Early genetic mutations associated with pathological complete response in triple-negative breast cancer and prognostic implications.

Abstract

e13132 Background: Triple-negative breast cancer (TNBC) with early-stage disease is typically treated with neoadjuvant chemotherapy with curative intent. Achieving a pathological complete response (pathCR) is associated with improved outcomes, yet predicting therapy response remains a challenge with distinct prognostic implications for node-positive and node-negative cases. We performed comprehensive genomic profiling to identify genetic mutations present in the tumor and blood of TNBC patients to attempt to detect biomarkers predictive of pathCR with the overriding goals to guide personalized therapeutic strategies and provide targets for longitudinal monitoring from diagnosis throughout the disease course. Methods: 25 TNBC patients provided matched, diagnostic solid tumor and blood samples prior to neoadjuvant treatment. Comprehensive genomic profiling was performed on DNA derived from tumors, plasma and buffy coat cells and the variants processed to identify tumor-specific and germline mutations for each patient. Results: 18 of the 25 TNBC patients achieved a pathCR while 7 had residual disease (RD). Comprehensive genomic profiling identified 5 tumor-specific mutations restricted to the pathCR group: CIC (22%; p = 0.17), NCOR1, NTRK1, PTEN, and RNF43 (17%; p = 0.24). Germline RUNX1 mutations were identified in 56% ( p = 0.06) and NOTCH4 in 50% ( p = 0.1) of the pathCR cases. A classifier was identified comprising above mentioned germline mutations plus 5 tumor-specific mutations that identified all pathCR patients while detecting only 1 with RD. We also compared the mutation profile of patients with (n = 10) or without nodal involvement (n = 15). 7 patients that achieved a pathCR were node positive as were 3 patients with RD. Discriminating mutations were predominantly specific to the node negative group including somatic mutations in 3 genes (PIK3R1 [27%; p = 0.07], PTEN and RNF43[20% each; p = 0.13]) along with 8 germline mutations specific to the node-negative group (ERBB2, ZNF703 [27%]; APC, BLM, CREBBP, DOT1L, FLT4, and NUTM1 [20%]). However there was no genomic signature that readily distinguished between these cohorts. These findings suggest that a limited number of germline mutations coupled with somatic mutations can predict which patients respond to neoadjuvant therapy irrespective of lymph node metastasis. Conclusions: This study identified blood-based genetic markers to predict pathCR among TNBC patients. This approach to characterizing TBNC tumors could help facilitate future therapeutic decisions and identification of targets for novel therapies. Our findings highlight differences in genetic profiles based on outcomes irrespective of other clinical characteristics of these patients. We are expanding our cohort and continuing our investigation of early markers, including the interplay between nodal involvement and genetic mutations.