Introduction Extensive genomic instability and heterogeneity are characteristics of high-grade serous ovarian cancer (HGSOC), with deficiency in homologous recombination (HR) repair has been reported for approximately 50% of HGSOC patients. Testing tumours for HR-deficiency (HRD) status is now a clinically applicable test, with a HR score of >42 suggestive of HRD and thus, platinum or PARP-inhibitor sensitivity in HGSOC. We aimed to determine the influence of spatial heterogeneity on HR scores in disseminated HGSOC.

Methods An algorithm detecting genomic scar HR scores was applied to genomic data from three cohorts of multi-site tumour samples – in-house Hammersmith hospital (HH) 45 patients n=5 tumours per patient; GSE38787 (n=24 patients) and GSE40546 (n=14 patients). A HR score is an unweighted sum of three independent measures of genomic instability: loss of heterozygosity (LOH), telomeric allelic imbalance (TAI), and large-scale state transition (LST). A cut-off >42 denotes HR-deficient.

Results Heterogeneity in tumour HR scores were detected in each cohort with patients presenting with either all HRD tumours, all HRP tumours or mixed HR scores, showing both HRD and HR-Proficient (HRP) tumour scores: HH (22%); GSE38787 (17%); and GSE40546 (28%). Within the HH cohort, survival analysis revealed that patients with all HRP tumours and mixed HR status had worse survival (progression-free survival p=0.0052; overall-survival p=0.00092) than patients with all HRD tumours.

Conclusion/Implications Our data demonstrating differences in HRD/HRP scores proposes that HR status may vary across disseminated HGSOC. Thus, implying that testing a single tumour biopsy may not accurately portray HGSOC tumour biology and could incorrectly guide clinical decision-making.