The value of non-enhanced magnetic resonance imaging based on T2-weighted imaging and apparent diffusion coefficient in the study of three-tiered HER2 status in patients with breast cancer

Background: Preoperative assessment of human epidermal growth factor receptor 2 (HER2) status relies on invasive biopsy, which is subject to sampling limitations. Meanwhile, conventional contrast-enhanced magnetic resonance imaging (MRI) and radiomics face barriers to widespread adoption due to high costs, complexity, and poor reproducibility. In contrast, non-contrast MRI based on T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI)/apparent diffusion coefficient (ADC) offers a non-invasive, cost-effective, and safe alternative. This study aimed to evaluate whether non-contrast T2WI combined with ADC can non-invasively stratify HER2-zero, HER2-low, and HER2-overexpressing invasive breast cancer.

Methods: In this retrospective single-center study, 361 women with surgically confirmed invasive breast cancer (February 2015–December 2023) were stratified into HER2-zero expressing (n=63), HER2-low expressing (n=181), and HER2-overexpressing (n=117) cohorts. All underwent preoperative 3 Tesla (T) breast MRI including fat-suppressed (FS) T2WI and DWI. Two readers assessed breast edema score (BES 1–4) and lesion-to-muscle T2 signal ratio (L/M). ADC maps were generated based on the b values of 50 and 800 s/mm² or 0 and 1,000 s/mm² DWI MRI data, and clinicopathologic variables were integrated into a predictive model. Independent predictors were identified by multivariable logistic regression and receiver operating characteristic (ROC) analysis.

Results: Tumor edema extent differed significantly between HER2-zero and HER2-low (P=0.017) and between HER2-low and HER2-overexpressing (P=0.021). T2 signal intensity (SI) also differed between HER2-zero and HER2-low (P=0.012) and between HER2-low and HER2-overexpressing (P=0.001). Multivariate logistic regression analysis revealed that intratumoral T2FS hyperintensity (P=0.001) and L/M ratio (P=0.001) were independent predictors for differentiating the HER2-zero group from the HER2-low group, whereas ADC values (P<0.001), L/M ratio (P=0.028) and Ki-67 levels (P<0.001) were independent predictors for distinguishing the HER2-low group from the HER2-high group. The predictive model achieved areas under the curve (AUCs) of 0.737 for differentiating HER2-zero versus HER2-low and 0.761 for differentiating HER2-low versus HER2-overexpressing.

Conclusions: T2WI-derived edema and SI, together with ADC, may serve as valuable noninvasive MRI biomarkers to complement existing approaches for HER2 level assessment, offering a simple and reproducible adjunctive screening method in clinical practice.