Lung cancer involving neighboring structures is classified

GENERAL THORACIC Subcategorization of Resectable Non-Small Cell Lung Cancer Involving Neighboring Structures Noriaki Sakakura, MD, Shoichi Mori, MD, Futoshi Ishiguro, MD, Takayuki Fukui, MD, Shunzo Hatooka, MD, Masayuki Shinoda, MD, Kohei Yokoi, MD, and Tetsuya Mitsudomi, MD Department of Thoracic Surgery, Aichi Cancer Center Hospital, and Division of Thoracic Surgery, Nagoya University Graduate School of Medicine, Nagoya, Japan Background. Although the prognoses of patients with resectable lung cancer involving neighboring structures vary, the current tumor-nodes-metastasis (TNM) classification system does not elucidate criteria for tumor subcategorization. Methods. We studied 196 consecutive patients who underwent resection of non-small cell lung cancer involving neighboring structures at the Aichi Cancer Center Hospital and were diagnosed as pathologic T3 diseases using the current staging system. Tumors were divided into six groups based on the involved neighboring structures: parietal or mediastinal pleura, subpleural soft tissue, ribs, main bronchus, pericardium, and diaphragm. Results. The overall 5-year survival rate was 39.8%. The survival rates for the six groups were: pleura (n 62), 54.8%; subpleural soft tissue (n 50), 30.0%; rib (n 25), 24.0%; main bronchus (n 33), 48.5%; pericardium (n 14), 21.4%; and diaphragm (n 12), 33.3%. The combined pleura and bronchus groups (n 95) demonstrated significantly better survival outcome than the other groups (n 101): 52.6% and 27.7%, respectively p ( 0.0002). Furthermore, among 108 patients with pt3n0 (stage IIB) disease, the prognostic difference between the pleura and bronchus groups (n 50) and the other groups (n 58) was significant: 64.0% and 25.9%, respectively p ( < 0.0001). Similar results were confirmed in patients with complete resection (n 159). Conclusions. Subcategorization of resectable lung cancer involving neighboring structures resulted in tumor groups infiltrating pleura or main bronchus, and those involving subpleural structures, pericardium, or diaphragm. (Ann Thorac Surg 2008;86:1076 83) 2008 by The Society of Thoracic Surgeons Lung cancer involving neighboring structures is classified as T3 or T4 disease in the current tumor-a heterogeneous group [11, 12]. Furthermore, although bronchus within 2 cm of the carina are considered to be nodes-metastasis (TNM) classification for lung cancer, tumors involving the diaphragm are defined as T3, some last revised in 1997 [1, 2]. The T3 tumors are defined as authors have postulated that this classification should be those involving resectable structures (chest wall, pericardium, diaphragm, or main bronchus within 2 cm of the The T3N0M0 tumors have been shifted from stage IIIA to revised [13 15]. carina), while T4 tumors are those invading organs con-iisidered to be unresectable (heart, great vessels, trachea neity of T3 tumors, however, the classification of all in the current staging system [1, 2]. Given the heteroge- and carina, esophagus, or vertebral body). T3N0M0 tumors into stage IIB has been considered to Numerous studies have reported on the various aspects of the prognoses of T3 diseases. Among the tumorssystem was finally proposed by the International Associa- require further examination [10]. The next TNM staging involving the chest wall, those with infiltration limited tion for the Study of Lung Cancer (IASLC) International to the parietal pleura reportedly have better prognoses Staging Committee in 2007 [16, 17]. In this system, current than those with involvement extending to subpleural soft T2 tumors larger than 7 cm and current T4 tumors with tissue or ribs [3 5]. In contrast, some reports suggest that pulmonary metastases in the same lobe are both categorized as T3 lesions, together with the current T3 tumors. The the depth of invasion does not influence prognoses [6 9]. Regarding tumors invading the mediastinum, it has been treatment strategies for lung cancer patients will be revised, suggested that patients with such tumors have worse in the near future, on the basis of this staging system. Here, prognoses than those with tumors invading the osseous chest wall [10]. Alternatively, tumors involving the main Accepted for publication June 9, 2008. Presented at the Forty-fourth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 30, 2008. Address correspondence to Dr Sakakura, Division of Thoracic Surgery, Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya, Aichi, 466-8550, Japan; e-mail: nskkr@med.nagoya-u.ac.jp. we focused on resectable lung cancer involving neighboring structures and developed a subcategorization criteria for such tumors. Patients and Methods This study was approved by the Institutional Review Board of the Aichi Cancer Center Hospital, and patient consent was waived because of the study s retrospective 2008 by The Society of Thoracic Surgeons 0003-4975/08/$34.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2008.06.034

