Preoperative Assessment Of The Thoracic Surgical Patient

PREOPERATIVE ASSESSMENT OF THE THORACIC SURGICAL PATIENT

Preoperative evaluation of patients who are candidates for thoracic surgery is a complex process that is essential in fulfilling a variety of objectives. The surgeon requires such assessment lo plan the operative approach, anticipate potential operative and postoperative complications, decide on the necessary level of postoperative care, and determine what resources might be required to support the patient until full recovery takes place. The patient requires such assessment so that he or she can ask relevant questions about the recommended procedure, gain an understanding of the short- and long-term consequences of having surgery, and make an informed decision about whether to proceed. The preoperative evaluation of candidates for thoracic surgery is an art as much as it is a science. Despite the plethora of noninvasive and invasive tests that is available for assessing operative risks and predicting outcomes, the final decision ultimately is based on the surgeon's impression of the likelihood of success of the planned operation. Success can be identified in a number of ways, such as absence of complications, survival until hospital discharge, correction of an underlying disorder, cure of a cancer, or improved long-term quality of life (QOL). This chapter focuses on the physiologic evaluation of patients and on the associations among surgery and perioperative complications, operative mortality, long-term survival, and postoperative QOL.

GENERAL STATUS Age Given the continued growth of the advanced age sector of the population, is it no surprise that surgeons are being referred a higher percentage of elderly patients for consideration for surgery. Seventy-five years was once considered a prohibitive age for aggressive intervention for intrathoracic problems, but it is now commonplace to recommend major surgery to such patients. In 1975, the average U.S. white male barely lived into his early 70s; in 2005, the life expectancy of a 75-year-old U.S. white male was more than 10 years. The realization that the aging population needs and desires continued aggressive surgical care for selected problems has resulted in a substantial increase in the percentage of elderly patients in an overall surgical practice. For example, in 2001, the percentage of patients older than 70 years of age undergoing major lung resection was in excess of 43%,' a 25% increase over the percentage of elderly in such a cohort only 2 decades earlier.2 In the 1970s and 1980s, advanced age was associated with a substantial increase in morbidity and mortality from thoracic surgery.2 In most reports, age

continues to be an important and independent determinant of operative mortality and morbidity for lung resection, although the relative increased risk of surgery-related death associated with advanced age has substantially decreased owing to improvements in patient selection and surgical and postoperative management (Berrisford et al, 2005; Ferguson et al, 1995}.3,4 In fact, some reports suggest that age is no longer an independent determinant of operative mortality.5,6 In contrast, advanced age universally remains an independent and strong factor associated with increased risk of mortality and morbidity after esophagectomy (Atkins et al, 2004).7-9 Advanced age by itself is not an absolute contraindication to major thoracic surgery. For example, disease-specific survival is unrelated to the patient's age at the time of resection for lung cancer.10 However, a patient's age must be considered carefully in deciding on major surgical intervention, particularly in light of other comorbid conditions. Age interacts with other factors to increase the risk of operative morbidity and mortality. For example, diffusing capacity and age have been shown to be independent predictors of morbidity and mortality after major lung resection.3 Whereas the presence of a high risk value for only one parameter moderately increases operative risk, mortality increases exponentially if both parameters are in the high risk zone. Similarly, combined increased risk values for age and renal function, or for age and cardiovascular function, substantially elevate the risk of postoperative morbidity and mortality.11 For this reason, never consider such values independently; rather, evaluate them collectively in the overall context of a patient's medical condition.

Performance Status Performance status is a general measure of a patient's overall ability to participate in activities of daily life. It is useful to routinely assess performance status as part of the overview of a patient's physiologic and psychological condition. Several scales have been developed for scoring performance status that are easy to use and are reproducible. The most commonly used scales are the Karnofsky score and the Zubrod (Eastern Cooperative Oncology Group [ECOG]] scale (Table 2-1). In the absence of specific risk factors, patients with an ECOG score of 0 to 1 or a Karnofsky score of 80% to 100% have a normal risk of complications and mortality after major thoracic surgery. Progressively worse performance status levels are associated with incremental operative risk.Performance status has been shown in a few studies to be an independent determinant of operative

outcomes. For example, mortality after esophagectomy has been shown to be predicted by age and performance status.7 Similarly, poor performance status is associated with an increase in the risk of operative mortality after resection for lung cancer in elderly patients.12 However, most studies assessing operative risk associated with thoracic surgery have not specifically evaluated performance status as a potential risk factor. In addition, specific risk factors that contribute to poor performance status are more likely to be statistically linked to adverse outcomes than is performance status itself.

