Experience has proven that pneumonectomy is a safe

Transcription

1 Pneumonectomy After Chemoradiation Therapy for Non-Small Cell Lung Cancer: Does Side Really Matter? Anthony W. Kim, MD, L. Penfield Faber, MD, William H. Warren, MD, Sanjib Basu, PhD, Sean C. Wightman, BA, James A. Weber, BA, Philip Bonomi, MD, and Michael J. Liptay, MD Division of Thoracic Surgery and Section of Medical Oncology, Rush University Medical Center, Chicago, Illinois Background. The long-term benefits and risks of pneumonectomy after neoadjuvant chemoradiation therapy remain controversial. This study evaluated our experience with pneumonectomy for advanced non-small cell lung cancer (NSCLC) after concurrent chemoradiation therapy. Methods. We reviewed medical records from patients undergoing concurrent chemoradiation therapy, followed by pneumonectomy (1983 to 2007). Clinical variables affecting Kaplan-Meier survival were analyzed. Results. After chemoradiation therapy, 129 pneumonectomies (right, 65; left, 64) were performed. Postoperative pathologic stages were complete responders (CR), 21; I, 23; II, 19; III, 62; and IV, 4. The 90-day perioperative mortality was 20% (13 of 65) after right-sided pneumonectomy vs 9% (6 of 64) after left-sided pneumonectomy (p 0.084). Complications occurred in 33% (43 of 129), including bronchopleural fistula in 12% (16 of 129) and acute respiratory distress syndrome in 2% (3 of 129). Overall 5-year survival was 33%. Survival was 32% for right-sided sections vs 34% for left-sided. CR patients had a 5-year survival of 48%. Survival of patients with postoperative N0, N1, and N2 nodes was 42%, 26%, and 28%, respectively. Multivariate analysis showed the development of major complications negatively affected 5-year survival for patients undergoing right-sided pneumonectomy (hazard ratio, 0.462; p ). Conclusions. Pneumonectomy after concurrent chemoradiation therapy achieved long-term survival. When neoadjuvant therapy resulted in complete response or nodal downstaging, survival was improved. The risk of early perioperative death and complications was higher for right-sided procedures, but long-term survival did not differ between right- and left-sided pneumonectomy. Major complications negatively affected 5-year survival with right-sided pneumonectomies. (Ann Thorac Surg 2009;88:937 44) 2009 by The Society of Thoracic Surgeons Accepted for publication April 22, Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26 28, Address correspondence to Dr Liptay, University Thoracic Surgeons, 1725 W Harrison St, Ste 774, Chicago, IL 60612; michael_liptay@rush.edu. Experience has proven that pneumonectomy is a safe thoracic operation. However, it remains an operation that carries a higher morbidity and mortality than a lobar or sublobar resection [1 3]. Certain investigators have suggested that it is the scale of the operation and not the neoadjuvant therapy that contributes to the postoperative morbidity and mortality [3, 4]. The performance of a right-sided pneumonectomy has been thought to be associated with a greater likelihood of morbidity and mortality than a left-sided pneumonectomy [5]. The results of the Intergroup 0139 trial have suggested that the benefits of a right-sided pneumonectomy may not be worthwhile in the setting of the added layer of complexity with neoadjuvant chemoradiation therapy [6]. This has led several clinicians to believe that right pneumonectomy after chemoradiation therapy should never be offered. The long-term benefits and risks of pneumonectomy (especially right-sided) after neoadjuvant chemoradiation therapy remain controversial. The purpose of this study was to evaluate the perioperative risks and longterm survival associated with pneumonectomy for advanced non-small cell lung cancer (NSCLC) after concurrent chemoradiation therapy. Patients and Methods From 1983 to 2007, the clinical records of all patients undergoing chemoradiation therapy, followed by pneumonectomy, were reviewed. Inclusion criteria were: 1. the identification of locally advanced disease that required neoadjuvant therapy to achieve complete resection, 2. completion of chemotherapy and radiation therapy in a concurrent fashion, and 3. nothing less than a pneumonectomy at the completion of the full course of neoadjuvant therapy. Institutional Review Board approval was obtained to perform this retrospective review. Individual patient consent was waived for this retrospective study. The age, gender, chemoradiation therapy regimen, histopathology, preoperative