Sequencing in the post-immunotherapy era: Does second-line tyrosine kinase inhibitor choice impact survival and safety in advanced HCC?

Abstract

4140 Background: IMbrave150 trial established atezolizumab/bevacizumab as first line (1L) treatment for advanced hepatocellular carcinoma (HCC) in 2020. Current guidelines for second line (2L) therapy in advanced HCC rely on pre-immunotherapy (IO) trials (Lenvatinib, REFLECT trial and Cabozantinib, CELESTIAL trial). We investigated whether Lenvatinib (potent FGFR 1–4 inhibitor) offers superior efficacy over Cabozantinib (MET/AXL inhibitor with minimal FGFR activity) in post-IO setting by targeting the FGF-driven angiogenic escape pathways. Methods: Using TriNetX global network, we identified advanced HCC patients who received 1L atezolizumab/bevacizumab followed by 2L Lenvatinib (n=436) or Cabozantinib (n=115) within 6 months. 1:1 propensity score matching (PSM) balanced demographics, liver function, viral hepatitis, alcohol-related disorders, chronic liver diseases and key comorbidities. Outcomes include 1-year overall survival (OS), hepatic decompensation, cardiovascular (CV) outcomes and 90-day healthcare utilization, GI toxicity and acute renal failure (ARF). Each cohort was analyzed for 1 year competing risk of transition to third line (3L) systemic therapy vs. death vs. hospice. Results: Baseline characteristics were balanced except for lower pre-existing hypertension in Lenvatinib cohort (64% vs 74%; p=0.02). Post matching (n=110/arm), Lenvatinib showed significantly higher 1-year overall survival (OS). Safety analysis revealed significantly higher rates of new onset hepatic decompensation and a trend toward increased acute healthcare utilization with Cabozantinib. Renal, GI, and CV outcomes were comparable between cohorts. Lenvatinib arm had lower competing risk of mortality and higher cumulative incidence of receiving 3L therapy. Conclusions: In the real-world post-IO setting, Lenvatinib conferred significant 1-year survival benefit with superior hepatic safety profile compared to Cabozantinib. Cabozantinib exhibited higher rates of new hepatic decompensation and a trend toward increased healthcare resource utilization. Beyond a favorable pharmacoeconomic profile, results suggest superior sustained clinical benefit while preserving eligibility for 3L therapy. Given sample size constraints, these hypothesis-generating signals warrant validation in larger, multi-institutional cohorts. Key clinical outcomes. Outcome Lenvatinib Cabozantinib HR/ RR (95% CI) p-value Median OS (months) 10.2 8.3 1 year OS 47.1% 31.9% 0.65 (0.45, 0.95) 0.026 Hepatic decompensation 22.1% 37.5% 0.59 (0.34, 1.01) 0.046 CV outcomes 18.2% 16.4% 1.11 (0.62, 1.98) 0.72 90-day outcomes Healthcare utilization 37.3% 50.0% 0.75 (0.55, 1.01) 0.057 ARF 12.7% 13.6% 0.93 (0.47, 1.84) 0.84 GI events 25.5% 23.6% 1.08 (0.68, 1.71) 0.75 1 year Competing risks Death 36.1% 48.7% 3L therapy 30.7% 27.9% Probability of being event-free 32.6% 23.4%