Grant Number:	2R01CA229409-06A1 Interpret this number
Primary Investigator:	Carey, Lisa
Organization:	Univ Of North Carolina Chapel Hill
Project Title:	Optimizing Prognostic and Predictive Algorithms Using Residual Disease (RD) Genomic Biomarkers in Her2+ and Tnbc
Fiscal Year:	2025

Abstract HER2-positive (HER2+) and triple negative breast cancer (TNBC) are the most aggressive clinical breast cancer subtypes and share similar treatment approaches and unmet needs. Both are treated with either anti- HER2 drugs (for HER2+), or immunotherapy (for TNBC), added to polychemotherapy prior to surgery, called “neoadjuvant” therapy. Residual disease (RD) at surgery is seen in 45% of HER2+ and 35% of TNBC and carries a significantly poorer prognosis. For this reason, both HER2+ and TNBC receive additional therapy postoperatively ("adjuvantly"), namely an anti-HER2 antibody-drug conjugate (ADC) in HER2+, and additional chemotherapy added to ongoing immunotherapy in TNBC. This approach has resulted in significantly improved outcomes in both of these subtypes but is expensive, toxic, and overtreats many patients. We aim to better understand the biology and heterogeneity of the residual disease setting, including prediction of response to therapy and prognosis, in order to better tailor adjuvant therapy. We and others have previously identified a common theme in neoadjuvantly treated HER2+ and TNBC in that immune activation assessed by T cell and B cell gene expression in untreated tumors is a strong predictor of complete response (pCR) to chemotherapy, HER2-targeting, and immunotherapy, resulting in improved survival. However, we still lack biomarker information to tailor treatment in RD. We now propose to focus on the unmet needs of patients with RD in both HER2+ (Aim 1) and TNBC (Aim 2) by integrating and analyzing data from >2,500 women participating in clinical trials from each subset, by performing genomic analysis of RD specimens and using these data to improve prognostication in the two clinical subsets. We will compare paired untreated versus post-treatment RD tumors to identify how tumor and immune biology is changed in residual disease. Importantly, we will determine if biomarkers assayed from RD specimens improves prognostication and identification of those who may need, or not need, additional postoperative ADC (in HER2+) or chemotherapy (in TNBC). We will use supervised learning to build multi-feature computational prognostic models using RD specimens. We will examine the nature and role of residual disease tumor cells and the microenvironment by single-cell spatial transcriptomics to further understand how the interaction between both elements affects survival outcomes in patients with HER2+ and TNBC.