European Journal of Cardio-thoracic Surgery 36 (2009) 1031—1036 www.elsevier.com/locate/ejcts
Applicability of the revised International Association for the Study of Lung Cancer staging system to operable non-small-cell lung cancers§ Makoto Suzuki *, Shigetoshi Yoshida, Hajime Tamura, Hironobu Wada, Yasumitsu Moriya, Hidehisa Hoshino, Kiyoshi Shibuya, Ichiro Yoshino Department of Thoracic Surgery, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8670, Japan Received 30 March 2009; received in revised form 30 March 2009; accepted 12 June 2009; Available online 18 August 2009
Abstract Objective: A new staging system for lung cancer has been proposed by The International Association for the Study of Lung Cancer Staging Committee. We assessed the feasibility of this system for surgical patients. Methods: We reviewed the surgical outcome of 1623 consecutive patients with non-small-cell lung cancer (NSCLC), who underwent pulmonary resection in our institution, with regard to the subpopulations categorised in the current and proposed (2009) systems for postoperative pathologic staging. Results: The proportion of patients staged as IIA, IIB, IIIA and IV increased, while those staged as IB and IIIB decreased. Diseases staged as IIIA or earlier were significantly increased in the new system (current system: N = 1281, 78.9% vs new system: N = 1457, 89.8%). The 5-year survival rates of patients with new stages IB and IIA were clearly dissociated with 72.5% and 51.3%, respectively (P < 0.0001). The 5-year survival rates of the newly classified T1 patients were 90.3% for T1aN0M0 and 81.5% for T1bN0M0 (P = 0.009). Re-classification of T2bN0M0 as stages IIA and T3 (same lobe nodules) N0M0 as stage IIB appropriately emphasised prognostic differences, while T4 (ipsilateral different lobe nodules) N2—3M0 (stage IIIB) and M1a (pleural effusion, stage IV) did not. Conclusions: This study demonstrated that the new system is superior to the current system in terms of the proportion and prognostic prediction of each stage, although it contains minor contradictions. Therefore, revision of the staging system will contribute to the decision for limited operation and adjuvant therapy of resected NSCLC. # 2009 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. Keywords: Non-small-cell lung cancer; Stage; Prognosis
1. Introduction Staging and classification of non-small cell lung cancers (NSCLCs) are important for appropriate patient management, prognosis estimations, facilitation of the exchange of information and comparison among institutions and disease treatment. In 1997, the International Union against Cancer/ American Joint Committee on Cancer (UICC/AJCC) published the fifth edition of the (Tumour, Node, Metastasis) TNM classification system for lung cancer [1]. The current (sixth) TNM Classification of Malignant Tumours, introduced in 2002, made no changes to the previous edition with regard to lung cancer. However, the sixth edition has some classification problems. The first involves the unbalanced patient numbers for each stage. The speed of diagnostic imaging and nationwide mass screening systems resulted in an increased number of patients with stage I NSCLC. In contrast, stage IIA NSCLC was diagnosed in only 1—4.4% [1—5]. The second drawback is the § This study was supported by a grant from the Smoking Research Foundation (M.S.). * Corresponding author. Tel.: +81 42 222 7171; fax: +81 43 226 2172. E-mail address: smakoto@faculty.chiba-u.jp (M. Suzuki).
