e12643 Background: Biodynamic signatures (temporal patterns of microscopic motion within a three-dimensional tumor explant) offer a phenomic biomarker that is highly predictive for therapeutic response. The purpose of this study is to evaluate predictive accuracy of a biodynamic drug response classifier in breast cancer patients receiving neoadjuvant chemotherapy (NAC). Methods: Consecutive breast cancer patients from 4 institutions were screened for enrollment in a prospective observational study (NCT03164863). Treatment-naïve needle biopsies were delivered to a central laboratory where biodynamic signatures were measured in living tumor fragments challenged by standard-of-care cytotoxins. Patients received NAC per institutional guidelines and were followed through surgical intervention. A four-point classifier was trained to predict pathologic complete response (CR) then prospectively validated. Results: Among patients completing neoadjuvant treatment and surgical intervention, 33% (24 of 72) achieved CR. The biodynamic classifier predicted CR with 96% sensitivity and 97% negative predictive value. Biodynamically “favored” (scoring ≥ 3) and “strongly favored” (scoring 4) regimens produced CR at rates of 56% (23 of 41) and 73% (19 of 26), respectively. Only 3% of patients (1 of 31) achieved CR from regimens scoring 1 or 2. Area under the receiver operating curve (AUC) was 87% (95% CI: 75% to 93%, p < .0001), with similar performance across all subtypes and therapy cohorts (range: 84% to 89%). Performance of the classifier on training and validation patients was statistically equivalent. Conclusions: Biodynamic scoring accurately predicts response in breast cancer patients receiving NAC and holds promise to substantially improve management of these patients. [Table: see text]
Surgical Intervention Following Neoadjuvant Chemotherapy in Breast Cancer
