic steps fail to improve the recipient's condition, support with partial femoral artery to femoral vein cardiopulmonary by-pass may be required.196 C. PERFUSION OF THE TRANSPLANTED LUNG WITHOUT VENTILATION
Finally, if vascular integrity of the transplant is achieved before the completion of the bronchial anastomosis, and blood flow is allowed to occur, a shunt may be created. During this period of perfusion without ventilation, ST-segment changes and tachycardia/other arrhythmias have been noted in a few patients; it is possible that the institution of transplanted lung blood perfusion causes the washout of cold transplant perfusate and/or air and/or a bolus of potassium from the transplant (Eurocollins has a potassium concentration of 107 mmol/L, and high serum potassium has been documented in some of these patients).233 This unstable phase is usually short lived, lasting the time it takes to complete the bronchial anastomosis and reinflate the transplanted lung. During this unstable phase, it may be necessary to reduce/eliminate the perfusate injection/shunt by clamping of the pulmonary artery to the transplant or by separate ventilation of the transplanted lung. C. Postoperative Considerations Patients undergoing lung transplantation face a multitude of hemodynamic, respiratory, metabolic, infectious, immunologic, and other complications. A high level of intensive care is necessary for the first few weeks, and sometimes much longer, if success is to be achieved. Protocols for surveillance and monitoring of patients undergoing a complex cardiopulmonary procedure and transplantation should be in place. After the critical postoperative period, these patients do surprisingly well in the long term, with levels of energy and activity very similar to those of the general population. Many patients emerge from the operation moderately severely hypothermic (32°C)233 in spite of the use of warmed gases and intravenous fluids and warming blankets. Lung-transplant recipients are then subjected to donor-transmitted infections in the early postoperative period and later from nondonor sources. Donor-transmitted infections are associated wifh a high mortality rate, whereas later infections of nondonor source appear to be tolerated better.196 The recipient may be treated expectantly with appropriate antibotics initially
based on donor Gram's stain findings and later by actual recipient culture.196 Intermittent fiberoptic bronchoscopy may be used to evacuate the airway, obtain culture specimens, as well as monitor/document a satisfactory anastomotic lumen. The transplanted lung is readily susceptible to fluid overload, as well as large fluid loss from exudation into the plerual cavity, mainly because of the total disruption of pulmonary lymphatics combined with the increase in extravascular lung water that accompanies preservation, reimplantation, and reperfusion, as well as perhaps loss of a peripheral lung cough reflex. Low-dose vasoconstrictor blood pressure support rather than volume infusion, once adequate filling pressures have been established, should be used in an effort to minimize lung water. Diuretics are frequently used.199 Despite these measures, the transplanted lung, especially when only single-lung transplantation is carried out, frequently demonstrates a diffuse interstitial infiltrate on chest radiograph. Patients are weaned from ventilatory support as early as possible, and this may be facilitated by using a rigorous preoperative rehabilitation program and an epidural narcotic analgesia.199·225 Clearly, the longer that endotracheal intubation is required, the higher is the likelihood of pneumonia developing and the greater is the likelihood of problems with airway healing. This is a particular concern in patients with infective end-stage lung disease whose upper and lower airways are colonized at the time of transplantation. From the point of view of treating both lung infection and lung water accumulation with intermittent fiberoptic bronchoscopy, ventilation with a large-diameter single-lumen tube is desirable. In patients with emphysema, the remaining native lung may show evidence of severe air trapping in the emphysematous lung and may need to be treated with deliberate hypoventilation (in an attempt to prevent the air trapping and preferential ventilation of the highly compliant native lung). It may be necessary to use differential lung ventilation postoperatively for the highly compliant native lung and the relatively normal single transplanted lung (see chapter 20).226· 234~236 The air trapping, if severe, may cause hypotension, a phenomenon that one group has come to call "pulmonary tamponade."226,235 Occasionally, a prolonged ileus may occur that mandates the use of postoperative intravenous total parenteral nutrition. Patients should walk as early as possible, using supplemental oxygen as necessary and monitoring transcutaneous oxygen saturation. Frequent chest physical therapy is essential for optimal mobilization of secretions and is begun while the patient is still intubated. The usual immunosuppression protocol consists
Anesthesia for Special Elective Therapeutic Procedures
of using cyclosporine (4 mg/hr), azathioprine (2 mg/kg/day), and antilymphocyte globulin (2 mg/ kg/day) intravenously. The intravenous cyclospor ine is started at the conclusion of the transplanta tion operation. The antilymphocyte globulin, cy closporine, and azathioprine are continued for the first 10 to 14 days after the transplantation; pred nisolone is then added (second to third postopera tive week; steroids before this time interfere with the healing of the airway anastomosis) and contin ued with azathioprine and cyclosporine for the next 6 months. Subsequently, patients are weaned from prednisolone. All patients continue on azathioprine and cyclo sporine for life (no human leukocyte antigen matching of donor and recipient is possible; there fore, all patients require immunosuppression for life). Suspected acute episodes of rejection may be controlled by pulsed methylprednisolone, 10 mg/ kg intravenously for 3 days, followed by aug mented oral corticosteroids for 1 month. Patients who do not respond to methylprednisolone may be given a monoclonal antibody directed against Τ cells intravenously over 7 to 10 days. However, the diagnosis of rejection is a difficult one and presumptive and consists of observing developing pulmonary infiltrates on radiography, fever, and leukocytosis and a decrease in arterial oxygen ten sion (unfortunately, these changes are also sugges tive of infection). Early acute rejection has been almost a constant feature of isolated lung trans plantation, with two or three rejection episodes usually occurring within the first month. Bron choscopy is usually performed to obtain material for culture to rule out infection, and transbronchial lung biopsies may be performed during suspected rejection episodes in an attempt to obtain tissue confirmation of rejection. The episodes of rejec tion are usually readily managed with pulsed ste roid therapy, as described previously. With the replacement of the diseased lung by the transplanted lung, resetting of the receptors that control oxygenation, pH, and carbon dioxide tension appears to take several weeks.196 Most pa tients have experienced mild to moderate hypercarbia for the first several weeks after transplanta tion that gradually normalizes later. Either because of inadequate feedback signals from respiratory muscles as a result of the reduced respiratory effort required after transplantation or because there is delay in resetting the central chemoreceptors, pa tients frequently experience a sense of hypoxia in the postoperative period and will often respond by hyperventilation. Repeated blood-gas determina tions during these episodes are frequently in the normal range and have not shown hypoxemia. These distortions in respiratory mechanisms sub side gradually with time.
567
X. TUMORS AT THE CONFLUENCE OF THE SUPERIOR, ANTERIOR, AND MIDDLE MEDIASTINA A. General Considerations The mediastinum is divided arbitrarily, for the purposes of description, into upper and lower parts at the upper level of the pericardium by a plane that extends from the sternal angle to the lower border of the fourth thoracic vertebra. The upper part is named the superior mediastinum, and the lower part is again subdivided into three parts: the anterior mediastinum, in front of the pericardium: the middle mediastinum, containing the heart and pericardium; and the posterior mediastinum, be hind the pericardium. At the confluence of the superior, anterior, and middle mediastina are the middle portion of the superior vena cava, the tra cheal bifurcation, the main pulmonary artery, the aortic arch, and parts of the cephalad surface of the heart (see Figs. 2-23 to 2-25, 14-7, 15-38, and 15-39). In adults, the majority of tumors in this region originate from involvement of the hilar lymph nodes with bronchial carcinoma or lym phoma, whereas in infants the masses are most often benign bronchial cysts, esophageal duplica tion, or teratoma. Tumors of this region can cause compression and obstruction of three of the vital mediastinal structures: the tracheobronchial tree in the region of the tracheal carina, the main pulmo nary artery and atria, and the superior vena cava. The spectrum of clinical syndromes that can be caused by sclerosing mediastinitis (an encasing fibrotic reaction to chronic inflammation) well illus 237 trates this point. Computed tomographic scan of the chest is probably the single most important diagnostic procedure because it defines the size and degree of compression of vital structures. The most common complication to occur during anesthesia for masses involving the three vital me diastinal structures (tracheobronchial tree, pulmo nary artery and vein, and superior vena cava) is airway obstruction; this was a feature in 20 of 22 patients traced (1969-1983) and succinctly sum marized.238 Although airway obstruction has been predominant in terms of symptomatology, it is not uncommon for compression of two or three of these three major organs to be present in varying degrees in the same patient.238· 2 3 9 Each of these complications is life-threatening and can cause acute deterioration and death during anesthesia if not handled with the most extreme caution and expertise. Each of these three major complications and anesthetic management problems is discussed separately. Figure 15-35 shows the overall strat egy for managing these three problems.
568
Anesthesia for Special Elective Therapeutic Procedures
B. Compression of the Tracheobronchial Tree Most anterior mediastinal masses that cause airway obstruction are lymphomatous in origin. However, a number of benign conditions such as cystic hygroma, teratoma, and thymoma and thyroid tumors can occur in a similar fashion. A tissue diagnosis, therefore, usually (see later discussion) must be made before radiation or chemotherapy can be undertaken. Thus, most patients with a mediastinal mass causing airway obstruction will first require anesthesia for a diagnostic procedure (cervical or scalene node biopsy, staging laparotomy for Hodgkin's disease). However, it should be noted that general anesthesia before radiation treatment is associated with a 16 per cent incidence of life-threatening respiratory complications; indeed, 70 per cent of those who experienced life-threatening respiratory complications were symptomatic preoperatively.240·241 Not all patients in whom severe intraoperative problems developed had respiratory systems and signs preoperatively (significant tracheal obstruc-
tion may not cause symptoms if the onset has been insidious and chronic). The strong concern about the wisdom of using general anesthesia before radiation therapy, especially when the radiation therapy field can create a window for later accurate tissue diagnosis,242,243 has been voiced.241·244 Nevertheless, some pathologists insist that any prebiopsy irradiation of tumor area distorts cellular morphology to the extent that it prevents recognition of cell type and use of specific cell-type chemotherapeutic agents.240 In fact, in one institution, in patients with mediastinal lymphomas, 2.2 procedures per patient were required to obtain enough tissue for immunophenotyping.245 1. Anesthetic Management The anesthetic management of these patients is based on two overriding considerations. First, the obstruction of major airways by a tumor is usually life-threatening because the obstruction usually occurs around the bifurcation of the tracheobronchial tree and is, therefore, distal to the endotracheal
Anesthesia for Special Elective Therapeutic Procedures
569
Figure 15-36 In noncritical stenosis (a), the normal laminar flow profiles (as are produced by spontaneous ventilation) are compressed but resume their usual pattern distal to the stenosis. In critical stenosis (b), laminar flow profiles are severely compressed. If the gas flow reaches critical velocity (as produced by vigorous positive-pressure ventilation), flow will be disrupted into turbulent eddies, and laminar flow cannot be restored. With critical stenosis, if either the density or the velocity of the gas is reduced (c), gas flow may be maintained with much less disruption of the laminar flow profiles. (From Sibert K. Biondi JW, Hirsch NP: Spontaneous respiration during thoracotomy in a patient with a mediastinal mass. Anesth Analg 66:904-907, 1987. Used with permission.)
tube. Inhalation induction precipitated obstruction in three cases, and in none of these did intubation completely relieve it. In one case, intubation made the obstruction complete until a long, thin tube was passed, producing partial improvement.238 In another case,246 ventilation was difficult until spontaneous ventilation returned with the tube still in place. In another case,246 only a 4.5-mm rigid bronchoscope eventually secured a patent airway in an 11-year-old boy. In eight cases, intubation itself precipitated or exacerbated a partial mainstem bronchial obstruction.246"249 The obstruction was thought to be relieved by the return of spontaneous respiration in some of the patients.246-248 It may be that loss of chest wall tone and the distending forces of active inspiration (loss of negative intrathoracic pressure gradient, which increases airway diameter and holds the airway open) after administration of muscle relaxants release extrinsic support of a critically narrowed airway. Alternatively, intubation in the presence of distortion or compression of the trachea may cause complete obstruction if the orifice of the tube impinges on the tracheal wall, or if the lumen of the tube is occluded where it passes a narrowed section or turns a sharp angle. Such obstruction has developed in many patients and was relieved only by a long, thin tube or bronchoscope passed beyond the stenosis.246·249"253 Other reasons for tracheobronchial tree compression after the induction of general anesthesia include loss of lung volume (the tumor remains constant in size), the presence of tracheomalacia,
increase in central blood volume in the supine position (the tumor remains constant in size), edema, bleeding and hematoma formation in the tumor after it has been manipulated, injury to the recurrent laryngeal nerve (vocal cord paralysis), and finally the creation of very high flow rates and turbulent flow profiles with vigorous positive-pressure ventilation (Fig. 15-36 and Table 15-16). In view of this reported seriousness of airway obstruction during general anesthesia, all possible at-
Table 15-16
POSSIBLE CONTRIBUTORY CAUSES OF TRACHEOBRONCHIAL TREE OBSTRUCTION FOLLOWING THE INDUCTION OF GENERAL ANESTHESIA AND TRACHEAL INTUBATION IN PATIENTS WITH MEDIASTINAL MASSES
1. Loss of lung volume, mass remains constant in size, pressure on vital structures increases 2. Increase in central blood volume, mass remains constant in size, pressure on vital structures increases 3. Loss of negative pleural pressure and distending transmural pressure gradient 4. Obstruction of tip of endotracheal tube by wall of tracheobronchial tree resulting from distortion (bending, unusual curves, etc.) 5. Presence of tracheobronchial tree malacia 6. Increase in size of tumor as a result of surgical manipulation (edema, hematoma) 7. Creation of turbulent flow caused by vigorous positivepressure ventilation 8. Injury to the recurrent laryngeal nerve
570
Anesthesia for Special Elective Therapeutic Procedures
Table 15-17 IMPORTANT MANAGEMENT PRINCIPLES FOR TUMORS AT THE CONFLUENCE OF THE SUPERIOR, ANTERIOR, AND MIDDLE MEDIASTINA 1. Perform all procedures under local anesthesia if at all possible 2. Use radiation and/or chemotherapy if at all possible prior to general anesthesia 3. If general anesthesia, then consider inspection of tracheobronchial tree with fiberoptic bronchoscope and intubate awake 4. If general anesthesia, then maintain spontaneous ventilation
tempts to perform the procedure under local anesthesia should be undertaken. Second, the response of lymphomatous tumors to radiation or chemotherapy is normally dramatic. Chest roentgenograms reveal a marked decrease in tumor size, and the symptoms are usually improved. Consequently, it behooves the treating physicians to use radio- or chemotherapy if at all possible (sometimes the cell type can be known with a reasonable degree of certainty without a biopsy) before general anesthesia. It should be noted that, during radiation, a small window can be created to spare some tissue for adequate histologic diagnosis.242,243 The following management plan is based on the two previous principles189 (Table 15-17 and Fig. 15-37). If a patient with a mediastinal mass near the confluence of the superior, anterior, and middle mediastina exhibits dyspnea and/or intolerance of the supine position and a biopsy is scheduled, it should be done under local anesthesia, if at all possible. If the cell type is thought to be radiosensitive or chemosensitive, then appropriate types of therapy should be undertaken before any further surgery is performed. As noted previously, an anatomic radiation treatment window can be created to spare tissue for accurate histologic diagnosis.242,243 Following these types of therapy, the radiologic and computed tomogram appearance of the tumor must be reviewed along with a dynamic evaluation of pulmonary function (see later discussion). If the patient does not have dyspnea or intolerance of the supine position (i.e., is asymptomatic), a series of noninvasive tests may be performed to evaluate the anatomic and functional position of the tumor. First, a flow-volume loop should be performed in the upright and supine positions. The flow-volume loop is an extremely sensitive tool for evaluating obstructive lesions of the major airways255 and can differentiate between extrathoracic and intrathoracic airway obstructions. With extrathoracic obstruction, the inspiratory limb of the
flow-volume loop will show a plateau, and with intrathoracic airway obstruction, the expiratory limb of the flow-volume loop will show a plateau (see Fig. 15-15). A disproportionate reduction in maximal expiratory flows should also alert the physician to the presence of tracheomalacia and its inherent risk of precipitating dynamic airway collapse after tracheal extubation. Second, the chest computed tomographic scan will best reveal the anatomic location of the tumor and, perhaps, show a static picture of airway obstruction. Third, echocardiography may be performed in both the upright and supine positions to determine the impact of the tumor on the geography of the heart. If any of these three tests have positive results, then local anesthesia should be used for biopsy even if the patient is asymptomatic. If all three of these major noninvasive diagnostic tests have negative results, the patient may be anesthetized with general anesthesia if necessary, but local anesthesia is still preferable. Again, once the biopsy has been taken and the tissue shown to be radio- or chemosensitive, appropriate therapies should be instituted and the patient re-evaluated radiographically and functionally before further surgery is attempted. If general anesthesia is to be used in adults, the! airway should be evaluated by fiberoptic bronchoscopy with topical anesthesia before the induction of general anesthesia, if possible.25^258 The fiberoptic bronchoscope should be jacketed with an endotracheal tube, and after the fiberoptic bronchoscopy examination has been completed, the patient may be intubated. It is not advisable to pass a flexible bronchoscope through an extremely narrowed airway segment because this may precipitate total obstruction (the patient must exhale around the bronchoscope, even if oxygen is delivered through the small suction channel). Also, even brief introduction and prompt withdrawal of a fiberscope from a narrow passage may produce enough local edema to convert a partial obstruction into total occlusion. In this situation, it is best to intubate, keeping the fiberoptic bronchoscope proximal to the obstruction. The endotracheal tube (or double-lumen tube with main-stem bronchial obstructions or rigic bronchoscope), however, should be placed distai to any tracheobronchial tree obstruction whenevei it is inserted.253 Indeed, in one report, the only wa> both lungs could be ventilated at the same time was to cannulate both main-stem bronchi independently.259 In some extremely compromised patients, and depending on the exact location of the mediastinal mass, a surgical airway with the patient awake may be the best first choice of airway General anesthesia should be induced with th(
571
Conservative approach to Extubation Figure 15-37 Algorithm for anesthetic management of a patient with a mediastinal mass and suspected tracheobronchial tree obstruction. (CPB = cardiopulmonary bypass.) (Adapted with permission from Neuman GG, Weingarten AE, Abramowitz RM, et al: The anesthetic management of the patient with an anterior mediastinal mass. Anesthesiology 60:144-147, 1984. Used with permission.)
patient in the semi-Fowler's position, the patient should be allowed to breathe spontaneously throughout the procedure, and muscle relaxants should be avoided256 (even with a thoracotomy, if at all possible).254 In children, slow intravenous administration of ketamine and administration of nitrous oxide with spontaneous ventilation fulfills these goals260 (this is certainly possible if the thorax does not have to be entered; if the thorax does have to be entered, positive-pressure ventilation may be required). Spontaneous ventilation, however, does not guarantee that airway obstruction will not occur (half the reported cases of perioperative obstruction have been during spontaneous ventilation).242· 261 Large swings in intrathoracic pressure, which may promote collapse of a weakened tracheobronchial
tree, must be avoided. The operating room team should retain the capability of changing the patient's position rapidly to the sitting or lateral or prone position (e.g., Fig. 15-38). In one case, the fortuitous application of a laryngoscope blade relieved a near-complete obstruction of the tracheobronchial tree perhaps by applying tension to one end of the trachea, which straightened out the trachea.262 A rigid ventilating bronchoscope should be on hand, and the appropriate personnel and equipment for cardiopulmonary by-pass also should be available. These patients must be closely monitored in the first few hours. Airway obstruction requiring reintubation and mechanical ventilation has occurred (40 per cent of all perioperative obstruction cases)242·261 possibly secondary to an increase in
572
Anesthesia for Special Elective Therapeutic Procedures
Figure 15-38 Fiberoptic bronchoscopic appearance of lower trachea with anesthetized patient with large anterior mediastinal mass in supine position (A), exhibiting almost total obstruction of the trachea in anteroposterior plane. With patient in sitting position (B), lumen appears normal. (From Prakash UBS, Abel MD, Hubmayr RD: Mediastinal mass and tracheal obstruction during general anesthesia. Mayo Clin Proc 63:1004-1011, 1988. Used with permission.)
tumor size resulting from tumor edema after instrumentation. In some patients, especially those requiring reintubation, it may be best to irradiate the central core of the tumor and put the patient on steroids before extubation is attempted again.256
arterial oxygenation.267 Large mediastinal lymphomas have also been associated with arrhythmias under anesthesia owing to pericardial or myocar-i dial involvement. 1. Anesthetic Management
C. Compression of the Pulmonary Artery and Heart Compression of the pulmonary artery and heart is rare because the pulmonary trunk is more or less protected by the aortic arch and tracheobronchial tree; there are only four definite case reports in the literature.263-266 However, in view of the lethal nature of this complication, it warrants discussion. Although experience with this problem is extremely limited, the cell type has been lymphomatous in three reports and a cyst in the fourth patient. In the patient with a large mediastinal cyst, it was thought that a double-lumen tube caused the left main-stem bronchus to put pressure on the cyst, which, in turn, compressed the pulmonary artery.266 In an animal model of an expansile anterior mediastinal mass (800-ml intravenous infusion bag), the most significant cardiopulmonary effect of the mass, irrespective of the type of ventilation used (positive- or negative-pressure breathing) was obstruction of the pulmonary artery, causing right ventricular enlargement, decrease in left ventricular volume through ventricular interdependence (septal shifting), and decrease in cardiac index and
Similar principles for compression of the tracheobronchial tree apply to compression of the pulmonary artery. Most patients have their firsi anesthetic experience because they require a diagnostic procedure (e.g., a biopsy). All diagnostic procedures should be performed under local anes thesia if at all possible. Since the symptoms usu ally worsen when the patient assumes the supin< position, an unusual position may need to be uti lized (see later discussion). Lymphomatous tumor; are usually radiosensitive. These patients should be evaluated preopera tively in a manner similar to that for those witl compression of the tracheobronchial tree. If then is absolutely no indication of the cell type, loca anesthesia should be tried, if at all possible, ti perform the biopsy. If the cell type is known or i highly suspected, preoperative irradiation shouli be seriously considered. Some tissue should b excluded from the radiation field for accurate his tologic diagnosis.242·256 If general anesthesia is required, and the tumc is anterior to the heart, the sitting, leaning forwarc or even face-down position is advised (because th tumor will not press on the pulmonary artery c
Anesthesia for Special Elective Therapeutic Procedures
heart [as it would in the supine position]), and spontaneous ventilation should be maintained throughout the procedure (as with the intrathoracic airways, negative pleural pressure will distend the vessels and keep them open). Measures to maintain venous return, pulmonary artery pressure, and cardiac output, such as volume loading and use of ketamine, should be considered. Arrangements for extracorporeal oxygenation should be completed preoperatively. The anesthetist has to be aware of the danger of air embolism in a patient in the sitting position having a vertically nondependent surgical procedure. D. Superior Vena Cava Syndrome The superior vena cava syndrome is caused by mechanical obstruction (compression, invasion, thrombosis) of the superior vena cava. The causes of superior vena caval obstruction, in order of rapidly decreasing incidence, are bronchial carcinoma (40-90 per cent) (Fig. 15-39),268 malignant lymphomas (5-15 per cent), and benign causes (1-10 per cent), such as pulmonary artery, central venous, hyperalimentation, and pacemaker-catheterinduced thrombosis of the superior vena cava,269 idiopathic mediastinal fibrosis, mediastinal granuloma, and multinodular goiter.270-273 The classic features of superior vena cava syndrome include dilated distended veins in the upper half of the body due to increased peripheral venous pressure (which can be as high as 40 mm Hg); edema of the head, neck, and upper extremities; dilated venous collateral channels in the chest wall; and cyanosis. Venous distension is most prominent in the recumbent position, but in most instances the veins do not collapse in the normal manner with the patient upright. The majority of patients have respiratory symptoms (shortness of breath, cough, orthopnea, nasal stuffiness), which are due to obstruction of the airways by engorged veins and mucosal edema and are ominous signs. Similarly, a change in mentation and the presence of headaches, caused by cerebral venous hypertension and edema, are also ominous signs. In some cases, the superior vena cava becomes occluded quite slowly, collateral vessels have a chance to develop, and the signs and symptoms may be insidious in onset.272 Collaterals drain to the azygos vein in cases of high obstruction or to the hemiazygos or through chest wall veins to the inferior vena cava in cases of obstruction below the azygos confluence point.272 When the occlusion occurs relatively rapidly, all clinical manifestations are more prominent; in this setting, facial edema may be so severe that it prevents patients from opening their eyes. Moreover, rapidly in-
573
creasing venous pressure in the cerebral circulation may lead to neurologic impairment as cerebral perfusion pressure is decreased. The most common radiologic sign is widening of the superior mediastinum.272· 273 Venography confirms the diagnosis (but not the cause).274 Determination of cause may require thoracotomy, sternotomy, bronchoscopy, lymph node biopsy, and so on.272-274 Transesophageal echocardiography has been used to delineate the mechanism of superior vena cava obstruction.275 Most patients with superior vena cava syndrome (i.e., those with incomplete obstruction) caused by a malignant process are treated with irradiation and chemotherapy.274 However, in patients with near-complete to complete obstruction (who usually have signs of cerebral venous hypertension and/or airway obstruction), or when irradiation or chemotherapy proves ineffective, surgical by-pass or resection of the lesion via median sternotomy is indicated.274 These operations are usually technically quite difficult because tissue planes are poorly delineated, anatomy is grossly distorted, and varying degrees of fibrosis are present. Good temporary results have been obtained with balloon angioplasty.276 1. Anesthetic Management
-
The preoperative anesthetic evaluation of a patient for superior vena caval decompression should include careful assessment of the airway. The same degree of edema that is present externally in the face and neck can be expected to be present in the mouth, oropharynx, and hypopharynx. In addition, the airway may be compromised by external compression, fibrosis limiting normal movement, or recurrent laryngeal nerve involvement. If tracheal compression is suspected, it should be evaluated by computed tomography. Premedication for these patients is light or deleted when there is concern about the integrity of the airway. A drying agent is helpful if a difficult intubation is anticipated, and it should be administered to all patients having difficulty swallowing. The patient is transported to the operating room in the head-up position in order to minimize airway edema. A radial artery catheter is inserted in all patients and, depending on the medical condition of the patient, a central venous or pulmonary artery catheter is inserted via the femoral vein prior to the induction of anesthesia. On one occasion atrial pacing via a pulmonary artery catheter was utilized in a patient with suspected cardiac involvement from a mediastinal tumor who was experiencing significant bradycardia preoperatively. At least one large-bore intravenous cannula should be inserted in the leg or femoral vein prior to
574
Figure report, central MI: A
Anesthesia for Special Elective Therapeutic Procedures
15-39 A common cause of superior vena cava (SVC) obstruction is thoracic malignancy. In the patient described in thi an adenocarcinoma arose from the left main bronchus, involving the tracheal cartilage, hilar lymph nodes, and aorta, ι venous catheter may have also contributed to the thrombosis in conjunction with the tumor. (From Smith S, Vacek JL, Dun sudden change in central venous pressure. Hosp Physician 41-45, 1987. Used with permission.)
operation; the upper extremities are not used for intravenous infusions because of the long and un predictable circulation time that results from the superior vena caval obstruction. Cross-matched blood should be available in the operating room at the time of sternotomy. The method chosen for induction of anesthesia and intubation depends on the preoperative airway evaluation. If it is necessary for the patient to maintain the sitting position in order to achieve adequate ventilation prior to induction, intubation with the patient awake may be facilitated by using a fiberoptic laryngoscope277 or bronchoscope.278 The most significant intraoperative problem en countered is bleeding. Substantial venous blood loss results from the abnormally high central ve nous pressure. Further, unexpected arterial bleed ing may occur because of the difficulty of dissect
ing in a distorted surgical field. Cross-matche blood should therefore be available in the opera ing room at the time of sternotomy. Postoperatively, especially after diagnostic pre cedures such as mediastinoscopy and broncho! copy, wherein the superior vena caval obstrucţie has not been relieved, acute severe respiratory fai ure requiring intubation and mechanical ventili tion may occur.271· 279~283 The mechanisms of tlj acute respiratory failure are obscure, but the mq likely ones that are unique to the superior vei cava syndrome are acute laryngospasm and/) acute bronchospasm (both due to continued an perhaps, increased obstruction of the superior vei cava), impaired respiratory muscle function (p tients with malignant disease may have an abnc mal response to muscle relaxants),283 and increas airway obstruction by the tumor (due to turn
Anesthesia for Special Elective Therapeutic Procedures
swelling). Consequently, these patients must be closely monitored in the first few postoperative hours.
XI. REPAIR OF THORACIC AORTIC ANEURYSMS Patients with thoracic aortic aneurysms may come to the operating room on either an elective or an emergency basis. Thoracic aortic aneurysms in patients who are hemodynamically stable and free of significant symptoms are repaired on an elective basis. Patients with thoracic aortic aneurysms that were caused by trauma, or those who are hemodynamically unstable with acute symptoms (e.g., chest pain) due to an acute dissection, come to the operating room on an emergency basis. Consequently, the anesthetic management of patients with thoracic aortic aneurysms could be logically discussed as either an elective therapeutic procedure or an emergency procedure. The anesthetic management of patients with thoracic aortic aneurysms is discussed with special emergency procedures (see Thoracic Aortic Aneurysms and Dissections/Disruptions, chapter 17). The reason for this is twofold. First, patients requiring emergency repair of thoracic aortic aneurysms outnumber patients requiring elective repair.284 Second, the emergency cases are more challenging, involve higher risks, and entail a greater number of anesthetic considerations. Indeed, there is a 16 per cent surgical mortality for emergency resection compared with a 5 per cent surgical mortality for elective resection. The discussion of the anesthetic management of patients with thoracic aortic aneurysms in chapter 17 includes all the anesthetic considerations that would apply to patients undergoing elective resection of thoracic aortic aneurysms.
XII. THYMECTOMY FOR MYASTHENIA GRAVIS A. General Considerations Myasthenia gravis is a disease of neuromuscular transmission characterized by weakness and easy fatigability of the voluntary muscles. The disease has a prevalence of about 1 in 30,000. If untreated, myasthenia gravis has a 40 per cent mortality rate in 10 years, but with modern treatment death is rare.285 Evidence indicates that the disease is caused by an autoimmune attack by the immunoglobulin G autoantibody on the postsynaptic acetylcholine receptors in skeletal muscle. There appears to be a
575
causal connection between the binding of antibody to receptors, a decrease in number of the receptors in motor end-plates, and interference with neuromuscular transmission286 (Fig. 15-40). The average life of an acetylcholine receptor in a normal individual is 7 days, whereas in a myasthenic individual it is 1 day. The place of thymectomy in the treatment of myasthenia gravis is well established. It has been reported to result in remission or clinical improvement in at least 80 per cent of patients.287 The diagnosis can usually be established by the typical historic and neuromuscular examination findings.288 The most frequent complaint of myasthenic patients is diplopia; ptosis, the second most common sign, may be missed if it is mild. Ptosis may be unilateral and characteristically may shift from side to side. Dysarthria is an early symptom of bulbar involvement, followed by difficulties in chewing and swallowing, leading to weight loss. In fact, almost all myasthenia gravis patients have abnormal esophageal manometric studies.288 In advanced myasthenia gravis, the most frequent bulbar sign is facial weakness. In 15 to 20 per cent of myasthenic patients, the chief complaint is extremity weakness and easy fatigability; the arms are affected more frequently. Respiratory muscle weakness may prompt the patient to visit a physician, but this is rarely the first symptom. Various tests can be used to confirm the diagnosis. Anti-acetylcholine receptor antibody is specific to myasthenia gravis and is thus very useful in diagnosis. The antibody titer is elevated in 90 per cent of patients with generalized myasthenia gravis and in approximately 70 per cent of patients with ocular (group 1) disease. Antistriated muscle antibody is detectable in more than 90 per cent of patients with thymoma and in approximately 30 per cent of other myasthenia gravis patients. Electromyography shows a decrementai response in repetitive nerve stimulation. Administration of edrophonium or neostigmine (Prostigmine) results in a transient increase in muscle strength. Computed tomography of the mediastinum is helpful in determining whether a thymoma is present (9 to 16 per cent of myasthenic patients). Finally, acetylcholine receptor antibody titers are elevated in a majority of patients. There are five classifications of the clinical status of the patient (Table 15-18): 289 1, ocular symptoms and signs only, without progression (these patients are not treated by thymectomy); IIA, generalized weakness, mild bulbar and skeletal symptoms, and IIB, generalized weakness, moderate to severe bulbar and skeletal symptoms; III, acute fulminating weakness with severe bulbar involvement; IV, late onset or exacerbation, severe bulbar symptoms, and severe generalized weakness; and V, myasthenia gravis with muscular atrophy.
576 Anesthesia for Special Elective Therapeutic Procedures
Pharmacology of Myasthenia Gravis
Figure 15—40 Schematic diagram of the motor end-plate. In myasthenia gravis, antibodies to the postsynaptic acetylcholine (Ach) receptors destroy and reduce the number of receptors; hence, the affected muscle fatigues easily. Muscle relaxants have a much more exaggerated effect because they easily bind the reduced number of receptors. Acetylcholinesterase (Achesterase) metabolizes Ach. Achesterase inhibitors increase the amount of end-plate Ach and increase Ach-receptor binding and thereby increase muscle strength.
Approximately 33 per cent of patients with thy momas have myasthenia gravis and 10 to 15 per cent of myasthenia gravis patients have thymomas. Patients with thymomas tend to be older, have higher acetylcholine receptor antibody titers, and have a more severe form of the disease, and the response to thymectomy is somewhat poorer than in patients with thymitis.290 Myasthenia gravis may Table 15-18 MAIN GROUPS OF ACQUIRED MYASTHENIA GRAVIS Group
Description
Group I
Ocular myasthenia gravis (symptoms may remain persistently confined to the ocular muscles, particularly when 2 years have elapsed since the onset) Mild or moderately severe generalized myasthenia gravis Acute severe (fulminating) myasthenia gravis with respiratory muscle involvement Late (chronic) severe disease Myasthenia gravis with muscular atrophy
Group IIA, Β Group III Group IV Group V
be associated with thymitis; these patients are usu ally younger than 40 years and often have other autoimmune diseases (Table 15-19), and the re sponse to thymectomy is good.290 The pure ocular form of the disease responds very poorly to thy mectomy and is best treated with steroids (see following discussion). The nonoperative therapy of myasthenia gravis arises from the apparent autoimmune cause of the known deficiency in acetylcholine receptors. The mainstay of treatment of patients with myasthenia Table 15-19 IMMUNE DISORDERS ASSOCIATED WITH MYASTHENIA GRAVIS Rheumatoid arthritis Hyperthyroidism Hypothyroidism Polymyositis Systemic lupus erythematosus Pernicious anemia Sjogren's syndrome Pemphigus
Anesthesia for Special Elective Therapeutic Procedures
gravis is use of the oral anticholinesterase pyridostigmine (Mestinon) (Table 15-20). Most patients tend to prefer pyridostigmine because it has fewer muscarinic side effects (i.e., sweating, salivation, abdominal pain, diarrhea, bradycardia) than neostigmine and a more prolonged length of action. The anticholinesterase maintains the local concentration of acetylcholine at the motor end-plate at a high level and increases the chance of acetylcholine binding to an acetylcholine receptor (Fig. 1540). Occasional patients with myasthenia gravis may not respond to anticholinesterase drug therapy, which may be due to complete absence of postsynaptic acetylcholine receptors so that even the increased amount of acetylcholine present has no place to act. In addition, myasthenics on prolonged chronic anticholinesterase drug therapy may show decreased sensitivity to the drug. Exacerbation of myasthenia gravis can occur secondary to emotional and/or surgical stress and/ or adverse drug interactions (antibiotics and antiarrhythmia drugs), with the subsequent development of a myasthenic crisis. During crisis, patients have decreased responsiveness to anticholinesterase. This situation must be differentiated from a cholinergic crisis, which is secondary to an anticholinesterase overdosage.291 In both situations, there is an increase in muscle weakness, which may involve the respiratory musculature and, thereby, necessitate respiratory support. A small dose of edrophonium (10 mg intravenously) will improve strength in a patient with myasthenic crisis but will have little or a negative effect in a patient with cholinergic crisis. In both myasthenic and cholinergic crisis, it is best to withhold anticholinesterase medications while providing mechanical supportive ventilation. Remission of myasthenia can be anticipated in 80 per cent of patients receiving corticosteroids. Doses of prednisolone, 30 to 40 mg daily, are usually recommended. Once remission is achieved, the dose can be steadily reduced and remission maintained in many patients on as small a dose as 10 mg on alternate days. Because the beneficial effects are usually achieved well within 2 to 3
Table 15-20
577
weeks, corticosteroids must act in some way other than by immunosuppression. It seems probable that they protect the acetylcholine receptor from immunologic attack. Patients taking corticosteroids require less anticholinesterase therapy. If the anticholinesterase drug dose is not reduced, an initial deterioration in the myasthenia may occur in the first week or 10 days of treatment. It is advisable to start corticosteroid treatment with the patient in the hospital under close supervision. Relapse is common after the withdrawal of corticosteroids, and the inevitability of adverse effects must be appreciated if the dose of prednisolone cannot be reduced below 10 mg daily. For patients in whom remission cannot be maintained on 15 mg of prednisolone on alternate days, azathioprine treatment is instituted (see later discussion). In treating myasthenia, corticosteroids are used for those patients who are seriously ill before thymectomy, those who are unsuitable for thymectomy, and those who are insufficiently improved postoperatively. They are also useful in patients with ocular myasthenia, who, as a group, respond poorly to anticholinesterase drugs and to thymectomy. Azathioprine (2.5 mg/kg/day) is also effective in reducing anti-acetylcholine receptor antibody levels and in producing clinical improvement. The response is slower than that of corticosteroids; improvement begins at 6 to 12 weeks and becomes maximal at a mean of 6 to 15 months. In one study of 78 patients treated with azathioprine over 11 years, 31 patients were in complete remission, 40 were much improved, and none were worse.292 A dramatic but short-lived improvement in myasthenia can be brought about by plasma exchange. It is useful as a short-term measure in seriously ill patients until other forms of therapy become effective. However, there is no evidence that repeated plasma exchange combined with immunosuppression confers any greater long-term benefit than immunosuppression alone. Thymectomy has become increasingly important in the management of myasthenia gravis and should be offered to all significantly affected pa-
ANTICHOLINESTERASE MEDICATIONS AVAILABLE IN THE UNITED STATES FOR THE TREATMENT OF MYASTHENIA GRAVIS Parenteral Dose (mg)
Drug Neostigmine Pyridostigmine Ambenonium
Trade Name
Oral Dose (mg)
Onset/Duration of Oral Route
Prostigmine Mestinon Mytelase
15 60 6
15 min/2 hours 30 min/3-6 hours 30 min/4-6 hours
Abbreviations: IV = intravenous; IM = intramuscular.
IV
IM
0.5 2
0.5-1.0 3-4 Not available
578
Anesthesia for Special Elective Therapeutic Procedures
tients who are fit for operation unless they have purely ocular disease or have minimal symptoms. Suffice it to say that the thymus contains acetyl choline receptors on the myoid cells, and it is probable that tolerance to these is broken (e.g., by thymitis) and the circulating antibodies cross react with acetylcholine receptors on skeletal muscle. Complete remission or substantial improvement can be expected in 80 to 90 per cent of patients without thymomas, although 5 to 8 years may elapse before the benefits of operation become ap parent.293-296 In patients with thymic tumors, sur gical excision is indicated to prevent local spread (thymomas can be malignant), but the prognosis is not as good. Because of the possibility of malignancy, most centers advocate the most complete resection pos sible, including the tumor, the thymus gland, lymph nodes, and the surrounding fatty tissue as well.294·297 To achieve complete resection and rad ical dissection, a median longitudinal sternotomy, a lateral thoracotomy, or a transverse sternotomy may be used.297-298
B. Anesthetic Considerations
1. Preoperative Considerations Physical examination should include airway evaluation and tests of muscle strength and the ability to cough, chew, and swallow. Vital capac ity, peak inspiratory force, and maximum breath ing capacity should be measured in every patient. Those with impaired respiratory function need complete respiratory function tests performed, in cluding analysis of arterial blood gases. A chest roentgenogram and computed tomogram should be reviewed to assess the presence and size of the thymic tumor. Myasthenic patients may also have associated myocardial degenerative changes, which make a preoperative electrocardiogram nec essary. Thyroid abnormalities may be associated with myasthenia gravis, and patients need to be evaluated clinically and by laboratory tests for ev idence of hypo- or hyperthyroidism. Nutritional status should be evaluated by measurement of serum electrolyte, albumin, globulin, and hemo globin levels. Special attention must be given to serum glucose and electrolyte concentrations in patients maintained with steroid preparations, since prolonged therapy may induce fluid and elec trolyte disturbances and hyperglycemia or glycos uria. All nutritional deficiencies, cases of dehydra tion, electrolyte imbalances, and respiratory tract infections must be treated preoperatively. Plasma
pheresis is commonly used in the preoperative p< riod to optimize the patient's physical status. At present, it is recommended that anticholine terase therapy at the regular dose be continue until the day of surgery. On the day of surgery, mild cases, none or half the amount of the mornir dose of pyridostigmine is administered, whereas severe cases the full morning dose is prescribe In patients receiving systemic steroids, suppressu of the pituitary-adrenal axis should be considere and the regular dose of steroid should be mai tained throughout the immediate perioperative ρ riod. Postoperatively, the cortisone administraţii may be tapered from the second to approximate the fifth postoperative day. In view of the propensity for stress to cause myasthenic crisis, and because many of these ρ tients demonstrate some emotional instability, sp cial attention must be paid to the psycholo^ preparation of these patients. Premedication is i dicated, but care should be taken not to depress already weakened respiratory apparatus.
2. Intraoperative and Postoperative Considerations Anesthesia may be induced with thiopental, f lowed by the administration of a halogenated dn Thoracic epidural anesthesia with 1.5 per cent docaine in addition to a light general anesthe has been used with good success (the epidu catheter was also used for postoperative pain lief)·299 Both anesthesia techniques allow avo ance of muscle relaxants (see following disc sion). The surgical approach is most commoi through a median sternotomy (but may be a tra verse sternotomy), but in patients who have a n» midline thymoma, a lateral thoracotomy may necessary. If the incision is a median sternotoi a single-lumen tube may be used, and if the ir sion is a lateral thoracotomy, a single- or dout lumen tube may be used. Muscle relaxation is a special problem in th patients. Myasthenia gravis patients are resistan succinylcholine (more than normal is needed; e 1.5-2.0 mg/kg may be needed to produce η onset of good intubating conditions), and succii choline is associated with an early onset of a pr II block, which has a prolonged duration.300·301 ' likely explanation for these abnormal response that there is a decreased number of acetylcho receptors at the motor end-plate. Consistent \ this contention is the observation that myasthi gravis patients in true remission (asymptom while receiving no therapy) may not demonsl resistance to succinylcholine.302 Furthermore, it is not surprising that pati
Anesthesia for Special Elective Therapeutic Procedures
with myasthenia gravis have a marked sensitivity to nondepolarizing relaxants, and all nondepolarizing muscle relaxants have an unacceptable duration of action in these patients if administered in usual doses (e.g., the ED50 and ED95 for vecuronium is anywhere from two to five times less than that for normal patients).303' 304 The decrease in vecuronium requirement is significantly related to the patient's acetylcholine receptor antibody titer.304 Consequently, if relaxation is required, small doses of nondepolarizing relaxants should be given. Atracurium in small doses (5-15 mg)305 appears to be the drug of choice because of its short duration of action and method of elimination by spontaneous decomposition (Hoffman elimination). Vecuronium in small doses (0.01 mg/kg) has been successfully used.305· 306 No matter what relaxant was used, or how much, neuromuscular blockade must be monitored with a nerve stimulator; indeed, with appropriate monitoring, a succinylcholinevecuronium sequence may be used,306 or vecuronium may be used for intubation as well as for the procedure itself.307 Myasthenia gravis patients are much more sensitive, with respect to neuromuscular blockade, to isoflurane and halothane.308 The time course of the neuromuscular blockade after small doses of nondepolarizing muscle relaxants is similar to that observed after larger doses in nonmyasthenic subjects.308 At the end of the operation, the residual effects of the muscle relaxant can be effectively antagonized by neostigmine. A scoring system to predict which patients will require postoperative ventilatory support has been devised.309 The components of the scoring system consist of duration of myasthenia gravis equal to or greater than 6 years (equals 12 points); other concomitant respiratory disease present (equals 10 points); pyridostigmine requirement greater than 750 mg/day (equals 8 points); and a vital capacity less than 2.9 L (equals 4 points). A score of 10 points or more predicts the need for ventilatory support. The scoring system was derived in patients undergoing trans-sternal thymectomy with halogenated drug anesthesia without muscle relaxation; the predictive accuracy of this scoring system was 80 per cent.309 However, in patients undergoing transcervical thymectomy with the same anesthetic, the predictive accuracy of the scoring system was only 13 per cent.310 The variation in predictability of the scoring system may be due to differences in stress between the two surgical approaches. Alternatively, postoperative mechanical ventilation of all patients with myasthenia gravis has also been recommended.287 Extubation can be accom-
579
plished in an unhurried fashion and according to such objective criteria as vital capacity and peak inspiratory force. In addition, this approach allows time to restart the anticholinesterase therapy after clinical examination of the patient, especially with regard to bulbar and respiratory function and testing of handgrip strength with a dynamometer. The aim of the immediate postoperative anticholinesterase therapy is to maintain adequate spontaneous respiratory exchange. If the spontaneous ventilation is adequate or if mechanical ventilation is instituted, the anticholinergics are not required in the postoperative period. However, if spontaneous respiratory exchange is not adequate, half the regular dose of anticholinesterases may be administered during the first 3 postoperative days. On the fourth postoperative day, regular doses of anticholinesterases may be started. If the patient cannot take oral medication, parenteral administration of neostigmine (0.5 mg to 1 mg intramuscularly) can be given every 2 to 3 hours until medication can be absorbed orally, at which time pyridostigmine can be restarted. Because the beneficial effects of thymectomy can be delayed for several weeks to several years postoperatively, it is necessary to reassess the doses for a prolonged postoperative period. Pain relief can be obtained by titrating small doses of commonly used narcotics to the desired endpoint. However, one report claimed that lumbar epidural morphine (14 ml of 0.5 mg/ml solution) provided superior pain relief compared with intravenous narcotics.3" All patients with myasthenia gravis should be monitored postoperatively in an intensive care unit setting.
