Objective/rationale
High-grade serous ovarian cancer (HGSOC) accounts for the majority of ovarian cancer deaths. ~20% of HGSOC patients have mutations in BRCA1 and BRCA2 that make them sensitive to the targeted PARP inhibitors. A major challenge is that most patients develop PARP inhibitor resistance, leading to the need for combination or subsequent therapies. Reliable HGSOC models are required to develop new therapies.
From an in silico study of 1065 cell lines, we found that cell lines with BRCA1/2 mutation are not more responsive to PARP inhibitors than cell lines that are BRCA1/2 wildtype, which limits their utility for drug combination studies with PARP inhibitors. We aimed to characterise representative in vitro models of HGSOC that demonstrate PARP inhibitor sensitivity that is dependent on BRCA1/2 mutation, and can be used to mimic patient response during treatment response and resistance.
Method
We used the ID8 immortalised surface epithelial mouse cells which forms HGSOC-like disease in vivo. We characterised the response of ID8 p53-/-, ID8 p53-/-Brca1-/- and ID8 p53-/-Brca2-/- to the PARP inhibitor olaparib. We next generated olaparib resistant cell lines of ID8 p53-/-Brca1-/- and ID8 p53-/-Brca2-/- to mimic drug resistance in the clinical setting. Using colony forming assays, western blotting, and synergy assays, we investigated the response of olaparib sensitive and resistant cell lines to olaparib alone or in combination with chemotherapy. Finally, we benchmarked the response of ID8 cell lines to the response seen to combination therapy of patients with or without BRCA1/BRCA2 mutation in clinical trials.
Results
Using the isogenic ID8 p53-/-, p53-/-Brca1-/- and p53-/-Brca2-/- models we demonstrated that BRCA1 and BRCA2 mutation induces sensitivity to olaparib in these isogenic models. We next identified that the combination of olaparib + chemotherapy was effective in p53-/-Brca1-/- and p53-/-Brca2-/- cells, but not in the p53-/- Brca1/2 wt model using colony forming assays, measures of DNA damage, and synergy assays. Synergy to olaparib + chemotherapy was lost in p53-/-Brca1-/- and p53-/-Brca2-/- models that were made resistant to olaparib. Our parallel meta-analysis of clinical trials showed that only patients with BRCA1/BRCA2 mutation have enhanced sensitivity to PARP inhibitor and chemotherapy combination treatment. Furthermore, emerging data shows that PARP inhibitor resistant patients lose sensitivity to chemotherapy.
Conclusion
We have generated and characterised models of HGSOC PARP inhibitor sensitivity and resistance that recapitulate the clinical disease. These models will be an invaluable resource in performing subsequent investigations for new therapies for HGSOC patients.