Preoperative electrolyte-based metabolic biomarkers for prognostic assessment in gastric cancer: a multicenter study

Background Gastric cancer remains a major cause of cancer-related mortality worldwide, with substantial heterogeneity in patient outcomes despite advances in treatment. Clinically accessible biomarkers are still needed to improve prognostic stratification. Serum electrolytes, as routine circulating parameters, may reflect systemic metabolic status, host physiological adaptation, and tumor-related stress. This study aimed to develop and externally validate an electrolyte-based metabolic biomarker score for prognostic assessment in gastric cancer. Methods This retrospective multicenter study included 1, 371 patients with gastric cancer from two independent cohorts: a training cohort (n = 1, 085) and an external validation cohort (n = 286). Seven preoperative serum electrolytes were analyzed. Univariate and multivariate Cox regression analyses were used in the training cohort to identify independent prognostic electrolytes and construct an Electrolyte Score. Patients were stratified into high- and low-risk groups according to the median score in the training cohort, and the same cutoff was applied to the validation cohort. Kaplan-Meier analysis, time-dependent ROC analysis, calibration curves, and decision curve analysis were used to evaluate prognostic performance and clinical utility. Results Sodium and calcium were identified as independent prognostic factors for overall survival and incorporated into the Electrolyte Score. Higher sodium and calcium levels were associated with better survival. Although the Electrolyte Score alone had limited discriminatory ability, it was able to stratify patients into prognostically distinct groups in both cohorts. A high Electrolyte Score was associated with worse overall survival (training cohort, P < 0.001; validation cohort, P = 0.004). Multivariate analysis confirmed the score as an independent prognostic factor. Importantly, a nomogram integrating the Electrolyte Score with age, TNM stage, and histological differentiation improved individualized prognostic prediction, demonstrating potential clinical utility for risk stratification and guiding patient management. Conclusions A preoperative electrolyte-based score derived from sodium and calcium provided independent prognostic information in gastric cancer. While its predictive value alone is limited, its combination with clinicopathological variables in a nomogram enhances individualized risk stratification, offering a practical tool to complement conventional prognostic assessment in clinical practice.