Ann Thorac Surg SAKAKURA ET AL 2008;86:1076 83 SUBCATEGORIZATION BY INVOLVED STRUCTURES Table 1. Patient Characteristics Variables All Patients (n 196) Patients With Complete Resection (n 159) Median age (years) (range) 63.0 (28 81) 63.6 (28 81) Gender: Male 166 135 Female 30 24 Histology: Adenocarcinoma 77 58 Squamous cell carcinoma 93 82 Large cell carcinoma 19 13 Adenosquamous carcinoma 7 6 Surgical procedure: Lobectomy 150 125 Peumonectomy 37 32 Segmentectomy 9 2 Chemotherapy and radiotherapy: Preoperative Yes 30 25 No 166 134 Postoperative Yes 42 27 No 154 132 Tumor size: 5 cm 107 101 5 cm 89 58 Node status: N0 108 91 N1 41 34 N2 47 34 T3 structures involved: Pleura (parietal or mediastinal) 62 53 Subpleural soft tissue 50 39 Ribs 25 19 Main bronchus 33 28 Pericardium 14 11 Diaphragm 12 9 design. We studied 196 consecutive patients who underwent resection of non-small cell lung cancer involving neighboring structures between January 1980 and December 2002 at the Aichi Cancer Center Hospital and were diagnosed as pathologic T3 (pt3) diseases using the current TNM staging system. The characteristics of the patients are shown in Table 1. Histologic typings were assigned according to the World Health Organization criteria [18]. Tumors were divided into six groups based on the involved neighboring structures that were pathologically determined: parietal or mediastinal pleura, subpleural soft tissue, ribs, main bronchus, pericardium, and diaphragm. Based on the neighboring structures involved, the groups were defined as follows: pleura group, tumors infiltrating parietal or mediastinal pleura without involvement of deeper structures; subpleural soft tissue 1077 group, tumors involving subpleural fat tissues, nerves, small vessels, and intercostal muscles; rib group, tumors involving ribs regardless of whether their involvement is limited to the periostea or extends to the bone cortex; main bronchus group, tumors invading main bronchus within 2 cm of the carina; pericardium group, tumors involving pericardium regardless of whether their involvement is limited to or penetrates the pericardium; and diaphragm group, tumors involving the diaphragm regardless of whether their involvement is limited to diaphragmatic pleura or extends to diaphragmatic muscle layer. The first three groups are categorized based on the degree of the depth of invasion, while the latter three groups are categorized based on the organs involved. When tumors involved two different structures, they were assigned to the group which would be considered to have the worse prognoses based on previous reports [10 15]. As a result, two cases that involved both the main bronchus and pericardium were included in the pericardium group, two other cases that invaded both the pericardium and diaphragm were included in the diaphragm group. All pathologic information was obtained by reviewing the pathologic reports of each patient. Concerning surgical procedures for chest wall involvement, when the pleura into which the tumor infiltrated was easily mobilized from the subpleural structures, extrapleural resection was selected. In other cases, combined en bloc resection was performed at the discretion of the operating surgeon. All patients routinely underwent systematic hilar and mediastinal lymph node dissection. Whether or not chemotherapy and radiotherapy were administered differed considerably according to the study period, and the chemotherapy regimen also varied widely. In cases of incomplete resection, radiation therapy was performed if possible. Taking the new TNM staging system proposed by IASLC [16, 17] into consideration, the prognoses of following additional patients were also studied. During the same study period at the Aichi Cancer Center Hospital, 62 patients with resected non-small cell lung cancer were diagnosed as pt2 greater than 7 cm, and 49 patients were identified as pt4 with pulmonary metastases in the same lobe using the current staging system. Their tumors did not invade neighboring structures, but they would be assigned to have pt3 disease according to the new staging system. Accordingly, we analyzed their survival for comparison. Statistical Analysis The overall survival rate was calculated using the Kaplan-Meier method. The survival duration was defined from the date of operation to the last known date of survival or all-cause death. The difference in survival rates was tested using the log-rank test. The Cox proportional hazards model was used for multivariate analysis. All statistical analyses were performed using StatView for Windows (version 5.0; SAS Institute, Cary, NC). GENERAL THORACIC

