Number of Metastatic Lymph Nodes in Resected Non Small Cell Lung Cancer Predicts Patient Survival Jin Gu Lee, MD, Chang Young Lee, MD, In Kyu Park, MD, Dae Joon Kim, MD, Seong Yong Park, MD, Kil Dong Kim, MD, and Kyung Young Chung, MD Department of Thoracic and Cardiovascular Surgery, Yonsei University College of Medicine, Seoul, and Department of Thoracic and Cardiovascular Surgery, Eulji University School of Medicine, Daejon, South Korea Background. This study was conducted to evaluate the prognostic significance of the number of lymph node metastases compared with the pathologic nodal stage (pn category) based on the anatomic extent of lymph node metastases in TNM classification of non small cell lung cancer. Methods. We reviewed 1,081 patients who underwent major pulmonary resection and were proven to be pathologic stage I through IIIA between 1990 and 2006. Patients were divided into four subgroups (nn category) according to the number of metastatic lymph nodes: those without nodal metastases were nn0, those with 1 to 3 metastatic lymph nodes were nn1 3, those with 4 to 14 were nn4 14, and those with 15 or more were nn>15. Results. The nn category followed a significant stepwise deterioration. The 5-year survival rate was 69.0% for nn0, 42.9% for nn1 3, 30.0% for nn4 14, and 11.5% for nn>15 (p < 0.001). Multivariate analysis showed that the nn category was a significant prognostic indicator similar to the pn category. Hazard ratios versus pn0 for pn1 and pn2 were 1.639 and 2.639, respectively, and 1.860, 2.029, and 4.758 for nn1 3, nn4 14, and nn>15, respectively. The nn category showed excellent agreement with the pn category ( 0.723; p < 0.001). Conclusions. We can predict patient prognosis after surgery for non small cell lung cancer according to the number of lymph nodes instead of the anatomic extent of lymph node metastases. At minimum, the number of metastatic lymph nodes adds more information to the pn category of the current TNM classification system. (Ann Thorac Surg 2008;85:211 5) 2008 by The Society of Thoracic Surgeons Nodal involvement is the most important prognostic factor in determining survival for many malignancies. These factors are expressed in the N category in the TNM classification and are grouped according to the anatomic extent and number of lymph node metastases. In the classification of lung cancer, only the anatomic extent of lymph node metastases defines the pathologic N category [1]. In the latest TNM classification, however, the number of metastatic lymph nodes is included in the definition of pn categories in breast, gastric, and colorectal cancer, and pn status shows significant correlation with prognosis [2 6]. In non small cell lung cancer (NSCLC), few studies have suggested the significance of the number of lymph node metastases [7]. In the present study, we evaluated two hypotheses regarding the prognostic significance of the number of metastatic lymph nodes: (1) the prognosis of patients can be predicted according to the number of metastatic lymph nodes instead of the anatomic extent of lymph node metastases as in other cancers such as breast, gastric, and colorectal cancer, and (2) the number of Accepted for publication Aug 9, 2007. Address correspondence to Dr Chung, 134 Sinchon-dong, Seodaemun-gu, CPO Box 8044, Seoul, South Korea, 120-752; e-mail: kychu@yumc. yonsei.ac.kr. metastatic lymph nodes can add new information to the pn category of the current TNM classification system. Patients and Methods Patients We reviewed the records of 1,478 patients with NSCLC who underwent major pulmonary resection and systematic node dissection of the hilar and mediastinal lymph nodes at our institute between 1990 and 2006. Patients with pathologic stage I through IIIA were included in this study. Patients who received preoperative induction therapy and those with fewer than 11 lymph nodes retrieved were excluded from this study because these cases could possibly have been inadequately staged. Patients who died within 1 month after surgery were also excluded. Ultimately, a total of 1,081 patients were entered in the study. Among them, lymph node metastases were found in 479 patients, and their lymph node status was assessed according to the system defined by Mountain and Dresler [8]. Pathologic staging was based on the 1997 TNM classification system [1]. The group contained 261 female and 820 male patients, with a median age of 62 years (range, 10 to 82 years). The median follow-up time was 29.3 months (range, 1.0 to 215.3 months). All patients were followed until either death or the last follow-up 2008 by The Society of Thoracic Surgeons 0003-4975/08/$34.