Poly (ADP-ribose) polymerase (PARP) inhibitor regimens for ovarian cancer in phase III randomized controlled trials: a network meta-analysis

Introduction We aimed to evaluate poly (ADP-ribose) polymerase (PARP) inhibitor (PARPi) regimens in BRCA-mutated ovarian cancer for patients responsive to front-line platinum (bevacizumab and olaparib, veliparib and chemotherapy, olaparib) or platinum-sensitive relapsed (olaparib, rucaprib, niraparib) patients in phase III randomized controlled trials. Methods A network meta-analysis was utilized to generate the direct and indirect comparisons. The primary outcomes for network meta-analysis were efficacy (hazard ratios for progression-free survival in BRCA mutation cohort) and toxicity (odds ratios for all grade 3-4 adverse events). The American Society of Clinical Oncology (ASCO) value framework was used to assess the cost-effectiveness of the PARPi regimens. Results Network meta-analysis indicated no statistically significant differences in efficacy and toxicity among the assessed upfront or relapsed PARPi regimens (95% CI included 1). The ASCO value framework indicated that current PARPi regimens were similar in clinical benefits, toxicity, and net health benefit in the upfront (bevacizumab and olaparib, veliparib and chemotherapy, olaparib) and relapsed setting (olaparib, rucaprib, niraparib). The addition of bevacizumab to olaparib ($353.72) increased the cost per unit net health benefit for patients compared with olaparib monotherapy ($260.57). The upfront PARPi regimens had lower toxic scores than the regimens used at relapse. Conclusions The choice of PARPi regimens both in the upfront and relapsed setting should consider not only efficacy and toxicity but also costs in BRCA mutation patients. Current combining PARPi regimens are not recommended for such patients in the upfront setting from the cost-effective perspective. Upfront PARPi regimens are less toxic than those used at relapse.