XIII. ONE-LUNG ANESTHESIA IN MORBIDLY OBESE PATIENTS Gastric stapling has become an accepted treatment for refractory morbid obesity.312·313 Although the procedure is usually performed through an abdominal incision, a transthoracic transdiaphragmatic approach with the patient in the right lateral decubitus position has been described and provides an improved operative exposure. As with any intrathoracic procedure, the transthoracic transdiaphragmatic surgical exposure can be greatly increased by using one-lung ventilation. Morbidly obese patients have heavy chest walls that cause a reduction in FRC below the closing volume of the lung, low ventilation-perfusion relationships, and hypoxemia. Consequently, one-lung ventilation in morbidly obese patients may be thought to be associated with an increased risk of hypoxemia. However, in two studies of morbidly obese pa-
580
Anesthesia for Special Elective Therapeutic Procedures
tients undergoing one-lung ventilation, no particular intraoperative or postoperative problems were encountered.314· 315 The patients were ventilated with a tidal volume of 15 ml/kg of ideal weight and an F,02 of 1.0; Pa02 ranged from 72 to 230 mm Hg during one-lung ventilation. The most probable reason good success was obtained during one-lung ventilation in these morbidly obese patients was that the lateral decubitus position allowed the panniculus to displace itself on the operating room table, thereby reducing abdominal pressure against the diaphragm (compared with the supine position) and allowing increased FRC and greater tidal diaphragmatic excursion.316 Postoperative depression of Pa02 and depression of postoperative pulmonary function test studies were slightly greater in the one-lung ventilation patients compared with a similar group of patients undergoing only an abdominal incision approach. Thus, it is reasonable to conclude that morbidly obese patients can tolerate one-lung ventilation/ anesthesia for transthoracic stapling surgery with safety as opposed to the abdominal approach. However, with either approach, special attention to many details is necessary to avoid problems with positioning (fitting on the operating room table, avoiding excessive pressure on nerves, elevation of the upper half of the body to facilitate respiration), monitoring (arterial catheter, access to central veins), the airway (acid aspiration, tracheal intubation, supplementary oxygen), choice of an-
esthesia (fat sequesters inhalation anesthetics and therefore metabolism is greater; postoperative respiratory depression), and postoperative care (positioning, ventilation, oxygenation, prevention of thromboembolism, pain control, and motivation/ ambulation). XIV. THORACIC OUTLET SYNDROMES A. General and Surgical Considerations The thoracic outlet syndromes are due to compression of any one or combination of lower brachial plexus (C-8, T-1), upper brachial plexus (C-5-C-7), the subclavian artery, and the subclavian vein (Table 15-21). The compression may be caused by the first rib, a cervical rib, an exostosis of a cervical vertebra, anomalies of the anterior scalene muscle, and cancer. 1. Benign Causes In the benign group of causes of the thoracic outlet syndrome, 70 to 80 per cent will have a history of trauma and 5 to 10 per cent will have a cervical rib.317·318 Compression of the lower brachial plexus causes pain in the supraclavicular and infraclavicular fossae, the back of the neck and rhomboid areas, and the axilla and inner arm, with
Table 15-21 THORACIC OUTLET SYNDROMES AND THEIR DIAGNOSIS AND TREATMENT
Thoracic Outlet Syndrome Compression of the lower brachial plexus (C-8, Tl) Compression of the upper brachial plexus (C-5-C-7) Compression of subclavian artery
Diagnosis History and physical (see text), nerve conduction/function studies; findings are in the distribution of these nerves History and physical (see text), nerve conduction/function studies; findings are in the distribution of these nerves History and physical (see text), arteriography
Compression of subclavian vein
History and physical (see text), venography
Superior sulcus (Pancoast) tumor causing compression of lower brachial plexus, subclavian vessels
All of the above and staging procedures (see Figs. 15-1 and 15-8)
Treatment (Transaxillary and/or Supraclavicular Approaches for Benign Disease) Removal of first rib, cervical rib (i.e., bony floor of thoracic outlet), and anomalous fibromuscular bands Removal of parts of the anterior scalene muscle Removal of first or cervical rib, various vascular surgery techniques (resection of aneurysm, endarterectomy) Removal of first rib, anomalous fibromuscular bands, possible fibrinolytic enzymes if thrombosis present Irradiation followed by en bloc removal of apical chest wall (including posterior portions of first three ribs, transverse processes of these thoracic vertebra, intercostal nerves, lower brachial plexus, stellate ganglion, and dorsal sympathetic chain), underlying lung, artery and vein procedures as dictated by tumor involvement
Anesthesia for Special Elective Therapeutic Procedures
numbness and tingling radiating through the ulnar nerve distribution from the axilla to the ring and small fingers.317 Symptoms may include dysesthesias of coldness and a dead feeling, with impaired strength and dexterity of the hand and fingers. The plexus is tender, and the 3-min elevated arm stress test quickly reproduces the symptoms. Treatment usually consists of removing the bony floor of the thoracic outlet, the first rib, and the cervical rib if present, along with the various types of anomalous fibromuscular bands, which usually provides remarkably good relief. Compression of the upper brachial plexus causes pain in the anterolateral aspect of the neck and in the region of the brachial plexus itself just behind the sternocleidomastoid muscle.317 Pain radiates up to the mandible and ear, posteriorly to the scapula, anteriorly into the upper chest, laterally across the top of the shoulder, and down the outer aspect of the arm in a C-5, C-6, and C-7 nerve distribution. Examination shows exquisite tenderness of the C-5 nerve and upper trunk of the plexus, especially with the head tilted to the opposite side, and much less incidence of edema, coldness, paresthesia, and motor impairment of the hand compared with the lower plexus pattern. Treatment usually consists of removing the anomalous encasing anterior scalene muscle. Compression of the subclavian artery (almost always caused by a cervical or first rib) causes ischemia of the upper extremity and may cause aneurysm formation and atherosclerotic plaques. Treatment consists of removal of the offending rib and appropriate vascular repair. Compression of the subclavian vein causes congestion and edema in the arm and hand, and venography will reveal whether the vessel is thrombosed. Treatment consists of removal of the offending rib and/or scalene muscle as well as perhaps fibrinolytic therapy. In general, using a combination of transaxillary first-rib resection, supraclavicular cervical rib resection, or scalenectomy, 70 per cent of patients will obtain long-lasting relief.318 The transthoracic route (anterolateral thoracotomy) has provided the same or better benefit with less risk of brachial plexus damage.3'9 With all approaches, a small percentage of patients may experience recurrence because of new scar tissue formation.317"319
2. Superior Sulcus (Pancoast) Tumor Although a variety of nonbronchogenic tumors may produce the clinical pattern peculiar to being located at the thoracic inlet, it is generally accepted that the most common cause of Pancoast's syn-
581
drome is a bronchogenic carcinoma arising in or near the superior pulmonary sulcus and invading the adjoining extrapulmonic structures by direct extension. It is the location of the tumor that is significant in producing the characteristic clinical pattern and not its pathologic structure or tissue of origin. Bronchogenic carcinomas situated in the narrow confines of the thoracic inlet invade the lymphatics in the endothoracic fascia and involve, by direct extension, the lower roots of the brachial plexus, intercostal nerves, stellate ganglion, sympathetic chain, and adjacent ribs and vertebrae, producing severe pain and Horner's syndrome (Pancoast's syndrome) (Fig. 15-41).320 The symptoms are characteristic of the location of the tumor in the superior pulmonary sulcus or thoracic inlet adjacent to the first and second thoracic and eighth cervical nerve roots, the sympathetic chain, and stellate ganglion. Initially, there is localized pain in the shoulder and vertebral border of the scapula. Later it may extend down the ulnar distribution of the arm to the elbow (involvement of T-l) and finally to the ulnar surface of the forearm and small and ring fingers of the hand (C-8). If the tumor extends to the sympathetic chain and stellate ganglion, Horner's syndrome and anhidrosis develop on the same side of the face and upper extremity. The complete diagnostic workup and staging and treatment plan is shown in Figure 15-42.320 Staging is determined by the location of the lesion and its metastases. The true superior sulcus tumor (Pancoast's tumor) is usually T3, which describes the extension of the tumor through the visceral pleura into the parietal pleura and the chest wall. No lymph node metastases (NO) or metastases only to the hilar nodes (Nl) are the usual criteria for operation. Mediastinal node metastases (N2) indicate "inoperability" (except with an N2 paratracheal mediastinal node on the ipsilateral side of the right chest that contains an intranodal metastasis, which may be considered "regional" spread). MO is the only stage for the operable group because peripheral metastases (Ml) signal a poor prognosis and contraindicate surgery. Previously, superior sulcus tumors were considered inoperable and were not often successfully palliated with irradiation alone. Best results seem to occur when the tumor and the localized adjacent area, including the superior mediastinal nodes, are treated preoperatively with 3000 rads given over 2 to 3 weeks. The purpose of the preoperative irradiation is to shrink the tumor, to weaken any cells that might be "spilled" at the time of surgery, and to block the lymphatics temporarily, which absorb cells
582
Anesthesia for Special Elective Therapeutic Procedures
Figure 15-41 Composite illustration of a patient with clinical manifestations, chest radiograph of tumor with superior sulcus location, and gross pathologic diagnostic demonstration of the lung carcinoma invading the chest wall, brachial plexus, and sympathetic nerves. (Copyright 1979 CIBA-GEIGY Corporation. Reproduced with permission from The CIBA Collection of Medical Illustrations by Frank H. Netter, M.D. All rights reserved.)
during the operative procedure. This is a reasonable approach considering the close margin of en bloc dissection in this type of tumor. Preoperative treatment of greater than 4000 rads may lead to poor healing after surgery. An interval of 2 to 4 weeks after radiation therapy allows the radiation to have maximal effect; resection of the tumor en bloc with the chest wall is then performed. Typical results of this treatment plan are a survival rate of 55 per cent at 3 years,321 27 to 35 per cent at 5 years,320·321 and palliation of pain in 72 per cent.321
B. Anesthetic Considerations There are several anesthetic considerations tha are unique to surgery for thoracic outlet problems First, if the surgical approach is extrapleural, on< needs to be especially watchful for an inadverten surgical entry into the operative pleural space causing either tension pneumothorax and/or col lapse of the lung. Second, the transthoracic ap proach is greatly facilitated by one-lung ventilatio! (collapse of the lung on the opposite side). Thirc if the disease and surgery involve the subclavia
Figure 15—42 Superior pulmonary sulcus carcinoma diagnosis (left column), staging (middle column), and treatment (right column). (CBC = complete blood count; SMA 20 = 20 blood chemistry tests; CT = computed tomography.) (From Urschel HC Jr: Superior pulmonary sulcus carcinoma. Surg Clin North Am 68:497-509, 1988. Used with permission.)
584
Anesthesia for Special Elective Therapeutic Procedures
vessels, large-bore intravenous access needs to be ensured to deal with major blood loss. Fourth, if the chest wall is unstable at the end of the proce dure, then a period of postoperative mechanical ventilation should be seriously considered.
REFERENCES 1. Dedhia HV, Leroy L, Jain PR, et al.: Endoscopic laser therapy for respiratory distress due to obstructive airway tumors. Crit Care Med 13:464-467, 1985. 2. Gelb AF, Epstein JD: Laser in treatment of lung cancer. Chest 86:662-666, 1984. 3. Unger M: Bronchoscopic utilization of the Nd:YAG laser for obstructing lesions of the trachea and bronchi. Surg Clin North Am 64:931-938, 1984. 4. Parr GVS, Unger M, Trout RG, et al: One hundred neodymium-YAG laser ablations of obstructing tracheal neo plasms. Ann Thorac Surg 38:374-381, 1984. 5. George PJM, Clarke G, Tolfree S, Garrett CPO, Hetzel MR: Changes in regional ventilation and perfusion of the lung after endoscopic laser treatment. Thorax 45:248253, 1990. 6. Gelb AF, Tashkin DP, Epstein JD, Szeftel A, Fairshter R: Physiologic characteristics of malignant unilateral mainstem bronchial obstruction. Am Rev Respir Dis 138: 1382-1385, 1988. 7. Mohsenifar Ζ, Jasper AC, Koerner SK: Physiologic as sessment of lung function in patients undergoing laser photoresection of tracheobronchial tree tumors. Chest 93:65-69, 1988. 8. Warner ME, Warner MA, Leonard PF: Anesthesia for neodymium-YAG (Nd-YAG) laser resection of major air way obstructing tumors. Anesthesiology 60:230-232, 1984. 9. Jacobson MJ, Lo Cicero J: Endobronchial treatment of lung carcinoma. Chest 100:837-841, 1991. 10. Goldberg M: Endoscopic laser treatment of bronchogenic carcinoma. Surg Clin North Am 68:635-644, 1988. 11. Snow JC: Fire hazard during C02-laser microsurgery on larynx and trachea. Anesth Analg 55:146, 1976. 12. Hirshman CA, Smith J: Indirect ignition of the endotra cheal tube during carbon dioxide laser surgery. Arch Oto laryngol 106:639, 1980. 13. Burgess GE, LeJeune FE: Endotracheal tube ignition dur ing laser surgery. Otolaryngology 5:561, 1979. 14. Patel KF, Hicks JN: Prevention of fire hazards associated with the use of carbon dioxide lasers. Anesth Analg 60:885, 1981. 15. Brutinel WM, McDougall JC, Cortese DA: Broncho scopic therapy with neodymium-yttrium-aluminum-garnet (Nd-YAG) laser during intravenous anesthesia. Chest 84:518-521, 1983. 16. Vourch G, Fischler M, Personne C, et al: Anesthetic management during Nd-YAG laser resection for major tracheobronchial obstructing tumors. Anesthesiology 61:150-151,1984. 17. Rudow M, Hill AB, Thompson NW, Finch J: Heliumoxygen mixtures in airway obstruction due to thyroid carcinoma. Can Anaesth Soc J 33:498-501, 1986. 18. Mizrahi S, Yaari Y, Lugassy JG, Cotev S: Major airway obstruction relieved by helium/oxygen breathing. Crit Care Med 14:986-987, 1986. 19. Pashayan AG, Gravenstein JS, Cassisi NJ, McLaughlin G: The helium protocol for laryngotracheal operations with CO : laser: A retrospective review of 523 cases. Anesthesiology 68:801-804, 1988. 20. Sosis M: Anesthesia for laser surgery. J Voice 3:163174, 1989.
21. Sosis MB: On the development of a new laser resistant endotracheal tube. J Clin Anesth 4:87-88, 1992. 22. Sosis MB: Evaluation of five metallic tapes for protection of endotracheal tubes during CO, laser surgery. Anesth Analg 68:392-393, 1989. 23. Green JM, Gonzalez RM, Sonbolian N, Rehkopf P: The resistance to carbon dioxide laser ignition of a new en dotracheal tube: Xomed Laser-Shield II. J Clin Anesth 4:89-92, 1992. 24. Rampil IJ: Anesthetic consideration for laser surgery. Anesth Analg 74:424^*35, 1992. 25. Lejeune FE, Guice C, LeTard F, Marice H: Heat sink protection against lasering endotracheal cuffs. Ann Otol Rhinol Laryngol 9:606-607, 1982. 26. Goldhill DR, Hill AJ, Whitburn RH, Feneck RO, George PJM, Keeling P: Carboxyhemoglobin concentrations, pulse oximetry and arterial blood gas tensions during jet ventilation for Nd:YAG laser bronchoscopy. Br J Anaesth 65:749-753, 1990. 27. Sosis M: Evaluation of a new laser resistant anesthesia circuit protector, drape and patient eye shield. Anesth Analg 72:S265, 1991. 28. Brutinel WM, Cortese DA, Edell ES, McDougall JC, Prakash UBS: Complications of Nd:YAG laser therapy. Chest 94:902-903, 1988. 29. Cavaliere S, Foccoli Ρ, Farina PL: Nd:YAG laser bron choscopy: A five year experience with 1,396 applications in 1,000 patients. Chest 94:15-21, 1988. 30. Arabian A, Spagnolo SV: Laser therapy in patients with primary lung cancer. Chest 86:519-523, 1984. 31. McDougall JC, Cortese DA: Neodymium-YAG laser therapy of malignant airway obstruction. Mayo Clin Proc 58:35-39, 1983. 32. McElvein RB, Zorn GL: Indications, results and compli cations of bronchoscopic carbon dioxide laser therapy. Ann Surg 199:522-525, 1984. 33. Chan AL, Tharratt RS, Siefkin AD, Albertson TE, Volz WG, Allen RP: Nd:YAG laser bronchoscopy: Rigid or fiberoptic mode? Chest 98:271-275, 1990. 34. Schneider M, Probst R: High frequency jet ventilation via a telescope for endobronchial laser surgery. Can Anaesth 37:372-376, 1990. 35. Personne C, Colchen A, Leroy M, Vourc'h G, Toty L: Indications and techniques for endoscopic laser resections in bronchology. J Thorac Cardiovasc Surg 91:710-715, 1986. 36. Dumas JF, Shapshay S, Bourcereau J, Cavaliere S, Meric Β, Garbi Ν, Beamis J: Principles for safety in application of neodymium-YAG laser in bronchology. Chest 86:163168, 1984. 37. McElvein RB: Treatment of malignant tracheobronchial tree obstruction. Ann Thorac Surg 48:463-464, 1989. 38. George RJM, Garrett CPO, Nixon C, Hetzel MR, Nanson E, Millard FJC: Laser treatment of tracheobronchial tu mors: Local or general anesthesia? Thorax 42:656-660, 1987. 39. Hetzel MR, Smith SGT: Endoscopic palliation of tracheo bronchial malignancies. Thorax 46:325-333, 1991. 40. Bloomquist S, Algotsson L, Karlsson SE: Anaesthesia for resection of tumours in the trachea and central bronchi using the Nd-YAG laser technique. Acta Anaesthesiol Scand 34:506-510, 1990. 41. Benumof JL: Another use for the plastic transparent dressing (letter). Anesthesiology 63:334, 1985. 42. Grant RP, White SA, Brand SC: Modified rigid broncho scope for Nd:YAG laser resection of tracheobronchial obstructing lesions. Anesthesiology 66:575-576, 1987. 43. Edell ES, Cortese DA: Bronchoscopic phototherapy with hematoporphyin derivative for treatment of localized bronchogenic carcinoma: A 5-year experience. Mayo Clin Proc 62:8-14, 1987.
CHAPTER
16
Anesthesia for Esophageal Surgery Preoperative Considerations A. History B. Nutritional Status C. Perioperative Regurgitation and Aspiration D. Preoperative Chemotherapy E. Diagnostic Workup Logic for Esophageal Lesions (With Special Emphasis on Esophageal Carcinoma) 1. Esophagoscopy a. Indications b. Anesthetic Technique c. Complications F. Benign Esophageal Diseases (Often Requiring Surgical Repair) 1. Hiatal Hernia 2. Benign Esophagorespiratory Fistula 3. Benign Esophageal Stricture 4. Achalasia
5. Esophageal Rupture and Perforation a. Clinical/Diagnostic Features b. Treatment of Esophageal Perforations G. Surgical Approaches/Incisions for Esophageal Surgery (With Emphasis on Esophageal Carcinoma) II. Intraoperative Anesthetic Considerations A. Monitoring B. Induction of Anesthesia C. Airway Considerations 1. General Airway Considerations 2. Airway Considerations in Cases of Esophagorespiratory Fistula D. Hemodynamic Considerations E. Esophagogastrointestinal Considerations III. Postoperative Considerations
592
Anesthesia for Esophageal Surgery
I. PREOPERATIVE CONSIDERATIONS
Table 16-1
A. History1 The most common symptom of esophageal disease is difficulty in swallowing, which is termed dysphagia. The difficulty may be characterized as various degrees of obstruction and/or pain during swallowing. Dysphagia that has been present for any significant length of time is almost always accompanied by weight loss, dehydration, prerenal failure, hypoalbuminemia, decreased plasma oncotic pressure, anemia, electrolyte abnormalities, and depressed immune mechanisms and muscle power (see Nutritional Status). Heartburn is also a common symptom and is due to reflux of gastric contents into the esophagus. Heartburn is often a postprandial symptom brought on by recumbency, leaning forward, belching, and severe exercise. In addition, many patients with esophageal disease (those with a significant pouch proximal to an obstruction and those with hiatal hernia) are prone to chronic regurgitation and aspiration and may have chronic lung disease (see Perioperative Regurgitation and Aspiration). Finally, patients with esophageal cancer are often treated with anticancer drugs that may cause considerable organ toxicity; doxorubicin (Adriamycin) can cause very severe refractory left ventricular myopathy, bleomycin can cause respiratory failure, and mitomycin can cause both pulmonary and nephrotoxicity (see Chapter 5 and Preoperative Chemotherapy discussed later). B. Nutritional Status
Patients with esophageal diseases present with dysphagia, nausea, and vomiting; poor nutrition may result, which, in turn, leads to many secondary metabolic and functional changes that have important anesthetic implications (Table 16-1).2 Dehydration resulting from poor fluid intake may be present, rendering the patient more susceptible to anesthesia-induced hemodynamic suppression. Dehydration may also cause prerenal failure as manifested by decreased urine output and increased blood urea nitrogen and creatinine. Hypoalbuminemia renders the patient more susceptible to an exaggerated response to drugs that are normally protein bound (thiopental, muscle relaxants, local anesthetics). Decreased plasma oncotic pressure will aîso make the patient more susceptible to pulmonary edema. Low levels of hemoglobin, resulting from poor iron intake or chronic bleeding from the site of the tumor, de-
POOR NUTRITION PROBLEMS CAUSED BY DYSPHAGIA AND NAUSEA AND VOMITING AND THEIR ANESTHETIC IMPLICATIONS
Dysphagia-lnduced Nutrition Problem Dehydration Hypoalbuminemia
Decreased hemoglobin Hypomagnesemia Malnutrition
Anesthetic Implication Hemodynamic instability, prerenal failure Exaggerated response to drugs, increased susceptibility to pulmonary edema Decreased 02 transport, tissue hypoxia Arrhythmias, altered neuromuscular transmission Depressed immune responses, sepsis, decreased muscle power, poor postoperative respiratory effort
crease oxygen transport potential and often result in preoperative elevated cardiac output and a rightshifted oxygen-hemoglobin dissociation curve. Subsequent decreased cardiac output caused by anesthesia and left shifting of the oxygen-hemoglobin dissociation curve resulting from hyperventilation may result in tissue hypoperfusion and hypoxia. Electrolytes may be altered secondary to starvation, commonly producing either hypokalemia or hypomagnesemia. Hypokalemia may be accentuated by unintentional hyperventilation or metabolic alkalosis (as may be caused by citrated blood, administration of sodium bicarbonate, loss of gastric acid, diuretics), which may cause cardiac arrhythmias, especially in a digitalized patient. Low levels of magnesium may cause alterations in the electrocardiogram and in neuromuscular transmission. Malnutrition also suppresses the immune response and impedes wound healing; these latter two abnormalities may increase the risk of wound infection. Malnutrition also decreases muscle power, and a poorly nourished patient is more likely to experience postoperative respiratory complications. All these metabolic and functional abnormalities can now be corrected by total parenteral nutrition before surgery.3 Consequently, parenteral nutrition is now frequently used preoperatively, and a typical daily intravenous nutrition prescription is shown in Table 16-2.4 The calorie source for parenterally fed patients is usually a combination of carbohydrate (dextrose) and fat. For most adult patients, 40 to 45 calories per kilogram body weight per day are adequate. Intravenous protein requirements are usually provided as synthesized amino acids.
Anesthesia for Esophageal Surgery
Table 16-2 DAILY INTRAVENOUS NUTRITION PRESCRIPTION Source 10% Fat solution Dextrose 50% (500 g) Amino acids 8.5% Multiple essential ions/ minerals and vitamins
Amount (ml)
Nutrition
500 1000 1000
550 calories 1700 calories 85 g
However, it is important for the anesthesiologist to realize that parenteral nutrition may pose two common special problems during anesthesia and surgery (Table 16-3). First, the risk of intraoperative hypoglycemia is increased in these patients.5 The islet cells of the pancreas are stimulated during the infusion of dextrose-rich solutions, and these patients may have high levels of endogenous insulin. If the dextrose-rich infusion is stopped when the patient arrives in the operating room, and the islet cells continue to secrete insulin, hypoglycemia may develop intraoperatively within 45 to 60 min. If the hypoglycemic episode is severe and prolonged, it may result in delayed awakening or coma in the postoperative period. This problem can be avoided either by infusing the parenteral nutrition solution throughout the intraoperative period at half the presurgical flow rate or by substituting 10 per cent dextrose in water at the same infusion rate.2 Either technique prevents the intraoperative hypoglycemia, but both techniques require perioperative monitoring of the blood sugar. Second, the preoperative administration of large amounts of glucose-rich parenteral nutritional solutions may result in Hpogenesis.6 Lipogenesis results in increased carbon dioxide production, which is removed in the awake patient by a spontaneous increase in ventilation. However, when these patients are paralyzed and mechanically ventilated intraoperatively, the use of a normal minute ventilation may result in severe hypercarbia.7 This can be prevented by hyperventilation, sometimes in excess of twice the normal minute ventilation. Postoperatively, the increase in C0 2 production (caused by lipogenesis) may be a critical factor inhibiting the weaning of a patient from mechanical ventilatory support. Decreasing the glucose load or changing to fat emulsions, which causes a respiratory quotient of 0.7, may be useful in these circumstances (see chapter 3).6 C. Perioperative Regurgitation and Aspiration Patients with esophageal disease are prone to regurgitation and aspiration. In very debilitated pa-
593
tients with poor laryngeal reflexes, regurgitation may lead to chronic aspiration, atelectasis, pneumonia, and chronic lung disease. In many patients, the aspiration of acid gastric contents may take place during sleep.8·9 It is thought that the gastric reflux may be the cause of asthma in some patients, and both diseases/problems may be cured by an antireflux surgical procedure.10 In patients with an esophageal obstruction, it may take many hours for the esophagus above the stricture to empty, and having the patient fast overnight before surgery does not guarantee an empty pouch above the stricture or obstruction. Consequently, if there is any question about the proximal esophageal pouch, it should be emptied by passing a large nasogastric tube and suctioning before induction of anesthesia. In spite of these attempts, the pouch may not be completely empty, and it is reasonable to consider these patients as having a full stomach. The gastroesophageal barrier consists of both a valve (caused by the angle at which the esophagus enters the stomach) and a sphincter (the lower esophageal sphincter)." Patients with hiatal hernia or esophageal carcinoma may have an obliterated esophagogastric angle and physiologically incompetent or dysfunctioning lower esophageal sphincter (often indicated by history of reflux).12 They are also thus prone to regurgitation and aspiration during induction of anesthesia. With these considerations in mind, intubation should be performed either in an awake patient with sedation and topical anesthesia or by rapid-sequence induction. A nasogastric tube should be available toward the end of the operation for threading through any esophageal anastomosis and into the stomach, if the surgeon wishes. The purpose of the tube is to prevent gastric distention and to indicate whether bleeding is continuing from the stomach. It provides no guarantee against aspiration of gastric contents into the lungs. Patients without significant pulmonary disease can be allowed to breathe spontaneously after an uncomplicated procedure, but they should not be extubated until they are alert and sitting upright because there may be no mechanical barrier to reflux and aspiration into the lungs postoperatively.
Table 16-3
INTRAOPERATIVE PROBLEMS THAT CAN BE CAUSED BY PREOPERATIVE TOTAL PARENTERAL NUTRITION
1. Hypoglycemia (reactive insulin secretion) 2. Hypercarbia (increased C0 2 production because of lipogenesis)
594
Anesthesia for Esophageal Surgery
D. Preoperative Chemotherapy (See Chapter 5)
Patients with esophageal cancer are often treated with chemotherapeutic drugs before surgery. The anticancer effect of these antibiotics is produced by formation of relatively stable complexes with deoxyribonucleic acid (DNA), which inhibits DNA and/or ribonucleic acid (RNA) function and synthesis.13 These drugs affect not only the cancer cells but also rapidly growing normal cells (erythropoiesis, leukocyte and platelet production, and gastrointestinal tract lining). Commonly used chemotherapeutic drugs in carcinoma of the esophagus belong to the antibiotic group, which includes doxorubicin, bleomycin, and mitomycin C. Toxicity secondary to doxorubicin includes severe cardiomyopathy, seen in 1.8 per cent of patients treated. When cardiomyopathy develops, it has been shown to be irreversible in 60 per cent of the patients, with death occurring within 3 weeks of the onset of the symptoms.14 The left ventricular failure that occurs with doxorubicin is refractory to inotropic drugs. Electrocardiogram abnormalities are also part of doxorubicin toxicity, but they resolve 1 to 2 months after cessation of therapy.15 Bleomycin's action is similar to the effect of radiation, and bleomycin and radiation may act synergistically during simultaneous therapy.16 Pulmonary toxicity is the most life-threatening, druglimiting effect, reported in 15 to 25 per cent of patients.17- I8 Predisposing factors include age greater than 20 years, dose greater than 400 units, underlying pulmonary disease, and prior radiation therapy. Signs and symptoms of pulmonary toxicity are cough, dyspnea, and basal rales. The disease may manifest itself with minimal radiologic changes and normal resting Pa02, or it may progress to severe hypoxemia at rest, with radiologic changes similar to severe adult respiratory distress syndrome. A controversial factor that may predispose patients to pulmonary toxicity is the administration of oxygen in high concentrations.1920 Mitomycin C is also highly toxic and can cause pulmonary fibrosis and nephrotoxicity.21 Thus, patients receiving mitomycin C should have their pulmonary and renal status fully evaluated in the preoperative period. E. Diagnostic Workup Logic for Esophageal Lesions22 (With Special Emphasis on Esophageal Carcinoma)
There are numerous esophageal lesions and disorders. In order of decreasing frequency of occurrence, they are tumors (squamous cell carcinomas and adenocarcinomas are most common), hiatal
hernia, benign strictures (ingestion of caustic fluids causes strictures in the cervical esophagus, whereas reflux esophagitis causes strictures in the lower third of the esophagus), foreign bodies, diverticula, achalasia, esophagorespiratory tract fistula, traumatic perforations, and various motility disorders (such as scleroderma). After routine screening chest X-rays, a barium swallow (esophagogram) under cinefluoroscopy should be performed (Fig. 16-1). The only contraindication to obtaining an esophagogram is when a fistula to the trachea or bronchus is suspected. The barium swallow in a very high percentage of cases gives a strong indication of diagnosis. If a structural lesion is suspected, the esophagogram should be followed by esophagoscopy. Esophagoscopy should be performed with a flexible fiberoptic instrument; however, when difficulty is encountered while obtaining an adequate biopsy specimen, a rigid esophagoscope may provide improved conditions for collecting a biopsy. Rigid esophagoscopy should be avoided as the first diagnostic test whenever possible because of the risk of inadvertent perforation of an esophageal tumor. Less often, the esophagogram shows that the symptoms are not due to a structural obstruction, and manometric motility studies, pH studies, and provocation studies (Bernstein test [reproduces heartburn and pain by dripping a gastric acid-like solution through a nasogastric tube 30 tc 35 cm from the nose, which is the distal esophagus] and the Tensilon test [reproduces chest pair by causing increased esophageal contractions]) are indicated. If the sequence of esophagogram anc esophagoscopy has not made possible diagnosis thoracotomy rarely will be required for diagnosis. If the diagnosis of carcinoma of the esophagu; is being entertained, strong consideration shouk be given to examination of the mediastinum (me diastinoscopy or computed tomographic scanninj and nuclear magnetic resonance imaging; how ever, the value of computed tomography and nu clear magnetic resonance imaging has been sen ously questioned)23 because 60 to 80 per cent o esophageal tumors may involve the mediastinum by the time surgery is performed24-27 (see discus sion of surgical approach to esophagectomy). Bronchoscopy should be performed in patient with upper or middle one third esophageal lesions because nearly one sixth of these lesions will hav tracheobronchial tree involvement or vocal cor paralysis.27 The staging of carcinoma of the esophagus i similar to that of the lung (Table 16—4). Based o preoperative staging, 60 per cent of these lesior are considered to be potentially resectable (Fij 16-2). Unfortunately, and perhaps because of th
Anesthesia for Esophageal Surgery
595
Figure 16-1 Preoperative evaluation logic of esophageal lesions. See text for full explanation.
limitations of mediastinal examination,23 final staging is often done at the time of surgery and is most often upgraded by the intraoperative and postoperative (histologic) findings.27 The vast majority of esophageal carcinoma is squamous cell, and a small percentage is adenocarcinoma and small-cell carcinoma. All esophageal cancers have a large male preponderance (approximately 3:1). Nonstructural lesions may be initially treated with antireflux (Table 16-5) or antimotility (Table 16-6) measures. The distinction between these two categories (reflux vs. motility disorder) is important because some of the treatments are contraindicated for the other disease (compare Table 16-5 with Table 16-6). The physiologic assessment of overall cardiorespiratory function of the patient requiring esophageal surgery is similar to that of a patient with lung cancer. As with resection of lung carcinoma, age per se is not a contraindication to esophagectomy (operative mortality of 13 per cent in patients older than 70 years compares favorably with the overall operative mortality of 10-20 per cent [see later discussion]). As with lung carcinoma, resection is the best hope of cure for patients with esophageal carcinoma and is frequently the best form of palliation.28
1. Esophagoscopy a. INDICATIONS
The major indication for esophagoscopy is demonstration of an esophageal lesion following contrast studies that needs either etiologic or anatomic clarification.29 Such lesions include all strictures, intraluminal filling defects, mucosal abnormalities, and upper gastrointestinal bleeding. In most of these situations, biopsy specimens and cytologic brushings will also need to be obtained. Candidates for esophagoscopy include patients who have symptoms of difficulty in swallowing and those with esophageal reflux. Finally, therapeutic esophagoscopy is required for removal of foreign bodies, dilatation of strictures, placement of plastic prosthesis across a malignant stricture, the injection of sclerosing agents into esophageal varices, and coagulation of bleeding lesions. Table 16-7 compares the advantages and disadvantages of fiberoptic esophagoscopy and opentube rigid esophagoscopy.29 Generally, fiberoptic esophagoscopy lends itself to greater patient comfort, the ability to examine the entire upper gastrointestinal tract, and greater safety. Consequently, flexible fiberoptic esophagoscopy is usually carried out under local anesthesia. Disadvantages of the
596
Anesthesia for Esophageal Surgery
Table 16-4 UICC/AJCC TNM STAGING SYSTEM FOR CARCINOMA OF THE ESOPHAGUS't
*From Muller JM, Erasmi H, Stelzner M, et al: Surgical therapy of oesophageal carcinoma. Br J Surg 77:845-857, 1990. Used with permission. tThe evaluation of the primary tumor (T) no longer considers anatomic location, length of tumor, or percentage of wall circumference involved. Only depth of esophageal wall invasion is assessed. The evaluation of regional lymph nodes (N) has also been simplified. Anatomic site and side of involvement are no longer considered. Whether a regional lymph node contains metastatic carcinoma is the only staging criterion. Sites of distant metastasis (M) have been redefined to include celiac lymph nodes.
fiberoptic instrument are the inability to obtain deep biopsy specimens and inadequate instrumen tation for foreign body removal. Specific indications for the use of open-tube
esophagoscopy include removal of a foreign body, dilation of strictures, and determination of the source of massive esophageal bleeding. Although open-tube esophagoscopy can be carried out under
Figure 16-2 Survival rate after resection according to stage (Union Internationale Contre Cancer; 1978) of the esophageal carcinoma. Stage 1 (ο): η - 739; Stage II (·): η = 1592; Stage III (ο): η = 2995; Stage IV (•): η - 1392. (From Muller JM, Erasmi H, Stelzner M, et al: Surgical therapy of oesophageal carcinoma. Br J Surg 77:845-857, 1990. Used with permission.)
J
Anesthesia for Esophageal Surgery
Table 16-5
INITIAL THERAPY FOR GASTROESOPHAGEAL REFLUX
1. Diet; avoid irritating substances (citrus juices, tomato products, coffee, spices). 2. Avoid substances that lower the lower esophageal sphincter pressure (nicotine, fatty foods, alcohol, peppermint, nitrates, anticholinergics, theophylline, calcium blockers). 3. Avoid nighttime sedatives; arousal from sleep may help to clear the esophagus. 4. Elevate head of bed. 5. Take antacids. 6. Take H:-blockers. 7. Take metoclopramide. 8. Undergo surgery.
topical anesthesia, it is more readily accomplished under general endotracheal anesthesia.30 Opentube esophagoscopy is to be avoided in the patient with a large thoracic aneurysm because instrumentation may cause rupture. The procedure is also contraindicated in patients with acute pharyngitis because infection may be aggravated. Finally, the risk of perforation during dilation of malignant strictures with a rigid esophagoscope has been approximately 10 per cent.31 b. ANESTHETIC TECHNIQUE
Topical anesthesia of the oropharynx for fiberoptic esophagoscopy can be achieved by having the patient gargle viscous lidocaine followed by a tetracaine or lidocaine spray. The patient is further sedated before the examination with small doses of a sedative and/or narcotic. General anesthesia is usually required for rigid esophagoscopy (certainly for children).32 Atropine premedication is useful in decreasing secretions and in preventing a vagal reaction to gastric distension. Oropharyngeal topical anesthesia (with lidocaine gargle and spray) is useful in minimizing the
Table 16-6
INITIAL THERAPY FOR ESOPHAGEAL MOTILITY DISORDERS OTHER THAN ACHALASIA
Mild Cases Nitrates Nitroglycerin (0.4 mg prn) Isosorbidc ( 10—20 mg qid) Anticholinergics Dicyclomine (20 mg tid) Tranquilizers/antidepressants Diazepam (5 mg tid) Doxepin (50 mg hs) Reassurance
Severe Cases Calcium channel-blocking agents Diltiazem (60-90 mg tid) Nifedipine (l0-20mg tid) Esophageal dilatation (50 bougie prn) Psychological evaluation
Esophagomyotomy (only in extreme cases)
Abbreviations: prn = as occasion requires; qid = four times daily; tid = three times daily; hs = at bedtime.
597
amount of general anesthesia required. Preoxygenation is followed by intravenous induction using cricoid pressure to control reflux of esophageal contents. After intravenous induction of anesthesia and paralysis, a small-sized endotracheal tube should be inserted to prevent anterior compression of the esophagus (as a large-sized endotracheal tube might do) and thereby allow more room for esophageal instrumentation and passage of the esophagoscope. The endotracheal tube should be positioned to the left side of the patient's mouth so that the esophagoscope can be introduced through the right side of the mouth. The connecting tubing should be fastened over the front of the patient's chest to allow the surgeon easy access to the mouth. The patient's eyes should be protected because they are likely to be concealed by a drape around the head. Full relaxation should be maintained at least until the esophagoscope has been passed through the cricopharyngeal sphincter. If this passage proves difficult, it may be necessary to deflate the cuff on the endotracheal tube for a few moments to provide a little more room for the esophagoscope. Although spontaneous ventilation can be permitted once the esophagoscope is through the cricopharyngeal sphincter, it is preferable to retain control of ventilation and maintain paralysis to prevent any coughing or bucking during the procedure, which may cause the esophagoscope to damage or tear the esophagus. Certainly, full relaxation should be maintained throughout the procedure if repeated instrumentation through the cricopharyngeal muscle is required. Arrhythmias may occur as the esophagoscope passes behind the heart, but they are rarely a source of actual concern and are usually only transitory. Recovery of consciousness should be supervised with the patient on the side, slightly head-down on a gurney that can be tipped. A postoperative chest X-ray should be obtained to rule out signs of esophageal tear, such as subcutaneous emphysema, pneumoperitoneum, pneumothorax, or pneumomediastinum. Emergency thoracotomy for primary repair or diagnosis may be required if esophageal rupture is suspected (see Esophageal Rupture Perforation below). C. COMPLICATIONS
The complications of esophagoscopy include perforation, hemorrhage, and cardiopulmonary complications.33 34 The most frequent site of perforation is the hypopharynx, which carries a mortality rate ranging from 34 to 84 per cent.33"37 Perforation of the esophagus is most likely to occur in patients with esophageal carcinoma (e.g., 12 per cent).38 Biopsy specimens that are taken too deeply
598
Anesthesia for Esophageal Surgery
Table 16-7
FIBEROPTIC VERSUS OPEN-TUBE ESOPHAGOSCOPY Advantages Patient comfort Ability to examine entire upper gastrointestinal tract General anesthesia not required Greater safety 1. 2. 3. 4. 5.
Deeper biopsy possible Greater aspiration capacity Dilation feasible/indicated Foreign body removal Ease of sterilization
Disadvantages Fiberoptic Esophagoscopy 1. Expensive equipment required 2. Greater occurrence of mechanical failure 3. Dilation not possible 4. Smaller biopsies Open-Tube Esophagoscopy 1. Patient discomfort 2. Safety 3. Less readily done on outpatient basis 4. Greater risk of perforation
(i.e., full thickness) can also result in perforations. Hemorrhage may occur from esophageal varices, lacerations at the gastroesophageal junction induced by forceful retching, and biopsy. Cardiopulmonary problems are due to aspiration, pneumonia, cardiac arrhythmias, and myocardial infarction. Septicemia may occur in immunosuppressed patients, and prophylactic antibiotics are indicated in this group. F. Benign Esophageal Diseases (Often Requiring Surgical Repair) The surgical management of complex benign esophageal disease is challenging and has no easy solution. Many patients with this condition have undergone a previous operation. Table 16-8 shows the wide variety in 63 operations initially per-
formed in 35 patients, and Table 16-9 shows the wide variety of subsequent reconstructive operations performed in the same 35 patients in one institution.39 1. Hiatal Hernia Two types of hiatus hernia have been described. Type I hernias, also called sliding hernias, make up approximately 90 per cent of esophageal hiatal hernias. In this type, the esophagogastric junction and fundus of the stomach have herniated axially through the esophageal hiatus into the thorax. The type II, or paraesophageal hiatus hernia, is characterized by portions of the stomach herniating into the thorax next to the esophagus. In the presence of a type II hernia, the esophagogastric junction is still located in the abdomen. The goal of surgical Table 16-9
Table 16-8
COMPLEX BENIGN ESOPHAGEAL DISEASE: PREVIOUS OPERATION*
COMPLEX BENIGN ESOPHAGEAL DISEASE: SUBSEQUENT "DEFINITIVE" OPERATION* No.
No. Fundoplication Nissen (17) Collis-Nissen (7) Collis-Belsey (2) Belsey (2) Thal-Nissen (2) Hill (1) Esophagomyotomy Diaphragmatic hernia repair Esophageal exploration and/or repair Esophageal exclusion Cardiectomy Diverticulectomy Miscellaneous Total
31
11 5 4 4 3 2 3 63
*From Ellis FH, Gibb SP: Esophageal reconstruction for complex benign esophageal disease. J Thorac Cardiovasc Surg 99:192-199, 1990. Used with permission.
Operation Esophagectomy Esophageal exclusion Cardioplasty No resection Total Reconstruction Acid suppression and alkaline diversion Colon interposition Substernal (5) Intrathoracic (3) Esophagogastrostomy Inkwell (3) End to side (2) Transhiatal (1) Total
27 4 2 _2 35 21 8
6
35
*From Ellis FH, Gibb SP: Esophageal reconstruction for complex benign esophageal disease. J Thorac Cardiovasc Surg 99:192-199, 1990. Used with permission.
Anesthesia for Esophageal Surgery
repair of a sliding hernia is to obtain gastroesophageal competence. Because restoration of the normal anatomy is not always successful in preventing subsequent reflux, several antireflux operations have been developed, the so-called wraparound procedures. For example, Nissen fundoplication, which can be performed via an abdominal or thoracic incision, entails wrapping the distal esophagus with the fundus of the stomach. Barrett's esophagus is the eponym commonly used to describe the presence of columnar epithelial lining within the distal tubular esophagus. The change in the lining of the esophagus may be due to chronic gastric reflux (as may be caused by a hiatal hernia). Unfortunately, there is a moderately strong association between the columnar-lined esophagus and esophageal adenocarcinoma. 2. Benign Esophagorespiratory Fistula Benign esophagorespiratory fistula is a rare condition and an unusual cause of chronic pulmonary suppuration in adults. Although the ratio of benign to malignant esophagorespiratory fistulas is approximately 1:5, benign esophagorespiratory fistulas are of greater surgical interest because of their potential for total curability. Benign esophagorespiratory fistulas may be either congenital or acquired; the latter is more common. Acquired benign esophagorespiratory fistulas are most commonly inflammatory or posttraumatic in origin, and an increasing number of the latter seem to have been recorded as a complication of blunt trauma to the chest in automobile accidents and secondary to cuffed endotracheal or tracheostomy tubes in patients requiring prolonged intubation and mechanical ventilation. The nonspecific nature of presenting symptoms makes the diagnosis sometimes difficult; the main classic finding of sudden coughing seconds after ingestion of fluids or solids (Ono's sign) should arouse suspicion of the diagnosis. The diagnosis (esophagotracheal or esophagobronchial communication) is established by means of contrast radiography. Esophagoscopy and bronchoscopy are not as accurate as radiography.40 3. Benign Esophageal Stricture Chronic reflux of acidic gastric contents leads to ulceration, inflammation, and, eventually, esophageal stricture. Reflux is the most common cause of benign stricture formation in the lower esophagus. The pathologic changes are reversible if the acidic gastric contents cease contact with the esophageal mucosa. Surgery may be necessary if medical treatment (diet, H2-blockers, antacids; see Table 16-5) and dilatations are inadequate.
