Postoperative cardiopulmonary problems are responsible

Transcription

1 Is Cardiopulmonary Exercise Testing a Useful Test Before Esophagectomy? Matthew J. Forshaw, FRCS, Dirk C. Strauss, FCS (SA), Andrew R. Davies, MRCS, David Wilson, BS, Boris Lams, MRCP, Adrian Pearce, FRCA, Abraham J. Botha, FRCS, and Robert C. Mason, FRCS(Ed) Departments of General Surgery and Anaesthesia, St. Thomas Hospital, and Department of Respiratory Medicine, Guy s Hospital, Guy s and St. Thomas NHS Foundation Trust, London, United Kingdom Background. Cardiopulmonary exercise (CPX) testing may identify patients at high risk of postoperative cardiopulmonary morbidity and mortality. This study aims to assess the utility of CPX testing before esophagectomy. Methods. Between January 2004 and October 2006, 78 consecutive patients (64 men) with a median age of 65 years (range, 40 to 81 years) underwent CPX testing before esophagectomy (50% transhiatal; 50% transthoracic). Measured variables included anaerobic threshold (AT) and maximum oxygen uptake at peak exercise (V O2 peak). Outcome measures were postoperative morbidity and mortality, length of hospital stay, and unplanned intensive therapy unit admission. Results. Cardiopulmonary complications occurred in 33 (42%) patients and noncardiopulmonary complications in 19 (24%). One in-hospital death (1.3%) occurred, and 13 patients (17%) required an unplanned intensive therapy unit admission. The level of V O2 peak was significantly lower in patients with postoperative cardiopulmonary morbidity (p 0.04). The area under a receiver operating characteristic curve was 0.63 (95% confidence interval [CI], 0.50 to 0.76) for the V O2 peak and 0.62 (95% CI, 0.49 to 0.75) for AT. An AT cutoff of 11 ml/kg/min was a poor predictor of postoperative cardiopulmonary morbidity. Conclusions. Although the V O2 peak was significantly lower in those patients who developed cardiopulmonary complications, CPX testing is of limited value in predicting postoperative cardiopulmonary morbidity in patients undergoing esophagectomy. (Ann Thorac Surg 2008;85:294 9) 2008 by The Society of Thoracic Surgeons Postoperative cardiopulmonary problems are responsible for a substantial proportion of the morbidity and mortality experienced by patients undergoing esophagectomy [1 3]. This morbidity can result in unplanned intensive therapy unit (ITU) admission and a prolonged hospital stay. Major surgery has been shown to place severe stress on a patient s cardiopulmonary reserve, requiring an increased oxygen demand of about 40% [4]. High-risk patients have traditionally been assessed using tests such as transthoracic echocardiography, dobutamine stress echocardiography, radionuclide ventriculography, and spirometry, although none of these have been validated as preoperative screening tests and most provide only static measures of cardiopulmonary performance [5 7].A patient s walking distance or ability to climb stairs has also been used as a subjective measure of exercise tolerance and has been shown to predict perioperative complications [8, 9]. The lack of objectivity and a failure to detect silent cardiopulmonary abnormalities remains a criticism of such subjective measures. A dynamic assessment of a patient s preoperative exercise capacity may therefore be a useful predictor of Accepted for publication May 22, Address correspondence to Mr Mason, Guy s and St. Thomas NHS Foundation Trust, Lambeth Palace Rd, London, SE1 7EH, United Kingdom; robert.mason@gstt.nhs.uk. postoperative morbidity and mortality. Cardiopulmonary exercise (CPX) testing measures oxygen uptake at increasing levels of work and objectively determines the cardiopulmonary performance under conditions of stress, thereby closely mimicking the postoperative situation. The results from CPX testing have been applied to elderly patients undergoing major abdominal and thoracic surgical procedures, and it has been shown that virtually all postoperative cardiopulmonary deaths occur in