Pretreatment Amide Proton Transfer-Weighted Imaging Histogram Analysis Combined With ER-Negative and HER2-Positive Expression Predicts Pathologic Complete Response After Neoadjuvant Chemotherapy in Breast Cancer

To evaluate the predictive value of pre-treatment histogram analysis using APTWI, diffusion-weighted imaging (DWI), and early contrast-enhanced silhouette imaging in determining pathological complete response (pCR) post-NAC in breast cancer, and to investigate whether combining immunohistochemical indicators enhances predictive accuracy.

A retrospective continuous collection of 108 females with breast cancer who underwent NAC and pre-treatment APTWI, DWI, and dynamic contrast-enhanced imaging at our hospital. Clinical, MRI imaging, and pathological characteristics were analyzed for patients. NAC response was divided into pCR and non-pCR. Tumor segmentation and histogram feature extraction were performed on APT, ADC, and early contrast-enhanced silhouette images, and combined them with clinical features to construct an NAC efficacy prediction model. Diagnostic performance was assessed using receiver operating characteristic curves, with DeLong's test employed to compare areas under the curve (AUC).

In pCR group, mean, root-mean-square deviation, and 5th, 10th, 15th, 25th, 50th, 75th, 85th percentile of MTRasym, along with 1st percentiles of ADC were significantly higher in the pCR group than in the non-pCR group (p < 0.05). Conversely, the interquartile range of MTRasym and 10th percentiles of ADC were significantly lower in the pCR group (p < 0.05). ER-negative, HER2-positive expression, and 5th percentile MTRasym value were identified as independent predictors of pCR post-NAC (odds ratios, 0.16, 7.25, and 1.35, respectively). The combined diagnostic model demonstrated an AUC of 0.844, significantly outperforming individual parameters (p < 0.05).

Pre-treatment histogram analysis of MTRasym values derived from APTWI provides significant predictive value for pCR post-NAC in breast cancer. The combined diagnostic model incorporating APTWI with ER and HER2 expression status further enhances diagnostic performance.