Second-line TKI after first-line immunotherapy-based treatment in advanced HCC: Reconstructed IPD meta-analysis.

BACKGROUND

The efficacy of second-line tyrosine kinase inhibitors (TKIs) after first-line immunotherapy-based (IO) treatment in advanced hepatocellular carcinoma (HCC) is not well-established.

METHODS

A systematic search was conducted to identify studies reporting outcomes with second-line TKIs after progression on first-line IO-based treatment. This reconstructed individual patient data (IPD) meta-analysis used survival data reconstructed from published Kaplan-Meier curves. Studies presenting or combining third-line-or-beyond data were excluded. Overall survival (OS) (primary endpoint) and progression-free survival (PFS) were analyzed using restricted mean survival time (RMST), and random-effects univariable and adjusted meta-regression analyses were performed to account for heterogeneity and potential confounding.

RESULTS

A total of 1,663 patients (16 studies) were included (sorafenib [n = 769], lenvatinib [n = 691], regorafenib [n = 105], and cabozantinib [n = 98]). Most patients received atezolizumab-bevacizumab in the first line. Cabozantinib was excluded from primary analyses as 83.7% of the data were derived from a single study, and baseline characteristics data were limited. The regorafenib group had significantly more Child-Pugh A and less macrovascular invasion. Median OS of all patients was 9.8 months (95% CI 9.4-10.2). The 12-month OS was significantly longer with lenvatinib or regorafenib compared with sorafenib (ΔRMST, 1.49 months [95% CI 1.10-1.88], p <0.001 and 1.16 months [95% CI 0.41-1.96], p = 0.002). The median PFS for all patients was 3.2 months (95% CI 3.0-3.4). The 6-month PFS with lenvatinib or regorafenib was significantly longer than with sorafenib (ΔRMST, 1.28 months [95% CI 1.08-1.49], p <0.001, and 0.73 months [95% CI 0.38-1.09], p <0.001). Meta-regression analyses suggested similar results.

CONCLUSIONS

In this reconstructed IPD meta-analysis of predominantly retrospective studies, survival outcomes with second-line TKIs after IO-based therapy were heterogeneous. Lenvatinib and regorafenib showed consistent survival compared with sorafenib; however, findings are exploratory, limited by the observational nature of the data and residual confounding. Prospective studies are needed to define optimal post-immunotherapy sequencing strategies.

IMPACT AND IMPLICATIONS: This study addresses a critical evidence gap in advanced hepatocellular carcinoma by synthesizing available real-world data on second-line TKIs after IO-based first-line treatment, using reconstructed individual patient data and restricted mean survival time to accommodate non-proportional hazards and heterogeneous follow-up. The findings should not be interpreted as a recommendation for any specific TKI. Instead, they describe current survival patterns across heterogeneous retrospective cohorts and underscore the absence of robust comparative evidence in the post-immunotherapy setting. The results are relevant for clinicians and multidisciplinary teams caring for patients who progress after immunotherapy, as they may help contextualize expectations, support shared decision-making with patients and caregivers, and highlight areas of doubt in routine practice. This work highlights the unmet need for prospective randomized trials and high-quality real-world registries to define optimal treatment sequencing, inform regulatory decisions, and ensure equitable access to evidence-based therapies.

SYSTEMATIC REVIEW REGISTRATION: This study was registered with PROSPERO (Protocol No.: CRD420251133124).