Subsequent Pulmonary Resection for Bronchogenic Carcinoma After Pneumonectomy

Transcription

1 ORIGINAL ARTICLES: GENERAL THORACIC Subsequent Pulmonary Resection for Bronchogenic Carcinoma After Pneumonectomy Jessica S. Donington, MD, Daniel L. Miller, MD, Charles C. Rowland, BS, Claude Deschamps, MD, Mark S. Allen, MD, Victor F. Trastek, MD, and Peter C. Pairolero, MD Division of General Thoracic Surgery and the Section of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota Background. Patients who have undergone a pneumonectomy for bronchogenic carcinoma are at risk of cancer in the contralateral lung. Little information exists regarding the outcome of subsequent lung operation for lung cancer after pneumonectomy. Methods. The records of all patients who underwent lung resection after pneumonectomy for lung cancer from January 1980 through July 2001 were reviewed. Results. There were 24 patients (18 men and 6 women). Median age was 64 years (range, 43 to 84 years). Median preoperative forced expiratory volume in 1 second was 1.47 L (range, 0.66 to 2.55 L). Subsequent pulmonary resection was performed 2 to 213 months after pneumonectomy (median, 23 months). Wedge excision was performed in 20 patients, segmentectomy in 3, and lobectomy in 1. Diagnosis was a metachronous lung cancer in 14 patients and metastatic lung cancer in 10. Complications occurred in 11 patients (44.0%), and 2 died (operative mortality, 8.3%). Median hospitalization was 7 days (range, 2 to 72 days). Follow-up was complete in all patients and ranged between 6 and 140 months (median, 37 months). Overall 1-, 3-, and 5-year survivals were 87%, 61%, and 40%, respectively. Five-year survival of patients undergoing resection for a metachronous lung cancer (50%) was better than the survival of patients who underwent resection for metastatic cancer (14%; p 0.14). Five-year survival after a solitary wedge excision was 46% compared with 25% after a more extensive resection (p 0.54). Conclusions. Limited pulmonary resection of the contralateral lung after pneumonectomy is associated with acceptable morbidity and mortality. Long-term survival is possible, especially in patients with a metachronous cancer. Solitary wedge excision is the treatment of choice. (Ann Thorac Surg 2002;74:154 9) 2002 by The Society of Thoracic Surgeons Patients treated surgically for bronchogenic carcinoma have a 1% to 5% risk per year for development of a second primary lung cancer [1, 2] and a 2% to 5% risk per year of developing recurrent carcinoma, depending on the initial tumor stage [3, 4]. Surgical resection is the preferred treatment for patients who develop a new lung lesion. However, further resection is not always feasible. Difficulty arises when the initial resection was a pneumonectomy. Many physicians consider resection of a lesion in the remaining lung contraindicated. Despite a significant number of pneumonectomies performed annually, few patients actually undergo additional lung resection after pneumonectomy. The information about this select group of patients is limited. The purpose of this review was to analyze our experience of subsequent resection for lung cancer in patients who have had a previous pneumonectomy. Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 30, Address reprint requests to Dr Miller, Division of General Thoracic Surgery, Mayo Clinic and Mayo Foundation, 200 First St SW, Rochester, MN 55905; miller.danielmd@mayo.edu. Material and Methods Between January 1, 1980 and June 30, 2001, 772 patients underwent pneumonectomy at the Mayo Clinic in Rochester, MN, for lung cancer. Twenty-four (3.1%) of these patients underwent subsequent resection for lung cancer in the contralateral lung. Patients who underwent pulmonary resection for benign disease or other malignancies were excluded. The records of these patients were analyzed for age, sex, preoperative conditions, pulmonary function, initial cell type and postsurgical stage, extent of initial procedure, adjuvant treatment, subsequent tumor cell type and postsurgical stage, type of subsequent resection, morbidity, and mortality. Using criteria previously described by Martini and Melamed [5], new tumors were classified as either metachronous primary lung cancer or metastatic lung cancer. Operative mortality included those patients who died within the first 30 days after operation and those who died later but during the same hospitalization. Survival was estimated by the Kaplan-Meier method using the date of the pulmonary resection after pneumonectomy as the starting point and the date of death or last follow-up as the end point [6]. All statistical tests were two-sided with the threshold of significance set at p less 2002 by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)