patients with an exercise anaerobic threshold (AT) of less than 11 ml/min/kg [10, 11]. The AT results from individual patients have been used to develop postoperative treatment strategies to minimize the risk of morbidity and mortality. CPX testing has similarly been used to assess whether borderline patients should undergo surgical procedures for lung cancer [12, 13]. Results from several studies have shown that CPX testing is simple to perform, noninvasive, applicable to both inpatients and outpatients, cost-effective, and has a low incidence of adverse events [10, 11]. To date, one study has examined the usefulness of CPX testing in patients undergoing esophagectomy [14]. This study from Japan showed that the maximum oxygen uptake (V o 2 max), but not the AT, correlated with postoperative cardiopulmonary complications after radical esophagectomy with a three-field lymphadenectomy. The purpose of our study was to assess the relationship 2008 by The Society of Thoracic Surgeons /08/$34.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg FORSHAW ET AL 2008;85:294 9 CARDIOPULMONARY EXERCISE TESTS 295 between values obtained from CPX testing and the incidence of postoperative cardiopulmonary complications in patients undergoing esophagectomy. Patients and Methods Patient Population Between January 2004 and October 2006, 78 consecutive patients (64 men, 14 women) with a median standard deviation age of 65 9 years (range, 40 to 81 years) underwent esophagectomy at our institution. Ethical committee approval was obtained for this study, and the need for individual patient consent was waived. The standard preoperative work-up included clinical evaluation, routine blood tests, chest roentgenogram, electrocardiogram (ECG), and echocardiogram. Pulmonary function tests, exercise ECG, stress echocardiogram, and thallium myocardial perfusion scans were performed selectively as determined by clinical evaluation. CPX testing was additionally performed in all 78 patients at the Respiratory Laboratory at Guy s Hospital. The underlying pathology in these 78 patients was adenocarcinoma in 58, squamous cell carcinoma in 13, adenosquamous carcinoma in 2, leiomyoma in 1, high-grade dysplasia in 3, and esophageal diverticulum in 1. According to the American Society of Anesthesiologists (ASA) classification [15], operative risk was scored as ASA-I in 1 patient, ASA-II in 54, or ASA-III in 23. A further 19 patients underwent CPX testing but did not undergo esophagectomy. This included 3 patients who were considered to be medically unfit on standard preoperative evaluation. CPX testing confirmed poor cardiopulmonary reserve (AT, 8.0 to 9.1 ml/min/kg; V o 2 peak, 11.2 to 13.9 ml/min/kg), and an operation was not offered to these patients. Seven patients declined operation or were advised to have nonsurgical treatments such as chemoradiotherapy. Further staging investigations suggested unresectable disease in 3 patients. Six patients were found to have unresectable tumors at laparoscopy or laparotomy, and no resection was performed. Fifty patients (64%) had received neoadjuvant chemotherapy, typically consisting of three cycles of combination epirubicin (50 mg/m 2 bolus day 1), cisplatin (60 mg/m 2 bolus day 1), and infusional 5-fluorouracil (200 mg/m 2 per day) given during a 3-week period, broadly following the Medical Research Council Adjuvant Gastric Infusional Chemotherapy (MAGIC) trial protocol [16]. The operative approach was determined by the location of the tumor and the individual surgeon s preference. The CPX values did not influence the operative approach. The types of operation performed included transhiatal esophagectomy in 39 patients, laparoscopicassisted two-stage esophagectomy in 23, two-stage esophagectomy in 6, three-stage esophagectomy in 5, and left thoracoabdominal esophagectomy in 5. Postoperatively, all patients were kept intubated and ventilated overnight in the recovery unit, followed by extubation and discharge to a surgical