ORIGINAL ARTICLES. Neoadjuvant Therapy: A Novel and Effective Treatment for Stage IIIb Non-Small Cell Lung Cancer

Transcription

1 ORIGINAL ARTICLES _ Neoadjuvant Therapy: A Novel and Effective Treatment for Stage IIIb Non-Small Cell Lung Cancer Valerie W. Rusch, MO, Kathy S. Albain, MO, John J. Crowley, PhD, Thomas W. Rice, MO, Vassyl Lonchyna, MO, Robert McKenna, [r, MO, Keith Stelzer, MO, Robert B. Livingston, MO, and the Southwest Oncology Group Thoracic Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York; Section of Hematology Oncology, Department of Medicine, Loyola University Chicago, Maywood, Illinois; Southwest Oncology Group Statistical Office, Seattle, Washington; Division of Thoracic Surgery, Cleveland Clinic, Cleveland, Ohio; Department of Cardiothoracic Surgery, Loyola University Chicago, Maywood, Illinois; Wilshire Oncology Medical Group, Marion, California; and Department of Radiation Oncology and Division of Medical Oncology, Department of Medicine, University of Washington, Seattle, Washington Neoadjuvant therapy has become an accepted treatment for stage IlIa, but not for stage IlIb, non-small cell lung cancer, which is usually considered incurable and treated nonsurgically. We determined the feasibility of neoadjuvant therapy in the setting of stage Illb non-small cell lung cancer in a prospective multiinstitutional trial. For patients to be eligible for entry into the study, they had to have pathologically documented Tl-4 N2-3 disease. Treatment consisted of: (1) cisplatin (50 mg/m") given on days 1, 8, 29, and 36 plus VP-16 (50 mg/nr') given on days 1 to 5 and 29 to 33, together with concurrent radiotherapy (4,500 cgy; 180 cgy per daily fraction); and (2) surgical resection performed 3 to 5 weeks after induction of medical therapy, if the response was stable, partial, or complete. Of the 126 total eligible patients entered into the study, 51 patients had stage Illb tumors (24 with T4 tumors and 27 with N3 disease). This consisted of 34 men and 17 women with a median age of 57 years. Thirty-two (63%) patients (18 with T4 tumors and 14 with N3 disease) underwent resection of the primary tumor, with a 5.2% operative mortality. There was no difference in the operative time, blood loss, and length of hospital stay for the T4 versus the N3 patients. For all 51 patients, survival at 2 years was 39%. Sites of relapse in all patients were mainly distant, even though patients with N3 disease did not initially have involved N3 nodes resected. Our experience shows the feasibility of instituting neoadjuvant therapy in patients with stage IIlb non-small cell lung cancer, and suggests that such patients could be included in future trials designed to evaluate the role of surgical resection in the combined-modality treatment of stage III non-small cell lung cancer. (Ann Thorae Surg ) Stage IlIb non-small cell lung cancers are usually considered unresectable tumors and, with rare exceptions, are treated with chemotherapy and irradiation, or with irradiation alone. Clinical trials of neoadjuvant therapy that combine surgical resection with induction chemotherapy or chemoradiotherapy have focused primarily on patients with stage IlIa N2 disease (1). Patients with stage IlIb non-small cell lung cancer have only occasionally been included in such trials. Because it was unclear whether patients with T4 tumors or N3 disease might be suitable candidates for neoadjuvant therapy, the Southwest Oncology Group (SWOG) decided to address this question within the context of a phase II trial of neoadjuvant therapy for selected groups of patients with stages IlIa and IlIb non-small cell lung cancer. Presented at the Thirtieth Annual Meeting of The Society of Thoracic Surgeons. New Orleans, LA, Jan 31-Feb 2,1994. Address reprint requests to Dr Rusch, Thoracic Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 127S York Ave, New York, NY by The Society of Thoracic Surgeons Material and Methods This prospective multiinstitutional trial was open to selected patients with biopsy-proven stage IlIa (Tl-3 N2) or IlIb (T4 or N3) non-small cell lung cancer who had not received prior treatment. Computed tomographic findings indicating mediastinal adenopathy or direct mediastinal invasion by the primary tumor were not accepted as criteria for entry into the study. The criteria required to establish a IlIb tumor stage included (1) positive N31ymph nodes shown by mediastinoscopy or by supraclavicular lymph node biopsy findings, and (2) lesions demonstrated to be T4 by virtue of recurrent laryngeal nerve involvement, extension into the trachea shown by bronchoscopy, or direct extension into the mediastinum found at mediastinoscopy or exploratory thoracotomy. Patients with tumors that were classified as T3 because of superior sulcus or direct chest wall involvement were eligible only if the N2 or N3 nodal involvement was also pathologically proved. Contraindications to eligibility included (1) small cell lung cancer; (2) bronchioloalveolar carcinoma with lobar /94/$7.00

