Long-term results of pathological stage I non-small cell lung cancer: validation of using the number of totally removed lymph nodes as a staging control

Objective: The number of totally removed lymph nodes during thoracotomy was used alternatively to represent the quality of lymphadenectomy in patients with pathologic stage I non-small cell lung cancer (NSCLC). We combined this new parameter with other well-established prognostic factors and performed multivariate survival analyses to validate its usage as a stage control. Methods: Three hundred and twenty-one patients who underwent complete surgical resection for stage I NSCLC were reviewed retrospectively. Aside from the number of lymph nodes removed during thoracotomy, other well-known clinical and histopathological factors were also included as possible prognostic factors for analysis. Two survival analyses, overall death and cancer-related death as study end-point, were performed, using the Kaplan-Meier method and multivariable Cox's proportional hazard regression analysis. Stepwise method of variable selection was employed to choose the 'best' Cox proportional hazard model in each survival analysis. Results: The overall 5- and 10-year survival rates were 48 and 35%, and the cancer-related 5- and 10-year survival rate was 63.3 and 58.3%, respectively. The number of totally removed lymph nodes during thoracotomy, tumor size and smoking history in multivariable analysis significantly affected both overall and cancer-related survival rates. Cell type of adenocarcinoma or large cell carcinoma was associated with a worse cancer-related survival compared with other histological types. Conclusions: The quality of lymphadenectomy, represented quantitatively by the number of totally removed lymph nodes during thoracotomy, may impact on a more accurate tumor stage, and will affect the survival rate for patients with stage I NSCLC as well as other well known clinical and histopathological factors. (C) 2003 Elsevier B.V. All rights reserved.