high-dependency unit the next day. Cardiopulmonary Exercise Tests All CPX tests were performed in the Respiratory Function Unit at Guy s Hospital with a doctor and full resuscitation equipment present. Between January 2004 and December 2005, the Medi-Soft Partn air 5400 system (MediSoft S.A., Dinant, Belguim) and Ergoline Ergometrics 900 cycle ergometer (Ergoline GmbH, Bitz, Germany) were used for the tests. From January 2006 onwards, the Med Graphics CPX Ultima system (Medical Graphics Corporation, St. Paul, MN) and Corvial cycle ergometer V2 (Lode BV, Groningen, The Netherlands) were used. A standard maximum incremental exercise test was used, consisting of a 3-minute resting period, followed by 3 minutes of unloaded cycling and then by incremental increases in work rate until the patient became symptomlimited and could not continue or the test was stopped for safety reasons. Table 1. Cardiopulmonary and Noncardiopulmonary Complications Occurring in 78 Patients Undergoing Esophagectomy Complication a THE (n 39) TTE (n 39) Mean AT b (ml/min/kg) Mean V o 2 Peak b (ml/min/kg) Cardiopulmonary complications c : Cardiac ischemia/infarction Arrhythmia Left ventricular dysfunction Respiratory failure requiring prolonged/reintubation Atelectasis/aspiration/pneumonia Pleural effusion Noncardiopulmonary complications: Sepsis Clinical anastomotic leak/conduit necrosis/chyle leak Reoperation a Some patients had more than one complication. b Per complication. c As defined by common terminology criteria for adverse events [20]. AT anaerobic threshold; THE transhiatal esophagectomy; TTE transthoracic esophagectomy; V o 2 maximum oxygen uptake at peak exercise.

3 296 FORSHAW ET AL Ann Thorac Surg CARDIOPULMONARY EXERCISE TESTS 2008;85:294 9 Table 2. Outcome Prediction Using Cardiopulmonary Exercise Variables in 78 Patients Undergoing Esophagectomy Outcome a CPX Variable Present Absent Difference (95% CI) Cardiopulmonary complications: V o 2 peak (ml/min/kg) 19.2 (5.1) 21.4 (4.8) 2.3 ( 0.06 to 4.5) 0.04 V o 2 peak % predicted 80.6 (22.9) 87.6 (20.2) 7.1 ( 2.7 to 16.8) 0.15 AT (ml/min/kg) b 13.2 (3.1) 14.4 (2.6) 1.2 ( 0.09 to 2.6) 0.07 Noncardiopulmonary complications: V o 2 peak (ml/min/kg) 20.4 (5.2) 20.7 (4.3) 0.27 ( 2.4 to 2.9) 0.84 V o 2 peak % predicted 86.4 (22.1) 84.1 (21.4) 2.3 ( 9.1 to 13.6) 0.69 AT (ml/min/kg) b 14.1 (3.0) 13.9 (2.9) 0.22 ( 1.3 to 1.8) 0.77 Unplanned ITU admission: V o 2 peak (ml/min/kg) 18.9 (5.1) 20.8 (5.0) 1.9 ( 1.1 to 4.9) 0.21 V o 2 peak % predicted 78.9 (25.2) 85.7 (20.7) 6.9 ( 6.2 to 19.9) 0.30 AT (ml/min/kg) b 12.6 (3.2) 14.2 (2.8) 1.6 ( 0.12 to 3.3) 0.07 p Value a Values presented as mean (SD). b Based upon 75 patients (3 patients with indeterminate ATs excluded). AT anaerobic threshold; CI confidence interval; CPX cardiopulmonary exercise; ITU intensive therapy unit; V o 2 maximum oxygen uptake at peak exercise. Breath-by-breath gas analysis was performed throughout all stages of the test, enabling the measurement of ventilation, oxygen uptake (V o 2 ), and carbon dioxide production. Oxygen saturations and 12-lead ECG were also monitored throughout, with blood pressure responses being measured at intervals. The gas analysis obtained during the CPX test was used to calculate values for V o 2 peak and AT. The AT was calculated using the V-slope method [17] and confirmed using the ventilatory equivalents method [18]. The percentage predicted V o 2 peak was calculated using standard formulas based on the patient s age, sex, weight, and height [19]. The AT was indeterminate in 3 patients who did not exercise for long enough to reach AT, or estimation of AT was impossible due to irregular breathing patterns. (ROC) curves were plotted to assess the predictive value of CPX measurements. Statistical analysis was performed with GraphPad Prism 4.0 (GraphPad Software, San Diego, CA). Results The mean anaerobic threshold and V o 2 peak in the 78 patients undergoing esophagectomy were and ml/min/kg, respectively. The