Development and Validation of a Predictive Model Combining [18F] FDG PET/CT and the Suidan Score for Primary Optimal Cytoreduction in Advanced High-Grade Serous Ovarian Cancer: A Retrospective Cohort Study
BACKGROUND For patients with advanced ovarian cancer undergoing cytoreductive surgery, achieving optimal surgical outcomes has significant clinical implications for the design of adjuvant therapy and surveillance strategies. This retrospective, single-center cohort study aimed to develop a radiomics model based on preoperative PET-CT imaging parameters to predict optimal cytoreductive surgery outcomes, thereby providing clinical decision-making support to improve the prognosis of patients with advanced ovarian cancer. MATERIAL AND METHODS Patients with clinically staged IIIC-IVB high-grade serous ovarian carcinoma (HGSOC) who underwent primary cytoreductive surgery between October 2020 and October 2023 were enrolled and divided into training and validation cohorts. A radiomics model based on preoperative PET-CT was developed, and its performance was quantitatively evaluated and compared within the validation cohort. Subsequently, the PET-CT radiomics model was combined with the Suidan score to generate a combined model. RESULTS Optimal cytoreduction rates were 72.86% (training) and 75.0% (validation). The PET-CT radiomics model demonstrated superior discriminative ability (AUC: 0.808, 95%CI: 0.665-0.951) compared to the Suidan score (AUC: 0.773, 95%CI: 0.615-0.931) in validation. The combined model achieved the highest predictive performance (AUC: 0.836, 95%CI: 0.706-0.966), with sensitivity 81.8%, specificity 83.9%, PPV 64.3%, NPV 92.9%, and accuracy 83.3%. CONCLUSIONS The combined model integrating PET-CT radiomics and Suidan score provides accurate preoperative prediction of cytoreductive outcomes, optimizing treatment strategies for advanced ovarian cancer.