Correlation analysis of multiparametric magnetic resonance imaging features and molecular subtypes of breast cancer.

Objectives: This study aims to evaluate the relationship between multiparametric magnetic resonance imaging (MRI) features - including T2-weighted imaging (T2WI), apparent diffusion coefficient (ADC), and dynamic contrast enhancement (DCE) - and molecular subtypes of breast cancer, to enhance non-invasive diagnostic stratification. Material and Methods: This retrospective study enrolled 134 consecutive patients with pathologically confirmed breast cancer. A comparative analysis was performed to evaluate intergroup variations in clinicopathological characteristics, morphological features, and multiparametric MRI parameters (including T2WI signal intensity, ADC value, early-phase enhancement rate, and time-intensity curve pattern) across the four molecular subtypes. Results: The cohort comprised 134 breast cancer patients stratified into molecular subtypes as follows: Luminal A (n = 22, 16.4%), Luminal B (n = 82, 61.2%), human epidermal growth factor receptor-2 (HER-2) (+) (n = 13, 9.7%), and triple-negative breast cancer (TNBC) (n = 17, 12.7%). Among the subtypes, there were statistically significant differences in terms of age, Ki-67 index, mass shape, margin, internal enhancement characteristic, T2WI signal, ADC value, early enhancement rate, and time intensity curve (TIC) pattern (P = 0.025; P < 0.001; P = 0.039; P < 0.001; P = 0.043; P = 0.014; P < 0.001; P = 0.009; and P = 0.020, respectively). Luminal subtypes predominantly exhibited irregular shapes, unclear/spiculated margins, heterogeneous enhancement, and uneven hypointense or isointense signal on T2WI. TNBC displayed regular shapes with smooth margins, ring enhancement, and uneven high signal on T2WI. The mean ADC value was significantly higher in HER-2 (+). Luminal A exhibited the highest early enhancement rate, while HER-2 (+) demonstrated the lowest. Analysis of TIC pattern revealed that type III curves were predominant across all subtypes, with a higher proportion observed in Luminal A and TNBC compared to Luminal B and HER-2 (+). Notably, no significant differences were observed between molecular subtypes in terms of menopausal status, axillary node metastasis, lesion type, number, size, and distribution, internal characteristics of non-mass enhancement lesions (P > 0.05). Conclusion: Multiparametric MRI features, particularly ADC values, DCE kinetics, and T2WI signals, demonstrate significant associations with breast cancer molecular subtypes. These imaging biomarkers offer potential for non-invasive subtype prediction, supporting more tailored diagnostic and treatment strategies.