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2007.08.020
212 LEE ET AL Ann Thorac Surg NUMBER OF METASTATIC LYMPH NODES 2008;85:211 5 Table 1. Clinical and Pathologic Patient Characteristics Variables No. of Patients (%) 62 y 475 (43.9) 62 y 606 (56.1) Male 820 (75.9) Female 261 (24.1) Type of resection Pneumonectomy 367 (34.0) Bilobectomy 139 (12.9) Lobectomy 575 (53.1) Histologic subtype Squamous cell carcinoma 530 (49.0) Adenocacinoma 411 (38.0) Other 140 (13.0) Pathologic tumor factor (pt) pt1 235 (21.7) pt2 670 (62.0) pt3 176 (16.3) Pathologic node factor (pn) pn0 602 (55.7) pn1 204 (18.9) pn2 275 (25.4) Pathologic stage IA 166 (15.4) IB 367 (34.0) IIA 34 (3.1) IIB 201 (18.6) IIIA 313 (29.0) Number of positive lymph nodes 0 602 (55.7) 1 3 289 (26.7) 4 14 167 (15.4) 15 23 (2.1) in which stratification was based solely on the number of positive nodes, ignoring the anatomic extent of the regional lymph node involvement. Although this study included only pn0 to pn2 patients (no pn3 patients), we decided to make four subgroups for the nn categories because some pn2 patients show poor survival, resembling pn3 patients, and we wanted to reflect this in the nn categories. Thus, the absence of nodal metastases was defined as nn0 (identical to pn0), the presence of 1 to 3 metastatic lymph nodes was defined as nn1 3, 4 to 14 nodes was defined as nn4 14, and 15 or more nodes was defined as nn 15. The division pattern was defined according to prognosis. Statistical Analysis The association between variables was analyzed by either 2 or Student s t test. The duration of survival was defined as the interval between the date of surgery and either the date of death or the last follow-up date. The degree of agreement between the pn and nn categories was analyzed using kappa statistics. The differences in survival among groups were examined using a log-rank test, and multivariate analyses were performed by means of the Cox proportional hazard model in variables that had probability values of less than 0.05 in the univariate analyses. A probability value of less than 0.05 was considered significant. The data were analyzed using SPSS for Windows (Statistical Package for Social Science, SPSS Inc, Chicago, IL). Results Characteristics of Lymph Nodes For the 1,081 patients, the mean number of retrieved lymph nodes was 29.6 11.5 (range, 11 to 61) per patient, and the mean number of metastatic lymph nodes was 4.4 5.0 (range, 1 to 32) in 479 patients with lymph node date (May 1, 2007). All patient characteristics are summarized in Table 1. The Institutional Review Board of Yonsei University College of Medicine approved this retrospective study. The need of a subsequent individual consent of patients whose records were evaluated was waived because individuals were not identified within the study. Analysis of Lymph Node Metastasis All retrieved lymph nodes were examined for metastases by light microscopy after being stained with hematoxylin and eosin. The number of metastatic lymph nodes from each defined anatomic region was recorded. Based on the data obtained, stratification was performed according to two different modes of lymph node status assessment: the anatomic extent of lymph node metastases (pn category) as defined by TNM classification, and the number of regional lymph nodes with metastases (nn category) Fig 1. Distribution of the number of metastatic lymph nodes in 479 pn( ) patients. Shaded bars are pn1 nodes, and striped bars are pn2 nodes.