GENERAL THORACIC 1078 SAKAKURA ET AL Ann Thorac Surg SUBCATEGORIZATION BY INVOLVED STRUCTURES 2008;86:1076 83 Results Surgical Outcomes The overall 5-year survival rate of the 196 patients was 39.8%. The survival rate was 41.6% for lobectomies, 37.8% for pneumonectomies, and 22.2% for segmentectomies. There were no significant prognostic differences according to gender, surgical procedure, and tumor histology. There were 6 hospital deaths; causes of death included bleeding, interstitial pneumonia, bacterial pneumonia, and bronchopleural fistula. Completeness of resection crucially affected the prognosis. The 5-year survival rates for 159 patients (81.1%) with complete resections and 37 patients (18.9%) with incomplete resections were 44.0% and 21.6%, respectively (p 0.0001). Thirty patients had received induction chemotherapy and radiation therapy. Among them, 9 patients had clinical N2 disease and the postinduction ycn0-1 was obtained in 4 patients. Postoperative adjuvant therapy was performed in 42 patients. Among 159 patients with complete resection, there was no significant prognostic difference between with (n 47) and without (n 112) preoperative or postoperative chemotherapy and radiotherapy (46.8% and 42.9%, respectively; p 0.7188). Among 37 patients with incomplete resections, 17 patients treated with chemotherapy and radiotherapy showed a better 5-year survival rate than those who did not receive these therapies (35.3% and 10.0%, respectively; p 0.0782). Prognoses Based on Tumor Size and Nodal Status The 5-year survival rates according to tumor size ( 5 cm and 5 cm) were 41.1% and 38.2%, respectively. The prognosis of patients with tumors measuring 5 cm or smaller tended to be better than those with tumors larger than 5 cm, but there was no significant statistical difference between the two groups (p 0.7229). The 5-year survival rates for patients with N0, N1, and N2 diseases were 43.5%, 51.2%, and 21.3%, respectively. Patients with N0 disease showed unexpectedly slightly worse outcomes than those with N1 disease, but there was no significant statistical difference between the N0 and N1 diseases. Alternatively, the difference of the survival rates between N0-1 and N2 diseases was significant (45.6% and 21.3%, respectively; p 0.0027). Prognoses Based on Involved Neighboring Structures The pleura group had a 5-year survival rate of 54.8%. Seventeen patients had N2 disease. Extrapleural resection was the most preferable intervention and was performed in 49 patients, and combined en bloc resection of the chest wall in 13 patients. Complete resection was achieved in 53 patients (85% of this group); their 5-year survival rate was 53.8%. In the subpleural soft tissue group, the 5-year survival rate was 30.0% and 17 patients had N2 disease. Tumor invasion was limited to the subpleural fat in 22 patients, whereas it extended to the intercostal muscle layer in 28 patients. Extrapleural resection was performed in 21 patients and en bloc resection in 29 patients, including 8 cases of superior sulcus tumor. Thirty-nine patients underwent complete resection (78% of this group), with a survival of 35.9% at 5 years. The rib group showed a 5-year survival rate of 24.0%. No patient had N2 disease. All patients underwent en bloc resection of the chest wall. There were 8 cases of superior sulcus tumor. Complete resection was achieved in 19 patients (76% of this group), with a survival rate of 26.3%. The 5-year survival rate was 48.5% in the main bronchus group, and 6 patients were identified with N2 disease. Operative procedures performed included lobectomy in 19 patients and pneumonectomy in 14. Bronchoplasty was performed in all cases of lobectomy (4 sleeve and 15 wedge resections) and 3 patients underwent Fig 1. Survival curves of all 196 patients based on the neighboring structures involved.

