Contrast-enhanced 3.0-T breast MRI for characterization of breast lesions: increased specificity by using vascular maps

Abstract

Purpose: To assess the diagnostic accuracy of contrast-enhanced 3.0-T breast magnetic resonance imaging (MRI) for differentiating benign from malignant breast masses and subsequently to test if specificity could be further improved by scoring of the overall ipsilateral breast vascularity.

Materials and methods: Fifty-four patients were prospectively enrolled in the study and underwent contrast-enhanced 3.0-T breast MRI. MR images were evaluated and classified according to the MRI BI-RADS lexicon criteria. Lesion size, number of lesions, and localization in the breast were systematically assessed. Maximum intensity projections (MIPS) were obtained by using high-resolution contrast-enhanced (0.1 mmol/kg gadobutrol) fat-saturated T1-weighted images. Breast vascularization was scored according to the methods from Sardanelli et al. by measuring the number, diameter, and length of the vessels on the MIPS. The score ranged from 0 (indicating absent or low breast vascularity) to 3 (indicating high breast vascularity).

Results: Final analysis of 56 lesions revealed 25 (45%) malignant lesions and 31 (55%) benign lesions. Correlation with the MRI BI-RADS classification revealed cancer in none (0%) of the BI-RADS II lesions, in 1 (12%) of the BI-RADS III lesions, in 5 (83%) of the BI-RADS IV lesions, and in 19 (100%) of the BI-RADS V lesions. Based on morphologic and kinetic data analysis, the sensitivity and specificity of 3.0-T breast MRI was 100% (25/25) and 74% (23/31), respectively. After adjustment for the breast vascularity score, specificity significantly (p = 0.048) increased to 87% (27/31) without affecting sensitivity.

Conclusion: Diagnostic accuracy of contrast-enhanced 3.0-T breast MRI increased significantly when the vascularity score was added to the standard morphologic and kinetic data analysis, resulting in a specificity of 87% without affecting sensitivity, which remained 100%.