Dynamic Inflammatory-Nutritional Biomarkers Predict Early Postoperative Complications in Gastric Cancer: A Retrospective Cohort Study

Abstract
Background: Early postoperative complications adversely affect outcomes following gastrectomy.
Objective: To evaluate whether dynamic inflammatory-nutritional biomarkers can predict early complications after radical gastrectomy.
Methods: In this retrospective analysis of 182 patients undergoing radical gastrectomy between July 2023 and May 2025, we assessed key inflammatory-nutritional biomarkers, including CAR and neutrophil count, on postoperative days 1 and 3. Independent predictors were identified using multivariate logistic regression and receiver operating characteristic (ROC) curve analysis.
Results: Early complications occurred in 37 patients (20.33%). Multivariate logistic regression revealed age >70 years (OR=5.59, 95% CI: 1.19, 29.78), CAR>1.942 on postoperative day 1 (OR=7.26, 95% CI: 1.46, 39.85), and neutrophil count on postoperative day 3 (OR=1.66, 95% CI: 1.21, 2.42) as independent predictors. Tumor location in the gastric body was associated with a reduced risk (OR=0.17, 95% CI: 0.03, 0.76). The combined model yielded superior predictive performance (AUC=0.796; sensitivity: 75.7%; specificity: 82.8%). An interactive web-based calculator was developed using the R Shiny framework to enable rapid individual risk assessment, available at https://gbdata.shinyapps.io/gastric_app/.
Conclusions: Postoperative CAR and neutrophil count, when integrated with key clinical variables, allow early risk stratification for gastric cancer surgery. This simple, cost-effective model may assist clinical decision-making and perioperative management.
Keywords: Gastric Neoplasms; C-reactive Protein-to-Albumin Ratio; Neutrophils; Postoperative Complications; ROC Curve