Non-mass Enhancement in Breast MRI: Characterization with BI-RADS Descriptors and ADC Values

Abstract

Objectives: The purpose of this study was to assess the accuracy of contrast-enhanced magnetic resonance imaging and diffusion-weighted imaging in distinguishing benign from malignant non-mass-like breast lesions. Methods: 103 lesions showing non-mass-like enhancement in 100 consecutive patients were analyzed. Distribution, internal enhancement patterns, and contrast kinetic curve patterns were classified according to the BI-RADS lexicon. Apparent diffusion coefficient (ADC) values were obtained from manually placed regions of interest (ROIs) on diffusion-weighted images. The optimal ADC value threshold for the distinction between benign and malignant lesions was determined by ROC analysis. Univariate and multivariate analyses were performed to identify independent predictors of malignancy, and the probability of malignancy was calculated for various combinations of findings. Histological diagnosis obtained by means of core needle biopsy was used as gold standard. Results: According to the univariate and multivariate analysis, odds ratios for malignancy were significantly elevated for clumped or clustered ring internal enhancement and low ADC values (p < 0.001), whereas distribution patterns and contrast kinetic patterns were not significantly correlated with benignity or malignancy. In non-mass lesions with homogeneous or heterogeneous internal enhancement and ADC values greater than 1.26×10-3mm2/s, no malignancy was detected, while all other combinations of findings had a probability of malignancy ranging from 22.2 to 76.6%. Conclusions: A combination of BI-RADS descriptors of internal enhancement and ADC values is useful for the differential diagnosis of lesions showing non-mass enhancement. Lesions with homogeneous or heterogeneous enhancement and high ADC can be followed up, while all other lesions should be biopsied. Doi: 10.28991/SciMedJ-2021-0302-1 Full Text: PDF