Optimal surrogate endpoint of survival for gastric cancer patients undergoing neoadjuvant chemotherapy.

Abstract

141 Background: Surrogate endpoint of survival for gastric cancer patients undergoing neoadjuvant chemotherapy (NAC) has not been established yet. To determine the most optimal surrogate endpoint, we compared the validity of outcomes according to the endoscopic or histological response criteria and other histological factors in gastric cancer patients who underwent NAC. Methods: The target population in this study was 77 gastric cancer patients who underwent NAC followed by curative gastrectomy between January 2010 and December 2016. In histological response criteria, if necrosis or disappearance of the tumor was present in more than 2/3 of the primary lesion, the case was diagnosed as a histological responder. Hazard ratio (HR) for RFS in responders according to each factor was estimated, and survival curves were compared between responders and non-responders using log-rank test. Results: The number of patients in each regimen was as follows: docetaxel/cisplatin/S-1, 40; docetaxel/oxaliplatin/S-1, 16; cisplatin/S-1, 15; S-1/oxaliplatin, 4; docetaxel/S-1, 2. The endoscopic and histological response rates were 73% (56/77) and 16% (12/77), respectively. The proportions of patients with ypT0-1 and ypN0 were 17% (13/77) and 23% (18/77), respectively. HRs were 0.761 (log-rank p = 0.43) in the endoscopic response criteria, 0.879 (log-rank p = 0.77) in the histological response criteria, 0.703 (log-rank p = 0.46) in ypT0-1, and 0.121 (log-rank p = 0.001) in ypN0, respectively. Three-year RFS rates in ypN status were ypN0/1/2/3 = 93%/48%/43%/14%. Conclusions: Pathological ypN status was the most optimal surrogate endpoint of RFS for gastric cancer patients undergoing NAC. Patients with ypN0 would be expected for long relapse-free survival.