Ann Thorac Surg SAKAKURA ET AL 1079 2008;86:1076 83 SUBCATEGORIZATION BY INVOLVED STRUCTURES Table 2. Univariate and Multivariate Analyses of Prognostic Factors Univariate Analysis Multivariate Analysis No. of Patients 5-Year Survival (%) p Value HR 95% CI p Value GENERAL THORACIC Resection Complete 159 44.0 Incomplete 37 21.6 0.0001 2.272 0.286 0.678 0.0002 Tumor size 5 cm 107 41.1 5 cm 89 38.2 0.7229 1.065 0.735 1.544 0.7395 Node status N0-1 149 45.6 N2 47 21.3 0.0027 1.698 0.398 0.871 0.0080 Invasion Pleura or Bronchus 95 52.6 Others 101 27.7 0.0002 1.901 0.361 0.767 0.0008 CI confidence interval; HR hazard ratio. Fig 2. Survival curves of all patients based on the subcategorization of pt3 tumors involving neighboring structures (A) and survival curves of patients with pt3n0 (stage IIB) disease (B).

GENERAL THORACIC 1080 SAKAKURA ET AL Ann Thorac Surg SUBCATEGORIZATION BY INVOLVED STRUCTURES 2008;86:1076 83 pneumonectomy (1 sleeve resection and 2 wedge resections). Complete resection was performed in 28 patients (85% of this group), and their survival rate was 54.7%. The pericardium group showed a 5-year survival rate of 21.4%. Five patients had N2 disease. Pneumonectomy was the most common intervention and was performed in 8 patients. Eleven patients (79% of this group) underwent complete resection with a survival of 27.3% at 5 years. The diaphragm group had a 5-year survival rate of 33.3%. Two patients had N2 disease. Combined en bloc resection of the diaphragm was performed in all patients. Complete resection was achieved in 9 patients (75% of this group); their survival rate was 40.4%. structures and those of the two new pt3 diseases: current T2 tumors larger than 7 cm and current T4 tumors with pulmonary metastases in the same lobe. The overall 5-year survival rates of all new pt3 (n 307) and pt3n0 (stage IIB) (n 165) diseases were 36.5% and 42.4%, respectively. The prognoses of pt3 diseases vary considerably (p 0.0023 between all groups) (Fig 3A). Among stage IIB diseases, patients with tumors infiltrating pleura or main bronchus had markedly better prognoses than the other three groups, while patients with tumors involving other neighboring structures demonstrated significantly worse outcomes than the other three groups (p 0.0004 between all groups) (Fig 3B). Subcategorization of Lung Cancer Involving Neighboring Structures Figure 1 shows the survival curves based on the neighboring structures involved. The 5-year survival rate was most favorable in the pleura group, followed in descending order by the main bronchus, diaphragm, subpleural soft tissue, rib, and pericardium groups. The prognoses of the six groups varied considerably (p 0.0100 between all groups), and the pleura and main bronchus groups demonstrated markedly better prognoses than the other groups. The combined pleura and main bronchus groups (n 95) demonstrated significantly better 5-year survival outcome than the other groups (n 101): 52.6% and 27.7%, respectively (p 0.0002) (Fig 2A). Among the patients without lymph node metastasis (pt3n0), who were classified as having stage IIB disease according to the current staging system, the survival rates for patients with tumors infiltrating pleura or main bronchus (n 50) and those with tumors involving other neighboring structures (n 58) were 64.0% and 25.9%, respectively (p 0.0001) (Fig 2B). Similar results were obtained among the 159 patients with complete resection. The survival rates for the combined pleura and main bronchus groups (n 81) and the other groups (n 78) were 54.3% and 33.3%, respectively (p 0.0044), and the survival rates for patients with N0 diseases infiltrating pleura or main bronchus (n 47) and those with N0 diseases involving other neighboring structures (n 44) were 61.7% and 29.5%, respectively (p 0.0009). Statistical analyses are summarized in Table 2. The disease could not be subcategorized on the basis of tumor size. In univariate analysis, the patients with N2 disease showed significantly worse prognosis than those with N0-1 diseases (p 0.0027), but the significance decreased in multivariate analysis (p 0.0080). Multivariate analysis revealed that both the completeness of resection (p 0.0002) and the subcategorization of the neighboring structures involved (p 0.0008) were much more significant and crucial predictors of pt3 lung cancer involving the neighboring structures. Prognoses of pt3 Diseases Defined Using the New TNM Staging System Figure 3 shows the survival curves of the presently discussed currrent pt3 diseases involving neighboring Comment Until now, analyses of T3 lung cancer