Background

An impaired DNA damage response (DDR) permits genomic aberration accrual and is a major chemosensitivity determinant. The DDR comprises single-strand break (SSB) (BER, NER, MMR) and double-strand break (DSB) (HR, NHEJ) pathways. Pathway redundancy is complex and rarely studied alongside mitochondrial dysfunction or ROS both of which influence chemoresistance. Ovarian cancer is hallmarked by chromosomal instability and is ideally positioned for DDR profiling. We developed a functional assay suite to monitor DDR, mitochondrial, and ROS capacities and trialled this on explants from 16 ovarian cancer patients undergoing primary or interval debulking surgery. Signatures were correlated with patient progression-free (PFS) and overall survival (OS).

Method

Solid tumour and ascitic explants were established from patients receiving platinum chemotherapy. HR was assessed by our Rad51 assay; NHEJ by mismatched expression plasmid DNA DSB repair; BER and NER by comet assays; MMR by exome sequence mutational impact. Carboplatin cytotoxicity (MTS), ROS (DCFDA), and mitochondrial membrane potential (JC-10) assays were conducted. Explants and patients were classified (discriminant analysis, neural networks, support vector machines).

Results

Patient DDR dysregulation was wide-ranging. Defective HR (HRD) and NHEJ were near-mutually exclusive. 44% patients were HRD while partially perturbed and basal competence signatures were evident for remaining pathways. HRD cells had increased SSB pathway abrogation vs NHEJ-defective cells. Homologous recombination-competent (HRC) phenotypes had increased inter-explant heterogeneity (44.4% vs 14.2% HRD), decreased intra-explant heterogeneity (0.71FC vs HRD), and associated with perturbed mitochondria (78% vs 57% HRD). A modest significant inverse correlation between explant cytotoxicity and patient survival occurred (r=-0.41, p=0.042) and every relapsed patient harboured perturbed mitochondria. DDR abrogation patterns permitted construction of DDR signatures which classify explant platinum cytotoxicity, mitochondrial membrane dysregulation, and ROS response (AUCs:1). Of importance, these signatures fully classify and predict patient PFS and OS (15 months to-date, AUC:1).

Conclusion

HRC-associated chemoresistance is established however emphases on individual pathways are mechanistically insufficient to capture DDR landscapes. Our functional assay suite provides a holistic view of explant DDR and platinum cytotoxicity signatures to classify and predict ovarian cancer patient survival.

Impact statement

Our real-time functional assay suite demonstrates promise as a translational patient chemosensitivity prediction tool.