Nomograms for Predicting Overall Survival and Cancer-Specific Survival of Small Cell Carcinoma of Ovary Patients: A Retrospective Cohort Study

Abstract

Background: This study aimed to develop functional nomograms to predict overall survival (OS) and cancer-specific survival (CSS) of small cell carcinoma of ovary (SCCO).

Methods: SSCO case data were recruited retrospectively from the Surveillance, Epidemiology, and End Results (SEER) database. Nomograms were constructed to predict the probabilities of OS and CSS in SCCO patients based on independent predictors. The predictive performance of nomogram was evaluated with the concordance index (C-index), area under the curve (AUC), calibration curves, and decision curve analysis (DCA).

Results: The independent risk factors affecting the prognosis of SCCO patients were older age, lower income, surgery, chemotherapy, radiation, advanced International Federation of Gynecology and Obstetrics (FIGO) stage, and number of primary tumors. The C-index for the OS nomogram was 0.78 (95% confidence interval (CI): 0.75 - 0.82), and AUCs for 1-, 3-, and 5-year OS were 0.861, 0.807, and 0.821, respectively. The C-index for the CSS nomogram was 0.79 (95% CI: 0.76 - 0.83), and AUCs for 1-, 3-, and 5-year OS were 0.873, 0.841, and 0.864, respectively. The calibration curves indicated reasonable agreement between the observed and predicted probabilities of the OS and CSS nomograms, which indicated a good degree of confidence. According to the C-index, ROC, and DCA, the prognostic nomograms of OS and CSS showed better prediction accuracy and clinical application value for SCCO than the FIGO staging system.

Conclusions: We constructed original nomograms that provided useful prediction of OS and CSS for patients with SCCO. These models could facilitate the postoperative personalized assessment and the identification of treatment strategy.