Survival Benefit of Surgery Following Chemoradiation Therapy for Stage III NSCLC Is Dependent on Achieving Pathologic Nodal Clearance

Abstract

Patients (pts) with stage III non-small cell lung cancer (NSCLC) experience high rates of locoregional (LR) and distance recurrence. Although surgery (S) following chemoradiation therapy (CRT) has been shown to improve LR control and progression-free survival (PFS), its impact on overall survival (OS) is unclear. Nodal pathologic complete response (PCR) at the time of surgery is a strong predictor of survival, but the optimal management of pts with residual nodal disease is debated. The objective of this retrospective study was to compare survival outcomes in pts treated with CRT and CRT + S, focusing on results as a function of nodal clearance. Pts with stage III NSCLC who were treated with CRT +/- S at our institution from Dec 2004 through Aug 2012 were included for analysis. For pts treated with CRT + S, nodal response was dichotomized into PCR vs not (N-PCR). Overall survival, PFS, and distant metastases-free survival (DMFS) were determined using Kaplan-Meier statistics and Cox regression analysis. We determined cumulative incidences of locoregional recurrence (LRR), with death as a competing risk. A total of 204 pts were eligible for analysis, (69% definitive CRT and 31% CRT + S). There was a slight female preponderance (51%). Median age of the whole cohort (WC) was 66 years. Stage distribution for WC was 52% IIIA and 48% IIIB, and for CRT + S it was 71% IIIA and 29% IIIB. Among CRT + S pts, there were 75% PCR and 25% N-PCR. Median follow-up for surviving pts was 37.3 months (mo) with a median OS for WC, CRT, and CRT + S pts of 26.3, 21.4, and 80.6 mo (log rank p < 0.0001). Median OS for PCR (83.2 mo) was superior to N-PCR (15.1 mo), p < 0.0001. CRT pts and N-PCR had no difference in OS (p = 0.79). On multivariate analysis (MVA) the difference between CRT and CRT + S remained significant (HR = 0.50, p = 0.0049) after adjusting for histology, stage, age, and gender. However, stratification of CRT + S pts by nodal response showed no difference between CRT and N-PCR (HR = 0.81 favoring CRT, p = 0.52). The PFS for WC, CRT, CRT + S, PCR, and N-PCR pts were 9.9, 9.1, 22.7, 49.2, and 7.1 mo, respectively. On MVA, the PFS for PCR pts was significantly better vs CRT (p < 0.0001), but there was no difference between CRT and N-PCR (p = 0.26). The benefit in DMFS was limited to PCR vs CRT (median 62.3 mo vs 15.6 mo, adjusted HR = 0.244, p < 0.0001), while the adjusted HR favored CRT vs N-PCR (HR = 0.63, p = 0.1151). There was a trend toward decreased LRR for CRT + S vs CRT (Gray’s p = 0.065), and no difference in LRR between CRT and N-PCR (p = 0.34). In agreement with previous studies, pts with PCR experienced markedly superior survival outcomes. However, pts who did not achieve nodal clearance fared no better than the CRT cohort, emphasizing the importance of preoperative nodal evaluation. Accurate assessment of nodal status prior to post-induction surgery may provide an opportunity to guide treatment decisions.