2 Ann Thorac Surg DONINGTON ET AL 2002;74:154 9 PULMONARY RESECTION AFTER PNEUMONECTOMY 155 than All analyses were conducted using SAS (SAS Institute Inc., Cary, NC). The study was granted approval by Mayo Foundation s Institutional Review Board. Results Table 1. Complications Complication No. (%) Respiratory complications 6 (25) Pneumonia 5 (21) Respiratory failure 4 (17) Bronchoscopy 3 (13) Tracheostomy 2 (8) ARDS 1 (4) Prolonged airleak ( 7 days) 1 (4) Cardiac complications 3 (13) Atrial fibrillation 2 (8) Ventricular arrhythmia 1 (4) Chylothorax 2 (8) Wound infection 1 (4) Renal failure 1 (4) Stroke 1 (4) ARDS adult resporatory distress syndrome. There were 24 patients (18 men and 6 women) who underwent 25 subsequent pulmonary resections for lung cancer after pneumonectomy. Median age at the subsequent resection was 64 years and ranged from 43 to 84 years. Twenty-three patients (95.8%) were current or former smokers. Nine patients (37.5%) had a completion pneumonectomy. Twelve patients (50%) had a previous right pneumonectomy. Cell type of the original cancer was squamous cell carcinoma in 10 patients, adenocarcinoma in 8, large cell carcinoma in 2, carcinoid in 2, and bronchoalveolar cell carcinoma and adenosquamous cell carcinoma in 1 patient each. Postsurgical stage of the initial cancer was stage IA in 3 patients, IB in 7, IIA in 3, IIB in 4, IIIB in 3, stage IV in 1, and unknown in 3. The initial pneumonectomy included resection of a portion of the superior vena cava in 2 patients and the diaphragm in 1. Eighteen of the subsequent tumors (72.0%) were found in asymptomatic patients by chest roentgenogram in 15, computed tomography in 2, and sputum cytology in 1. The remaining 7 patients were symptomatic and presented with hemoptysis in 4, cough in 2, and progressive dyspnea in 1. Median forced expiratory volume in 1 second was 1.47 L (range, 0.66 to 2.55 L), median percent predicted forced expiratory volume in 1 second was 48% (range, 29% to 83%), and median diffusion capacity for carbon monoxide was 63% (range, 45% to 98%). No patient was using oxygen before the subsequent resection. Six patients (24.0%) had previous adjuvant therapy that included radiation therapy in 3 and a combination of chemotherapy and radiation in 3. Subsequent pulmonary resection was performed 2 to 213 months (median, 23 months) after pneumonectomy through a posterolateral thoracotomy in all patients. Twenty-five pulmonary resections were performed in the 24 patients. One patient had two separate resections (solitary wedge excision and multiple wedge excisions) for metastatic squamous cell cancer. Fourteen resections were for metachronous lung cancer, and 11 were for metastatic lung cancer. The type of resection performed was a solitary wedge excision in 14 patients, multiple wedge excisions in 7, segmentectomy in 3, and lobectomy in 1. Resection was considered complete in all patients. Cell type was squamous cell carcinoma in 10 patients, adenocarcinoma in 7, large cell in 4, and bronchioalveolar in 4. Fifteen patients (62.5%) had the same cell type as the original lung cancer. Complete lymphadenectomy was performed in 6 patients and lymph node sampling in 7. Metachronous tumors were stage IA in 9 patients and stage IB in 5. There were two operative deaths (8.3%), one caused by a stroke with subsequent multisystem organ failure after a segmentectomy and the other secondary to a ventricular arrhythmia after multiple wedge excisions. Complications occurred in 11 patients (44.0%; Table 1). Twenty patients (80.0%) were extubated immediately after the operation, and 3 patients were extubated the next morning. Two patients required a tracheostomy for prolonged ventilatory support. Median hospitalization was 7 days (range, 4 to 72 days). Six patients (25.0%) required supplemental oxygen at discharge; only 2, however, required supplemental oxygen long-term. Fourteen patients (58.3%) underwent a single wedge excision, and 10 had more extensive resections. No difference in preoperative patient characteristics was observed between patients who had a solitary wedge excision and those patients who had a more extensive resection (Table 2). No operative deaths occurred in the 14 patients who had a solitary wedge excision, but 4 experienced complications (28.6%), 1 of which