Over the past 7 years, four PARP (poly[ADP-ribose] polymerase) inhibitors, niraparib, olaparib, rucaparib, and talazoparib, have been approved in the United States and have led to clinically significant improvements in progression-free survival in patients with recurrent or primary ovarian cancer (OC). However, recent data have come to light that warn of potentially serious adverse events associated with this class of medications.

Investigators conducted a safety meta-analysis of randomized, controlled trials (RCTs) and a retrospective study of the World Health Organization’s (WHO’s) pharmacovigilance database, VigiBase, to determine the risk of developing myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) related to PARP inhibitors to assess the incidence and risk of PARP-inhibitor-related MDS and AML in randomized clinical trials and to describe the clinical features of these hematological disorders based on data found in VigiBase.

The authors gathered RCT data through May 31, 2020 from Medline, Cochrane Central Register of Controlled Trials, and the ClinicalTrials.gov on RCTs comparing PARP inhibitors to placebo or nonplacebo controls in adult patients with cancer for the meta-analysis. A stepwise approach was undertaken to identify rare and delayed adverse events, utilizing data from Clinicaltrials.gov, the published literature, and finally, if needed, from the clinical investigators of PARP trials. Studies had to be available in English and contain information on adverse events to be included in the meta-analysis. The pharmacovigilance study involved searching VigiBase for reports of MDS or AML through to May 2020. The VigiBase database is home to over 21 million anonymized cases of adverse events from 130 countries worldwide.

The investigators calculated a Peto odds ratio (OR), which is the most accurate method employed to analyze data from binary studies with rare events or events that occur in <1% of the population.

Twenty-eight studies were included in the safety meta-analysis, 18 placebo and 10 nonplacebo RCTs, which enrolled 9,099 patients, 5,693 and 3,406, respectively. Analysis of these data revealed that PARP inhibitors were associated with a significantly increased risk of MDS and AML compared with placebo (Peto OR) 2.63, P = .026). Although RCTs included studies with various types of malignancy, all MDS and AML cases were reported in RCTs involving patients with OC. The median latency period for the development of these hematological abnormalities was 20.3 months and ranged from 18.4 to 26.6 months. The overall incidence of MDS and AML associated with PARP inhibitors was 0.73% compared with 0.47% for the control group across placebo-controlled RCTs. When both placebo-controlled and nonplacebo RCTs were included, these incidences were 0.83% and 0.66%, respectively.

A review of VigiBase data revealed a total of 178 cases—99 cases of MDS and 79 cases of AML—associated with PARP-inhibitor therapy. Of these cases, 85% occurred in women with OC. The median treatment duration of the PARP inhibitor was 9.8 months (range: 0.2-66.8 months). The latency period for MDS and AML from first exposure to the PARP inhibitor was 17.8 months (range: 0.6-66.8 months); MDS occurred after a median of 17.8 months and AML after 20.6 months. In 8% of patients, MDS progressed to AML.

Cytopenia were also reported in 40% of MDS and AML patients, with one-quarter of these occurring after a median of 7 months following PARP-inhibitor initiation. Outcomes were available for 104 of the 178 cases; of these, 9% recovered or were recovering from MDS/AML, 46% had persistent disease, and 45% died from MDS/AML.

Pharmacists need to be vigilant and closely monitor OC patients on PARP inhibitors for signs and symptoms of delayed hematological toxicities.

The content contained in this article is for informational purposes only. The content is not intended to be a substitute for professional advice. Reliance on any information provided in this article is solely at your own risk.

 « Click here to return to Specialty Pharmacy Update.