Introduction/Background Platinum sensitivity and homologous recombinaison deficiency (HRD) are predictive biomarkers for PARP inhibitors (PARPi) benefit in HGOC patients. The only validated assays to assess HRD are BRCA-mutation (BRCAmut) analysis and genomic instability scores (GIS) designed to detect genomic ‘scars’ in tumor DNA. However, these tests require large samples and yield non-contributive (NA) in 15% of cases. Bevacizumab and PARPi are approved as maintenance therapy regardless HRD status and the optimal maintenance strategy in case of non-contributive HRD test is a major unmet medical need. We aim to report the clinical characteristics and behavior under chemotherapy of NA HGOC pts.

Methodology This is a retrospective analysis of all pts tested for GIS by myChoice HRD Plus assay (Myriad Genetic Laboratories). Pts included presented HGOC with advanced FIGO III/IV diseases and treated according to guidelines. GIS was performed on baseline pretreatment samples, preferably. Platinum-free interval (PFI) was calculated from the date of last platinum-based chemotherapy to the date of relapse.

Results 210 patients were recruited: 100 were classified HRD negative (HRD-, score <42), 81 HRD positive (HRD+, score ≥42) and 29 NA (14%). HRD+ cohort was significantly enriched with BRCAmut pts (21/81 = 27%) compared to HRD- and NA. In the NA cohort, median age was 64 years, 86% had an high-grade serous tumor and 10% presented germinal BRCAmut. With a median follow-up of 39 months, median PFI in the overall population was 19.8 months (95%CI 16.7–24.4). In the HRD+, HRD- and NA cohorts (excluding BRCAmut), median PFI were 34.0 (95%CI 16.7–64.4), 14.6 (95%CI 12.0–20.9) and 37.3 months (95%CI 21.0-NA) respectively (P=0.004).

Conclusion Our results suggest that patients with NA GIS results behave like HRD+ tumors harboring high platinum-sensitivity and therefore may benefit from PARPi maintenance. The reason for non-contributivity in the first place is unknown and may explain these observations.