Added Value of Quantitative Analysis of Diffusion-Weighted Imaging in Ovarian-Adnexal Reporting and Data System Magnetic Resonance Imaging

BACKGROUND: The ovarian-adnexal reporting and data system-magnetic resonance imaging (O-RADS-MRI) score decreases the incidence of indeterminate adnexal masses from 18% to 31% with ultrasound till 10.8% to 12.5% with MRI. Further improvement of this score may be useful to improve patients' management.

PURPOSE: To evaluate the added value of quantitative diffusion-weighted imaging (DWI) in the diagnosis of adnexal masses classified according to the O-RADS-MRI score.

STUDY TYPE: Prospective cohort study with retrospective DWI analysis.

POPULATION: Among 402 recruited patients, surgery was done only in 163 women (median-age: 51 years) with 201 indeterminate adnexal masses, which were included in the final analysis.

FIELD STRENGTH/SEQUENCE: Standardized MRI (1.5 and 3-T) including diffusion and dynamic contrast-enhanced sequences (diffusion-weighted single-shot spin-echo echo-planar imaging) were used.

ASSESSMENT: Two radiologists classified the adnexal masses according to O-RADS-MRI and they were blinded to the pathology report. Two methods of quantitative analysis were applied using region-of-interest apparent-diffusion-coefficient (ROI-ADC) and whole-lesion ADC-histogram (WL-ADC).

STATISTICAL TESTS: Fisher's exact and Mann-Whitney-U tests were used to compare variables among malignant and benign lesions. Receiver-operating-characteristic (ROC) curves were constructed to examine the sensitivity/specificity of each parameter. ROI-ADC and WL-ADC of lesions with O-RADS-MRI score-4 were plotted to identify thresholds of malignant lesions. The improvement of the O-RADS-MRI score after adding these thresholds was assessed by two ROC-curves. A P < 0.05 was considered to be statistically significant.

RESULTS: Fifty-eight of the 201 lesions (28.9%) were malignant. The ROI-ADC and the WL-ADC means of malignant lesions were significantly lower than those of benign lesions. Forty-two lesions (20.9%) had an O-RADS-MRI score-4. In this subgroup, 76% of lesions with ROI-ADC < 1.7 × 10-3 mm2 /sec and WL-ADC < 2.6 × 10-3 mm2 /sec were malignant, whereas only 11.8% with ROI-ADC ≥ 1.7 × 10-3 mm2 /sec or a WL-ADC ≥ 2.6 × 10-3 mm2 /sec were malignant. The overall performance of the O-RADS-MRI score combined with these thresholds was improved.

DATA CONCLUSION: Integrating ADC-thresholds in O-RADS-MRI score-4 may discriminate low-to-intermediate and intermediate-to-high malignancy risk groups.

LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY STAGE: 2.