PULMONARY FUNCTION A general assessment of pulmonary function is appropriate in every patient undergoing thoracic surgery. The risk of pulmonary complications after major thoracic surgery is as high as 25%, and preoperative pulmonary function is an important predictor of such complications. Many patients who are candidates for thoracic surgery have had extensive exposure to tobacco smoke, putting them at high risk for emphysema and other forms of chronic obstructive lung disease. Assess the patient's smoking status during the initial evaluation, and provide smoking cessation advice as part of the initial encounter. It is often appropriate to remind patients of the substantial increase in risk of pulmonary complications for those patients who are unable or unwilling to stop smoking before major thoracic surgery. Initial screening consists of taking a history focused on the patient's respiratory status, including symptoms such as shortness of breath, dyspnea on exertion, the presence of a cough, whether the cough is productive, hemoptysis, and limitations in exercise capacity related to breathlessness. Additional informal evaluation in an outpatient clinic setting might include measurement of oxygen saturation during exercise, such as walking for a measured distance on flat ground or climbing a specified number of stairs. Failure to maintain adequate oxygen saturation during such maneuvers may indicate the need for more formal testing of pulmonary function.'3,14 Formal pulmonary function testing is appropriate in patients undergoing certain types of thoracic surgery in whom surgical recommendations would be altered based on the results of such testing. The finding of poor spirometry values may not influence the decision to perform limited wedge resection for diagnosis of diffuse pulmonary disease or thoracoscopic excision of a small, peripheral lung nodule. In contrast, elective major lung resection should virtually always be preceded by a formal assessment of pulmonary function to help determine operative risks and enable the surgeon to hold an informed discussion with the patient. Assessment of pulmonary function is also appropriate in many instances for preoperative evaluation before nonpulmonary surgery. For example, the risk of pulmonary complications is predicted by spirometry in patients undergoing

esophagectomy,15"17 and operative mortality after esophagectomy may similarly be related to preoperative pulmonary disease.18

Risk Factors Specific risk factors for major thoracic surgery related to pulmonary function include chronic pulmonary disease (emphysema, chronic bronchitis, asthma} and any condition that limits lung volume, including a large pleural effusion, a large diaphragmatic hernia, and prior major lung resection. Interstitial lung disease that interferes with gas exchange may be associated with hypoxia. Induction chemotherapy and radiotherapy result in measurable decrements in lung function. Similarly, distant prior radiotherapy to the lung or mediastinum can cause considerable impairment of pulmonary function as well as decreasing chest wall mobility and limiting mediastinal motion. In addition to these conditions, many of which cause chronic changes in lung function, performance of a thoracotomy has acute detrimental effects on spirometry that persist for up to 8 to 12 weeks postoperatively. Functional residual capacity drops by 35% on the first postoperative day. Sixty percent decreases in forced vital capacity (FVC) and forced expiratory volume in 1 second (FEVi) also occur during this period.19,20 The addition of a major lung resection substantially further decreases spirometric values and gas exchange parameters to an extent directly correlated with the amount of functional lung tissue that is resected. Furthermore, these reductions persist because of permanent loss of lung volume and are sometimes associated with impaired exercise capacity, particularly in patients who have undergone pneumonectomy. At 6 to 12 months postoperatively, patients who have undergone lobectomy have a 5% to 15% reduction in FVC and a 10% to 25% reduction in FEV|. Corresponding values for pneumonectomy are a 35% to 40% reduction in FVC and a 35% to 50% decrease in FEVi.21,22 Interestingly, in highly selected patients with severe heterogeneous emphysema who undergo major lung resection, it is possible to demonstrate an improvement in spirometric functionthat is similar to that seen in patients undergoing lung volume reduction surgery (LVRS) (Baldi et al, 2005).23"25 Standard calculation of expected postoperative function in such patients may substantially underestimate their actual postoperative function. In order to properly select patients who can be shepherded through the acute recovery period after major lung resection and some other types of thoracic surgery, a careful preoperative assessment of lung function and estimation of expected postoperative function is essential in the evaluation of the lung resection candidate.

