Nomogram for predicting gastric cancer recurrence using biomarker gene expression

Abstract

Background: Recently, researchers have tried to predict patient prognosis using biomarker expression in cancer patients. The aim of this study was to develop a nomogram predicting the 5-year recurrence-free probability (RFP) of gastric cancer patients using prognostic biomarker gene expression. Methods: We enrolled 360 patients in the training data set to develop the predictive model and nomogram. We analyzed the patients' general variables and the gene expression levels of 10 prognostic biomarker candidates between the nonrecurrence and recurrence groups. We also performed external validation using 420 patients from the validation data set. Results: The final nomogram was composed of age, sex, and the expression levels of CAPZA, Pease, OCT-1, PRDX4, gamma-enolase, and c-Myc. The five-year RFPs were 89%, 75%, 54% and 32% for the patients in the low-risk, intermediate-risk, high-risk and very-high-risk groups in the development cohort, respectively. In the external validation cohort, the 5-year RFPs were 89%, 75%, 63% and 60%, respectively. The areas under the curve were 0.718 (95% CI, 0.65-0.78) and 0.640 (95% CI, 0.57-0.70) for the training and validation data sets, respectively. The RFP Kaplan-Meier curves were significantly different among the 4 groups in the training and validation data sets (p < 0.0001). Conclusion: This newly developed nomogram using gene expression can predict the 5-year RFP for gastric cancer patients after surgical treatment. We hope that this nomogram will help in the therapeutic decision between endoscopic treatment and gastrectomy. (C) 2019 The Authors. Published by Elsevier Ltd.