Abstract
IMPORTANCE: Treatment of resectable gastric cancer (RGC) uses a multimodal approach, including surgical treatment and chemotherapy with or without radiation therapy, and the optimal treatment strategy and timing of each of these modalities is unknown. OBJECTIVE: To investigate the association of various neoadjuvant and adjuvant treatment modalities with pathologic complete response (pCR), surgical margin status (SMS), and overall survival (OS) in RGC. DESIGN, SETTING, AND PARTICIPANTS: For this comparative effectiveness study, the National Cancer Database was interrogated to identify patients with RGC diagnosed from 2004 to 2015. Patients with gastric adenocarcinoma that was cT2-T4b, any N, and M0 and who underwent definitive surgical treatment were included. MAIN OUTCOMES AND MEASURES: The association of 9 treatment groups (ie, neoadjuvant chemoradiation only [nCRT], neoadjuvant chemotherapy only, adjuvant chemotherapy only [aCT], adjuvant chemoradiation only [aCRT], neoadjuvant chemotherapy and adjuvant radiation, chemotherapy with timing unknown [CTTU], chemoradiation therapy with timing unknown, radiation therapy with timing unknown (RTTU), and no perioperative therapy [NT]) with 3 end points (ie, pCR, SMS, and OS) was analyzed. The analysis was done using logistic regression and Cox proportional hazards models with adjustment for baseline characteristics. Data were analyzed from September 2019 through February 2020. RESULTS: Among 183 204 patients with RGC who were screened, 3064 patients were included in the analysis (median [IQR] age, 68 [57-77] years; 1764 [57.6%] men). There were 1584 tumors (51.7%) located in the antrum and 1539 stage 2 tumors (50.2%). On multivariable analyses among 1939 patients (owing to 137 patients with missing data for pCR and the exclusion of 988 patients with aCT and aCRT from pCR analysis), nCRT was associated with increased odds of pCR compared with NT, with the greatest odds ratio (OR) among all treatments (OR, 59.55; 95% CI, 10.63-333.56; P < .001). RTTU had the next highest OR (29.96; 95% CI, 2.92-307.53; P = .004). In multivariable analysis for OS among 3061 patients (owing to missing data for OS), CTTU was associated with decreased risk of death compared with NT (hazard ratio, [HR], 0.41; 95% CI, 0.35-0.48; P < .001), with the lowest HR, as was nCRT (HR, 0.48; 95% CI, 0.35-0.66; P < .001), with the next lowest HR. Median OS was greatest among patients treated with CTTU (53.9 months; 95% CI, 44.5-61.0 months), followed by nCRT (39.1 months; 95% CI, 26.9 months-not applicable) and aCT (36.1 months; 95% CI, 28.88-49.18 months), while 2-year OS rates were 65.6% (95% CI, 61.3%-69.5%) for CTTU, 63.6% (95% CI, 52.3%-73.0%) for nCRT, and 59.7% (95% CI, 54.2%-64.7%) for aCT. CONCLUSIONS AND RELEVANCE: This study found that nCRT was associated with the highest pCR rate, while CTTU (ie, neoadjuvant or adjuvant therapy) was associated with the greatest OS.