Comparative efficacy of novel combination immunotherapy strategies for unresectable hepatocellular carcinoma (HCC): A network metanalysis of landmark phase III trials.

Abstract

4098 Background: After over a decade of stagnation, therapeutic options for unresectable HCC (uHCC) are expanding. In 2020, the results of IMbrave150 established the combination of atezolizumab and bevacizumab (A+B) as the novel standard of care for patients with uHCC. Parallel reporting of novel immunotherapy combinations tested in phase III trials against sorafenib limits therapeutic decision making in clinical practice, given direct comparison between novel first line treatment options does not exist. We conducted a network metanalysis (NMA) to compare A+B with other first line systemic therapies that reached their primary endpoint in phase III trials. Methods: After performing a literature review from January 2008 to February 2022, we identified 13709 studies for screening, 70 for revision, and the following 9 phase III trials for the analysis: SHARP, Asia Pacific, REFLECT, CheckMate459, IMbrave150, ORIENT32, HIMALAYA, COSMIC312 and Qin et al.2021, which tested respectively: sorafenib (Sor) vs placebo (SHARP and Asia Pacific), lenvatinib (Len) vs Sor, nivolumab (Nivo) vs Sor, A+B vs Sor, Sintilimab+IBI305 vs Sor, Durvalumab+Tremelimumab (D+T) vs Sor, atezolizumab+cabozantinib (A+C) vs Sor, and Donafenib vs Sor, as first line systemic treatments for uHCC. Hazard ratios(HR) and 95% confidence intervals(95%CI) for overall (OS) and progression free survival (PFS) were extracted for each study. A frequentist network metanalysis, with fixed effect multivariable meta-regression models to estimate the indirect pooled HRs and corresponding 95%CI, was performed. CheckMate459 was the only trial testing PD-1 monotherapy and was included as a reference despite not reaching its primary endpoint. Results: In total, 6272 patients were included in the analysis, among them, 5896 received active treatment and 376 had placebo. Amongst analyzed treatment regimens, A+B reduced the risk of death by 60% compared to placebo (HR 0.40; 95%CI 0.28-0.57), and by 42%, 37%, 36% and 32% compared to Sor (HR 0.58; 95%CI 0.43-0.79), Len (HR 0.63; 95%CI 0.45-0.89), A+C (HR 0.64; 95%CI 0.43-0.97) and Nivo (HR 0.68; 95%CI 0.48-0.98), respectively. With regards to OS, D+T was not significantly inferior to A+B (HR 0.74; 95%CI 0.52-1.06) and the efficacy of sintilimab+IBI305 was similar to A+B (HR 1.02; 95%CI 0.67-1.54). Considering PFS, A+B was significantly superior to placebo, Sor, donafenib and Nivo. Conclusions: In this network metanalysis comparing 9 landmark phase III trials in uHCC, we confirmed combination of immunotherapy with PD-1 pathway plus VEGF blockade (A+B, sintilimab+IBI305) to be associated to the highest reduction in the risk of death compared to other regimens. Within the methodological limits of this NMA, we provide evidence for the first time of comparable efficacy in terms of OS and PFS for D+T and A+B.