and postoperative T N M 2009 by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 938 KIM ET AL Ann Thorac Surg PNEUMONECTOMY AFTER CHEMORADIATION 2009;88: Table 1. Complications Associated With Pneumonectomy After Neoadjuvant Chemoradiation Therapy a Complication Right-Sided, N 65 Left-Sided, N 64 BPF 11 (4) 5 (2) ARDS 3 (3) 0 Empyema without BPF 1 1 Secretions/pneumonia/aspiration 2 (1) 2 (1) Cardiac arrhythmia 4 (2) 6 (1) RLN injury 1 1 Esophageal perforation 0 1 Myocardial infarction 0 1 Intraoperative hemorrhage 0 1 (1) a Numbers in parenthesis are complications that were known to directly contribute to the 90-day perioperative mortality rate. ARDS acute respiratory distress syndrome; BPF bronchopleural fistula; RLN recurrent laryngeal nerve injury. stage, 90-day mortality, major complications, and survival were recorded. Major complication were defined as bronchopleural fistula (BPF), cardiac arrhythmia, acute respiratory distress syndrome (ARDS), respiratory secretions requiring bronchoscopic intervention, pneumonia, myocardial infarction, pulmonary embolism, and recurrent laryngeal nerve injury. Clinical Considerations Locally advanced disease was defined as the presence of central disease (T3 N0), or positive mediastinal lymph nodes (N2 disease). Mediastinal lymphadenopathy was confirmed by cervical mediastinoscopy or radiographic evidence of disease, including (1) the presence of lymphadenopathy in the mediastinum greater than 2.0 cm on short axis, or (2) the presence of the aforementioned lymphadenopathy with positive findings on positron emission tomography (PET), defined as more than maximum standard uptake value of 2.5. Radiographic staging alone was relied on more frequently in the earlier time period, whereas PET scan and mediastinoscopy were used more systematically later in the study period. Patients with contralateral mediastinal lymph node involvement (N3 status 1) and limited stage IV disease (isolated brain 4) were also included but the number was very small. The study excluded patients who were initially given chemoradiation therapy in a neoadjuvant setting but in whom progressive disease developed that precluded surgical resection. Chemotherapy was given concurrently with splitcourse radiotherapy in all of the regimens used. Chemotherapy consisted of platinum-based regimens. Before 1994, the regimen consisted of cisplatin and 5-fluorouracil. Etoposide was added later. After 1994, the regimen changed to carboplatin, paclitaxel, and etoposide; eventually, etoposide was eliminated. There were some variations in these regimens. The mean split-course radiation dose was 4299 cgy (range, 4000 to 4500 cgy), although there were minor variations. Right-sided and left-sided pneumonectomies were performed in a standard fashion. No special distinction was given to pneumonectomies requiring intrapericardial dissection. Carinal pneumonectomies were not included in this analysis. Pneumonectomies performed for other malignancies such as mesothelioma, metastatic disease, and neuroendocrine malignancies were excluded. Lymphadenectomy, and not lymph node sampling, was performed for all of the pneumonectomies. The right-sided resections underwent dissection of the level 2, 4, 7, 8, 9, 10, and 11 lymph node stations. The left-sided resections underwent dissection of the level 5, 6, 7, 8, 9, 10, and 11 lymph node stations. For right-sided resection, routine stump coverage with a pleural or pericardial fat pad buttress was performed. Closure of the bronchus was routinely accomplished by stapling. Statistical Methods Kaplan-Meier 5-year survival curves were generated based on the time from pneumonectomy. All univariate comparisons of survival were made with log-rank tests and represent virtually all of the p values in this article. The exception to this was in the use of 2 analysis to generate a p value in comparing the differences in the 90-day perioperative mortality between left-sided and right-sided resections. A multivariate analysis using a Cox proportional hazard regression model was performed to evaluate the effect of several covariates on the side of resection. Additional clinical considerations and values of p 0.1 on univariate analysis were determined the factors that would be used as the variables. Statistical analysis was performed using SAS 9.1 software (SAS Institute Inc, Cary, NC). Significance was defined as p Fig 1. The overall 5-year survival for the entire cohort was 33%. The numbers in the parentheses represent the patients at risk.