Ann Thorac Surg LEE ET AL 2008;85:211 5 NUMBER OF METASTATIC LYMPH NODES 213 Table 2. Relationship Between pn and nn Categories Number of Patients by nn Category pn Category nn0 nn1 3 nn4 14 nn 15 pn0 602 (100%) 0 0 0 pn1 0 168 (58.1%) 36 (21.6%) 0 pn2 0 121 (41.9%) 131 (78.4%) 23 (100.0%) metastasis. The distribution of the number of lymph node metastases in pn( ) patients is shown in Figure 1. Agreement Between nn and pn Categories Excellent agreement was observed between nn and pn categories, as indicated by a kappa value of 0.723 (p 0.001). Table 2 shows the relationship between the three pn subgroups and the four nn subgroups. The mean number of metastatic lymph nodes was 2.3 1.6 in patients classified as pn1 and 6.0 6.0 in patients classified as pn2, which was a significant difference (p 0.001). Survival Rate According to Lymph Node Status The overall 5-year survival rate of all 1,081 patients was 54.4%. The 5-year survival rate was 69.0% for pn0, 46.3% for pn1, and 30.2% for pn2 (p 0.001; Fig 2). The 5-year survival rate was 69.0% for nn0, 42.9% for nn1 3, 30.0% for nn4 14, and 11.5% for nn 15 (p 0.001; Fig 3). Table 3 shows the 5-year survival rates of all pn and nn subgroups. Within the nn categories, there were no significant survival differences among the pn subgroups. With regard to pn category, pn1 classification had no survival difference among nn subgroups, but pn2 classification showed distinct survival differences among Fig 3. Overall survival curves for patients according to the number of positive lymph nodes: nn0 is solid line, nn1 3 is long dashed line, nn4 14 is short dashed line, and nn 15 is mixed short and long dashed line. (5YSR 5-year survival rate.) subgroups (nn1 3 versus nn4 14, p 0.022; nn4 14 versus nn 15, p 0.026). Prognostic Significance of nn and pn Categories In the multivariate analysis, the nn category, defined by the number of metastatic lymph nodes, was an independent prognostic factor similar to the pn category, defined by the anatomic extent of lymph node metastases. Hazard ratios versus pn0 for pn1 and pn2 were 1.639 and 2.639, respectively (Table 4), and 1.860, 2.029, and 4.758 for nn1 3, nn4 14, and nn 15, respectively (Table 5). The number of metastatic lymph nodes was analyzed as a continuous variable in multivariate analysis instead of the pn or nn category; it was a significant prognostic factor (hazard ratio, 1.067; 95% confidence interval, 1.050 to 1.084; p 0.001). Comment The number of metastatic lymph nodes was a significant prognostic factor. The nn category followed a significant stepwise deterioration and had excellent agreement with Table 3. Five-Year Survival Rates According to pn and nn Categories Survival Rate (%), Number of Metastatic Lymph Nodes pn Status nn1 3 nn4 14 nn 15 Fig 2. Overall survival curves for patients according to pn status: pn0 is solid line, pn1 is long dashed line, and pn2 is short dashed line. (5YSR 5-year survival rate.) pn1 46.0 47.6 0.735 pn2 39.0 25.4 11.5 0.001 a 0.163 0.052 a nn1 3 vs. nn4 14, p 0.022; nn4 14 vs. nn 15, p 0.026.