involving neighboring structures have often been conducted for each subgroup independently. Therefore, prognostic variation in each subgroup has been demonstrated; however, the optimal method of integrating or subcategorizing these subgroups through which the staging classification adequately reflects the characteristics of tumors has not yet been determined. Detterbeck and Socinski [10] suggested that the heterogeneity of these tumors should be fully recognized in order to establish a more appropriate staging system. Based on their careful review of many previous reports, they concluded that the central-type tumors invading the mediastinal structures and the superior sulcus tumors should be classified as stage IIIA disease even if there was no lymph node metastasis. Table 3 shows the survival rates of patients with complete resection in previous representative reports of T3 lung cancer involving neighboring structures. Concerning tumors invading the chest wall, Magdelienat and colleagues [4] and Facciolo and colleagues [5] reported that tumors with infiltration limited to the parietal pleura have better prognoses than those extending to subpleural structures. In our study, similar results were obtained. In the study by Burkhart and colleagues [7], although there was no significant prognostic difference, the 5-year survival rate tended to be better in tumors with infiltration limited to pleura than those involving deeper structures. Alternatively, Downey and colleagues [6] and Doddoli and colleagues [9] reported that the degree of invasion hardly had any effect on the prognoses, but that the latter heavily depended on the mediastinal nodal involvement. It would be necessary to recognize that the depth of invasion may be an important prognosticator. Regarding tumors involving the main bronchus within 2 cm of the carina, it has been reported that prognoses of this tumor group are relatively favorable [11, 12]. Our surgical outcome was more favorable, which may indicate that the operative techniques for bronchial surgery have been improving. There are few studies dealing with tumors invading only the pericardium. These tumors are often analyzed among those involving mediastinal structures. Patients with such tumors reportedly have a relatively worse prognosis than those with tumors involving the osseous

Ann Thorac Surg SAKAKURA ET AL 2008;86:1076 83 SUBCATEGORIZATION BY INVOLVED STRUCTURES 1081 GENERAL THORACIC Fig 3. Survival curves for pt3 (A) and pt3n0 (stage IIB) (B) diseases defined using the new TNM staging system proposed in 2007. chest wall [10]. This may be attributed to the fact that pneumonectomy is likely to be the most frequent intervention for this group [11]. In our cases with complete resection, pneumonectomy was performed for 63% (7 of 11) of this group. Although the tumors involving the diaphragm are defined as T3, this tumor group reportedly has a worse prognosis than the other T3 diseases. Rocco and colleagues [14] and Yokoi and colleagues [15] pointed out that the current staging system for tumors involving the diaphragm should be revised. In the present study, a relatively better outcome was obtained; this may be attributed to the fact that the N2 disease was confirmed in only a few patients in the diaphragm group. Considering the previous reports, it is reasonable to classify these tumors as worse prognostic T3 disease in this study. From the viewpoint of the depth of invasion, it would be considered important to discern whether invasion is limited to or penetrates the pericardium to analyze the pericardium group, and whether invasion is limited to diaphragmatic pleura or reaches the subpleural diaphragmatic muscle layer to analyze the diaphragm group. In this study, however, due to the small sample size of these groups, the depth was not considered for subcategorizing the pericardium and diaphragm groups. Taking an overall view of Table 3, although different studies have different outcomes, we can see that the pleura and main bronchus groups tend to have better outcomes than the other groups, indicating a subcategorization criteria of T3 lung cancer involving neighboring structures. It would be also important to consider that the number of patients with N2 disease in each study may affect the prognoses. In this study, the reason for the slightly worse prognoses of patients with N0 disease compared with those with N1 disease may be due to the small number of patients with N1 disease. However, no prognostic difference between N0 and N1 diseases was also observed in other reports [3, 8]. Currently, a proper clinical evaluation of the depth of invasion is difficult and this tumor characteristic cannot easily be incorporated into the lung cancer staging sys-

