Comparative Effectiveness of Neoadjuvant Chemoradiation Therapy Versus Chemotherapy Alone Followed by Surgery for Patients With Stage III Non-small Cell Lung Cancer (NSCLC)

Abstract

The optimal neoadjuvant therapy prior to surgical resection of stage III NSCLC is controversial, as randomized data and guidelines support both chemoradiation therapy (N-CRT) and chemotherapy (N-CTX) as preoperative treatment before surgery. Whether survival and pathologic outcomes following these two treatments are comparable in the real-world setting is unknown. In this study, we evaluated the comparative effectiveness of N-CRT versus N-CTX in stage III patients in the National Cancer Database (NCDB). Patients in the NCDB with stage III NSCLC treated with N-CRT or N-CTX and surgery between 2003-2005 were eligible for analysis. Endpoints included overall survival (OS), nodal pathologic complete response (PCR), positive surgical margins (PSM), and 30-day postoperative mortality (POPM). Stepwise Cox and logistic regression analysis were used to determine independent predictors of OS, PCR/PSM/POPM, respectively. A total of 1,559 patients were included in this cohort: 1,126 (72.2%) underwent N-CRT and 433 (27.8%) were treated with N-CTX. Patients treated with N-CRT were younger (median age 59.0 vs. 63 years, p < 0.0001), with more T3-4 tumors (51.2% vs. 38.1%, p < 0.0001). Most patients underwent lobectomy (1,069, 68.6%), and the remainder were treated with pneumonectomy (21.6%) or sublobar resection (SLR, 9.8%); surgery type was not associated with preoperative treatment. There were no difference in nodal PCR (N-CRT odds ratio, OR = 0.98, p = 0.9) or PSM (N-CRT OR = 0.79, p = 0.23). The 5-year OS for the entire cohort was 39.7% (39% N-CRT vs. 40% N-CTX, p = 0.40). On multivariable regression, factors independently associated with improved OS were: younger age (hazard ratio, HR, 0.69 below vs. above 60 years, p < 0.0001), female gender (HR = 0.82, p = 0.0024), nodal status (N0-1 vs. N2 and N3, HR = 0.83 and 0.69, respectively p = 0.024) and lobectomy versus SLR and pneumonectomy (HR = 0.55 and 0.64, respectively, p < 0.0001). Neoadjuvant CRT was not associated with improved survival (HR = 1.07, 95% confidence interval 0.93-1.23). When including nodal PCR in the model, nodal clearance replaced clinical nodal status in the model (HR = 0.72, p < 0.0001). Thirty-day POPM was 4.6%. Multivariable predictors of POPM included older age (OR = 1.68, p = 0.0392), male gender (OR = 1.89, p = 0.019), and SLR or pneumonectomy (OR = 2.99 and 2.87 versus lobectomy, respectively, p < 0.0001). Neoadjuvant chemoradiation therapy did not increase early post-operative mortality. There was no difference in OS, PSM or nodal PCR between patients treated with N-CRT or N-CTX in this large national database. Although selection bias may influence these results, this analysis suggests both treatment approaches are viable strategies. Lobectomy was associated with improved survival and postoperative mortality, reaffirming that it is the preferred post-induction surgical approach.