Spirometry Spirometry has been used to assess operative risk in

Section 1 Introduction

lung resection candidates for more than 5 decades. FVC was initially used to assess risk, and subsequently FEV] was considered the optimal parameter for assessing the likelihood of postoperative respiratory complications (Table 2-2). Calculation of a predicted postoperative value for FEVi (ppoFEVi) has proved to be very useful in estimating a patient's postoperative risk.26 Patients with normal risk have a ppoFEVi of 800 to 1000 mL or greater. Maximum voluntary ventilation (MVV) has also been used as a measure of risk associated with major lung resection; patients with an MW less than 50% of predicted are at increased risk for postoperative complications after major lung resection. However, this parameter is strongly dependent on patient effort and therefore is subject to tremendous variability. Traditional cutoff values for FVC and FEV! that are used to differentiate between low and high risk for major pulmonary resection are relatively inaccurate at the extremes of the body mass spectrum. In consideration of this fact, spirometric values expressed as a percentage of the predicted value based on age, gender, and height have more commonly been used to assess operative risk.13,26,27 In general, patients with a preoperative FEV| of at least 60% of predicted have a normal risk profile for major lung resection excluding pneumonectomy. Further refinement has included calculation of a ppoFEVi; values of 40% or greater are generally thought to indicate normal operative risk for major lung resection. The calculation of predicted postoperative values is sometimes challenging. In patients with normal lung function (who do not often need major lung resection), the simplest method is to multiply the preoperative spirometric value by the fraction of functional lung segments expected to remain postoperatively. For example, assuming 19 functional lung segments, a patient who is undergoing right upper lobectomy (losing 3 segments) would be expected to retain 16/19 of original lung function. Another simple method of estimation is to subtract 5% from original lung function for each functioning segment that is to be removed. The calculation becomes more important in patients with marginal lung function, especially those who have areas of functional heterogeneity, and in patients who have undergone prior lung resection. Lung segments that are obstructed are eliminated from calculations in order to more accurately assess predicted postoperative lung function. Lobes that are affected by emphysema to a greater extent than the remaining lung are not considered fully functional for purposes of calculating estimated postoperative function. Several techniques are available that enable refinement of the calculation of estimated postoperative function. Quantitative pulmonary scintigraphy, using the perfusion phase of the examination as the best estimate of regional function, effectively estimates regional lung function assessed per quadrant or per lung. A newer method,

quantitative computed tomography (CT], provides similar or greater accuracy through measurement of relative lung density as an estimate of pulmonary vasculature (Bolliger et al, 2002).28 By using one or more of these techniques for estimating regional lung function, and thus the amount of functional lung expected to remain after major lung resection, one can calculate a ppoFEVj that closely parallels the measured postoperative function. In addition to the utility of spirometry in estimating postoperative risk after major lung resection, it is also effective in predicting the risk of pulmonary complications after esophagectomy.17,29,30 Pulmonary complications are more than four times more likely to occur in patients with abnormal spirometry results than in those with normal spirometry.31 These findings do not suggest that spirometry be performed in all patients undergoing esophagectomy. Rather, spirometry may be appropriate to perform in patients who have clinical evidence of underlying lung dysfunction as a means to estimate the risk of postoperative pulmonary complications. If that risk is high, interventions such as preoperative cardiopulmonary rehabilitation may be appropriate, and a more accurate informed discussion can take place with the patient.

Diffusing Capacity Until the late 1980s, the only reliable method of assessing lung function as a means for predicting complications in patients undergoing thoracic surgery was spirometry. The measured and postoperative estimated values failed to predict most pulmonary complications and postoperative mortality, particularly in patients undergoing major lung resection. Subsequent studies identified diffusing capacity as an independent and important predictor of incremental risk of postoperative pulmonary morbidity and overall mortality after major lung resection.32"3,1 The highest risk group initially was identified as having a preoperative carbon monoxide diffusing capacity (DLCO) of less than 60% of predicted. The identification of high-risk patients is more accurately achieved by calculating the ppoDLCO based on the amount of lung to be resected; the highest risk group includes those patients with ppoDLCO less than 40% of predicted.3 In addition to predicting perioperative complications, DLCO also predicts long-term outcomes after major lung resection. Patients with a preoperative DLCO less than 50% of predicted who underwent lobectomy or less than 60% of predicted who underwent pneumonectomy had a worse QOL, an increased need for supplemental oxygen, and a greater frequency of hospital readmission during the first postoperative year after resection, compared to patients with normal DLCO.35 DLCO is also an important predictor of outcomes in patients undergoing LVRS for emphysema. DLCO is

