Unmatched Cell Line Collections Are Not Optimal for Identification of PARP Inhibitor Response and Drug Synergies

Abstract

PARP inhibitors show great efficacy in BRCA1/2-mutated patients, but many preclinical studies, including combination therapies, fail to translate clinically, likely due to the limitations of preclinical models. This brief report aimed to identify appropriate cell line models to investigate PARP inhibitor sensitivity and synergies. An in silico study of cell line collections was performed to assess the correlation between BRCA1/2 mutations and sensitivity to PARP inhibitor monotherapy or with platinum-based chemotherapy. Subsequently, we characterised an isogenic model containing Brca1 and Brca2 mutations and investigated treatment response. Using cell line collections, BRCA1- and BRCA2-altered cell lines were not associated with increased sensitivity to PARP inhibitors. Other factors, including high PARP1 expression and low-level genome alterations, showed correlation with increased sensitivity to a PARP inhibitor. Furthermore, cell line collections did not reflect the improved patient outcomes arising from combination PARP inhibitor and platinum-based chemotherapy. In contrast, the ID8 isogenic model, with specific Brca1 and Brca2 mutations, reflected patient tumour-like responses to PARP inhibitor monotherapy and combination therapy. This study suggests exercising caution when using cell line collections as part of model selection when investigating PARP inhibitor sensitivity and synergy. Our data propose that using an isogenic preclinical model is more likely to accurately reflect patient tumour response. However, as this study was limited to a single isogenic model, further validation in additional systems will be required to broaden the scope of these observations.