2 Ann Thorac Surg RUSCH ET AL 291 Selected Stages IliA and 1118 NSCLC ~ 2 cycles COOP, VP-16 + concurrent RT (4500 cgy) t CR, PR, Stable disease + Thoracotomy, Complete resection (-) nodes + No treatment Fig 1. Treatment schemafor Southwest Oncology Group trial (COOP = cisplatin; CR = complete response; NSCLC = non-small cell lung cancer; PR = partial response; RT = radiotherapy.) or multilobar involvement; (3) other primary cancer diagnosed within the 5 years before diagnosis; (4) tumors classified as T4 by virtue of a malignant pleural effusion; and (5) a clinically evident superior vena cava syndrome. Computed tomographic scans of the chest, abdomen, and brain as well as a bone scan were required in order to exclude the presence of distant metastases. Hematologic, hepatic, and renal function, as determined by a screening complete blood count, the chemical profile, and the calculated creatinine clearance had to be adequate to permit cisplatin chemotherapy. A predicted postresection forced expiratory volume at 1 second of 800 ml/s, based on the results of pulmonary function tests and a quantitative ventilation perfusion lung scan, was required. Preregistration evaluation of the patient by the thoracic surgeon, medical oncologist, and radiation oncologist was mandated to make certain that all physicians agreed that the patient could safely tolerate the combined-modality regimen. This trial was open only to SWOG institutions that had thoracic surgeons who were willing to abide by the stringent staging criteria required in the protocol. The 25 participating institutions included 15 academic centers and 10 community-based groups. The trial was approved by the institutional review board at each of these institutions. Patients who met the eligibility criteria and agreed to participate in the trial gave written informed consent. Patients were classified at the time of initial registration according to the cell type, stage (Illa versus Illb), and performance status (SWOG scale 0 to 1 versus 2). After being registered in the study, patients were started on induction treatment consisting of concurrent chemotherapy and radiation therapy (Fig 1). Cisplatin (50 mg/m 2 given intravenously) was administered on days 1, 8, 29, and 36. VP-16 (etopaside: 50 mg/m 2 given intravenously) was administered on days 1 to 5 and days 29 to 33. Continuous chest radiotherapy to the primary tumor, ~ Re-evaluation t Unresectable Incomplete resection (+) nodes + t Progressive disease Off-study 2 cycles COOP, VP-16 + RT (1400 cgy) From July 1988 to March 1992, 156 patients with stage lila or IIlb non-small cell lung cancer were entered into SWOG trial No A total of 126 patients were ultimately deemed eligible for entry. Preliminary data regarding the response rates and toxicity of the induction chemoradiomediastinum, and bilateral supraclavicular regions was delivered through anteroposterior-posteroanterior fields at a daily dose of 180 cgy, for a total dose of 4,500 cgy. Two to 4 weeks after completion of the induction treatment, the patients' disease was restaged and they were evaluated for their clinical response by physical examination, complete blood count, chemical profile, chest radiographs, bone scan, and computed tomographic scans of the chest, abdomen, and brain. Patients were removed from the study if distant metastases or local disease progression developed, or if they had persistently involved supraclavicular nodes. All other patients, including those who had radiographically confirmed stable disease, underwent thoracotomy with the intent to completely resect the primary tumor. Mediastinal nodal sampling or dissection with mapping performed according to the American Thoracic Society lymph node map was required. Patients who underwent a grossly complete resection of the primary tumor, with microscopically negative margins and negative mediastinal nodes, did not receive any additional treatment. Patients who exhibited microscopically positive