First-line therapies for unresectable hepatocellular carcinoma: a network meta-analysis of immune checkpoint inhibitors and transarterial therapies based on 35 randomized trials.

Objective: Unresectable hepatocellular carcinoma (uHCC) is a major global health challenge. Current first-line strategies involve immune checkpoint inhibitors (ICIs), transarterial therapies, and anti-VEGF agents. However, comprehensive comparative efficacy and safety, including certainty assessments, for diverse combination interventions are lacking. This study systematically compared 30 first-line uHCC interventions using network meta-analysis, integrating evidence certainty, to inform clinical decisions and refine guideline recommendations. Methods: Following a preregistered protocol (PROSPERO: CRD420261303389), a systematic search of major databases (until Jan 10, 2026) identified randomized controlled trials (RCTs) on first-line uHCC monotherapies or combinations (ICIs, transarterial therapies). Primary outcomes were overall survival (OS) and progression-free survival (PFS); secondary outcomes included objective response rate (ORR) and grade ≥3 adverse events (AEs). Screening, data extraction, and risk of bias (Cochrane RoB 2.0) were conducted by two reviewers. A Bayesian random-effects network meta-analysis, with certainty of evidence assessed by CINeMA and GRADE, synthesized direct and indirect evidence, reporting HRs/ORs (95% CrIs) and ranking interventions (SUCRA). Results: Thirty-five RCTs (13,595 participants, 30 interventions) were analyzed. Most studies had low bias risk; overall evidence certainty was moderate. HAIC-Sora significantly improved OS (HR 0.57; 95% CrI 0.36-0.89; moderate certainty) and PFS (HR 0.46; 95% CrI 0.28-0.75; moderate certainty) versus sorafenib. HAIC notably improved ORR (OR 29.94; 95% CrI 3.38-265.11, high certainty). TACE-Lenv significantly reduced grade ≥3 AEs (OR 0.23; 95% CrI 0.06-0.87; moderate certainty) versus sorafenib. Conclusion: For uHCC patients, HAIC-Sora demonstrated superior OS and PFS with an acceptable safety profile, representing the only comparison achieving statistical significance against sorafenib monotherapy for survival outcomes. While SUCRA rankings provided a relative hierarchy of other treatments, their interpretation should be approached with caution due to predominantly indirect evidence and lack of statistical significance in most pairwise comparisons. Limitations include blinding challenges and a reliance on indirect evidence for many comparisons, contributing to moderate evidence certainty. High-quality future studies, especially head-to-head comparisons, are crucial to solidify findings and guide clinical practice. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/, identifier CRD420261303389.