one of the components that helps identify patients who belong to the so-called prohibitive risk category

for LVRS, which is characterized by an FEV! of less than 20% of predicted and either a diffusing capacity less than 20% of predicted or homogenous distribution of emphysema.36 DLCO also predicts the likelihood of pulmonary morbidity after LVRS in the lower-risk groups.37 The data are sufficiently compelling that diffusing capacity be measured routinely in candidates for major lung resection or LVRS. In the absence of severe pulmonary dysfunction, DLCO assessment in patients undergoing lesser lung operations is of questionable value; DLCO measurement in patients with severely compromised lung function may assist the physician in having an informed discussion with the patient about potential risks and outcomes. In addition to its utility in assessing risk related to major lung resection and LVRS, the DLCO predicts the incremental risk of pulmonary complications in patients undergoing esophagectomy. In the predictive model that was developed from this analysis, patients with a DLCO less than 80% of predicted had a 1.7-fold increased risk of pulmonary complications, compared to patients with a DLCO of 100% of predicted or better/8 The predictive capacity of this value, although strong, is probably overshadowed by several other physiologic predictors in candidates for esophagectomy. Therefore, routine measurement of DLCO is not generally indicated in this patient population.

Exercise Capacity and Oxygen Consumption Another method of assessing operative risk for major lung resection is measurement of exercise capacity. This is accomplished with simple techniques such as the 6-minute walk distance, stair climbing ability, and assessment of arterial oxygen saturation (Pa0 2) during walking on flat ground or during stair climbing.39''10 Patients with a very limited ability to exercise and those who experience a substantial drop in Pa02 during exercise are considered to be at high risk for postoperative complications.41 These techniques are inexpensive and are reasonably reliable for estimating whether a patient's risk is normal or substantially increased. However, incremental risk is difficult to establish using these semiquantitative methods. It is often appropriate to further evaluate patients who are deemed to be at substantially increased risk for complications after major lung resection by measuring maximum oxygen consumption (Vo2max) during exercise. This technique is expensive and labor intensive, and its accuracy depends to some extent on the patient's willingness to exercise to capacity and on the ability of the physician who is supervising the; test to determine when the point of maximum exercise has been achieved. With these caveats in mind, the objective data that result from this test provide estimates of risk that are similar or greater in accuracy to those provided by more standard measurements such as spirometry and DLCO.42 The limiting value of Vb2max for prohibitive risk is

10mL/kg/ min; values in excess of 15 to 20 mL/kg/min are indicative of normal risk. Values between 10 and 15 mL/kg/min must be interpreted clinically because the risk level associated with this range of oxygen consumption is variable and often is not prohibitive. Efforts have been made to correlate risk with V02max expressed as a percentage of the predicted value; the results suggest that values less than 50% to 60% of predicted are indicative of much higher than average risk, although the accuracy of such predictions is poor at the extremes of the functional spectrum [Win et al, 2005).43'46 An algorithm for the stepwise pulmonary assessment of candidates for major lung resection is presented in Figure 2-1.

Lung Function and Long-Term Outcomes In addition to the immediate postoperative risk of morbidity and mortality after major thoracic surgery, long-term QOL and overall survival must be considered when making surgical recommendations to patients. The influence of pulmonary function on long-term outcomes has been best defined for patients undergoing major lung resection and often reflects processes that are characteristic of a general population. Impaired short- and intermediate-term QOL is related to reduced DLCO after major lung resection (Hardy et al, 2002).47,4S Spirometric values do not appear to have an important influence on QOL in this time frame. It has been known for centuries that life expectancy in the general population is inversely related to FVC, and insurance companies have recently begun to use spirometry as part of their actuarial analyses in setting life insurance rates. Similarly, long-term survival in patients with lung cancer is related to the severity of chronic obstructive pulmonary disease (LopezEncuentra et al, 2005).49 In patients who undergo major lung resection, long-term survival is inversely related to FEV1( with incremental mortality occurring as a result of intercurrent disease rather than recurrent cancer.50'53 In patients with severely impaired spirometry results, give careful consideration to the impact of major lung resection on QOL and long-term survival. Weight this factor against the relative risk of death from recurrent cancer based on the type of lung resection performed.