overlapping prognosis of stages IB and IIA NSCLC patients. Survival rates for stages IB and IIA patients were 57% and 55%, respectively, 5 years after surgery [1]. The third problem is the inconsistent prognosis of cases with pulmonary metastasis [6]. Thus, there is a need to refine the international classification system based on recent trends in patient characteristics. The seventh edition of the TNM Classification for Lung Cancers, scheduled in 2009, incorporates the new staging system proposed by the International Association for the Study of Lung Cancer (IASLC) [7]. Major changes include the definition of T factor according to tumour size and transfer of the status of malignant pleural effusion or dissemination to M1a disease from T4. It has been recommended that large tumours (>7 cm) be classified as T3, that additional nodules within the same lobe as the primary tumour should also be categorised as T3, that ipsilateral intrapulmonary nodules in other lobes become T4 and that all pleural effusions, regardless of cytology, should be M1 (Table 1). In addition, T2bN0M0 cases might be moved from stage IB to stage IIA, T2aN1M0 cases from stage IIB to stage IIA and T4N0—1M0 cases from stage IIIB to stage IIIA. This retrospective study reviewed the postoperative survival of patients based on the new pathologic staging
1010-7940/$ — see front matter # 2009 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.ejcts.2009.06.025
1032
M. Suzuki et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 1031—1036
Table 1 Proposed TNM system by the International Association for the Study of Lung Cancer (IASLC).
Parentheses indicate stage of the sixth edition. * Changes which reflect the survival rate — appropriate (details in the text). # Changes which reflects the survival rate — inappropriate.
proposed by the IASLC, and investigated the feasibility of this system for surgical management of patients with NSCLC.
tional hazards model with two different models: the current (1997) and the proposed (2009) staging systems. In each analysis, P < 0.05 was deemed significant.
2. Materials and methods 3. Results 2.1. Patients 3.1. Distribution of pathologic stages This retrospective study included 1632 patients with pathologic stages IA—IV NSCLC who underwent complete resection at the Chiba University Hospital from January 1990 to December 2007. All patients underwent more than a segmentectomy with lymph nodal dissection. The clinical records of these patients were reviewed to obtain routine demographic information and to confirm the pathology staging. Pathology diagnoses were classified according to the latest TNM staging system (fifth edition of UICC/AJCC, 1997). Data extracted from each medical record included age, sex, smoking history, tumour histology and location, adjuvant therapies, date of last follow-up and death from any cause. This study was approved by the Institutional Review Board, and written informed consent was obtained from all the participants (Chiba University, IRB No. 204). The study group included 1081 (67%) men and 542 (33%) women ranging in age from 16 to 89 years (mean standard deviation, 64 9.7 years). There were 1043 (64%) adenocarcinomas, 452 (28%) squamous cell carcinomas, 75 (5%) large cell carcinomas and 53 (3%) other cancer types. Of those treated, 125 patients received adjuvant treatments consisting of chemotherapy, chemo-immunotherapy or radiotherapy. 2.2. Survival analysis The Kaplan—Meier method was used to plot survival curves, and the log-rank test was used to evaluate differences between subgroups. To evaluate the significance of the new classification as an independent prognostic factor, we performed multivariate analysis using the Cox propor-
According to the 1997 system, the group of 1623 patients fractionated as 597 (36.8%) in stage IA, 257 (15.8%) in stage IB, 50 (3.1%) in stage IIA, 175 (10.8%) in stage IIB, 202 (12.4%) in stage IIIA, 282 (17.4%) in stage IIIB and 60 (3.7%) in stage IV. Under the 2009 system, they were categorised as 597 (36.8%) in stage IA, 199 (12.3%) in stage IB, 165 (10.2%) in stage IIA, 195 (12.0%) in stage IIB, 301 (18.5%) in stage IIIA, 74 (4.6%) in stage IIIB and 92 (5.7%) in stage IV (Fig. 1). Comparing the two staging systems, the proportion of patients with stage IIA disease was markedly increased by the re-classification of stages IB and IIB based on the re-definition of tumour size (Figs. 1 and 2). Fifty-eight patients with stage IB disease were divided into the new IIA (N = 41) and IIB (N = 17) stages, and 78 patients with stage IIB disease were divided into the new IIA (N = 74) and IIIA (N = 4). Of the patients with stage IIIB disease, 170 were placed in the new IIB (N = 81) and IIIA (N = 89) stages due to the consideration of ipsilateral pulmonary metastasis of the same lobe as T-factor. Forty-three patients classified as stage IIIB due to the presence of malignant pleural effusion or dissemination were redefined as stage IV. Eleven patients diagnosed as stage IV due to pulmonary metastasis of ipsilateral different lobes were re-defined as stage IIIA (N = 6) and IIIB (N = 5). 3.2. Survival analysis Survival curves and the 5-year survival rates of each stage according to the 1997 and 2009 staging systems are shown in
M. Suzuki et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 1031â&#x20AC;&#x201D;1036
Fig. 1. Distribution of patients according to the current (1997) and proposed (2009) staging systems. The proportion of patients with stages II and IV diseases increased, while those with stages I and III diseased decreased after revision.