mean percentage predicted V o 2 peak was 83.0% 21.2%. There was no significant difference in the mean AT and V o 2 peak according to operative approach: the transhiatal esophagectomy mean AT was ml/min/kg, and the mean V o 2 peak was ml/min/kg; the transtho- Outcome Measures Length of hospital stay, unplanned ITU admission (defined as the need for unplanned reintubation and mechanical ventilation), and the outcome of the operation, including mortality and morbidity, were recorded. Postoperative morbidity was divided into cardiopulmonary and noncardiopulmonary complications. Cardiac and pulmonary complications were defined according to the Common Terminology Criteria for Adverse Events [20]. Noncardiopulmonary complications included septic, anastomotic, and operative complications. Statistical Methods The CPX measurements were summarized as mean standard deviation for each outcome measure under study. Continuous variables were assessed by using the Student t test. Categoric variables were assessed by using the Fisher exact test. A value of p 0.05 was regarded as significant. The relationship between length of hospital stay and CPX measurements was assessed by using linear regression. Receiver operating characteristic Fig 1. Outcome prediction using anaerobic threshold measurements in 75 patients undergoing esophagectomy. (Horizontal lines refer to mean values, solid squares are cardiopulmonary events [CR], triangles are non-cr events, and open squares are admission to the intensive therapy unit [ITU].)

4 Ann Thorac Surg FORSHAW ET AL 2008;85:294 9 CARDIOPULMONARY EXERCISE TESTS 297 Fig 2. Outcome prediction using maximum oxygen uptake at peak exercise (V o 2 peak) measurements in 78 patients undergoing esophagectomy. (Horizontal lines refer to mean values, solid squares are cardiopulmonary events [CR], triangles are non-cr events, and open squares are admission to the intensive therapy unit [ITU].) racic esophagectomy mean AT was ml/min/ kg, and the mean V o 2 peak was ml/min/kg. In patients undergoing esophagectomy, cardiopulmonary complications occurred in 33 (42%) and noncardiopulmonary in 19 (24%; Table 1), and 13 patients (17%) required unplanned ITU admission. The mean length of hospital stay was days (range, 8 to 191 days). Performing a thoracotomy or receiving neoadjuvant chemotherapy did not produce any significant differences in the incidence of cardiopulmonary and noncardiopulmonary complications, unplanned ITU admission rates, or length of hospital stay. Only one in-hospital death (1.3%) occurred. The patient was a 74-year-old man with a medical history significant for pneumonectomy for lung cancer and a previous myocardial infarction (V o 2 peak, 20.1 ml/min/ kg; AT, 15.1 ml/min/kg). Respiratory failure developed after a transhiatal esophagectomy and the patient required prolonged ITU admission and respiratory support. He died from a myocardial infarction on day 44. Table 2 and Figures 1 and 2 summarize the differences in CPX variables according to the presence or absence of cardiopulmonary or noncardiopulmonary complications Table 3. Value of AT Cutoff Set at 11 ml/min/kg in 75 Patients Undergoing Esophagectomy Variable AT 11 ml/min/kg (n 12) AT 11 ml/min/kg (n 63) p Value Cardiopulmonary 7 (58%) 23 (37%) 0.20 complications Noncardiopulmonary 2 (17%) 15 (24%) 0.72 complications Unplanned ITU admission 3 (25%) 10 (16%) 0.42 LOS (days), mean SD AT anaerobic threshold; ITU intensive therapy unit; LOS length of stay; SD standard deviation. Fig 3. Receiver operator curve (ROC) to predict cardiopulmonary complications from measurements of maximum oxygen uptake at peak exercise (V o 2 peak) and anaerobic threshold (AT). The area under the ROC curves for both V o 2 peak; (squares) peak and AT (diamonds) were 0.63 (95% confidence interval, 0.50 to 0.76; p 0.02) and 0.62 (95% confidence interval, 0.49 to 0.75; p 0.03) respectively, both of which were significantly greater than 0.5. The diagonal line indicates no discrimination. and