Outcomes for patients with platinum-sensitive recurrent ovarian cancer treated with prolonged PARP inhibitor therapy.

e17607 Background: PARP inhibitors (PARPi) can be an effective tool for treatment and maintenance in platinum-sensitive recurrent epithelial ovarian cancer (PSROC), especially in those with somatic and germline BRCA mutations. In patients with PSROC with prolonged response to PARPi greater than 24 months without progression, little data is available to guide length of therapy and toxicity counseling. This retrospective cohort study aimed to describe survival and toxicity outcomes of these prolonged responders. Methods: The electronic medical records of a single tertiary academic center were searched to identify all patients with platinum-sensitive recurrence of epithelial ovarian cancer treated with at least 24 months of a PARP inhibitor. Individuals who progressed before 24 months of therapy were excluded. Medical records were then manually abstracted for data on demographics, cancer pathology, treatment and survival details, and toxicity. Descriptive statistics were performed with survival analysis using log-rank testing for comparisons. Results: 41 individuals were identified as having a prolonged response to a PARPi. Median age at PARPi initiation was 64 years (range 45 – 80), with the majority having had only two prior lines of chemotherapy (n=34, 82.9%). 24 individuals carried a germline or somatic pathogenic variant in BRCA1 , BRCA2 , or RAD51C (58.5%), and an additional 4 had tumors deficient in homologous recombination repair (HRD). Olaparib was most prescribed (n=19, 46.3%) followed by niraparib (n=14, 34.2%). After at least 24 months of therapy, 7 individuals discontinued initial PARPi therapy for patient or provider preference (17.1%), 4 for toxicity (9.8%), and 14 for progression (34.2%), though 13 (31.7%) had a PARPi rechallenge for subsequent therapy. In total, 23 individuals (56.1%) progressed during follow-up, with a median progression-free survival (PFS) of 50 months (range 24-81 months), and 5-year PFS of 38.8% (95% CI 22.4 - 55.0%). 5- and 10-year overall survival was 87.5% (95% CI 69.9 – 95.2%) and 82.4% (95% CI 61.6 – 92.5%), respectively. HRD status did not affect PFS (log-rank p=0.83), but patients younger than 65 at initiation of PARPi had decreased PFS compared to older individuals (median 40 months vs. not reached, log-rank p=0.04). Including subsequent line treatments with PARPi, the median length of PARPi therapy in the cohort was 46 months (range 24-125 months). Two individuals (4.9%) were diagnosed with myelodysplastic syndrome during the follow-up period (PARPi exposure 29 and 59 months, respectively). Conclusions: Individuals with PSROC who display a prolonged response (at least 24 months) to PARPi therapy still face a significant recurrence risk of over 60% at five years, suggesting that indefinite treatment may be warranted despite the risks of hematopoietic toxicity. Comparative trial data will be useful to define the optimal duration of therapy.