DefinePK

Abstract

Background: Accurate characterization of adnexal masses is critical in gynecological practice. Ultrasound (US) is the frontline tool, yet indeterminate lesions remain common. The American College of Radiology Ovarian‑Adnexal Reporting and Data System for MRI (O‑RADS MRI) was developed to standardize post‑US risk stratification. Evidence on its prospective performance in routine practice—particularly across the full reproductive spectrum—remains limited.
Methods: In this single‑centre prospective study (January 2023 – December 2024), consecutive reproductive and post‑menopausal women with sonographically indeterminate adnexal masses underwent 3 T pelvic MRI with a dedicated O‑RADS protocol. Two fellowship‑trained radiologists assigned O‑RADS MRI scores (1–5) blinded to histology. Reference standards were surgical histopathology or ≥ 6‑month imaging follow‑up. Diagnostic metrics were calculated for the threshold O‑RADS ≥ 4 (high risk). Sub‑analysis examined performance in reproductive (< 50 y) versus post‑menopausal (≥ 50 y) cohorts.
Results: One‑hundred‑twenty women (mean age ± SD 48.9 ± 13.8 y) with 120 lesions were analysed; 42 (35%) were malignant, 8 (7%) borderline, and 70 (58%) benign. O‑RADS MRI yielded sensitivity 88.1%, specificity 85.7%, positive predictive value 78.0%, negative predictive value 92.4%, and overall accuracy 86.7%. Area under the ROC curve was 0.93 (Figure 1). Performance was similar in reproductive and post‑menopausal groups (p = 0.46). Lesions with lipid‑rich solid enhancement (newly classified as O‑RADS 4 in the 2022 revision) accounted for 10% of malignancies.
Conclusion: A prospectively applied O‑RADS MRI algorithm provides high accuracy for differentiating benign from malignant adnexal lesions after indeterminate US, irrespective of menopausal status. Adoption could reduce unnecessary surgery while expediting oncologic referral. Future multicentre work should evaluate cost‑effectiveness and integrate advanced diffusion and radiomics biomarkers.