The Impact of Radiotherapy for Post-operative Patients with Stage II or Stage III Non-Small Cell Lung Cancer Utilizing the Surveillance, Epidemiology, and End Results Database

Abstract

Purpose/Objective: We investigated the impact of post-operative radiotherapy (PORT) on overall survival (OS) and disease specific survival (DSS) for patients with Stage II or III Non-Small Cell Lung Cancer (NSCLC) who had undergone a definitive surgical procedure utilizing the Surveillance, Epidemiology, and End Results Database (SEER). Materials/Methods: We selected (within SEER) adult patients diagnosed between and including 1988 and 2001 with Stage II or III NSCLC who underwent a lobectomy or pneumonectomy. Patients were excluded from analysis if they received radiotherapy other then PORT or if they had N3 nodal disease. Patients were also excluded if complete information was not available for extent of disease, course of therapy, and cause of death, if applicable. To account for peri-operative mortality, patients were excluded if survival was 3 months or less. Categorical variables included T-stage (T1/T2 vs. T3/T4), size of tumor (< 3cm vs. > 3 cm), N-stage (N0, N1, and N2), number of positive lymph nodes (less then 3 vs. 3 or more), histology (brochioalveolar carcinoma vs. other), and the use of PORT. OS and DSS were the study endpoints. Results: A total of 6953 patients with a median age of 66 years met our selection criteria, or which 3390 (48.76%) received PORT. The median follow-up for patients still alive was 3.25 years. On univariate analysis, size greater then 3 cm (p<0.0001), T3/T4 T-stage (p<0.0001), non N0 N-stage (p<0.0001), 3 or more positive lymph nodes (p<0.0001), non-brochioalveolar histology (p=0.0075), and the use of PORT (p=0.0003) were associated with worse survival. When a multivariate model was constructed utilizing all categorical variables, only non-bronchioalveolar histology (p=0.0901) and use of PORT (p=0.5548) failed to maintain significance. When DSS was examined, patients who received PORT were more likely to die of lung cancer in both univariate (p<0.0001) and multivariate analysis (p=0.0419). Subgroup analysis was performed for patients with N2 nodal disease with univariate analysis showing that patients who received PORT had increased OS (p=0.0029) and DSS (p=0.0336). The 5 year OS for patients with N2 nodal disease who received PORT was 26.9 ± 1.4%, while for other patients it was 18.7 ± 2.0%. The 5 year DSS for patients with N2 nodal disease who receive PORT was 35.0 ± 1.6%, while for other patients it was 25.8 ± 2.4%. When subgroup multivariate analysis for patients with N2 nodal disease was performed with the other categorical variables, the use of PORT maintained prognostic significance for both improved OS (HR=0.826; CI:0.736-0.926; p=0.0011) and improved DSS (HR=0.854; CI:0.751-0.970; p=0.0150). Conclusions: In our large retrospective series of almost 7000 patients with stage II and III NSCLC, PORT does not appear to impact on OS and has a significant decrease in DSS. This decrease in DSS may likely be a reflection of standard clinical practice to offer PORT to patients with early stage disease with factors associated poorer prognosis. Importantly, PORT was associated with improved outcomes for both OS and DSS in patients with N2 disease.