Objective: To estimate the risk of myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) secondary to PARP inhibitors (PARPi), based on the line of treatment, in real-world ovarian cancer (OC) patients.
Methods: Using the TriNetX platform, we compared a cohort (experimental A) of 3402 OC patients treated with first-line maintenance PARPi to a control cohort of 1653 OC patients treated with platinum-based chemotherapy without PARPi. Experimental group B included 356 OC patients treated with PARPi after a platinum-sensitive relapse and was compared to a control cohort of 1503 patients who had not received PARPi after 2 lines of platinum-based chemotherapy. The cohorts were propensity score matched (PSM) 1:1 (experimental A vs control 1 and experimental B vs control 2) for age, race, bevacizumab treatment, and genetic susceptibility to neoplasms. A hazard ratio (HR) was used to compare the incidence of MDS and AML between the matched cohorts.
Results: In the first-line setting, 2 groups of 1346 matched OC patients (mean age 59.8 ± 10.2 SD) were evaluated. The overall incidence of MDS or AML was 1.9% and 0.1% in the experimental A and control groups, respectively (HR = 2.46; 95% CI 1.27-4.75, P = .006). For the platinum--sensitive relapse setting, the HR was 1.76 (95% CI 0.42-7.37, P = .432). No significant differences were observed between the various PARPi used.
Conclusions: Our study indicates that PARPi may increase the risk of MDS or AML after first-line maintenance treatment. No significant differences were found across the types of PARPi used.
Keywords: PARP inhibitor; acute myeloid leukemia; myelodysplastic syndrome; ovarian cancer; real-world evidence.
© The Author(s) 2025. Published by Oxford University Press.