was respiratory (7.1%). Three patients (21.4%) were discharged on supplemental oxygen. In contrast, 2 of the 10 patients with more extensive resection died (operative mortality, 20.0%), and 7 experienced complications (70.0%), 5 of which were respiratory (50.0%). Three patients (30.0%) were discharged on supplemental oxygen. Median hospitalization in the solitary wedge excision patients was 7 days (range, 4 to 22 days) as compared with 13 days Table 2. Preoperative Data Solitary Wedge Extended Resection p Value Number of patients Age (y, median, range) (61) (61) Tobacco use (%) (79) (82) 0.39 COPD (%) (36) (55) 0.35 Preop FEV 1 (L, median) Cardiac comorbidity (%) (21) (27) 0.73 Preop steroids (%) (14) (0) 0.19 Adjuvant therapy (%) (29) (18) 0.55 COPD chronic obstructive pulmonary disease; FEV 1 forced expiratory volume in 1 second; Preop preoperative.

3 156 DONINGTON ET AL Ann Thorac Surg PULMONARY RESECTION AFTER PNEUMONECTOMY 2002;74:154 9 Table 3. Postoperative Data Value Solitary Wedge Extended Resection p Value Mortality (no. of 0 (0) 2 (18) 0.10 patients, %) Complications 4 (29) 7 (64) 0.08 Respiratory 1 (8) 5 (45) 0.03 complications Length of stay (d, (7) 4 22 (13) median, range) Home oxygen (no. of patients, %) 3 (25) 3 (38) 0.55 (range, 4 to 72 days) for patients with more extensive resections (Table 3). Follow-up was complete in all 22 operative survivors and ranged from 6 to 140 months (median, 37 months). No patients received adjuvant treatment during the follow-up period. Currently, 8 patients are alive. The cause of death in the remaining 14 patients included recurrent lung cancer in 7, other cancers in 2, respiratory failure in 2, other medical causes in 2, and unknown in 1. Overall 1-, 3-, and 5-year survival was 87%, 61%, and 40, respectively (Fig 1). Overall median survival was 39 months. One-, 3-, and 5-year survival after resection for metachronous cancers were 86%, 71%,, and 50%, respectively, as compared with 90%, 41%, and 14%, respectively, for metastatic disease (p 0.14; Fig 2). Median survival for metachronous cancer was 73 months compared with 27 months for recurrent cancer (p 0.14) One-, 3-, and 5-year survival after a solitary wedge excision was 93%, 64%, and 46%, respectively, as compared with 80%, 57%, and 25%, respectively, for a more extensive resection (p 0.54; Fig 3). Median survival for patients undergoing a solitary wedge excision was 59 months compared with 39 months for extended resection patients (p 0.53). Comment Patients with a completely resected bronchogenic carcinoma have a 5% to 10% risk of having a new lung cancer Fig 2. Probability of survival (death of any cause) in 14 patients who underwent resection for metachronous primary lung cancer as compared with the 10 patients who had resection for metastatic lung cancer. Zero time on the abscissa represents the date of thoracotomy (p 0.14). [3] and a 30% to 80% risk of having recurrent disease [2, 4]. This percentage will continue to increase as patient survival improves. Nevertheless, surgical resection for bronchogenic carcinoma of the residual lung after pneumonectomy is rarely reported. Fewer than 100 cases have been published in the literature during the last 35 years [5, 7 14]. Factors limiting surgical intervention in these patients are poor pulmonary reserve, the presence of distant metastasis, and a misconception by many physicians that a previous pneumonectomy is an absolute contraindication for subsequent lung resection. Our results demonstrate that further resection in selected patients can be performed with acceptable morbidity and mortality and good long-term survival. Recurrent lung cancer that develops after successful resection of bronchogenic carcinoma usually occurs early and rarely is found after 5 years [1]. In contrast, a new primary lung cancer can occur at any time after curative resection and occurs at a rate of 1% to 5% per year [1, 3, 8]. There are no definitive criteria to differentiate between a metachronous primary cancer and metastatic disease. According to Martini and Melamed [5], a second Fig 1. Probability of survival (death of any cause) in 24 patients who underwent subsequent resection for recurrent lung cancer after pneumonectomy. Zero time on the abscissa represents the date of thoracotomy. Fig 3. Probability of survival (death of any cause) in 14 patients who underwent a solitary wedge excision as compared with the 10 patients who underwent a more extensive resection. Zero time on the abscissa represents the date of thoracotomy (p 0.54).