3 Ann Thorac Surg KIM ET AL 2009;88: PNEUMONECTOMY AFTER CHEMORADIATION 939 Table 2. Variables Compared in the Univariate Analysis of 5-Year Survival Variable % p Value Gender Female 28 Male 36 Laterality of resection a Right 32 Left 34 Postoperative stage a b CR (0-T0 N0) 48 I 43 II 26 III 27 IV 25 Pathologic node status a b N0 42 N1 26 N2 28 Nodal downstaging a b 0 (No change) 29 1 (N2 N1, N1 N0) 33 2 (N2 N0) 42 Major complications a No 39 Yes 21 a Variables included in multivariate analysis. CR Complete response. Results b Pooled analysis. A total of 129 pneumonectomies, comprising 65 right and 64 left, were performed after concurrent chemoradiation Fig 3. Survival stratified according to the pathologic stage of complete response (CR, solid line), stage I (small dashed line), stage II (dotted line), stage III (dottted-dashed line), and stage IV (large dashed line). The numbers representing patients at risk were not included for clarity. therapy. Patients were a mean age of 57 9 years (range, 30 to 72 years). The gender distribution was 50 women and 79 men. Histologic subtypes were divided into squamous carcinoma in 77, adenocarcinoma in 44, large cell carcinoma in 5, adenosquamous in 2, and carcinosarcoma in 1. Pretreatment stages of the carcinomas were IB, 2; IIB, 11; IIIA, 95; IIIB, 18; IV, in 3. Postoperative pathologic stages were complete responders (CR), 21; IA, 11; IB, 12; IIA, 10; IIB, 9; IIIA, 46; IIIB, 16, and IV, 4. Of the pathologic Fig 2. Survival stratified according to right-sided (solid line) and left-sided (dashed line) pneumonectomy demonstrated no statistically significant difference at 5 years. The numbers in the parentheses represent the patients at risk. Fig 4. Survival stratified according to the pathologic lymph node status of N0 (solid line), N1 (dashed line), and N2 (dotted line). The numbers in the parentheses represent the patients at risk.

4 940 KIM ET AL Ann Thorac Surg PNEUMONECTOMY AFTER CHEMORADIATION 2009;88: Table 3. Downstaged Grid a Variable Pre-op Post-op Stage 0 (CR) (21) I 2 23 (22) II (18) III IV 3 4 T (22) (30) (29) (1) N (41) (12) a Numbers in parenthesis are patients who were truly downstaged to this postoperative stage with neoadjuvant chemoradiation therapy. stage IV disease, resection was used to treat 3 isolated brain metastases and 1 adrenal metastasis. Overall, complications occurred in 43 of the 129 patients (33%), including BPF in 16 (12%) and ARDS in 3 (2%; Table 1). Other complications were cardiac arrhythmias, empyema without BPF, respiratory secretions or pneumonias requiring bronchoscopic intervention, recurrent laryngeal nerve injury, myocardial infarction, esophageal injury, intraoperative hemorrhage. Nineteen patients died within 90 days of pneumonectomy. Right pneumonectomy was associated with a higher 90-day perioperative mortality of 20% (13 of 65) compared with 9% (6 of 64) for left pneumonectomy (p 0.089). Of these 19 perioperative 90-day deaths, 14 were secondary to the complications experienced after the pneumonectomy. BPF accounted for 6 perioperative deaths (right, 4; left, 2), and the 3 right-side patients with Table 4. Distribution of 61 Patients Who Experienced Downstaging With Neoadjuvant Chemoradiation Therapy and Their Associated 5-Year Survival Downstaged a Patients, No. 5-Year Survival, % By 1 stage 23 III II II I 5 By 2 stages 21 III I II CR 4 By 3 stages 17 III CR a The sample size was too small to calculate those that were downstaged from stage IV to pathologic complete response status. CR complete response. Fig 5. The 5-year survival stratified according to nodal downstaging. Node downstage 0 (solid line) was equivalent to no change in pathologic node status from preoperative staging. Node downstage 1 (dashed line) was equivalent to N2 N1 or N1 N0 downstaging. Node downstage 2 (dotted line) was equivalent N2 N0 downstaging. The numbers in the parentheses represent the patients at risk. ARDS died. An additional 5 patients died of other causes, including stroke, 1; massive aspiration, 1; pulmonary embolism, 1; failure to thrive, 1; unknown, 1. Overall 5-year survival was 33% (Fig 1). Univariate analysis of several factors on 5-year survival was performed (Table 2). Survival of right-sided (32%) vs left-sided resections (34%) was not significantly different (Fig 2). The 5-year survival according to the postoperative stages was CR, 48%; I, 43%; II, 26%; III, 27%; 4, 25% (Fig 3). Differences between the individual stages were not significant, although the observed survival for the CR and stage I patients individually approached significance compared with stage III patients (p and p , respectively). The 5-year survival of patients according to pathologic N status only was N0, 42%; N1, 26%; N2, 28%. Despite the relatively small sample size, differences in survival between N0 and N1, and N0 and N2 approached significance (p and p , respectively; Fig 4). Of the 54 patients who had N2 disease, 34 were right-sided and 20 were left-sided. The percentage of patients alive at 5 years after resection with positive N2 disease was 21% (7 of 34) for right-sided pneumonectomy and 30% (6 of 20) for left-sided pneumonectomy. Tumor downstaging occurred in 61 patients (Table 3). The 5-year survivals improved according to degree of downstaging that occurred (Table 4). When nodal downstaging was specifically evaluated, 5-year survivals according to nodal downstaging were downstaged from N2 to N0, 43%; downstaged from N1 to N0 or N2 to N1, 33%; no change in N2 or N1 status, 29% (Fig 5). Again,

