Introduction/Background Although randomised controlled trials (RCTs) have demonstrated the benefit of PARP inhibitors and bevacizumab as monotherapies and combination therapies, there is limited direct head-to-head evidence of their relative clinical efficacy.

In the PRIMA study, niraparib demonstrated a clinically significant improvement in progression-free survival (PFS) compared with placebo, as a first-line (1L) ovarian cancer (OC) maintenance therapy.

The objectives of the study were to assess feasibility of an indirect treatment comparison (ITC) and a population-adjusted indirect treatment comparison (PAIC) for estimating the relative efficacy of niraparib compared with olaparib, olaparib plus bevacizumab, and bevacizumab as maintenance following 1L chemotherapy in OC. The study focused on fully powered statistical cohorts.

Methodology Trials included in the ITC analysis were based on a systematic literature review conducted in February 2020.

Guidelines from the Cochrane Handbook for Systematic Reviews of Interventions were used to assess the level of heterogeneity across the studies in terms of designs, population characteristics, treatment arms and outcome measures.

The feasibility of PAIC for the PRIMA and PAOLA-1 trials was assessed based on assumptions outlined in the guidance by the Decision Support Unit in NICE DSU Technical Support Document 18; PFS was the outcome for the analysis.

Results All 12 RCTs assessed for ITC feasibility were excluded based on various factors including: the lack of a common comparator with PRIMA within the network (ICON-7, GOG-0218, PAOLA-1, VELIA/GOG-3005); differing measurement of PFS and overall survival starting timepoint due to trial design (ICON-7, GOG-0218, VELIA/GOG-3005); inclusion of stage III patients with no visible residual disease following debulking surgery (PAOLA-1, SOLO-1, VELIA/GOG-3005); disparity between disease biomarker (SOLO-1).

For the PAIC, three fundamental differences between the PRIMA and PAOLA trials were identified; inclusion criterion related to residual disease was wider in PAOLA meaning that the ‘conditional constancy of absolute effects’ was violated; receipt of neoadjuvant chemotherapy was identified as a confounding factor that would bias a PAIC; discrepancies in the assessment of type and frequency of measurement of the PFS outcome.

Conclusions Based on the evidence currently available, neither an ITC nor PAIC would meet current guidelines, such as those outlined by the International Society for Pharmacoeconomics and Outcomes Research, for these analyses. Their results would not be considered appropriate evidence for use in clinical decision making or reimbursement decisions. The extent of imbalance caused by differences in the patient inclusion/exclusion criteria for intended comparisons is unknown and a recognised limitation of PAICs.

Disclosures This study was funded by GlaxoSmithKline.

Due to lengthy author disclosures, author COI information will be provided directly to the congress.