Prediction of Survival Outcomes Based on Preoperative Clinical Parameters in Gastric Cancer

Few current preoperative risk assessment tools provide essential, optimized treatment for gastric cancer. The purpose of this study was to develop and validate a nomogram that uses preoperative data to predict survival and risk assessments.

A survival prediction model was constructed using data from a developmental cohort of 1251 patients with stage I to III gastric cancer who underwent curative resection between January 2005 and December 2008 at Ajou University Hospital, Korea. The model was internally validated for discrimination and calibrated using bootstrap resampling. To externally validate the model, data from a validation cohort of 2012 patients with stage I to III gastric cancer who underwent surgery at multiple centers in Korea between January 2001 and June 2006 were analyzed. Analyses included the model's discrimination index (C-index), calibration plots, and decision curve that predict overall survival.

Eight independent predictors, including age, sex, clinical tumor size, macroscopic features, body mass index, histology, clinical stages, and tumor location, were considered for developing the nomogram. The discrimination index was 0.816 (adjusted C-index) in the developmental cohort and 0.781 (adjusted C-index) in the external validation cohort. Additionally, in both the developmental and validation datasets, age and tumor size were significantly correlated with each other and were independent indicators for survival (P < 0.05).

We developed a new nomogram by using the most common and significant preoperative parameters that can help to identify high-risk patients before treatment and help clinicians to make appropriate decisions for patients with stage I to III gastric cancer.