Fig. 3. The survival curves of patients with stage IIA and stage IIB in the current system are superimposed with the respective 5-year survival rates of 57.6% and 53.1% (P = 0.5779 by log-rank test). In addition, the respective 5year survival rates of patients with stages IIIA and IIIB disease were similar as 41.0% and 42.8% (P = 0.8909). In the 2009 system, the 5-year survival rate for patients with stage IB disease was 72.5% and 51.3% for those with IIA disease (P < 0.0001). However, the survival curve of patients with stage IIA (51.3%) was inferior to that of those with stage IIB (64.8%), although a statistical significance was not found (P = 0.0833, Fig. 3B). The survival curves of stages IIIA and IIIB
Fig. 2. Comparison between the sixth edition TNM and proposed pathologic stage groupings (IASLC). Closed circle indicates current stages that are divided into particular new stages (see Fig. 4). Dotted circle indicates new stages that are derived from particular current stages (see Fig. 5). IASLC, International Association of the Study of Lung Cancer.
1033
Fig. 3. Overall survival, expressed as the number of patients with a 5-year survival rate by pathologic stage using the sixth edition of TNM (1997 system, A) and recommendations from the International Association of the Study of Lung Cancer (2009 system, B). Survival curves for patient subgroups of stage IA stratified according to T factors (2009 system, C).
in the new system were clearly dissociated (P = 0.0169, Fig. 3B). Stage IA was not changed by the revision; however, newly fractionated T1aN0M0 and T1bN0M0 showed statistically distinguishable survival profiles (P = 0.009, Fig. 3C). The results of Cox proportional hazards regression modelling for each pathological stage of the two systems are listed in Table 2. Each stage was parameterised both as a set of indicator variables and by ordered variables adjusted for cell type, sex and age (younger than 65 vs 65 and older). Each stage grouping was tested independently and in combination with other variables. Hazard ratios for stage IIA referred to as stage IB in the 1997 and for that in 2009 systems were 1.54 (P = 0.07) and 2.12 (P < 0.0001), respectively. In addition, the hazard ratios for stage IIIB referred to as stage IIIA in each system were 0.99 (P = 0.96) and 1.48 (P = 0.0169), respectively. However, the hazard ratios for stage IIB referred to as stage IIA remained low, and the hazard ratio for stage IV referred to as stage IIIB in the new system (P = 0.82) was inconsistent with that of the current system (P < 0.0001). Further, we analysed how each stage of the current system is fractionated by the new system (Fig. 4). Four current stages (i.e., IB, IIB, IIIB and IV) are divided into several different new stages (Fig. 2; closed circles). Many survival curves were well separated, although they were not significant (Fig. 4B and D). Among them, tumour size (>5â&#x20AC;&#x201D; 7 cm) with N0, which is T2N0M0 (stage IB) in the 1997 system, was upstaged to T2bN0M0 (stage IIA) in the 2009 system (Table 1; *1). Survival of this population was poorer than that
M. Suzuki et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 1031—1036
1034
Table 2 Cox proportional hazards regression models for the sixth edition TNM and proposed pathologic stage groupings (IASLC). Comparisons
Pathologic stage groupings (IASLC) Sixth edition
IB versus IA IIA versus IB IIB versus IIA IIIA versus IIB IIIB versus IIIA IV versus IIIB
IASLC
Hazard ratio
95% CI
P
Hazard ratio
95% CI
P
1.78 1.54 1.12 1.45 0.99 2.21
1.31—2.42 0.96—2.45 0.70—1.80 1.09—1.95 0.79—1.27 1.58—3.10
0.0002 0.07 0.63 0.0120 0.96 <0.0001
1.59 2.12 0.76 2.07 1.48 0.95
0.99—1.97 1.49—3.01 0.55—1.04 1.57—2.72 1.07—2.03 0.63—1.43
0.0593 <0.0001 0.0846 <0.0001 0.0169 0.82
Adjusted for cell type (squamous/nonsquamous), sex (male/female), and age (<65/>65 years: divided by median age). IASLC, International Association of the Study of Lung Cancer; CI, confidence interval.