margins or positive N2 or N3 nodes, or who proved unresectable, received additional chemotherapy and radiotherapy starting 3 to 6 weeks after thoracotomy. The chemotherapy included two additional cycles of cisplatin and VP-16 administered in a manner identical to that done preoperatively. The radiotherapy was started at the same time as the chemotherapy, and consisted of an off-cord boost of 1,400 cgy limited to the area of known disease, delivered in 180-cGy daily fractions. After the completion of all treatment, the patients' courses were evaluated by physical examination, complete blood count, and chemical profile every 2 months for the first year, then every 3 months for the next 2 years, then every 6 months for 2 years. Computed tomographic scans of the chest, abdomen, and brain, as well as a bone scan, were obtained at 6, 12, and 18 months after completion of the treatment, then annually for 3 years. Follow-up in all patients is maintained until the time of death. Review of all the data forms for each patient was carried out jointly by the medical oncology (K.A.) and thoracic surgery (V.R.) principal investigators of this trial. The operative reports for the prestudy surgical staging procedures and for subsequent thoracotomy, and the associated pathology reports for each patient, were reviewed to verify that the information coded on the data forms was correct. Overall survival was calculated using the product-limit method according to Kaplan and Meier [2]. Perioperative data for the T4 versus N3 patients were compared by Wilcoxon's rank-sum test, and survival was compared using the log-rank test [3]. Differences were considered significant if the two-tailed p value was less than Results

3 292 RUSCH ET AL Ann Thorac Surg Table 1. Sites of Tumor Involvement Type of Disease T4 tumors Right lung Left lung N3 disease Site of Involvement Main carina or trachea Mediastinum (pericaval region) Spine Aortopulmonary window Spine Contralateral mediastinal Supraclavicular No. of Patients therapy in the first 75 patients entered in the study have been reported [4]. However, that interim analysis included only 28 patients with stage IIIb tumors. We report here the results of treatment for the entire group of 51 eligible patients who had stage IIIb non-small cell lung cancer, and describe the findings during a longer follow-up (a minimum follow-up of 24 months in all patients). There were 34 men (66%) and 17 women (34%) with a median age of 57 years (range, 41 to 70 years). In 24 patients, the indication for entry in the study was a T4 tumor; in 27 patients, it was the presence of N3 nodal disease. Of the 24 patients who had T4 tumors, nine were right-sided tumors involving the proximal airway, the mediastinum, or the spine, and 15 were left-sided tumors encasing the subaortic region or involving the spine (Table 1). Of the 27 patients with N3 disease, 9 had involved contralateral nodes and 18 had involved supraclavicular nodes. Twenty-two tumors (43%) were squamous cell cancers, 22 tumors (43%) were adenocarcinomas, and seven tumors (14%) were large cell cancers. After induction treatment, progressive disease developed in 7 patients, 1 had an intercurrent myocardial infarction, 2 had persistent N3 disease in the supraclavicular nodes, and 1 did not recover from the toxicity of induction. Therefore, 40 patients (78%) were potential candidates for surgical resection. Of these, 1 surgeon refused to operate, and 1 patient refused to have the operation. Thus, 38 patients ultimately underwent thoracotomy. Eighteen of the 24 patients (75%) with T4 tumors and 14 of the 27 patients (52%) with N3 disease underwent resection of the primary tumor. The types of resections performed are shown in Table 2. They are classified as either standard or "complex" procedures, the latter including intrapericardia1resections as well as en bloc resections of the chest wall, vertebral bodies, or mediastinal vessels. Fifteen of the 18 patients (83%) with T4 tumors who underwent resection required a complex operation, with the most common procedure an intrapericardial pneumonectomy. In contrast, only 3 of the 14 patients (21%) who had N3 disease underwent a complex resection. Of the 3 patients who originally had T4 tumors by virtue of involvement of the proximal airway (main carina, 2 patients; lateral tracheal wall, 1 patient), 1 patient did not undergo surgical resection because of progressive disease and the primary tumor was downstaged in 2, and thus carinal or