599
There are two types of surgical repair, both of which are usually approached via a left thoracoabdominal incision. Gastroplasty after esophageal dilatation interposes the fundus of the stomach between esophageal mucosa and the acidic milieu of the stomach. The remaining fundus may be sewn to the lower esophagus to create a valve-like effect. The second type of repair is resection of the stricture and the creation of a thoracic end-to-side esophagogastrostomy. Vagotomy and antrectomy are performed to eliminate stomach acidity, and a Roux-en-Y gastric drainage procedure is performed to prevent alkaline intestinal reflux. 4. Achalasia Achalasia is a rare motor disorder characterized by almost complete disruption of the primary peristalsis function of the esophagus. The neurologic defect that is responsible for the development of achalasia remains poorly understood. Consistent pathologic findings include degeneration of ganglion cells in the myenteric plexus, occasionally accompanied by chronic inflammatory cell infiltrates, and loss of nerves innervating the smoothmuscle cells of the lower esophageal sphincter.41 Esophageal manometry remains the best means for diagnosing achalasia. Initial therapy can include either pneumatic dilation or esophagomyotomy. Symptomatic improvement occurs in 71 per cent of patients after pneumatic dilation, with a risk of perforation of 1.4 per cent.41 Surgical procedures for achalasia can be performed through either an abdominal or a thoracic incision. Nearly all authors favoring an abdominal approach add an antireflux operation to esophagomyotomy, whereas many authors advocating a transthoracic esophagomyotomy believe that an antireflux wrap is unnecessary. Overall results for the various surgical approaches used as initial therapy are excellent, with symptomatic improvement in 89 per cent of patients, a mortality rate of less than 1 per cent, and development of gastroesophageal acid reflux in less than 10 per cent.41 5. Esophageal Rupture and Perforation A rupture is a bursting injury and may be a spontaneous event as from uncoordinated vomiting, straining associated with weightlifting, childbirth, and defecation, or it may be externally caused as in crush injuries to the chest and abdomen. These ruptures are due to a sudden increase in abdominal pressure with a relaxed lower esophageal sphincter and an obstructed esophageal inlet, and the tear is usually located within 2 cm of the gastroesophageal junction. In contrast to a perforation, in the presence of a rupture, the stomach
600
Anesthesia for Esophageal Surgery
contents enter the mediastinum under high pressure and the patient becomes symptomatic from mediastinitis much more abruptly (i.e., septic syndrome). Perforations are usually unaccompanied by high intraluminal pressure and are usually due to iatrogenic instrumentation (the most common cause) (the origin of an iatrogenic tear may be either intraluminal or extraluminal; see Table 16-10), trauma, or ingestion of a perforating substance/ material. Causes of iatrogenic esophageal perforation include endoesophageal intubation with any catheter/instrument (e.g., endotracheal tubes, laryngoscopes, esophageal obturator airways, esophageal balloon tamponade, and nasogastric tubes). Table 16-10 lists the causes of esophageal rupture and perforation. The relative frequency of these causes of esophageal rupture and perforation from reported series is depicted in Figure 16-3,42 and the relationship of cause to location is shown in Figure 16-4.42 In 60 per cent of patients, there is esophageal disease (benign stricture [most common cause], diverticulum, tumor, achalasia) underlying the perforation.42 a. CLINICAL/DIAGNOSTIC FEATURES
There is a sharp rise in mortality rates when treatment is delayed. The mortality rate for patients treated in fewer than 24 hours is near 10 per Table 16-10 CAUSES OF ESOPHAGEAL RUPTURE AND PERFORATION* I. Instrumentation/iatrogenic A. Intraluminal 1. Esophagoscopy 2. Bougienage 3. Pneumatic dilation 4. Sclerosis of esophageal varices 5. Placement of intraesophageal tubes (nasogastric, Sengstaken-Blakemore, prostheses) 6. Endotracheal tube placement B. Extraluminal 1. Mediastinoscopy 2. Intraoperative injury a. Thyroid resection b. Leiomyoma enucleation c. Proximal gastric vagotomy d. Pneumonectomy 3. Radiation therapy II. Traumatic A. Blunt B. Penetrating C. Ingestion of caustic substance III. Spontaneous (barogenic) A. Postemetic B. Straining (weightlifting, bowel movement, childbirth) IV. Ingestion of foreign body V. Tumor VI. Surrounding infections *Based on data from Jones and Ginsberg.42
cent, but it rises sharply to greater than 50 per cent for those treated later.43"*5 The major reason for this sharp increase in mortality rate is the rapid development of necrotizing mediastinitis, combined with the inability to close the perforation surgically. Because of the propensity to develop mediastinitis, thoracic perforations have three times the mortality of cervical perforation,45 and spontaneous ruptures are much more lethal than perforations because of the explosive contamination of the mediastinum as well as the delay in diagnosis, because they most often occur unobserved outside the hospital and may be clinically silent initially. Cervical tears may be due to either instrumentation or trauma and occur at the level of the upper esophageal sphincter (the cricopharyngeal muscle), which is the narrowest point in the esophagus (and therefore this is the area where foreign bodies most frequently impact in children). In addition, the upper esophagus is compressed by the cervical vertebrae (exacerbated by hyperextension of the neck) and at the level of the C-5 and C-6 vertebrae, the posterior esophageal mucosa is covered only by fascia. The early signs of cervical esophageal perforation include neck stiffness and a dull neck ache, regurgitation of blood material, and the finding of cervical subcutaneous emphysema. Inflammatory changes in the neck may not develop for several hours; signs of systemic sepsis often do not occur for up to 24 hours. In contrast, perforations of the thoracic esophagus result directly in mediastinal contamination, leading to a more rapid development of pneumomediastinum and mediastinitis than after cervical perforations. The thin mediastinal pleura is usually ruptured by the inflammatory process, producing contamination of the pleural space and a pleural effusion. Gastric contents and fluids are then drawn into the pleural space by the negative intrathoracic pressure, resulting in further inflammation and fluid sequestration, hypovolemia, and the early appearance of tachycardia and systemic sepsis. Chest pain and subcutaneous emphysema are usually present, and dyspnea is often prominent even in the absence of pneumothorax. Intra-abdominal esophageal perforations occur into the free peritoneal cavity and result in peritonitis. A dull, retrosternal ache in association with epigastric pain radiating to the shoulders is characteristic because of the relationship of the intraabdominal esophagus to the diaphragm. Systemic signs such as tachycardia, tachypnea, and fever develop early; progression to sepsis and shock occurs within hours. Although contrast esophagograms are the stan-
Anesthesia for Esophageal Surgery
601
Figure 16-3 04) Cause of 511 cases of esophageal perforation in recent series and (B) types of instrumental perforations among those cases. Other causes of iatrogenic esophageal perforation include endoesophageal intubation with any catheter/instrument such as endotracheal tubes, laryngoscopes, esophageal obturator airways, esophageal balloon tamponade, and nasogastric tubes. Data are taken from References 49, 50, 51, 72, 112 and 113 of Jones (my Reference 42).
dard diagnostic procedure in cases of suspected esophageal perforation, it should be noted that the false-negative rate of these examinations can exceed 10 per cent. Thus, "negative" studies may not completely eliminate the possibility of a perforation. When positive, contrast studies have the advantage of demonstrating the level of the perforation and the presence of extension into the pleural cavity. In the patient in whom esophageal perforation is highly suspected clinically but contrast esophagograms are negative, flexible esophagoscopy and computed tomography can be useful diagnostic adjuvants.
b. TREATMENT OF ESOPHAGEAL PERFORATIONS
Once the diagnosis of esophageal perforation is established, multiple factors must be considered in selecting the appropriate therapy. The cause of the perforation, its location, the presence of underlying esophageal disease, and the interval between perforation and diagnosis are critical factors in the determination of treatment. In addition, the condition of the esophagus and the extent of soilage or injury to adjacent organs and tissues as well as the age and general condition of the patient must also
Figure 16-4 Relationship of cause to location of esophageal perforations in recent series.
be considered. Treatment options for esophageal perforations are shown in Table 16-11. The surgical approach is dependent on the location of the perforation. Cervical esophageal tears are best exposed through an incision parallel to the left sternocleidomastoid muscle and anterior to the carotid artery and internal jugular vein. Perforations in the upper two thirds of the thoracic esophagus can be reached through a right posterolateral thoracotomy in the fourth or fifth intercostal space, whereas those in the lower third are best approached through a posterolateral thoracotomy in the left sixth or seventh interspace. Abdominal esophageal injuries usually require an upper midline laparotomy. Once the esophagus is exposed, localization of subtle perforations may require instillation of methylene blue into the esophageal lumen or the insufflation of air into the esophagus immersed in saline solution. Figure 16-5 shows the variety of diversion, drainage, and feeding
Table 16-11 VARIOUS POSSIBLE TREATMENTS OF ESOPHAGEAL PERFORATION 1. Operative management a. Primary closure b. Primary closure reinforced with tissue (pleural flap, omentum, intercostal flap, pericardial fat, diaphragm, gastric wall) c. Resection (transhiatal or transthoracic) d. Drainage alone e. T-tube drainage (creating esophagocutaneous "venting" fistula) f. Exclusion and diversion (surgical closure of proximal and distal esophagus with cervical esophagostomy and gastrostomy drainage tubes) g. Intraluminal stent (for carcinoma cases) 2. Nonoperative management
tubes a patient with an esophageal tear may require. The following guidelines have been suggested for selecting nonoperative treatment: (1) clinically stable patients; (2) instrumental perforations detected before major mediastinal contamination has occurred or perforations with such a long delay in diagnosis that the patient has already demonstrated tolerance for the perforation without the need for surgery; and (3) esophageal disruption that is well contained within the mediastinum or a pleural loculus. Nonoperative management is much more feasible in children because mediastinal tissues seem more resistant and esophageal perforations appear more contained in children than adults.46 Nonoperative management consists of nothing by mouth, total parenteral nutrition, broad-spectrum antibiotics, and some of the various tubes shown in Figure 16-5. Final outcome is related to cause (e.g., spontaneous ruptures have a worse prognosis compared with iatrogenic instrumental causes), underlying disease (e.g., underlying carcinoma has a much worse prognosis than benign disease), location (e.g., thoracic and abdominal tears have a much worse prognosis than a cervical tear), and type and timing of treatment (late treatment has a much worse prognosis than early treatment). G. Surgical Approaches/Incisions for Esophageal Surgery (With Emphasis on Esophageal Carcinoma)47
The distribution of esophageal carcinoma between the upper, middle, and lower thirds of the esophagus is approximately 17 to 27 per cent, 40
Anesthesia for Esophageal Surgery·
603
Figure 16—5 Diagram of esophagogastric-intestinal temporary diversion, drainage, and feeding tubes.
to 54 per cent, and 20 to 30 per cent, respectively.27· 48~50 Resectional surgery of the midthoracic esophagus is technically very complicated (modern immediate mortality is high but with a wide reported range of approximately 3-20 per cent). 25 · 28 ' 45 · 51-53 The main causes of mortality are anastomotic leaks and cardiorespiratory complications. Surgery for midthoracic esophageal lesions usually involves an additional proximal cervical incision (to approach the upper third of the esophagus) and a distal abdominal incision (to approach the lower third of the esophagus and/or mobilize the stomach). The midthoracic esophagus can be approached through either left-sided or right-sided thoracotomy. Tumor resection is performed through the thoracotomy. The stomach is then pulled up into the chest and anastomosed to the proximal esophagus. For esophageal tumors involving the stomach, an esophagogastrectomy is performed, and the jejunum is anastomosed to the proximal esophagus. Advantages of a left thoracotomy include the ability to split open the left diaphragm, mobilize the stomach, and bring it up into the chest without having to change the patient's position. A right thoracotomy has the advantage of good exposure because the midthoracic portion of the esophagus, except for its most distal part, is primarily on the right side of the mediastinum. Also, if the azygos vein is infiltrated by tumor, it is safer to dissect it from the right side. The disadvantages of a right thoracotomy are that the position of the patient must be changed after the abdominal part of the operation (per-
formed first) is completed. In addition, after an extensive procedure in the abdomen and an opening of the chest, the lesion may be found to be nonresectable. Finally, esophagectomy can be performed without thoracotomy through an abdominal transhiatal approach (requires only abdominal and neck incisions and the esophagus is freed by blunt finger dissection). The purported advantages of this approach are that it (1) is associated with decreased cardiopulmonary mobidity and (2) places the anastomosis in the neck. The surgical literature is very controversial with respect to the appropriate surgical approach to esophageal cancer. The issues involved in the controversy are whether the goal is cure versus palliation (i.e., whether the resection is complete and includes lymph nodes; the transthoracic approach can potentially be curative, whereas the transhiatal approach is only palliative), ease of taking care of postoperative anastomotic leaks (transthoracic approaches usually result in an intrathoracic anastomosis; an intrathoracic anastomotic leak is much more difficult to take care of than a cervical anastomotic leak and has a 40 per cent mortality),54 the length of surgery and the amount of intraoperative bleeding (transhiatal pull-through is blind and may be associated with mediastinal bleeding), and postoperative esophageal motility. Esophageal surgery involves a number of other controversies, including questions of delayed versus immediate reconstruction and the use of colon or stomach to restore continuity.49 Without quoting a large number of references, it is my impression that the left thoracoabdominal incision is most commonly used today.
604
Anesthesia for Esophageal Surgery
Unfortunately, it is not the operation but rather the stage and biological behavior of the tumor at the time of esophagectomy that determine sur vival. Only about 10 per cent of patients have stage I disease (tumor that is superficial and con fined to the mucosa and submucosa and therefore potentially curable by virtually any extirpative technique).26 Perhaps in the 10 to 25 per cent of the patients with stage II disease, a traditional "cancer operation"—transthoracic esophagec tomy with regional lymph node dissection—offers a chance for better survival than does the transhia tal approach, but this has by no means been estab lished. The finding of mediastinal lymph node me tastases from esophageal carcinoma carries an ominous prognosis. The 5-year survival for stage III tumors without lymph node metastases (T3, NO, MO) was 34 per cent compared with 8 per cent for patients with lymph node involvement (T3, Nl-3, MO).53 Unfortunately, the majority of pa tients with esophageal carcinoma have mediastinal lymph node metastases, and that is the primary determinant of their survival. Overall, the 5-year survival rate is 10 to 15 per cent (Fig. 16-6)26 but has ranged as high as 54 per cent. 2 6 · 5 5 · 5 6 The preponderance of evidence indi cates that combined modality therapy (surgery, chemotherapy, and radiation) slightly and moder ately increases survival (from 5-10 per cent to ΙΟ Ι 5 per cent) compared with surgery alone.57-59 Ob viously, there is agreement that we need to find a way to recognize esophageal cancer early. Encour aging reports from China, where carcinoma of the esophagus is endemic (there is a 50-fold greater incidence of esophageal carcinoma in China com pared with the United States), have described mass screening techniques and early esophageal resec tions.60 The resectability rate was 100 per cent, hospital mortality was 2.5 per cent, and the 5-year survival rate was 85.9 per cent.
56 CONSIDERED POTENTIALLY RESECTABLE
As with obstruction of the tracheobronchial tree by lung carcinoma, obstruction of the esophagus by esophageal carcinoma may be palliated (unob structed) by endoscopic laser61 and photodynamic therapy.62 With these therapies, malignant stenoses have been recanalized or widened in all patients, and the ability to swallow was improved in about 80 per cent with no procedure-related deaths.61·62 Other forms of palliation of a totally obstructed esophagus, but with significant procedure-related mortality, are by-pass operations (with a hospital mortality of 20 per cent) and esophageal dilation and intubation (stenting with a catheter; 30-day mortality is 10 per cent, with much of it resulting from perforation38).63· M With all forms of pallia tion (laser therapy, bypass, intubation) median sur vival is approximately 4-7 months.62
II. INTRAOPERATIVE ANESTHETIC CONSIDERATIONS A. Monitoring In addition to all the considerations discussed in chapter 7, an arterial catheter is indicated if (1) one-lung ventilation is to be used, (2) blood loss will be significant, (3) the heart will be manipu lated (by retraction or passage of gastrointestinal tract behind the heart), (4) an esophagorespiratory fistula exists, and/or (5) chronic lung disease is present. Central venous access is required for mon itoring the effects of manipulating the heart (see later discussion), preload measurement, blood loss replacement, and drug administration.
B. Induction of Anesthesia Because patients presenting for esophageal sur gery may be at risk for aspiration, either an awake
49 SURVIVE THE OPERATION
20-30 ALIVE AT 1 YEAR
10-20 ALIVE AT 5 YEARS
7 DEATHS FROM PERIOPERATIVE COMPLICATIONS
100 PATIENTS
44 UNRESECTABLE
Figure 16-6 The fate of 100 patients with esophageal carcinoma. Statistics are based on treatment outcome from I20l articles published between 1980 and 1988 and include the use of adjuvant chemotherapy and radiation. (From Muller JM, Erasmi H, Stelzner M, et al: Surgical therapy of oesophageal carcinoma. Br J Surg 77:845-857, 1990. Used with permission.)
Anesthesia for Esophageal Surgery
intubation or a rapid-sequence induction with cricoid pressure may be indicated. However, rarely, neck, trachea, or an esophageal foreign body may prevent the application of cricoid pressure (necessitating an awake intubation).65 Additionally, a patient with mediastinal lymphadenopathy may have tracheal compression and collapse of the airway with the onset of muscle relaxation. Ventilation may be possible only by passage of an endotracheal tube beyond the obstruction (see chapter 15, Mediastinal Mass).
C. Airway Considerations 1. General Airway Considerations There are several important airway considerations that are generally related to the performance of esophageal surgery. First, for lower esophagogastric resections/procedures via a left thoracoabdominal incision or an abdominal incision, it is not necessary to collapse the left lung using a doublelumen endobronchial tube. A single-lumen endotracheal tube can be placed, and surgical exposure can be obtained by gentle retraction of the left lung. For esophageal surgery via a standard thoracotomy alone, it is usually necessary to place a double-lumen endobronchial tube or bronchial blocker tube to collapse the ipsilateral lung. However, a patient with relatively normal lungs undergoing esophageal surgery may be more likely to experience hypoxemia during one-lung ventilation than a patient presenting for lung resection. This is because the patient presenting for lung surgery may already have limitation of blood flow to the diseased lung and, thus, less ventilation/perfusion mismatching during one-lung anesthesia. Also, during lung resection, the surgeon ligates the pulmonary artery or a branch thereof, which decreases the shunt. Second, as large a tracheal tube (either doublelumen tube, single-lumen tube, Univent bronchial blocker tube) as possible should be used if the patient has chronic lung disease in order to effectively suction secretions. Chronic lung disease in these patients is due to chronic aspiration, and the secretions are frequently copious. In addition, patients with chronic aspiration have a greater risk of intraoperative bronchospasm secondary to increased airway reactivity caused by chronically irritated airways and may require bronchodilation ^ 2 -agonist, adequate anesthesia). Third, in thoracotomy cases, one-lung ventilation (as created by either a double-lumen tube or bronchial blocker) will facilitate exposure of the esophagus. However, when the thoracotomy is fol-
605
lowed by a cervical anastomosis, it should be remembered that a very large tracheal tube or overinflated cuff may cause the posterior membrane of the trachea to bulge into the anterior surface of the esophagus and interfere with the surgical field. Fourth, high concentrations of nitrous oxide are contraindicated with bowel present in the chest because the resultant bowel distention may cause respiratory impairment and possible interference with surgical exposure. Fifth, patients who have chronic lung disease and copious secretions should be mechanically ventilated and extubated only when they demonstrate an adequate cough. If the patient is to remain intubated at the end of the procedure in which a double-lumen endobronchial tube has been used, reintubation with a single-lumen endobronchial tube must be done very carefully (see Jet Stylet Method, chapter 9). If the neck is going to be fixed in a hyperflexed position (either by mandibularmanubrium wires or chin to sternal skin sutures) to avoid/minimize tension on an esophageal anastomosis,66·67 then the single-lumen tube should be in place before fixation of the neck in a hyperflexed position (placement of the single-lumen tube may require neck extension). Sixth, in nonthoracotomy cases (without a chest drain), a chest X-ray should always be obtained immediately postoperatively to exclude a pneumothorax from inadvertent entry into either pleural cavity. Finally, after extubation, aspiration may occur because of recurrent laryngeal nerve injury from cervical dissection.
2. Airway Considerations in Cases of Esophagorespiratory Fistula™ Anesthetizing the patient with an esophagorespiratory tract fistula carries with it unique considerations, distinguishing it from other types of esophageal surgery. The level of the fistula should be identified preoperatively (esophography or endoscopy) to determine the most appropriate tracheal tube (double-lumen tube, single-lumen tube, bronchial blocker) (see later discussion). If the fistula is very large, positive-pressure ventilation may result in massive abdominal distention (resulting in secondary respiratory insufficiency and cardiovascular compromise), and loss of inspired gas (with resultant inadequate ventilation) through a gastric tube (if present). Thus, spontaneous ventilation should be maintained, if possible. Awake intubation is, therefore, an appropriate method of securing the airway, if possible, in the presence of a very large fistula. If the fistula is in the mid to proximal trachea, a single-lumen tube can be used with the tip placed distal to the fistula so that the cuff of the endotracheal tube may oc-
606
Anesthesia for Esophageal Surgery
elude the fistula. If the fistula is in the distal trachea or main-stem bronchus, a double-lumen tube will be necessary to isolate the fistula. A rightsided double-lumen tube should be used for a fistula in the distal trachea, left main-stem bronchus, or left lung, and a left-sided double-lumen tube should be used for a fistula in the distal trachea, right main-stem bronchus, or right lung because, in either case, the endobronchial cuff guarantees isolation of the fistula (Fig. 16-7). If the site of the fistula cannot be determined preoperatively, a right-sided double-lumen tube should be placed because, statistically, the fistula is most likely to communicate with the trachea or left main-stem bronchus. Right-sided ventilation should be attempted first. Either gastric distention or loss of delivered tidal volume indicates presence of a right-sided fistula, and left-lung ventilation should be used instead. If the patient must be generally anesthetized before intubation, spontaneous ventilation should be maintained until gentle (low positive pressure) ventilation by mask has been shown to provide effective ventilation. If the stomach distends with positive-pressure ventilation, then it may need to be vented (nasogastric tube, gastrostomy) (Fig. 16-8). If venting the stomach results in loss of
tidal volume, then appropriate one-lung ventilation must be instituted (see prior discussion). The amount of tidal volume loss is a function of size of the fistula, compliance of the lungs, peak inspiratory pressure, and whether there is a stomach tube or not. The lung on the side of the fistula is likely to have a decreased compliance; consequently, the contralateral lung tends to receive most of the tidal volume. When one-lung ventilation is used and chronic lung disease in the ventilated lung prevents adequate arterial oxygenation, operative lung continuous positive airway pressure (5-10 cm H20) is indicated. When two-lung ventilation is used after the esophageal portion that contained the fistula has been excluded (i.e., the creation of a blind esophageal pouch), the excluded portion must be drained to protect against proximal or distal pouch suture line disruption resulting from distention of the pouch by air delivered through the fistula during positive-pressure ventilation (Fig. 16-9). If the pouch is not drained, it should not be exposed to positive-pressure ventilation (i.e., use one-lung ventilation, spontaneous ventilation, block fistula, and so on).68,69 In view of these considerations, spontaneous ventilation at the end of the procedure is desirable.
Isolation of Esophago-Respiratory Tract Fistula By Double-Lumen Tube Intubation Fistula in Distal Trachea, Left Mainstem Bronchus or Lung
Fistula in Distal Trachea, Right Mainstem Bronchus or Lung
Figure 16-7 Either a right- or left-sided double-lumen tube can be used to isolate a distal tracheal fistula. A left-lung fistula is isolated by a right-sided double-lumen tube, and a right-lung fistula is isolated by a left-sided double-lumen tube.
Anesthesia for Esophageal Surgery 6 0 7
Figure 16-8 Algorithm for the management of the airway during the induction of anesthesia in a patient with an esophagorespiratory tract fistula.
Surgical Treatment of Esophago-RespiratoryTract Fistula
Figure 16-9 Exclusion of malignant esophagorespiratory fistula procedure.
608
Anesthesia for Esophageal Surgery
In this circumstance, meticulous suctioning of se cretions, avoidance of sedation, and provision of appropriate analgesia are not only desirable but necessary. However, the timing of extubation should be conservative if chronic lung disease and secretions are present.
D. Hemodynamic Considerations Two major hemodynamic considerations are unique to major esophageal surgery. First, trans thoracic or transhiatal tumor resection may lead to surgical compression and/or manipulation of vital organs (vena cava, heart), as may passing of distal gastrointestinal tract (stomach, small intestine, co lon) retrosternally into the chest. The compression/ manipulation of the central cardiovascular struc tures may cause arrhythmias, decreased cardiac output, hypotension, and perhaps myocardial is chemia (in patients with pre-existing coronary ar tery disease).70· 7I Each time the surgeon's hand enters the thorax, it may be expected that the mean arterial pressure and cardiac index will decrease by an average of 46 per cent (Figs. 16-10 and Ιο ί 1 ), and premature ventricular contractions occur in 50 per cent of patients.71 Pulmonary artery occluded pressure signifi
cantly increases with each intrathoracic manipula tion (see Figs. 16-10 and 16-11). All hemody namic parameters return to control values once manipulation is stopped (see Figs. 16-10 and lo ll). The simultaneous decrease in mean arterial pressure and cardiac index are clearly related to impaired venous return and/or ventricular ejection from manual compression. From these findings, it can be assumed that long-lasting alterations in hemodynamic status could be detrimental to patients with cardiac dis ease. Therefore, careful intraoperative hemody namic monitoring is advisable. The obvious re sponse to the mechanical/surgical interference of cardiac function is to terminate the offending stim ulus (see Figs. 16-10 and 16-11), but in situations in which an arrhythmia is prolonged, antiarrhythmia drugs may need to be administered. If the electrocardiogram indicates ischemia, nitroglyc erin and calcium channel-blocking drugs (for cor onary artery vasospasm) may need to be adminis tered. Second, the average blood loss associated with esophagectomy is approximately 2000 ml.72 The negative hemodynamic consequences of manipu lation of the heart described previously will be greatly magnified if the patient is concomitantly hypovolemic. The preload monitoring suggested
HEMODYNAMIC CHANGES OBSERVED DURING MANUAL DISSECTION Figure 16-10 A typical recording of the hemodynamic events observed during intrathoracic esophageal resection. " I n " repre sents the moment the surgeon's hand entered the posterior mediastinum. " O u t " represents the moment the surgeon's hand was removed. Total time elapsed between the two points in and out was 55 sec in this case. (A.P. = aortic pressure; P.C.W.P. = pulmonary capillary wedge pressure; ECG = electrocardiogram.) (From Yakoubian K, Bougeois B, Marty J, et al: Cardiovascular responses to manual dissection associated with transhiatal esophageal resection. J Cardiothorac Anesth 4:458—461, 1990. Used with permission.)
Anesthesia for Esophageal Surgery
609
ing air to define the esophageal anatomy better. Air may also be injected through a retrograde gas troesophageal catheter. Similarly, to test the integ rity of the anastomosis or repair, methylene blue may need to be injected via a naso- or oroesophageal tube, the lower end of which has been care fully positioned by the surgeon in relation to the anastomosis. Finally, it may be necessary to pass bougies per os to dilate strictures and/or define anatomy. Dila tions may progressively increase in size to the 50 to 60 French range. The final bougie may be used as a stent around which to perform a fundoplication. Whenever a catheter or bougie is passed into the esophagus and resistance at the esophageal in let is felt, which may be due to collapse of the esophagus by the tracheal tube and/or its cuff, it may be necessary to pull the trachea anteriorly by grabbing it at the right and left lateral margins (if the neck is not sterile) or by transiently deflating the cuff of the tracheal tube.
III. POSTOPERATIVE CONSIDERATIONS
Figure 16-11 Changes in mean arterial pressure (MAP), cardiac index (CI), and pulmonary capillary wedge pressure (PCWP) during intrathoracic manipulations, -p < .05 versus before. (From Yakoubian K, Bougeois B, Marty J, et al: Car diovascular responses to manual dissection associated with transhiatal esophageal resection. J Cardiothorac Anesth 4:458461, 1989. Used with permission.)
The major routine postoperative problems con sist of obtaining satisfactory gas exchange and, related to gas exchange, postoperative analgesia. Postoperative mechanical ventilation is discussed in chapter 20 and postoperative analgesia in chap ter 21. The major nonroutine complication of esophageal surgery is leakage from an esophageal to gastrointestinal tract anastomosis. Leakage from a cervical anastomosis can usually be easily han dled by incision and drainage. Leakage from an intrathoracic anastomosis is the most feared com plication of esophageal surgery and carries a 50 per cent mortality rate. The treatment options are listed in Table 16-11, and Figure 16-5 shows the variety of diversion, drainage, and feeding tubes that are often required to deal with an intrathoracic anastomotic leak.
REFERENCES previously and serial hemoglobin and hematocrit determinations allow fairly precise replacement of blood and fluid losses.
E. Esophagogastrointestinal Considerations The anesthesiologist may be asked to insert a nasogastric tube (the exact distance is usually de termined by the surgeon) for the purpose of inject-
1. Hardy JD: Diseases of the esophagus: An overview. In Hardy JD (ed): Textbook of Surgery. Philadelphia, JB Lippincott Co, chapter 37, 1977. 2. Roa TLK, El-Etr AA: Esophageal and mediastinal surgery. In Kaplan JA (ed): Thoracic Anesthesia. New York. Churchill Livingstone, 1983, chapter 14, pp ΑΑΊ-ΑΊΑ. 3. Ruberg RL, Dudrick SJ: Intravenous hyperalimentation in head and neck tumor surgery: Indications and precautions. Br J Plast Surg 30:151 -153, 1977. 4. Weiss SM: Nutritional aspects of preoperative manage ment. Med Clin North Am 71:369-375, 1987. 5. Reinhardt GF, DeOrio AJ, Kaminski MV Jr: Total paren teral nutrition. Surg Clin North Am 57:1283-1301, 1977. 6. Askanazi J, Carpentier YA, Elwyn DH, et al: Influence of
CHAPTER
17
Anesthesia for Emergency Thoracic Surgery I. Introduction II. Massive Hemoptysis A. General Considerations B. Surgical Considerations C. Anesthetic Considerations 1. Preoperative Considerations a. Prevent Asphyxiation b. Prevent Contamination of Normal Lung c. Correct Hypovolemia 2. Intraoperative Considerations III. Thoracic Aortic Aneurysms and Dissections/Disruptions A. General Considerations 1. Spontaneous Dissection of Thoracic Aortic Aneurysms 2. Traumatic Disruption 3. Signs and Symptoms of Any Type (Traumatic or Spontaneous) of Dissection B. Surgical Considerations C. Anesthetic Considerations 1. Preoperative Considerations 2. Intraoperative Considerations IV. Bronchopleural Fistula A. General Considerations B. Surgical Considerations C. Anesthetic Considerations V. Lung Abscesses and Empyema A. General Considerations B. Surgical Considerations C. Anesthetic Considerations VI. Chest Trauma A. Overview of the Management of the Patient With Extensive Trauma 1. Primary Survey a. Airway and Ventilation b. Circulation c. Neurologic Status d. Mechanism of Injury 2. Secondary Survey 3. Surgical Priorities B. Specific Chest (Noncardiac) Injuries: General Considerations 1. Chest Wall Fractures (Flail Chest) 2. Pleural Space (Hemothorax, Pneumothorax) and Pulmonary Parenchyma (Contusions)
612
a. Hemothorax b. Pneumothorax c. Pulmonary Contusion 3. Tracheobronchial Disruptions 4. Esophageal Disruptions 5. Diaphragmatic Disruptions C. Specific (Noncardiac) Injuries: Surgical Considerations 1. Chest Wall Fractures (Flail Chest) 2. Pleural Space (Hemothorax, Pneumothorax) and Pulmonary Parenchyma (Contusions) a. Hemothorax b. Pneumothorax c. Pulmonary Contusion 3. Tracheobronchial Disruptions 4. Esophageal Disruptions 5. Diaphragmatic Disruptions D. Specific Chest (Noncardiac) Injuries: Anesthetic Considerations 1. Chest Wall Fractures (Flail Chest) 2. Pleural Space (Hemothorax, Pneumothorax) and Pulmonary Parenchyma (Contusions) a. Hemothorax b. Pneumothorax c. Pulmonary Contusion 3. Tracheobronchial Disruptions 4. Esophageal Disruptions 5. Diaphragmatic Disruptions VII. Transvenous Pulmonary Embolectomy A. General and Surgical Considerations B. Anesthetic Considerations VIII. Emergency Room Thoracotomy in the Management of Trauma A. General Considerations B. Surgical Considerations C. Anesthetic Considerations IX. Removal of Tracheobronchial Tree Foreign Bodies A. General Considerations B. Surgical Considerations C. Anesthetic Considerations
Anesthesia for Emergency Thoracic Surgery
I. INTRODUCTION
Table 17-1 CAUSES OF MASSIVE HEMOPTYSIS*!
The vast majority of cases requiring emergency thoracic surgery consist of massive hemoptysis, thoracic aortic aneurysms and dissections/disruptions, bronchopleural fistula (BPF), lung abscess and empyema, chest trauma (chest wall fractures, pulmonary parenchymal contusions, tracheobronchial disruption, and esophageal and diaphragmatic trauma), and tracheobronchial tree foreign bodies. Since chest trauma can cause massive hemoptysis and thoracic aortic disruptions and necessitate emergency room thoracotomy, it is obvious that this classification is an arbitrary one. However, since each of these indications for emergency thoracic surgery commonly occurs independently of chest trauma, they warrant separate consideration.
I. Infection (45-90%) Tuberculosis Bronchiectasis Bronchitis Lung abscess Necrotizing pneumonia II. Neoplasm (7-19%) Bronchogenic carcinoma Metastatic carcinoma Endobronchial polyp III. Cardiovascular disease Mitral stenosis Pulmonary arteriovenous malformation Pulmonary embolus Pulmonary vasculitis IV. Miscellaneous causes Pulmonary artery catheterization Exploratory needling Cystic fibrosis (5-10%) Pulmonary contusion, laceration, aneurysm Reperfusion of pulmonary vasculature after pulmonary embolectomy and after cardiopulmonary by-pass
II. MASSIVE HEMOPTYSIS A. General Considerations Massive hemoptysis is uncommon, occurring in less than 0.5 per cent of patients admitted to a large pulmonary medicine service.1 It has been arbitrarily defined, on the basis of the amount of daily volume of blood expectorated, as 200 ml,2 more than 300 ml,3 more than 500 ml,4 more than 600 ml in 24 to 48 hours,1·5"7 more than 600 ml within 16 hours,5 and more than 1000 ml in 24 hours.8·9 However, others evaluated hemoptysis not in terms of rate of bleeding but from the standpoint of its threat to vital functions. Thus, a life-threatening hemoptysis was one that caused acute airway obstruction or hypotension severe enough to require blood transfusion (although, in general, the greater the rate of bleeding, the greater the mortality).1·8· I0 In all of these reports, massive hemoptysis often occurred abruptly, unexpectedly, and without prodromal symptoms. The differential diagnosis of massive hemoptysis is shown in Table 17-1."· I2 On occasion, aspirated and/or swallowed epistaxis and hematemesis can be mistaken for hemoptysis. The large majority of reported cases of massive hemoptysis have had a chronic infectious cause."· I3 14 · This is because chronic inflammation leads to profuse vascularization of the high-pressure bronchial artery system. Subsequently, any erosion or rupture of enlarged bronchial arteries will result in massive hemoptysis. Bronchitis is the most common, active tuberculosis the second most common, and bronchiectasis the third most common infection causing massive hemoptysis. 12~14 Of patients
613
*Based on data from Garzon & Gourin.1 Yeoh et al.,2 Stern et al.,3 Ehrenhaft & Taber,4 Crocco et al.,s Gourin & Garzon,h Conlan and Hurwitz,7 Corey & Hla,* Muthuswamy et al.." Thorns et al.,10 Wedel," Bone,12 Johnston & Reisz,13 Wedzicha & Pearson,14 Panos et al.,15 Cervenko et al.,'6 and Rice et al." tRange of percentage of incidence is given for the major causes.
with pulmonary tuberculosis, 1 to 5 per cent will have massive hemoptysis.9 The majority of the remaining causes of hemoptysis are due to bleeding neoplasms. Although the major cause of spontaneous bleeding with neoplasms is the same as with the infectious causes (erosion into bronchial arteries), in present-day clinical practice massive hemoptysis from neoplasms increasingly occurs during diagnostic fiberoptic bronchoscopic manipulations of exophytic airway tumors. Hemoptysis from neoplasm has a worse prognosis than hemoptysis from infection/ The prognosis is especially poor in carcinoma patients if there is concomitant fungal infection (36 per cent motility).15 A mechanism for hemoptysis in both cancer and fungal infections, in addition to direct invasion of blood vessels, is distal ischemia and avascular necrosis.1,5 Massive hemoptysis may also occur when the low-pressure pulmonary circulation is transected or ruptured, as may occur with accidental trauma with sharp or blunt objects and pulmonary artery catheter trauma. The likelihood of massive endobronchial hemorrhage following pulmonary artery catheterization is increased by subsequent heparinization.16 I7 Thus, massive hemoptysis may arise from either the pulmonary or the systemic circulation.
614
Anesthesia for Emergency Thoracic Surgery
Death from massive hemoptysis usually results from asphyxiation, rarely from exsanguination. However, the amount of blood that has actually been lost may be seriously underestimated based on the history since at least some of the coughedup blood is swallowed. In addition, the accurate measurement of expectorated blood is frequently difficult. B. Surgical Considerations In a series of 55 pulmonary resections performed for massive hemoptysis (600 ml/16 hours),5 there was a mortality of 18 per cent reported; this was markedly better than with conservative treatment, which resulted in a mortality of 75 per cent in patients who bled 600 ml or more in 16 hours and of 54 per cent in those who bled 600 ml or more in 48 hours. 5,6 However, routine use of surgery has been debated because other authors have found that somewhat lesser degrees of hemoptysis may be successfully managed conservatively regardless of the amount of bleeding in the first 24 hours.2·9· l8· l9 The intensive care unit medical protocol used in one institution is shown in Table 17-2.9 Nevertheless, surgery (resection) is probably indicated in patients who require multiple transfusions, in those in whom bleeding results in progressive impairment of pulmonary function (aspiration should be evaluated by freTable 17-2 INTENSIVE CARE UNIT MEDICAL PROTOCOL FOR MASSIVE HEMOPTYSIS* 1. 2. 3. 4. 5. 6.
Complete bed rest Postural management Nothing per mouth Establishment of large-bore intravenous catheters Correction of abnormalities of coagulation Oxygen therapy to keep arterial oxygen pressure at 75 mm Hg range 7. Accurate record of volume and rate of hemoptysis 8. Antituberculosis chemotherapyt 9. Broad-spectrum antibioticst 10. Judicious transfusion and fluid therapy 11. Cough suppression (codeine and similar drugs) 12. Stool softeners (to avoid straining) 13. Coordination with thoracic surgeons and angioradiologists 14. Bronchoscopy (see Fig. 17-1) 15. Radionuclide scanning for localization selectively 16. Vasopressin intravenously§ *Based on data from Muthuswamy et al.9 tThis consists of four tuberculocidal drugs, including isoniazid, rifampin, streptomycin, and pyrazinamide. This combination is used because of the high incidence of primary drug resistance in their population of patients. $ Third-generation cephalosporins are not used to avoid complicating coagulopathies. §This may be used in those patients who are at high risk for invasive procedures or who are being prepared for surgery.
quent serial chest roentgenograms), and in those in whom hemoptysis persists for several days despite optimum medical management.19 Contraindications to surgery include inoperable carcinoma of the lung, inability to localize the bleeding site, and presence of severe bilateral pulmonary disease and systemic disease (debilitation). These patients are candidates for bronchial artery embolization (Fig. 17-1) (see later). The chest X-ray may contain strong clues (evidence of tuberculosis or opacifications) as to the source of bleeding. However, considering that blood can be, and probably has been, aspirated into the nonbleeding lung, it cannot be safely assumed that the chest X-ray pathologic findings correspond to the site of bleeding.20 Indeed, in many patients with massive hemoptysis, the chest X-ray may even be normal.21 Bronchoscopy during active bleeding is the single most important technique for determining the cause and location of bleeding and should be performed in all patients. The procedure should be done, if possible, in the operating room so that immediate resection can be performed (see Fig. 17-1).22 Most bronchoscopists will use a rigid bronchoscope because of the much greater suctioning and ventilating capability. However, the flexible fiberoptic bronchoscope may be used if there is no active bleeding and/or the site of bleeding is thought to be in the upper lobes. The combination of fiberoptic bronchoscopy through a rigid bronchoscope may be a safer alternative (see chapter 15). If localization is still uncertain after bronchoscopy and if the clinical situation permits, bronchial and pulmonary arteriography can be helpful. When bronchoscopy is coupled with selective angiography, a high rate of accurate localization can be achieved. When bronchoscopy has caused the bleeding (as in diagnostic procedures), the site of bleeding is known and the patient can proceed to definitive therapy (see later discussion) without delay. The endoscopist may be able to control bleeding and the spread of bleeding during bronchoscopy (see Fig. 17-1). Topical iced saline and vasoconstrictors can be administered through the bronchoscope to control bleeding, provided the bleeding is not so massive as to preclude visualization of the origin.7 Bronchodilator treatment should not be administered because these may have vasodilator actions and precipitate renewed bleeding. The spread of blood from one lung to the other can be controlled by the use of a bronchial blocker (Univent tube or balloon-tipped Fogarty catheter) in the main bronchus of the bleeding side or a gauze packing at the bleeding segment or side.22-26 Yttrium aluminum garnet (YAG) laser has proved efficacious in the treatment of hemoptysis in pa-
Anesthesia for Emergency Thoracic Surgery-
615
Figure 17-1 Treatment algorithm for massive hemoptysis. The most important function of emergency bronchoscopy is to establish or to diagnose the cause of bleeding. However, the amount and spread of bleeding can also be controlled during emergency bronchoscopy (see items under Diagnosis of Cause and Treatment). (PEEP = positive end-expiratory pressure: CPAP = continuous positive airway pressure.)
tients with lung cancer.27·28 In a very exciting development in pulmonary medicine, two reports described complete cessation of bleeding in almost all patients by selective intrabronchial spraying of fibrin precursors through a catheter in the suction port of a fiberoptic bronchoscope without any adverse effect.29·30 This selective intrabronchial local coagulative method seems to be the most logical, physiologic, and effecti ve one of all the conservative medical therapies. During bronchoscopy, the endoscopist should frequently restore adequate oxygenation and ventilation by intubating the uninvolved main-stem bronchus. In some of the patients with a contraindication to resectional surgery, control of hemorrhage can
be achieved by selective bronchial artery embolization using resorbable material.31"34 The major risk of bronchial artery embolization is spinal cord injury due to spinal cord embolization via arterial collaterals to the spinal cord. The presence of spinal cord collaterals on scout arteriograms is an absolute contraindication to bronchial, embolization. The important causes of initial failure of bronchial artery embolization are extensive and bilateral disease, technical failure to catheterize or achieve a secure catheter position in the bronchial artery for safe bronchial artery embolization, and pulmonary artery origin of the bleeding. In fact, several authors strongly advocated both pulmonary and bronchial arteriography (and embolization,
616
Anesthesia for Emergency Thoracic Surgery
when appropriate), especially in patients with tuberculosis and other suppurative pulmonary disease who experience massive hemoptysis.9·35 Recurrence of hemoptysis after an initially successful bronchial artery embolization is not uncommon. The most important causes of recurrence of bleeding are incomplete embolization, progression of the native disease, and recanalization of the embolized vessels. Probably all three factors operate in each patient to a varying degree. Finally, it must be realized that bronchial artery embolization is only a palliative procedure and does not cure the underlying disease that caused hemoptysis in the first place; therefore, surgery is often indicated once the patient has been stabilized by embolization (see Fig. 17-1).9 In one patient with hemorrhage from the left lung, the combined occlusion of left pulmonary and bronchial arteries was used to control repeated hemorrhage.36 In this patient, a Swan-Ganz catheter was guided by fluoroscopy into the left main pulmonary artery while bronchial arteries were selectively embolized. When recurrent hemorrhage occurred, the balloon in the pulmonary artery was inflated with 4 ml of saline. The decrease in pulmonary artery blood flow to the bleeding area was believed to be instrumental in stopping the second episode of hemorrhage. Figure 17-1 summarizes most of the medical/surgical considerations involving massive hemoptysis. The site of bleeding must be known, and in the vast majority of patients, this will require bronchoscopy. The rigid bronchoscope is preferred (see chapter 15). Unstable patients must be stabilized. If they ultimately require intubation while still bleeding, a double-lumen tube is preferred (see Anesthetic Considerations later). During bronchoscopy, the surgeon may help to control bleeding by performing regional ice lavage; applying topical vasoconstrictors, fibrin seal, YAG laser; and instituting tamponade of the bleeding region. If flexible fiberoptic bronchoscopy has to be performed at the bedside and major life-threatening obstructing blood clots have to be removed, a Fogarty balloon-tip embolectomy catheter can be passed via the suction/biopsy channel to a position beyond the clot. After balloon inflation, partial withdrawal of the catheter and the bronchoscope may dislodge the clot, allowing its subsequent suctioning and removal.37 Of course, the great danger of this procedure is restarting the pulmonary hemorrhage that ceased because of the airway clot. If the patient can withstand surgery and has an operable lesion, surgery should be performed. If the patient cannot withstand surgery and/or has an inoperable lesion, bronchial embolization should be tried.