unplanned ITU admission. The V o 2 peak was significantly lower (p 0.04), and there was a trend towards a lower AT (p 0.07) in those patients with cardiopulmo- Table 4. Positive and Negative Predictive Values for V o 2 Peak and Anaerobic Threshold Cutoff Value Cardiopulmonary Complication if CPX Variable Is Cutoff, (%) No Cardiopulmonary Complication if CPX Variable Is Cutoff, (%) V o 2 peak (ml/min/kg) 14 5/9 (56) 41/69 (59) 16 7/13 (54) 39/65 (60) 18 11/24 (45) 32/54 (59) 20 19/35 (54) 29/43 (67) 22 24/46 (52) 23/32 (72) AT (ml/min/kg) 10 4/6 (67) 42/69 (61) 11 7/12 (58) 39/63 (62) 12 10/17 (59) 37/58 (64) 13 15/28 (54) 31/47 (66) 14 20/39 (51) 25/36 (69) Data are presented as number of patients/total patients (%). AT anaerobic threshold; CPX cardiopulmonary exercise; V o 2 maximum oxygen uptake at peak exercise.

5 298 FORSHAW ET AL Ann Thorac Surg CARDIOPULMONARY EXERCISE TESTS 2008;85:294 9 nary complications compared with those without. There was also a trend toward a lower AT (p 0.07) in those with an unplanned ITU admission. No apparent correlation was evident between length of hospital stay and CPX measurements: r 2 values for AT and V o 2 peak were and 0.063, respectively. No significant increase was found in the incidence of cardiopulmonary or noncardiopulmonary complications, unplanned ITU admission rates, or length of hospital stay in patients with an AT of less than 11 ml/min/kg (Table 3). Figure 3 shows the ROC curves for V o 2 peak and AT as predictors of cardiopulmonary complications. The area under the ROC was 0.63 (95% confidence interval [CI] 0.50 to 0.76; p 0.02) for V o 2 peak and 0.62 (95% CI, 0.49 to 0.75; p 0.03) for AT, both of which were significantly greater than 0.5. Table 4 shows the positive and negative predictive probability of cardiopulmonary complications from CPX variables at a series of cutoffs. The positive and negative predictive values of both V o 2 peak and AT were unhelpful over a series of cutoffs. Comment This study investigated the relationship between preoperative CPX indicators and the development of postoperative complications, unplanned ITU admission, and length of stay in 78 patients undergoing esophagectomy. The variables derived from CPX testing included V o 2 peak, which is the maximum oxygen uptake at peak exercise. A related measurement, V o 2 max, has previously been found to be the most useful predictor of postoperative cardiopulmonary complications both in patients undergoing radical esophagectomy with threefield lymphadenectomy [14] and after surgical procedures for lung cancer [21 25]. Also assessed was the AT, which is defined as the point during exercise at which oxygen demand outstrips oxygen delivery and metabolism starts to become anaerobic. AT is a measure of the ability of the cardiopulmonary system to deliver adequate oxygen to tissues and has the advantage of being independent of patient motivation, does not require high levels of physical stress, and occurs well before the V o 2 peak [19]. The AT has been assessed predominantly in relation to major surgical procedures in elderly patients, allowing the development of an operative risk grading and treatment protocol [10, 11]. An AT cut off of 11 ml/min/kg is internationally recognized and is currently being used to select patients in enhanced recovery programs for colorectal surgery. To date, the AT has not been found to be useful in the assessment of cardiopulmonary fitness of patients undergoing exclusively esophagectomy [14]. We have demonstrated that there is a significantly reduced V o 2 peak (p 0.04) and a nonsignificant trend to a reduced AT (p 0.07) in those patients in whom postoperative cardiopulmonary complications develop after esophagectomy. We also found a nonsignificant trend to a reduced AT (p 0.07) in those patients who had an unplanned postoperative admission to the ITU. When AT and V o 2 peak were analyzed by ROC curves, both had an area under the curve that was significantly different