Correlation between mutation landscape and clinical outcomes of neoadjuvant therapy in HER2-positive breast cancer patients

579 Background: The standard management of early stage human epidermal growth factor receptor 2 (HER2) positive (+) breast cancer (BC) involves neoadjuvant therapy with combination of chemotherapy and HER2-targeted therapy followed by surgery. However, diverse pathologic responses were observed. We interrogated whether baseline genomic heterogeneity contributes to the varied therapeutic responses. Methods: Capture-based targeted sequencing using a panel consisting of 520 cancer-related genes, spanning 1.6MB of human genome, was performed on tissue biopsy samples, obtained prior to neoadjuvant therapy, of 33 HER2+ women with stage I-III BC. The median age of the cohort was 53. The correlation between genomic alterations and pathologic response were analyzed by multivariate analysis. Results: A majority of them was diagnosed with stage II (67%, 22/33), while 30% (10/33) had stage III and 3% (1/33) had stage I disease. 58% (19/33) were HR+ and 42% (14/33) were HR-. Mutation profiling of baseline samples revealed 349 mutations spanning 145 genes, with TP53, CDK12 and PIK3CA being the top 3 most frequently mutated genes. Neoadjuvant regimen was comprised of trastuzumab and HER2 inhibitor (i.e. pertuzumab or lapatinib). 15 patients used single HER2 inhibitor;18 used dual HER2 inhibitors. Endocrine therapy was also administered to HR+ patients (19/33) in combination with trastuzumab and HER2 inhibitor. Complete pathologic response (pCR) was observed in 45.5% (15/33) of patients. Interestingly, ROS1 copy number amplifications (CANs) were only identified in patients achieved pCR (p = 0.033). In contrast, missense mutations in PIK3CA and CNAs in CCND1, FGF19, FGF3, FGF4, SPOP, HNF1B and BRIP1 showed a trend of being less likely to mutate in pCR patients (p values between 0.05-0.1). Previous reports have suggested that pCR rates in HER2+ patients are associated with HR status. However, our data revealed comparable pathologic response of patients based on either HR status or neoadjuvant regimen. Conclusions: Our data revealed a distinct mutational profile between patients achieved pCR vs patients did not. Further studies with a larger cohort are required to confirm these findings.