Predicting Pathological Complete Response in Neoadjuvant Dual Blockade With Trastuzumab and Pertuzumab in HER2 Gene Amplified Breast Cancer

BACKGROUND: Dual-targeted therapy is the standard treatment for human epidermal growth factor receptor 2 (HER2)-positive breast cancer, and effective biomarkers to predict the response to neoadjuvant trastuzumab and pertuzumab treatment need further investigation. Here, we developed a predictive model to evaluate the dual-targeted neoadjuvant treatment efficacy in HER2 gene-amplified breast cancer. METHOD: This retrospective study included 159 HER2-amplified patients with locally advanced breast cancer who received neoadjuvant trastuzumab, pertuzumab, and chemotherapy. The correlation between clinicopathological factors and pathological complete response (pCR, in the breast and axilla) was evaluated. Patients were randomly assigned into the training set (n=110) and the testing set (n=49). We used an independent cohort (n=65) for external validation. We constructed our predictive nomogram model with the results of risk variables associated with pCR identified in the multivariate logistic analysis. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve, decision curve analysis, and calibration curves were employed to assess the nomogram’s performance. RESULTS: We revealed that the HER2/CEP17 ratio (p=0.001), CD8 levels (p=0.005), and histological grade (p=0.007) were independent indicators for pCR in dual-targeted neoadjuvant treatment after multivariate adjustment. The combined prediction efficacy of the three indicators was significantly higher than that of each single indicator alone. The AUCs were 0.819, 0.773, and 0.744 in the training, testing, and external validation sets, respectively. CONCLUSIONS: The HER2/CEP17 ratio, CD8 levels, and histological grade were significantly correlated with pCR in dual-targeted neoadjuvant treatment. The combined model using these three markers provided a better predictive value for pCR than the HER2/CEP17 ratio, CD8 levels, and the histological grade alone, which showed that an immunological effect partially mediates the predictive impact of neoadjuvant treatment.