4 Ann Thorac Surg DONINGTON ET AL 2002;74:154 9 PULMONARY RESECTION AFTER PNEUMONECTOMY 157 lung tumor in a patient previously operated on for lung cancer should be considered a second primary (metachronous) and not a metastasis when the cell type of the two lesions is different, the time interval between the two lesions exceeds 2 years, and there are no other distant metastases. Accordingly, 14 (58.3%) patients in this series had a metachronous lung cancer. Seventy percent to 80% of patients who have metachronous lung cancers are asymptomatic; these new cancers are usually discovered on surveillance chest roentgenograms [5, 15 17]. Symptoms, if present, usually include cough, hemoptysis, and dyspnea [1, 15, 17]. Because the majority of patients are asymptomatic at the time of diagnosis, a long-term follow-up program of lung cancer patients who undergo successful resection is necessary. However, several recent studies have shown that routine surveillance using conventional diagnostic studies does not impact long-term survival [18 20]. Therefore, it is impossible to recommend an aggressive surveillance program after resection for lung cancer. Also, it is difficult to recommend a surveillance program after resection because there is no standard screening method for detection of new lung cancers in the United States. Hopefully, current ongoing high-resolution computed tomography screening studies may change our philosophy about screening and surveillance [20]. Preoperative workup of a lesion suggestive of a second lung cancer is identical to the workup of the initial cancer. Extensive testing is recommended to determine whether the patient has advanced disease within the chest and or extrathoracic sites. Radiologic evaluation should include standard chest roentgenogram and thoracic and upper abdomen computed tomography. Computed tomography of the brain and a bone scan should be considered in selected patients. As positron emission tomography scans become more available and reliable, they may become the diagnostic tool required before proceeding with resection in patients with recurrent disease. None of our patients had a positron emission tomography scan before subsequent resection. If there is evidence of mediastinal lymphadenopathy, mediastinoscopy should be performed before resection. The main concern in these patients is their pulmonary reserve and quality of life after undergoing operation on their only lung. Not all patients who present with a new lesion in the contralateral lung after pneumonectomy are candidates for resection. Only 3% of patients who had a pneumonectomy for lung cancer at our institution went on to subsequent resection during a 21-year period. Median forced expiratory volume in 1 second in our select group of patients was 1.47 L, with a range of 0.66 to 2.55L; only 2 patients had a forced expiratory volume in 1 second of less than 800 ml. Unfortunately, one of these patients died postoperatively, but not because of respiratory failure. The extent of pulmonary resection a patient can tolerate is determined preoperatively on the basis of both underlying pulmonary function and the patient s performance status. The size and the location of the lesion are also helpful in determining how much lung parenchyma will be removed with complete resection of the tumor. The need for a lobectomy to resect a deepseated tumor may be a limiting factor that prevents a patient from undergoing a potentially curative resection. Success of pulmonary resection in a patient who has had a previous pneumonectomy is dependent on a team approach consisting of an experienced thoracic surgeon, anesthesiologist, pulmonologist, respiratory therapist, dedicated nursing staff, and pulmonary rehabilitation specialists. New techniques in anesthesiology and critical care have enabled patients with poor pulmonary function to achieve a better outcome after pulmonary resection. Anesthetic agents with minimal respiratory and cardiac depression have led to early extubation, which avoids tracheobronchitis and nosocomial infections from prolonged mechanical ventilation. In our series, 92% of our patients were extubated by the morning after the procedure. Epidural anesthesia has helped to reduce postoperative pain and to enable patients to clear secretions by improved cough, deeper inspiration, and