GENERAL THORACIC 1082 SAKAKURA ET AL Ann Thorac Surg SUBCATEGORIZATION BY INVOLVED STRUCTURES 2008;86:1076 83 Table 3. Survival of Patients with Completely Resected T3N0-2M0 Lung Cancer T3 Organs First Author [Ref. No.] Year No. of Patients 5-Year Survival (%) Pleura Downey [6] 1999 96 37 Magdeleinat [4] 2001 66 37 Facciolo [5] 2001 28 79 Burkhart [7] 2002 29 50 Doddoli [9] 2005 138 10 45 Present study 2008 53 b 54 Subpleural tissue a Downey [6] 1999 79 34 35 Magdeleinat [4] 2001 101 15 Facciolo [5] 2001 76 54 Burkhart [7] 2002 66 32 35 Doddoli [9] 2005 171 13 41 Present study 2008 58 26 36 Main bronchus Pitz [11] 1996 46 40 Riquet [12] 2002 68 35 Present study 2008 28 55 Pericardium Pitz [11] 1996 21 c 25 Present study 2008 11 27 Diaphragm Rocco [14] 1999 15 20 Yokoi [15] 2000 55 23 Present study 2008 9 40 a Including subpleural soft tissue and ribs. b including patients with tumors infiltrating parietal or mediastinal pleura. c including patients with tumors invading mediastinal structures. tem. However, it should be emphasized that the prognostic difference between patients with tumors infiltrating pleura or main bronchus and those with tumors involving other neighboring structures may be more significant than tumor size or even distinction between N0-1 and N2 diseases. In the new TNM staging system proposed by the IASLC in 2007, two different subcategories are subsumed under the new T3 category together with current T3 tumors. In other words, the proposed T3 grouping would include more tumors than the current classification, and would constitute a collection of tumors with quite different characteristics. Because the new staging system has already been formally proposed, our investigation regarding the TNM staging system may likely present both supporting and detracting perspectives. However, the main discussion point arising from our study is that the prognoses of pt3 non-small cell lung cancer involving neighboring structures vary considerably, and it is very important to recognize that this prognostic variation remains significant even after adopting the new staging system (Fig 3). Figure 3 also indicates the reasonableness of reclassifying such T2 and T4 tumors into in the new T3 disease category. We conclude that the subcategorization of resectable lung cancer invading neighboring structures resulted in tumor groups infiltrating pleura or main bronchus, and those involving subpleural structures, pericardium, or diaphragm. Our study, of course, has limitations because of its retrospective design and the relatively small number of cases from a single institution. However, we believe that our proposed subcategorization criteria could provide more information for tumor characteristics, and hence should be validated using larger samples. References 1. International Union Against Cancer. Lung tumours. In: Sobin LH, Wittekind CH, eds. TNM classification of malignant tumours. 5th ed. New York: Wiley-Liss; 1997:91 7. 2. Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest 1997;111:1710 7. 3. Chapelier A, Fadel E, Macchiarini P, et al. Factors affecting long-term survival after en-bloc resection of lung cancer invading the chest wall. Eur J Cardiothorac Surg 2000;18: 513 8. 4. Magdeleinat P, Alifano M, Benbrahem C, et al. Surgical treatment of lung cancer invading the chest wall: results and prognostic factors. Ann Thorac Surg 2001;71:1094 9. 5. Facciolo F, Cardillo G, Lopergolo M, Pallone G, Sera F, Martelli M. Chest wall invasion in non-small cell lung carcinoma: a rationale for en bloc resection. J Thorac Cardiovasc Surg 2001;121:649 56. 6. Downey RJ, Martini N, Rusch VW, Bains MS, Korst RJ, Ginsberg RJ. Extent of chest wall invasion and survival in patients with lung cancer. Ann Thorac Surg 1999;68:188 93. 7. Burkhart HM, Allen MS, Nichols FC III, et al. Results of en bloc resection for bronchogenic carcinoma with chest wall invasion. J Thorac Cardiovasc Surg 2002;123:670 5. 8. Matsuoka H, Nishio W, Okada M, Sakamoto T, Yoshimura M, Tsubota N. Resection of chest wall invasion in patients with non-small cell lung cancer. Eur J Cardiothorac Surg 2004;26:1200 4. 9. Doddoli C, D Journo B, Le Pimpec-Barthes F, et al. Lung cancer invading the chest wall: a plea for en-bloc resection but the need for new treatment strategies. Ann Thorac Surg 2005;80:2032 40. 10. Detterbeck FC, Socinski MA. IIB or not IIB: the current question in staging non-small cell lung cancer. Chest 1997; 112:229 34.

