Nivolumab plus ipilimumab in advanced hepatocellular carcinoma with Child-Pugh B: A multicenter retrospective study.

551 Background: The phase 3 CheckMate 9DW trial proposed that nivolumab plus ipilimumab (Nivo/Ipi) significantly improved the survival of patients with advanced hepatocellular carcinoma (aHCC) compared with sorafenib or lenvatinib as first-line treatment, but it only included patients with Child-Pugh A. Given the limited treatment options for patients with poor liver function, we aimed to assess the efficacy and safety of Nivo/Ipi in aHCC patients, stratified by liver function. Methods: This study included patients with aHCC who received Nivo/Ipi treatment at three referral hospitals between March 2020 and September 2023. Patients received nivolumab (1 mg/kg) plus ipilimumab (3 mg/kg) every 3 weeks (four doses), followed by nivolumab (240 mg) every 2 weeks. Results: Of 106 patients treated with Nivo/Ipi, 96 evaluable patients were included in the final analysis. Among them, 76 patients had Child-Pugh A, while 20 had Child-Pugh B. The baseline characteristics were generally comparable between the two groups, except for a significantly higher rate of extrahepatic metastasis in patients with Child-Pugh A (86.8% vs. 65.0%, p = 0.043) and a higher incidence of macrovascular invasion in patients with Child-Pugh B (55.0% vs. 27.6%, p = 0.032). The objective response rate (ORR) was 28.9% (22/76, 4 complete responses [CR] and 18 partial responses [PR]) in the Child-Pugh A group and 10.0% (2/20; 0 CR and 2 PR) in the Child-Pugh B group. Disease control rates were 42.1% in the Child-Pugh A group and 25.0% in the Child-Pugh B group. In a median follow-up duration of 29.3 months (95% confidence interval [CI], 28.8-29.8), the median overall survival (OS) was 8.9 months (95% CI, 5.5-17.6) in the Child-Pugh A group and 3.9 months (95% CI, 2.1-4.8) in the Child-Pugh B group (p < 0.001). The median progression-free survival (PFS) was 1.6 months (95% CI, 1.2-3.5) in the Child-Pugh A group and 1.2 months (95% CI, 0.8-2.1) in the Child-Pugh B group (p = 0.003). Adverse events of any grade occurred at a comparable rate between the two groups, although immune-related adverse events were significantly more frequent in the Child-Pugh A group. Multivariable analysis indicated that Child-Pugh B was consistently associated with a worse PFS and OS, while prior exposure to immune checkpoint inhibitors (ICIs) was associated with a poorer PFS. Notably, both responders with Child-Pugh B were both ICI naïve. Conclusions: While Nivo/Ipi treatment showed reduced efficacy in patients with Child-Pugh B, identifying potential responders beyond the current clinical trial criteria may offer opportunities to improve outcomes in this challenging population.