Purpose Identification of biomarkers to predict recurrence risk is essential to improve adjuvant treatment strategies in stage II/III gastric cancer patients. This study evaluated biomarkers for predicting survival after surgical resection. Materials and Methods This post-hoc analysis evaluated patients from the CLASSIC trial who underwent D2 gastrectomy with or without adjuvant chemotherapy (capecitabine plus oxaliplatin) at the Yonsei Cancer Center. Tumor expressions of thymidylate synthase (TS), excision repair cross-complementation group 1 (ERCC1), and programmed death-ligand 1 (PD-L1) were evaluated by immunohistochemical (IHC) staining to determine their predictive values. Results Among 139 patients, IHC analysis revealed high tumor expression of TS (n=22, 15.8%), ERCC1 (n=23, 16.5%), and PD-L1 (n=42, 30.2%) in the subset of patients. Among all patients, high TS expression tended to predict poor disease-free survival (DFS; hazard ratio [HR], 1.80; p=0.053), whereas PD-L1 positivity was associated with favorable DFS (HR, 0.33; p=0.001) and overall survival (OS; HR, 0.38; p=0.009) in multivariate Cox analysis. In the subgroup analysis, poor DFS was independently predicted by high TS expression (HR, 2.51; p=0.022) in the adjuvant chemotherapy subgroup (n=66). High PD-L1 expression was associated with favorable DFS (HR, 0.25; p=0.011) and OS (HR, 0.22; p=0.015) only in the surgery-alone subgroup (n=73). The prognostic impact of high ERCC1 expression was not significant in the multivariate Cox analysis. Conclusion This study shows that high TS expression is a predictive factor for worse outcomes on capecitabine plus oxaliplatin adjuvant chemotherapy, whereas PD-L1 expression is a favorable prognostic factor in locally advanced gastric cancer patients.
Bibliographical noteFunding Information:
This study was supported by grants from the Health & Medical Technology R&D Program, Ministry of Health and Welfare, Republic of Korea (A101232, Minkyu Jung) and the National R&D Program for Cancer Control, Ministry of Health and Welfare, Republic of Korea (1520190, Sun Young Rha).
Copyright © 2019 by the Korean Cancer Association.
All Science Journal Classification (ASJC) codes
- Cancer Research