Prognostic Significance of the Triglyceride-to-High-Density Lipoprotein Cholesterol Ratio (TG/HDL-C) in Breast Cancer Patients: A Real-World Retrospective Study

Abstract

Objective This study investigates the prognostic utility of lipid biomarkers in breast cancer (BC) and constructs nomograms incorporating the triglyceride-to-high-density lipoprotein cholesterol (TG/HDL-C) ratio to facilitate individualized outcome prediction. Methods A total of 563 patients with stage I–III breast cancer who underwent mastectomy at Hebei General Hospital between July 2017 and July 2020 were retrospectively analyzed in this single-centre study. Patients were randomly allocated into a training cohort (N=394) and a validation cohort (N=169). Associations between lipid biomarkers, clinicopathological factors, and patient outcomes—overall survival (OS) and disease-free survival (DFS)—were evaluated using Cox proportional hazards models. Nomograms were subsequently constructed from variables identified in multivariate analyses. Predictive performance was assessed via Akaike Information Criterion (AIC), time-dependent ROC curves, calibration plots, decision curve analysis (DCA), concordance index (C-index), and risk group stratification. Results Preoperative TG/HDL-C, TNM stage, and molecular subtype emerged as significant predictors of OS in BC patients, whereas TG/HDL-C and TNM stage demonstrated prognostic value for DFS. The derived and internally validated nomograms displayed robust predictive performance. For OS prediction, the 1-, 3-, and 5-year time-dependent AUCs reached 0.868, 0.913, 0.966 in the training cohort and 0.877, 0.921, 0.933 in the validation cohort, with corresponding C-indexes of 0.892 (95% CI: 0.868–0.916) and 0.907 (95% CI: 0.874–0.939), respectively. For DFS prediction, the 1-, 3-, and 5-year time-dependent AUCs were 0.797, 0.760, 0.788 for the training cohort and 0.644, 0.700, 0.755 for the validation cohort, accompanied by C-indexes of 0.737 (95% CI: 0.682–0.792) and 0.664 (95% CI: 0.544–0.784). The nomograms had better discrimination, calibration and clinical utility than TNM staging, with stable performance in the validation cohort. A user-friendly web calculator was built for clinical use. Conclusion This study performed the development and internal validation of prognostic nomograms. The preoperative TG/HDL-C ratio serves as a potential prognostic indicator for OS and DFS in stage I–III BC. The nomogram integrating this ratio, TNM stage and molecular subtype represents an exploratory tool for personalized risk stratification. These findings warrant further validation in independent prospective multicenter cohorts before any clinical application.