early ambulation. Also, because of the localized pain control there are fewer sedation problems, which allows patients to participate in an inpatient pulmonary rehabilitation program. Extended pulmonary resections, such as lobectomy, segmentectomy, and multiple wedge excisions, have been associated with a poor outcome in patients undergoing pulmonary resection after pneumonectomy [10, 13, 14]. This was also evident in our study. Postoperative respiratory function is dependent on the extent of resection and restrictive effect of the thoracotomy [21]. Also, manipulation and contusion of the residual lung may significantly impair the already limited pulmonary reserve in these patients. Therefore, resection should be performed using decreased tidal volumes to minimize barotrauma and parenchymal damage related to extensive manipulation of a hyperinflated lung. Periods of apnea are also frequently used to facilitate deeper wedge excisions and more extensive procedures. More than 40% of our patients had a more extensive procedure. Although preoperative risk factors were similar between the patients undergoing a more extensive resection and those undergoing a solitary wedge excision, the patients undergoing an extended resection had significantly more respiratory complications. Although lobectomy is considered the treatment of choice for early stage non small cell lung cancer because of less recurrence and improved survival compared with limited resection [22, 23], it is not usually an option for most patients after pneumonectomy, unless the middle lobe is the lobe considered for resection. Only 1 patient (4.2%) in our series had a lobectomy performed. Overall survival was not significantly influenced by limited resection in this select group of patients. Actually, patients who underwent a solitary wedge excision had a better 5-year survival than the patients who had a more extensive resection. When considering both the technical and oncologic aspects in this high-risk group of patients, limited resection provides the best risk-benefit ratio. A solitary wedge excision when possible should be consid-

5 158 DONINGTON ET AL Ann Thorac Surg PULMONARY RESECTION AFTER PNEUMONECTOMY 2002;74:154 9 ered the gold standard for patients with a lung cancer in the remaining lung. Because of limited cardiopulmonary reserve after pneumonectomy, one has to carefully evaluate the riskbenefit ratio of an additional resection in patients with metastatic disease. Our study showed that patients who underwent resection for metachronous cancer had an improved 5-year survival compared with patients who had resection for metastatic disease. Five-year survival after resection for metastatic disease was 14%. The problem is that one cannot differentiate whether the lesion is a metastatic tumor or a metachronous primary cancer until pathologic confirmation at the time of the thoracotomy. If the diagnosis of metastatic disease can be established preoperatively, other treatment modalities may be warranted. In conclusion, pulmonary resection of bronchogenic carcinoma after pneumonectomy is associated with acceptable morbidity and mortality. Long-term survival can be achieved in selected patients, especially in patients with a metachronous primary cancer. Wedge excision when feasible is the treatment of choice. References 1. Deschamps C, Pairolero PC, Trasteck VF, et al. Multiple primary lung cancers. J Thorac Cardiovasc Surg 1990;99: Fontana RS, Sanderson DR, Woolner LB, et al. Lung cancer screening: The Mayo program. J Occup Environ Med 1986; 28: Pairolero PC, Williams DE, Bergstralh EJ, et al. Postsurgical stage I bronchogenic carcinoma: morbid implications of recurrent disease. Ann Thorac Surg 1984;38: Fleisher AG, McElvaney G, Robinson CLN. Multiple primary bronchogenic carcinomas: treatment and follow-up. Ann Thorac Surg 1991;51: Martini N, Melamed M. Multiple primary lung cancers. J Thorac Cardiovasc Surg 1975;70: Kaplan E, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Shields TW, Drake CT, Sherick JC. Bilateral primary bronchogenic carcinoma. J Thorac Cardiovasc Surg 1964;48: Salerno TA, Munro DD, Blundell PE, et al. Second primary bronchogenic carcinoma: life-table analysis of surgical treatment. Ann Thorac Surg 1979;27: Kittle CF, Faber LP, Jensik RJ, et al. Pulmonary resection in patients