Abstract
Background: Poly ADP-ribose polymerase inhibitors (PARPi) have improved outcomes in cancers, particularly ovarian and breast cancer. With their increasing use, it is important to track and manage potential adverse events (AEs). One important AE pattern with PARPi therapy is hematologic toxicity. We aim to review real-world data on hematologic toxicities with PARPi using the FDA Adverse Event Reporting System (FAERS) database.
Methods: We queried the FAERS database from 2015 to 2025 using a search-by-product strategy and retrieved 12,413 adverse events. We employed 4 PARPi drugs (included generic and brand names), olaparib, niraparib, rucaparib, and talazoparib were coded as a primary suspect drug. Hematologic AEs were defined using a curated MedDRA preferred term dictionary including cytopenias, bone marrow suppression/failure, myelodysplastic syndrome (MDS), acute myeloid leukemia (AML), and coagulation disorders; thrombotic and non-hematologic bleeding events were excluded. Descriptive statistics were performed, and disproportionality analysis was conducted by calculating the reporting odds ratio (ROR) with 95% confidence intervals (CI) and the information component (IC) with 95% credibility intervals. ROR was considered significant when the lower limit of the 95% CI was >1. RORs were calculated for each hematologic AE category by agent. We evaluated trends across two eras (2015–2021 vs 2022–2025) and modeled serious outcomes, including death, hospitalization, and life-threatening events.
Results: Total AEs from all included PARP were 44916, and the hematologic AEs were 27.64% of all AEs. ROR for any event was highest with anemia with niraparib (54.03, 95% CI 51.00–57.25), followed by thrombocytopenia with niraparib (33.28, 95% CI 32.24–34.35) and MDS with olaparib (32.53, 95% CI 28.62–36.99). For olaparib, myelosuppression had an ROR of 11.97 (95% CI 10.63–13.46) and pancytopenia 5.15 (95% CI 4.53–5.86). Rucaparib showed the highest anemia signal (ROR 26.36, 95% CI 23.38–29.71) among its categories, along with thrombocytopenia (ROR 8.87, 95% CI 8.22–9.57). Talazoparib demonstrated notable signals for pancytopenia (ROR 10.33, 95% CI 7.22–14.77) and MDS (ROR 15.26, 95% CI 7.61–30.61), despite lower total counts. Cytopenia signal strength remained stable across eras, while RORs for MDS modestly increased for olaparib and niraparib. AML and MDS were associated with the highest fatality proportions (up to 38% and 31%, respectively). Anemia and thrombocytopenia were more often linked to hospitalization and life-threatening events. Across the two eras, the AEs remained consistent, but the severity profile shifted slightly in the latter era.
Conclusions: This is the largest real-world study to date demonstrating specific hematologic AEs with PARPi. Our results show distinct and varying patterns of hematologic toxicity across PARPi, with the strongest disproportionality observed for cytopenias, particularly thrombocytopenia, leukopenia, and anemia, and notable signals for MDS. Recognizing these patterns will help clinicians better monitor and manage patients receiving PARPi therapy.