An approach to the clinical management of massive hemoptysis in patients with cystic fibrosis has been recommended. It is similar to the prior algorithm but differs in one key respect: Embolization is the first choice, and surgery is a second choice of treatment.38 The difference is because patients with cystic fibrosis have generalized lung disease, and preservation of tissue is a much more crucial issue. The site of bleeding is first identified by bronchoscopy, ideally under general anesthesia. Then selective bronchial arteriography is performed instead of surgery. If there are no collaterals to the spinal cord, then bronchial embolization is performed. If collaterals to the spinal cord are visualized, arterial embolization is abandoned and pulmonary resection is undertaken within the limits dictated by the patient's overall pulmonary function. Hemoptysis has a 37 per cent fatality rate; therefore, it behooves all users of pulmonary artery catheters to know how to manage pulmonary artery rupture caused by a pulmonary artery catheter.39 If the hemoptysis is minor, it usually ceases after the pulmonary artery catheter is moved to a more proximal position. If the hemoptysis is still present, gentle inflation of the balloon with 2 ml of air may stop blood flow in the pulmonary artery branch and therefore tamponade the hemorrhage from a more distal perforation. General anticoagulation should be reversed if present. Preservation of oxygen transport may require emergency transfusion, oxygen therapy, or endotracheal intubation. A double-lumen endotracheal tube or bronchial blocker may facilitate ventilation and prevent blood aspiration in the unaffected lung (occasionally in dire circumstances, an endobronchial intubation with a single-lumen tube may be life-saving [see chapter 9]). Artificial ventilation with positive end-expiratory pressure (PEEP) has been advocated to decrease the pressure gradient between damaged vessel and the surrounding lung parenchyma. With a double-lumen tube, a high level of selective bleeding lung continuous positive airway pressure (CPAP) may be used. If all these previous measures fail, emergency thoracic surgery would be the last therapeutic possibility.
C. Anesthetic Considerations 1. Preoperative Considerations The most important preoperative priorities are to prevent asphyxiation, localize site of bleeding, prevent contamination of normal lung if possible, and correct hypovolemia. It must be realized that many of these priorities should be fulfilled simul-
Anesthesia for Emergency Thoracic Surgery
taneously. For example, at the same time a doublelumen tube is being inserted to prevent asphyxiation by drowning in blood, large-bore intravenous cannulas must be inserted to begin to rapidly correct hypovolemia. Antibiotics should be administered preoperatively, and antituberculous drugs should be started in patients with tuberculosis. a. PREVENT ASPHYXIATION
An increased F,0 2 should be immediately administered as continuously as possible (oxygen administration may be interrupted by episodes of hemoptysis). Oxygenation and ventilation should be monitored, when possible, with arterial blood-gas determinations. If the site of bleeding is known, the bleeding lung should be placed in a dependent position to prevent soiling of the nonbleeding lung. Patients who cannot cough out the blood effectively enough to prevent contamination of the nonbleeding lung must have the lungs separated as soon as possible. In the large majority of these patients, this can be done most expeditiously and effectively with a double-lumen tube; however, in the unusual case in which the bleeding is intermittent and a fiberoptic bronchoscope that was being used for diagnostic purposes is positioned near the site of bleeding, placement of a bronchial blocker is a viable alternative (see chapter 9). Rarely, insertion of a single-lumen tube down a main-stem bronchus must suffice as a life-saving measure when the other technically more difficult procedures cannot be accomplished (see chapter 9). Ventilatory support (intermittent positive-pressure breathing and PEEP) must be provided as needed. Suctioning of the tracheobronchial tree must be aggressive. b. PREVENT CONTAMINATION OF NORMAL LUNG
Coughing may increase bleeding. The advisability of using sedatives and cough suppressants is time dependent.19 In the unintubated patient, the ability to cough may be life-saving, and suppressants should be avoided. The patient should be at strict bed rest in the semi-Fowler's position or with the radiologically normal lung in a nondependent position. In the intubated patient, suctioning can replace the cough mechanism, and suppression of cough may decrease bleeding. If possible, the initial intubation should be with a double-lumen tube. A coagulation profile should be drawn early; if any abnormalities are noted, they should be corrected. During and after bronchoscopy, the surgeon internist can control bleeding by iced-saline lavage; placement of topical vasoconstrictors, fibrin seal, YAG laser; placement of a bronchial blocker or gauze packing; and use of bronchial artery embolization (see prior discussion).
617
c. CORRECT HYPOVOLEMIA
As soon as possible, large-bore intravenous cannulas should be inserted. The patient's blood should be typed and screened and cross matched for adequate amounts of blood products (whole blood, packed red blood cells, platelets, fresh frozen plasma). Transfusion should begin if appropriate. Finally, the appropriate monitoring (e.g., arterial line, central venous line) should be instituted.
2. Intraoperative Considerations The patient with massive hemoptysis can come to the operating room without an endotracheal tube in place, with a single-lumen endotracheal tube in place, or with a double-lumen endotracheal tube in place (Fig. 17-2). In deciding what type of airway the patient needs, it is important to remember several general principles. First, the important advantages of a double-lumen tube include separation of the two lungs in order to prevent the spread of blood from one lung to the other lung; improved surgical exposure; the ability to ventilate, provide CPAP, and sigh the operative lung when desired; and the ability to safely inspect an open bronchus during resection. Second, the potential disadvantages of a double-lumen tube include the increased technical difficulty in intubating a bloody trachea in a hypoxic patient, and, if a large quantity of blood has already spread to the contralateral lung, it may be impossible to adequately ventilate and oxygenate the patient using only the now-soiled contralateral lung. Third, if a single-lumen tube is already placed or is inserted, the lungs can still be separated by a subsequent endobronchial intubation (right main-stem bronchus blindly and left main-stem bronchus with the aid of a fiberoptic bronchoscope) or by insertion of a bronchial blocker along the side of the single-lumen tube into the appropriate main-stem bronchus (see chapter 9). Fourth, the Univent bronchial blocker tube may be a good alternative in cases in which rapid intubation and single-lung blockage are required. If the patient with massive hemoptysis is without an indwelling endotracheal tube, preoxygenation should be instituted immediately. Adequate suctioning must be available. It may be necessary for the patient to be awake for intubation during massive, active, spontaneous bleeding in order to avoid the hazard of trying to visualize a bloodobscured airway in a paralyzed patient. Intubation performed in the semiupright position may minimize coughing that results in the presence of blood in the upper airway and, thereby, may provide a clearer field of vision.
618
Anesthesia for Emergency Thoracic Surgery
Figure 17-2 Endotracheal intubation options in massive hemoptysis and the sequence of conversion of one option to another. Double-lumen tubes are the intubation option of choice in the majority of patients.
If the patient without an indwelling endotracheal tube is to be put to sleep, aspiration precautions (e.g., cricoid pressure) should be utilized. Because these patients are likely to be hypovolemic, anesthesia should be induced with a small dose of short-acting barbiturate or ketamine or with narcotics followed in rapid sequence by relaxation. If the larynx can be visualized, insertion of a double-lumen tube is preferable to insertion of a single-lumen tube (see General Considerations presented earlier and Fig. 17-2). It should be remembered that if the patient is not actively bleeding but has a blood-filled cavity (i.e., a hemorrhagic lobe) this cavity will likely empty its contents into the dependent, unsoiled lung when the patient is turned to the lateral decubitus position; therefore, this situation is a strong indication for placement of a double-lumen tube. If a singlelumen tube is inserted, it might as well be the Univent bronchial blocker tube, which additionally allows lung separation. If the Univent bronchial blocker tube is not available, then serious consideration should be given to converting a standard single-lumen tube to an endobronchial position or using a bronchial blocker either inside the lumen or alongside it before turning the patient to the lateral decubitus position. If a single-lumen tube is already in place, the same conversion considerations to bronchial blocker and endobronchial position apply (Fig. 17-2). In addition, consideration should be given to converting the single-lumen tube to a double-
lumen tube (Fig. 17-2). If a double-lumen tube is in place, it should remain and be utilized (provided serious bilateral lung contamination has not occurred). In all situations, if active tuberculosis is present or suspected, contamination precautions should be utilized. Once the airway has been established, the patient must be placed in the lateral decubitus position with the bleeding lung in the nondependent position. Use of this position, of course, emphasizes the importance of separating the lungs. Blood products must be administered according to the continued loss of blood, updated coagulation profiles, and hemodynamic monitoring findings. Insertion of an arterial line greatly facilitates monitoring of cardiovascular status and allows repetitive sampling for arterial blood-gas analysis during the operation. The use of blood warmers is strongly recommended. At the end of the case, the endotracheal tube should be left in place and the patient should be ventilated mechanically. Most of these patients will have impaired gas exchange postoperatively owing to pre-existing lung disease, the probability that the nonbleeding lung has been soiled by the recent hemoptysis from the diseased lung, and the physiologic consequence of having just undergone a major anesthetic and surgical experience. Coagulation profiles, electrolyte concentrations, and acid-base status must be monitored in the immediate postoperative period, and abnormalities must be treated quickly.
Anesthesia for Emergency Thoracic Surgery·
III. THORACIC AORTIC ANEURYSMS AND DISSECTIONS/DISRUPTIONS A. General Considerations A true thoracic aortic aneurysm is a dilatation of all three layers of the aortic wall. Dissection of the thoracic aorta is a propagation of hematoma between the intima and the adventitia (i.e., in the media). Aortic dissection is caused by the sudden development of a tear in the aortic intima, opening the way for blood to enter the media of the aortic wall, thereby separating intima from the adventitia for variable distances along the length of the aorta. The most important forces acting to propagate the dissecting hematoma are the steepness of the as cending pressure pulse wave (dp/dt) and the aortic blood pressure. Spontaneous dissections occur in aneurysms, and traumatic dissections occur with out being preceded or accompanied by an aneu rysm.
7. Spontaneous Dissection of Thoracic Aortic Aneurysms Spontaneously occurring aneurysms are usually repaired electively, and spontaneous acute dissec tion in a spontaneously occurring aneurysm usu ally requires emergency surgical repair.40 Sponta neous dissection is approximately four times as common as a spontaneous rupture.41 Spontaneous rupture is usually a fatal event. There are two classifications of thoracic aortic dissections, and they are based on anatomic loca tion. The DeBakey classification42 consists of types I, II, and III (Fig. 17-3). Type I dissection begins in the ascending aorta and extends for vary ing distances past the aortic arch and below the diaphragm. Type II dissection also starts in the ascending aorta but ends proximal to the left sub clavian artery. Type III aortic dissection generally starts just distal to the left subclavian artery, ex tends for varying distances (type 3A remains above the diaphragm and type 3B extends below the diaphragm), and may even include the iliac arteries. Type I aortic dissection is the most com mon, occurring in approximately 70 per cent of all cases of spontaneous thoracic aortic dissection. 43 The Stanford classification consists of types A and Β (Fig. 17-3). Type A encompasses DeBakey types I and II, and type Β consists of DeBakey type III. The Stanford classification emphasizes the all-important point that Stanford type A and DeBakey types I and II require immediate surgery. Numerous conditions predispose to aneurysmal dilatation and spontaneous dissection of the tho racic aorta. All of these conditions cause degener-
619
ation of the aortic wall. Hypertension, found in 90 per cent of patients with acute aortic dissection,44 mechanically weakens the thoracic aortic wall be cause of chronic wall stress and shear forces. Sim ilarly, the mechanical effect of a jet stream due to aortic valve stenosis and the proximal hyperten sion due to a coarctation of the aorta are associated with aneurysm of the thoracic aortic arch. Turner's syndrome, which is associated with coarctation of the aorta, is also associated with thoracic aortic aneurysms. Inflammation of the aorta will weaken the wall, and, not surprisingly, giant cell aortitis and syphilitic aortitis are associated with thoracic aortic aneurysms. Marfan's and Ehlers-Danlos syndromes are associated with thoracic aortic aneurysms owing to hereditary defects in connec tive tissue strength. The hormonal changes of pregnancy can cause a loss of thoracic aortic wall structural integrity. Finally, although spontaneous internal tears are seldom seen through atheroma tous lesions, thoracic aortic dissections can occur around atheromatous plaques following retrograde catheterization of the central arteries. Patients with thoracic aortic aneurysms have a high incidence of other arterial diseases, such as cerebrovascular and other vessel occlusive disease, abdominal aortic aneurysms, hypertension, and coronary artery dis ease.
2. Traumatic Disruption Thoracic aortic injury caused by blunt objects results from differential rates of deceleration of the fixed aortic arch relative to the more mobile car diac chambers and descending thoracic aorta. Con sequently, traumatic disruption of the thoracic aorta usually occurs in the region of the ligamentum arteriosum just distal to the left subclavian artery. A less common site is just above (cephalad to) the aortic valve. The tear may involve the in tima, intima and media, or the entire aortic wall. Traumatic dissection and rupture of the thoracic aorta have been estimated to be associated with 10 to 16 per cent of all automobile accident 43 45 fatalities. · When a traumatic thoracic aortic tear occurs, 80 to 90 per cent of the patients die at the scene of the accident. 4 3 · 4 5 · 4 6 Among the 10 to 20 per cent of patients who survive an initial trau matic thoracic aortic tear, the leak is temporarily controlled by the adventitia of the aorta. Even though the adventitia is strong, it is not capable of resisting the same bursting pressure as was the intact aorta, and increases in blood pressure above normal limits must be avoided. Most of these pa tients will have associated injuries, and a priority system must be established; often head and ab dominal injuries take precedence. In the group of
620
Anesthesia for Emergency Thoracic Surgery
Figure 17—3 The classification of thoracic aortic dissections is based on anatomic location. The DeBakey classification consists of types I, II, and III, and the Stanford classification consists of types A and B. DeBakey types I and II and Stanford type A involve the ascending aorta and aortic arch, and DeBakey type III and Stanford type Β originate distal to the left subclavian artery.
patients arriving at the hospital alive, as many as 80 per cent may survive postoperatively.43 47
3. Signs and Symptoms of Any Type (Traumatic or Spontaneous) of Dissection The most common presenting symptom (94 per cent of patients)44 is the sudden onset of severe, unremitting, tearing, or ripping chest or back pain. Blood pressure is increased in two thirds of pa tients (especially if the dissection involves the renal arteries), and the patients appear cool, clammy, and vasoconstricted, but hypotension re sulting from rupture of the aorta may be present. Pseudohypotension may be caused by compromise of the flow through either or both subclavian arte ries, and a difference in blood pressure between the arms is not uncommon. In dissections involving the ascending aorta, aortic valvular insufficiency may be present in two thirds of the patients (which may cause acute congestive heart failure) as a result of stretching of the aortic annulus by the dissecting hematoma. Other presenting signs and symptoms of thoracic aortic dissection/disruption result from the ana tomic course that the propagating hematoma takes.48 As the medial hematoma dissects or as the intimai flap peels off, partial or complete obstruc tion of the lumen of an artery arising from the aorta can occur, and the signs and symptoms re flect the distribution of the obstructed artery. In volvement of the coronary arteries causes acute
myocardial infarction and frequently sudden death. Involvement of the innominate and/or common ca rotid artery may cause syncope, confusion, stroke, or coma. Involvement of the innominate and/or subclavian artery can cause upper limb gangrene and paralysis and disparity in the pulses in the extremities. Involvement of the intercostal and/or lumbar arteries can cause spinal cord ischemia and paraplegia. Involvement of the renal, mesenteric, and common iliac arteries can cause oliguria and renal vascular hypertension, bowel ischemia (nau sea and vomiting, abdominal pain, hematemesis, melena), lower leg gangrene and paralysis, and disparity in the pulses in the extremities, respec tively. Through-and-through rupture of aorta (the most common cause of death) can cause extravasation of blood into any adjacent region; consequently, blood may be found in the pericardial sac (most common), pleural space (second most common), mediastinum (which can cause superior vena cava and tracheal compression),49 retroperitoneum, wall of the pulmonary trunk and/or main right and left pulmonary arteries (because of common adventitia with aorta), lung parenchyma, and esophagus. Rupture might even occur back into the aortic lu men, leading to a "spontaneous cure." 4 4 · 5 0
B. Surgical Considerations Since there are tremendous differences in mor tality with surgical versus medical treatment of
Anesthesia for Emergency Thoracic Surgery
proximal aortic lesions48· 5I · 52 the indications for surgery are based on the location of the dissec tion/disruption. Stanford type A and complicated Stanford type Β lesions require immediate surgery. Complications of type Β dissection consist of any of the following: failure to control hypertension, continued pain, expanding aortic diameter, devel opment of neurologic deficit, evidence of compro mise of a major subdiaphragmatic aortic branch vessel, and development of aortic valvular insuffi ciency. On the basis of widespread experience, a con sensus has emerged concerning major principles and management of suspected acute aortic dissec tion. 5 3 " 5 8 It is most important to decrease blood pressure and velocity of ventricular contraction. During the first few hours after admission, the patient's blood pressure should be stabilized with sodium nitroprusside (to systolic blood pressure of 100 to 110 mm Hg). Because sodium nitroprusside causes reflex tachycardia and an increase in dp/dt, beta blockade (propanolol, esmolol, labetalol) is also required to decrease contractility and heart rate (to 60-70 beats/min). Nifedipine may be use ful in patients with asthma. An upright chest radio graph with a nasogastric tube in place should be obtained. The chest radiograph criteria for pro ceeding to aortic angiography include the presence of any of the following: mediastinal widening, the loss of normal contour of the aortic knob, opacifi cation of the aortopulmonary window, depression of the left main bronchus, deviation of the trachea or nasogastric tube to the right, and the apical cap sign.58 Ratios of mediastinal width to chest width at the aortic knob greater than 0.28 have been found to be accurate predictors of thoracic aortic 59 rupture. If the patient is stable, the patient should be transferred for aortography.47 Aortic angiography 50 has a diagnostic accuracy of 95 to 99 per cent ; it is best performed by retrograde aortic catheteriza tion. While the patient is in the aortography suite, use of pain medications or sedatives should be restricted. The persistence of pain despite an ade quate decrease in arterial blood pressure is a sign of continuing hematoma formation. Similarly, depression of mental status is an ominous clinical finding, which may indicate involvement of the major extracranial cerebral vessels in the dissec tion of the hematoma. Both these valuable and important signs (pain and depressed mental status) are masked by overmedication with narcotics or sedatives. Small doses of a mild hypnotic agent such as diazepam should provide adequate relief of patient anxiety. Oxygen should be administered by face mask. If the patient is hemodynamically unstable, the patient should be immediately trans ferred to the operating room.
621
On the basis of the aortography results (which can show the intimai tear, the false channel, com pression of true lumen, patency of major vessels arising from the aorta), the lesion can be classified as Stanford type A or B. This information, coupled with the patient's in-hospital course and response to emergency therapy, dictates whether immediate surgery (type A or complicated type B; e.g., leak ing or ruptured vessel, compromise of a branch so that an organ is jeopardized 60 ) or further stabiliza tion (e.g., repair intracranial and/or bleeding ab dominal lesions or torn airway) with later surgery (some type B) or no surgery (uncomplicated type B) will be required. The operative mortality of type A dissections is approximately 15 to 20 per cent. 50 In-hospital survival of stable type Β treated either medically or surgically is 80 per cent. 50 The overall 10-year survival is 40 per cent.50 Unfortunately, there is approximately a 2 per cent incidence of false-negative aortography re sults in aortic dissection.61 Transesophageal echo cardiography has a diagnostic sensitivity and spec 62 ificity near 100 per cent and therefore has been recommended as the primary bedside test that can identify patients who need surgery without delay for any other test. Indeed, there are cases in which transesophageal echocardiography has detected an asymptomatic descending aortic intimai tear while aortography was negative.63 The diagnostic accu racy of computed tomographic scanning and mag netic resonance imaging is 90 per cent or better. 64 · 65 These noninvasive techniques should be used when aortography is negative, for serial measurements, and perhaps as a screen when sus picion of aortic disruption is low or diagnosis is needed immediately at the bedside (i.e., with trans esophageal echocardiography). The surgical approach (Fig. 17-4) is dictated by several considerations. First, and foremost, is the site of the lesion. Proximal repairs (Stanford type A) require cardiopulmonary by-pass and a median sternotomy (approaches A and Β of Fig. 17-4). Distal repairs (Stanford type B) utilize a left tho racotomy and usually some form of partial by-pass (approaches C and D of Fig. 17-4). With partial by-pass the organs proximal to the clamp area are perfused by the patient's heart and oxygenated by the patient's lungs, whereas the organs distal to the clamped area are perfused and oxygenated by a pump-oxygenator system. The second consideration is control of proximal hypertension during partial by-pass. Approach C decreases proximal hypertension by inserting a proximal-distal shunt (Gott, TDMAC-heparin shunt). The heparin-bonded shunt (systemic heparinization is not used) is inserted proximally in the ascending aorta, or apex of the left ventricle, or, if neither is available, the left subclavian artery and
622
Anesthesia for Emergency Thoracic Surgery
A. Ascending Aorta
Figure 17-4
B. Aortic Arch
See legend on opposite page
Anesthesia for Emergency Thoracic Surgery
distally into the aorta or femoral artery. Approach D decreases proximal hypertension by a left ven tricular, or left atrial, or femoral vein-femoral ar tery partial by-pass. The left ventricular, or left atrial, or femoral vein drain decreases venous re turn to the left side of the heart, thereby limiting left-sided heart preload, cardiac output, and hyper tension proximal to the descending thoracic aortic clamp. It is logical that the patients who would benefit the most by partial cardiopulmonary by pass, with respect to preservation of cardiac func tion, are those with depressed left ventricular func tion preoperatively (e.g., ejection fraction less than 20 per cent). 66 Associated injuries may contraindi cate the use of heparinization required for ap proaches C and D. The third consideration is minimizing neuro logic (spinal cord) damage. Approaches C and D provide some form of distal perfusion. Although there has been one report of no paraplegia in 168 patients shunted with a Gott shunt (cross-clamp time was 37 min), 67 in general, there has been no difference in the incidence of paraplegia between these approaches (C and D) and simple aortic cross clamping (approach Ε of Fig. 17-4). The inci dence of paralysis has been as high as 24 per cent in acute traumatic disruption, as high as 20 to 40 per cent with extensive thoracoabdominal disease, but as low 0 to 2 per cent in patients with elective repair of chronic aneurysms. 68-71 To reduce the incidence of neurologic damage further, other adjunctive methods are being evalu ated clinically or experimentally. These efforts in clude monitoring of evoked potentials,72 monitor ing cerebrospinal fluid pressure and attempting to reduce it if elevated during aortic cross clamping,73 74 use of intrathecal vasodilators, monitoring distal arterial pressures, use of intravenous steroids,75 use of local hypothermia,76 and pharmacologic suppression of spinal cord activity (pentothal, ste roids, magnesium sulfate, mannitol). To date, none of these methods (shunting or otherwise) have been shown to be a means of avoiding paraplegia.70 This is because the most important common denominators for spinal cord injury are prolonged cross-clamp time (e.g., greater than 30 min), prolonged systemic hypoten sion, removal of long segments of aorta, and re moval of aortic segments from which a collateral circulation to the spinal cord has not had time to
623
develop. The presence of these denominators has little to do with the type of by-pass. 6 8 · 6 9 · 7 7 - 7 8 Of these factors, aortic cross-clamp time is probably the most important.79 In addition, distal perfusion of the aorta only provides adequate lumbar spinal cord protection, but the arrangement of the vascu lar anatomy still leaves the lower thoracic spinal cord at risk from ischemia (vascular resistance up the anterior spinal artery is 50 times greater than down the anterior spinal artery).80 It is possible that, aside from short aortic cross-clamp time, scavenging oxygen free radicals during reperfu sion (e.g., with recombinant superoxide dismutase) may prove to be a useful way in limiting ischemiainduced neurologic damage.81 The fourth consideration is maintenance of renal function, and partial by-passes are theoretically most protective.82 Pharmacologic methods such as mannitol and dopamine have not proved to be helpful.83 Other factors that may cause renal dys function/failure (aside from aortic cross clamping with or without shunting) are hypotension before and during operation, amount of operative blood loss, degree of associated injuries, or a combina tion of these factors.82 The fifth consideration is associated injuries. For example, systemic heparinization is contraindicated with closed-head injuries and would elim inate use of pump by-pass. Simple aortic cross clamping avoids heparinization and allows for more prompt management of other injuries owing to shorter surgical time, there are no cannulation site complications, and postoperative renal func tion is normal if clamping time is less than 30 min. Use of the TDMAC-heparin-bonded shunt also avoids the use of heparin.
C. Anesthetic Considerations This disease has an extremely high early mortal ity, and the aim of preoperative medical manage ment is to limit the extent of dissection or prevent frank aortic rupture. The time spent in establishing monitoring should be inversely proportional, and the speed of induction of anesthesia should be directly proportional, to the likelihood of further dissection/rupture and the hemodynamic stability of the patient. Proper judgment as to the appropri-
Figure 17-4 The repair of the various thoracic aortic dissections requires different surgical approaches. A, Complete cardiopul monary by-pass (CPB); cardioplegia for myocardial preservation. B, Cardiopulmonary by-pass for cooling prior to cross clamping, then cessation of by-pass with profound hypothermia during cross clamping. C, TDMAC-heparin shunt for descending aortic dissection. D. Left ventricular, atrial, or femoral vein-femoral artery partial by-pass (PB) for descending aortic dissection. E. Simple aortic cross clamping for descending aortic dissection.
624
Anesthesia for Emergency Thoracic Surgery
ate balance between preoperative preparation and patient risk is extremely critical with this disease. 1. Preoperative Considerations Preoperative anesthetic considerations consist of correction of hypovolemia, institution of monitoring, drug therapy, and laboratory analysis. The extent of correction of hypovolemia will depend on the need for speed of control of the thoracic aorta. At a minimum, adequate intravenous access must be ensured by at least two large-bore intravenous catheters connected to blood warmers, and the patient's blood must be typed and cross matched for adequate amounts of blood products (10 units of whole blood, platelets, fresh frozen plasma). Consideration should be given to use of autotransfusion (i.e., the cell saver). Monitoring should be instituted as permitted by the overall hemodynamic status of the patient. If possible, the V5 or equivalent leads (CMV5 or CB5) should be used to monitor myocardial ischemia. A right radial (or brachial) intra-arterial catheter should be inserted because the left subclavian artery and, therefore the left radial artery, might be compromised by aortic dissection and/or the proximal aortic clamp. If right arm arteries cannot be cannulated, then insertion of a needle into the proximal aorta in the operative field can be substituted during the aortic cross-clamping period. In addition, the left subclavian artery is more likely to be compromised by intraoperative clamping than the innominate artery. A second arterial catheter distal to the cross clamp (e.g., femoral artery, dorsalis pedis artery) may be useful for measuring distal perfusion pressure during cross clamping.84 Urine output should be monitored with a urinary catheter. If time permits, a pulmonary artery catheter or central venous pressure catheter should be inserted. If access to the central circulation is obtained, a pulmonary artery catheter is desirable because of the known discrepancies between central venous pressure and the pulmonary artery occluded pressure in patients with poor ventricular function (e.g., ejection fraction less than 40 per cent) (which many patients with this problem may have).85·86 Both central venous pressure and pulmonary artery occluded pressure correlate well with cerebrospinal fluid pressure (which is the back pressure for spinal cord perfusion).87 Transesophageal echocardiography, either in use to make the diagnosis of aortic disruption or inserted after the induction of anesthesia, can certainly monitor on-line, ventricular performance (see prior discussion and chapter 7). Before the induction of anesthesia, cerebral function should be noted simply by briefly talking with the patient and by observing the pupils and
motor function. Somatic sensory evoked potential monitoring of spinal function has been used, but unfortunately it has not had a significant impact on the prevention of neurologic deficit. Specifically, localization of critical spinal arteries for reattachment has not been possible, and the incidence of false-negative results has been 13 per cent, and of false-positive results, 67 per cent.88 Drug therapy consists of controlling blood pressure with sodium nitroprusside and heart rate with propranolol. The sodium nitroprusside should be titrated to an end point of systemic arterial blood pressure of 100 to 110 mm Hg, and propranolol should be titrated to the end point of a heart rate of 60 to 70 beats/min. Laboratory analysis should consist of complete blood count, determination of electrolytes and blood urea nitrogen/creatinine concentrations, clotting studies (prothrombin time [PT]/partial thromboplastin time [PTT]/bleeding time/platelet count), and a 12-lead electrocardiogram (EKG). Oxygen should be administered by a face mask throughout all these procedures. 2.
Intraoperative Considerations
The important intraoperative considerations are mentioned in approximately the order that they arise during the repair of the thoracic aortic aneurysm (Table 17-3). The aim of induction of anesthesia is to prevent both hypertension and hypotension. In patients with acute dissection, the induction should be smooth and rapid but controlled. In these patients, anesthesia should be induced intravenously with small to moderate doses of narcotic, ketamine, or thiopental (or some combination of these drugs, depending on the starting blood pressure and the degree of hypovolemia thought to be present), followed by profound mus-
Table 17-3
IMPORTANT ANESTHETIC CONSIDERATIONS FOR DESCENDING THORACIC AORTIC ANEURYSMS/DISRUPTIONS
I. Control hemodynamic responses: A. Nitroprusside (systolic blood pressure 110 to 100 mm Hg) B. Propranolol or labetalol (heart rate 60 to 70 beats/min) II. One-lung ventilation (collapse left lung, ventilate right lung) III. Following clamping control proximal hypertension: A. Nitroprusside, B. Increase partial by-pass drainage, C. Decrease intravenous infusion IV. Prior to and during unclamping avoid hypotension: A. Infuse volume, B. Taper nitroprusside, C. Discontinue by-passes, D. Administer bicarbonate as indicated, E. Restore coagulation
Anesthesia for Emergency Thoracic Surgery
cle relaxation, cricoid pressure, and rapid endotracheal intubation. For elective aneurysm repair, higher induction doses of narcotic are usually used, and the induction pace can be more leisurely. Adjuncts for control of hypertension include administration of increasing sodium nitroprusside dose, intravenous lidocaine (Xylocaine) bolus administration, intravenous propranolol bolus administration, further intravenous anesthetic drug administration, and the initiation of inhalation anesthesia. Adjuncts for control of hypotension consist of volume infusion and vasopressor drug administration. The goal of the maintenance of anesthesia is to control systemic blood pressure, which can be done in a variety of ways (halogenated drug anesthesia, high-dose narcotic anesthesia with sodium nitroprusside). Unless there is a strong contraindication, these cases should be done with a double-lumen tube and collapse of the left lung and ventilation of the right lung (with patient in the right lateral decubitus position for left thoracotomy). Ordinarily, a left-sided double-lumen tube should be used (see chapter 9), but if the aneurysm distorts the left main-stem bronchus, a right-sided double-lumen tube should be used. In fact, a dilated aorta may impinge on the trachea and carina sufficient to make ventilation difficult89 and may indicate the use of cardiopulmonary by-pass.90 Examination of a preoperative chest film will often permit the correct initial decision of which bronchus to intubate. If it is anticipated that differential lung ventilation will not be needed, the Univent tube (or a single-lumen tube with an independent bronchial blocker) will provide satisfactory one- and two-lung ventilation. There are three important reasons why one-lung ventilation and double-lumen tube intubation are particularly advantageous for these cases. First, as in many other thoracic surgery cases, surgical exposure is greatly improved, and collapse of the adjacent left lung is particularly important during the difficult period of initial aortic mobilization. Second, initial surgical dissection can cause bleeding into the bronchi of the left (nondependent) lung because the aorta is often adherent to the adjacent lung tissue. The double-lumen tube protects the right (dependent) lung, which might otherwise be flooded by this blood traversing the carina. Third, collapse of the left lung protects the left lung from damage during periods of heparinization. If the left lung were being ventilated during this period, considerable retraction would have to be placed on it to allow for adequate surgical access. This, in a fully heparinized patient, could again initiate bleeding in the left lung. One-lung ventilation, therefore, confers several benefits to the patient. However, it should be noted
625
that these patients are likely to have a sodium nitroprusside infusion, which is a potent inhibitor of hypoxic pulmonary vasoconstriction in the collapsed left lung. Consequently, use of sodium nitroprusside during one-lung ventilation requires that arterial blood-gas concentrations be monitored frequently (pulse oximetry is useful in this situation). If severe hypoxemia occurs, nondependentlung CPAP or temporary clamping of the left pulmonary artery to diminish the shunt flow through the left lung should be instituted (see chapter 11). Blood loss may be considerable during the initial mobilization of the aorta, and adequate volume replacement is a high priority. Once the section of the aorta that is diseased has been freed and mobilized, arterial cross clamps will be placed proximal and distal to the lesion. Prior to placement of the cross clamps, a sample of unheparinized blood must be drawn from the patient to preclot the prosthetic graft, and the patient must be adequately anticoagulated with intravenous heparin; the activated clotting time (ACT) should be approximately four times the control value.91 From the point of view of both possible preclamp hypovolemia and postclamp hypoperfusion, the wellbeing of the kidneys must be monitored closely by the urine output. If oliguria or anuria is present, an osmotic diuretic should be administered. Placement of an aortic cross clamp causes an increase in left ventricular afterload. If the ascending aorta or aortic arch is cross clamped, the heart must be fibrillated and cardiopulmonary by-pass must be used (Fig. 17-4, approach A or B). When the descending aorta is cross clamped, structures that are proximal to the cross clamp are still perfused by the heart, and structures distal to the cross clamp are either unperfused or perfused via a shunt or partial by-pass pump (see Fig. 17-4, approaches C, D, and E). However, because the heart is now pumping into a markedly reduced vascular bed, and vasoconstrictor substances are released by aortic cross clamping,92 hypertension proximal to the clamp occurs, which is dangerous because the proximal hypertension may cause the heart to become ischemic and/or excessively distended. In this context, pulmonary artery catheter monitoring (pulmonary artery occluded pressure, giant CV waves, and so on) can be very useful.85 Consequently, proximal hypertension must be controlled by infusing vasodilator drug, moderately increasing the inhaled halogenated drug concentration,93 increasing the drainage of the heart by the partial by-pass (see Fig. 17-4, approach D), and/or decreasing the intravenous volume infusion (let the systemic blood pressure drift downward). Structures distal to the cross clamp, especially the kidneys and lower spinal cord, must also receive blood flow. In this regard, a distal perfusion
626
Anesthesia for Emergency Thoracic Surgery
pressure of 40 to 70 mm Hg must be maintained.94 During the period of cross clamping, the heart should not be depressed by very high doses of inhaled halogenated drugs. During the period of cross clamping, intraoperative monitoring of somatosensory evoked potentials as a guide to the state of spinal cord perfusion has been used. Ideally, an early warning of cord ischemia may allow for modification of surgical technique, such as reanastomosis of sacrificed intercostal artery or repositioning of hemostatic clamps, in an attempt to restore spinal cord perfusion.95 Unfortunately, the incidence of false-positive and false-negative results has been high with somatosensory evoked potential monitoring.88 Pharmacologic agents used to decrease spinal injury, in addition to sodium nitroprusside, are steroids, pentathol, magnesium sulfate, and mannitol. The intravascular volume management before unclamping is critical. Just as the placement of the cross clamp decreased the arterial space and vascular compliance, unclamping will have the opposite effect. To prepare for this sudden return to normal vascular size and compliance and to avoid precipitous hypotension, the anesthesiologist should begin transfusion to high-normal cardiac filling pressures as the final vascular anastomosis is performed. If hypotensive agents have been necessary to treat proximal hypertension, careful drug infusion tapering should begin. Adequate amounts of blood and blood products should be available if further transfusion is necessary once the cross clamp is removed. Unclamping the descending thoracic aorta must be accompanied by simultaneous discontinuance of the distal perfusion mechanism. With unclamping, bleeding may occur along fresh suture lines and must be monitored and replaced. Metabolic acidosis after unclamping should be expected in unshunted patients (even with a mean aortic cross-clamp time of just 32 min the pH decreases by a mean of 0.2 units)96 and should be corrected as indicated by arterial bloodgas tensions. A small to moderate transient increase in mean pulmonary artery pressure should be expected (approximately 4-5 mm Hg) within the first 5 min after abdominal aortic declamping perhaps because of a sudden decrease in Sv02 or increase in PaC02 (causing hypoxic pulmonary vasoconstriction).97 Urine output should be monitored closely in order to detect oliguria early. After aortic unclamping, the coagulation status should be managed in the following sequence: reverse heparin with protamine; check ACT (or equivalent method to detect residual'heparin effect); give additional protamine as indicated to return ACT to control level; perform coagulation tests (laboratory), including PT, PTT, and platelet count; and treat ac-
cording to results of clotting studies. After aortic unclamping, the goal of anesthesia should be to end the case so that the patient is paralyzed and adequately sedated. As the patient emerges from deeper levels of anesthesia, hypertension must be treated to prevent suture line stress. When the pa- ; tient is in the supine position, the double-lumen tube should be converted to a single-lumen tube and the patient should be ventilated mechanically until all vital functions are adequate and have stabilized. IV. BRONCHOPLEURAL FISTULA A. General Considerations A BPF may be caused by the rupture of a lung abscess, bronchus, bulla, cyst, or parenchymal tissue (as in the case of high levels of PEEP during mechanical ventilation) into the pleural space; erosion of a bronchus can occur by a carcinoma or chronic inflammatory disease and by breakdown of a bronchial suture line following pulmonary resection. Other abnormal connections between distal portions of the respiratory airways and other anatomic structures are abdominobronchial and bronchovascular, and they also usually produce significant pulmonary compromise. A variety of abnormal connections between the lung and abdominal organs occur usually as a result of infection, trauma, or malignancy. These fistulas most often produce pulmonary rather than visceral symptoms even though the primary disease initiating the fistula originates in the abdominal organ. It is postulated that inflammation, infection, or tumor invasion within the abdomen causes diaphragmatic and lung penetration. The pulmonary consequences are nonspecific and include consolidation, atelectasis, and hypoxemia. Fistulas between the . lung and blood vessels or the heart usually present clinically with life-threatening hemoptysis or air embolization. Two examples of such acute and catastrophic connections include blunt and penetrating wounds to the lung and rupture of the pulmonary artery into the airway as caused by a balloon flotation catheter.
B. Surgical Considerations The diagnosis of BPF is usually made clinically. In early postpneumonectomy patients, the diagnosis is based on sudden dyspnea, subcutaneous emphysema, contralateral deviation of the trachea, and disappearance of the fluid level on roentgenograms of the chest. In patients after lobectomy,
Anesthesia for Emergency Thoracic Surgery
persistent air leak, purulent drainage, and expectoration of purulent material are usually diagnostic. When the fistula appears after removal of the chest tube, the diagnosis of a BPF is made on the basis of fever, purulent sputum, and a new airfluid level in the pleural cavity on the roentgenogram of the chest. The diagnosis is confirmed by bronchoscopic examination in most, bronchography in a few, and sinograms (of the fistula) in an occasional patient.98 Other methods consist of injection of an indicator such as methylene blue into the pleural space with its recovery from the sputum, accumulation of radionuclide in the pleural space after inhalation of xenon or when bronchography shows spillage of contrast into the vacant hemithorax. In an effort to eliminate the dreaded complication of postsurgical (resection) leak, tissue adhesive has been prophylactically applied at the time of bronchial closure." In postpneumonectomy patients, if the disruption occurs early, it is possible to resuture the stump. Late postpneumonectomy bronchial disruption is usually associated with empyema and has been managed by conservative drainage, but definitive operative closure is now considered the treatment of choice. Operations for late postpneumonectomy disruption consist of closure with pedicled flaps of muscle and omentum using a lateral thoracotomy and staple closure through an anterior transpericardial approach via median sternotomy.100 In non-postpneumonectomy cases, if the lung expands to fill the thoracic cavity, the leak can usually be controlled with chest tube drainage alone. The proven rationale for thoracostomy suction is to evacuate the pleural space and seal the parietal and visceral pleurae against one another so that the BPF is occluded. With time, most fistulas scar and heal. However, if the fistula is large, and a large leak through a large persistent pleural space occurs, it is unlikely that the fistula will close, and surgical resection is usually necessary.98 The most commonly performed operations are staged multiple-rib resection thoracoplasties (to obliterate the pleural space), resuturing of bronchial stump, and suture of intercostal muscle pedicle (or pleuras, thymus, pericardium, or omentum) over the open bronchial stump.98 If the lung remaining in the hemithorax is unable to expand because of pleural thickening, a decortication procedure may also be performed. An empyema complicating a bronchopleural fistula should, if possible, be drained before surgery. This is done under local anesthesia with the patient in an upright position to minimize the possibility of contaminating healthy lung tissue with the cavity's purulent material.
627
Finally, a spontaneous pneumothorax (no prior lung resection involved) is pathophysiologically similar to a BPF. A spontaneous pneumothorax is due to a ruptured pseudocyst: either a bleb or a bullae. Blebs are subpleural and small, less than 1 cm in diameter. There are three types of bullae. Type I is a thinwalled cyst with little communication with the bronchial system that is basically extrapulmonary in nature. It is usually a few centimeters in diameter but sometimes grows to a large size: 15 to 25 cm. At this size, such pseudocysts are readily visible on a plain chest X-ray film and cause compressive bullous emphysema. Surgical bullectomy for compressive bullous emphysema has been most successful for type I bullae. A type II pseudocyst is a medium-sized bulla and has thick, fibrous walls. It is located deep in the lung parenchyma. Several type II bullae may be found together in one lobe. Patients with type II bullae usually remain asymptomatic, and the bullae are not visible on plain chest films. When they rupture, causing a spontaneous pneumothorax, however, surgical intervention is frequently required. Type III is a large pseudocyst and is usually found in more than one lobe (diffuse bullous emphysema). They are the most common cause of diffuse bullous emphysema. Bronchial communications are abundant. When they rupture, surgical intervention is required. There are three situations in which definitive surgical treatment is indicated in treating a spontaneous pneumothorax. First, surgery is required when conventional tube drainage and suction have been unsuccessful in clearing the pleural space (approximately 80 per cent of patients with spontaneous pneumothorax have cure without recurrence with simple chest tube drainage),101 and when, in effect, a BPF has formed. Second, surgical intervention is usually indicated when a second ipsilateral or first contralateral spontaneous pneumothorax occurs. Third, considering that a spontaneous pneumothorax has a recurrence rate of 20 per cent, if after the initial event the patient's lifestyle is such that a recurrence might be life-threatening or highly inconvenient, definitive treatment is indicated. In younger patients with a simple pneumothorax, pleurectomy is the prophylactic procedure of choice and results in a low recurrence rate.102·I03 In older patients with severe, chronic airflow obstruction or young patients with cystic fibrosis, chemical pleurodesis with talc, dextrose, tetracycline, AgN0 3 , or other irritant is safer but also less effective.104 There are a number of closed-chest closure methods of a BPF or a ruptured pseudocyst (Table 17-4).105 Most of these methods are based on a
628
Anesthesia for Emergency Thoracic Surgery
Table 17-4 CLOSED-CHEST CLOSURE METHODS FOR BRONCHIAL FISTULAS AND RUPTURED PSEUDOCYST (BLEBS, BULLAE)* I. Differential lung ventilation (see Table 17-5, 13) II. Bronchoscopic 1. Nd:YAG and C0 2 laser 2. Lead fishing weights or shot 3. Balloon occlusion (septostomy catheter) 4. Tissue adhesive (bucrylate, Histoacryl) 5. Silver nitrate to directly visualize stump fistulas 6. Fibrin glue 7. Autologous blood instillation to form an obstructive clot and doxycycline as an irritant 8. Absorbable gelatin sponge (Gelfoam) III. Chest tube chemical pleurodesis IV. Thoracoscopic 1. Instillation of talc or other irritant chemical into fistula and/or pleural space (pleurodesis) 2. Nd:YAG pleurodesis 3. Fibrin glue 4. Tissue glue—Histoacryl 5. CO, laser, Nd:YAG laser ] of blebs and 6. Electrocautery J bullae ^Modified from Powner DJ, Bierman MI: Thoracic and extrathoracic bronchial fistulas. Chest 100:480-486, 1991. Used with permission. Abbreviation: Nd:YAG = neodymium: yttrium aluminum garnet.
temporary physical occlusion of the airway until an inflammatory response to a foreign material effects a permanent seal. All of the occlusion methods carry a risk that the obstructed segment may become infected or the obstructing object may act as a one-way valve, leading to distal expansion rupture of other previously normal lung units. Simultaneous use of a bronchoscope and thoracoscope combines the advantages of two of the methods.
