Immune checkpoint inhibitors (ICIs) combined with other agents have emerged as the standard first-line treatment for metastatic hepatocellular carcinoma (HCC), replacing tyrosine kinase inhibitors (TKIs). However, the comparative efficacy and safety of different ICI-based combinations remains unclear. To evaluate the efficacy and safety of ICI-based combinations versus TKIs across different regimens through a systematic review and network meta-analysis (NMA) of phase III randomized controlled trials (RCTs). A comprehensive literature search was conducted across major databases and conference proceedings between 2019 and 2024. Eligible studies included phase III RCTs that evaluated ICI combinations in the first-line setting for metastatic HCC. Pairwise meta-analysis and Bayesian NMA were performed to assess overall survival (OS), progression-free survival (PFS), overall response rate (ORR), treatment-related adverse events (TRAEs), grade 3–4 TRAEs, and therapy discontinuation owing to toxicity. Six RCTs comprising 3937 patients were included in the study. Compared with TKIs, ICI-based combinations improved OS (OR, 0.72; 95% CI: 0.58–0.91) and ORR (OR: 3.13; 95% CI: 2.07–4.47), without significantly increasing TRAEs. No significant benefit was observed in PFS (OR, 0.81; 95% CI: 0.56–1.19). The NMA rankings suggested camrelizumab plus rivaroceranib (CAM+ RIVO), nivolumab plus ipilimumab (NIVO + IPI), and durvalumab plus remelimumab (DURVA + TREME) as the most effective regimens for OS, PFS, and ORR, respectively. DURVA + TREME appeared to have the best safety profile, and CAM + RIVO was associated with higher rates of treatment discontinuation due to toxicity. ICI-based combinations are more effective than TKIs in improving the OS and ORR in patients with metastatic HCC, with an acceptable safety profile. CAM + RIVO, NIVO + IPI, and DURVA + TREME have emerged as promising first-line options, although direct comparisons in future trials are warranted to confirm these findings.
Moura et al. (2026) studied this question.