Development of specific biomarkers is necessary for individualized management of patients with gastric cancer. The aim of this study was to design a simple expression panel comprising novel molecular markers for precise risk stratification. Patients (n = 200) who underwent gastrectomy for gastric cancer were randomly assigned into learning and validation sets. Tissue mRNA expression levels of 15 candidate molecular markers were determined using quantitative PCR analysis. A dual-marker expression panel was created according to concordance index (C-index) values of overall survival for all 105 combinations of two markers in the learning set. The reproducibility and clinical significance of the dual-marker expression panel were evaluated in the validation set. The patient characteristics of the learning and validation sets were well balanced. The C-index values of combinations were significantly higher compared with those of single markers. The panel with the highest C-index (0.718) of the learning set comprised SYT8 and MAGED2, which clearly stratified patients into low-, intermediate-, and high-risk groups. The reproducibility of the panel was demonstrated in the validation set. High expression scores were significantly associated with larger tumor size, vascular invasion, lymph node metastasis, peritoneal metastasis, and advanced disease. The dual-marker expression panel provides a simple tool that clearly stratifies patients with gastric cancer into low-, intermediate-, and high risk after gastrectomy.
Keywords: Biomarker; expression panel; gastric cancer; prognosis; recurrence.
© 2018 The Authors. Cancer Medicine published by John Wiley & Sons Ltd.