tracheal resection was not required. All patients with N3 Table 2. Types of Resection Types of Resection Operated Resected Standard resection Segmentectomy Lobectomy Pneumonectomy Complex resection Intrapericardial pneumonectomy Lobectomy and spine Lobectomy and chest wall LobectomyI 2 pneumonectomy and vessels N3, N3, T4 (n = 24) Contralateral (n = 9) Supraclavicular (n = 18) 19 (79%) 9 (90%) 8(47%) 18 (75%) 8 (80%) 6 (35%) Table 3. Perioperative Information in the 36 Patients Undergoing Thoracotomy" 2 disease were approached by means of a standard posterolateral thoracotomy to remove the primary tumor and the ipsilateral mediastinal lymph nodes. No attempt was made to resect the contralateral or supraclavicular lymph nodes. The operating time, intraoperative blood loss, and length of hospital stay were not significantly different between the patients with T4 tumors and those with N3 disease (Table 3). Two of the 38 patients (5.2%) who underwent thoracotomy died postoperatively of multisystem failure. Both patients had undergone long and technically complicated resections for T4 tumors, one involving the spine and chest wall, the other involving the aorta. Perioperative complications in the remaining 34 patients were infrequent, and included nonfatal arrhythmias (3 patients), wound infection (1 patient), and atelectasis (5 patients). At 2 years postoperatively, the overall survival rate for all 51 patients entered in study was 39%, with a median survival of 17 months. For the 24 patients with T4 tumors, the overall survival rate at 2 years was 53%, and, for the 27 patients with N3 disease, it was 25% (p = 0.09). The median survival for patients with T4 tumors was 27 months versus 13 months for patients with N3 disease. Twenty-five patients did not relapse. Four had initial relapse in the ipsilateral lung or lymph node only, 8 had locoregional plus simultaneous distant relapse, and 14 had only distant metastases. There was no significant difference in the Variable T4 (n = 19) N3 (n = 15) P Value Operating time (h) 3.3 ( ) ) 0.56 Blood loss (ml.) 600(50-2,300) ,400) 0.15 Hospital stay (days) 80-19) 7 (4-22) 0.83 a The data are expressed as medians, with the ranges in parenthesis. There were no significant differences in the data between the patients with T4 tumors and those with N3 disease.

4 Ann Thorac Surg 1994;58:290-5 RUSCH ET AL 293 distribution of relapse sites between T4 and N3 patient subsets. Comment When the International Staging System for Lung Cancer was introduced in 1986, the stage IlIb classification was used to designate locally advanced and usually unresectable tumors [5]. For a few small series, 5-year survival rates as high as 20% have been reported in highly selected patients who have undergone complete resection of tumors that are T4 NO-l by virtue of carinal involvement or direct extension into the mediastinum [6-8]. The presence of mediastinal nodal involvement confers a dismal prognosis if the treatment in such patients consists primarily of surgical resection [6-8]. In general, patients with stage IlIb non-small cell lung cancer have been treated nonsurgically with radiation therapy alone, or, more recently, with chemotherapy and irradiation. Several prospective randomized trials conducted in patients with stage III nonsmall cell lung cancer have revealed that combinedmodality therapy consisting of chemotherapy and irradiation (either sequential or concurrent) is well tolerated and leads to better overall survival than does radiation therapy alone [9-13]. However, it is difficult to assess specifically the outcome of patients with stage IIIb tumors in these trials because most of the patients entered are only staged clinically. The lack of pathologic confirmation of the stage before entry in the study makes it impossible to distinguish definitively which patients have stage IlIa and which have stage IlIb non-small cell lung cancer. The concept of neoadjuvant therapy, namely surgical resection performed after induction chemotherapy or chemoradiotherapy, has gained widespread acceptance for patients with stage IlIa tumors because the results from numerous phase II trials have demonstrated the feasibility of this approach, and have suggested the survival rates are improved compared with the historical experience [1, 4, 14-21]. Recently, the findings from two small randomized trials have shown that the overall survival rate in patients with stage IlIa N2 disease is significantly better in association with induction