Development and Validation of a KELIM-Based Prognostic Nomogram for Long-Term Survival in Epithelial Ovarian Cancer: A 10-Year Single-Center Study

Epithelial ovarian cancer (EOC) remains the most lethal gynecologic malignancy worldwide, with late-stage diagnosis and poor long-term survival being major clinical challenges. This study aimed to identify key prognostic factors associated with long-term survival in patients with EOC and to develop a predictive nomogram for individualized risk assessment. A total of 100 patients diagnosed with EOC between April 2014 and October 2019 at the Second Hospital of Hebei Medical University were retrospectively enrolled. All patients underwent primary comprehensive staging or cytoreductive surgery followed by platinum-based chemotherapy. Clinical, surgical, and biochemical variables, including age, FIGO stage, histologic subtype, ascites volume, residual tumor size, preoperative CA125 levels, and the CA125 kinetic parameter KELIM, were collected and analyzed. Survival data were followed up through October 2024, with overall survival as the primary endpoint. Univariate and multivariate Cox regression identified FIGO stage, postoperative residual tumor size, and KELIM score as independent predictors of overall survival. A nomogram integrating these three factors was constructed and internally validated using bootstrapping. The model demonstrated excellent discrimination (C-index = 0.86) and calibration, outperforming FIGO stage alone. Time-dependent ROC analysis demonstrated that the nomogram achieved AUCs of 0.808, 0.945, and 0.950 for predicting 1-, 3-, and 5-year overall survival, respectively, outperforming FIGO stage alone (AUCs of 0.714, 0.824, and 0.889). These findings highlight that integrating the KELIM score with conventional staging and postoperative residual disease significantly improves prognostic accuracy in EOC and enables tailored therapeutic decision-making based on individual tumor response.