C. Anesthetic Considerations Preoperatively, it is useful to estimate the loss of tidal volume through the BPF. This may be done in two ways (Fig. 17-5). First, one should determine whether air bubbles intermittently or continuously through the chest tube. If air bubbles intermittently, it means the BPF is small. A small BPF only allows air to enter the pleural space with large transpleural pressure swings (deep inhalation, during cough against a partly closed glottis, or during positive-pressure ventilation). With quiet breathing, one generally sees air bubbling through the water-seal chamber when the patient exhales and intrapleural pressure rises. In contrast, when one has a large, low-resistance BPF or bronchial rupture, air may bubble continuously through the water-seal chamber of the chest
tube drainage system. A constant air leak with quiet respiration suggests that the damage is so great that ordinary transpleural pressures drive air into the pleural space during most of the ventilatory cycle. The application of pleural suction may actually increase the transpleural pressure gradient and increase gas flow from the airway to the drainage system. Consequently, the lung will not expand to contact the chest wall, and closure of such a large BPF may be difficult. Second, the size of the BPF may be quantitated by the difference between inhaled and exhaled tidal volumes. In the nonintubated patient, this may be determined with a tight-fitting mask and a fast-responding spirometer, and in the intubated patient by direct attachment of the spirometer to the endotracheal tube. The larger the leak, the greater the need to isolate the BPF (double-lumen tube, bronchial blocker). During the induction of anesthesia, suction on the chest tube should be discontinued to decrease the loss of tidal volume with the initiation of positive-pressure ventilation.106 The presence of the chest tube still protects against the development of pleural air trapping and a tension pneumothorax. The bubbling through the chest tube usually increases with the initiation of positive-pressure ventilation (even though the suction is discontinued). Although virtually all surgical procedures, including and especially thoracoscopic procedures, require a double-lumen tube and one-lung ventilation, selective lobar bronchial blockade of a lobe with a BPF (achieved by placing a fiberoptic bronchoscope and Fogarty catheter down one of the lumens of an in-situ double-lumen tube) has been described.107 At the conclusion of surgery, the integrity of the airway may be tested by pressurizing the airway and looking for bubbling of air in a saline-filled hemothorax. Nevertheless, whether the chest is open or closed, the lung should be expanded slowly to decrease the risk of re-expansion pulmonary edema.108 Therapy to decrease gas flow via the fistula in anticipation of spontaneous closure may be used when urgent intervention is needed, multiple fistulas are present, or full closure techniques fail. Such treatment is directed toward reducing the pressure gradient across the fistula by changing gas pressure at one of both ends of the BPF. This is accomplished by either reducing the force created by positive airway pressure or altering the subambient pressure in the pleural space. These techniques have been extensively reviewed and are summarized in Table 17-5. There have been several medical (nonsurgical) approaches (use of various mechanical ventilation-chest tube drainage systems) to the treatment
Anesthesia for Emergency Thoracic Surgery
QUANTIFICATION OF TIDAL VOLUME LOSS THROUGH BRONCHOPLEURAL FISTULA (2) Spirometer Connected to Mask (non-intubated patient) or Endotracheal Tube ( ETT): Difference between Inhaled (I) and Exhaled (E) Tidal Volume = Tidal Volume Loss
Mj Intermittent or Continuous Bubbling = Tidal Volume Loss
Figure 17-5 The loss of tidal volume through a bronchopleural fistula (BPF) may be semiquantitated in two ways: (l) the amount of bubbling through a chest tube and (2) the difference between inspired and expired volume.
of patients with BPF that are important for the anesthesiologist and intensive care physician to know about' 09- " 8 (see Fig. 12-9). In all of these approaches, initial adequate conventional positivepressure ventilation through a single-lumen endotracheal tube was generally difficult to achieve because of loss of much of the tidal volume through the fistula. This, in turn, retarded healing and may have even made the fistula larger. One solution was to intubate selectively the bronchus of the
Table 17-5 TREATMENT TO DECREASE GAS FLOW THROUGH A BRONCHOPLEURAL FISTULA* I. Reduce positive airway pressure 1. Minimize tidal volume, positive end-expiratory pressure (PEEP), expiratory retard, inspiratory flow, and inspiratory time during mechanical ventilation 2. Emphasize spontaneous breathing modes of ventilation 3. Independent lung ventilation via a double-lumen endobronchial tube 4. High-frequency jet ventilation II. Change pleural pressure 1. Increase or decrease chest tube suction 2. Application of PEEP or inspiratory closure valves on chest tubes 3. Decubitus body position with the fistula dependent *From Powner DJ, Bierman MI: Thoracic and extrathoracic bronchial fistulas. Chest 100:480-486, 1991. Used with permission.
contralateral normal lung (a double-lumen tube is the best way to do this) and provide ventilation to only one lung, thereby allowing the fistula to heal in a relatively quiet environment109 (see Fig. 129). However, this approach may not be applied to patients with a BPF who have associated pulmonary pathologic changes and are, therefore, unable to tolerate the large transpulmonary shunt inherent in this method of treatment. In addition, this type of patient may require PEEP during mechanical ventilation to maintain functional residual capacity, and unilateral PEEP may increase the shunt through the nonventilated lung. It is difficult to apply PEEP effectively in the presence of a large BPF because a constant air leak is present." 2 " 3 Furthermore, the suction chest drainage, which is normally present in these patients, may cause the ventilator to cycle if it is set in the assist mode."2· "3 One solution to the problem of maintaining PEEP in the presence of a large BPF has been the application of positive intrapleural pressure equal to the end-expiratory pressure desired during mechanical ventilation110 (see Fig. 12-9). This method is effective in maintaining PEEP and in preventing gas leak during exhalation, but it does not prevent an air leak during positive-pressure inspiration. Alternatively, the insertion of a unidirectional valve into the chest tube drainage apparatus solves the problems of avoid-
Anesthesia for Emergency Thoracic Surgery
ing an air leak during positive-pressure inspiration and using PEEP.111 The unidirectional chest tube valve is closed by the inspiratory phase of the ventilator and opens during expiration. A small tension pneumothorax is created with each mechanical inspiration, but it is immediately relieved when the chest tube valve opens. Deleterious cardiovascular effects have not been noted with this technique. However, the chest tube valve must frequently be checked because chest tube drainage fluids may cause the valve to either stick closed or remain permanently open. One report describes a conservative approach to managing a BPF by decreasing airway pressure delivered to the diseased lung using differential lung ventilation.114 A double-lumen endotracheal tube was placed in the patient, and independent volume settings for each lung were achieved by using two synchronized ventilators. The nondiseased lung was ventilated normally (tidal volume 800 ml, PEEP 12 cm H 2 0, rate 10/min), while tidal volume was kept low in the diseased lung (tidal volume 150 ml, PEEP 5 cm H 2 0, rate 10/min). Oxygenation and ventilation were improved over that provided by conventional mechanical ventilatory support, although the patient eventually died. The authors stated that, in retrospect, they should not have ventilated the diseased lung at all; rather, a level of CPAP just below the critical opening pressure of the fistula might have been more beneficial. High-frequency ventilation (HFV) has been proposed as treatment for a large BPF (see chapter 12 and Fig. 12-9). The purported advantages of highfrequency ventilation for BPF included the following: (1) There is minimal loss of tidal volume through the BPF; (2) the fistula may heal more quickly; (3) the airway resistance (or lack of) and pulmonary compliance have minimal influence on ventilation; (4) low airway pressures result in minimal effects on cardiac output; and (5) spontaneous ventilatory efforts are usually abolished, thereby lowering oxygen consumption from respiratory muscles and eliminating the need for paralysis and excessive sedation. However, the gas-exchange effect of HFV in patients with BPF has not been predictable, and in some patients119 and in wellcontrolled animal studies120 HFV has increased gas leak through the BPF. The first report of the successful treatment of a BPF with high-frequency jet ventilation appeared in 1980.'I5 This article described a patient in whom a BPF occurred along with an empyema several months after a right upper lobectomy. The patient was ventilated (rate 115/min) for several weeks until the fistula closed and weaning was possible. The use of dual-mode independent lung ventilation utilizing both conventional and high-frequency jet
ventilation for the intraoperative repair oï a BPF has also been reported."6 This article described a patient who was managed during thoracoplasty with a double-lumen endotracheal tube; the nondiseased lung was ventilated in a conventional manner (tidal volume 500 ml, rate 9/min), while the lung with the BPF received high-frequency jet ventilation (rate 150/min). This method was continued for 2 hours following successful repair of the BPF until extubation was accomplished. Finally, high-frequency jet ventilation has been used in the nonoperative management of a patient with bilateral BPF." 7 This report described a patient who required high-frequency jet ventilation for almost 2 months; at the time of extubation, small fistulas were still present but eventually healed. Thus, although the eventual role of high-frequency jet ventilation for treating BPF is not fully established,"8 it is emerging as the best nonsurgical method to treat this condition when mechanical ventilatory support is required and the proper equipment and personnel experienced with its use are available. No matter which ventilation method and chest tube drainage system are used, the entire approach should be evaluated by continuously measuring the volume (flow) of gas passing from the chest tube by inserting a sterile "flow-tube" sensor into the chest tube drainage system.106 The usual methods of visually assessing the amount of bubbling in the water-sealed chamber or subtracting the measured exhaled lung volume from the present inspired tidal breaths, although useful for quick preoperative assessment (see prior discussion and Fig. 17-5), are only intermittent semiquantitative measures of lost volume, which are especially inaccurate if the patient is breathing spontaneously. In addition, flow through the fistula may frequently be changed during the patient's clinical course. Continuously measuring fistula flow allows on-line titration of tidal volume, PEEP, inspiratory occlusion of the chest tube, and balancing airway/chest tube PEEP.106 In addition, the response time of the flowmeter is short enough to separate fistula flow during the inspiratory phase of ventilation from that of exhalation in most patients. Gas recovered from the fistula may have participated in physiologic gas exchange, as evidenced by an increased amount of carbon dioxide in gas leaving the chest. The magnitude of such gas exchange was dramatically demonstrated by a patient who had a long-term thoracostomy window and, with mouth and nose occluded, could meet his respiratory needs by ventilating only through the bronchocutaneous fistula.106 Obviously from the prior discussion, and for the preoperative nonintubated patient, the ability to
Anesthesia for Emergency Thoracic Surgery
deliver positive-pressure ventilation adequately to a patient with a BPF and chest tube drainage, or to a patient with a bronchopleurocutaneous fistula, must be carefully considered preoperatively. If a fistula is obviously small, chronic, and uninfected, a standard endotracheal tube can likely be used safely. When in doubt, positive-pressure ventilation can be tested, and, if found inadequate, the standard endotracheal tube can be replaced with a double-lumen endotracheal tube. If, after the chest is opened, an excessive leak is encountered when using a standard endotracheal tube, ventilation can be improved by lung packing and manual control of the air leak.121 For large fistulas or fistulas of unknown size, and for those in which an associated abscess or empyema is known or suspected to be present, intraoperative use of a double-lumen endotracheal tube has the great triple benefits of permitting positive-pressure ventilation of the normal lung without loss of the minute ventilation through the fistula, avoiding the great hazard of contamination of the uninfected lung with infected material when the patient is turned to the lateral decubitus position,122"124 and facilitating surgical exposure of the operative lung. Indeed, in one series of 22 patients undergoing operations for BPF following pulmonary resection for tuberculosis or tuberculous empyema, management with a single-lumen endotracheal tube, despite intubation in the head-up position and frequent suctioning, resulted in extensive contamination of normal lung in two patients.125 In one patient, the operation had to be terminated, and in the other emergency bronchoscopy had to be performed. The use of a doublelumen endotracheal tube effectively isolates the leaking and perhaps infected lung cavity from the normal lung. For patients who cannot have a double-lumen tube (e.g., small children, patients who cannot tolerate being taken off the ventilator, those with anatomic difficulties, postpneumonectomy patients,126) bronchial blockade and endobronchial intubation are less satisfactory alternatives.
V. LUNG ABSCESSES AND EMPYEMA A. General Considerations Aspiration secondary to alcoholic stupor has classically been reported as the factor that most commonly precipitates lung abscess. Empyema is the accumulation of pus in the pleural cavity. Any lung abscess may produce empyema. In descending order of frequency, causes of empyema are postpneumonic (abscess), postoperative (any thoracic surgery), thoracic trauma (infection of resid-
631
ual clotted blood after a hemothorax that was treated by chest tube placement), septicemia, and iatrogenesis (entry into the pleural space by needle, scope, and so on, especially if there is a concurrent intra-abdominal injury or infection).127-129 Both a lung abscess and an empyema may erode a bronchus and cause a BPF (see Chest Trauma and Fig. 17-8). The precise mechanism by which empyema develops is still obscure, but the following scenario seems likely.130 Inflammation of the visceral pleura occurs as a result of pulmonary inflammation. This encourages an increase in the production of pleural fluid; in time, an expanding pleural effusion, which may be sterile, develops. Unfortunately, pleural fluid is an excellent culture medium, and organisms reaching the pleural space from the lung probably contribute the greatest number of cases of empyema. However, any organisms iatrogenically introduced into the pleural space by unsterile tapping are likely to flourish. Continued irritation of the pleura not only results in the production of pleural fluid, which may or may not be infected, but also edema and thickening of the pleura itself. Fibrin within the pleural fluid forms a coagulum. Coalescence of this coagulum will form septations and hence loculation of pus within the pleural space, the presence of which will make drainage of the pleural space more difficult to accomplish. In addition, partial or complete collapse of the lung may be caused by pressure from the accumulating fluid. Thickened visceral pleura will then trap the lung in this position. The symptoms of lung abscess and empyema are similar and include cough, purulent sputum production, fever, chest pain, and dyspnea. Many patients have some form of systemic illness such as rheumatoid disease, diabetus, chronic alcoholism, advanced malignancy, chronic malnutrition and debility, renal failure, drug abuse, or acquired immunodeficiency syndrome.128·I31 The most frequent physical findings include decreased breath sounds, rales, tachypnea, rhonchi. dullness on percussion, and tachycardia.129 I31 The diagnosis of thoracic empyema is most often first strongly suspected by the finding of a pleural effusion on chest roentgenogram and is confirmed by diagnostic pleural aspiration from which either frank pus was obtained or organisms were grown.
B. Surgical Considerations As with all abscesses, drainage of the pus will dramatically alleviate symptoms and signs and speed the patient's recovery. Table 17-6 lists all the drainage methods in increasing order of inva-
632
Anesthesia for Emergency Thoracic Surgery
Table 17-6
METHODS FOR DRAINING AN EMPYEMA IN ORDER OF INCREASING INVASIVENESS*
1. Needle aspiration. 2. Tube drainage, either with a fine bore percutaneous tube placed accurately with radiographic or ultrasound control, or a large bore tube (26-30 French gauge). Absolute control over the insertion of any of these tubes is mandatory if subdiaphragmatic placement with injury to the abdominal organs is to be avoided. 3. Tube drainage plus irrigation of the pleural space plus instillation of uro- or streptokinase. 4. Thoracoscopy, through which adhesions and septae may be broken down, and irrigation through a tube placed at the time of surgery. 5. Thoracotomy and decortication with excision of the empyema. 6. Rib resection and insertion of an empyema drainage tube or the creation of an Eloesser flap. 7. Thoracoplasty. 8. Resection of lung. *From Wells FC: Empyema thoracis: What is the role of surgery? Respir Med 84:97-99, 1990. Used with permission.
siveness. The choice of treatment should be tailored to the stage of development that the empyema has reached (Table 17-7). Various combinations of the therapies listed in Table 17-6 may be used for the different classes of empyema in Table 17-7. Obviously, the success rate for a given therapy will depend on the class of empyema. For example, aspiration or closed-chest tube drainage canTable 17-7
Class II:
Class III:
C. Anesthetic Considerations
In patients with lung abscess and empyema, some of the preoperative findings may be very unusual. First, the ipsilateral lung may be collapsed. Second, because the infectious process may diminish hypoxic pulmonary vasoconstriction and hypoxemia may be profound. Third, the mediastinal structures may be shifted toward the diseased side. Fourth, the mediastinal shift may cause changes on the EKG. If a surgical procedure is going to be done under general anesthesia in a patient with either a lung abscess or an empyema, then double-lumen tube intubation is absolutely indicated to prevent contamination of the uninfected lung by the infected
A CLASSIFICATION AND TREATMENT PLAN OF EMPYEMA THORACIS1
Class of Empyema Class I:
not help a loculated class IN empyema and multiple reports have urged surgical aggressiveness when confronting class II and class III empyemas.128 129· 132-134 For one dramatic example, the mortality rate for patients with class II and class III empyemas managed completely nonoperatively has ranged from 58 to 100 per cent, whereas it has been 16 to 19 per cent for those receiving operative drainage.128 129,132 Of course, many nonoperative treatment failures have become operative treatment successes.128·129· 132-134 The overall mortality for all classes of empyema has ranged from 5 to 19 per cent.128·129132-135
Description of Class
Treatment of Class
Post pneumonic pleural effusion —bacteria may/or may not be isolated; usually no pus —no loculi within the pleural space —pleural fluid usually of low viscosity Classic uniloculatef empyema —pus in the pleural space —positive cultures of pleural aspirate —absence of multiple locations —pus may be very thick and viscous with sediment Complicated empyema —multiple loculations —grossly thickened pleura —pus within empyema cavity which may be sterile, but is usually very viscous —trapped lung
Simple aspiration (thoracentesis) Percutaneous catheter drainage to underwater seal
1. Tube thoracostomy for continuous drainage (may need multiple tubes) 2. Irrigation with saline, antibiotics, enzymes 3. Thoracoscope to break empyema down to one space (break loculations) plus 2 above
1. Thoracoscope as above 2. Thoracotomy for a. decortication and breakdown of loculations b. excision of empyema cavity c. rib resection for extensive and dependent drainage d. closure of bronchopleural fistula e. lung resection 3. Thoracoplasty
*From Wells FC: Empyema thoracis: What is the role of surgery? Respir Med 84:97-99, 1990. Used with permission. tThere may be fibrinous strands criss-crossing the pleural space but discrete loculi are not present.
Anesthesia for Emergency Thoracic Surgery
Table 17-8 CAUSE OF INJURIES LEADING TO ACCIDENTAL DEATH IN THE UNITED STATES (PER CENT) IN 1985* Motor vehicle accidents Suicide Homicide Miscellaneous
48 29 22 1
*From LoCicero J, Mattox KL: Epidemiology of chest trauma. Surg Clin North Am 69:15-19, 1989. Used with permission.
lung. In addition, if an empyema is going to be treated with thoracoscopy, collapse of the diseased lung greatly aids access to the empyema.136·I37 The seal of the endobronchial cuff should be tight and can be quantitated by the bubble-under-water technique described in chapter 9. The position of the tube should be determined by fiberoptic bronchoscopy. Both of these maneuvers (checking of tube by fiberoptic bronchoscopy and determination of pressure level of cuff seal) should be done before the patient is turned to the lateral decubitus position. As in many thoracic surgery cases, one-lung ventilation facilitates surgical exposure (although the surgeon will usually also be intent on distinguishing fibrous adhesions and in expanding the diseased lung), and lung separation allows differential lung ventilation, as might be needed with the presence of a BPF. During the surgical procedure, the diseased side should be suctioned frequently (fiberoptically if necessary). Whenever possible, but particularly at the end of the procedure, the diseased lung should be fully expanded manually. When the lung has been fully expanded, the surgeon should check carefully for the presence of a BPF. The pleural cavity may be irrigated with antibiotics at the end of the procedure, and one must be mindful of the neuromuscular blockade effects of antibiotics.
VI. CHEST TRAUMA A. Overview of the Management of the Patient With Extensive Trauma Motor vehicle accidents are the most common cause of accidental death, followed by suicide and Table 17-9 FREQUENCY OF VARIOUS INJURIES (PER CENT) IN MOTOR VEHICLE ACCIDENTS* Extremities Head and neck Chest Abdomen
34 32 25 15
*From LoCicero J, Mattox KL: Epidemiology of chest trauma. Surg Clin North Am 69:15-19, 1989.
633
homicide (Table 17-8).138 In a motor vehicle accident, multiple organs may be injured (Table 179).138 Extensive trauma has been defined as an injury severity score (ISS) greater than 25 (Table 17-10).139 The principles of management in all cases of extensive trauma are simultaneous assessment and resuscitation, complete physical examination and diagnostic studies if the patient becomes hemodynamically stable, and life-saving surgery (the timing of surgery depends on the patient's stability). Resuscitation has two components: the primary survey and initial resuscitation, and the secondary survey and continuing resuscitation (Figs. 17-6 and 17-7). All patients undergo the primary survey of airway, breathing, circulation, and neurologic status. Only patients who become hemodynamically stable will progress to the secondary survey, which focuses on a complete physical examination that directs further diagnostic studies. The great majority of patients who remain hemodynamically unstable require operative intervention immediately. 1. Primary Survey The goals of the primary survey are, in order, to establish a patent airway and adequate ventilation, maintain the circulation (including cardiac function and intravascular volume), assess the global neurologic status, and determine the mechanism of injury. a. AIRWAY AND VENTILATION
Patients with extensive trauma who are unconscious or in shock benefit from immediate endotracheal intubation. To prevent injury to the spinal cord, the cervical spine must not be excessively flexed or extended during intubation. Oral endotracheal intubation is successful in the majority of injured patients. On rare occasions, bleeding, neck deformity, airway laceration, or edema from maxillofacial injuries necessitates cricothyroidotomy or tracheostomy. b. CIRCULATION
A patient who is in shock with flat neck veins is assumed to have hypovolemic shock until proved otherwise (see later discussion). Indeed, in one large series of acute blunt thoracic trauma patients (n = 303), the cause of all but one intraoperative deaths (n = 15) was exsanguination.141 If the neck veins are distended and the blood pressure is low, there are five possibilities: (1) myocardial contusion, (2) myocardial infarction, (3) tension pneumothorax, (4) air embolism, and (5) pericardial tamponade. Myocardial contusion is an increasingly recognized cause of arrhythmias and cardiac failure in patients with trauma. Indeed, the
634
Anesthesia for Emergency Thoracic Surgery
Table 17-10 THE INJURY SEVERITY SCORE* The injury severity score is based on the abbreviated injury score (AIS), a score on a numerical scale ranging from l (indicating minor injury) to 6 (virtually unsurvivable injury). An AIS is assigned to each of six regions of the body: head or neck, face, chest, abdomen or pelvic contents, extremities, and body surface. The injury severity score is defined as the sum of the squares of the AISs for the three most severely injured body regions. The procedure can easily be seen in the following example: Body Region Head Face Chest Abdomen Extremities
Injury Epidural hematoma Ear laceration Rib fractures Ruptured spleen Fractured femur
AIS
AIS2 16 1 4 9 9
In this hypothetical case, the sum of the square of the three highest AISs (those for the most severely injured body regions) is 16 + 9 + 9, or 34. The ISS is thus 34. The AIS for flail chest is 4, moderate pulmonary contusion 3, simple pneumothorax 3. *From Trunkey D: Initial treatment of patients with extensive trauma. Ν Engl J Med 324:1259-1263, 1991. Used with permission
incidence can be as high as 70 per cent when it is searched for with thallium 201 scintigraphy.142 The clinical diagnosis is made by EKG changes consis tent with acute injury, increased creatine kinaseMB greater than 5 per cent of the total creatine phosphokinase (CPK), or an abnormal echocardio gram consistent with acute injury such as hypokinesis, pericardial effusion, valvular injury, apical thrombus, or wall thickening with edema and/or hemorrhage.143 Myocardial infarction from coronary occlusion
is not uncommon in elderly trauma patients. Ten sion pneumothorax should be ranked high in thi physician's differential diagnosis of shock witl distended neck veins because it is the easiest life threatening injury to treat in the emergency depart ment (needle the chest anteriorly and apically). / simple tube thoracostomy is the definitive methoi of management. Air embolism has come to be appreciated as aj important complication in injured patients (esti mated as high as 4 per cent of patients with majc
Anesthesia for Emergency Thoracic Surgery
635
Figure 17-7 The algorithm for management of thoracic shotgun wounds is a specific application of Figure 17-6. (O.R. = operating room; IV = intravenous.) (From Walker ML, Poindexter JM, Stovall I: Principles of management of shotgun wounds. Surg Gynecol Obstet 170:97-105. 1990. Used with permission.)
trauma). 139 I44 It consists of air in the systemic circulation and is caused by a bronchopulmonary fistula near the hilum of the lung (pulmonary veins are in close proximity to their bronchi only in the hilar region; see chapter 2).144 Not surprisingly, therefore, airway bleeding is a common finding in patients with air embolism.144 A high index of suspicion is necessary to make a diagnosis of air embolism. Any patient with chest trauma (especially penetrating) who has no obvious head injury but has focal or lateralizing neurologic signs may have air bubbles occluding the cerebral circulation. Theoretically, the observation of air in the retinal vessels on funduscopic examination will confirm the presence of cerebral air embolism, but in reality the diagnosis is most often made only at the time of thoracotomy by direct visualization of air in the coronary vessels or by aspiration of air from the left ventricle. Any intubated patient on positive-pressure ventilation who has a sudden cardiovascular collapse should be presumed to have either tension pneumothorax or air embolism in the coronary circulation. The definitive treatment is immediate thoracotomy in the steep head-down position, clamping of the hilum of the injured lung to prevent further embolism, expansion of the intravascular volume, and introducing proximal aortic hypertension (manual occlusion of aorta) to
increase coronary blood flow. Open cardiac massage, intravenous administration of epinephrine, and venting of the left side of the heart and the aorta with a needle to remove residual air may be required. The pulmonary injury is definitively treated by oversewing the laceration or resecting the offending lobe. Pericardial tamponade is most commonly encountered in patients with penetrating injuries to the torso. The diagnosis is often obvious. The patient has distended neck veins with poor peripheral perfusion; a few have pulsus paradoxus. Ultrasonography and pericardiocentesis are occasionally useful. The proper treatment is immediate thoracotomy, preferably in the operating room, although thoracotomy in the emergency room can be life-saving. If the primary cause of shock is blood loss, the four steps to be taken are to (1) gain access to the circulation, (2) obtain a blood sample from the patient, (3) determine where the volume loss is occurring, and (4) give fluids for resuscitation. The fastest and most reliable way to gain access to the circulation is by a surgical cut down on the saphenous vein at the ankle. Incision in this anatomically constant vein allows the physician to achieve access quickly with a large-diameter tube. Alternatively, the resuscitating physician can perform
636
Anesthesia for Emergency Thoracic Surgery
Table 17-11 FLOW RATES OF CATHETERS (ml/min) THROUGH 6 FEET OF TUBING WITH PERFUSION PRESSURE OF 300 mm Hg* Catheter
Crystalloid
Whole Blood
158 387 492 486
60 200 265 248
16 g central line 14 g IV 8F introducer IV tubing
*Reprinted with permission from Millikan JS, Cain T, Hillsbrough J: Rapid volume replacement for hypovolemic shock: A comparison of techniques and equipment. J Trauma 24(5):428-431, 1984. Abbreviation: IV = intravenous.
ongoing hemorrhage in the abdomen or pelvis. Most such patients require immediate laparotomy if death from hemorrhage is to be averted. The fourth step for the resuscitating physician is to administer fluids for resuscitation, beginning with balanced salt solution and adding type-spe cific whole blood as soon as possible. For the patient who is exsanguinating, the order of blood administration, in order of increasing preference, is type O (e.g., the type has not yet been deter mined), type specific, and finally in the case of a stable patient, it is prudent to wait for typed and cross-matched blood. c. NEUROLOGIC STATUS
bilateral percutaneous femoral vein cannulation with a large-bore catheter or an 8 French intro ducer, more commonly used for passing a pulmo nary artery catheter. Some experienced physicians prefer to gain access through the subclavian or internal jugular vein. Table 17-11 shows the pres surized flow rates from various sized catheters. As soon as the first intravenous catheter has been established, base-line blood work should be undertaken; it should include the hematocrit, toxi cologic screening, blood typing and cross match ing, and a screening battery of laboratory tests if the patient is elderly or has premorbid conditions. Blood-gas determinations should be performed early in the course of resuscitation. Next, it must be determined where occult blood loss is occur ring. Three sites of hidden blood loss are the pleural cavities (a possibility that can be elimi nated by rapid chest radiography), the thigh, and the abdomen, including the retroperitoneum and pelvis. In most cases, a diagnostic peritoneal la vage will be performed. Table 17-12 shows the interpretation of the results, which has approxi mately an 85 per cent success rate in detecting all abdominal injuries and close to 100 per cent suc cess in intra-abdominal hemorrhage. Common sense dictates that if the patient's chest film is normal and the femur is not fractured, the patient who remains in shock must be suspected of having
Table 17-12 BLUNT TRAUMA: INTERPRETATION OF PERITONEAL LAVAGE Parameter RBCs WBCs Amylase Bile
Negative
Indeterminate
Positive
175 U/dl Present
Abbreviations: RBCs = red blood cells; WBCs = white blood cells.
The next task during the primary survey is to assess neurologic status and to initiate diagnostic and therapeutic procedures. The simplest evalua tion is alertness and response to verbal and painful stimuli. The complete Glasgow Coma Scale is shown in Table 17-13. It should be remembered that apnea occurs with cerebral and cord lesions above C-3, and intercostal paralysis and diaphrag matic breathing occurs with cord lesions above C-6. A decreasing level of consciousness is the single most reliable indication that the patient has a seri ous head injury or secondary insult (usually hy poxic or hypotensive) to the brain. An improving neurologic status reassures the physician that re suscitation is improving cerebral blood flow. Neu rologic deterioration is strong presumptive evi dence of either a mass lesion or important neurologic injury. Computed tomographic scan ning of the head is the preferred technique for diagnosing head injury and should be performed as soon as possible.
Table 17-13 GLASGOW COMA SCALE* I. Eye opening Spontaneous To voice To pain None II. Verbal response Oriented Confused Inappropriate words Incomprehensible words None III. Motor response Obeys commands Purposeful movements (pain) Withdraws (pain) Flexion (pain) Extension (pain) None
4 3 2 l 5 4 3 2 l 6 5 4 3 2 1
Total Glasgow Coma Scale points = I + II + III.
Anesthesia for Emergency Thoracic Surgery d. MECHANISM OF INJURY
The primary survey must take into account the mechanism of injury. There are distinct patterns of injury according to whether the patient has suffered blunt injury (high velocity, low velocity, or crush injury; Table 17-14) or penetrating trauma.
2. Secondary Survey If the patient becomes hemodynamically stable during the initial phase of assessment and resuscitation, it is then appropriate to perform diagnostic studies as indicated to determine surgical priorities.
3. Surgical Priorities (Fig. 17-6) The priorities for surgical intervention are straightforward. A mass lesion in the cranial vault must always be treated immediately. Injuries to the torso are next in line for surgical repair, but they can be treated at the same time as head injuries. As stated previously, the surgeon can easily determine from a chest film whether there is hemorrhage into the chest. Obviously, impalement injuries to the thorax (the imbedded piece should always remain in situ from the field) must go directly to the operating room.147 A patient in shock who has a normal chest film and no other obvious source of blood loss must have hemorrhage within the peritoneal cavity or the retroperitoneum. Such a patient requires immediate laparotomy to control the hemorrhage. Peripheral vascular injuries are assigned the next priority for surgery, and orthopedic injuries the last priority.
637
B. Specific Chest (Noncardiac) Injuries: General Considerations Blunt thoracic injury is usually associated with other injuries to the head, abdomen, or limbs, and it is the associated injuries that often determine the prognosis for the patient. Associated head injury is often severe, and severe head injury is the single most common cause of death in patients with chest injury. The presence of associated abdominal injuries makes respiratory failure and the need for mechanical ventilation more likely. Orthopedic injuries cause patient discomfort and require immobility, which makes management of the chest injury technically difficult. Overall, less than 15 per cent of all patients with chest trauma will require a thoracotomy. The management of those patients near death from chest trauma requiring immediate thoracotomy upon arrival at the hospital is covered under Emergency Room Thoracotomy in the Management of Trauma. Patients requiring less urgent thoracotomy after chest trauma may have a variety of thoracic injuries. Figure 17-8 shows the variety of primary and secondary thoracic injuries, and the inter-relationships of the injuries, as a traumatic force penetrates inward. Most of the initial primary injuries can result in major chronic secondary complications.
1. Chest Wall Fractures (Flail Chest) Flail chest may be defined as an abnormal movement of the chest wall occurring as a result of fractures of two or more ribs in two places on the same side. If a scapula or sternum is fractured, high energy transfer has occurred to the thorax, and one must suspect that deeper structures have
Table 17-14 MECHANISMS AND PATTERNS OF BLUNT THORACIC TRAUMA* Type of Impact
Chest Wall Injuries
Possible Thoracic Visceral Injuries
Common Associated Injuries
High Velocity (Deceleration)
Chest wall often intact or fractured sternum or bilateral rib fractures with anterior flail (caused by steering wheel)
Ruptured aorta Cardiac contusion Major airways injury Ruptured diaphragm
Head and faciomaxillary injuries Fractured cervical spine Lacerated liver or spleen Long bone fractures
Low Velocity (Direct Blow)
Lateral-unilateral fractured ribs Anterior-fractured sternum
Pulmonary contusion Cardiac contusion
Lacerated liver or spleen if ribs 6-12 are fractured
Ruptured bronchus Cardiac contusion
Fractured thoracic spine Lacerated liver or spleen
Crush
Anteroposterior crush—bilateral rib fractures with or without anterior flail Lateral crush—ipsilateral fractures with or without flail Possible contralateral fractures
Pulmonary contusion
Lacerated liver or spleen
'From Westaby S, Brayley N: Thoracic trauma—I. Br Med J 300:1639-1643, 1990. Used with permission.
638
Anesthesia for Emergency Thoracic Surgery
been injured. With a flail chest, each rib section between the two fracture sites essentially floats free. With spontaneous inspiration and the development of a more negative intrathoracic pressure, the injured segment moves inward upon the lung (paradoxic chest wall movement), preventing lung expansion and impairing oxygenation. The reverse occurs during a spontaneous exhalation. The underlying lung is frequently contused, which greatly contributes to the poor gas exchange.148 This injury most commonly involves the anterolateral aspect of the chest. The posterior wall is heavily fortified with muscle and the scapula and, therefore, is rarely involved. The thoracic cage in older age groups is more calcified and brittle and, therefore, more susceptible to flail and other serious injuries. In contrast, the thoracic cage in pediatric patients is extremely elastic and resilient and affords much greater protection to underlying structures. An opening in the chest wall allows atmospheric air to enter the pleura, permitting the intrapleural pressure to rise toward atmospheric pressure and producing pneumothorax and collapse of the lung. The result is the familiar "sucking wound," with air entering the pleural space during inspiration and exiting during exhalation.
Laceration of intercostal vessels may produce hemothorax; bleeding may continue until thoracotomy becomes necessary. When the wound is parasternal, the internal mammary vessels may be injured. Bleeding from the intercostal or internal mammary arteries, being under systemic pressure, is not tamponaded by hemothorax and usually continues until shock or frank pulmonary insufficiency occurs. Surgical ligation is usually required. The major concerns with chest wall and sternal injuries are correction of inadequate gas exchange resulting from flail chest, the provision of adequate analgesia (to decrease splinting), and the diagnosis of possible underlying injury. Indications for institution of mechanical ventilation are severe hypoxemia, severe hypercarbia, clinically obvious excessive work of breathing (usually present with major flail and paradoxic chest wall movement), and severe associated injuries (e.g., parenchymal contusion and hemothorax). Associated injuries, especially parenchymal contusions, are likely to be present with sternal fractures and fractures of the upper ribs because these fractures require a large force in order to occur. For example, with sternal injuries, one should always suspect rupture of the trachea, rupture of major arteries (Fig. 17-9), and,
Anesthesia for Emergency Thoracic Surgery
639
sive hemothorax have severe hypotension.157 In addition, as blood progressively accumulates in the pleural space, the underlying lung is compressed. The chest is dull to percussion, the breath sounds are diminished, and a shift in the mediastinum toward the uninvolved hemithorax (displaced apical cardiac impulse and trachea) may compress the uninvolved lung; the compression of the contralateral lung will exacerbate the ventilatory impairment. Thus, these patients also can have severe hypoxemia.157 b. PNEUMOTHORAX
Figure 17-9 Rupture of the trachea and avulsion of the innominate artery by a fractured manubrium. (From Pate JW: Tracheobronchial and esophageal injuries. Surg Clin North Am 69:111. 1989. Used with permission.)
especially, myocardial injury, because the heart is compressed between the sternum and the vertebrae.149 Some patients can be effectively treated without intubation and mechanical ventilation if they are provided with adequate analgesia (see Anesthetic Considerations and chapter 21); in addition to increasing lung volume (by decreasing splinting), adequate analgesia decreases the respiratory rate, which, in turn, makes the overall flail of the chest much less pronounced.150·I51 In summary, the approach to flail chest injuries should be flexible.151 152 In a prospective study of 36 patients with this injury,151 only 30 per cent required mechanical ventilation, even though 69 per cent developed pneumonia and about 30 per cent developed the adult respiratory distress syndrome. In the ventilated patients, the mean duration of ventilation was 10.5 days. The overall mortality was approximately 10 per cent. Depending upon the series, the mortality may vary between 6 and 50 percent. 153 154
2. Pleural Space (Hemothorax, Pneumothorax) and Pulmonary Parenchyma (Contusions) a. HEMOTHORAX155
l56
The hemithorax is a potential space that can easily accommodate 30 to 40 per cent of a patient's blood volume (greater than 2000 ml in a 70-kg adult). Consequently, patients with a mas-
Pneumothorax occurs secondary to blunt or penetrating trauma to the chest wall. With penetrating trauma the mechanism of air escape is obvious; the most common cause of penetrating pneumothorax is iatrogenic and is due to subclavian puncture, which has an overall incidence of pneumothorax of 2 per cent.158 Penetrating knife wounds and gunshot wounds have close to a 100 per cent incidence of pneumothorax. With blunt trauma the mechanism of air escape may be tearing of the lung by the edge of a rib fracture or development of an alveolar bursting force (with a closed glottis) during the trauma. With a pneumothorax, ipsilateral breath sounds are decreased, and the chest is tympanic to percussion. A tension pneumothorax occurs when air enters the pleural cavity during inspiration but, owing to a ball-valve action, cannot escape during exhalation. A tension pneumothorax can cause circulatory embarrassment because of decreased venous return and mediastinal shift and causes respiratory embarrassment via compression of ipsilateral lung by direct tension and contralateral lung by mediastinal shift. An open pneumothorax communicates with the environment and has been referred to as a "sucking" chest wall defect. The sucking refers to air being drawn into the hemithorax from the environment during a spontaneous inspiration; the physiology of the situation is identical to the paradoxic respiration and mediastinal flap described in chapter 4. Pneumomediastinum and pneumopericardium in the adult trauma patient are often incidental findings of air that has tracked along the bronchial or vascular sheaths of the thorax into the mediastinum and pericardium.159 Occassionally, this can be a grave sign of a ruptured bronchus or esophagus, uncontrolled pneumothorax, or ventilator barotrauma. Air within the pericardial space can develop into cardiac tamponade. c. PULMONARY CONTUSION
Pulmonary contusions occur either with penetrating lung injury or as the result of rapid decel-
640
Anesthesia for Emergency Thoracic Surgery
eration forces (blunt injury), which cause the lung to forcibly impact upon the chest wall.155, 156 Rib fractures occur in approximately 50 per cent of the cases of pulmonary contusion. Deceleration forces cause rupture of alveoli and vessels with hemor rhage, and increased endothelial and epithelial permeability causes interstitial and intra-alveolar edema. The edema phase develops progressively with time (hours) and with fluid administration and is accompanied by progressive chest X-ray changes and a decrease in P a 0 2 . Consequently, the initial chest roentgenogram is often not a good indicator of the severity or extent of the contusion. However, with computed tomographic scanning, the lesion can be seen immediately.160
3. Tracheobronchial Disruptions Tracheobronchial disruptions should be sus pected in any patient with either penetrating or blunt trauma to the neck or chest, especially if accompanied by subcutaneous or mediastinal em physema (escape of air into adjacent tissues), he moptysis (bleeding within the bronchus), pneumo thorax (if no chest tube is in place), and/or BPFs (escape of air into the pleural space and out the chest tube, if present). It should be noted that a lung that is lacerated by a knife may have many small transected bronchioles and behave like a Bpp iei although the trachea and major bronchi may be involved at any level (especially with pen etrating trauma), greater than 80 per cent of inju ries with blunt objects are within 2.5 cm of the carina (see later discussion). 162-165 Penetrating trauma to the neck may cause combined tracheo esophageal injuries.166 Tracheobronchial disruptions due to blunt ob jects require a large force and are, therefore, often associated with trauma to adjacent structures.167 In fact, in one series of tracheobronchial disruptions due to blunt trauma, there was an average of 3.8 injuries per patient.167 The major late complica tions of tracheobronchial disruption consist of BPF, empyema, and mediastinitis. Proximal tra cheobronchial tree injuries, which cause dramatic clinical symptoms, are often recognized and treated earlier and, therefore, have a better out come than distal injuries, which involve a smaller volume of air, go undetected longer, and have higher incidence of major late complications. Flex ible fiberoptic (or rigid) bronchoscopy should be performed if the diagnosis of tracheobronchial tree disruption is being considered.168 There are three reasons why tracheobronchial disruptions anatomically localize around the ca rina. 169 First, anterior chest compression causes rapid lateral movement of the lungs (pulls the lungs apart), which creates a shearing force on the
carina. Second, an increase in intrathoracic pres sure with the glottis closed causes the greatest wall tension (and, therefore, risk of tear) to occur in the airways with the largest diameter (Laplace's rela tionship Τ - Ρ X R, in which Τ = tension, Ρ = transmural pressure difference, and R = radius of curvature). Third, the fixation of the trachea above (cricoid cartilage) and below the carina predis poses the more mobile carina to shear forces.
4. Esophageal Disruptions^5 l56 17° (see Chapter 16) Esophageal ruptures occur from internal trauma (after medical instrumentation of any type, inges tion of penetrating or corrosive material); as a re sult of external trauma (blunt and penetrating); spontaneously (postemetic); and as part of the nat ural history of a pre-existing esophageal lesion (e.g., tumor and caustic burns) (see chapter 16). Unrecognized and untreated traumatic esopha geal injury has an extremely high late mortality (20 per cent) due to mediastinitis, empyema, and sepsis. 171 · 172 In a recent series, all patients operated on within 24 hours of perforation survived, whereas there was a 33 per cent mortality when an operative delay of more than 24 hours occurred.173 Unfortunately, esophageal injury may go unrecog nized for hours to days in a multiply traumatized patient when attention is not focused on the esoph agus. The diagnosis should be suspected in pa tients who have chest pain, dysphagia, hematemesis, cervical or mediastinal emphysema, and fever (especially after an endoscopic procedure, removal of a foreign body, or trauma). Pain is the most striking and consistent symptom. The chest X-ray is often suggestive by revealing subcutaneous emphysema, pneumothorax (rupture of mediastinal pleura), pleural effusion (also caused by tear of the mediastinal pleural, resulting in esophageal contents irritating the pleural space), pneumoperitoneum, and retropharyngeal swelling. However, chest X-ray findings take several hours to develop and depend on the site of perforation.174 If a chest tube has been inserted because of asso ciated injuries, continuous bubbling may reflect an air leak from the esophagus, and drainage of par ticulate matter should raise suspicion of an esoph ageal tear. The diagnosis can be confirmed with meglumine diatrizoate (Gastrografin) roentgenographically (esophagogram). Esophagoscopy is frequently unnecessary. If the tear is diagnosed early (less than 6 hours), primary repair is accom plished, often with good results.
5. Diaphragmatic Disruptions Diaphragmatic injuries can occur secondary to either penetrating or nonpenetrating (blunt)
Anesthesia for Emergency Thoracic Surgery·
trauma.175-177 Penetrating trauma as high as the fourth intercostal space can penetrate the diaphragm (at the domes) and cause intra-abdominal injury. Alternatively, stab wounds of the abdomen (subcostal) can penetrate the diaphragm (at the lower circumference border) and cause intrathoracic damage. Extreme forces are necessary to cause diaphragmatic rupture with blunt trauma; therefore, there is a high incidence of associated injuries and deaths with blunt diaphragmatic trauma (median ISS = 34).I78 Seventy-five to 95 per cent of diaphragmatic tears after blunt trauma occur on the left side because of the protection afforded the right side by the liver. There are several important pathophysiologic consequences of diaphragmatic tears. First, if the disruption is extensive, diaphragmatic movements will be ineffectual and the thoracic cage will behave as a flail chest. Second, abdominal pressure is greater than thoracic pressure; therefore, there is always the potential for abdominal viscera to work their way into the thorax through the diaphragmatic defects. Small diaphragmatic tears are more likely to capture and strangulate abdominal organs. The organs that most commonly herniate (in decreasing frequency) are stomach, small bowel, spleen, omentum, liver, and kidney.I78· l79 Third, herniation of abdominal contents into the thorax can cause lung compression, mediastinal shift, and decreased venous return.