after pneumonectomy. Ann Thorac Surg 1985;40: Levasseur P, Regnard JF, Icard P, et al. Cancer surgery on a single residual lung. Eur J Cardiothorac Surg 1992;6: Westermann CJ, van Swieten HA, Brutel de la Riviere A, et al. Pulmonary resections after pneumonectomy in patients with bronchogenic carcinoma. J Thorac Cardiovasc Surg 1993;106: Massard G, Wihlm JM, Monard G. Surgical management for metachronous bronchogenic carcinoma after pneumonectomy. J Thorac Cardiovasc Surg 1995;109: Spraggiari L, Grunenwald D, Girard P, et al. Cancer resection on the residual lung after pneumonectomy for bronchogenic carcinoma. Ann Thorac Surg 1996;62: Terzi A, Furlan G, Gorla A, et al. Lung resection on single residual lung after pneumonectomy for bronchogenic carcinoma. Thorac Cardiovasc Surg 1997;45: Mathisen DJ, Jensik RJ, Faber LP, et al. Survival following resection for second and third primary lung cancers. J Thorac Cardiovasc Surg 1984;88: Abbey-Smith R, Nigam BK, et al. Second primary lung cancer carcinoma. Thorax 1976;31: Jensik RJ, Faber LP. Kittle CF. Survival following resection for second primary bronchogenic carcinoma. J Thorac Cardiovasc Surg 1981;82: Hiebert CA. The cured lung cancer patient: is follow up by the surgeon worthwhile? Ann Thorac Surg 1991;60: Naunheim KS, Virgo KS. Postoperative surveillance following lung cancer resection. Chest Surg Clinic North Am 2001; 11: Midthun DE, Jett JR. Early detection of lung cancer: today s approach. J Respir Dis 1998;19: Cerfolio RJ, Allen MS, Trastek VF, et al. Lung resection in patients with compromised pulmonary function. Ann Thorac Surg 1996;62: Lung Cancer Study Group (prepared by Ginsberg RJ and Rubinstein LV). Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Ann Thorac Surg 1995;60: Miller DL, Rowland CM, Deschamps C, et al. Surgical treatment of non-small cell lung cancer one centimeter or less in diameter. Ann Thorac Surg 2002, in press. DISCUSSION DR LESLIE J. KOHMAN (Syracuse, NY): Very nicely presented. I have three questions. You had 24 patients of 700 some who had pneumonectomies. How many patients came back with a metachronous lung cancer and were turned down? You had all diffusing capacities greater than 50%, so this was obviously selected. I am just curious as to what percent of these patients actually did get into the surgeon s hands. The second question is, the patient who had a lobectomy, which lobe was it? The third question is how did you determine whether a tumor on the other side was a metastasis or a metachronous lung cancer of the same cell type? DR DONINGTON: I will start with the first one, which is we do not know the denominator in this group of patients. I cannot tell you how many of those 772 came back for evaluation and what percentage this 24 represents. The one lobe that was resected was a left lower lobe in a patient with a previous right pneumonectomy. As regards our criteria for a metachronous versus a primary lung cancer, we used the criteria as described by Doctor Martini in his article from 1975, which stated that a second tumor in a patient with a previous lung cancer would be a metachronous primary lung cancer and not a metastasis, one, if the histologies were different, or, two, if the interval between the two lesions was greater than 24 months with no evidence of distant metastatic disease. DR THOMAS R. TODD (Abu Dhabi, United Arab Emirates): I would like to follow up on what Leslie said. She noted that your diffusing capacities were high at 65, so this is probably a selected group of patients. Nonetheless, you do have forced expiratory volumes in 1 second as low as Did you report your data in your manuscript with the forced expiratory volume in 1 second being a percent of the forced vital capacity so that the number is a little more meaningful? Second, given the fact that we have

