Exploratory evaluation of a CEUS-derived substitution for solid-component measurement in the ADNEX model

Although the Assessment of Different Neoplasias in the adnexa (ADNEX) model has been widely validated, there remains challenges in differentiating true solid components from benign mimics. We aimed to evaluate the feasibility and value of substituting conventional ultrasound-derived solid component size with CEUS-derived measurements in the ADNEX model for adnexal lesions discrimination. This multicenter prospective study recruited consecutive patients with suspected adnexal lesions with solid components detected at ultrasound between January 2022 and June 2024. All patients underwent preoperative CEUS examination. Pathological results serve as the reference standard. Each adnexal lesion was evaluated using ADNEX, a CEUS-based modified ADNEX score (CEUS-derived substitution for solid-component measurement without updating model coefficients) and Ovarian-Adnexal Reporting and Data System (O-RADS) Ultrasound (US) score. A total of 204 patients (median age, 47.0 years [IQR: 33.0-56.0 years]) with 204 lesions were included. The area under the receiver operating characteristic curve (AUROC) for CEUS-based modified ADNEX was 0.906 (95% CI: 0.867-0.944), which was superior to that of ADNEX (AUROC:0.865; 95% CI, 0.814-0.916; P = 0.019) and O-RADS US (AUROC: 0.783; 95% CI, 0.726 -0.840; P < 0.001) in classifying adnexal lesions with solid components. CEUS-based modified ADNEX demonstrated greater specificity (0.664) compared to ADNEX (0.283; P < 0.001) and O-RADS US (0.221; P < 0.001). CEUS-derived modification for solid-component measurement in the ADNEX model showed potential improvements in the diagnostic specificity and clinical benefit compared with the original ADNEX framework in the classification of adnexal lesions with solid components.