Determinants of PARP-inhibitor response in Breast Cancer Patients without germline mutations in BRCA1/2 - Summary. This R21 project aims to identify molecular features predisposing to response, or resistance, to PARP-inhibition in metastatic breast cancer (MBC) not related to a germline (g) BRCA1/2 mutation. Toward that goal, we will analyze pre-treatment tumor DNA and pre- and post-treatment circulating tumor DNA (ctDNA) in collaboration with the BROAD Institute. The study “Olaparib Expanded,” NCT03344965, accrued MBC patients with gene mutations related to homologous recombination deficiency (HRD) other than gBRCA1/2. It evaluated the response to PARP-inhibition in an initial (n=54) and an expansion (n=54) cohort. In the initial cohort, 87% of the patients had somatic BRCA1/2 mutations (sBRCA1/2), PALB2, ATM, or CHEK2 mutations. Confirmed responses in this first cohort were seen in patients with gPALB2 MUT (overall response rate (ORR) 82%) and sBRCA1/2 MUT (ORR, 50%). In the expansion, ORR for gPALB2MUT BC was 75%, and 37% for sBRCA1/2MUT BC. Median progression-free survival (mPFS) for gPALB2MUT was 13.3 months initially, and 9.6 months in the expansion cohort. MPFS for sBRCAMUT BC was 6.3 months initially and 5.6 months subsequently. This landmark study established PARP-inhibition as an effective treatment for patients with MBC and gPALB2MUT or sBRCA1/2MUT, significantly expanding the population of BC patients likely to benefit from PARP-inhibition beyond gBRCA1/2MUT1. Mutant allele variant fraction (MAF) was evaluated for 33 who had sBRCAMUT; the mean MAF was 43% in responders and 39% in non-responders (p=0.7), and thus did not explain the difference in response. Therefore, the trial left critical unanswered questions: Why did more than half of the patients with sBRCAMUT not respond to PARP-inhibition, and why are their responses shorter than historically observed for gBRCAMUTrelated BC? Why do some tumors with initial response develop resistance much more rapidly than others? Are there co- mutations that, if targeted, could extend PFS? For the analysis of this trial, we have pre-study tissue biopsies for whole exome sequencing (WES) and pre- and post-study samples of circulating tumor DNA (ctDNA). Initial sequencing has indicated high quality DNA. We aim to analyze tumor tissue and ctDNA for 108 patients from both the initial and the expansion cohort and test putative mechanisms that modulate PARP-inhibitor sensitivity ex vivo. This project will be conducted as a collaboration between Drs. Wulf, Tung, Heng (all BIDMC), Garber (DFCI), Getz (BROAD institute). After this project, we will know if and how frequently reversion mutations in PALB2 occur, something that is currently unclear; we will understand to what extent reversion mutations in sBRCA1MUT BC are responsible for the relatively short-lived responses to PARP-inhibition. We will also understand which somatic molecular alterations sensitize a tumor to PARP-inhibition in tumors without reversion mutations, and whether these are causally related. Data will be shared via dbGAP and will be hypothesis- generating for novel PARP-inhibitor combinations.
