Ultrasonographic assessment of pediatric adnexal lesions: validation of the O-RADS system and development of a simplified C-CAS framework

Background: The Ovarian-Adnexal Reporting and Data System (O-RADS) ultrasound (US) is a standardized tool for assessing adult adnexal lesions, but its performance in the pediatric population has not been fully validated. This study aimed to validate O-RADS US in children and develop a simplified assessment framework for pediatric adnexal lesions.

Methods: This retrospective study included a training cohort of 301 adnexal lesions from three centers and an independent validation cohort of 95 lesions from a fourth center. All lesions were pathologically confirmed. In the training cohort, lesions were classified according to the O-RADS V2022 criteria, and the diagnostic performance of the system was evaluated. Independent predictors of malignancy identified by firth penalized regression were used to construct a simplified framework [Cysts-Color-Ascites-Shadowing (C-CAS)]. This C-CAS framework was subsequently tested in the validation cohort.

Results: Of the 396 lesions, 44 (11.11%) were malignant. In the training cohort, O-RADS showed excellent diagnostic performance [area under the curve (AUC) =0.931, 95% confidence interval (CI): 0.898–0.964]. Using a cutoff of category ≥4, sensitivity was 97.2% (95% CI: 91.9–100.0), specificity 89.1% (95% CI: 85.3–92.8%), positive predictive value 54.7% (95% CI: 42.5–66.9%), and negative predictive value 99.6% (95% CI: 98.8–100.0%). Firth regression analysis confirmed that “color score 4” [odds ratio (OR) =21.78, 95% CI: 1.80–255.38, P=0.012] and “ascites” (OR =49.13, 95% CI: 4.25–417.96, P<0.001) were strong independent predictors of malignancy, whereas “uni- or bilocular cyst without solid component” (OR ≈0.00, 95% CI: 0.00–0.04, P<0.001) and “acoustic shadowing” (OR =0.04, 95% CI: 0.00–0.45, P=0.004) were significant protective factors. The derived C-CAS framework achieved an AUC of 0.932 (95% CI: 0.911–0.953) in the training set and maintained high diagnostic accuracy in the independent validation set (AUC =0.948, 95% CI: 0.916–0.980).

Conclusions: O-RADS US is a clinically applicable tool for pediatric adnexal lesions. The proposed C-CAS framework offers a validated, simplified alternative which maintains high diagnostic accuracy.