214 LEE ET AL Ann Thorac Surg NUMBER OF METASTATIC LYMPH NODES 2008;85:211 5 Table 4. Multivariate Analysis Including pn Category Variable Hazard Ratio 95% Confidence Interval 62/age 62 1.513 1.241 1.844 0.001 Male/female 1.212 0.933 1.576 0.150 Operation Pneumonectomy/ 1.641 1.328 2.028 0.001 lobectomy or bilobectomy T stage T2/T1 1.711 1.238 2.363 0.001 T3/T1 2.778 1.933 3.991 0.001 pn status pn1/pn0 1.639 1.252 2.146 0.001 pn2/pn0 2.639 2.096 3.323 0.001 the pn category. Based on these results, we can assume that the nn and pn categories have at least a similar significance for the prediction of patient prognosis after surgery. The current TNM staging system for NSCLC has served us well for a number of years. This system has helped us design a treatment plan and discuss patient prognosis. However, the most complex and unsatisfactory aspect of the current TNM staging system is the method of assessing nodal disease. The current nodal system, which is based on the anatomic extent of metastatic lymph nodes, can result in different prognoses even in the same nodal stage, especially with mediastinal (N2) lymph node involvement. Several subclassifications have been proposed [9 11], and the definition of the border between N1 and N2 has been challenged because of its complexity and ambiguity [12, 13]. In the latest TNM classification system, as revised in 1997, the number of metastatic lymph nodes was included in the definition of pathologic node stages in breast, gastric, and colorectal cancers [2 4]. In NSCLC, Fukui and colleagues [7] suggested the significance of the number of metastatic lymph nodes in resected NSCLC, but their evaluation was limited to pn2 lymph nodes. This study was conducted to resolve the inaccuracy of the current N staging system and to find the significance of the number of metastatic lymph nodes. An adequate number of retrieved lymph nodes is an essential component in evaluating the prognostic significance of the number of metastatic lymph nodes. Analysis of data for cancers of the colon, breast, and bladder demonstrate that the number of lymph nodes evaluated during staging is associated with postoperative survival [14 19]. In gastric cancer, the number of retrieved lymph nodes has an influence on the frequency of metastatic lymph nodes, and the number of retrieved lymph nodes reflects the reliability of the pn classification [20]. To define nodal stage in gastric cancer, at least 15 retrieved lymph nodes are necessary [3]. In lung cancer, Ludwig and associates [21] recommend that an evaluation of nodal status should include somewhere between 11 to 16 lymph nodes, based on an analysis of 16,800 patients. In accordance with this recommendation, we excluded patients with fewer than 11 lymph nodes retrieved and obtained a mean of 29.6 lymph nodes per patient. Fukui and colleagues [7] defined their four N category subgroups based on the number of lymph nodes by considering presence of mediastinal (N2) lymph node involvement and in the middle of the number regardless of their prognosis. We divided nn categories according to their prognosis. The population in this study did not include pn3 patients, and only had three pn subgroups (pn0, pn1, and pn2). We decided to make four nn subgroups because many studies indicate that some pn2 patients have poor prognosis similar to pn3 patients, and we wanted to reflect this poor survival in the new nn category. Meanwhile, similar survival differences among several divisions such as nn0, nn1 3, nn4 9, nn 10 or nn0, nn1 4, nn4 10, nn 11 were observed (data not shown), but the final nn categories of this study (nn0, nn1 3, nn4 14, nn 15) showed the most definitive differences in survival. Further discussion and study of additional populations will be needed to help clarify the best category divisions. This nn category is simple and easy to apply compared with the current pn category. Even so, the nn category system has a few limitations. It is not helpful in designing treatment preoperatively because it is based on pathologically proven node status. To date, chest computed tomograms and positron emission tomography are not able to detect positive lymph nodes, although this may change in the future with the development of new imaging devices. In addition, although the nn category showed no statistical different survivals among the pn categories, it can be heterogeneous, especially in the nn4 14 subcategory. Further discussion and study will be necessary to evaluate additional subdivisions or pattern of the divisions. Table 5. Multivariate Analysis Including nn Category Variable Hazard Ratio 95% Confidence Interval 62/age 62 1.539 1.261 1.878 0.001 Male/female 1.168 0.898 1.519 0.246 Operation Pneumonectomy/ 1.582 1.276 1.962 0.001 lobectomy or bilobectomy T stage T2/T1 1.746 1.263 2.414 0.001 T3/T1 2.768 1.923 3.984 0.001 Number of positive lymph nodes 1 3/0 1.860 1.468 2.357 0.001 4 14/0 2.029 2.029 3.474 0.001 15/0 4.758 2.912 7.774 0.001