C. Specific (Noncardiac) Chest Injuries: Surgical Considerations 1. Chest Wall Fractures (Flail Chest) The treatment varies with the severity of the injury, ranging from simple supportive therapy such as oxygen enrichment, physical therapy, and pain management to CPAP by mask to full ventilatory support. Although mechanical ventilation is obviously required for patients in whom acute or subacute respiratory failure develops, most patients can be managed by 5 cm H 2 0 CPAP via mask.'48 CPAP by mask is effective because in many patients it is the underlying contused lung that is primarily responsible for the respiratory difficulty. Control of pain by epidural narcotics is an important component of effective treatment of this situation. Other much less frequently used treatments consist of external stabilization of the flail segment by traction on the injured segment'80 and stabilization of the chest wall by intramedullary pinning of fractured ribs' 8 '; indications for these procedures are gross instability of a large segment of the chest wall and unremitting pain caused by the movement of the rib fracture.'82
641
2. Pleural Space (Hemothorax, Pneumothorax) and Pulmonary Parenchyma (Contusions) a. HEMOTHORAX About 500 ml of blood must accumulate in the chest cavity before it is detectable radiologically. Upright chest films are preferable, if possible. Tube thoracostomy allows immediate confirmation of the diagnosis. Hypovolemia from blood loss is the most common presenting problem in patients with significant hemothorax. Therefore, the immediate treatment should be directed toward restoring blood volume. Thoracostomy tube insertion (usually in the sixth intercostal space in the midaxillary line) alone is the only surgical treatment required in more than 80 per cent of patients with hemothorax.'83 In the majority of cases, the source of bleeding is the pulmonary vessels, which normally have low perfusion pressures. Although chest tube drainage is the first step in the treatment of hemothorax, one must remember that chest tube drainage can occasionally release a major vessel tamponade and cause rapid exsanguination. Patients who have responded adequately to volume replacement, have less than 150 ml/hour thoracostomy tube output after initial drainage, or have reexpanded the injured lung on chest X-ray without reaccumulation of hemothorax do not require emergency thoracotomy. In contrast, bleeding from systemic vessels is usually much more persistent and voluminous and usually requires thoracotomy (for blood loss greater than 300 ml/hour for 4 hours). Sometimes residual clot remains in the thorax following tube thoracostomy drainage of a hemothorax. The residual clotted blood remaining in the pleural cavity has the potential for causing empyema or pleural fibrosis with lung entrapment. Consequently, some authors recommended early thoracotomy to evacuate the residual blood and decorticate any early fibrosis process.127 b. PNEUMOTHORAX
Pneumothoraces of less than 20 per cent are usually not detectable clinically. Patients with larger pneumothoraces (20 to 40 per cent) usually complain of chest pain, which is accentuated by deep breathing. With even larger pneumothoraces (40 to 60 per cent), cyanosis may be evident and the trachea may be deviated. The presence of rib fractures or tissue emphysema should suggest the diagnosis. Radiologic examination is the best diagnostic aid available, and all films should be taken during expiration. A simple closed pneumothorax (greater than 10 per cent on chest X-ray) requires only a chest tube;
642
Anesthesia for Emergency Thoracic Surgery
further surgical intervention is not required unless there are underlying injuries. Minor-sized penetrating defects can be sealed with Vaseline gauze, dressing, and tape and with chest tube evacuation of the pleural space. A suspected tension pneumothorax (by clinical signs or chest X-ray) requires immediate decompression; this should first be done by insertion of a needle in the second intercostal space in the midclavicular line. A tension pneumothorax is a dire emergency that may get worse, and valuable time should not be wasted seeking radiologic confirmation. If the diagnosis was correct, rapid improvement will be observed. The needle should be followed by chest tube insertion. Large-sized open defects often require debridement and primary closure. It is especially important to decompress even a very small pneumothorax if the patient is going to be transported by air. Most commercial aircraft are pressurized to 5000 to 6000 feet. At these altitudes, Boyle's law (the volume of a gas is inversely proportional to pressure) becomes important because gases will expand 20 to 25 per cent at this altitude compared with their volume at sea level. Decompression of existing pneumothoraces is extremely important. Consideration should be given to the effect of trapped gases in other areas of the body (e.g., pass a nasogastric tube) or in medical devices. c. PULMONARY CONTUSION
There are no special surgical considerations for pulmonary contusions (they are not resected). The contusion must be watched for the possible development of infection and abscess formation within the contusion.
3. Tracheobronchial Disruptions A persistent pneumothorax and air leak after tube thoracostomy and the presence of subcutaneous and mediastinal emphysema are suggestive of a tracheobronchial tree disruption. The few cases that can be treated nonsurgically include small distal tears with minimal air leak, those in a major bronchus that are less than one third of the circumference and have no air leak, and small tracheal wounds with good apposition of the margins. Small- to moderate-sized high cervical tracheal wounds may be treated by endotracheal intubation, with the cuff of the endotracheal tube placed below the wound site.184 Tracheostomy is indicated when there is extensive injury to the cervical trachea and larynx or when endotracheal intubation cannot be performed. Surgery is indicated for most tracheobronchial disruptions. Surgery should be preceded, if possible, by diagnostic bronchoscopy to identify the site
and severity of lesion. The surgical approach is a right thoracotomy for right-sided and tracheal lesions and a left-sided thoracotomy for left-sided lesions. The procedures consist of primary suture repair if technically feasible and, if not, lung resection.167 The long-term sequela of these injuries is airway stenosis.
4. Esophageal Disruptions (see Chapter 16) All traumatic (caused by external causes), mitotic, and spontaneous esophageal tears must be treated surgically as soon as they are diagnosed; any delay results in a sharp increase in mortality. Because there are false-negative results both with esophagography and esophagoscopy, surgical exploration is sometimes undertaken empirically just on the basis of a very high index of suspicion.174· 185 Surgical procedures range greatly in aggressiveness and consist of primary repair and drainage (adjacent wall tissue normal), pure drainage procedure (adjacent wall tissue injured), esophageal exclusion, and esophagogastrectomies173 (see chapter 16). Tears of the upper and middle thirds of the esophagus are repaired through a right thoracotomy, and tears of the lower third are repaired through a left thoracotomy (see chapter 16). Unfortunately, extensive esophageal injuries often leak after repair and result in the late complications of tracheoesophageal fistula, mediastinal abscess, wound infection, and carotid artery blowout.166 Patients with a small, well-contained perforation due to a medical instrument (internal trauma) can be treated nonoperatively with antibiotics and intravenous alimentation.173 Nonoperative management is successful in these cases because the perforations are small, and frequently patients undergoing medical instrumentation have esophageal disease with periesophageal inflammation and fibrosis, which limit the spread of the mediastinal contamination.
5. Diaphragmatic Disruptions Patients with massive herniation of intra-abdominal contents into the thorax with obvious embarrassment of pulmonary function ("tension enterothorax"186) require immediate operative intervention. If the diaphragmatic injury is not lifethreatening, repair can be undertaken after diagnostic confirmation by chest X-ray (observation of elevated diaphragm, hollow viscus in the chest ["loculated pneumothorax"], obscured diaphragmatic shadow, and a pneumothorax or hemothorax when the entry wound is abdominal), diagnostic
Anesthesia for Emergency Thoracic Surgery
pneumoperitoneum (which causes a pneumothorax), and hemodynamic stabilization. Similarly, abdominal findings with a chest entry wound indicate a diaphragmatic tear. Finally, if a chest tube is in place, appearance of peritoneal lavage fluid in the chest tube drainage and a nasogastric tube or radiocontrast material that curls back up into the chest after entering the abdomen are virtually diagnostic of a diaphragmatic tear. Nevertheless, in both acute and subacute situations, diaphragmatic tears are still most often discovered at the time of surgical exploration of associated injuries.178·I86 However, undue delay is associated with an increased incidence of organ strangulation and perforation. In fact, patients with a chronic diaphragmatic tear most often have symptoms of intestinal obstruction. The surgical approach is most often abdominal (the source of associated injuries and bleeding (which is the reason a body cavity is being explored) is usually in the abdomen178·l86 but may be thoracic, or both, depending on the nature of the associated or concomitant injuries.178 ISf^188
D. Specific Chest (Noncardiac) Injuries: Anesthetic Considerations
643
nous catheters and finding one good one for drugs; (5) inserting blood warmers in the intravenous catheter lines; (6) getting blood into the room, verifying patient identity, and beginning transfusion; (7) starting an arterial catheter; (8) drawing blood gases and a hematocrit; (9) titrating the anesthetic if possible; (10) checking temperature; (11) measuring urine output; and (12) inserting a central venus or pulmonary arterial catheter. This last procedure should be done only after initial resuscitation has been completed or when time permits. The choice between crystalloid and colloid infusion is discussed at length in chapter 13. Successful fluid resuscitation is indicated by a decreasing pulse rate (below 100 beats/min), a pulse pressure of greater than 30 mm Hg, adequate urine output (greater than 1 ml/kg/hour), minimal metabolic acidosis, and the lack of a large respiratory swing in arterial pressure. Other intraoperative concerns in extensively traumatized patients are persistent hypotension in spite of apparently adequate volume replacement (which may be due to an occult blood loss, air embolism, pericardial tamponade, myocardial contusion, head injury, and so on), hypothermia, acidosis, and coagulopathy (see chapter 13). •
The anesthetic management of patients with thoracic trauma is complicated by the fact that there are often multiple, life-threatening, interrelated problems. In one large series of acute, blunt, thoracic trauma patients, three major problems/concerns were very apparent.'41 First, virtually all intraoperative deaths are due to exsanguination. Second, many of the injuries indicated that the patients should be intubated awake (e.g., there was a possible cervical spine injury, facial trauma, severe shock). Third, increased airway pressure was a common intraoperative problem and, in order of decreasing frequency, the increased airway pressure was due to hemorrhage into the tracheobronchial tree, pulmonary edema, intraoperative pneumothorax, lung collapse, and aspiration. In the vast majority of these patients, chest tubes should be inserted before the initiation of positivepressure ventilation. Finally, patients who have a myocardial contusion are at high risk for lifethreatening arrhythmias and hemodynamic instability unrelated to bleeding (6 per cent incidence).189 The anesthesiologist must do many things very quickly when beginning to administer an anesthetic to a patient with extensive thoracic trauma, and these tasks must be done according to priority: (1) ventilation; (2) measurement of blood pressure; (3) attachment of an EKG; (4) sorting out intrave-
1. Chest Wall Fractures (Flail Chest) Patients with chest wall injuries that require surgical repair also require general anesthesia, intubation, and mechanical ventilation. Analgesia for continued mechanical ventilation is most effectively provided by continuous lumbar or thoracic epidural narcotics and can be a major determinant of morbidity and mortality (see chapter 21).
2. Pleural Space (Hemothorax, Pneumothorax) and Pulmonary Parenchyma (Contusions) a. HEMOTHORAX Intravascular volume should be replaced through large-bore intravenous lines. Preoperatively, the potential for sudden massive blood loss with chest tube insertion should be remembered. Intraoperatively, the use of an autotransfuser should be strongly considered. Acute respiratory failure may occur prior to surgery and may require intubation and mechanical ventilation. A doublelumen tube should be considered if there is a large concomitant air leak out the chest tube (raising the possibility of a tracheobronchial tree disruption) or if there is hemoptysis or a significant amount of blood in the airway. Other specific anesthetic considerations depend upon associated injuries and the
644
Anesthesia for Emergency Thoracic Surgery
specific site of bleeding (which may not be diagnosed until later; for example, following artériographie demonstration of an aortic disruption). b. PNEUMOTHORAX
The only surgical response to a pneumothorax that requires general anesthesia is debridement and primary closure of a large open pneumothorax. Of course, if a pneumothorax is associated with other injuries, general anesthesia will likely be required for the treatment of the other injuries. In all cases involving a potential pneumothorax, the anesthesiologist must consider the possibility of converting a small, untreated, simple, closed pneumothorax to a large tension pneumothorax during the induction of anesthesia and the initiation of intermittent positive-pressure breathing. If chest tubes have been inserted, they must be monitored for continued function (i.e., if they fail to function properly, tension pneumothorax can still occur). The signs of the development of a tension pneumothorax consist of decreased breath sounds, decreased compliance, deteriorating arterial bloodgas values, tracheal deviation, and cardiovascular collapse. As soon as a tension pneumothorax is suspected, a large-bore needle should be inserted into the pleural cavity, allowing air to drain freely. Nitrous oxide should be avoided if a pneumothorax is a possibility.190
tion, partial tracheal injuries can be converted to full tears by such an intubation. Separation of the lungs is often required and is often life-saving.161 Because positive-pressure ventilation may convert a simple mucosal tear to a major BPF or pneumothorax, consideration should be given to maintaining spontaneous ventilation during the induction of anesthesia, endotracheal intubation, and the maintenance of anesthesia if a single-lumen tube is used. Double-lumen tubes should be used for injuries located at or below the carina. Alternatively, small catheters can be passed below injuries near the carina for HFV and highflow apneic ventilation (see chapter 12). Injuries above the carina are best handled by passing single-lumen or double-lumen tubes past the injury and ventilating one (or both) of the lungs. With high cervical tracheal transections, it is advisable to use a fiberoptic bronchoscope as a stent across the transection. It should be remembered that tracheostomy under local anesthesia may be the safest way to establish the airway in severely traumatized patients, and that gentle positive-pressure ventilation will be necessary when the chest is opened. The anesthesiologist should have sterile endotracheal tubes of various sizes available for bronchial placement from within the chest during airway repair.
C. PULMONARY CONTUSION
4. Esophageal Disruptions (see Chapter 16)
It is important to remember that the edema phase of a pulmonary contusion may coincide with the intraoperative period when fluid is necessarily administered. The edema phase will be accompanied by a progressive decrease in Pa02 and compliance and should be treated with PEEP and perhaps fluid restriction and diuretics. The type of fluid infused (colloid versus crystalloid) is not a critical matter since the permeability characteristics of the involved areas are grossly deranged and the area will become edematous regardless of the type of fluid infused.
Anesthesia for esophageal surgery involves respiratory, hemodynamic, and gastrointestinal considerations (see chapter 16). Most important, surgical exposure can be greatly facilitated by doublelumen tube insertion and use of one-lung ventilation. In patients with esophageal disruption, further esophageal instrumentation, such as insertion of esophageal stethoscope and nasogastric tube, is contraindicated. At the end of the case, the surgeon may gently guide a nasogastric tube or esophageal stent past the injured area. 5. Diaphragmatic Disruptions
3. Tracheobronchial Disruptions Whether the patient is intubated awake or anesthetized, with or without spontaneous ventilation, using a single-lumen tube versus a double-lumen tube, inspection of the trachea with a fiberoptic bronchoscope and passage of the endotracheal tube over the direct vision fiberoptic bronchoscope guide are good ideas. In cases of complete tracheal transection, blind' endotracheal intubation (nonvisualization of the trachea per se) may be hazardous because it may be difficult to slide by the tear and intubate the distal trachea in this manner. In addi-
Patients, whether having laparotomy or thoracotomy as the initial approach, may be prepared and draped into the "corkscrew" position to allow exploration of the second body cavity if appropriate.191 The pelvis lies flat on the operating table, but the thorax and shoulders are rotated, and the uppermost arm is brought across the table on a support. Optimal access to both the chest and abdomen is afforded by appropriate lateral tilt of the operating table. If the injury is approached by thoracotomy, then surgical exposure can be greatly facilitated by dou-
Anesthesia for Emergency Thoracic Surgery
ble-lumen tube insertion and use of one-lung ventilation. Decompression of the stomach may significantly improve the hemodynamic and ventilatory status. The approach to weaning and extubation should be conservative in view of the fact that chest and diaphragm trauma can induce a marked decrease in diaphragmatic displacement on the injured side.192
VII. TRANSVENOUS PULMONARY EMBOLECTOMY A. General and Surgical Considerations
Greenfield and Stewart classified patients suffering from pulmonary embolism into two groups based on physiologic parameters (Table 17-15).193 The first consists of class I, II and III patients who can be managed medically. The second group includes classes IV and V who have either a massive pulmonary embolus or a pulmonary embolus associated with pulmonary or myocardial disease194 and demonstrate a very high morbidity and mortality (50 per cent dead in 30 min, 70 per cent in 1 hour, 85 per cent by 6 hours).195·196 Several methods of treatment for acute pulmonary embolism have been described, including heparin therapy, streptokinase, and surgery (open pulmonary embolectomy). In class IV and V patients, medical therapy has little to offer because of the associated high early mortality. Those patients who are treated in the acute situation with open embolectomy continue to demonstrate a significant morbidity and mortality (approaching 50 per cent).197 As a result, the technique of transvenous pulmonary embolectomy has been advocated by Stewart and Greenfield.198
645
Indications for transvenous pulmonary embolectomy are similar to those for open pulmonary embolectomy (i.e., class IV and V patients with a strong clinical suspicion for pulmonary embolism). These patients, if identified at an early stage, could be saved by rapid pulmonary embolectomy. The technique of transvenous pulmonary embolectomy has been described by Stewart and Greenfield.198 The catheter is steerable, measuring 100 cm in length, with a suction cup at the end. It is inserted through the femoral venotomy to the right atrium and ventricle to the pulmonary outflow tract. After artériographie localization of the embolus, the distal cup is placed in close proximity to the embolus, and suction is applied. The embolus is then extracted via the femoral venotomy along with the catheter. Frequently, multiple attempts are necessary to remove enough clot to achieve hemodynamic stability, as evidenced by recovery from shock and a significant improvement in the Pa02. Complications of the procedure include ventricular arrhythmia, inability to extract the embolus, perforation of the pulmonary artery,198 and fatal hemorrhage197 after reperfusion of the lungs. At the completion of the pulmonary embolectomy, a Greenfield filter is inserted through the femoral venotomy. The filter is inserted to protect against further embolization in an already hemodynamically compromised patient. The filter is inserted via the femoral or jugular route over a guidewire using fluoroscopy. It is positioned in the inferior vena cava at the level of the third lumbar vertebra. Systemic anticoagulation should be continued after the conclusion of the procedure because the filter protects against further embolization but does not treat the initiating problem (i.e., deep venous thrombosis). Complications of Green-
Table 17-15 CLASSIFICATION OF PULMONARY THROMBOEMBOLISM1
*From Stewart JP, Greenfield LJ: Transvenous vena cava filtration and pulmonary embolectomy. Surg Clin North Am 62:411430, 1982. With permission of the publisher. Abbreviations: CVP = central venous pressure; PA = mean pulmonary artery pressure; BP = blood pressure; CO = cardiac output.
646
Anesthesia for Emergency Thoracic Surgery
field filters include recurrent embolism, retroperi toneal hemorrhage, recurrent thrombophlebitis, pe ripheral edema, and filter migration.
VIII. EMERGENCY ROOM THORACOTOMY IN THE MANAGEMENT OF TRAUMA
A. General Considerations B. Anesthetic Considerations1 Appropriate monitoring should include arterial catheter placement for the second-to-second obser vation of blood pressure and to obtain arterial blood gases as indicated. A large-bore peripheral or central intravenous catheter is indicated for rapid infusion of volume and/or vasoactive drugs. Pulmonary artery pressure monitoring is desirable, and the values are obtained from the angiographic catheter. A pulmonary artery catheter should be placed subsequently through the internal jugular vein (because the insertion of a Greenfield filter is via the femoral vein) as early as possible after the embolectomy both to determine pulmonary artery pressures and to aid in evaluation of hemodynamic status. Monitoring should also include an EKG, which should be able to give a lead II reading. In more than one third of patients with pulmonary embo lism, the EKG is normal, whereas other patients will show various transient abnormalities of the ST segment and Τ wave in the precordial leads.'99 However, tachycardia, both supraventricular and ventricular, and bradycardia often result from pul monary embolus manipulation and therefore a lead II or transesophageal lead is best elected. Anesthetic management requires a motionless patient. If paralysis were used alone, awareness might result; therefore, some form of sedation or general anesthesia without cardiovascular depres sion is desirable. Pulmonary embolectomy can be associated with a significant blood loss. Large amounts of blood loss may not be readily evident because of inaccessibility to the patient's lower extremity. This patient will generally be on an anticoagulant or thrombolytic agent and therefore may bleed profusely from the cut-down site. Bleeding secondary to heparin can be treated with protamine, and bleeding secondary to streptoki nase may respond to treatment with epsilon aminocaproic acid. There appear to be two main periods when large-volume infusions are necessary. The first is when the embolus is removed, with reperfusion and its associated complications. The second is when the Greenfield filter is being inserted, be cause of the surgical incision and bleeding in an anticoagulated patient.
The use of emergency room thoracotomy in the management of trauma has increased greatly in the last decade. Thoracotomy in the emergency room is used when the patient's condition is thought to be so dire that it precludes taking the time neces sary to transfer the patient to the operating theater. Emergency room thoracotomy has been performed by surgeons as well as emergency room physicians in a wide variety of clinical conditions. As experi ence with the procedure accumulates, it has be come apparent that the mode of injury (blunt ver sus penetrating trauma), site and extent of injury, the condition of the patient at the scene of injury (e.g., presence of vital signs, reactivity of pupils), and the rapidity of transport are the most important determinants of the outcome. The procedure is most useful in penetrating lung injuries (the lung hilum can be clamped so as to control lung bleed ing and prevent pulmonary venous air embolism), is progressively less useful in penetrating cardiac injuries (pericardial tamponade can be relieved, cardiac bleeding can be controlled with digital pressure and internal defibrillation, and cardiac massage can be performed) and penetrating ab dominal wounds (thoracic aorta can be occluded to provide proximal control of massive abdominal arterial bleeding), and is least successful in blunt thoracoabdominal injuries200 (Table 17-16). Although penetrating lung injuries continue to have the best results (survival) with emergency thoracotomy, the results of the most recent large series are not nearly as impressive as those in Table 17-16, perhaps because patient selection was very liberal.201 If the patient is alive at the scene of injury, the ultimate survival of critically injured patients will be increased by minimizing the use of time-wasting field stabilization measures (e.g., starting many intravenous lines202 and apply ing military antishock trousers203) and by trans porting the patient as rapidly as possible to the hospital.204 Lack of ventricular activity or a palpa ble pulse and the absence of reactive pupils at the scene of injury carry an almost uniformly fatal prognosis.201 Patients who are clinically dead at the scene of injury and who remain so throughout transport do not benefit from heroic emergency room thoracotomy.201,205 In penetrating injuries of the thorax, the lung is the most frequently injured organ.206 The majority of these injuries may be treated with just tube thoracostomy206 (see Chest Trauma). However,
Anesthesia for Emergency Thoracic Surgery·
647
Table 17-16 RESULTS OF IMMEDIATE EMERGENCY ROOM THORACOTOMY*
Type of Injury Penetrating lung Penetrating cardiac Penetrating abdominal Blunt thoracoabdominal
Number of Cases (Number of Reports) 1699 (7) 324 (14) 194 (6) 252 (7)
Number of Thoracotomies (%)
Mortality (%)
321 (19) All
9 71
All
95
All
96
*Based on data from Bodai et al.20"
with penetrating thoracic battle injuries, the number of patients requiring thoracotomy is significantly increased.207 Still, of those who require thoracotomy for penetrating injury, only a small fraction ever require thoracotomy in the emergency room. These patients are the ones who arrive in extremis, and thoracotomy performed in the emergency room may be life-saving200·2()8 (Table 17-16). The prehospital mortality of penetrating cardiac wounds is high (38 to 83 per cent).200·208 Patients with penetrating cardiac injuries who survive the initial injury but show signs of acute decompensation on or just before arrival in the emergency room may also benefit from emergency room thoracotomy. The majority of these survivors have cardiac tamponade (as has been demonstrated by previous immediate thoracotomy experience); simple pericardiocentesis is recommended only to transiently improve hemodynamic function, allowing the necessary time for transfer to the operating room. The more aggressive approach of performing thoracotomy in the emergency room has resulted in a significant decrease in the mortality due to this injury209·210 (Table 17-16). Thoracotomy in the emergency room has been used for injured patients with arterial abdominal bleeding. The rationale for this practice is that proximal control of the bleeding by thoracic aortic cross clamping prior to release of the tamponading effects of blood in the intact abdominal cavity would increase survival following abdominal decompression. However, experience with patients with penetrating abdominal injuries subjected to emergency room thoracotomy discloses an extremely low survival rate200·208 (see Table 17-16). The difference between the relatively high success rate with lung injunes compared with that of abdominal injuries can probably be explained by two factors. First, emergency thoracotomy and aortic occlusion will not substantially affect the rate or volume of bleeding from major abdominal venous injuries. Such venous injuries include penetrating
injuries to the liver, the vena cava, and major portal venous structures. The second factor is that many of these patients with penetrating abdominal injury did not have demonstrable vital signs before the performance of thoracotomy. The thoracotomy was undertaken as a "last-ditch," heroic attempt in the emergency room to save the patient. Thus, the trauma victim who has suffered a cardiac arrest from abdominal hemorrhage has a slim chance of recovery and, in most instances, is not a candidate for emergency room thoracotomy.200·208 Only patients in whom there is a strong suspicion of major intra-abdominal arterial injury, and who still have signs of life, are candidates to survive emergency room thoracotomy. The alive but severely hypovolemic patient with penetrating abdominal injury and massively distended abdomen should go immediately to the operating room, where the surgeon may elect to perform a left anterior thoracotomy with aortic cross clamping in a thorough manner before opening the abdomen.2" Patients with blunt injury severe enough to necessitate emergency room thoracotomy also demonstrate uniformly poor results200·206 2()8 (see Table 17-16). The reasons for these dismal results probably include widespread damage, venous bleeding, and much higher incidence of head trauma.212
B. Surgical Considerations A left anterolateral thoracotomy is most often performed for aortic cross clamping and left hilar clamping. The pericardial sac should be inspected and, if the sac is bulging or dark blue, or if there are no signs of underlying cardiac activity, it should be immediately opened. Care should be taken during the opening of the pericardial sac to avoid injury to the coronary arteries. Once the pericardial sac has been opened, clots should be removed, bleeding should be controlled by digital pressure, and. if there is no cardiac activity, internal defibrillation and cardiac massage should be
648
Anesthesia for Emergency Thoracic Surgery
instituted. Finally, a left thoracotomy provides good access for descending thoracic aortic occlusion to provide proximal control of massive abdominal arterial injuries. A right lateral thoracotomy will be performed only for right-sided lesions (i.e., suspected right hilar damage). C. Anesthetic Considerations The patient must be intubated immediately and ventilated with 100 per cent oxygen. If a doublelumen tube is not available or cannot be easily inserted by the available personnel or if a singlelumen tube is in situ, selective ventilation of the right lung may be reliably achieved by advancing the single-lumen tube in 1-cm steps until left lung breath sounds disappear (mean ± standard deviation of length of single-lumen tube required to achieve a right main-stem bronchial intubation is 30 ± 1.1 cm).213 Unfortunately, the right upper lobe may be expected to be obstructed nearly 100 per cent of the time by this technique. Multiple large-bore intravenous catheters must be inserted, and intravascular volume repletion should be begun immediately. As soon as a peripheral pulse can be palpated, an arterial line should be inserted for pressure measurement and blood sampling. As intravascular volume is being repleted, a central venous catheter should be inserted to measure central venous pressure. Paralysis should be instituted if the patient is moving at all. No, or extremely small doses of, intravenous anesthetics should be administered. A Foley catheter should be inserted as soon as possible. The patient's blood should be typed and cross matched for appropriate number of units of blood. Use of the autotransfuser should be considered. IX. REMOVAL OF TRACHEOBRONCHIAL TREE FOREIGN BODIES A. General Considerations The majority of foreign bodies are inhaled by children younger than 3 years.214 The lack of molars at an early age enhances potential aspiration (e.g., molars grind food into small particles that are easily swallowed with the saliva stream).214 In addition, adults who have an acutely (alcoholism) or chronically (dementia) depressed sensorium also are at risk for foreign body inhalation. Acute total upper airway obstruction is the immediate hazard, but far more commonly the foreign body is inhaled deeper into the tracheobronchial tree, where it impacts and initiates a local inflammatory
reaction. The subsequent local inflammatory reaction holds the foreign body more securely in place (perhaps preventing removal).215 Without removal, the foreign body ultimately causes distal collapse and infection. Organic material is more dangerous than inorganic material because it swells after inhalation and may fragment, even before attempts are made to remove it. Peanuts are particularly liable to swell and fragment, and they liberate an irritant oil that causes severe local inflammation. Sweets are also a special problem because they dissolve in the tracheobronchial secretions, forming a viscous hypertonic solution that predisposes to obstruction and collapse. Sharp foreign bodies, such as bones, needles, pins, and glass may also be difficult to remove if they stick into the wall of the tracheobronchial tree. In one large series of inhaled foreign bodies, 66 per cent were peanuts, 16 per cent were vegetable derivatives, and 17 per cent were inert objects.214 The majority of foreign bodies lodge in the right lung because of the disposition of the right main-stem bronchus relative to the trachea, but approximately 20 to 44 per cent can be found on the left side.214-216 B. Surgical Considerations The history remains the most reliable indicator of aspiration of a foreign body. Paroxysmal cough and wheeze are the common symptoms in children, but dyspnea, stridor, fever, and vomiting also occur. Superglottic foreign bodies often cause gurgling, drooling, and inspiratory stridor; tracheal foreign bodies cause both inspiratory and expiratory stridor; and subcarinal foreign bodies cause wheezing. On auscultation, a very high percentage of patients with subcarinal foreign bodies will have decreased breaths sounds and a wheeze in the affected area217 and often the wheeze is generalized.218 In fact, administration of theophylline, epinephrine, and steroids will decrease wheezing in the uninvolved lung but not in the involved lung; this differential lung response has been used as a diagnostic test in cases in which the differential diagnosis is between bronchial foreign body and asthma.218 Table 17-17 shows the differential diagnosis (based on history and physical examination) of the typical presentations of the major causes of airway obstruction in children. A chest infection that fails to clear in spite of antibiotic treatment in an otherwise healthy child warrants chest radiography and diagnostic bronchoscopy to exclude the presence of an inhaled foreign body. This is particularly true in patients with recurrent one-sided infection and/or persistent unilateral wheezing
Anesthesia for Emergency Thoracic Surgery
without prior evidence of underlying allergic diathesis. Unfortunately, a specific history of inhalation is unavailable in approximately 20 per cent of all patients with this problem; these patients consist of some children, adults after alcoholic excess, adults with dementia, and patients under general anesthesia having dental surgery without endotracheal intubation.216 In approximately 70 to 80 per cent of the patients, the initial chest roentgenogram will be positive and show the foreign body, distal atelectasis, air trapping (foreign body acts as a one-way ball valve in a bronchus), or mediastinal shift.214-216 In those with initially normal chest radiographs, approximately half will develop very suggestive "positive" X-ray findings (air trapping, atelectasis, infiltrates, the foreign body itself) before the removal of the foreign body. However, the foreign body will not be directly visible most of the time. Thus, approximately one eighth of patients will come to bronchoscopy on the basis of the history and physical examination alone and have a normal chest roentgenogram. If fluoroscopy is added to routine radiographic techniques, a much higher yield of positive findings on initial radiologic examination is obtained.216 Positive fluoroscopic findings include a mediastinal shift away from the affected side consistent with air trapping on expiration (i.e., the foreign body acts as a one-way ball valve).214 In a small percentage of patients, the foreign body may change location between the time of admission to the hospital and bronchoscopy.215 Because foreign bodies may wander, it is essential to obtain a chest X-ray just before bronchoscopy.215
649
In summary, a history compatible with foreign body aspiration dictates diagnostic and therapeutic endoscopy with or without radiologic confirmation. To decrease the approximately one fourth failure rate of plain films in the first 24 hours, fluoroscopy must be strongly considered as an initial diagnostic technique in foreign body evaluation. In many instances, a 24-hour interval, a safety zone, can be observed prior to endoscopy. The safety zone ensures adequate gastric emptying and unhurried and thorough preparation for the procedure. During this period of time, postural drainage, chest percussion and vibration, and administration of bronchial dilators (including corticosteroids) may be used; in an occasional patient these measures may result in a spontaneous extrusion of the foreign body. However, these maneuvers should not be regarded as a substitution for endoscopic removal219 and, on rare occasions, have caused increased airway obstruction and cardiac arrest.22a-222 If a foreign body is not readily removed at the first endoscopic procedure, chest physical therapy and bronchodilatation should be resumed. Vegetable foreign bodies, including nuts, show potential for self-extrusion, particularly postbronchoscopy.215 A spontaneously extruded foreign body always has the potential for causing a major (proximal) obstruction of the tracheobronchial tree, causing hypoxia.2'5 In an effort to avoid thoracotomy, repeated endoscopy should be used when clinically indicated. However, bronchotomy, or even lobectomy, may be necessary if the foreign body becomes deeply embedded in a mass of in-
650
Anesthesia for Emergency Thoracic Surgery
flammatory tissue or erodes through the wall of a bronchus.215 As discussed under Laser Resection of Major Airway Obstructing Tumors (chapter 15), a rigid bronchoscope was considered preferable to a fiberoptic bronchoscope because it allowed better control of secretions and blood, permitted better removal of larger pieces of necrotic material, and provided a better view. Similarly, foreign bodies to the tracheobronchial tree should be removed with a rigid bronchoscope because it allows passage of much larger instruments and foreign bodies, and the open-ended rigid instrument also makes ventilation easier in small children.215·223,224 In fact, it may be impossible for a small patient to breathe around a fiberoptic bronchoscope because even the narrowest fiberoptic bronchoscope (1.8mm outside diameter) takes up most of the area of the trachea. Furthermore, and of special relevance to small patients, rigid bronchoscopes range in nominal size from 2.5 to nearly 10 mm. The nominal size of a rigid bronchoscope refers to the smallest internal diameter through which instruments may be passed, not to the outer diameter, which may be several millimeters greater. For example, the 3.5-mm flexible pediatric bronchoscope will pass through a 3.5-mm Storz rigid bronchoscope. A large foreign body cannot be removed through a small bronchoscope and must, therefore, be grasped with forceps and extracted while the bronchoscope is withdrawn simultaneously. Repeated instrumentation may be necessary, and even in skilled hands this will traumatize the vocal cords and upper respiratory tract. An alternative solution that has been reported is to pass a Fogarty embolectomy catheter beyond the foreign body; the balloon is inflated and the catheter is withdrawn so that the foreign body is trapped against the tip of the bronchoscope and the entire assembly is removed together.217,225 If thoracotomy for bronchotomy is required, the usual method of localizing the foreign body is by palpating the object through the firm bronchial wall. However, the palpation may not be positive or precise. During thoracotomy, a fiberoptic bronchoscope can be a useful guide for the rapid and precise determination of the site of bronchotomy.223 The fiberoptic bronchoscope finds the object and shines its light on it.223 Dimming the lights of the operating room is helpful in order to better visualize the light at the end of the bronchoscope. A small well-placed bronchotomy will predictably result in lower morbidity. This technique should be considered for all patients undergoing open removal of a foreign body or bronchial tumor in whom the precise placement of the bronchotomy is in question.
C. Anesthetic Considerations (Table 17-18) If a child has nondistressed respiration and crying does not cause an embarrassment of respiration, the child can be heavily premedicated intramuscularly and brought to the operating room for a smooth inhalation induction. If the child has respiratory distress and/or crying causes wheezing and/or cyanosis, premedication should be withheld. In this type of patient, anesthesia should be induced with intramuscular ketamine in the operating room. Although an intramuscular injection may cause crying, it is only for a short period of time, especially if the child is held and comforted. Since bronchoscopy involves intense airway stimulation, atropine or glycopyrrolate should be included in the intramuscular injection to decrease vagal reactions and secretions. As soon as the child appears sedated, the induction of anesthesia can be continued by inhalation, and intravenous access can be established as quickly as possible. When adequately anesthetized, the patient undergoes a laryngoscopy, and the larynx and tracheobronchial tree are sprayed with 4 per cent lidocaine. If the response to laryngoscopy and spraying is minimal, the airway may be released to the endoscopist. The eyes are taped shut, and soft pads are placed over the globes. The patient is then positioned for endoscopy. Although succinylcholine is kept available, we prefer spontaneous respiration until at least the nature and location of the foreign body are known; avoidance of relaxants requires the anesthesiologist to achieve adequate levels of anesthesia so that the insertion of the bronchoscope does not cause laryngospasm, bronchospasm, and chest wall rigidity as a result of energetic active exhalation. Spontaneous ventilation may also be protective of not forcing the
Table 17-18
IMPORTANT ANESTHETIC CONSIDERATIONS WITH TRACHEOBRONCHIAL TREE FOREIGN BODIES
1. Minimize preanesthetic crying-induced respiratory distress 2. Administer atropine prior to passage of bronchoscope 3. Provide depth of anesthesia to prevent reaction to bronchoscope 4. Administer lidocaine to decrease reaction to bronchoscope 5. Try to maintain spontaneous ventilation 6. Paralyze if excessive reaction to bronchoscope (straining, bucking, laryngospasm, bronchospasm) 7. Beware of mechanical problems when the foreign body is being removed (lodge in trachea, fragment) 8. Consider endotracheal intubation and/or administration of steroids postoperatively if bronchoscope was passed several times 9. Some cooperative adults can undergo bronchoscopy under local anesthesia
Anesthesia for Emergency Thoracic Surgery
foreign body deeper into the bronchial tree.217 Intravenous lidocaine may diminish any coughing or straining reaction to the insertion of the rigid bronchoscope. However, the inability to quickly reverse any of these aforementioned problems at any time during the case may require the use of relaxants (small succinylcholine bolus, succinylcholine drip, atracurium, vecuronium). Because these cases may end suddenly, the use of nondepolarizing relaxants may be problematic. Controlled ventilation without relaxants is an acceptable middle ground. On occasion, major airway obstruction may be relieved by adding helium to the inspired gas.226 During the actual bronchoscopy, one overhead spotlight is directed onto the anesthesia machine, and another is directed onto the patient's feet to assess perfusion and color without interfering with the surgeon's vision. Following passage of the rigid bronchoscope and during removal of the foreign body, the anesthesiologist should be aware of several important mechanical problems related to the physical characteristics of the foreign body (softness, smoothness, size).227 First, beans and nuts can fragment and obstruct both main-stem bronchi. If this is thought to be a possibility, removal of the foreign body with the patient in the lateral decubitus position (with the foreign body in a dependent position) should be considered; if this is not practical, a thoracic surgeon and thoracotomy instruments need to be available. Second, plastic and smooth inert objects (beads) may be hard to grasp, and during removal they can slip out of the grasp of the forceps and block larger and more proximal airways. If a large airway does become occluded, the object should be pushed back to its previous location to re-establish ventilation and reassess the situation. Emphasis is necessary on previous location because inflammation in new areas of the lung may occur quickly following relocation. Third, if vocal cord motion during extraction causes significant obstruction, succinylcholine should be administered. Fourth, the foreign body may sometimes be larger than the bronchoscope, and the bronchoscope and the foreign body may need to be removed at the same time. Fifth, infection may already be well established, and vigilance in suctioning infected secretions is important; rarely, an abscess may be present distal to the foreign body. In all of these situations, if the bronchoscope is withdrawn and the foreign body is not seen, the pharynx should be quickly inspected. If the object is not visible, the bronchoscope should be quickly reintroduced to make certain that the object is not occluding the trachea because the narrowing of the airway at the glottis makes this a most probable location. The advantage of spontaneous respiration becomes manifest during this situation because tra-
651
cheal obstruction is instantly obvious (tracheal tug, intercostal retractions). If the trachea is occluded, the object is quickly pushed back to its previous location to re-establish ventilation, and the situation is reassessed. If positive-pressure ventilation is used, it should be possible to maintain adequate gas exchange without excessive inflation pressures, even when a small bronchoscope is in use. However, extreme care must be taken to avoid overinflation of the lungs when either an instrument or the foreign body is blocking the bronchoscope, and ventilation may need to be interrupted from time to time to prevent the foreign body or its fragments from being blown deeper into the respiratory tract when they are loose within the bronchial lumen. It should be remembered that some cooperative adult patients can have the foreign body removed under local anesthesia.215 The patient should be supervised closely until fully conscious. Laryngeal stridor is relatively common, particularly after a difficult and prolonged procedure (with many passages of the bronchoscope through the vocal cords) in a small child. Consequently, use of steroids and bronchodilators in the postoperative period is common. In addition, after a traumatic endoscopy, nasotracheal intubation is indicated with a tube 0.5 mm smaller than indicated by age; this is left in place until chest X-ray and "air leak" both confirm that swelling is decreased. Chest physiotherapy should be used in patients who have postoperative atelectasis and in patients with pneumonia who require antibiotics. The patient is then extubated and may be placed in a croup tent for up to 24 hours, if necessary. Racemic epinephrine may also be useful after extubation.228 Patients who are extubated in the operating room should breathe humidified oxygen by face mask or croup tent postoperatively.
REFERENCES 1. Garzon AA, Gourin A: Surgical management of massive hemoptysis. Ann Surg 187:267-271, 1978. 2. Yeoh CB, Hubaytar RT, Ford JM, et al: Treatment of massive hemorrhage in pulmonary tuberculosis. J Thorac Cardiovasc Surg 54:503-510, 1967. 3. Stern RC, Wood RE, Boat TF, et al: Treatment and prognosis of massive hemoptysis in cystic fibrosis. Am Rev Respir Dis 117:825-828, 1978. 4. Ehrenhaft JL, Taber RE: Management of massive hemoptysis not due to pulmonary tuberculosis or neoplasm. J Thorac Cardiovasc Surg 30:275-287, 1955. 5. Crocco JA, Rooney JJ, Fankushen DS, et al: Massive hemoptysis. Arch Intern Med 121:495-^98, 1968. 6. Gourin A, Garzon AA: Operative treatment of massive hemoptysis. Ann Thorac Surg 18:52-60, 1974. 7. Conlan AA, Hurwitz SS: Management of massive haemoptysis with the rigid bronchoscope and cold saline lavage. Thorax 35:901-904, 1980.
CHAPTER
18
Anesthesia for Pediatric Thoracic Surgery I. Introduction II. Special Problems Related to Premature and Newborn Infants A. Persistence of the Fetal Circulation B. Respiratory Distress Syndrome C. Retinopathy of Prematurity D. Periodic Breathing and Apnea E. Thermoregulation F. Vitamin, Caloric, Electrolyte, and Fluid Requirements G. Airway Anatomy III. Congenital Diaphragmatic Hernia A. General Considerations B. Surgical Considerations C. Anesthetic Considerations IV. Esophageal Atresia and Tracheoesophageal Fistula A. General Considerations B. Surgical Considerations C. Anesthetic Considerations V. Ligation of Patent Ductus Arteriosus in Premature Infants A. General Considerations B. Surgical Considerations C. Anesthetic Considerations VI. Vascular Rings A. General Considerations B. Surgical Considerations C. Anesthetic Considerations VII. Congenital Parenchymal Lesions: Lobar Emphysema, Cysts, Sequestrations, and Cystic Adenomatoid Malformations A. General Considerations 1. Congenital Lobar Emphysema
VIII.
IX.
X. XI.
2. Congenital Bronchogenic Cysts 3. Pulmonary Sequestration 4. Cystic Adenomatoid Malformations B. Surgical Considerations 1. Congenital Lobar Emphysema 2. Congenital Bronchogenic Cysts 3. Pulmonary Sequestration 4. Cystic Adenomatoid Malformations C. Anesthetic Considerations 1. Congenital Lobar Emphysema 2. Congenital Bronchogenic Cysts 3. Pulmonary Sequestration 4. Cystic Adenomatoid Malformations Thoracic Surgical Procedures Requiring One-Lung Ventilation A. Bronchial Blocker B. Main-Stem Bronchial Intubation C. Bronchial Blockade Plus MainStem Bronchial Intubation D. Prone Positioning Diagnostic Bronchoscopy A. General Considerations B. Surgical Considerations C. Anesthetic Considerations Bronchography Asphyxiating Thoracic Dystrophy (Jeune's Syndrome) A. General and Surgical Considerations B. Anesthetic Considerations
The Fetal Circulation*
Anesthesia for Pediatric Thoracic Surgery
I. INTRODUCTION This chapter has two major sections. The first section highlights the important physiologic and anatomic problems of premature infants and neonates that have special relevance to anesthesia for thoracic surgery (as opposed to a broader discussion of all anesthesia problems for all pediatric surgery). These problems consist of persistence of the fetal circulation; the respiratory distress syndrome; retrolental fibroplasia; periodic breathing and apnea; impaired thermoregulation; precarious fluid, electrolyte, and caloric balance; and altered airway anatomy. The second section discusses the major pediatric thoracic surgical cases, which consist of congenital diaphragmatic hernia, tracheoesophageal fistula, vascular rings, congenital lobar emphysema, conditions necessitating bronchoscopy, and various pulmonary parenchymal problems requiring lung separation. Anesthesia for resection of mediastinal tumors, thymectomy for myasthenia gravis, and removal of foreign bodies are procedures that also frequently involve pediatric patients but are discussed in chapters 15 and 17.
II. SPECIAL PROBLEMS RELATED TO PREMATURE AND NEWBORN INFANTS The unique anesthetic problems with the pediatric patient occur primarily in the neonatal period and with the premature infant. These patients may have persistence of the fetal circulation, the respiratory distress syndrome, greatly increased sensitivity to increased arterial oxygen tension, periodic breathing and apneic spells, difficulty in regulating temperature and the internal chemical and caloric status, and an airway anatomy different from that of the adult.
