Accurate prediction of benign and malignant adnexal tumors in surgical resection and conservative treatment: construction and external validation of a diagnostic model based on CEUS, HE4, and O-RADS US v2022 evaluation.

Purpose: To establish a diagnostic model combining contrast-enhanced ultrasound (CEUS), human epididymis protein 4 (HE4), and Ovarian-Adnexal Reporting and Data Systems (O-RADS) US v2022, verify its diagnostic efficacy, and compare it with subjective evaluation.

Methods: From January 2018 to August 2021 (the test group) and from September 2021 to September 2022 (the validation group), the data of patients classified as O-RADS US v2022 categories 2 to 5 who underwent adnexal ultrasound examinations were prospectively and continuously collected. In the test group, univariate and multivariate analyses were used to explore the relationship between age, body mass index (BMI), maximum diameter of the lesion, menopausal status, HE4, cancer antigen 125 (CA125), and the characteristics of CEUS and malignant lesions. Selecting independent influencing factors to construct diagnostic model, which was validated in the external validation group and compared with subjective evaluation.

Results: The test group included 563 patients (mean age, 48.7 ± 13.2), and the validation group included 246 patients (mean age, 47.6 ± 12.9). Univariate and multivariate analyses showed that enhancement time, enhancement intensity, dynamic changes, and HE4 were independent influencing factors for predicting adnexal malignant tumors. In the validation group, the sensitivities and specificities of O-RADS US v2022, O-RADS US v2022 + CEUS, O-RADS US v2022 + CEUS + HE4, and subjective assessment were 88.89% and 70.69%, 94.44% and 79.31%, 91.67% and 92.53%, and 93.09% and 89.66% respectively. In addition, the combined diagnostic performance of O-RADS US v2022, CEUS and HE4 (AUC = 0.980) was higher than that of O-RADS US v2022 alone (AUC = 0.876, P < 0.001) and the combination of O-RADS US v2022 + CEUS (AUC = 0.908, P < 0.001), and was comparable to the subjective evaluation (AUC = 0.963, P = 0.192).

Conclusions: The combined diagnostic model of O-RADS US v2022, CEUS and HE4 can improve the specificity of adnexal ultrasound diagnosis without sacrificing sensitivity, and it has high reliability.