CARDIOVASCULAR STATUS Patients who have disease requiring major thoracic surgery frequently have risk factors for pulmonary disease, as described earlier, and many of those risk factors are also associated with cardiovascular disease. As part of the initial evaluation of such patients, a careful history and a thorough physical examination are vitally important in identifying problems that portend an increased risk of postoperative cardiovascular complications, including stroke, myocardial infarction, and arrhythmia. It is estimated that between one quarter and one third of patients undergoing general anesthesia have known

Section 1 Introduction

cardiac disease or known risk factors (Box 2-1) and that almost 5% of all patients will experience a postoperative cardiac complication.54 The risk of possible neurovascular and peripheral vascular complications is also substantial. In general, the risk of cardiovascular complications is much higher in patients undergoing major thoracic surgery than in those undergoing less stressful types of general surgical procedures.

Coronary Artery Disease Risk factors for postoperative coronary artery complications include ischemic heart disease, congestive heart failure, diabetes mellitus, renal insufficiency, and poor overall functional status (Fleisher and Eagle, 2001}.54 In the absence of any such risk factors, patients proceed directly to surgery without any specific evaluation of their coronary arterial anatomy. Patients who have unstable angina or recent myocardial infarction must undergo a thorough evaluation, and any elective surgery is postponed until such conditions are stabilized. An algorithm for managing patients with one or more risk factors is outlined in Figure 2-2. Specific testing is performed when the clinical situation indicates that changes in management would occur if the test returned positive, suggesting that the algorithm is cost-effective. Further testing is not performed if the results would not influence a patient's overall management strategy. The likelihood of perioperative complications in patients with these risk factors may be reduced through revascularization for coronary artery disease, including use of such techniques as angioplasty, stenting, and coronary artery bypass grafting (CABG). Stenting requires administration of antiplatelet agents, including aspirin and clopidogrel, for a period of at least 4 to 12 weeks after stent placement and aspirin indefinitely afterward. Performance of major surgery before the end of the 4- to 1 2-week period leads to unacceptable risks of bleeding if antithrombotic agents are not discontinued or to myocardial infarction in those patients in whom antithrombotic agents are stopped preoperatively. Some newer drug-eluting stents require intensive antithrombotic therapy for even longer periods before the risk of stent thrombosis is sufficiently small to permit discontinuation of these medications preoperatively. Both aspirin and clopidogrel must be discontinued for 5 to 7 days before major surgical intervention to reduce the risk of surgical bleeding. CABG before thoracic surgical intervention is appropriate in patients in whom important coronary artery disease is not amenable to percutaneous revascularization techniques. There is no specified interval that must be observed between successful coronary artery surgery and subsequent major thoracic surgery. The surgeon's clinical judgment about the patient's condition and ability to withstand further major surgery is the best means for determining suitable timing. It remains to be seen whether the routine use of minimally invasive approaches to off-

pump coronary artery bypass surgery can meaningfully decrease the necessary time interval between operations. Of note, there is rarely an imperative to perform a major thoracic procedure under the same anesthesia used for CABG. Extensive operations for lung resection are usually performed less thoroughly through a median sternotomy than they are through a transthoracic approach, which may potentially compromise the therapeutic efficacy of interventions for oncologic problems. In addition, manipulations of tumor tissue before or during periods when patients are on cardiopulmonary bypass theoretically increase the risk of bloodborne distant metastatic disease. Finally, the use of anticoagulation, which is frequently necessary for performing coronary artery bypass, increases the risk of bleeding from the thoracic surgical sites, and these sites may not be easy to identify or control if such bleeding occurs. Additional medical management suitable for most patients with at least one risk factor includes administration of |3-blockers in the perioperative period. The medication is begun 2 to 7 days preoperatively and is continued for at least 1 week postoperatively. The dose is titrated to reduce resting heart rate to about 60 beats per minute. Patients with risk factors for coronary artery disease treated within these guidelines experience a reduction of up to 90% in the incidence of myocardial infarction or cardiac death after major noncardiac surgery.55 ^-Blockers must be administered carefully in patients with important lung disease such as reactive airways disease or emphysema; however, use of highly selective P-blockers is appropriate in most patients in this group.

