Development and external validation of a multivariable nomogram predicting cancer specific survival in patients with serous ovarian carcinoma using SEER and TCGA cohorts

Purpose Serous ovarian carcinoma is the most lethal gynecologic malignancy. Existing prognostic nomograms frequently lack independent external molecular validation. The present study aimed to develop and externally validate a parsimonious five-variable Cox nomogram predicting cancer-specific survival (CSS) in serous ovarian carcinoma using national population-based registries. Methods Patients diagnosed with primary serous ovarian carcinoma (2000–2022) were retrieved from the SEER database and allocated (7:3 ratio) to training (n = 34,402) and internal validation (n = 14,744) cohorts. The Cancer Genome Atlas Ovarian Serous Cystadenocarcinoma (TCGA-OV, n = 583) cohort served as an independent external validation set. Predictors were selected via ten-fold cross-validated LASSO-Cox regression. Discrimination, calibration, decision curve analysis (DCA), and Kaplan-Meier risk stratification were appraised across all cohorts. Results Five routine clinical variables were retained: age group, race, laterality, histologic grade, and summary stage. Distant stage (HR = 6.45, 95% CI: 5.62–7.42) and age ≥ 80 years (HR = 3.17, 95% CI: 2.99–3.35) were the strongest mortality determinants. The nomogram yielded a C-index of 0.6700 in the training cohort, 0.6690 in the internal validation cohort, and 0.6081 in the external TCGA-OV cohort. Five-year calibration curves demonstrated close agreement between predicted and observed survival. DCA verified net clinical benefit across 0.05–0.45 threshold probabilities, and risk stratification significantly segregated patients into high- and low-risk subgroups across all cohorts (all log-rank P < 0.001). Conclusion A practical five-variable Cox nomogram was established and externally validated to accurately estimate 1-, 3-, and 5-year CSS in serous ovarian carcinoma, providing a standardized tool for individualized post-treatment surveillance.