Abstract #2738

Breast MRI for early prediction of residual disease following neoadjuvant chemotherapy: optimization of response cut-point by tumor subtype

Wen Li¹, Vignesh Arasu¹, Ella F Jones¹, David C Newitt¹, Lisa J Wilmes¹, John Kornak², Laura Esserman³, and Nola M Hylton¹

¹Radiology & Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, ²Epidemiology and Biostatistics, University of California San Francisco, San Francisco, CA, United States, ³Surgery, University of California San Francisco, San Francisco, CA, United States

This study demonstrated the effect of changing the cut-point of the functional tumor volume measured in breast MRI on the prediction of pathologic complete response (pCR) for breast cancer patients undergoing neoadjuvant chemotherapy. The study was performed using the retrospective data of a multi-center clinical trial as a full cohort and in subsets defined by clinically-relevant breast cancer subtypes. Optimal cut-point was selected by minimizing a penalty equation that considered different relative consequences of false negative and false positive predictions. Results showed that the optimal cut-point chosen in subtype had superior negative predictive value than using the one chosen from the full cohort.

This abstract and the presentation materials are available to members only; a login is required.

Join Here