Risk Factors for Postoperative Arrhythmias Cardiac arrhythmias, particularly supraventricular arrhythmias, occur commonly after major thoracic surgery. Often they are transient, but frequently they are persistent and difficult to manage. In efforts to prevent such complications, which develop most frequently after pneumonectomy and esophagectomy, prophylactic regimens are sometimes recommended for patients at increased risk. Elevated risk is associated with advanced age, greater extent of lung resection, mediastinal surgery (thymus, mediastinal tumor, esophagectomy), and possibly a low DLCO (Vaporciyan et al, 2004).56,57 One regimen used after major lung resection that has been shown to reduce the risk of supraventricular arrhythmias (including atrial fibrillation) by 50% is diltiazem given intravenously (IV) on arrival in the postanesthetic care unit and continued thereafter IV or orally for a period of 2 weeks.58

Systemic Anticoagulation in the Perioperative Period Conditions requiring preoperative anticoagulation are

not uncommon among thoracic surgical patients. Anticoagulation is necessitated most commonly by acute conditions, including venous thrombosis and pulmonary embolism, and by chronic conditions such as recurrent venous thrombosis, a mechanical heart valve, or atrial fibrillation. In the setting of chronic conditions such as prior venous thrombosis, atrial fibrillation, or distant prior pulmonary embolism, anticoagulation is usually safely discontinued 1 week preoperatively and is resumed after the risk of postoperative bleeding is normal. In contrast, patients with mechanical artificial valves or more acute thrombotic problems require anticoagulation until the day of the operation. This is most easily achieved by using either IV heparin in an inpatient setting or enoxaparin injections until 8 to 12 hours before the planned incision time. Anticoagulation therapy is resumed as soon as the risk of bleeding is substantially reduced, typically not until the day after surgery.

PREOPERATIVE EVALUATION OF OTHER SYSTEMS Patients undergoing major thoracic surgery who have underlying diabetes mellitus are at increased risk for a variety of complications, including myocardial infarction (see earlier discussion), wound infection, bronchial stump leak, and a variety of other wound healing complications.59,60 In patients undergoing cardiac surgery, assiduous control of blood glucose levels perioperatively appears to improve overall outcomes.61,152 Similar benefits may occur in general thoracic surgical patients, although this has not yet been established. In any case, assessment of the increased risks associated with diabetes enables the surgeon to have an informed discussion . with the patient regarding surgical outcomes and to prepare necessary resources to permit optimal perioperative management. Impaired renal function poses important challenges during the preoperative evaluation of thoracic surgical patients. Use of contrast material as part of staging studies is often contra-indicated, reducing the accuracy of such studies and adding potential uncertainty to the outcome of any operation. Perioperative management in such patients requires a careful review of medications to be used, with appropriate dose reduction or altered dose scheduling based on the degree of functional renal impairment. For patients who are undergoing hemodialysis, arrangements must be made for this to be performed on the day before surgery, so that dialysis on the day of surgery is avoided, reducing the risk of bleeding associated with heparin needed for hemodialysis. Patients who are receiving peritoneal dialysis and who require a laparotomy must be converted for the short term to hemodialysis, usually through a temporary venous catheter rather than a shunt or fistula. The presence of renal failure is associated with poorer outcomes for most important general thoracic procedures, including major lung resection,63,6'' and it is appropriate that this be discussed as part of the

informed consent process. Hepatic insufficiency presents considerable challenges for performing thoracic surgery, including increased risks of bleeding from coagulopathy, hemorrhage from esophageal varices, hepatic encephalopathy, and uncontrollable ascites. Patients with suspected cirrhosis are evaluated according to standard systems such as the Child classification, which requires assessment of serum bilirubin and albumin, prothrombin time, degree of encephalopathy, and amount of ascites. Carefully selected patients in Child's group A or possibly group B may be candidates for major lung resection or esophagectomy, with the anticipation that their risks of operative complications are considerably increased.65 The finding of cirrhosis also portends a reduced long-term survival after potentially curative oncologic thoracic surgery because of the increased risk of death from intercurrent causes. General physical limitations sometimes become important in the preoperative evaluation of the thoracic surgery patient. Patients with lower extremity amputations (e.g., for sarcoma] sometimes develop a need for thoracotomy or sternotomy, often for resection of pulmonary metastases. Patients who cannot ambulate independently using a limb prosthesis must be assessed with regard to their ability to ambulate as part of their recovery from surgery. This may not be an important issue if a musclesparing thoracotomy is performed because this procedure preserves shoulder girdle musculature and function and does not affect the use of walking aids. However, it may be a complicating factor if a sternotomy or transverse sternothoracotomy is performed because ambulation using a walker or crutches places unusual stresses on the reapproximated sternum, possibly leading to dehiscence and infection or simple malunion. Airway issues affect any patient who requires lung isolation as part of a thoracic surgical procedure. Lung and esophageal cancers share common risk factors with head and neck cancer, and it is not uncommon for a patient to require surgery for more than one of these conditions over time. Patients who have undergone laryngectomy for head and neck cancer present unique challenges for obtaining lung isolation. This must be considered before major thoracic surgery is recommended. The collective effect of comorbidities has an important influence on both short-term and long-term outcomes after thoracic surgery. Higher comorbidity scores are associated with an increased risk of postoperative complications after major lung resection.66,67 In addition, elevated comorbidity scores are linked to increased long-term mortality after lung cancer resection.68,69