Fig. 4. Validity studies of dividing a single stage of the 1997 system into several stages in the 2009 system. Overall survival was calculated from the number of patients and surviving 5 years post-surgery in the new pathologic stages IB, IIA, and IIB that changed from stage IB of the current system (A); new stages IIA, IIB, and IIIA from current stage IIB (B); new stages IIB, IIIA, IIIB, and IV from current stage IIIB (C); and new stages IIIA, IIIB, and IV from current stage IV (D).
of stage IB patients in the 2009 system (Fig. 4A, P = 0.0006 by log-rank test), both of which come from the same stage IB in the 1997 system. Furthermore, same lobe nodules with N0, which is T4N0M0 (stage IIIB) in the 1997 system, was downstaged to T3N0M0 (stage IIB) in the 2009 system (Table 1; *2). Survival of this population was significantly better than that of stage IIIB (2009 system) patients (Fig. 4C, P < 0.0001), both of which come from the same stage IIIB in the 1997 system. The survival curves of several subpopulations are categorised to one stage of the 2009 system (Fig. 5). Five new stages (i.e., IIA, IIB, IIIA, IIIB and IV) consist of several current stages (Fig. 2; dotted circles). The new stage IIA, which consists of subpopulations of current stages IB and IIB, and the entire population of current stage IIA, showed similar survival curves, although stages IIB and IB tended to be inferior to stage IIA (Fig. 5A). In contrast is the new stage IIB, which consists of subpopulations of current stages IB, IIB and
IIIB. Survival curves of subpopulations of stages IB and IIIB tended to be higher than the original subpopulation of stage IIB (T2bN1M0 and T3N0M0, T3; invasion, N = 97) in the 1997 system, although these differences were not significant (Fig. 5B). In results, the survival of new stage IIA worsened while that of new stage IIB improved; thus the two curves were reversed (Fig. 3B). New stage IIIA also consists of subpopulations of current stages IIB, IIIB and IV, as well as the entire population of current stage IIIA. The survival curves were very similar (Fig. 5C). Current stages IIIB and IV had two populations with distinct survival profiles (Fig. 5D and E). The subpopulation of current stage IV that shows ipsilateral different lobe nodule with N2—3 (Table 1; #1), and subpopulation of current stage IIIB, have been categorised into a single stage IIIB in the 2009 system (Fig. 5D), but the survival of this population was quite different (P = 0.0056). The subpopulation of current stage IIIB due to pleural disease (Table 1; #2) and the subpopulation of current stage IV are
M. Suzuki et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 1031—1036
1035
Fig. 5. Validity studies of categorising subdivided stages of the 1997 system into a single stage in the 2009 system. Overall survival, expressed as the number of patients surviving 5 years post-surgery, based on the current pathologic stages IB, IIA, and IIB categorised to stage IIA of the 2009 system (A); current stages IB, IIB, and IIIB categorised to new stage IIB (B); current stages IIB, IIIA, IIIB, and IV categorised to new stage IIIA (C); current stages IIIB and IV categorised to new stage IIIB (D); and current stages IIIB and IV categorised to new stage IV (E).
grouped into a single stage IV in the 2009 system (Fig. 5E). However, the survival of these populations was different (P = 0.0524).