from 0.5 (p 0.02 and p 0.03, respectively), indicating that they both had a predictive value in determining which patients would subsequently present with cardiopulmonary complications. The actual values for the area under the curve for both of these tests were small (0.63 and 0.62, respectively), however, suggesting that in this group of patients, CPX indicators did not perform well as clinically useful tests for predicting postoperative cardiopulmonary complications. Furthermore, having an AT of less than 11 ml/min/kg, a cutoff suggested by Older and colleagues [10, 11], did not predict which patients would subsequently present with postoperative cardiopulmonary complications or require an unplanned ITU admission or increased length of hospital stay. Why has this study failed to show a stronger correlation between the results of CPX testing and cardiopulmonary complications, as has been suggested by other studies? First, with only one in-hospital death (1.3%), mortality could not be used as an outcome measure in this study. Although we used standardized criteria for reporting cardiopulmonary complications, many studies have not defined their criteria used for reporting postoperative complications. Second, a much smaller proportion of patients (16%) with an AT of less than 11 ml/min/kg underwent operation in the current study compared with previous studies, such as 29% in the one by Older and colleagues [10]. This suggests that higher-risk patients were not considered for resection by the surgeons and that the surgical cohort was therefore by definition a fitter group. However, there was no evidence of bias towards a particular operative approach on the basis of CPX test results. Formal measures of preoperative risk analysis have previously been shown to significantly reduce mortality rates from esophagectomy [26, 27]. Third, patients with primary cardiopulmonary complications, which should hopefully have been detected by CPX testing, were not separated from patients with cardiopulmonary complications secondary to noncardiopulmonary complications such as anastomotic leakage, sepsis, and multiorgan failure, which could not have been predicted by CPX testing. Fourth, all patients at our institution were kept intubated and ventilated in a level 3 area (equivalent to ITU) for the first night after their esophagectomy. In the study by Older and colleagues [10], caring for high-risk patients (identified though having an AT 11 ml/min/kg) in a level 3 environment reduced their mortality compared with controls cared for in a regular surgical ward. Thus, any increased risk in our patients with an AT of less than 11 ml/min/kg may have been masked by postoperative fluid optimization and respiratory support occurring in all patients in the level 3 environment, thus protecting them against developing complications. In conclusion, we have demonstrated that although there was a reduction in measured CPX indicators in those patients who presented with postoperative cardiopulmonary complications after esophagectomy, these same indicators did not perform well as tests for predict-

6 Ann Thorac Surg FORSHAW ET AL 2008;85:294 9 CARDIOPULMONARY EXERCISE TESTS 299 ing the development of cardiopulmonary complications when assessed by using a ROC curve. The results generated from this study merit further investigation in a larger population, multicenter study with mortality or failure to rescue rates as the primary end point. This will allow definitive conclusions to be drawn on the role of CPX testing in esophageal surgery. We wish to thank Mr Aadil Khan for his invaluable biostatistical advice. References 1. Scottish Intercollegiate Guidelines Network. Management of oesophageal and gastric cancer. Edinburgh, UK: NHS Quality Improvement Scotland, McCulloch P, Ward J, Tekkis PP. Mortality and morbidity in gastro-oesophageal cancer surgery: initial results of ASCOT multicentre prospective cohort study. BMJ 2003;327: Bailey SH, Bull DA, Harpole DH, et