RISK ASSESSMENT An important focus of clinical research is the development of organized methods of assessing preoperative risks for surgical procedures. Risk assessment tools in thoracic surgery are in their

Section 1 Introduction

infancy, compared with the robust tools available for risk assessment in adult cardiac surgery. The use of such algorithms is potentially important in informing individual patients about risk levels, in determining the potential utility of preoperative interventions for lowering risk, and in assessing the need for enhanced resources during the postoperative care of such patients. Various scoring systems have been used to provide a reasonably accurate quantitative estimate of risk for patient populations undergoing major lung resection and other operations. These systems include the Physiological and Operative Severity Score for the Enumeration of Mortality and Morbidity (POSSUM), the Cardiopulmonary Risk Index (CPRI), the Predictive Respiratory Quotient (PRQ), the Predicted Postoperative Product (PPP), APACHE II (adapted from a trauma scoring system), the Estimation of Physiologic Ability and Surgical Stress (E-PASS), and evaluation of the three most important predictors of outcomes (expiratory volume, age, and diffusing capacity: EVAD).70"75 Unfortunately, such systems have not been adequately assessed with regard to their utility in estimating risk for individual patients. They are currently most useful for estimating outcomes in populations undergoing major lung resection that are stratified according to standard risk profiles.

DECISION MARKING PROCESS After completing the preoperative evaluation of candidates for thoracic surgery, the surgeon offers recommendations regarding possible operative intervention. In making such recommendations, the goals of the surgeon and of the patient must be explicitly expressed and assessed; they are not always similar, and in many cases they are quite disparate. Patients tend to follow their self-interest by seeking procedures and outcomes that minimize discomfort and optimize QOL; death as an outcome of surgery does not pose nearly as great a concern as does permanent postoperative disability (Cykert et al, 2000).'6 Surgeons, serving in part their own self-interest, tend to focus on minimizing postoperative complications and maximizing long-term survival, especially for oncologic conditions. Most risk factors described in this chapter are well understood and are indelibly etched into the minds of surgeons who deal with these patients on a daily basis. The scoring systems mentioned assess only short-term risk for groups of patients; none purports to assess risk for an individual patient. Potential methods to do this include the use of artificial intelligence software to train a neural network based on actual outcomes.77,78 As large quantities of data are entered, the neural network identifies risk patterns and modifies these patterns as new outcomes are included in the database. After sufficient learning and validation have taken place, the accuracy of neural network prediction of complications can exceed 95%. However, at present, a trained neural network is site specific, making its use feasible only in high-volume centers in which infrastructure is available to manage

