CT-based radiomics nomogram analysis for assessing BRCA mutation status in patients with high-grade serous ovarian cancer

BACKGROUND: Radiomics nomogram analysis is widely preoperatively used to assess gene mutations in various tumors. PURPOSE: To explore the value of computed tomography (CT)-based radiomics nomogram analysis for assessing BRCA gene mutation status of patients with high-grade serous ovarian cancer (HGSOC). MATERIAL AND METHODS: In total, 96 patients with HGSOC were retrospectively screened and randomly divided into primary (n = 68) and validation cohorts (n = 28). The clinical model was constructed based on clinical features and CT morphological features using univariate and multivariate logistic analyses. Maximum-relevance and minimum-redundancy (mRMR) and least absolute shrinkage and selection operator (LASSO) were performed for feature dimensionality reduction and radiomics score was calculated. The nomogram model combining the clinical model and the radiomics score was constructed using multivariate logistic regression. Receiver operating characteristic (ROC) curves were generated to assess models' performance. The calibration analysis and decision curve analysis (DCA) were also performed. RESULTS: The clinical model consisted of CA125 level and supradiaphragmatic lymphadenopathy and yielded an area under the curve (AUC) of 0.69 (primary cohort) and 0.81 (validation cohort). The radiomics model was built with seven selected features and showed an AUC of 0.87 (primary cohort) and 0.81 (validation cohort). The nomogram finally showed the highest AUC of 0.89 (primary cohort) and 0.87 (validation cohort). The nomogram presented favorable calibrations in both the primary and validation cohorts. DCA further confirmed the clinical benefits of the constructed nomogram. CONCLUSION: CT-based radiomics nomogram provides a non-invasive method to discriminate BRCA gene mutation status of HGSOC and potentially helps develop precise medical strategies.