OBJECTIVE: In advanced ovarian cancer (OC), poly(ADP-ribose) polymerase inhibitors (PARPi) have provided new maintenance therapy options with improved outcomes. Some patients on PARPi experience undesirable clinical events. Our aim was to assess the real-world clinical event experiences of patients on PARPis.
METHODS: A chart review using electronic medical records (EMR) was conducted in US adult OC patients at community and academic centers who initiated 2L/3L PARPi maintenance (olaparib, niraparib, rucaparib) between 03/2017–03/2019. EMR data were abstracted using electronic case report forms by a convenience sample of physicians from a research panel. Clinical events of interest (CEI) during the first 3 months of treatment, including anemia, neutropenia, thrombocytopenia, fatigue, insomnia, dysgeusia, nausea, vomiting, hypertension, rash, and diarrhea, were captured. Descriptive statistics were used to evaluate baseline demographic and clinical characteristics. Logistic regression controlled for potential prior confounders (see table footnote) associated with CEIs.
RESULTS: EMR data for 824 patients was abstracted by 119 physicians across the US. The most common specialties were medical/clinical oncology (43.7%, 52), hematology/oncology (29.4%, 35), and gynecologic oncology (19.3%, 23). Various practice settings were represented with 30 providers (25.2%) at academic hospitals, 11 (9.2%) at comprehensive cancer centers, and the remainder in community or private practice. Of 824 patients, 417 received olaparib, 269 niraparib, and 138 rucaparib. Most patients (65.5%, 540) were advanced or metastatic at diagnosis, with 62.4% (514) BRCA or other HRD positive, and most PARPi use was 2L maintenance (96.1%, 792). Differences were seen in baseline characteristics between PARPis including comorbidities and prior clinical events which were adjusted for in the analysis below. In particular 59.7% (249) of olaparib patients experienced a hematologic event on previous chemotherapy before initiating a PARPi compared to 39.8% (107) of niraparib, and 48.6% (64) of rucaparib patients. Among 2L PARPi patients, any CEI was reported in 29.1% (117) of olaparib, 38.9% (98) of niraparib, and 39.9% (55) of rucaparib patients. After adjusting for patient demographics, clinical characteristics, treatment prior to PARPi, patient starting dose, and history of clinical events, patients receiving niraparib were twice as likely to have a CEI (OR [odds ratio] 1.9, 95% CI 1.2–2.9) compared to olaparib. The likelihood of experiencing a CEI was similar between rucaparib and olaparib (OR: 1.3, 95% CI: 0.8-2.2).
CONCLUSIONS: In real world treatment of recurrent OC patients, substantial tolerability differences were observed between PARP inhibitors. These results may help providers select the most appropriate therapy for their patients and identify groups who may benefit from CEI management.