Ann Thorac Surg SAKAKURA ET AL 2008;86:1076 83 SUBCATEGORIZATION BY INVOLVED STRUCTURES 11. Pitz CC, Brutel de la Rivière A, Elbers HR, Westermann CJ, van den Bosch JM. Results of resection of T3 non-small cell lung cancer invading the mediastinum or main bronchus. Ann Thorac Surg 1996;62:1016 20. 12. Riquet M, Lang-Lazdunski L, Le PB, et al. Characteristics and prognosis of resected T3 non-small cell lung cancer. Ann Thorac Surg 2002;73:253 8. 13. Inoue K, Sato M, Fujimura S, et al. Prognostic assessment of 1,310 patients with non-small-cell lung cancer who underwent complete resection from 1980 to 1993. J Thorac Cardiovasc Surg 1998;116:407 11. 14. Rocco G, Rendina EA, Meroni A, et al. Prognostic factors after surgical treatment of lung cancer invading the diaphragm. Ann Thorac Surg 1999;68:2065 8. 15. Yokoi K, Tsuchiya R, Mori T, et al. Results of surgical treatment of lung cancer involving the diaphragm. J Thorac Cardiovasc Surg 2000;120:799 805. 1083 16. Rami-Porta R, Ball D, Crowley J, et al; on behalf of the International Staging Committee; Cancer Research and Biostatistics; Observers to the Committee; Participating Institutions. The IASLC Lung Cancer Staging Project: proposals for the revision of the T descriptors in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol 2007;2:593 602. 17. Goldstraw P, Crowley J, Chansky K, et al; on behalf of the International Staging Committee; Cancer Research and Biostatistics; Observers to the Committee; Participating Institutions. The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM Classification of malignant tumours. J Thorac Oncol 2007;2:706 14. 18. The World Health Organization histological typing of lung tumors. 3rd ed. Geneva: World Health Organization; 1999. GENERAL THORACIC DISCUSSION DR GAETANO ROCCO (Naples, Italy): I just have one question. What were your criteria for surgical exploration of the mediastinal nodal stations in these patients? Did you perform mediastinoscopy in these patients? DR SAKAKURA: Thank you for your question. We generally use only CT [computed tomographic] scans for clinical N2 diagnosis. Mediastinoscopy is not commonly performed in Japan. We diagnosed a tumor as cn2 when the short-axis diameter of the enlarged lymph node observed in an enhanced CT scan was greater than 1 cm. Thank you. DR FRANK D. DETTERBECK (New Haven, CT): I enjoyed your presentation very much. I can t shake the feeling that there is some confounding factor between the nodal status and the T subgroups. It certainly is odd that the N0 category was worse than the N1 category, and it makes me worried that there are T substages that were more commonly associated with that N0 category, and vice versa for the T stage. I think it s difficult to sort out the 6 different T subgroups when different nodal categories are involved. I know you did a multivariate analysis, but when you did the multivariate analysis, you only did it for the two groupings of T subclassifications. If you redo the multivariate analysis with the 6 different categories of T, I wonder whether you would get a different result. Have you tried that? DR SAKAKURA: Thank you for your valuable comments. There were no significant prognostic differences according to gender, surgical procedures performed, tumor histologies, and postoperative chemotherapy and/or radiation therapy. The reason why the degree of invasion and involved organs takes precedence over the nodal status is very important and essential. I think that the small number of cases may be one reason. We did not perform multivariate analysis as per your suggestion. However, our study findings suggested that the current T3 disease may greatly depend on the degree of invasion and involved organs. Thank you.