The role of surgery and radiation in advanced gastric cancer: A population-based study of Surveillance, Epidemiology, and End Results database

Abstract

Background Chemotherapy is the standard approach for advanced gastric cancer, while the role of local therapy such as surgery and radiation for this population remains controversial. Our purpose is to evaluate the effect of local therapies on cancer specific survival (CSS) for advanced gastric cancer patients. Methods Four subgroups of patients in different treatment strategies: surgery, radiation (RT), surgery and radiation (Surgery+RT), no surgery/no radiation (No Surgery/No RT) were identified from the Surveillance, Epidemiology, and End Results (SEER)-registered database. The risk factors and the survival outcomes were analyzed by multivariable Cox regression models and Kaplan-Meier methods. Results A total of 10,354 patients were eligible with 6658 males and 3696 females. The 5-year CSS in the four subgroups of “Surgery”, “RT”, “Surgery+RT” and “No Surgery/No RT” were respectively 8.9%. 5.7%, 19.8% and 3.2%, which were significantly different in multivariate Cox regression (P<0.001) and univariate log-rank test (P<0.001). Advanced stage categories were defined as stage I, II and III of T/N category according to different initial T and N status following American Joint Committee on Cancer (AJCC) staging principle. Further analysis showed that patients in the group of “Surgery+RT” have significant benefits of survival specifically on stage II and III of T/N category. “Surgery+RT” group and “Surgery” group patients have similar survival time in stage I of T/N category. Moreover, we also found CSS benefits from the administration of “Surgery+RT” in the patients aged both 75 and <75 years. Remarkably, patients in “Surgery” group have no different survival time with “RT” group in age category of 75 years and older. Conclusions Local therapies, including surgery, radiation, and combination of both might associate to improve survival in advanced gastric cancer patients, but confounding due to disease extent and physical status cannot be excluded.