5 Ann Thorac Surg KIM ET AL 2009;88: PNEUMONECTOMY AFTER CHEMORADIATION 941 Fig 6. Survival stratified according to the development of complications (dashed line) or no complications (solid line). The numbers in the parentheses represent the patients at risk. despite the obvious trend, significant differences were not observed. The 5-year survival was stratified according to the development of complications. When patients did not experience any complications, the 5-year survival was 39%. When any major complication occurred, the 5-year survival was 21%. This difference was significant (p 0.002; Fig 6). When the major complications were limited to BPF and ARDS, the 5-year survival of this cohort was 17%. This decreased survival was significantly different compared with those who did not experience a complication (p ). This was in contrast to those with non-bpf or non-ards complications, whose 5-year survival was 28%. Unlike the BPF or ARDS complications, this difference in survival was not significant (p ). The ARDS patients all died within 90 days. Pathologic node status and major complications were included into the multivariate analysis based on the univariate statistical analysis. Age, postoperative stage, and lymph node downstaging were additional covariates that were included because of their perceived clinical importance. Multivariate analysis showed that no variables negatively effected the 5-year survival for patients undergoing left-sided pneumonectomy. Conversely, multivariate analysis demonstrated that only the development of major complications negatively effected the 5-year survival for patients undergoing right-sided pneumonectomy (hazard ratio, 0.462; p ). Comment Chemoradiation therapy in the neoadjuvant setting for lung cancer is widely used. The Intergroup 0139 trial suggested a mild benefit of neoadjuvant chemoradiation therapy, followed by surgery. However, the subset analysis demonstrated a higher perioperative mortality rate of 26% (14 of 54) was associated among patients undergoing a pneumonectomy. More specifically, the analysis demonstrated that a large percentage of the deaths occurred in patients undergoing a right-sided pneumonectomy. The overall mortality associated with pneumonectomy has been reported to be 6% to 12% [7 10]. Despite a few reports citing 30-day and 90-day mortality rates within ranges similar to when no neoadjuvant chemotherapy has been used [11 14], the addition of neoadjuvant therapy to pneumonectomy has been thought to increase the already elevated mortality [5, 15]. Right-sided pneumonectomy specifically has been associated with a mortality of 12% to 37%, even in the absence of neoadjuvant chemoradiation therapy [7, 10, 16, 17]. When distinguished by laterality, right-sided pneumonectomy has been associated with increased mortality rates of 24% to 36% [5, 18, 19]. On the other hand, there has been evidence to the contrary as other investigators have found that after neoadjuvant chemotherapy, right-sided pneumonectomy can have either a significantly lower mortality or no increased mortality compared with pneumonectomies in general [1, 13, 20 22]. The overall perioperative mortality in this study was 16%. The definition of perioperative mortality occurring within a 90-day window from the date of the operation rather than a 30-day window was because several of the deaths within 90 days were considered related to the pneumonectomy. Furthermore, the 90-day mortality rate has been substantially greater than 30-day rate and is believed to be a more accurate representation of the sequelae associated with pneumonectomies [5, 13, 15]. Perioperative deaths appeared to be intimately associated with complications that developed during the postoperative stay or during the close postoperative follow-up period. Although long-term survival did not differ, the difference in outcomes was fairly dramatic when the 90-day perioperative mortality rate between left- and right-sided pneumonectomies was evaluated. From our extensive clinical experience, we believe that the higher 90-day perioperative mortality rate for right-sided pneumonectomy simply does not represent an anecdotal finding that only approaches significance; rather, we believe this is a very real and meaningful concern that warrants more attention than the statistical analysis supports. This series represents one of the larger neoadjuvant chemoradiation therapy followed by pneumonectomy-only experiences; however, it still may have had insufficient power to drive the observed difference to significance. A review of the data (not shown) revealed that the date of the operation did not affect the morbidity and mortality observed for the entire series. Although there was some minor variability among the surgeons performing the pneumonectomies, these resections were performed, largely, in a similar manner using a stapled bronchial closure for either left- or right-sided resection, with routine pericardial fat pad or pleural flap coverage on the right side.