chemotherapy and surgical resection than with resection alone [22, 23]. However, the concept of neoadjuvant therapy for stage IlIb non-small cell lung cancer has not been systematically evaluated. At least three neoadjuvant trials have included patients with stage IIIb tumors, but the numbers of such patients have been too small to permit any firm conclusions to be drawn [17-19]. Therefore, we decided to investigate the feasibility of neoadjuvant therapy for stage IlIb non-small cell lung cancer within the context of a larger trial aimed at identifying a treatment regimen that was safe and highly effective in the multiinstitutional setting. An induction program of concurrent chemoradiotherapy was theoretically ideal for these patients, who usually present with very bulky intrathoracic tumors and also face the imminent threat of systemic disease. This analysis of all 51 eligible patients with stage IlIb tumors entered into the SWaG No trial confirms our initial impression that neoadjuvant treatment is feasible, both in patients with T4 tumors and in patients with N3 disease [4]. Sixty-three percent of the patients were able to undergo resection of their primary tumor after successfully completing the planned induction chemoradiotherapy. The operative mortality was 5.2%, and the operative morbidity was low, even though this was a multiinstitutional trial and 44% of these locally advanced tumors were resected by a pneumonectomy. These results challenge our traditional concepts about the management of patients with stage IIIb non-small cell lung cancer, and suggest that such patients should be considered for future trials of neoadjuvant therapy. The difference in the overall survival rates for patients with T4 tumors versus the rates in patients with N3 disease, though not statistically significant, suggests that it may be appropriate to stratify patients by these two categories in future neoadjuvant trials. The results of this trial raise several issues regarding the surgical management of stage IlIb tumors after induction therapy. The operative approaches to T4 tumors are well defined. Most of them can be removed through a posterolateral thoracotomy, although, as our experience indicates, complex resections are often necessary. It is intuitively obvious that T4 tumors considered technically unresectable at presentation might be amenable to complete surgical resection after induction chemoradiotherapy, and this might improve the chances of survival. In contrast, the optimal technical approach for patients with N3 disease has not been defined, and the potential role of a neoadjuvant regimen in such patients is less clear. One approach is to regard the surgical resection only as a form of adjuvant treatment for control of the primary tumor, which may be too large to be sterilized by chemoradiotherapy alone. When this is the strategy, the technical approach would be a posterolateral thoracotomy with no attempt to resect or sample contralateral or supraclavicular nodes. Alternatively, complete surgical resection of all known tumor may lead to improved overall survival, as has been suggested by the findings from previous neoadjuvant trials that used chemotherapy alone for induction treatment [14]. When adopting this strategy, the technical approach would be a median sternotomy or clamshell thoracotomy, which would allow resection of the primary tumor and bilateral mediastinal lymph node dissection with or without supraclavicular lymph node sampling. This approach has been advocated by some centers in Japan [24, 25] but has not been widely applied in North America. However, the clamshell thoracotomy has recently been popularized for lung transplantation, bilateral pulmonary metastasectomy, and the resection of extensive mediastinal and pulmonary tumors [26, 27]. The SWaG No trial was not designed to answer this surgical question, and the number of patients with N3 disease who underwent surgical resection is too small to permit any firm conclusions to be drawn about the advisability of a posterolateral thoracotomy for resection. Nonetheless, it is interesting that the predominant form of relapse in patients with N3 disease was distant metastatic disease, just as it was for other patients entered in this trial. Only 11% of the patients with N3 disease suffered a loco regional recurrence. If future neoadjuvant trials focus on patients with N3 disease, it would be important to determine the optimal surgical approach, or at least to stratify for it.