al. Outcomes after esophagectomy: a ten-year prospective cohort. Ann Thorac Surg 2003;75: Older P, Smith R. Experience with the preoperative invasive measurement of haemodynamic, respiratory and renal function in 100 elderly patients scheduled for major abdominal surgery. Anaesth Intensive Care 1988;16: Halm EA, Browner WS, Tubau JF, Tateo IM, Mangano DT. Echocardiography for assessing cardiac risk in patients having noncardiac surgery. Study of Perioperative Ischemia Research Group. Ann Intern Med 1996;125: Dunselman PH, Kuntze CE, van Bruggen A, et al. Value of New York Heart Association classification, radionuclide ventriculography, and cardiopulmonary exercise tests for selection of patients for congestive heart failure studies. Am Heart J 1988;116: Mangano DT, London MJ, Tubau JF, et al. Dipyridamole thallium-201 scintigraphy as a preoperative screening test. A reexamination of its predictive potential. Study of Perioperative Ischemia Research Group. Circulation 1991;84: Reilly DF, McNeely MJ, Doerner D, et al. Self-reported exercise tolerance and the risk of serious perioperative complications. Arch Intern Med 1999;159: Girish M, Trayner E Jr, Dammann O, Pinto-Plata V, Celli B. Symptom-limited stair climbing as a predictor of postoperative cardiopulmonary complications after high-risk surgery. Chest 2001;120: Older P, Hall A, Hader R. Cardiopulmonary exercise testing as a screening test for perioperative management of major surgery in the elderly. Chest 1999;116: Older P, Smith R, Courtney P, Hone R. Preoperative evaluation of cardiac failure and ischemia in elderly patients by cardiopulmonary exercise testing. Chest 1993;104: BTS guidelines: Guidelines on the selection of patients with lung cancer for surgery. Thorax 2001;56: Beckles MA, Spiro SG, Colice GL, Rudd RM. The physiologic evaluation of patients with lung cancer being considered for resectional surgery. Chest 2003;123(1 suppl):105s 114S. 14. Nagamatsu Y, Shima I, Yamana H, Fujita H, Shirouzu K, Ishitake T. Preoperative evaluation of cardiopulmonary reserve with the use of expired gas analysis during exercise testing in patients with squamous cell carcinoma of the thoracic esophagus. J Thorac Cardiovasc Surg 2001;121: Owens WD, Felts JA, Spitznagel EL Jr. ASA physical status classifications. A study of consistency of ratings. Anesthesiology 1978;49: Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med 2006;355: Beaver WL, Wasserman K, Whipp BJ. A new method of detecting anaerobic threshold by gas exchange. J Appl Physiol 1986;60: American Thoracic Society, American College of Chest Physicians. ATS/ACCP statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med 2003;167: Weisman IM. Cardiopulmonary exercise testing in the preoperative assessment for lung resection surgery. Semin Thorac Cardiovasc Surg 2001;13: NCI. Common Terminology Criteria for Adverse Events v 3.0 (CTCAE). Available at forms/ctcaev3.pdf. Accessed May 16, Win T, Jackson A, Sharples L, et al. Cardiopulmonary exercise tests and lung cancer surgical outcome. Chest 2005;127: Nagamatsu Y, Terazaki Y, Muta F, Yamana H, Shirouzu K, Ishitake T. Expired gas analysis during exercise testing pre-pneumonectomy. Surg Today 2005;35: Smith TP, Kinasewitz GT, Tucker WY, Spillers WP, George RB. Exercise capacity as a predictor of post-thoracotomy morbidity. Am Rev Respir Dis 1984;129: Bechard D, Wetstein L. Assessment of exercise oxygen consumption as preoperative criterion for lung resection. Ann Thorac Surg 1987;44: Bolliger CT, Soler M, Stulz P, et al. Evaluation of high-risk lung resection candidates: pulmonary haemodynamics versus exercise testing. A series of five patients. Respiration 1994;61: Zhang GH, Fujita H, Yamana H, Kakegawa T. A prediction of hospital mortality after surgical treatment for esophageal cancer. Surg Today 1994;24: Bartels H, Stein HJ, Siewert JR. Preoperative risk analysis and postoperative mortality of oesophagectomy for resectable oesophageal cancer. Br J Surg 1998;85:840 4.