Survival Benefit from Adjuvant Radiotherapy by Gastric Cancer Histologic Type: A National Cohort Study

Abstract

Per INT0116, adjuvant chemoradiotherapy (CRT) after curative resection for gastric cancer (GC) has been shown to improve survival. ARTIST-2 challenged the role of adjuvant radiotherapy (RT) after D2 dissection. Few studies evaluate the prognostic impact of histologic subtype on survival and benefit from adjuvant RT. We hypothesized that survival would be different among adenocarcinoma [AC], signet ring carcinoma [SRC], and linitis plastica/diffuse type carcinoma [LPD], and for each, adjuvant RT would improve survival. The National Cancer Database (2004—2015) identified patients with AJCC stage I to III GC (AC, SRC, and LPC) who underwent upfront surgery, with and without adjuvant chemotherapy (CHT), RT, or CRT. On multivariable analysis (MVA), Cox proportional adjusted hazard ratios (aHR) compared overall survival (OS) adjusting for histology, RT (no RT vs 4500-5400cGy), stage, resection margins (positive, negative, or not evaluated), number of lymph nodes dissected (0, 1-14 nodes, 15-25, ≥26), CHT, year, age, sex, race, urban vs. rural, facility type, region, Charlson-Deyo comorbidity coefficient (CDCC), and insurance. Cox MVA for the subset of patients who received adjuvant CHT (CHT only or CRT) assessed survival by histology and receipt of CRT. Among 12,506 patients (AC: 9,353; SRC: 2,458; LPD: 695; median follow-up 5.5yrs) on MVA, OS was worse for SRC and LPD vs. AC (three-year overall survival [3yrOS] for AC = 57.2%, for SRC = 50.3%, for LPD = 46.6%; SRC vs AC aHR 1.26, P<0.001; LPD vs AC aHR 1.30, P<0.001). Receipt of adjuvant RT (N = 2,096) vs. no RT (N = 10,410) was associated with improved OS (3yrOS for RT: 55.9%; 3yrOS for no RT: 55.2%; RT vs no RT aHR 0.79, P<0.001). Other factors associated with improved OS were receipt of CHT (HR 0.91, P = 0.009), extent of LN dissection (15-25 nodes vs no nodes HR 0.85, P = 0.004; ≥26 vs no nodes HR 0.77, P<0.001), negative margins (positive margins HR 2.15, P<0.001), lower stage (stage II vs I HR 1.97, P<0.001; stage III vs I HR 2.73, P<0.001), younger age, female sex (HR 0.87, P<0.001), Asian race (vs white, HR 0.68, P<0.001), treatment at an academic center (vs. community HR 0.76, P<0.001), and lower CDCC (CDCC3 vs 0 HR 1.40, P<0.001). For patients who had adjuvant CHT and CRT (N = 3,859), OS was worse for SRC and LPD vs AC (3yrOS AC = 54.2%, SRC = 46.3%, LPD = 44.7%; SRC vs AC aHR 1.38, P<0.001; LPD vs AC aHR 1.33, P<0.001). CRT (N = 2,004, 51.9%) was associated with improved OS compared to CHT (N = 1,855) (3yrOS CRT: 56.2%; 3yrOS CHT: 46.3%; aHR 0.75, P<0.001). Stratified by histology (AC: 2,568; SRC: 971; LPD: 320), adjuvant CRT had improved OS vs. CHT (AC CRT vs CHT aHR 0.74, P<0.001; SRC CRT vs CHT aHR 0.79, P = 0.008; LPD CRT vs CHT aHR 0.73, P = 0.047). Survival differences existed among different histologic subgroups, with SRC and LPD conferring lower OS compared to AC, even when adjusting for stage and treatment patterns. AC, SRC, and LPD may benefit from adjuvant RT.