Exploiting the spectrum of response of BRCA-associated PDAC to platinum/PARPi treatment

Background BRCA1/2-associated pancreatic ductal adenocarcinoma (PDAC) tumors have DNA-damage repair deficiency rendering them sensitivity to platinum/PARPi. The vast heterogeneity in response to treatment exemplified in our POLO study by extended progression-free survival with maintenance PARPi versus placebo (7.4 vs. 3.8 months, respectively). There are patients demonstrating durable response >24 months (two patients with complete response) and patients with limited response and disease progression within 6 months.

Methods We have generated unique BRCA-associated PDAC patient-derived xenograft (PDX) models from metastatic lesions (n=17) at distinct time points of disease, closely mimicking patient`s clinical scenario. In collaboration with Prof. Gallinger from the OICR, we have performed whole-genome sequencing (WGS) analysis of >40 BRCA1/2-associated PDXs and tumor/metastatic samples.

Results We demonstrated heterogeneity in BRCA-associated tumors` genomic features; tumors are divided into HR-deficient (HRD) and HR-proficient (HRP) subgroups based on WGS analysis using a unique algorithm developed by Gallinger et al. HRD subgroup demonstrated beneficial median overall survival. Notably, none of the HRP tumors showed somatic inactivation of the second copy of BRCA1/2, including loss of heterozygocity (LOH), whereas in the HRD group majority of the tumors underwent LOH or other type of somatic inactivation. Hence, we assume that lack of LOH can serve as a surrogate marker for HRP status of the tumor in the majority of cases.

Conclusions Classification of the BRCA-associated PDAC tumors into HRD and HRP groups based on their genomic background might predict response to treatment and thus allow choosing alternative therapeutic combinations in addition to standard of care protocols.