A clinical model incorporating early lymphocyte dynamics in HER2-positive breast cancer

Background Immune activation is central to response to HER2-targeted therapy in HER2-positive breast cancer; however, early and scalable biomarkers that capture this biology remain limited. Methods We developed a predictive model that used early dynamic changes in routinely obtained peripheral blood absolute lymphocyte count (ALC) alongside standard clinical and pathologic features to predict pathologic complete response (pCR) after neoadjuvant therapy in HER2-positive breast cancer. Results In a discovery cohort of patients with stage I–III HER2-positive breast cancer treated per routine care with trastuzumab-based neoadjuvant therapy (n = 154), higher baseline ALC and greater decline in ALC after the first cycle were independently associated with increased likelihood of pCR (p < 0.001 for both). A predictive model using these two ALC variables alongside age, percentage of cells expressing estrogen receptor, progesterone receptor status, HER2 immunohistochemistry score, and tumor stage, demonstrated strong discrimination: the area under the curve (AUC) in the discovery cohort was 0.86, and 0.82 in a separate validation cohort (n = 83). In the validation cohort, the pCR rate was 12% among patients in the discovery model-predicted quartile with lowest pCR probability, 33% in the second quartile, 69% in the third quartile, and 79% in the highest probability quartile. Conclusion These findings suggest that early measures of peripheral blood ALC, alongside standard clinical and pathologic features, offer a simple, scalable biomarker for anticipating response to HER2-targeted neoadjuvant therapy, with potential utility for early treatment adaptation and trial stratification.