Ann Thorac Surg LEE ET AL 2008;85:211 5 NUMBER OF METASTATIC LYMPH NODES 215 In summary, our results indicate that nn category followed a significant stepwise deterioration and had excellent agreement with pn category. Although more studies are needed, our data support the conclusion that we can predict patient prognosis after surgery for NSCLC according to the number of lymph node metastases instead of the anatomic extent of lymph node metastases. At the very least, the number of metastatic lymph nodes can add new information to the pn category of the current TNM classification system. We thank Song Vogue Ahn, MD, PhD, Department of Preventive Medicine, Yonsei University College of Medicine, for advice on statistical analyses. References 1. Mountain CF. Revision of the international system for staging lung cancer. Chest 1997;111:1710 7. 2. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual, 6th ed. New York: Springer, 2002:99 106. 3. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual, 6th ed. New York: Springer, 2002:113 24. 4. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual, 6th ed. New York: Springer, 2002:171 80. 5. Kodera Y, Yamamura Y, Shimizu Y, et al. The number of metastatic lymph nodes: a promising prognostic determinant for gastric carcinoma in the latest edition of the TNM classification. J Am Coll Surg 1998;187:597 603. 6. Ichikura T, Tomimatsu S, Uefuji K, et al. Evaluation of the New American Joint Committee on Cancer/International Union Against Cancer classification of lymph node metastasis from gastric carcinoma in comparison with the Japanese classification. Cancer 1999;86:553 8. 7. Fukui T, Mori S, Yokoi K, Mitsudomi T. Significance of the number of positive lymph nodes in resected non-small cell lung cancer. J Thorac Oncol 2006;1:120 5. 8. Mountain CF, Dresler CM. Regional lymph node classification for lung cancer staging. Chest 1997;111:1718 23. 9. Andre F, Grunenwald D, Pignon JP, et al. Survival of patients with resected N2 non-small-cell lung cancer: evidence for a subclassification and implications. J Clin Oncol 2000; 18:2981 9. 10. Ruckdeschel JC. Combined modalities therapy of non-small cell lung cancer. Semin Oncol 1997;24:429 39. 11. Osaki T, Nagashima A, Yoshimatsu T, Tashima Y, Yasumoto K. Survival and characteristics of lymph node involvement in patients with N1 non-small cell lung cancer. Lung Cancer 2004;43:151 7. 12. Okada M, Sakamoto T, Yuki T, et al. Border between N1 and N2 stations in lung carcinoma: lessons from lymph node metastatic patterns of lower lobed tumors. J Thorac Cardiovasc Surg 2005;129:825 30. 13. Zielinski M, Rami-Porta R. Proposals for changes in the Mountain and Dresler mediastinal and pulmonary lymph node map. J Thorac Oncol 2007;2:3 6. 14. Compton C, Fielding L, Burgart L, et al. Prognostic factors in colorectal cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000;124:979 94. 15. Tepper J, O Connell M, Niedzwiecki D, et al. Impact of number of nodes retrieved on outcome in patients with rectal cancer. J Clin Oncol 2001;19:157 63. 16. Herr H, Bochner B, Dalbagni G, Donat M, Reuter V, Bajorin D. Impact of the number of lymph nodes retrieved on outcome in patients with muscle invasive bladder cancer. J Urol 2002;167:1295 8. 17. Herr H. Extent of surgery and pathology evaluation has an impact on bladder cancer outcomes after radical cystectomy. Urology 2003;61:105 8. 18. Polednak A. Survival of lymph node-negative breast cancer patients in relation to number of lymph nodes examined. Ann Surg 2003;237:163 7. 19. Weir L, Speers C, D yachkova Y, Olivotto I. Prognostic significance of the number of axillary lymph nodes removed in patients with node-negative breast cancer. J Clin Oncol 2002;20:1793 9. 20. Hermanek P. ptnm and residual tumor classification: problems of assessment and prognostic significance. World J Surg 1995;19:184 90. 21. Ludwig MS, Goodman M, Miller DL, Johnstone PA. Postoperative survival and the number of lymph nodes sampled during resection of node-negative non-small cell lung cancer. Chest 2005;128:1545 50.