EORTC 2016-Patient-derived xenograft models of BRCA+/BRCA-like ovarian tumors reflect clinical disease responses to PARP inhibition

Title

Patient-derived xenograft models of BRCA+/BRCA-like ovarian tumors reflect clinical disease responses to PARP inhibition

Authors

Angela Davies1, June Hou2, le-Ming Shih3 Jason Wright2, Daniel Ciznadija1, Amanda Katz1, and David Sidransky1 Champions Oncology, Baltimore, MD, USA. 2 Columbia University Medical Center, New York, NY. 3 Johns Hopkins University School of Medicine, Baltimore, MD, USA

Presented at

EORTC 2016

Abstract

PARP inhibition in BRCA-deficient/BRCA-like ovarian tumors can lead to synthetic lethality. The FDA has approved a PARP inhibitor to treat BRCA-deficient ovarian carcinomas and other drugs targeting PARPs (PARPi) are in development, generating further interest for this target. PDX models could enable evaluation of PARPi resistance/response mechanisms, improving patient stratification protocols and optimizing therapeutic strategies. In this pilot study, we evaluated responses to olaparib and carboplatin/paclitaxel in BRCA-deficient/BRCA-like ovarian PDX models and correlated responses to clinical trial outcomes.
PDX models developed from ovarian patient tumors were subjected to next-generation sequencing to identify genomic alterations in BRCA 1/2 and regulators of alternative DNA repair processes that contribute to a BRCA-like phenotype (e.g. ATM, RAD51, FANCA, and FANCD2). Models were screened against the FDA-approved PARPi, olaparib and carboplatin/paclitaxel. Tumor regression (TR) values were calculated, aligned to clinical RECIST criteria, and correlated with known literature-based response rates (RR).
In this pilot study, RR to olaparib in ovarian PDX models with BRCA mutations/BRCA-like phenotype was similar to clinical trial outcomes. In keeping with literature reports, platinum pretreatment did not affect model responsiveness to olaparib. Nevertheless, whilst the majority of models responding to olaparib harbored mutations in BRCA/BRCA-like genes, one was wild type with respect to these pathways. Further, other BRCA mutated/BRCA-like models failed to respond to PARP inhibition. This suggests a need for further interrogation of de novo resistance/sensitivity mechanisms, which the PDX platform is well suited to uncover.