4. Discussion Following the re-classification of the TNM, the proportion of stages II and IV patients increased while the proportion of stages I and III patients decreased. In addition, the proportion of patients more advanced than stage IIIB decreased to 10.3% from 21.1%. Stage re-classification in our resected cases revealed that the proportion of cases at stage IIIA or earlier, which are generally indicated for resection, had increased (current system; N = 1281, 78.9% vs the new system; N = 1457, 89.8%). Therefore, the new staging system could feasibly be used to select operable cases. The new classification, which reduced the proportion of advanced cases that are usually evaluated as inoperable and systematic diseases, more
appropriately assessed the resected NSCLC cases as localised diseases. These observations suggest that appropriate changes have been made in the proportion of each stage. In the new staging system, T1 was split into T1a and T1b. The 5-year survival rate of stage IA NSCLC with tumours of 2.1—3.0 cm was reported to be significantly worse than when tumours were of 2 cm or less (68.8% and 84.4%, each P < 0.0001) [8]. In addition, intentional sublobar resection has been suggested as an alternative lobectomy for patients with c-stage IA NSCLC of 2 cm or smaller [8]. A meta-analysis of postoperative adjuvant chemotherapy with tegafur—uracil in NSCLC showed the effectiveness of the chemotherapy for patients with stage I adenocarcinoma, especially in cases with tumours of 2 cm or more [9,10]. In this study, we demonstrated that stage IA NSCLC can also be divided into two populations in terms of prognosis. The subdivision of cases with resected stage IA NSCLC by size at 2 cm suggests a useful classification for the basis of a decision on limited surgery and candidate selection for adjuvant therapy.
1036
M. Suzuki et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 1031—1036
Subdivision of T2 into T2a, T2b and T3 based on tumour size increased the proportion of new stage IIA, and made a significant difference in the survival rate between stage IB and stage IIA in the 2009 system. In detail, one portion of the new stage IIA (T2bN0M0, N = 41) was separated as the poorer prognosis group from the current stage IB (Fig. 4A). Furthermore, one portion of new stage IIIA (T3N1M0, T3: >7 cm), although only four cases, was separated as a group with more of a tendency towards poorer prognoses than the current stage IIB (Fig. 4B). Thus, the new system succeeded in up-staging cases with large-sized tumours, which show poorer prognoses. Subdivision and re-classification of large tumours according to size will aid in the selection of N0 candidates for adjuvant chemotherapy, since the subpopulation of current stage IB (T2 >3 cm) appears to benefit from postoperative adjuvant chemotherapy using carboplatin/ taxol [11]. The new classification seems to succeed in enrolling cases with good surgical outcome into earlier stages by reclassifying pulmonary metastases. In our series, additional nodules in the primary lobe was a particularly favourable prognostic factor, with a survival rate that was significantly better than cases in the proposed stage IIIB (Fig. 4C), very similar to current stage IIB (Fig. 5B). It has been reported that cases with pulmonary nodules have a better prognosis than other cases in stage IIIB, and that it is a good indication for resection, since such a metastatic process is considered a type of local extension [6]. Therefore, with this revision, hazard ratios are more informative with statistical significance between stages IIIB and IIIA in the 2009 system. However, cases with nodules in ipsilateral different lobes and with lymph node involvement (N2 or N3) showed poor prognosis (N = 5, Fig. 5D), so these cases may be better classified as stage IV rather than stage IIIB. Further research will be needed to address this issue. In contrast to the advantages and benefits conveyed to several stages by the new system, the revision also has minor problems. In our study, the hazard ratio between new stages IV and IIIB was nearly 1.0, and other investigators have reported that the new pathologic stage IIIB tends to have a poorer prognosis than the new pathologic stage IV [7,12]. The heterogeneous populations included in new stage IV could accounts for this problem. In fact, the chance detection of pleural effusion during surgery sometimes shows long survival after resection of the primary tumour with intrapleural chemotherapy, and contralateral pulmonary nodules occasionally include multiple primary lesions. Another problem is the reversal of the hazard ratio between stages IIB and IIA, although the change was not significant in our data. The survival of subpopulations in the two stages is relatively diverse (Fig. 5A and B). That is, prognoses of new T2bN0M0 and T2aN1M0 tend to be poorer than current stage IIA, while prognoses of T3N0M0 (T3: >7 cm and T3: same lobe nodules)
tend to be better than some of the current stage IIB (T2bN1M0 and T3N0M0, T3; invasion) cases. The biological behaviours of these stages should be evaluated for further study. In conclusion, this study demonstrated that, although there are minor contradictions, the new 2009 TNM classification system is almost superior to the currently used system in terms of proportion represented by each stage, acceptability to surgeons and prediction of prognosis at each stage. The new TNM system will facilitate decisions for limited surgery and adjuvant therapy.