6 Ann Thorac Surg DONINGTON ET AL 2002;74:154 9 PULMONARY RESECTION AFTER PNEUMONECTOMY 159 learned from the lung volume reduction literature that the absolute forced expiratory volume in 1 second probably does not play a big role in the kind of operation that we perform, did you assess these people with either exercise testing or quantitative ventilation-perfusion scans to determine the function of the specific areas of lung you were going to resect? DR DONINGTON: As for your first question, yes, in the paper we present both the absolute volumes and the percents. As for exercise testing, yes, there actually was very good exercise testing documented in the majority of the charts on these patients, and they were not quantitative ventilation-perfusion scans but something a little more simple, as in most of the patients were walked up a flight of stairs or two. DR LARRY R. KAISER (Philadelphia, PA): Jessica, you mentioned staging the mediastinum. Your conclude that a wedge excision is adequate yet you did at least perform lymph node sampling or lymph node dissection in some of these patients. Do you stage these patients preoperatively? Would you recommend performing mediastinoscopy on these patients? If you did mediastinoscopy and found N2 disease, would you still recommend a resection in that particular patient? Is it feasible to do mediastinoscopy in these patients? Were all of these pneumonectomies performed at your institution or were some of these from outside and thus your denominator is a little bit misleading? DR DONINGTON: Three of the procedures were performed at an outside institution. Not all the patients had had a mediastinoscopy either at the initial or at the second resection. We would recommend staging the mediastinum and the patients as carefully as possible, including, obviously, chest computed tomography and probably positron emission tomography scan in all patients. We did not perform mediastinoscopy before any of these procedures, but we do not consider that as something that could not be done in a suspicious mediastinum. DR ROBERT J. CERFOLIO (Birmingham, AL): I have a question about the lymph nodes, I thought Doctor Kaiser was going to ask that, but he did not go where I thought he was going to go. I notice that you did not stage 12 patients, ie, 12 patients do not have their lymph nodes removed and were not adequately staged. I know that the Mayo Clinic preaches and practices outstanding staging having been there and been trained by Doctors Pairolero, Allen, Deschamps, and Trastek, most of you perform a complete thoracic lymphadenectomy, as I currently do now at University of Alabama Brimingham. Was the decision not to remove the nodes intentionally made because of the fear of pulmonary edema? For example, if the initial pneumonectomy was performed and the patient had a complete thoracic lymphadenectomy on one side, did you think it was contraindicated to do another thoracic lymphadenectomy because of some of the reports of pulmonary edema; is that why you did not stage these 12 patients? If so can you elaborate on that phenomenon, its incidence, and treatment and prevention. It seems very unusual to me that the patients did not have some type of intraoperative lymph nodes staging performed at least a biopsy of each nodal station. Was this done intentionally to avoid complications from bilateral thoracic lymphadenectomy? DR DONINGTON: I do not believe so. The one thing I can tell you about the lymphadenectomies is that the timing of them is a very long retrospective review. Many of the older subsequent resections did not have lymph node dissections that went along with them. Many of the more recent ones did. So I think that there has probably been a change in the philosophy at the institution. DR RICHARD I. WHYTE (Stanford, CA): Jessica, I enjoyed your presentation. I found it very interesting. At the end you make the conclusion that it is justifiable to do these operations on certain individuals with recurrent lung cancer. My question is, are you including in the term recurrent both the metastatic as well as the metachronous lesions? When you look at the metastatic tumors, you quoted a 5-year survival rate of 14%. Actually the data showed it to be a 3.5-year survival rate of 14%. So you actually observed not a very good survival for some of these patients, and perhaps some alternative measures should be considered instead of operating on these highrisk individuals. DR DONINGTON: We agree, but the one thing that we have to note is that until you operate, you do not always necessarily know whether this is a metachronous lung cancer or a metastatic disease. Even many patients who had a short, what we would call disease-free, interval, when we went back the second time, it was actually a different histology and therefore a different lung cancer.