6 942 KIM ET AL Ann Thorac Surg PNEUMONECTOMY AFTER CHEMORADIATION 2009;88: The 30% incidence of morbidity in this series was similar to that associated with pneumonectomies after neoadjuvant therapy [11]. The findings from the multivariate analysis underscored the negative effect of the development of major perioperative complications demonstrated by the univariate analysis. This investigation supports previous observations and provides additional information with respect to the use of concurrent neoadjuvant chemoradiation therapy that suggests the development of BPFs negatively affect survival [15, 23]. This complication is more prone to develop in right-sided pneumonectomies even without neoadjuvant chemotherapy [17]. Reports showing an absence in an increased incidence of BPFs between right and left-sided pneumonectomies [13] notwithstanding, most clinicians intuitively believe and others have, in fact, demonstrated a higher BPF incidence in right-sided pneumonectomies after neoadjuvant chemoradiation therapy [15]. The multivariate analysis performed by Gudbjartsson and colleagues [24] demonstrated an association between BPF formation and right pneumonectomy after preoperative chemoradiation therapy. Observations such as these, as well as the results reported in this article, support the role of any reasonable measures to prevent bronchial stump disruption. Also, the right-sided stumps in this series were routinely buttressed with pleura or pericardial fat pad. Although it could be argued that this technique did not result in a significant decrease in the incidence of BPFs after pneumonectomy, it is possible that the incidence of BPFs could have been higher without this added measure. The 2% incidence of ARDS in this series was on the lower end of spectrum of the 0.5% to 11% incidence that has been reported after pneumonectomy, either with or without induction therapy [3, 13, 17, 24 27]. However, the 90-day mortality rate of 100% that we observed is consistent with the high perioperative mortality rate reported in other studies [6, 25 27]. All of the ARDS cases occurred after right-sided pneumonectomy, supporting the recommendations of judicious administration of intravenous fluids, prevention of mediastinal shifting, and promotion of aggressive pulmonary toilet [28]. The overall 5-year survival in this series was 33%. This figure is consistent with that reported for pneumonectomies performed after neoadjuvant therapy [21]. Downstaging of disease suggested a survival benefit in this series. This effect was greatest among patients who achieved a complete response of the primary tumor and lymph nodes (T0 N0) with neoadjuvant therapy. Even without complete tumor sterilization, the results of this study suggest that when the mediastinal lymph nodes alone were downstaged, especially to an N0 status, there was a survival benefit. This finding was consistent with other investigations of nodal downstaging positively affecting long-term survival [29, 30]. As expected, the presence of pathologically positive N2 disease in this series was associated with a worse survival compared with N0 disease. Nonetheless, survival after resection of N2 disease was still acceptable given that patients had locally advanced disease at the onset of treatment. Also, when patients with N2 disease were evaluated by laterality of resection, 6 of 20 (30%) who underwent left-sided and 7 of 34 (21%) who underwent right-sided pneumonectomy were still alive 5 years after their resections. Although not an ideal rate of survival, this suggests that pneumonectomy in the presence of pathologically positive N2 disease should not be considered as devastating as is often believed if discovered either intraoperatively or postoperatively, even when performing a right-sided pneumonectomy. However, given the high perioperative mortality rate and the lower 5-year survival, when right-sided pneumonectomy is being contemplated after neoadjuvant chemoradiation therapy, we recommend invasive mediastinal restaging to confirm the absence of persistent N2 lymph node disease. Having the benefit of this retrospective study, we believe that right-sided pneumonectomy should not be offered in the subset of patients with positive N2 involvement after therapy. Understood in this recommendation is that extenuating circumstances may justify a right-sided pneumonectomy. Despite the possibility of a potentially fatal complication, if this operation is to be considered, a