References [1] Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710—7. [2] Yang P, Allen MS, Aubry MC, Wampfler JA, Marks RS, Edell ES, Thibodeau S, Adjei AA, Jett J, Deschamps C. Clinical features of 5,628 primary lung cancer patients: experience at Mayo Clinic from 1997 to 2003. Chest 2005;128:452—62. [3] Naruke T, Goya T, Tsuchiya R, Suemasu K. Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg 1988;96:440—7. [4] Inoue K, Sato M, Fujimura S, Sakurada A, Takahashi S, Usuda K, Kondo T, Tanita T, Handa M, Saito Y, Sagawa M. Prognostic assessment of 1310 patients with non-small-cell lung cancer who underwent complete resection from 1980 to 1993. J Thorac Cardiovasc Surg 1998;116:407—11. [5] Jassem J, Skokowski J, Dziadziuszko R, Jassem E, Szymanowska A, Rzyman W, Roszkiewicz A. Results of surgical treatment of non-small cell lung cancer: validation of the new postoperative pathologic TNM classification. J Thorac Cardiovasc Surg 2000;119:1141—6. [6] Yoshino I, Nakanishi R, Osaki T, Hasuda S, Taga S, Takenoyama M, Yoshimatsu T, Yasumoto K. Postoperative prognosis in patients with non-small cell lung cancer with synchronous ipsilateral intrapulmonary metastasis. Ann Thorac Surg 1997;64:809—13. [7] Goldstraw P, Crowley J, Chansky K, Giroux DJ, Groome PA, Rami-Porta R, Postmus PE, Rusch V, Sobin L. 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. [8] Okada M, Yoshikawa K, Hatta T, Tsubota N. Is segmentectomy with lymph node assessment an alternative to lobectomy for non-small cell lung cancer of 2 cm or smaller? Ann Thorac Surg 2001;71:956—60 [Discussion 61]. [9] Hamada C, Tanaka F, Ohta M, Fujimura S, Kodama K, Imaizumi M, Wada H. Meta-analysis of postoperative adjuvant chemotherapy with tegafur— uracil in non-small-cell lung cancer. J Clin Oncol 2005;23:4999—5006. [10] Kato H, Tsuboi M, Kato Y, Ikeda N, Okunaka T, Hamada C. Postoperative adjuvant therapy for completely resected early-stage non-small cell lung cancer. Int J Clin Oncol 2005;10:157—64. [11] Strauss GM, Herndon JE, Maddaus MA, Johnstone DW, Johnson EA, Harpole DH, Gillenwater HH, Watson DM, Sugarbaker DJ, Schilsky RL, Vokes EE, Green MR. Adjuvant paclitaxel plus carboplatin compared with observation in stage IB non-small-cell lung cancer: CALGB 9633 with the Cancer and Leukemia Group B, Radiation Therapy Oncology Group, and North Central Cancer Treatment Group Study Groups. Int J Clin Oncol 2008;26:5043—51. [12] Fukui T, Mori S, Hatooka S, Shinoda M, Mitsudomi T. Prognostic evaluation based on a new TNM staging system proposed by the International Association for the Study of Lung Cancer for resected non-small cell lung cancers. J Thorac Cardiovasc Surg 2008;136:1343—8.