Imaging Features Derived From Dynamic Contrast-Enhanced Magnetic Resonance Imaging to Differentiate Malignant From Benign Breast Lesions: A Systematic Review and Meta-Analysis.

Abstract

The PubMed, Web of Science, Embase, and the Cochrane Library databases were searched up to January 2021 to identify original studies that investigated the accuracy of individual DCE-MRI features in differentiating malignant from benign breast lesions. Pooled sensitivity, specificity, and area under the curve were calculated by STATA software based on the data extracted from included studies. Moreover, quality assessment, subgroup analysis, and publication bias evaluation were performed. Twenty-nine studies comprising 2976 patients and 3365 suspicious breast lesions were included. Malignant breast lesions tended to present irregular shapes (83.59%), noncircumscribed margins (85.50%), mass enhancement (52.31%), heterogeneous internal enhancement (71.72%), and type II or III time intensity curve (TIC) patterns (91.17%), showing significant differences compared with benign breast lesions (P < 0.05). For differentiating malignant from benign breast lesions, the area under the curve values of irregular shape, noncircumscribed margin, mass enhancement, heterogeneous internal enhancement, and type II or III TIC patterns were 0.79 (0.76-0.83), 0.87 (0.84-0.90), 0.63 (0.58-0.67), 0.82 (0.78-0.85), and 0.89 (0.86-0.92), respectively. Imaging features derived from DCE-MRI, especially TIC patterns, are important for diagnosing and differentiating malignant from benign breast lesions.