Utility of diffusion-weighted signal intensity ratio measurements for predicting malignancy in adnexal mass lesions and its comparison with ORADS MRI scoring: Can an abbreviated MRI protocol predict malignancy accurately?

Abstract

Aim: To assess if diffusion-weighted signal intensity ratio (DW-SIR) is a good parameter for differentiating malignant from benign and borderline ovarian lesions.

Materials and methods: The MRI studies of histopathologically diagnosed cases of ovarian tumors were reviewed by two groups of radiologists, both calculating the DW SIR for individual cases. The correlation between the DW SIR and ORADS scoring was tested statistically and the positive likelihood ratio of ORADS scoring for predicting malignancy when used solely as well as in conjunction with DW-SIR was calculated.

Results: A total of 184 cases were included, of which 74 cases were Benign/borderline (40 %) and 110 cases were malignant (60 %). A DW SIR cut off of 0.375 calculated after the first cycle of reading had an accuracy of 68.5 % and positive predictive value (PPV) of 73 % in predicting malignancy. The PPV for malignancy for each ORADS score increased especially in the ORADS 3 group when the DW SIR cut off was applied. Higher DW SIR values were associated with a higher ORADS score (p<0.001). A marginal increase in the positive likelihood ratio for malignancy was observed in the combined ORADS 4 and 5 groups upon using the DW-SIR cut off.

Conclusion: DW-SIR can serve as an adjunctive tool to improve confidence in interpreting intermediate-risk ORADS cases or as a standalone triage tool when contrast MRI is contraindicated, keeping in mind its limitations in sensitivity.