Risk prediction of second primary malignancies after gynecological malignant neoplasms resection with and without radiation therapy: a population-based surveillance, epidemiology, and end results (SEER) analysis

Abstract

Abstract Purpose The association between post-resection radiotherapy for primary gynecological malignant neoplasms (GMNs) and the development of secondary primary malignancies (SPMs) remains a subject of debate. This study represents the first population-based analysis employing a multivariate competitive risk model to assess risk factors for this relationship and to develop a comprehensive competing-risk nomogram for quantitatively predicting SPM probabilities. Materials and methods In our study, data on patients with primary GMNs were retrospectively collected from the Epidemiology, Surveillance and End Results (SEER) database from 1973 to 2015. The incidence of secondary malignant tumors diagnosed at least six months after GMN diagnosis was compared to determine potential risk factors for SPMs in GMN patients using the Fine and Gray proportional sub-distribution hazard model. A competing-risk nomogram was constructed to quantify SPM probabilities. Results A total of 109,537 patients with GMNs were included in the study, with 76,675 and 32,862 GMN patients in the training and verification sets, respectively. The competing-risk model analysis identified age, primary tumor location, tumor grade, disease stage, chemotherapy, and radiation as risk factors for SPMs in GMN patients. Calibration curves and ROC curves in both training and verification cohorts demonstrated the predictive accuracy of the established nomogram, which exhibited a good ability to predict SPM occurrence. Conclusions This study presents the nomogram developed for quantitatively predicting SPM probabilities in GMN patients for the first time. The constructed nomogram can assist clinicians in designing personalized treatment strategies and facilitate clinical decision-making processes.