5 294 RUSCH ET AL Ann Thorae Surg The apparently improved survival rates seen after combined-modality therapy with or without surgical resection in patients with stage IlIa non-small cell lung cancer have raised the question of how much surgical resection contributes to local control and overall survival. Indeed, the role of surgical resection in the management of stage IlIa non-small cell lung cancer is the subject of a new prospective randomized trial comparing the regimen used in the SWOG No with the same concurrent chemoradiotherapy regimen using a higher total dose of radiation but without surgical resection. Patients with stage I1Ib disease have been excluded from this trial because there were insufficient preliminary data regarding the feasibility of neoadjuvant therapy in this group of patients. Our results suggest that it may now be appropriate to include patients with stage I1Ib disease in such a randomized trial, particularly if they have T4 tumors. Should such a multiinstitutional trial be undertaken, it would be important to incorporate into it investigations that would answer the questions about the optimal technical approach to surgical resection raised by the findings yielded by the SWOG No trial. Supported by NIH grants CA 32102, 46282, 37429, 04919, and References 1. Rusch VW. Neoadjuvant therapy for Stage III lung cancer. Semin Thorac Cardiovasc Surg 1993;5: Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Statist Assoc 1958;53: Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50: Rusch VW, Albain KS, Crowley JJ, et al. Surgical resection for stage IIIA and stage IIIB non-small-cell lung cancer after concurrent induction chemoradiotherapy. A Southwest Oncology Group trial. J Thorac Cardiovasc Surg 1993;105: Lung tumours. In: Hermanek P, Sobin LH, eds. TNM classification of malignant tumours. 4th ed. Berlin: Springer-Verlag, 1987: Dartevelle PG, Chapelier AR, Pastorino U, et al. Long-term follow-up after prosthetic replacement of the superior vena cava combined with resection of mediastinal-pulmonary malignant tumors. J Thorac Cardiovasc Surg 1991;102: Dartevelle PG, Khalife I, Chapelier A, et al. Tracheal sleeve pneumonectomy for bronchogenic carcinoma: report of 55 cases. Ann Thorac Surg 1988;46: Mathisen DI, Grillo He. Carinal resection for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1991;102: Albain KS. Induction therapy followed by definitive local control for stage III non-small-cell lung cancer. A review, with a focus on recent trimodality trials. Chest 1993;103:435-50S. 10. Schaake-Konig C, van den Bogaert W, Dalesio 0, et al. Effects of concomitant cisplatin and radiotherapy on inoperable nonsmall-cell lung cancer. N Engl J Med 1992;326: Mattson K, Holsti LR, Holsti P, et al. Inoperable non-small cell lung cancer: radiation with or without chemotherapy. Eur J Cancer Clin Oncol 1988;24: Dillman RO, Seagren SL, Propert KI, et al. A randomized trial of induction chemotherapy plus high-dose radiation versus radiation alone in stage III non-small cell lung cancer. N Engl J Med 1990;323: Le Chevalier T, Arriagada R, Quoix E, et al. Radiotherapy alone versus combined chemotherapy and radiotherapy in nonresectable non-small cell lung cancer: first analysis of a randomized trial in 353 patients. J Natl Cancer Inst 1991;83: Martini N, Kris MG, Flehinger BI, et al. Preoperative chemotherapy for stage IIIA (N2) lung cancer: the Sloan-Kettering experience with 136 patients. Ann Thorac Surg 1993;55: Burkes RL, Ginsberg RI, Shepherd FA, et al. Induction chemotherapy with mitomycin, vindesine, and cisplatin for stage III unresectable non-small cell lung cancer: results of the Toronto phase II trial. J Clin Oncol 1992;10: Eagan RT, Ruud C, Lee RE, Pairolero PC, Gail MH. Pilot study of induction therapy with cyclophosphamide, doxorubicin, and cisplatin (CAP) and chest irradiation prior to thoracotomy in initially inoperable stage II M, non-small cell lung cancer. Cancer Treat Rep 1987;71: Wagner H, Lad T, Piantadosi S. Randomized phase II evaluation of preoperative radiation therapy and preoperative chemotherapy with mitomycin-c, vinblastine, and cisplatin in patients with technically unresectable stage IIIA and IIIB non-small cell cancer of the lung. Lung Cancer 1991;7: Weiden PL, Piantadosi S. Preoperative chemotherapy (cisplatin and fluorouracil) and radiation therapy in stage III nonsmall cell lung cancer: a phase II study of the lung cancer study group. J Natl Cancer Inst 1991;83: Faber LP, Kittle CF, Warren WH, et al. Preoperative chemotherapy and irradiation for stage III non-small cell lung cancer. Ann Thorac Surg 