Implications of Lymph Node Staging on Selection of Adjuvant Therapy for Gastric Cancer in the United States: A Propensity Score-matched Analysis.

Objective: To compare the efficacy of adjuvant chemoradiotherapy (CRT) and chemotherapy alone (CA) in gastric adenocarcinoma patients undergoing gastrectomy in the United States (US).
Background: A majority of US gastric adenocarcinoma patients are inadequately staged (<15 nodes examined). Despite this, and limited data comparing adjuvant CRT with CA in US patients, national guidelines endorse CA in selected patients undergoing D2 lymphadenectomy.
Methods: Resected stage IB-III gastric adenocarcinoma patients receiving adjuvant CRT or CA (n = 3008) were identified in the National Cancer Database (1998-2006). Cox regression identified covariates associated with overall survival (OS). CRT and CA cohorts were matched (3:1) by propensity scores based on the likelihood of receiving CA. OS was compared by Kaplan-Meier estimates.
Results: Adjuvant CA was associated with an increased risk of death (HR 1.29, P < 0.001) relative to CRT. Inadequate lymph node staging (LNS) and nodal positivity were strong predictors of risk-adjusted mortality (P < 0.001). After propensity score-matching, CRT demonstrated superior median OS compared with CA (36.1 vs 28.9 m; P < 0.0001), regardless of stage. CRT was superior to CA in inadequately staged patients (33.1 m vs 24.5 m; P < 0.001); this benefit was less pronounced with increasing nodal examination. CRT improved OS in node-positive disease (29.8 vs 22.2 m; P < 0.001), regardless of LNS adequacy. In node-negative disease, OS did not differ significantly between CRT and CA cohorts; however, node-negative patients undergoing inadequate LNS benefited from CRT.
Conclusions: CRT is associated with improved stage-stratified OS compared with CA. Lymph node status and adequacy of surgical staging should influence adjuvant therapy selection in the United States.