Dynamic contrast enhanced MRI and diffusion weighted imaging features of granulomatous mastitis: compared with breast cancer

Objective: To investigate the radiological features of granulomatous mastitis (GM) in dynamic contrast enhanced MRI (DCE-MRI) and DWI and to differentiate it from the breast cancer in diagnose. Methods: Forty five cases of GM and 64 cases of breast cancer confirmed by surgical histopathology or biopsy were retrospectively analyzed in the study. All of the patients were examined preoperatively by DCE-MRI and DWI. Including lesion type, location, enhancement pattern, nipple retraction, supplying arteries, skin thickening and axillary adenopathy in the two groups were evaluated and analyzed by using χ² test. One-way ANOVA was employed to compare the ADC values between the abscess area of GM and non-abscess area, and the difference among the breast cancer lesion area. Dynamic enhancement MR pharmacokinetic parameters were used to measure including early-phase enhancement rate (EER), peak enhancement ratio (E_max), and time to peak ehhancement(T_max). The statistical differences of EER, E_max and T_max between the two groups were calculated by using Wilcoxon test. Results: In 45 cases of GM, DCE MR images showed nonmass-like lesions (43 patients) and mass-like lesions (2 patients); the nipple involved(16 patients) and segment involved (29 patients); rim-like with heterogeneous enhancement (40 patients) and heterogeneous enhancement (5 patients); nipple retraction (24 patients) supplying arteries dilatation (42 patients), skin thickening (29 patients), and axillary adenopathy (17 patients). Corresponding to the radiological features above, in the 64 breast cancer cases, it showed 54, 10, 5, 59,30, 34, 16, 51, 12 and 20, respectively. There were statistical significance between GM and breast cancer in lesion type, location, enhancement pattern, and nipple retraction (χ² = 67.574, 13.075, 20.297, 20.398 and 23.510, respectively, all P < 0.01). But no differences were existed between 2 groups in supplying arteries and axillary adenopathy (χ² = 3.928 and 0.502, P > 0.05). EER, E_max and T_max GM were 146.58%, 191.13%, 195.00 s in GM and 118.13%, 159.43%, 183.33 s in breast cancer, respectively. Significant statistic differences between GM and breast cancer were found in EER and E_max (Z values were -2.271 and -2.948, P < 0.01). But it did not show significant difference in T_max(Z = -0.548, P > 0.05). The ADC values of GM on abscess area, non-abscess area, and breast cancer lesion area were (8.0 ± 2.6) × 10^-3, (11.3 ± 1.7) × 10^-3 and (8.2 ± 1.1) × 10^-3 mm²/s, respectively .There were significant differences in the groups (F = 52.167, P < 0.01). Conclusions: The characteristic of radiological findings can be found in GM by using advanced MR imaging techniques. DCE-MRI combined with DWI is useful in the differential diagnosis between GM and breast cancer. (authors)