lower yet still acceptable 5-year survival can still be achieved. It is up to the discretion of the clinician to select the optimal patient to undergo a pneumonectomy under these specific conditions. This study has several limitations. First, as with all retrospective studies, the data were collected post hoc. Second, a tissue diagnosis for mediastinal lymph node involvement was not secured in most patients. The argument against this was that we showed a survival with positive N2 disease to be acceptable. These findings were consistent with Mansour and colleagues [22] reported 5-year survivals for the patients who underwent induction chemotherapy. Clinical criteria of N2 disease using imaging studies alone have been reported in investigations that have demonstrated the absence of any detrimental effect of neoadjuvant therapy [3, 14]. Third, an attempt was made to collect pulmonary function studies of the patients undergoing surgical resection. However, these data were incomplete or not collected in approximately one-third of the study population. Therefore, it was not possible to determine preoperative pulmonary condition of the patients. Similarly, a fourth limitation was that it was difficult to determine if other preoperative comorbidities that were documented in patients preoperatively contributed to the deaths observed in the postoperative period. Therefore, our survival did not represent a disease-specific mortality. In reality, had the patients who died from their other comorbidities been excluded, this would have only increased the observed survival. Finally, the downstaged groups that were compared represented a heterogeneous group of patients, and this heterogeneity may have diluted the comparisons made regarding downstaging. In conclusion, this study demonstrated that with respect to pneumonectomy, side does matter under certain circumstances. In general, long-term survival can be

7 Ann Thorac Surg KIM ET AL 2009;88: PNEUMONECTOMY AFTER CHEMORADIATION 943 achieved after concurrent chemoradiation therapy irrespective of side of pneumonectomy. The greatest survival benefit derived from neoadjuvant therapy resulted when there was a complete response or nodal downstaging. The risk of early perioperative death, BPF, and ARDS was higher for right-sided pneumonectomy, but overall, 5-year survival did not differ between right-sided and left-sided resections. The development of any major perioperative complication, however, negatively affected long-term survival after right-sided pneumonectomies. Even when positive mediastinal lymph nodes were removed, reasonable long-term survival was achieved. However, we believe that when right-sided pneumonectomy is being contemplated after neoadjuvant therapy, the presence of persistent positive N2 disease should preclude resection due to the high perioperative morbidity and lower long-term survival. This reason alone, above all others, is justification for performing invasive mediastinal restaging before resection is considered. References 1. Brouchet L, Bauvin E, Marcheix B, et al. Impact of induction treatment on postoperative complications in the treatment of non-small cell lung cancer. J Thorac Oncol 2007;2: Kim DJ, Lee JG, Lee CY, Park IK, Chung KY. Long-term survival following pneumonectomy for non-small cell lung cancer: clinical implications for follow-up care. Chest 2007; 132: Siegenthaler MP, Pisters KM, Merriman KW, et al. Preoperative chemotherapy for lung cancer does not increase surgical morbidity. Ann Thorac Surg 2001;71: ; discussion Venuta F, Anile M, Diso D, et al. Operative complications and early mortality after induction therapy for lung cancer. Eur J Cardiothorac Surg 2007;31: Martin J, Ginsberg RJ, Abolhoda A, et al. Morbidity and mortality after neoadjuvant therapy for lung cancer: the risks of right pneumonectomy. Ann Thorac Surg 2001;72: Albain KS, Swann RS, Rusch VR, et al. Phase III study of concurrent chemotherapy and radiotherapy (CT/RT) vs CT/RT followed by surgical resection for stage IIIA(pN2) non-small cell lung cancer (NSCLC): outcomes update of North American Intergroup 0139 (RTOG 9309) ASCO Annual Meeting Proceedings. J Clin Oncol 2005;23(16 suppl): Bernard A, Deschamps C, Allen MS, et al. Pneumonectomy for malignant disease: factors affecting early morbidity and mortality. J Thorac Cardiovasc Surg 2001;121: Ginsberg RJ, Hill LD, Eagan RT, et al. Modern thirty-day operative mortality for surgical resections in lung cancer. J