A. Persistence of the Fetal Circulation In the fetal circulation well-oxygenated blood flows from the umbilical vein through the fetal liver and the ductus venosus to the inferior vena
659
cava. The mixing of well-oxygenated umbilical vein blood with poorly oxygenated inferior vena cava blood results in an inferior vena cava oxygen tension of about 35 to 40 mm Hg (Fig. 18-1). Inferior vena cava blood then flows into the right atrium. Since fetal right atrial pressure is greater than left atrial pressure, the "trap door" flap of the foramen ovale in the atrial septum remains constantly open prior to birth; thus, the foramen ovale directs most of the relatively oxygen-rich inferior vena cava blood across to the left atrium. The relatively well-oxygenated blood that enters the left atrium from the right atrium goes into the left ventricle, is then ejected into the aorta, and perfuses the coronary arteries, the aortic arch, and the brain. Reversal of the right-left atrial pressure gradient after birth (caused by a marked decrease in pulmonary vascular resistance after initiation of breathing; Figure 18-2) allows functional closure of the foramen ovale, even though it may not anatomically close for many years. Venous blood from the superior vena cava with a relatively low oxygen tension enters the right atrium, mixes minimally with blood from the inferior vena cava, and crosses the tricuspid valve into the right ventricle. Right ventricular ejection of this low-oxygen-tension blood into the pulmonary circulation causes pulmonary vasoconstriction and elevation of pulmonary vascular resistance. The blood in the pulmonary artery subsequently follows the path of least resistance and passes from the pulmonary artery into the aorta via the ductus arteriosus to perfuse the lower portion of the systemic vascular bed. A small amount of pulmonary circulation blood (10 per cent) mixes with bronchial artery blood to drain directly into the left side of the heart. Persistence of the fetal circulation means that right-to-left shunting occurs through a patent ductus arteriosus and/or patent foramen ovale. Anything that increases pulmonary vascular resistance (hypoxia, hypercarbia, acidosis, stress, vasoactive drugs) and/or decreases systemic vascular resistance (inhalation anesthetics, vasodilators) will increase shunting through the ductus arteriosus.1 Hypoxemia, in particular, may also cause ductal dilatation. In addition, anything that increases pulmonary vascular resistance will increase right ventricular and right atrial pressures and increase
Figure 18-1 This schematic diagram shows the main central blood flow pathways in the fetal circulation. Oxygen-rich blood is delivered from the mother to the fetus via the umbilical vein. The oxygen-rich blood in the umbilical vein mixes with the oxygenpoor blood draining the lower fetal body in the inferior vena cava and increases the inferior vena cava oxygen tension up to 35 to 40 mm Hg. This now relatively oxygen-rich inferior vena cava blood flow streamlines through the right atrium into the foramen ovale into the left side of the circulation, where it perfuses the coronary arteries, brain, and upper body. These areas are drained by the superior vena cava, which contains oxygen-poor blood. Superior vena cava blood flow streamlines through the right atrium into the right ventricle, which then ejects the oxygen-poor blood into the pulmonary artery. The oxygen-poor blood causes pulmonary vasoconstriction, which directs the blood flow through the ductus arteriosus to the descending aorta to perfuse the lower body as well as to return to the mother via the umbilical artery. The dashed arrows represent minor blood flow pathways.
660
Anesthesia for Pediatric Thoracic Surgery
Figure 18-2 Schematic diagram of the phases of the transitional circulation and their relative time course. Note that the time line is discontinuous. The period before birth represents minutes to hours. In utero, pulmonary vascular resistance (PVR) slowly decreases in the days to weeks before birth and thus is not represented in this diagram. The immediate phase is due to the initiation of breathing (mechanical lung expansion and introduction of oxygen into the alveolar space), the fast phase may be related to prostacyclin production (as well as perhaps bradykinin and angiotensin as well as decreasing levels of leukotrienes),' and the final phase may be the result of remodeling of the pulmonary circulation. (From Clarke WR: The transitional circulation: Physiology and anesthetic implications. J Clin Anesth 2:196, 1990. Used with permission.)
right-to-left shunting through the patent foramen ovale. The shunting through the ductus arteriosus and foramen ovale results in a vicious cycle of increased hypoxemia, pulmonary vasoconstriction, pulmonary vascular resistance, shunting, and hy poxemia. Patients with a persistent fetal circulation who often require surgery and anesthesia are those undergoing ligation of the ductus arteriosus, repair of diaphragmatic hernia, and resection of cystic 2 malformations. It should be remembered that the foramen ovale is a potential site of a paradoxic air embolus should air enter the venous circulation, and for this reason it is essential to exclude all air from intravenous fluid administration sets. The ductus arteriosus enters the aorta just distal to the left subclavian artery; however, there is enough mixing of ductal and left subclavian blood flow to make the left arm arterial oxygen tension significantly less than the preductal arterial oxygen tension. Consequently, preductal arterial oxygen tension must be obtained from the right radial ar tery and postductal arterial oxygen tension can be obtained from the umbilical artery. Simultaneous analysis of the pre- and postductal arterial oxygen tensions indicates the degree of ductal shunting. When the preductal arterial oxygen tension is at least 15 mm Hg higher than the postductal value,
significant ductal shunting is present.3 Indeed, a right-to-left shunt of 20 per cent is usually consid ered a "normal" amount of ductal shunting.4 When the arterial oxygen tension in the preductal samples is below the value predicted for a 20 per cent shunt (see Fig. 3-33), preductal shunting (through the lungs and/or patent foramen ovale) is present. If the amount of preductal shunting is severe, it becomes impossible to detect ductal shunting. Successful treatment of pulmonary vasocon striction and ductal and foramen ovale shunting involves avoidance of the events that can increase pulmonary vascular resistance (acidosis and hy poxemia: Fig. 18-3) 5 and/or decrease systemic vascular resistance. Metabolic acidosis is treated with sodium bicarbonate infusions. Endotracheal tube suctioning is minimized to avoid even tran sient changes in the alveolar and arterial oxygen partial pressure. The most consistently successful therapeutic modality is the alkalosis achieved with hyperventilation (if it can be achieved; VC/VŢ may be so high that the P a C0 2 cannot be lowered less than 40 to 50 mm Hg). 6 7 Frequently, a threshold effect is observed such that the arterial oxygen tension does not rise until arterial pH reaches 7.55 to 7.60.7 Generally, ventilatory rates of 60 to 120
Anesthesia for Pediatric Thoracic Surgery
Figure 18-3 Change in pulmonary vascular resistance (PVR) with oxygenation and hydrogen ion concentration. There is a sharp increase in PVR when P a 0 2 falls below 6.7 kPa (50 mm Hg). Note also the synergistic effect when hypoxemia is accompanied by acidosis. (Reproduced by permission of Rudolph AM, Yuan S: Response of the pulmonary vasculature to hypoxia and H+ ion concentration changes. J Clin Invest 45:399-411, 1966.)
breaths/min are necessary to achieve this level of hyperventilation. Table 18-1 shows the methodology and common pitfalls in mechanically ventilating patients with persistent fetal circulation.2 To achieve the goals in Table 18-1 (high respiratory rate and airway pressures), the anesthesiologist must manually ventilate the patient in the operating room because few ventilators compatible with anesthesia machines are capable of producing the required rates. Once the proper level of ventilation is determined ( S P 0 2 / P E T C 0 2 ) , it must be maintained. The temp-
661
tation is to back off a bit from high pressures and rates to avoid damaging the lung. This temptation must be resisted because such attempts can cause marked lability of the pulmonary vasculature. Small increases in P a C0 2 or decreases in P a O ; or pH can lead to sudden increases in pulmonary vascular resistance, resulting in renewed right-toleft shunting, with worsening of the arterial blood gases and a rapid downward spiral. Mechanical ventilation of the patient is considerably aided by paralysis. A number of pharmacologic pulmonary vasodilators have been tried, including morphine, prednisolone, chlorpromazine, tolazoline, bradykinin. and acetylcholine.3·8 No agent has consistently improved pulmonary hypertension, although tolazoline has been most frequently used. If systemic hypotension occurs, volume repletion (may require as much as 50-75 ml/kg for a 2-hour diaphragmatic hernia repair) and/or dopamine infusion is indicated because the reduction in systemic pressure increases ductal shunting. The systemic blood pressure is probably the best indicator for volume replacement because the central venous pressure may not be very useful in the presence of very high intrathoracic pressure. Prostaglandin E, (PGE,), which is a pulmonary and ductal dilator, has been used to decrease pulmonary vascular resistance in patients with persistent fetal circulation (the ductus is thought to be already fully patent in these patients and cannot dilate any more) and to dilate the ductus arteriosus in patients with ductal blood flow-dependent congenital heart defects (e.g., pulmonary atresia).9 Figure 18-4 is a flow chart for a suggested stepwise administration of therapy for patients with persistent fetal circulation.2 On the basis of these considerations, a reasonable anesthetic for these patients would consist of mainly narcotics, paralysis, low isoflurane concentration to control
Table 18-1 GUIDELINES FOR OPERATION OF MECHANICAL VENTILATOR IN PERSISTENT PULMONARY HYPERTENSION OF THE NEONATES* Factor
Goal
Inspiratory pressure
Level necessary to decrease Pco 2
Ventilator rate
Rapid rates (up to 150 breaths per minute) to decrease Pco2, with lowest inspiratory pressure Low values unless pulmonary parenchymal disease exists P a 0 2 lOOmm Hg Short inspiratory time to prevent air trapping at rapid rates
Positive end-expiratory pressure Inspired 0 : concentration Inspiratory-expiratory ratio
Common Error Inadequate inspiratory pressure (no movement of chest wall or increasing Pco 2 ); reading on hand manometer may differ from that on ventilator Use of ventilator with maximum rate below level needed to decrease Pco2; may require hand ventilation to achieve possible high rates Failure to try different levels that improve PaCO; as patient condition changes Weaning too rapidly Inadequate flow rates to achieve short inspiratory time
* Adapted from Clarke WR: The transitional circulation: Physiology and anesthetic implications. J Clin Anesth 2:192-211, 1990. Used with permission.
662
Anesthesia for Pediatric Thoracic Surgery
Figure 18—4 Suggested sequential implementation of therapy for infants with known or suspected persistent pulmonary hypertension of the newborn. Details are given in the text. (E.C.M.O. = extracorporeal membrane oxygenation.) (From Clarke WR: The transitional circulation: Physiology and anesthetic implications. J Clin Anesth 2:192, 1990. Used with permission.)
response to surgical stimulation, F,02 of 1.0, no nitrous oxide, and liberal fluid administration. The role of patent ductus arteriosus ligation in the infant with severe persistent fetal circulation is unclear. The patent ductus arteriosus, although a major source of right-to-left shunting and, therefore, of systemic hypoxemia and acidosis, is also a relief valve for the pressure-overloaded right ventricle. Ligation of the patent ductus arteriosus might cause acute right ventricular failure (and subsequently left ventricular failure because of interventricular septal shifting; see Fig. 18—5)10 and, therefore, should not be performed on a routine
basis.8 Ligation of the ductus arteriosus is most commonly performed under local anesthesia, moderate levels of narcosis, and neuromuscular blockade in the neonatal intensive care unit, which eliminates the precarious transportation of these very ill and tiny infants to the operating room. B. Respiratory Distress Syndrome The respiratory distress syndrome is a disease of atelectasis in premature infants caused by 2 failure of the type II alveolar cells to secrete 2
Anesthesia for Pediatric Thoracic Surgery
663
Figure 18-5 Cardiovascular ab normalities associated with the de velopment of pulmonary hyperten sion culminating in decreased CBF and CO. (RVEDP = right ventricu lar end-diastolic pressure; RVEDV = right ventricular end-diastolic vol ume; CBF = coronary blood flow; CO = cardiac output; LVEDV left ventricular end-diastolic volume; LVEDP = left ventricular end-dia stolic pressure; RV = right ventri cle.) (From Burrows FA, Klinck JR. Rabinovitch M, Bohn DJ: Pulmonary hypertension in children: Periopera tive management. Can Anaesth Soc J 33:606, 1986. Used with permis sion.)
normal amount of surfactant. Surfactant acts to reduce surface tension in the alveolus, allowing for expansion during inhalation and preventing collapse during exhalation (see Fig. 3-19). When surfactant is absent or reduced in quantity, the respiratory distress syndrome occurs with atelec tasis and ventilation-perfusion abnormalities. The presence of respiratory distress syndrome may cause increased pulmonary vascular resistance. In creased pulmonary vascular resistance may cause persistence of the fetal circulation. The more severe the respiratory distress syn drome, the higher the pulmonary vascular resis tance is in relation to the systemic vascular resis tance. In severe forms of the respiratory distress syndrome, pulmonary resistance is higher than systemic resistance, and right-to-left shunting oc curs through the ductus arteriosus, resulting in a lower arterial oxygen tension below the level of the ductus (Fig. 18—6). As the respiratory distress syndrome improves, pulmonary resistance falls be low systemic levels, the shunt is reversed, and congestive heart failure may ensue and warrant closure of the ductus either pharmacologically or surgically." -13 The primary aim of the treatment of the respi ratory distress syndrome is to open atelectatic al veoli with positive-pressure ventilation and then to keep the alveoli open using positive end-expira tory pressure. Casual removal of the positive endexpiratory pressure during transport to and in the operating room may result in profound alveolar collapse and the development of marked hypoxia, acidosis, and hypercarbia. Consequently, the man agement of ventilation during anesthesia of a child
with respiratory distress syndrome is critical; it is important to continue to provide the same ventila tory support during the operative period as was provided in the preoperative period. The important mechanical ventilation concerns are listed in Table 18-1. In many instances, it may be simplest to bring the infant's own ventilator into the operating room to ventilate the patient during surgery and to use intravenous anesthesic agents and muscle re laxants. Surfactant replacement is effective in prevention and amelioration of the respiratory distress syn drome. The essential differences between the ear lier unsuccessful studies and more recent studies that have met with unequivocal success (84 per cent of patients) relate to the means of administra tion and the nature of the surfactants delivered.14 The best results (increased P a 0 2 , lung compliance during spontaneous ventilation,15 and radiographic clearing) are achieved when 60 to 120 mg of sur factants derived from bovine lungs or human am niotic fluid are introduced into the lung through an endotracheal tube as a bolus of 3 to 5 ml. 14 Compared with matched controls, respiratory distress patients treated with surfactant have a lower incidence of death (16 per cent vs. 52 per cent, ρ < .001), bronchopulmonary dysplasia (16 per cent vs. 31 per cent, ρ < .001), and acute complications (p < .001).16 Good results have been evident when surfactants were instilled up to 12 hours afterbirth.' 4
C. Retinopathy of Prematurity Retinal damage due to inhaling an increased partial pressure of oxygen is unique to the prema-
Figure 18-6 The lack of surfactant in the respiratory distress syndrome (RDS) causes atelectasis and low ventilation-perfusion regions. The atelectatic and low ventilation-perfusion regions cause an increase in pulmonary vascular resistance (PVR), which directs pulmonary artery blood flow through the ductus arteriosus, thereby increasing right-to-left (R —> L) shunting. The ductus arteriosus enters the descending thoracic aorta just distal to the left subclavian artery. Blood flow to the retina is preductal, and oxygen tension in preductal blood flow must be measured by a right-arm arterial blood sample. Postductal oxygen tension can be measured from an umbilical artery blood sample.
ture newborn. The premature infant's retinal arteries react to increased partial pressure of oxygen with vasoconstriction, which induces injury, neovascularization, and retrolental fibroplasia that causes retinal detachment and destruction.17·18 Premature infants are at risk and remain at risk for the retinopathy of prematurity until their retinal vasculature completes its growth at approximately 35 to 40 weeks of age after conception. If gestational age cannot be determined accurately, a birth weight of less than 1500 g seems to be the best indicator of risk for the retinopathy of prematurity.19 Although there is no correlation between Pa02 levels and blindness risk,20·21 there is an association between the incidence and severity of retinopathy of prematurity and the duration of exposure to Ptc02 levels of 80 mm Hg or higher.22 However, in recent years, it has become clear that factors other than high inspired oxygen concentration may be involved in the pathogenesis of the retinopathy of prematurity.18·20 For example,
the retinopathy of prematurity occurs in some infants who have never been treated with supplemental oxygen, the disease does not occur in all premature infants treated with high oxygen concentration, and the incidence of prematurity appears to be rising at a time when there has been pronounced improvement in the technical methods of oxygen delivery and monitoring.18·22 In fact, it does not appear possible to prevent the disease with current (as of 1992) methods of monitoring.18· 22 It may be that the apparent increase in incidence may be due to increased survival of very small infants (less than 1 kg). Some other factors may be the absolute birth weight (risk is strongly inversely proportional to birth weight, indicating that this is a disease primarily caused by prematurity, with high inspired oxygen concentrations a secondary factor), maternal diabetes, frequent apneic spells, bronchopulmonary dysplasia, degree of illness, need for blood transfusions, and sepsis.23-25
Anesthesia for Pediatric Thoracic Surgery
The anesthesiologist's responsibility is to maintain the arterial and/or transcutaneous oxygen tension between 60 and 80 mm Hg.26 In patients without a ductus arteriosus, the arterial and transcutaneous oxygen tension can be measured anywhere. In patients with a ductus arteriosus, the retinal arteries are supplied by pre-ductus arteriosus blood flow (Fig. 18-6). Consequently, in patients with a ductus arteriosus, the arterial oxygen tension must be measured from the right radial artery and the transcutaneous oxygen tension must be measured from the right arm. The arterial and/or transcutaneous oxygen tension can be controlled by using nitrous oxide as a diluting gas or by using an air-oxygen blender to control the inspired oxygen concentration. Careful control of the inspired oxygen concentration and retinal artery oxygen tension is particularly important and difficult during thoracic surgery in which the patient may be either hypoxic or hyperoxic and may vary between these two states quickly, depending on whether or not the operative lung is compressed (retracted) or ventilated.
D. Periodic Breathing and Apnea Premature infants often have short periods of apnea (10 sec) that are not associated with hypoxemia or bradycardia. Long periods of apnea (30 sec) that are associated with hypoxemia and bradycardia in the premature infant are distinctly abnormal and can be lethal (sudden infant death syndrome [SIDS]).27 Indeed, irregular and periodic breathing is commonly seen in infants who have survived a "sudden infant death" episode (aborted SIDS), suggesting that immaturity of medullary and/or peripheral control of breathing (loss of responsiveness to hypoxemia and hypercarbia) is an important cause of this syndrome.28 The risk of SIDS may be increased by in situ exposure to smoke.29 Theophylline and caffeine have been shown to increase the central chemoreceptor ventilatory response and decrease the number of apneic spells in premature infants.30 In preparation for extubation after surgery, the breathing pattern of premature infants should be carefully observed for significant periodic breathing, the presence of which is a contraindication for the removal of the endotracheal tube. Immediately following extubation, an oral airway, if tolerated, may be especially helpful since many infants (60 per cent) may remain obligate nose breathers for as long as 2 months of age and may not open their mouths to breathe even if the nares are totally obstructed for any reason (e.g., secretions or edema).31
665
E. Thermoregulation The relative surface area of the newborn is one ninth that of an adult, whereas the weight of a newborn is approximately 20 times less than that of an adult. Thus, the ratio of surface area to weight for infants is approximately two times greater than for adults. Consequently, an infant loses heat much more rapidly than an adult. Hypothermia in premature and newborn infants is a dangerous condition; there is 90 per cent mortality in newborns whose temperatures are allowed to decrease below 32°C.32 Figure 18-7 shows the symptoms and signs of progressive hypothermia (Wilson WC, personal communication).33 Obviously, hypothermia can cause a severe metabolic acidosis and cardiovascular depression below 32°C. Oxygen consumption has been shown to double in the immediate postoperative period if the child is allowed to emerge from anesthesia even in a mildly to moderate hypothermic state.34 Hypothermia below 35°C is a relative contraindication to extubation. Hypothermia in the newborn and small child undergoing thoracic surgery is particularly likely to occur because there is a large surface area exposed to environmental air, allowing high evaporative and convective heat loss. It is far better to prevent this heat loss rather than to correct it after it occurs. Accepted techniques to maintain body temperature include raising the ambient temperature of the operating room to 30°C to 34°C (neutral thermal environment); warming and humidifying anesthetic gases35; using only warm scrub and irrigating solutions; using radiant heat lamps, portable heated isolettes, and thermal mattresses; and warming all intravenous fluids (Table 18-2). In most pediatric thoracic surgery cases, most or all of these heat loss prevention measures should be used. Core temperature (either nasopharyngeal, esophageal, or rectal) should be measured in every case.
F. Vitamin, Caloric, Electrolyte, and Fluid Requirements The newborn may be deficient in vitamin K, stores upon which the synthesis of coagulation factors II, VI, IX, and XI is dependent. Therefore, vitamin K, (0.5 to 1.0 mg) should be given intramuscularly or subcutaneously prior to surgery during the newborn period, especially if the child has never been fed. Hypoglycemia is common in the newborn period and is defined as a blood glucose level of less than 40 mg per cent during the first 72 hours of life, and less than 50 mg per cent after 72 hours of otherwise normal neonatal life. In low-
666
Anesthesia for Pediatric Thoracic Surgery
Signs and Symptoms of Hypothermia Normal core temperature Vasoconstriction begins Shivering prominent (BMR increased sixfold) Maximum ventilatory response to pC0 2 changes, confusion and disorientation (lowest measured temperature on many thermometers) Amnesia, lethargy, apathy, cardiac arrhythmias, bradycardia, prominent shivering stops Osborn waves regularly seen on ECG
Spontaneous respirations slow Coma, pupils dilate, tendon reflexes absent, heart rate half of baseline Risk of ventricular fibrillation high Primary vascular paralysis Venous stasis (blood viscosity 173% normal) Spontaneous respirations cease
EEG flatline
Figure 18-7 Signs and symptoms of hypothermia. (BMR = basal metabolic rate; ECG = electrocardiogram; EEG = electroencephalogram.) (Based on data from Wilson WC [personal communication] and Reuler.33)
Anesthesia for Pediatric Thoracic Surgery
Table 18-2
1. 2. 3. 4. 5. 6. 7. 8.
TECHNIQUES TO PREVENT HEAT LOSS IN INFANTS AND NEONATES Use portable heated isolettes Warm operating room to 30 to 40°C Use thermal mattresses Warm intravenous fluids Use radiant heat lamps Warm and humidify anesthetic gases Warm scrub and irrigating solutions Put hat on infant
birth-weight and premature infants, hypoglycemia is defined as a blood glucose level of less than 30 mg per cent. Infants of diabetic mothers (high fetal insulin production) and infants experiencing perinatal distress of any kind can be hypoglycemic. The presence of any condition causing stress or a history of a diabetic mother should make the anesthesiologist aware of the possibility of hypoglycemia and the need to determine blood sugar levels in the perioperative period. Normal glucose requirements for premature and normal newborns are 4 mg/kg/min. Neonatal hypocalcemia is related to low parathyroid hormone activity and decreased stores of available calcium.36 A rapid infusion of 10 mg/kg of calcium chloride or 30 mg/kg of calcium gluconate is usually effective in reversing hypocalcemia. Sodium, potassium, and chloride requirements are all similar to those of adults at 1 to 3 mEq/kg/24 hours. Using a solution of 5 per cent dextrose and 0.2 per cent saline with 20 mEq of KC1 added per liter, maintenance fluids can be approximated by administration of 4 ml/kg/hour for the first 10 kg of body weight and an additional 2 ml/kg/hour for the second 10 kg of body weight.37 If maintenance fluids are all that is required in a specific patient, it is advisable to calculate the preoperative fluid deficit and correct half of the deficit in the first hour, with further administration of fluid at 120 per cent of the base-line maintenance rate until the entire deficit is corrected. Generally, this simple approach is complicated by the trauma of surgery, requiring the intraoperative administration of substantially more fluid.37 The increased requirement may vary from 2 ml/kg/hour to 10 ml/kg/hour above the maintenance level for most surgery (however, major surgery, like repair of diaphragmatic hernia, may require much more; see previous discussion) and should consist of either a lactated Ringer's solution or saline. Blood should be infused if blood loss exceeds 15 per cent of the blood volume or reduces the hematocrit to less than 30.38
667
G. Airway Anatomy Anatomically, the child, and especially the infant, has an airway that is different enough from that of the adult to make intubation more difficult (Fig. 18-8). The head and occiput are relatively larger than the rest of the body, which requires the neck to be flexed on the chest if the face is in a midline sagittal plane; this is why a newborn lies with the face turned to one side. However, if the child's head is put into a good "sniffing" position (neck flexed on chest and head extended on neck) by extending the head on the neck, the exposure of the larynx will be facilitated. Still, during and after intubation, the large occiput causes the head to be unstable, with a propensity to roll to one side or another; this can be remedied by placing the occiput inside a doughnut-shaped stabilizing sponge pillow. There are other differences in the anatomy of the newborn airway compared with that of the adult airway. The narrowest portion of the infant's airway is the subglottic area at the level of the cricoid cartilage (in the adult the narrowest area is at the rima glottidis). The functional significance of the small diameter is enormous; if a neonate's larynx measures 4 mm at the level of the cricoid cartilage, a 1-mm reduction in this diameter, caused by either trauma or infection, will reduce the overall cross-sectional area of the airway by approximately 50 per cent. Table 18-3 shows guidelines for selection of endotracheal tubes in terms of appropriate internal diameter (i.e., an in-
668
Anesthesia for Pediatric Thoracic Surgery
Figure 18—8
The head and airway anatomy of the infant differs from that of an adult in the ways that are listed in the figure.
ternal diameter of an uncuffed endotracheal tube results in a leak around the endotracheal tube at 15-20 cm H20).39 In a normal newborn, the neck is short in comparison to the adult. The larynx is more cephalad in the child, lying approximately at the level of C2-C-3 compared with C-5 in adults. The larynxto-carina distance is only 4 cm (range = 2.5-5.0 cm)40 in the infant, and consequently great care must be taken to pass the endotracheal tube beyond the vocal cords by only 1.5 to 2.0 cm and to avoid severe head flexion (moves endotracheal tube caudad by 1 cm), head extension (moves endotracheal tube cephalad by 1 cm), and wide mouth opening (can move endotracheal tube caudad by 1 cm) to avoid bronchial cannulation or tracheal decannulation.40 Table 18-4 shows guidelines for the proper depth of insertion of endotracheal tubes as judged from the level of the anterior gum line. The infant's epiglottis projects cephalad at 45 degrees or more "from the anterior wall of the larynx, and it is relatively stiffer, longer, and omegashaped than an adult epiglottis. Although this makes the epiglottis easier to visualize, it also
makes the epiglottis more difficult to displace so that the laryngeal aperture may be visualized. For these reasons, a Miller blade is preferable to a Mackintosh blade for intubation of infants. The tongue, like the head, is relatively larger in children than in adults and may cause obstruction during induction of anesthesia and ventilation by mask and may impair visualization of. the larynx. Although the triad of more cephalad displacement of the larynx, increased difficulty in displacing the epiglottis, and the presence of a large tongue may make the larynx seem to be more anterior compared with that of the adult, it is not. All of the air passages (nose, nasopharynx, trachea) are small. Consequently, they may be easily obstructed, and at this small size, any obstruction can cause a catastrophic increase in resistance to airflow (see cricoid diameter discussed previously). There are also a number of physiologic differences between the infant and adult that make the infant more susceptible to hypoxia (Table 18-5).41 Finally, there are two primary determinants ol which main-stem bronchus an endotracheal tube will enter: namely, the angle of branching from the trachea (left greater than right) and the side. t
) and Effects ( ( = » Obstruction and/or Distortion of Superior and/or Inferior Vena Cava, Pulmonary Veins, Trachea and Mainstem Bronchi Coughing
Negative Pressure (Suction) To Empty Hemithorax
Constriction of Atrioventricular Groove EKG-Dramatic Axis and/or Ischemic Changes
Figure 19-2 Following a transpericardial radical pneumonectomy, the heart may herniate through a pericardial defect into the empty hemithorax. Factors that can cause this complication (solid arrows) are placing the empty hemithorax in a dependent position, which allows gravity to pull the heart through the defect; applying a high level of negative-pressure suction to the empty hemithorax, which also pulls the heart through the defect; applying a high level of positive-pressure ventilation to the remaining nondependent lung, which pushes the heart through the defect; and coughing, which also increases nondependent hemithorax pressure. The effects of herniation of the heart (open arrows) consist of obstruction and/or distortion of the mediastinal attachments (superior and inferior vena cava, pulmonary veins, and trachea and main-stem bronchi) and constriction of the atrioventricular groove by a tightly fitting pericardial defect. The electrocardiogram (EKG) will show dramatic axis and/or ischemic changes.
more likely to demonstrate ischemic electrocardiographic patterns with atrial and ventricular arrhythmias, and the changes in the electrocardiogram may precede clinical symptoms.2 Constriction of the atrioventricular groove may cause acute cardiovascular collapse. The chest roentgenogram usually shows that the heart has assumed a spheric shape and has been displaced laterally and at a right angle to the mediastinum and that the apex is against the left chest wall.8 In most reports, herniation of the heart has occurred either immediately after the patient is turned from the lateral decubitus to the supine position (75 per cent) or during the first few hours of postoperative mechanical ventilation.2,7·l4, l6 In
one case, a late intraoperative change in the patient's position from supine to a flexed position precipitated torsion of the great veins.17 Obviously, the patient will still be under the care of the anesthesiologist in many of these situations.14·17 Adhesions form between the heart and pericardium by the third day after surgery, thereby decreasing the chance of herniation.1 However, this complication has occurred as late as several days after surgery.18 Events that increase intrapleural pressure in the nonsurgical (ventilated) hemithorax or that decrease intrapleural pressure in the surgical (empty) hemithorax may predispose the patient to cardiac herniation. Placing the patient with the empty hemithorax in a dependent position allows the
Early Serious Complications Specifically Related to Thoracic Surgery
heart to be pulled by gravity into the empty hemithorax (see Fig. 19-2). Use of high levels of pressure and volume in the remaining lung can push the heart into the empty hemithorax (see Fig. 192). Similarly, coughing can increase pleural pressure in the remaining lung and thereby promote displacement of the heart into the empty hemithorax (see Fig. 19-2). Conversely, inadvertently applying suction to the chest drain in the empty hemithorax can pull the heart through a pericardial defect (tension vacuthorax)7 l0 (see Fig. 19-2). The heart can also herniate through a small defect (3 cm) without any of these factors being present.14 The differential diagnosis consists of myocardial infarction, cardiac tamponade, massive pulmonary embolus, airway obstruction, and collapse of the remaining lung. When circulatory signs permit, the diagnosis can be confirmed by portable chest roentgenography. A diagnosis of cardiac herniation almost always requires immediate re-exploration5-7; however, there is one report of herniation of the heart that caused only roentgenologic changes and no immediate clinical signs,13 and there is only one report in which herniation of the heart caused only electrocardiographic changes for several hours.2 There are conservative measures that may improve cardiopulmonary function before and during transfer to the operating room (they are the reverse of the causes of herniation of the heart shown in Figure 19-2), and they should be instituted as soon as the diagnosis is entertained. First, the patient should be positioned so that the ventilated, nonsurgical side is in a dependent position and the empty surgical side is in a nondependent position.4 I8 l9 Gravity may return the heart and mediastinum to their normal anatomic positions.14 Even if the heart does not re-enter the pericardium, the repositioning may decrease atriocaval kinking and increase cardiac output. Second, avoidance of high levels of pressure and volume in the ventilated lung may allow the return of the heart to the pericardium. Consequently, the tidal volume should be reduced, positive end-expiratory pressure (PEEP) should be removed, and the respiratory rate should be increased slightly. Third, suction to the empty hemithorax should be discontinued. Fourth, the pharmacologic support of the circulation should be provided as needed. Fifth, injecting 1 to 2 L of air into the surgical hemithorax may push the heart and mediastinum back to normal anatomic positions and may perhaps return the heart to the pericardium. Success with this latter technique has been variable. 1618 · 20-22
III. PULMONARY TORSION Pulmonary torsion refers to parenchymal rotation on its bronchovascular pedicle and is thought
699
to be due to increased mobility of a lobe (as may occur to the remaining lobes in one hemithorax after a lobectomy). Torsion after an intrathoracic surgical procedure has been described in the right middle lobe,23·24 in the left lower lobe,25,26 in the left upper lobe, 27 · 28 and in an entire lung.29 Any patient with atelectasis or an expanding intrathoracic mass (and perhaps with severe chest pain) after thoracotomy should have lung torsion included in the differential diagnosis (which also includes intrathoracic bleeding and progressive atelectasis). Pulmonary torsion may also occur after blunt chest trauma, nonpulmonary thoracic procedures (see Kucich et al.28 and Berkmen et al.30 for a detailed list of references), and pneumothorax (upper lobe hangs down over the lower lobe).30 It is possible that infiltrates seen in an upper lobe after re-expansion of a collapsed lung resulting from a pneumothorax is not re-expansion pulmonary edema but hemorrhage caused by venous obstruction. Because torsion compromises the pulmonary vasculature (both the arteries and veins) as well as the bronchi, prompt recognition and surgical intervention are required to avoid the attendant morbidity and mortality (i.e., infarction and gangrene).31 A double-lumen endotracheal tube should be inserted before surgery if lung torsion is expected because, during surgical correction of this condition (either completion pneumonectomy or untwisting and stabilization of the lobe to another lobe or chest wall), massive hemorrhage may occur into the airways (as a result of the release of entrapped blood and continuing bleeding from necrotic lung tissue [i.e., vessels and airways]), drowning the dependent lung and producing severe hypoxia or death.27·28 Other intraoperative therapeutic modalities that the anesthesiologist should consider include the use of intravenous steroids to decrease any reactive pulmonary inflammation, use of PEEP to reexpand atelectatic lung tissue, and the use of saline lavage for the removal of any excess blood, which might form clots in the normal airways, thereby obstructing good lung parenchyma.
IV. MAJOR HEMORRHAGE Postoperative bleeding that necessitates emergency thoracotomy may occur in as many as 3 per cent of all thoracotomy patients.32 The mortality of major postoperative bleeding is approximately 23 per cent.32 There are several potential sources of major postoperative hemorrhage (Table 19-1), and these consist of bleeding from divided pulmonary arteries and veins due to slippage of surgical ligatures, diffuse bleeding from raw surfaces, and sys-
700
Early Serious Complications Specifically Related to Thoracic Surgery
Table 19-1 POTENTIAL SITES OF MAJOR HEMORRHAGE AFTER THORACOTOMY 1 Slippage of pulmonary artery and/or vein sutures or ligatures 2. Large raw surfaces 3. Bronchial arteries 4. Intercostal arteries
temic arterial bleeding (bronchial and intercostal arteries). Bleeding may be catastrophic if ligatures slip off a pulmonary artery or vein. Although these vessels are not highly pressurized, hemorrhage from these vessels is into a low pressure-high vol ume space and, therefore, may be massive. The risk of postoperative hemorrhage can be greatly diminished by adequate intraoperative exposure, and adequate intraoperative exposure may be greatly aided in most cases by one-lung ventilation and in a few cases by an intrapericardial approach to the hilum. In addition, branches of the pulmo nary artery or vein should be directly oversewn with vascular suture material and/or should be doubly ligated, and then a transfixion suture or pursestring suture with a nonabsorbable material should be placed between the double ligation.32 Hemorrhage from raw surfaces is more common after pneumonectomy than after lobectomy be cause following lobectomy the apposition of the remaining lobes of lung against the chest wall and mediastinum may greatly diminish bleeding from these surfaces. Bleeding from raw pleural surfaces is especially likely when vascular adhesions be tween the visceral and parietal pleura have been divided. Hypertension is contributory to diffuse bleeding from raw surfaces. Postoperative bleeding may be from systemic arteries, especially the bronchial arteries. Bleeding from bronchial arteries is encouraged by extensive mediastinal dissection that damages bronchial and mediastinal arteries. Bronchial artery bleeding will also occur when the bronchial artery is not in cluded in the closure of a bronchial stump. If the free bronchial artery goes into spasm, then it may retract proximally. Subsequently, when the spasm is released postoperatively, the vessel will hemor rhage. In addition, the intercostal arteries may bleed if there has been inadvertent placement of periosteal sutures through an intercostal vessel.33 The drainage from a patent chest tube is an excellent monitor of the amount of intrathoracic bleeding (Table 19-2). In addition, the hematocrit of the usual fluid* in the chest drainage is always less than 20 per cent, and intrathoracic bleeding will increase the hematocrit to some higher value (see Table 19-2). Significant hemorrhage, of
course, will be accompanied by the hemodynamic signs of hypovolemia, such as tachycardia, hypo tension, and decrease in vascular filling pressures and urine output (see Table 19-2). Significant hemorrhage cannot always be ruled out by the absence of chest tube drainage, since the chest tube may be blocked by clotted blood or otherwise obstructed. Nevertheless, the patient will still show signs and symptoms of hypovolemia. In addition, if the chest tube is not draining well, there may be signs of significant shift in the me diastinum to the opposite side. In extreme cases exsanguination may occur rap idly unless the bleeding is promptly controlled; indeed, immediate thoracotomy in the recovery room in order to obtain manual control of the bleeding site may be life-saving. Once this has been accomplished, the patient can be returned to the operating room for closure under aseptic con ditions. In all cases, blood, preferably fresh whole blood, should be administered. If time permits, the coagulation profile should be obtained, and spe cific coagulation factors and/or platelet concen trates should be administered, as indicated by the laboratory findings.
V. BRONCHIAL DISRUPTION Development of a bronchopleural fistula is a serious complication of pulmonary resection and as recently as 1978 carried a mortality of 23 per cent.34, 35 The term refers to any communication between the tracheobronchial tree and the pleural cavity and can result from dehiscence of the bron chial stump after lobectomy or pneumonectomy, from rupture of an inflammatory lesion or cavity within the lung into the pleura, or from trauma or neoplasm erosion. The symptoms and signs are variable and depend upon the size of the commu nication, the presence or absence of a chest drain, and whether or not there is fluid of any şort within the pleural space. A. Acute Ipsilateral Bronchial Disruption Gross disruption of a bronchial stump in the early postoperative period may result from techniTable19-2 SIGNS OF POSTOPERATIVE INTRATHORACIC HEMORRHAGE 1. Volume of blood in chest tube drainage 2. Hematocrit of blood in chest tube drainage 3. Signs of hypovolemia (tachycardia and decreased systemic blood pressure, systemic pulse pressure, central filling pressures, and urine output) 4. Signs of tension hemithorax (with nonfunctional chest tube)
Early Serious Complications Specifically Related to Thoracic Surgery
cal error in bronchial closure, along with the use of high airway pressure for mechanical ventilation. The escape of gas is usually signaled by massive bubbling of gas through the chest tube drainage system. The patients quickly become hypoxemic and hypercarbic and develop a sympathomimetic cardiovascular response to these gas-exchange changes. Immediate re-exploration and revision of the bronchial closure are required. As soon as gross bronchial disruption is recognized, the chest tube should be removed from suction and left to underwater seal only. If the leak continues to be large, strong consideration should be given to reinsertion of a double-lumen tube to prevent the escape of gas through the disrupted bronchial stump. The incidence of this postoperative catastrophe can be diminished intraoperatively (prior to chest closure) by the simple technique of filling the pleural space with irrigating solution and having the anesthesiologist exert 35 to 40 cm H 2 0 of pressure to demonstrate an airtight bronchial closure. If the chest tube is obstructed, a tension pneumothorax with mediastinal shift will occur; the presence of a chest tube does not always prevent tension pneumothorax (e.g., the chest tube may be in a loculated compartment). High inflation pressures will be required for ventilation, and subcutaneous emphysema may become manifest (Fig. 19-3). A tension pneumothorax will shift the mediastinum, interfere with venous return, and decrease the cardiac output and blood pressure (see Fig. 19-3). The chest roentgenogram is diagnostic, but if the clinical situation is critical, therapy should be instituted without taking one. Treatment consists of re-establishing the patency of the chest tube or inserting a needle or a trocar and cannula into the pleural cavity through the second intercostal space in the midclavicular line on the affected side.
B. Acute Contralateral Pneumothorax It should be remembered that a tension pneumothorax may occur in the contralateral side if very high inflation pressures are used to expand a previously collapsed lung (as commonly occurs for one-lung ventilation); the tension pneumothorax may become apparent only after closure of the chest.36 Other causes of acute postoperative contralateral pneumothorax relevant to thoracic surgery include damage to the contralateral pleura during surgery, puncture of the pleura during insertion of a central venous cannula or thoracic epidural needle, or damage to the bronchus by use of an endobronchial catheter.
701
C. Chronic Ipsilateral Fistula Development of a chronic bronchopleural fistula may be expected to occur in 1 to 3 per cent of pulmonary resection patients.34 A chronic bronchopleural fistula most frequently becomes evident within the first 2 weeks after operation, may be large, and may run a fulminating course characterized by sepsis, empyema, purulent sputum, and respiratory insufficiency. In a much smaller number of patients, bronchopleural fistulas may develop even later, usually are smaller, and usually occur in an insidious fashion, with malaise, fever, and the presence of multiple air-fluid levels on chest roentgenograms. Factors that predispose to the formation of a bronchopleural fistula after pulmonary resection are preoperative irradiation, infection, residual neoplasm at the site of the closure, and presence of a long or avascular stump. A large bronchopleural fistula that develops several days after pneumonectomy occurs dramatically. The onset is often abrupt, with a bout of coughing and dyspnea precipitated by a change of posture that allows the operated side to be uppermost. Thin red-brown fluid that is characteristic of the contents of the empty hemithorax is both expectorated and inhaled into the remaining lung. This fluid is often infected and always irritant, so that it provokes bronchoconstriction and local inflammation, causing severe respiratory distress and hypoxemia. Signs of circulatory failure are usual too, probably caused by both hypoxemia and septicemia. The diagnosis is usually clear on clinical grounds alone. If confirmation is required, a chest roentgenogram will show a considerable fall in the fluid in the pneumonectomy hemithorax (fluid is replaced by air) and consolidation and collapse in the remaining lung. Prompt resuscitation is essential and may need to include mechanical and differential lung ventilation with a double-lumen tube and support of the circulation. A functioning chest tube must be inserted immediately to evacuate the air (to prevent a tension pneumothorax) and fluid (to prevent further contamination of the opposite lung) from the pneumonectomy cavity. Postural drainage and chest physiotherapy have no place in the management of a large bronchopleural fistula. The patient should be recumbent, with the normal side in a nondependent position. The stump must be closed surgically.34 At surgery, a double-lumen tube should be inserted to separate the two main-stem bronchi. The separation of the two main-stem bronchi prevents further contamination of the remaining or unoperated lung and allows positive-pressure ventilation of the remaining lung without loss of tidal volume through the fistula. The loss of tidal volume
702
Early Serious Complications Specifically Related to Thoracic Surgery
Signs of Tension Pneumothorax
Figure 19—3 A tension pneumothorax will cause a mediastinal shift. The mediastinal shift compresses the contralateral lung (which causes poor compliance and high inflation pressures) and decreases venous return, cardiac output, and systemic blood pressure. Since both lungs are under pressure, the diaphragms are depressed and flat, and if the tension pneumothorax occurs postpneumonectomy, the free air in the pleural space will cause a fall in the postpneumonectomy fluid level.
through the fistula can prevent effective ventilation in the remaining lung in three ways. First, tidal volume will be lost through the chest tube if the chest tube is functioning. Second, tension pneumothorax will develop if the chest tube is not functioning, which will cause mediastinal shift and compression of the remaining lung. Third, the escaping gas will pressurize the fluid in the empty hemithorax and force it to enter the fistula, which can cause contamination of the remaining lung. From the foregoing considerations, it is obvious that the lumen to the resected side must be occluded before positive-pressure ventilation is begun. The endobronchial lumen of the double-lumen tube should be in the contralateral main-stem bronchus. Surgical procedures for chronic bronchopleural fistula may include open drainage by rib resection, multiple-rib thoracoplasty, coverage of the bronchial leak with muscle flaps, and additional pulmonary resection; many patients require multiple procedures.34 If the chronic fistula is small, the symptoms are unobtrusive at first and consist of fever, tachycar-
dia, and persistent cough with expectoration of small amounts of sputum, which is almost always blood tinged. Episodes of coughing at night and bouts of wheezy dyspnea are particularly suggestive of bronchopleural fistula. The chest roentgenogram at this stage (postpneumonectomy) may show a decrease in the volume of fluid in the pleural space and an increase in the amount of air at the apex, or the fluid level may fail to continue rising as it normally does. A small fistula after pneumonectomy may heal without surgery. Medical management includes appropriate use of antibiotics, nutrition, chest tube placement (Table 19-3), selection of the drainage device (at - 2 0 cm H 2 0, suction pumps vary in capacity, from 2 to 35 L/min [Table 19-4]),38 ventilator selection and use (Table 19-5) (consider high-frequency ventilation in a patient with a proximal bronchopleural fistula and a normal lung compliance), and diagnostic and therapeutic bronchoscopy (balloon occlusion, use of sealants).37 With respect to therapeutic bronchoscopy, there are now many reports of sealing all sizes of bron-
Early Serious Complications Specifically Related to Thoracic Surger\·
Table 19-3
CHEST TUBE IN BRONCHOPLEURAL FISTULAS (BPF)*
I. Large diameter chest tube for A. High-flow BPF B. Drainage of infected pleural space II. Therapy (ventilated patient) A. Positive intrapleural pressure (expiratory phase) = to desired positive end-expiratory pressure level B. Chest tube occlusion during inspiratory phase C. Combination of above D. Application of sclerosing agent to pleural space (perhaps aided by thoracoscopy) III. Negative effects of chest tube (ventilated patient) A. Loss of tidal volume B. Some of the lost tidal volume has participated in gas exchange, making control of ventilation more difficult C. Negative suction pressure may cause ventilator cycling *Modified from Baumann MH, Sahn SA: Medical management and therapy of bronchopleural fistulas in the mechanically ventilated patient. Chest 97:721-728, 1990. Used with permission.
chopleural fistulas with tissue glue or other bioadhesives (e.g., clotted blood), and the sealant is most often applied through a fiberoptic bronchoscope (see chapter 17). However, surgery is still sometimes necessary,34 and a double-lumen tube (with the endobronchial lumen in the contralateral main-stem bronchus) should be used to separate the two lungs during surgery. Surgical procedures are as previously described.
VI. RESPIRATORYFAILURE There are a number of events/consequences of general anesthesia for thoracic surgery procedures that predispose a routine patient to postoperative respiratory failure. Dependent-lung atelectasis/infection may occur in any generally anesthetized patient. In addition, there are some unusual causes of atelectasis/edema/infection in both the operative and nonoperative lung that are specific to the thoracic surgery patient. Finally, the development of Table 19-4
DRAINAGE SYSTEMS*
System 1. 2. 3. 4.
Emerson pump Pleur-Evac A4000 Thora-Klex Sentinel seal
Maximal Flow (L/min) with - 20 cm H20 35.5 34.0 19.7 2.3
* Adapted from Capps. JS, Tyler ML, Rusch VW, Pierson DJ: Potential of chest drainage units to evacuate bronchopleural air leaks. Chest 88:57S, 1985. Used with permission.