the network. No current scoring or artificial learning systems can provide insight into long-term outcomes, including QOL and long-term survival. In fact, the necessary tools to measure QOL in the specific context of thoracic surgical procedures have not yet been devised or validated. Generic QOL tools have been applied to outcomes for lung resection, esophagectomy, and LVRS. Examples of tools that assess overall QOL include the Short-Form 36 (SF-36] derived from the Medical Outcomes Study, the Health Related Quality of Life Measure (HRQOL-14) of the Centers for Disease Control and Prevention, the Sickness Impact Profile, and the Nottingham Health Profile.79 There are numerous QOL measures for chronic lung disease, including those that measure baseline function, function during exercise, general fatigue, and responsiveness to interventions.80 No measure to date has sought to incorporate issues such as QOL during the postoperative recovery period, postoperative and chronic incisional pain, swallowing impairment after esophageal surgery, maintenance of normal body weight, or the impact of surgery on specific vocational and avocational activities. Until such measures are developed, surgeons and their patients will not have the ability to make truly informed decisions about the utility of surgery. Decision analysis models are being developed as methods to appropriately weigh risks and benefits for patients undergoing thoracic surgery. Some issues that have been assessed include the utility of various treatments (surgical and nonsurgical) for achalasia, whether to perform routine mediastinoscopy for staging of surgical candidates for lung cancer resection, and the choice between sleeve lobectomy or pneumonectomy for centrally located lung cancers.sl"83 Future possibilities for similar models include the selection of optimal therapy for medically marginal candidates for lung resection and esophagectomy. Such models, using data relevant to individual patients, may prove very useful in providing patient-specific risk estimates and guidelines for recommendations. Despite the promising work that is being done on risk analysis and decision-making algorithms, the evaluation of potential thoracic surgical patients currently remains an art that ultimately is dependent on the experience and judgment of the surgeon. The assessments outlined in this chapter provide useful algorithms for consideration of risks and outcomes in patient populations and in individual patients. Use of these algorithms must be tempered by the surgeon's knowledge of an individual patient's risks, needs, and desires. It is unlikely that this vital judgment function will ever be completely subsumed by technological advances.

COMMENTS AND CONTROVERSIES As pointed out by Doctor Ferguson, the preoperative evaluation of potential candidates for major thoracic procedures is a complex but important process that must be done for every

patient. One must try to establish a reliable patient profile (low

have been prevented if the problem had been identified

risk, high risk, prohibitive risk) to ensure that no individual is

preoperatively.

denied surgery while minimizing postoperative morbidity. Most importantly, the appreciation of such considerations allows the surgeon and other members of the team (anesthetist, intensivlst) to use a number of prophylactic measures intended to decrease morbidity in high-risk patients. Where indicated, pulmonary rehabilitation, smoking cessation, optimization of medical treatment of chronic obstructive pulmonary disease, and treatment of cardiac disease decrease the risks associated with pulmonary or esophageal resection. Although scoring systems are available to predict operative risk, numbers do not tell everything, and nothing replaces good clinical judgment. Moreover, none of those systems has been validated with large numbers of patients, and none provides insight into long-term outcomes such as QOL and cardiorespiratory function 5 years after surgery. In this excellent chapter by Doctor Ferguson, a number of risk factors for operative morbidity and mortality are analyzed. In general, advanced age (70 years or older) and comorbidities are intimately related and act as dependent variables in increasing the risk of postoperative events, especially In patients

undergoing pneumonectomy or esophagectomy.

Indeed, older patients are more likely to lose their ability to cooperate postoperatively (increased risk of delirium), a feature that may add significantly to the operative risk. No patient should have pulmonary surgery, no matter how limited, without preoperative pulmonary function testing. In many cases, a simple spirometric test provides enough information to determine that the pulmonary function is normal and that the patient can tolerate pneumonectomy if necessary. Exercise testing (measurements of Vo2max, Pao2, and Paco2, both at rest and during exercise) measures the ability of the whole organism to perform well because it assesses the interaction of pulmonary function, hemodynamic performance, and peripheral tissue oxygen use. A Paco2 that rises on minimal exercise, for example, is a strong indicator of inadequate reserve, and such patients must be looked at very carefully before surgery. Indeed, several authors have shown that exercise testing is the only objective measurement of cardiopulmonary significant

reserve

difference

postoperative

courses

to

demonstrate

between and

those

patients with

a

statistically

with

benign

cardiorespiratory

complications. For most patients undergoing thoracotomy, the greatest cardiac risk arises from the presence of coronary artery disease. Operations performed within 3 months after a myocardial infarction, for instance, result in a 27% incidence of recurrent infarction. This incidence decreases to about 15% if the infarction occurred 4 to 6 months previously and to 6% if the operation is delayed for 6 months or longer. Similar risks have been identified for patients with angina. For these reasons, an accurate cardiac history and evaluation are of utmost importance. A screening exercise test is recommended for all patients who are smokers and older than 45 years of age, and for those with significant other risk factors for coronary artery disease. Overall, it is important to remember that, in the practice of thoracic surgery, technical misadventures do occur but seldom account for significant postoperative morbidity. On the other hand, the majority of postoperative complications and deaths are related to cardiopulmonary events, most of which could

J. D.