Use of Radiographic Features for Risk Stratification of Ovarian Masses in Children and Adolescents

Introduction: Radiographic imaging is used to identify and characterize ovarian tumors. Radiographic features suspicious for malignancy should prompt further evaluation including a multidisciplinary discussion for consideration of ovarian sparing surgery versus oophorectomy. This study assessed the diagnostic capability of radiographic features alone and in combination to risk stratify between benign and malignant ovarian tumors in pediatric and adolescent patients. Methods: We performed a planned secondary analysis of a prospective multi-institutional, interventional cohort study using a consensus-based preoperative risk stratification algorithm to decrease unnecessary oophorectomies in patients 6-21 years at 11 children's hospitals between August 2018 and January 2021. All patients underwent radiographic imaging. Radiographic features were analyzed individually and in combination. A priori clinical consensus prioritized maximizing negative predictive value to minimize risk of missed malignancy. Results: Imaging was performed in 519 patients with ovarian tumors. The presence of a complex mass, solid component, or tumor size >8 cm was reported in 26 of 27 malignant tumors. All 26 malignant tumors were identified when one of these radiographic features was present, conferring a negative predictive value of 100% with 65.1% accuracy. The combination of radiographic evidence of metastatic disease and increased β-human gonadotropin, cancer antigen 125, or inhibin A identified all