Thorac Cardiovasc Surg 1983;86: Roxburgh JC, Thompson J, Goldstraw P. Hospital mortality and long-term survival after pulmonary resection in the elderly. Ann Thorac Surg 1991;51: Wahi R, McMurtrey MJ, DeCaro LF, et al. Determinants of perioperative morbidity and mortality after pneumonectomy. Ann Thorac Surg 1989;48: Perrot E, Guibert B, Mulsant P, et al. Preoperative chemotherapy does not increase complications after nonsmall cell lung cancer resection. Ann Thorac Surg 2005;80: Van Schil P, Van Meerbeeck J, Kramer G, et al. Morbidity and mortality in the surgery arm of EORTC trial. Eur Respir J 2005;26: Mansour Z. Kochetkova EA, Ducrocq X, et al. Induction chemotherapy does not increase the operative risk of pneumonectomy! Eur J Cardiothorac Surg 2007;31: Alifano M, Boudaya MS, Salvi M, et al. Pneumonectomy after chemotherapy: morbidity, mortality, and long-term outcome. Ann Thorac Surg 2008;85: ; discussion Doddoli C, Barlesi F, Trousse D, et al. One hundred consecutive pneumonectomies after induction therapy for nonsmall cell lung cancer: an uncertain balance between risks and benefits. J Thorac Cardiovasc Surg 2005;130: Au J, el-oakley R, Cameron EW. Pneumonectomy for bronchogenic carcinoma in the elderly. Eur J Cardiothorac Surg 1994;8: Darling GE, Abdurahman A, Yi QL, et al. Risk of a right pneumonectomy: role of bronchopleural fistula. Ann Thorac Surg 2005;79: Uy KL, Darling G, Xu W, et al. Improved results of induction chemoradiation before surgical intervention for selected patients with stage IIIA-N2 non-small cell lung cancer. J Thorac Cardiovasc Surg 2007;134: Allen AM, Mentzer SJ, Yeap BY, et al. Pneumonectomy after chemoradiation: the Dana-Farber Cancer Institute/Brigham and Women s Hospital experience. Cancer 2008;112: Sonett JR, Suntharalingam M, Edelman MJ, et al. Pulmonary resection after curative intent radiotherapy ( 59 Gy) and concurrent chemotherapy in non-small-cell lung cancer. Ann Thorac Surg 2004;78:1200 5; discussion Daly BD, Fernando HC, Ketchedjian A, et al. Pneumonectomy after high-dose radiation and concurrent chemotherapy for nonsmall cell lung cancer. Ann Thorac Surg 2006;82: Mansour Z, Kochetkova EA, Santelmo N, et al. Persistent N2 disease after induction therapy does not jeopardize early and medium term outcomes of pneumonectomy. Ann Thorac Surg 2008;86: Alexiou C, Beggs D, Rogers ML, Beggs L, Asopa S, Salama FD. Pneumonectomy for non-small cell lung cancer: predictors of operative mortality and survival. Eur J Cardiothorac Surg 2001;20: Gudbjartsson T, Gyllstedt E, Pikwer A, Jonsson P. Early surgical results after pneumonectomy for non-small cell lung cancer are not affected by preoperative radiotherapy and chemotherapy. Ann Thorac Surg 2008;86: Ruffini E, Parola A, Papalia E, et al. Frequency and mortality of acute lung injury and acute respiratory distress syndrome after pulmonary resection for bronchogenic carcinoma. Eur J Cardiothorac Surg 2001;20:30 6, discussion Alam N, Park BJ, Wilton A, et al. Incidence and risk factors for lung injury after lung cancer resection. Ann Thorac Surg 2007;84: ; discussion Tang SS, Redmond K, Griffiths M, Ladas G, Goldstraw P, Dusmet M. The mortality from acute respiratory distress syndrome after pulmonary resection is reducing: a 10-year single institutional experience. Eur J Cardiothorac Surg 2008;34: Deslauriers J, Aucoin A, Gregoire J. Postpneumonectomy pulmonary edema. Chest Surg Clin N Am 1998;8:611 31, ix. 29. Bueno R, Richards WG, Swanson SJ, et al. Nodal stage after induction therapy for stage IIIA lung cancer determines patient survival. Ann Thorac Surg 2000;70: Betticher DC, Hsu Schmitz SF, Tötsch M, et al. Mediastinal lymph node clearance after docetaxel-cisplatin neoadjuvant chemotherapy is prognostic of survival in patients with stage IIIA pn2 non-small-cell lung cancer: a multicenter phase II trial. J Clin Oncol 2003;21:

