Clinical value of atezolizumab + bevacizumab for first-line unresectable hepatocellular carcinoma (HCC): A network meta-analysis.

4585 Background: The IMbrave150 pivotal study in unresectable HCC showed superiority of atezolizumab + bevacizumab (atezo + bev) vs sorafenib for OS and PFS. Based on these data supporting first-line atezo + bev for HCC, we conducted a network meta-analysis (NMA) to compare the efficacy of atezo + bev with other systemic and local therapies approved for HCC. Methods: A systematic literature review identified randomized controlled trials in adults with locally advanced or metastatic HCC and no prior systemic therapy for HCC. Studies of therapies now approved for any line of HCC treatment with data reported for first-line treatment since sorafenib approval in 2007 were eligible. Screening of 8783 records yielded 55 trials for inclusion; 9 studies were eligible for the evidence network. Reported hazard ratios (HRs) for OS and PFS were extracted from published studies. The IMbrave150, REFLECT and CheckMate-459 study populations were considered sufficiently similar to compare. A generalized linear model with random effects was used to estimate indirect treatment effects. Informative priors for the heterogeneity of treatment effects across trials were adopted given the limited number of trials to inform each pairwise comparison. HRs with 95% credible intervals (CrIs) and Bayesian posterior probability of atezo + bev being superior to other treatments were calculated for each treatment comparison. The base case NMA compared the relative efficacy of atezo + bev vs sorafenib observed in the IMbrave150 study with the relative effect of other therapies. Sensitivity analyses were performed to compare subgroup results as appropriate based on disease etiology, extrahepatic spread and geography. Results: NMA results suggested improved OS with atezo + bev vs lenvatinib (HR, 0.63; 95% CrI: 0.32, 1.25; probability of atezo + bev being superior to lenvatinib: 93.7%) or nivolumab (HR, 0.68; 95% CrI: 0.35, 1.38; probability of atezo + bev being superior to nivolumab: 90.3%) and improved PFS with atezo + bev vs lenvatinib (HR, 0.91; 95% CrI: 0.23, 3.65; probability of atezo + bev being superior to lenvatinib: 61.5%) or nivolumab (HR, 0.63; 95% CrI: 0.16, 2.59; probability of atezo + bev being superior to nivolumab: 85.5%). Conclusions: This NMA suggested greater OS and PFS benefits with first-line atezo + bev treatment vs other therapies approved for treatment of unresectable HCC.