Retrospective Cohort Study of Tumor Size and Survival in Gastric Cancer

Abstract

Abstract Introduction: Tumor size constitutes a very important staging factor among patients afflicted by solid tumors, and was closely associated with the overall survival (OS). Nonetheless, the prognostic import of tumor size in gastric cancer (GC) remains shrouded in uncertainty. Method: A multivariable-adjusted hazard ratio (HR) along with a 95% confidence interval (CI) was computed for GC using Cox proportional hazard regression models. To assess the non-linear association between tumor size and OS, we employed a restricted cubic spline regression analysis. Additionally, a two-piece-wise Cox proportional hazard model was utilized to determine the threshold effect. The efficacy prediction of tumor size was tested by ROC curve. Results: A cohort comprising 2,012 gastric cancer patients who had undergone gastrectomy was included in our secondary analysis, sourced from a multicenter study conducted in Korea. Also, we found a nonlinear association between tumor size and OS at the turning point as 5.7 (95%CI: 5.1, 6.8). The HR for mortality was 1.50 (95%CI: 1.38, 1.64) for tumors size <5.7, while 1.09 (95%CI: 1.05, 1.13) with size ≥ 5.7. There was still a significant non-linear relationship between OS and size after adjusting for potential confounding factors (P=0.001). In addition, Besides, a significantly higher HR was found in young GC patients(≤45: 1.33; 95%CI:1.24, 1.41; >45, 1.16; 95%CI: 1.13,1.19; P for interaction = 0.0004 ). Conclusions: Tumor size was non-linear associated with survival for patients receiving gastrectomy. It might have the higher predictive power in young GC.