8 944 KIM ET AL Ann Thorac Surg PNEUMONECTOMY AFTER CHEMORADIATION 2009;88: DISCUSSION DR SCOTT J. SWANSON (Boston, MA): Great paper. Given the length of time over which the cases were performed, has anything changed in your management? Were there any timerelated factors in this analysis? DR KIM: The most consistent and constant change throughout the experience was probably in the surgeons performing the operations. If you look at the entire experience, we are talking about four surgeons. I m sure that how each individual surgeon approached their pneumonectomy varied to some degree. That being said, we stapled the bronchus in all of these cases. For the right side, it was routine that we performed stump coverage with either a pleural flap or pericardial fat pad. So there were certain constancies in that respect. From the chemotherapy standpoint, there were variations in the management as the regimens evolved. Ultimately, there were basically four iterations total, but they were all platinumbased regimens. Lastly, the administration of the split course radiation therapy was modified. Early in the series, our radiation oncologists would use a shorter course of therapy with higher fractionated dosages. Over time, the interval for the administration of radiation therapy increased with a concomitant lowering of the dosage. The use of three preoperative cycles did not vary. DR SWANSON: And the tests didn t cluster by time? DR KIM: No, they did not and that is actually something we reviewed. It appeared that there was an even distribution over the entire time period. DR PAUL DE LEYN (Leuven, Belgium): Thank you for this very nice presentation. It clearly shows that you can indeed perform right pneumonectomy in selected cases after chemoradiotherapy. I have two questions. My first question is your incidence of bronchopleural fistula after right pneumonectomy is 12%. This is quite high despite buttressing the bronchus. Can you tell us something about the technique, and do you use nowadays another technique to buttress the bronchial stump? And then the second question. In patients with persistent N2 disease, you had 5-year survival of 24%, which is very good. How were these patients selected? How did you stage your patients? Did you use PET? How did you restage your patients? DR KIM: With respect to your first question, technically not much has changed. We are still buttressing with pericardial fat, pleura, and mediastinal tissue. When we looked at our numbers, the incidence of BP fistulas on the right surprised us. We didn t expect it to be that high. A review of our operative records indicated that, technically, each pneumonectomy was performed in a similar manner with the aforementioned stapled closures and buttressing with pericardial fat pad, pleura, or mediastinal tissue. Muscle flaps and intercostal muscle were not used. Many of these cases were advanced cancers and extensive dissection was required. With respect to your second question regarding our 5-year survival of 28% for resections with positive N2 lymph nodes, we did not employ any special algorithm in patient selection. We relied on restaging with the CT scan after the third cycle of neoadjuvant therapy. In the early part of the series, clinically positive N2 disease was defined by CT evidence of nodes larger than 2 cm in largest dimension. Selected cases had mediastinoscopy. PET scan coupled with CT scan was a major factor in the later portion of the series. DR JEAN DES LAURIERS (Quebec, Canada): I would like to know how reliable is a mediastinoscopy after you have done a previous one and the patient has had radiotherapy over the mediastinum. I ask this question because you recommend that patients be restaged surgically which usually means by way of mediastinoscopy. From my experience, it would be very difficult to have accurate sampling and do a safe operation in such patients. DR KIM: Based on our data, we now recommend that invasive mediastinal staging be considered. Cervical mediastinoscopy does not necessarily have to be the approach used. Bronchoscopic-guided needle biopsy or EUS for certain nodal stations is certainly an acceptable alternative. Repeat cervical mediastinoscopy following neoadjuvant therapy, like repeat cervical mediastinoscopy in any other setting, warrants greater caution in avoiding the pitfalls. In the future, EBUS may be the method to initially stage mediastinal nodes with cervical mediastinoscopy reserved for restaging. DR DES LAURIERS: Is Dr Faber your mediastinoscopist? I don t know if he is he s right here, yeah. DR JOSHUA R. SONETT (New York, NY): I enjoyed your presentation. I think it s another third or fourth series that bespeaks to the safety of pneumonectomy after chemo and radiation since the Intergroup trial. I would just say that the buttressing is not all equal, and most of your complications and probably mortality were related to the BPF on the right. And if you switch to muscle flaps, either intercostal muscle or serratus, you may see that BPF rate go to an amazingly low rate. That s all, no other question.