Table 19-5
703
CONVENTIONAL VENTILATION IN BRONCHOPLEURAL FISTULAS*
I. Decrease fistula flow (reduce airway pressure) A. Lowest effective tidal volume B. Least number of mechanical breaths (allow spontaneous ventilation if possible; use intermittent mandatory ventilation) C. Lowest positive end-expiratory pressure and avoid expiratory retard D. Shorten inspiratory time II. Other maneuvers A. Selective intubation of both lungs (i.e., double-lumen tube) B. Differential lung ventilation C. Put the bronchopleural fistula in a dependent position *From Baumann MH, Sahn SA: Medical management and therapy of bronchopleural fistulas in the mechanically ventilated patient. Chest 97:721-728, 1990. Used with permission.
lung infection with gram-negative organisms may cause/contribute to the development of adult respiratory distress syndrome (ARDS) and sepsis.
A. Atelectasis and Lung Infection Acute respiratory insufficiency (within 30 days of resection) is one of the most common serious complications after pulmonary resection. Bronchial carcinoma resection data from a large thoracic surgery service indicate an incidence of approximately 4 to 8 per cent,39·40 and in one study, the patients in whom acute respiratory failure developed had a mortality of 50 per cent.39 Mortality from respiratory failure after right-sided pneumonectomy is greater than after pneumonectomy on the left side because the remaining functioning lung is smaller.39 Chapter 3 discussed the mechanisms of hypoxemia and hypercarbia during anesthesia and surgery, and many of these mechanisms may continue to function in the postoperative period. In addition, pulmonary resection per se involves additional new mechanisms of hypoxemia and hypercarbia. Most of the mechanisms of hypoxemia during anesthesia and surgery involve a decrease in the functional residual capacity below the closing volume of the lung, causing either low ventilationperfusion regions or atelectasis (Fig. 19-4). First, postoperative pain will cause splinting of the chest wall, compression of the lungs, and a decrease in functional residual capacity. Second, splinting will also prevent coughing and deep breathing, which will promote retention of secretions, which, in turn, will decrease the functional residual capacity. These first two mechanisms of postoperative hypoxemia can be minimized by appropriate control of pain (see chapter 21). Third, the remaining lung in either hemithorax may be edematous and/or
704
Early Serious Complications Specifically Related to Thoracic Surgery
Major Mechanisms of Lung Infection and Impaired Pulmonary Function Specific to Postoperative Thoracic Surgery Patients
Figure 19-4 The major mechanisms of atelectasis and lung infection that are specific to postoperative thoracic surgery patients. The inclusion of intrinsic positive end-expiratory pressure (PEEP) is based on Schmidt and Hall.42 The notion that atelectasis per se predisposes to infection is thought to be due to altered immune function in the atelectatic area and is based on data from Schmidt and Hall.42 The notion that lung infection may be due to aspiration of gram-negative organisms from the nasopharynx and gastrointestinal tract is based on data from Niederman et al.44
soiled with blood. The lung that was dependent during surgery may be edematous and/or may have aspirated blood (if a double-lumen tube was not used) owing to gravitational effects (in zones 3 and 4); the lung that was nondependent during surgery may be edematous and hemorrhagic owing to surgical compression and trauma and perhaps re-expansion of previously collapsed lungs (see next section). One-lung ventilation can minimize surgical trauma to the nondependent lung, and lung separation with a double-lumen tube will prevent aspiration of blood into the dependent lung. All of the remaining lung may be edematous because of a decreased vascular bed, increased cardiac output, decreased lymph clearance capability, and perhaps, but not necessarily, infusion of excessive amounts of fluid (see chapter 13 for a full discussion of postpneumonectomy pulmonary edema). The impact of diminishing the size of the pulmonary vascular bed should be predictable on the basis of preoperative pulmonary circulation and right ventricular testing results (see chapter 5). The
major mechanism of carbon dioxide retention is fatigue of the respiratory muscles and the diaphragm (which has decreased function after pulmonary resection)41 in attempting to move lungs that are now stiff (and may be hyperexpanded in areas that have intrinsic PEEP),42 have a high airway resistance, and have a diminished gasexchange surface. The lung involved in routine postoperative thoracic surgery may become infected via several mechanisms (see Fig. 19-4). First, for all the reasons cited previously, atelectasis may develop; atelectatic lung is susceptible to infection because of impaired atelectatic lung immunologic function.43 Second, major stress such as trauma and surgery may decrease immune function. Third, indwelling devices, such as endotracheal tubes, which are connected to a whole host of respiratory devices, and chest tubes, may cause lung infection. Fourth, aspiration of gram-negative bacilli from the nasopharynx and gastroesophageal area may initiate lung infection.44 Respiratory insufficiency is treated by instituting mechanical ventilation, and,
_J
Early Serious Complications Specifically Related to Thoracic Surgery
while the patient is being mechanically ventilated (see chapter 20), the underlying causes of the failure should be diagnosed and reversed.
B. ARDS The presence of lung infection may progress to ARDS and may be the cause of the septic syndrome, and the septic syndrome may independently cause ARDS (Fig. 19-5). Lung infection and the septic syndrome may cause ARDS by activation of complement and neutrophils. This mechanism is illustrated in Figure 19—6.45 46 ARDS may also be produced by a whole host of complement- and neutrophil-independent mechanisms (see Fig. 19—5). The pathologic hallmark of ARDS is lung edema despite low or normal hydrostatic pressures in the pulmonary microvasculature (i.e., permeability edema); the edema causes small, stiff lungs, which have a high right-to-left shunt. The specific diagnostic criteria for ARDS and the septic syndrome are defined in Tables 19-647 and 19-7,4*·49 respectively. Despite the increasing ability to define the mechanisms of ARDS and its occurrence,
705
the mortality from ARDS has been and still is approximately 40 per cent.45-51 The pathologic findings in the lungs of patients with ARDS are remarkably similar regardless of the underlying cause.50 Widespread alveolar and interstitial edema resulting from damage to the epithelial and endothelial cell layers is present within the first 24 hours of clinical symptoms. Within the next 24 to 48 hours, thickening of the alveolar walls caused by capillary congestion and alveolar hemorrhage occurs. Hyaline membranes, composed of cellular debris, protein, and fibrin line the respiratory bronchioles and alveolar ducts. Type I alveolar epithelial cells, which cover 95 per cent of the alveolar surface, are extensively involved. Damage may range from slight swelling to total cell destruction, leading to denudement of the basement membrane. The pulmonary circulation is constricted/obstructed both mechanically (endothelial cell edema, thrombi, platelet, fibrin, leukocyte aggregates) as well as actively (hypoxic pulmonary vasoconstriction [HPV], acidosis, vasoconstrictor amines, and peptides) (see Fig. 710).50·52·53 With the very first signs of respiratory distress, the chest roentgenogram is most frequently normal
706
Early Serious Complications Specifically Related to Thoracic Surgery
Figure 19-6 The complement ac· tivation-neutrophil-dependent adult respiratory distress syndrome (ARDS^ pathway. (Based on data from Rin aldo and Rogers45 and Said anc Foda.46)
and usually remains normal for the first 12 to 24 hours after the initial insult. The first abnormalities found in the early stages of ARDS are patchy, diffuse, symmetric, bilateral interstitial and alveolar infiltrates. The patchy zones of alveolar infiltrates rapidly coalesce to a more massive airspace consolidation. The cardiac silhouette is not enlarged, and the costophrenic and cardiophrenic angles are clear. This pattern of chest roentgenogram abnormalities is most common and constitutes the "progressive pulmonary edema" type. The lack of redistribution of blood flow to the upper lung zones, sparing of the costophrenic angles, absence Table 19-6
CRITERIA FOR DIAGNOSIS OF THE ADULT RESPIRATORY DISTRESS SYNDROME*
1. An appropriate risk factor (e.g., sepsis, aspiration, or trauma) 2. Severe hypoxemia while breathing increased oxygen concentrations 3. Increased pulmonary shunt fraction 4. Reduced lung compliance and volume 5. Radiographic evidence of pulmonary edema 6. Previously normal lung 7. No evidence of heart failure (Ppa030°C or 5 mmol (mEq)/L B. Systemic vascular resistance 65 with an F,0 2 3.5 mg/dl or patient requiring dialysis or urine output l .5 the control Glasgow coma score 20 mg/day for at least one month or other known immunosuppressive agent or patient with AIDS
*From Suchiya MR, Clemmer TP, Elliott CG, et al: adult respiratory distress syndrome: A report of survival modifying factors. Chest 101:1074-1079, 1992. Used with mission. Abbreviation: AIDS = acquired immunodeficiency drome.
The and per syn
active pulmonary vasoconstriction (such as caused by hypoxia, acidosis, and vasoactive amines and peptides) are placed on the right side the heart. Intrinsic PEEP (trapped gas at end-exhalation), just like extrinsic PEEP, increases intrathoracic pressure and adds yet another load on the right ventricle.42 Certainly, the development of ARDS will always cause a significant increase in pulmo nary vascular resistance, 52 · 53 and sepsis is regularly associated with or is causative of both right and 77 left ventricular dysfunction. In addition, pulmonary hypertension and right ventricular afterload will be increased by increased pulmonary capillary and venous pressures, which may result from fluid overload and/or decreased left ventricular compliance. Decreased left ventric ular compliance may be due to left ventricular ischemia and/or failure or interventricular septal shifting (i.e., right ventricular afterload stress may produce general dilatation of the right ventricle and cause interference with left ventricular per formance through ventricular interaction). 78 · 79 Right-sided heart failure may also be manifested
709
by increasing venous hypertension and peripheral edema, pulsatile liver, hepatojugular reflux, marked elevation of the jugular venous waveform, and tricuspid insufficiency. A pulmonary artery catheter should be used in any patient undergoing major pulmonary resection who is believed to be at risk for postoperative right-sided (or left-sided) cardiac failure; patients at risk for right-sided heart failure include those with an inferior wall myocardial infarction, which often involves the right ventricle, and those with pre-existing pulmonary hypertension. The diagnosis of selective right-sided heart fail ure is established when the right atrial pressure exceeds the left atrial pressure (i.e., the pulmonary artery wedge pressure) in the presence of an ab normally low cardiac output. In addition, pulmo nary hypertension with a large pulmonary artery diastolic to wedge pressure gradient will usually be present, along with the systemic signs of heart failure (oliguria, decreased mentation, and periph eral edema). One should remember that the pres ence of significant amounts of intrinsic PEEP might confound interpretation of the pulmonary capillary wedge pressure.42 The treatment of acute right-sided heart failure follows the same principles used in treating leftsided failure: Control heart rate, optimize preload and the inotropic state of the ventricle, and reduce afterload (see chapter 13). Preload, as measured by central venous pressure, can be reduced by fluid restriction, by diuretics, or by a venous vasodilator such as nitroglycerin. Inotropic drug support should take into account the response of the pul monary vasculature to the drug chosen. Thus, dobutamine, a drug that tends to reduce pulmonary vascular resistance, is a reasonable choice in this 80 setting. In fact, in this setting, a change from dopamine to dobutamine results in an increase in
Table 19-10 CAUSES OF INCREASED RIGHT VENTRICULAR AFTERLOAD AND FAILURE AFTER PULMONARY RESECTION 1. Resection of pulmonary vascular bed 2. Requirement for increased pulmonary blood flow (sympathetic nervous system activation caused by pain, stress, etc.) 3. Development of remaining lung infection, intrinsic PEEP, ARDS, HPV 4. Right ventricular dilation —* interventricular septal shifting —» decrease left ventricular compliance —» pulmonary hypertension 5. Decrease left ventricular compliance caused by left ventricular ischemia -» pulmonary hypertension Abbreviations: PEEP = positive end-expiratory pressure; ARDS = adult respiratory distress syndrome; HPV = hypoxic pulmonary vasoconstriction.
710
Early Serious Complications Specifically Related to Thoracic Surgery
stroke index and a significant decrease in right atrial pressure.81 Vasodilators that may be effective in reducing pulmonary vascular resistance include nitric oxide, nitroglycerin, nitroprusside, phentolamine, hydralazine, and prostaglandin E,.82 Of these drugs, nitric oxide and prostaglandin E, have the greatest pulmonary specificity (spares the systemic circulation), nitroglycerin and nitroprusside have intermediate pulmonary specificity, and hydralazine the least (causes only a moderate decrease in pulmonary artery pressure and resistance and a large decrease in systemic vascular resistance).83 In addition to these drugs, treatment should include still other measures aimed at reducing pulmonary arteriolar constriction. These include oxygen administration to reduce hypoxic pulmonary vasoconstriction, administration of bronchodilators to relieve bronchospasm, antibiotics and chest physical therapy to clear infection (to diminish the size of the hypoxic compartment), mechanical ventilation, and normalization of acid-base balance. Stabilization of cardiovascular function after initiation of respiratory therapy may be attributed, in part, to the disappearance of detrimental heartlung interaction.79
rate are a potent combination for producing myocardial ischemia/infarction. The diagnosis of myocardial infarction is made by history (pattern of pain), evidence of autonomic nervous system activation in the form of pallor, sweating and peripheral vasoconstriction, electrocardiogram changes and serial cardiac enzymes. Creatine kinase (CK) activity rises most rapidly (peak of 6-9 per cent occurs between 6 and 24 hours) and is cleared most rapidly (decreases to less than 1 per cent by 48 hours). The specific isoenzyme CK-MB helps to identify cardiac rather than skeletal muscle damage. Aspartate aminotransferase activity is intermediate (peaks at 48 hours), and lactic dehydrogenase or alpha-hydroxybutyric dehydrogenase activity rises later (peaks at 72 hours) and is cleared more slowly. The principal aspects of the management of acute myocardial infarction are administration of oxygen, nitroglycerin, analgesia (morphine), and aspirin, thrombolysis, and treatment of arrhythmias (beta blocker) and heart failure (oxygen, morphine, diuretic, venous and arterial vasodilators, ionotrope).
X. ARRHYTHMIAS IX. MYOCARDIAL ISCHEMIA/ INFARCTION -
Myocardial infarction and cardiac failure associated with noncardiac thoracic operations are important causes of mortality and serious morbidity.84-87 Transient ischemic electrocardiographic changes have been documented in 3.8 per cent of these patients and myocardial infarction in 1.2 per cent.88 The large majority of ischemic episodes (and arrhythmias) occur on the second and third postoperative days (Fig. 19-8).88 This, of course, corresponds with the time period of maximum decrement to postoperative pulmonary function. In fact, with close monitoring, a direct temporal relationship between decreases in Sp02 and decreases in ST-segment level, as well as both severity and duration of the ST-segment decreases, can be observed89 (Fig. 19-9). Postoperative decreases in Sp02 are usually accompanied by increases in heart rate, but the change in heart rate is usually proportionally less than the changes in Sp02 (see, e.g., Fig. 19-9).89·90 Increases in heart rate are less than the concomitant decreases in Sp02 postoperatively perhaps because the heart rate is usually already increased postoperatively as a result of pain/stress, and the potential for further increases is therefore reduced. Nevertheless, decreased Sp02 and increased heart
Supraventricular arrhythmias, primarily sinus tachycardia and atrial fibrillation and flutter, are frequent complications after major pulmonary resections, especially pneumonectomy,91·92 even in patients without significant pre-existing cardiac disease.88, 93 Atrial tachycardias occur in 16 per cent of patients; atrial fibrillation is preponderant (64-87 per cent), followed by supraventricular tachycardia (23 per cent) and atrial flutter (13 per cent).88·93 The potential causes of these arrhythmias after pulmonary resection include retraction and trauma to the heart (e.g., arrhythmias occur more frequently after intrapericardial dissection93 and in patients who have a pneumonectomy vs. a lesser procedure)88,93 and distension of the right ventricle and atrium (caused by increased pulmonary artery pressure). Arrhythmias occur more frequently in patients in whom postoperative pulmonary interstitial edema or perihilar infiltrates develop.93 The propensity for arrhythmias to occur in a traumatized distended atrium may be compounded by pre-existing cardiovascular disease, poor gas exchange, and sympathetic nervous system stimulation caused by pain. These arrhythmias may contribute greatly tc perioperative morbidity and mortality,8 especially in the aged.94 Indeed, patients with recurrent epl· sodes of arrhythmias have a significantly highe: mortality than those without episodes or with {
Figure 19-8 Distribution of post operative cardiac events in 53 pa tients by days after operation (From Von Knorring J, Lepantalo M, Lindgren L, Lindfors O: Cardiac arrhyth mias and myocardial ischemia after thoracotomy for lung cancer. Ann Thorac Surg 53:642-647, 1992. Used with permission.)
single episode only (17 per cent vs. 2.4 per cent; ρ < .Ol).88 In pneumonectomy patients, the occur rence of an arrhythmia is associated with a signif icant increase in mortality (25 per cent vs. 7 per cent; ρ < .Ol).93 In survivors, the occurrence of an arrhythmia doubles the length of stay in the inten sive care unit.95 There are no generally accepted preventive reg imens for these arrhythmias. Digoxin has been the only extensively studied drug to prevent atrial ar rhythmias after thoracic operations. However, be cause of positive as well as negative reported re sults, this regimen remains controversial (see chapter 5).88 It was demonstrated that intravenous infusion of the antiarrhythmic drug flecainide ace tate for 3 days after thoracotomy prevented or re
duced atrial and ventricular arrhythmias without any side effects.96 The preventive effect of verap amil in one study also seems promising.97 Once the arrhythmia occurs, general treatment consists of adequate sedation/analgesia and correc tion of hypoxia, hypercarbia or hypocarbia, and right-sided heart failure (see Right-Sided Heart Failure). Specific pharmacologic antiarrhythmia treatment is dependent on the particular arrhyth mia98 (see Table 13-19). Sinus tachycardia is treated with propranolol (0.5 mg every 2 min up to 5 mg). Paroxysmal atrial tachycardia can be treated with verapamil or adenosine. Treatment of atrial fibrillation is with digitalis (0.25-0.5 mg in travenously) if the patient is reasonably stable, ver apamil (5 mg/dose) if the patient is moderately
Figure 19—9 Patient No. 2: ST segment level, saturation, and heart rate (HR) plotted during the first night after operation. Removal of oxygen was associated with severe decreases in S P 0 2 and periods of ischemia. No ischemic episode occurred while the patient received oxygen. (From Reeder MK, Muir AD, Foex P, et al: Postoperative myocardial ischaemia: Temporal association with nocturnal hypoxaemia. Br J Anaesth 67:626-631, 1991. Used with permission.)
712
Early Serious Complications Specifically Related to Thoracic Surgery
Figure 19-10 Among the general population, 20 to 34 per cent have a patent foramen ovale. Normally, the patent foramen ovale remains functionally closed because left atrial pressure (LAP) exceeds right atrial pressure (RAP). When pulmonary vascular resistance is increased, such as following a pneumonectomy or lobectomy, right ventricular (RV) and right atrial pressures may be increased. If right atrial pressure exceeds left atrial pressure, then the foramen ovale opens and permits right-to-left shunting. (LV = left ventricle; PVR = pulmonary vascular resistance; COPD = chronic obstructive pulmonary disease; CHF = congestive heart failure; ARDS = adult respiratory distress syndrome; HPV = hypoxic pulmonary vasoconstriction.)
unstable, and externally synchronized cardioversion if the patient is extremely unstable. Treatment of atrial flutter consists of ouabain (0.1-0.2 mg intravenously) or verapamil (5-10 mg intravenously) (equal first choices), propranolol (second choice), and external cardioversion (third choice). Ventricular arrhythmias, usually premature ventricular contractions, can be controlled with intravenously administered lidocaine and/or betablocking drugs. Ventricular fibrillation and tachycardia must be controlled by electric defibrillation, followed by lidocaine (Xylocaine) and, perhaps, propranolol and, rarely, bretylium. Temporary pacemaker insertion should be strongly considered for patients in whom complete heart block has developed. XI. RIGHT-TO-LEFT SHUNTING ACROSS A PATENT FORAMEN OVALE Many adults have a patent foramen ovale; at autopsy, the incidence of a probe-patent foramen ovale is highest at 34 per cent during the first three decades of life and decreases to 20 per cent by the ninth and tenth decades." There is normally no right-to-left shunting across the foramen ovale because left atrial pressure exceeds right atrial pres-
sure, which keeps the one-way flap valve of the foramen ovale pressed against the foramen ovale, resulting in a functionally competent seal. If the right atrial pressure exceeds left atrial pressure (as might occur during the use of PEEP100·l01; in the course of pulmonary embolization,102, 103 pulmonary hypertension,102 high-altitude HPV,104 chronic obstructive pulmonary disease,105 ARDS,106 pulmonary valvular stenosis,100· 107 congestive heart failure,100·108 right ventricular infarction,109 or cardiopulmonary by-pass102, "°· '"; and during neurosurgical procedures complicated by air embolization"2), the one-way flap valve can open, and right-to-left shunting can occur through the foramen ovale (Fig. 19-10). In terms of commonly occuring events, right-toleft shunting at the atrial level may occur with reaction to the tracheal tube during emergence from anesthesia113 and during the application of PEEP.101 The mechanism of increase in right atrial pressure associated with reaction to the tracheal tube is a decrease in lung volume that causes increased pulmonary vascular resistance, pulmonary artery pressure, and right atrial pressure.113 Several reports documented new-onset right-toleft shunting across a patent foramen ovale or atrial septal defect after pneumonectomy114-120 and lobectomy121· 122 as a cause of otherwise unexplained postoperative dyspnea and systemic oxy-
Early Serious Complications Specifically Related to Thoracic Surgery
gen desaturation (Figs. 19-10 and 19-11).120 Mechanisms for the shunting across the patent foramen ovale include an increase in right atrial pressure to greater than left atrial pressure, which is caused by an increased pulmonary vascular resistance (due to the resection of vascular bed and/ or anesthesia- and surgery-induced lung disease) and/or a shifting in the anatomic relationship of the inferior vena cava, right atrium, and intra-atrial septum resulting from the pulmonary resection, so that the one-way flap valve is functionally less competent and permits shunting across the foramen ovale (see Figs. 19-10 and 19-11).120 The diagnosis of right-to-left shunting across a patent foramen ovale should be suspected in any patient who is hypoxemic (with relatively normal chest roentgenogram findings) or who has a progressive PEEP-induced decrease in the arterial oxygenation tension (these latter patients likely have abnormal chest roentgenogram findings).101 In the latter circumstance, PEEP is presumably further increasing pulmonary vascular resistance and right ventricular and right atrial pressures, thereby increasing the right-to-left shunt across the foramen. In extubated patients, the diagnosis should be suspected by otherwise unexplained clinically significant dyspnea and hypoxemia. Contrast two-dimensional and transesophageal echocardiography are excellent, increasingly readily available methods by which the diagnosis of this condition, if suspected, may be made.101·106' " 3 · l 2 0 · l 2 2 Contrast angiography and dye dilution curve analysis are the best definitive invasive diagnostic methods.110 Since surgical closure is stressful to the patient (it requires cardiopulmonary by-pass) and, in the presence of pulmonary hypertension, may result in acute right ventricular failure, therapy should always first be nonsurgical and aimed at decreasing pulmonary vascular resistance and right ventricular and right atrial pressures. Thus, nonsurgical treatment consists of decreasing pulmonary vasoconstriction by clearing pulmonary infections, administering oxygen, reversing acidosis, eliminating conditions that release pulmonary vasoactive amines and peptides (such as hypotension and infection), using pulmonary vasodilators, and decreasing right-sided heart pressures by preload manipulation. These nonsurgical treatment modalities usually permit functional closure of the foramen ovale in the majority of patients. In one postpneumonectomy patient123 and one patient with a right ventricular infarction,109 interatrial shunting was successfully managed for a time by occluding the foraminal orifice with a balloon-tip angiography catheter. Rarely, a patient may undergo surgical closure, even though he or she may still be on a mechanical ventilator and critically ill.106·109 The general care of a patient with right-to-left
713
intracardiac shunts must be meticulous. Avoidance of air bubbles in the intravenous lines is mandatory. Forceful flushing of poorly infusing venous catheters may result in systemic embolization.108 Blood to be transfused should be filtered through a micropore filter to exclude particulate material. Since postoperative patients have an increased propensity to venous stasis, prophylactic anticoagulation should be considered to reduce the possibility of systemic thromboembolization.108
XII. MASSIVE POSTPNEUMONECTOMY MALIGNANT PLEURAL EFFUSION AND CHYLOTHORAX At the end of a pneumonectomy, care should be taken to ensure that the remaining lung is well expanded and the mediastinum is near the midline (see chapter 13). Excessive midline shift can compromise venous return to the heart. During the weeks to months that follow surgery, fluid gradually accumulates and fills the initially air-filled hemithorax. The mediastinal structures slowly shift toward the side of the resection, the ipsilateral diaphragm progressively elevates, and the remaining lung herniates anteriorly across the midline. An early mediastinal shift toward the remaining lung is a cause for concern and may indicate atelectasis, postoperative hemorrhage, or infection in the residual cavity with possible bronchopleural fistula. A late mediastinal shift toward the remaining lung is unusual, and the possibility of a malignant effusion or chylothorax should be considered. Malignant pleural effusions and chylothorax account for substantial morbidity. Treatment of malignant effusions is directed toward palliation because life expectancy usually is limited and curative therapy does not exist. Ordinarily, the most popular methods are repeated thoracentesis, chest tube drainage (chemical pleurodesis is not an option because, with the lung removed, there is no visceral pleura to oppose the parietal surface and thereby obliterate the pleural space), and re-exploration.124 The same approach may be used for chylothorax, but parenteral nutrition may allow a small thoracic duct to close (low fat intake decreases the formation of chylous fluid), and re-exploration is necessary much more often because the large thoracic ducts may have been torn.124 Pleuroperitoneal shunting for postpneumonectomy malignant effusion125 and chylothorax126 has proved to be an effective alternative treatment. A unidirectional shunt valve with manual pump is built into the shunt and ensures clearance of fluid into the abdominal cavity regardless of the près-
714
Early Serious Complications Specifically Related to Thoracic Surgery
Figure 19-11 Transesophageal echogram showing a patent foramen ovale (PFO). Micro air bubbles were observed to easily cross the PFO. This patient had normal oxygenation before right pneumonectomy and became severely hypoxemic after right pneumonectomy. Surgical closure of the PFO resulted in a dramatic improvement in oxygenation. (RA = right atrium; LA = left atrium; AO = aorta.) (From Berry L, Braude S: Refractory hypoxaemia after pneumonectomy: Diagnosis by transesophageal echocardiography. Thorax 47:60-61, 1992. Used with permission.)
sure gradient between the thorax and abdomen. The advantage of this remedy is that it can be placed with the patient under local anesthesia. An obvious concern is causing symptomatic intra-abdominal tumor implantation, but this complication has not been reported. Malignant effusions have been successfully shunted for periods exceeding 1 year.127
XIII. SYSTEMIC TUMOR EMBOLISM Tumor emboli to peripheral vessels usually arise from left atrial myxomas. However, the second most common source of seeding is bronchogenic lung carcinoma; primary lung neoplasms account for 11 of the 29 cases of systemic tumor emboli reported in the largest series.128 Tumor invasion of the pulmonary vein with subsequent dislodgment during surgical manipulation leads to systemic embolization. The common femoral artery is most often the site of arterial tumor embolism,129 but unexpected occlusion of any artery after thoracotomy for tumor should raise suspicions of tumor emboli. For example, intraoperative arterial embolization of a bronchogenic tumor has led to occlusion of the axillary artery (which caused immediate damping of the radial artery trace during surgery).130
XIV. NEURAL INJURIES During radical hilar dissection or excision of mediastinal tumors, phrenic, vagus, and recurrent
laryngeal nerves may be injured accidentally or may be excised deliberately. Phrenic nerve injury causes respiratory embarrassment by a flail chest effect and also causes elevation of the ipsilateral hemidiaphragm. The diagnosis should be suspected in patients who have relatively clear chest roentgenograms, have adequate gas exchange, and cannot be weaned from the ventilator.131 The diagnosis can be confirmed by paradoxic movements of the diaphragm on fluoroscopy. Injury to the vagus nerve causes gastric and intestinal atony, which usually is not problematic in the first few postoperative days. Bilateral partial injury of the recurrent laryngeal nerve causes adductor spasm of the vocal cords, which, following extubation. may result in upper airway obstruction. This musl be diagnosed promptly and treated with immediate reintubation and possible tracheostomy until the dysfunction is resolved. The recurrent laryngeal nerve or phrenic nerve usually shows return of function within 2 to Ç months after injury. If vocal cord function doe; not return after recurrent laryngeal nerve injury the involved vocal cord may be injected with ste roids, which often causes a return of function. Ir rare cases, surgical procedures, including the injec tion of Teflon, may have to be performed on th< involved cord to improve its function. If phreni< nerve paralysis persists and continues to caus< marked respiratory embarrassment due to paralysii of the diaphragm, it is possible to implant j phrenic nerve stimulator132 or plicate the paralyze< diaphragm133 to correct the respiratory insuffi ciency. Aside from clamping of the thoracic aorta, then are three other causes of paraplegia after thoracot
Early Serious Complications Specifically Related to Thoracic Surgery
omy. First, damage to the spinal branches of intercostal arteries by dissection or diathermy at the posterior end of a rib will cause spinal cord ischemia.133 This is most likely to occur with damage to the intercostal arteries of the left lower lobe.134 Second, surgical dissection may create a communication between the epidural space and the pleural cavity. Clotted blood, or blood that can later clot, can then enter the epidural space and cause postoperative spinal cord compression and ischemia.135 Third, packing hemostatic gauze into the posterior end of a thoracotomy incision to stop intercostal artery bleeding and leaving the packing in place at the end of the thoracotomy has caused paraplegia in three patients; subsequent neurosurgery showed that the hemostatic gauze disrupted the dura and was intradural.136 Presumably, the hemostatic gauze was forced/propelled by tissue forceps through an intervertebral foramen, subsequently swelled when left in situ, and compressed the spinal cord.136 There are other less serious neural injuries that are also specifically related to thoracic surgery. The brachial plexus is especially vulnerable to trauma during thoracic surgery and anesthesia.137· I38 The brachial plexus has a potentially long superficial course in the axilla between two points of fixation: the vertebrae above and the axillary fascia below. If the two points of fixation are separated widely, stretching of the plexus will occur. Suspension of the arm from an ether screen in an excessive cephalad direction (see Fig. 10-1) in a paralyzed patient without muscle tone, which increases the distance between the two points of fixation, is the most common cause of stretching and damage to the brachial plexus. Stretching of the plexus is also promoted by extreme abduction, external rotation, and dorsal extension of the arm on an arm board. The brachial plexus may also be injured by a malpositioned chest tube (pushed in too far cephalad).139 The intercostal nerves are most exposed to injury during intrathoracic surgical procedures. A rib fracture occurring at the time of thoracotomy can compress the intercostal nerve, resulting in an intercostal neuritis, which is manifested by radicular pain in the postoperative period. When excessive stretching and tension are placed on the intercostal nerve roots while the chest is being opened, postoperative intercostal neuralgia may be quite severe. Median sternotomy may cause a brachial plexopathy,140,141 and the sternal wires may cause a parasternal intercostal neuropathy. The latter is very effectively treated by local blocks.142 In addition, the chest tubes can produce a neuroma or neuritis by compression of an intercostal nerve.
715
XV. COMPLICATIONS OF INTRATHORACIC INTERCOSTAL NERVE BLOCKS The performance of intrathoracic intercostal nerve blocks under direct vision just prior to closure of the chest is a common method of securing immediate postoperative analgesia. In one large institution it has been estimated that these blocks are performed in 10 to 15 per cent of thoracotomy patients.143 However, there have been case reports of total spinal anesthesia following this procedure.143 144 In these patients, following placement of intrathoracic intercostal nerve blocks at several different levels, there was an immediate profound decrease in blood pressure and pulse, persistent apnea, failure to regain consciousness, absence of reflexes, and dilated pupils, all of which are consistent with total spinal anesthesia. In addition, the temporal sequence of clinical recovery145 and a clear-cut sensory level when the patient awakened also strongly indicate the occurrence of total spinal anesthesia. There are several possible mechanisms for this complication. First, there can be outward prolongation of the subarachnoid space along the nerve roots. These durai cuffs can extend as far as 8 cm past the intervertebral foramen.145·146 It is possible, then, that the local anesthetic was introduced into one or more durai cuffs surrounding the intercostal nerves. A second possibility is that the local anesthetic can spread centrally to the spinal fluid via the perineural spaces. A third possibility is that the local anesthetic could have been introduced directly into the subarachnoid space through an intervertebral foramen because of an unnoticed illplaced angulation of the needle. These reports constitute a relative contraindication to the performance of intercostal nerve blocks near the spinal cord with local anesthetics when the patient is under general anesthesia and when it is not possible to assess accurately the effect of the nerve blocks. These reports also emphasize the necessity of including spinal anesthesia in the differential diagnosis of postoperative cardiopulmonary and central nervous system depression when local anesthesia has been used in the vicinity of the spinal cord during general anesthesia.
REFERENCES 1. Allison PR: Intrapericardial approach to the lung root in the treatment of bronchial carcinoma by dissection pneumonectomy. J Thorac Cardiovasc Surg 15:99, 1946. 2. Baaijens PFJ, Hasenbos AWM, Lacquet LK, Dekhuijzen PNR: Cardiac herniation after pneumonectomy. Acta Anaesthesiol Scand 36:842-845, 1992. 3. Yacoub MH, Williams WG, Ahmad A: Strangulation of the heart following intrapericardial pneumonectomy. Thorax 23:261-265, 1968.
CHAPTER
20
Mechanical Ventilation and Weaning
720
Mechanical Ventilation and Weaning
I. INTRODUCTION The clear-cut indications for postoperative mechanical ventilation (at least for a few hours) are numerous and include severe pre-existing lung disease (see chapter 5), ventilatory depression due to residual anesthesia or paralysis (for example, the peak inspiratory force is less than 20 cm H 2 0), presence of a significant amount of blood in the airway, massive intraoperative transfusion following many pneumonectomies, large air leak following any resection (possible bronchopleural fistula), presence of flail chest, extensive surgical trauma to the lung remaining in the operative hemithorax, multiple and severe organ trauma, sepsis, and unacceptable arterial blood-gas concentrations (perhaps due to the factors just listed). In addition to these obvious indications, continuing intubation and mechanical ventilation facilitates a smooth transition from the operating room to the intensive care unit (provided the patient is adequately anesthetized/sedated); allows the intraoperative use of high-dose narcotic anesthesia; enables adequate postoperative analgesia without concern for depression of respiration; allows a more aggressive approach to hemodynamic support (including intravascular volume loading) without excessive concern for its effects on pulmonary function; avoids the need for application of intensive respiratory therapy in exhausted, stressed patients in the early postoperative period; allows restoration of functional residual capacity (FRC) prior to extubation; and facilitates return to the operating room if there are concerns about surgical complications. If postoperative mechanical ventilation is planned, even for only a few hours, the patient should continue to receive narcotics and be paralyzed, and the double-lumen tube should be changed to a single-lumen tube (except perhaps when there is a large bronchopleural fistula; see Differential Lung Ventilation later) prior to leaving the operating room. The use of positive end-expiratory pressure (PEEP), continuous positive airway pressure (CPAP), intermittent mandatory ventilation (IMV), of either a spontaneously breathing or apneic patient, pressure support ventilation of a spontaneously breathing patient, and an increased understanding of the toxicity of unnecessarily high inspired concentrations of oxygen (F,0 2 ) constitute the cornerstones of modern clinical mechanical ventilation management. This chapter describes, in a step-by-step manner, the current practice of mechanical ventilation, weaning, and extubation.'^ The recommendations for initial ventilator settings (see later) are based on the assumption that the patient's lungs are disordered and, therefore, require a considerable amount of support. In addi-
721
tion, it is assumed that a volume-limited ventilator is being used, the patient is adequately sedated, and the endotracheal tube and chest tubes are functioning properly.
II. INITIAL VENTILATOR SETTINGS: INTERMITTENT MANDATORY VENTILATION IMV means that the ventilator delivers a positive-pressure breath a mandatory number of times per minute while the patient does as much spontaneous breathing as desired between these mechanical breaths. The IMV mode of ventilation is at present the most commonly used and preferable ventilation mode for several reasons.5 First, because the patients are allowed to do, and usually are doing, some spontaneous ventilation, they are able to generate a more negative pleural pressure, which better maintains venous return and cardiac output (compared with simple intermittent positive-pressure ventilation.6-9 The descent of the diaphragm into the abdomen with spontaneous ventilation also increases abdominal pressure, squeezes the splanchnic circulation, and contributes to venous return.10 Second, and consequently, less intravenous fluid is usually required to maintain systemic pressure and perfusion. Third, since it is not necessary to eliminate these spontaneous breaths, sedation requirements are minimized, and paralysis requirements are eliminated. Fourth, the patients benefit psychologically because they know they are breathing on their own. Fifth, the respiratory muscles remain coordinated because they are exercised. Finally, when the weaning stage is reached, IMV allows a gradual transition from ventilator dependence to independence. There are a number of other mechanical ventilation modes that also facilitate mechanical ventilation and/or the transition from mechanical ventilation to weaning, such as pressure support (PS), mandatory minute ventilation (MMV), inverse ratio ventilation (IRV), and airway pressure release ventilation (APRV), which have been introduced; these are discussed under Other New Mechanical Ventilation and Weaning Modes. With conventional IMV, it is possible for the ventilator to begin to deliver a mechanical breath just as the patient is completing a spontaneous inhalation. This results in the stacking of a machine breath on top of a spontaneous breath and can lead to overdistension of the lungs and high peak inspiratory pressures. This problem can be overcome by use of synchronized IMV (SIMV). With SIMV, the ventilator intermittently senses the onset of a spontaneous breath a mandatory
722
Mechanical Ventilation and Weaning
number of times per minute and immediately causes a machine (positive-pressure) breath to occur. Consequently, the machine breath overrides the spontaneous breath, and stacking of breaths cannot take place. Mechanical ventilation is initiated with a tidal volume of 12 ml/kg of lean body weight, an IMV respiratory rate so that the arterial carbon dioxide tension (PaC02) is 40 mm Hg (which usually requires an initial rate of 8 to 12 breaths/min), and an inspired oxygen fraction (F,02) of 1.0 (Table 20-1). Although patients may breathe spontaneously during IMV (which is especially helpful with regard to lowering pleural pressure and maintaining cardiac output), IMV should be regarded as initially providing total and full ventilatory support (i.e., the IMV breaths, not the spontaneous breaths, determine the minute ventilation)."·12 One hundred per cent oxygen should initially be used in order to update and document the level of rightto-left shunting and alveolar-arterial oxygen tension difference while minimizing the risk of unexpected hypoxemia. Administration of 100 per cent oxygen for a short period of time is not toxic (see later). A tidal volume of 12 ml/kg is considered large tidal volume ventilation. Large tidal volume ventilation is desirable because it prevents the development of miliary atelectasis or microatelectasis that is associated with constant small tidal volume ventilation (for example, 6 ml/kg).13 The miliary atelectasis or microatelectasis that is associated with small tidal volume ventilation results in a
Table 20-1
INITIAL VENTILATOR SETTINGS
1. IMV or SIMV mode 2. Tidal volume 12 ml/kg 3. Respiratory rate so that P a C0 2 = 40 mm Hg (usually 8 to 12 breaths/min) 4. F,0 2 = 1.0 5. I:E = 1:2 to 3 6. Inspiratory flow rate 30 L/min Abbreviations: IMV = intermittent mandatory ventilation; SIMV = synchronized IMV; I:E = inspiratiomexpiration ratio.
progressive increase in the alveolar-arterial oxygen tension difference and a progressive decrease in compliance. These deleterious changes associated with small tidal volume ventilation can be reversed by intermittent sighs. Large tidal volume ventilation, therefore, functions as frequent intermittent sighs and, thereby, minimizes the development of atelectasis. However, tidal volumes in excess of 15 ml/kg may cause the production of a permeability (high-protein) pulmonary edema caused by stretching of extra- and intra-alveolar vessels (causing a decreased extra-alveolar vessel perimicrovascular pressure)14-16 and alveolar epithelium17-19 (Fig. 20-1). A small tidal volume should be used in patients with bullous or cystic disease of the lung in an effort to minimize peak inspiratory pressures and risk of tension pneumothorax. The same considerations apply to patients with severe airflow obstruction (i.e., low tidal volume minimizes air trapping, hyperinflation, barotrauma).20
Figure 20-1 Scanning electron micrograph of lung capillaries and epithelium ruptured at high lung volume. Arrows show edges of epithelial breaks. Photomicrographs supplied by Drs. John B. West and Odile Mathieu-Costello. (From Parker JC, Hernandez LA, Peevy KJ: Mechanisms of ventilator-induced lung injury. Crit Care Med 21:131-143, 1993. Used with permission.)
Mechanical Ventilation and Weaning
There is no evidence that the type of inspiratory flow pattern (square wave, sinusoidal, accelerating, decelerating) makes any difference in any important respiratory or hemodynamic parameter.5 However, the physics of lung time constants mandate that the greater the time interval from gas delivery to end inspiration, the more complete the distribution of gas to areas of low ventilation/perfusion. Thus, an inspiratory hold is usually desirable and, with a ventilator rate of 10 breaths/min (6-sec frequency) and an inspiratory delivery time of 1 sec, an inflation hold of up to a 1 sec may be given while still maintaining an inspiratiomexpiration (I:E) ratio of 2:4 or 1:2. However, in patients with severe airflow obstruction (perhaps manifested by hyperinflation- [diagnosis made by chest X-ray and/or measurement of intrinsic PEEP] induced circulatory depression), I:E ratios of 1:3 or 1:4 are more appropriate.20 The respiratory rate should be set so that the patient has a normal arterial carbon dioxide tension (this rate is usually 8 to 12 breaths/min). Certainly, hypercarbia should be avoided in the vast majority of patients since the accompanying acidosis is dangerous, and the hypoventilation necessary to produce hypercarbia may also cause hypoxemia. Mild degrees of hypercarbia (P a C0 2 40 to 50 mm Hg) are indicated for patients who are chronic carbon dioxide retainers; mild hypercarbia helps to avoid changing the arterial and cerebrospinal fluid pH and bicarbonate levels from the values that the patient normally lives with. Hypocarbia should be avoided for multiple reasons. First, hypocarbia may cause a decrease in cardiac output via several mechanisms. In order to produce hypocarbia, an increased amount of positive-pressure ventilation will have to be used, which will decrease venous return. Hypocarbia also causes a withdrawal of sympathetic tone, which will decrease the inotropic state of the heart. Hypocarbia and alkalosis may decrease the level of ionized calcium, which will also decrease the inotropic state of the heart.21 Second, hypocarbia will also shift the oxygenhemoglobin dissociation curve to the left, which increases the hemoglobin affinity for oxygen; at the tissue level, the increased hemoglobin-oxygen binding impedes the delivery of oxygen to the tissues. Third, hypocarbia and alkalosis will increase oxygen consumption. The increase in oxygen consumption is due to a pH-mediated uncoupling of oxidation from phosphorylation; the uncoupling requires more oxygen to be passed along the respiratory chain to generate the same amount of adenosine triphosphate.22· 23 Quantitatively, at a P a C0 2 of 40 mm Hg, oxygen consumption is 250 ml/min, whereas at a P a C0 2 of 20 mm Hg, oxygen
723
consumption is 500 ml/min. Thus, hypocarbia may increase tissue oxygen demand (as a result of increased oxygen consumption) at a time when it is not possible to increase tissue oxygen supply (oxygen-hemoglobin dissociation curve is shifted to the left and cardiac output is decreased). Fourth, hypocarbia may cause ventilation-perfusion abnormalities by inhibiting hypoxic pulmonary vasoconstriction (HPV) and/or by causing bronchoconstriction and decreased lung compliance. Fifth, hyperventilation with alkalosis may cause ventricular irritability (perhaps due, in part, to a concomitant hypokalemia). However, the benefit of hyperventilation with hypocarbia to decrease an elevated intracranial pressure in headinjured patients may supersede the previously cited physiologic disadvantages. Fast respiratory rates should be avoided in patients with severe airflow obstruction (i.e., avoid hyperinflation).20 If the measured P a C0 2 differs from the desired level, one can easily calculate the amount of change in minute ventilation needed to produce the desired change in P a C0 2 by using the relationship: VE2 = (P a C0 2 '/P a C0 2 2 ) X VE> where VE2 = the desired minute volume, VE' = the actual minute volume, P a C0 2 2 = the desired P a C0 2 , and P a C0 2 ' = the actual P a C0 2 . Use of total minute ventilation, which includes both alveolar ventilation and dead space ventilation, instead of using only alveolar minute ventilation is clinically acceptable since the dead space remains relatively constant with moderate changes in tidal volume.24 A change in either respiratory rate or tidal volume, or both, can be used to change minute ventilation. The use of large tidal volume ventilation usually makes it possible to achieve normocarbia with relatively slow respiratory rates (8 to 12 breaths/min). This is highly desirable in two ways. First, slow rates allow a decrease in the (I:E) ratio. A long expiratory time increases venous return and augments cardiac output; it facilitates full expiration and avoids air trapping and obstructive lung disease. The I:E ratio should be approximately 1:2 or 3 (except when airway resistance is high and an even longer exhalation period and lower I:E ratio are required). Second, slow respiratory rates also allow slower inspiratory flow while maintaining an appropriately low I:E ratio. The slower the inspiratory flow, the more laminar and less turbulent is the air stream. This decreases airway resistance, reduces airway trauma, and allows for a more even distribution of inspired gas to lung regions with different time constants. The inspiratory flow should be approximately 30 L/min.
724
Mechanical Ventilation and Weaning
III. GOAL #1:|F,Q 2