1989;47: Strauss GM, Gerndon JE, Sherman DD, et al. Neoadjuvant chemotherapy and radiotherapy followed by surgery in stage IIIA non-small cell carcinoma of the lung: report of a Cancer and Leukemia Group B phase II study. J Clin OncoI1992;10: Yashar I, Weitberg AB, Glicksman AS, et al. Preoperative chemotherapy and radiation therapy for stage IIIA carcinoma of the lung. Ann Thorac Surg 1992;53: Roth JA, Fossella FV, Komaki R, et al. A randomized trial comparing perioperative chemotherapy and surgery with surgery alone in resectable stage IIIA non-small cell lung cancer. J Natl Cancer Inst 1994;86: Rosell R, Gomez-Codina J, Camps C, et al. A randomized trial comparing preoperative chemotherapy plus surgery with surgery alone in patients with non-small cell lung cancer. N Engl J Med 1994;330: Hata E, Hayakawa H, Miyamoto, et al. The incidence and the prognosis of the contralateral mediastinal node involvement of the left lung cancer patients who underwent bilateral mediastinal dissection and pulmonary resection through the median sternotomy. Lung Cancer 1988;4(Suppl):A Watanabe Y, Ichihashi T, Iwa T. Mediasternotomy as an approach for pulmonary surgery. J Thorac Cardiovasc Surg 1988;26: Egan TM, Detterbeck Fe. Technique and results of double lung transplantation. Chest Surg Clin North Am 1993;3: Bains MS, Ginsberg R, Jones WG II, et al. The clamshell incision: an improved approach to bilateral pulmonary and mediastinal tumor. Ann Thorac Surg 1994;58:30-3. DISCUSSION DR 1. PENFIELD FABER (Chicago, ILl: Doctor Rusch and the Southwest Oncology Group are to be congratulated and complimented on the completion of this study. It entails tremendous cooperation among radiation, medical, and surgical oncologists, as well as data collectors and statisticians; and last, but not least, on the part of the patients themselves. It is a rigorous treatment program. The definition of a central tumor as being a T3 or T4 lesion frequently lies in the eye of the beholder. A lesion may be designated as T4 by one clinician and as T3 by another. This is a

6 Ann Thorae Surg RUSCH ET AL 295 very important consideration when stratifying these patients for analysis, as the central lesion I clinically stage as T3 puts it into the stage lila category. However, another group may clinically stage the exact same lesion as T4, making it a stage lilb tumor. This clinical staging thus becomes quite important when discussing stage lila and lilb lesions. Doctor Rusch described nine right-sided T4 lesions, three around the superior vena cava and three extending into the trachea or carina. However, there were no carinal or caval resections performed in this series. Among the lesions on the left side, the tumors in 14 patients were deemed T4 because of aortopulmonary involvement. Did all 14 have vocal cord paralysis? It would be helpful for Dr Rusch to expand on their precise definition of T4 disease. Fifteen patients of this group did not come to surgery, and this number seems high. If the patients were initially deemed eligible for the program and completed the pretreatment, why did more patients not undergo thoracotomy? I believe that, in such a program for the management of clinically advanced cancer, it is important to make every attempt to resect the disease if there is no evidence of metastatic cancer. There were 12 patients, or 32%, who had an incomplete resection. I would assume these 12 resections were in the patients with T4 tumors and not the ones with N3 disease, as no attempt was made to resect N3 nodes. This would indicate an approximate 50% incomplete resection rate for the T4 patients. Can Dr Rusch define any specific sites of cancer involvement in the mediastinum that would rule out thoracotomy, so that a higher complete resection rate can be achieved. The key to survival is complete resection. I would like to know the survival status of the patients who received additional chemotherapy because of residual disease. These data are important in helping us determine whether additional chemoradiotherapy is of significant benefit at a cost in terms of both dollars and patient morbidity. It is my understanding that all of the N3 patients in this series received additional treatment. The efficacy of this form of neoadjuvant therapy is clearly demonstrated by a response rate of 62%, an acceptable morbidity, and a mortality of 5.5%. A 2-year survival of 38% and locoregional recurrence in only 12 patients who have stage lilb disease further support the Rush group's long-held rationale for administering combined radiotherapy and chemotherapy for the control of stage lila non-small cell