Segmental Non-Mass Enhancement Features in Breast Magnetic Resonance Imaging: A Multicenter Retrospective Study of Histopathologic Correlations.

Background/Objectives: Segmental non-mass enhancement (NME) is the breast MRI distribution pattern with the highest positive predictive value (PPV) for malignancy. Despite its diagnostic relevance, its imaging characteristics have rarely been examined in isolation, leaving uncertainty in clinical practice. This multicenter retrospective cohort study aimed to evaluate multiparametric MRI features-including internal enhancement pattern, dynamic contrast-enhanced (DCE) kinetics, and diffusion restriction-in segmental NME to identify malignancy predictors. Methods: This retrospective cohort review included 14,834 breast MRI reports from five institutions (September 2017-February 2024), identifying 103 women (mean age, 44.4 ± 9.9 years) with segmental NME (70 malignant, 33 benign). MRI was performed at 1.5 T or 3 T using standardized protocols. Two breast radiologists, blinded to pathology, assessed internal enhancement, DCE kinetics, diffusion restriction, and short tau inversion recovery (STIR) features according to BI-RADS. Statistical analyses included chi-square/Fisher's tests and logistic regression. Results: Clustered ring enhancement (CRE) was significantly associated with malignancy (p = 0.004). Fast initial-phase enhancement (p < 0.001) and delayed-phase washout (p = 0.011) also correlated with malignancy. On multivariate analysis, fast initial-phase enhancement remained an independent predictor (odds ratio [OR] = 5.133, p = 0.031), whereas slow enhancement predicted benignity (OR = 0.194, p = 0.020). Histologies included ductal carcinoma in situ, invasive ductal carcinoma, granulomatous mastitis, and benign hyperplastic lesions. Conclusions: This study, focusing exclusively on segmental NME, identifies CRE, fast initial-phase enhancement, and washout kinetics as reliable imaging biomarkers. Incorporating these features into breast MRI interpretation may improve diagnostic accuracy, risk stratification, and management decisions.