lung cancer. DR HANI SHENNIB (Montreal, Que, Canada): I congratulate you, Dr Rusch, on an excellent presentation, and I congratulate the Southwest Oncology Group for their outstanding results. I have a question that relates to the biologic characteristics of cancer. I have always wondered whether it really made a difference to the outcome in lung cancer patients, if in those who have subcarinal lymph node involvement, the ipsilateral or contralateral lymph nodes eventually become involved? Why then should the results from the surgical resection of stage lila lymph nodes be better than that for stage lilb lymph nodes if they are completely resected? Perhaps we are biased toward resecting ipsilateral disease over contralateral disease because of its accessibility? The question I have for Dr Rusch has to do with the median survival reported. I noted in the abstract that the median survival for combined stages lila and lilb disease is 14 months. You now present data on the median survival for stage lilb disease of 17 months. Does that imply that, in the Southwest Oncology Group's experience, the median survival for stage lilb disease is better than that for stage lila disease? DR JOSEPH B. ZWISCHENBERGER (Galveston, TX): Doctor Rusch, I thought the length of hospitalization for the patients undergoing lung resection after neoadjuvant treatment was reasonable. However, please comment on the total length of hospitalization for the entire protocol, including the time in the hospital for the induction chemotherapy and radiation therapy, as well as the management of related complications. DR JOSEPH LoCICERO III (Boston, MA): Doctor Rusch, I want to congratulate you on a fantastic effort. I would like to comment on our preliminary pilot study. We have 10 patients who received the same chemotherapy but a total of 56 Gy of radiation. Of these 10 patients, significant complications developed in 2 patients before operation: a myocardial infarction in one and a stroke in the other. Would you please comment on the morbidity associated with your neoadjuvant therapy before resection? DR RUSCH: I would like to thank all of the discussants for some very incisive questions. With regard to Dr Faber's remarks, one of the important aspects of this trial is that we did not accept computed tomographic scan criteria for the purpose of staging. Pathologic verification was required. Therefore, the 14 patients with T4 tumors involving or encasing the aortopulmonary window all had documented recurrent nerve paralysis based on the findings from indirect laryngoscopy. Similarly, in the patients who had T4 tumors by virtue of involvement of the spine or of the superior vena cava, this was documented pathologically at thoracotomy. Thus, we are confident that tumors classified as T4 are not T3. Patients who had phrenic nerve involvement, which according to the international staging system is T3, were indeed classified as T3. Doctor Faber questioned the number of patients who did not go on to have a thoracotomy. These were patients who had documented progression of their disease. Most of them had involved supraclavicular lymph nodes at diagnosis, and new areas of disease in the cervical or supraclavicular region or distant metastases developed. In my review of the operative notes from this group of patients, I found that all of the patients who had T4 tumors, except for the 2 who died postoperatively, underwent grossly complete resections of their primary tumor. Doctor Faber also asked about the impact of postoperative chemoradiotherapy on the survival of patients. We have not analyzed for this, and I suspect that our biostatistician would be reluctant to do so, because it would involve analyzing small subsets of patients. Moreover, the trial was not designed to ask that specific question. With respect to the question about the length of hospital stay and whether it reflected that of patients who received induction treatment, all of the patients entered in this trial received induction treatment. Therefore, the reported length of hospital stay is that observed in patients who had induction treatment followed by surgical resection. With respect to Dr LoCicero's question, the morbidity associated with the neoadjuvant therapy has been described in our initial report of this trial. We limited the dose of induction radiotherapy to 45 Gy because of concern about the risk of precipitating either the adult respiratory distress syndrome or a bronchopleural fistula at higher doses. The high percentage of patients who went on to undergo thoracotomy indicates that this was a well-tolerated regimen.