Atezolizumab (ATEZO) plus bevacizumab (BEVA) represents the new standard of care for the treatment of advanced hepatocellular carcinoma (HCC). However, the choice of the second-line treatment after the failure of immunotherapy-based first-line remains elusive. Taking into account the weaknesses of the available evidence, we developed a simulation model based on available phase III randomized clinical trials (RCTs) to identify optimal risk/benefit sequential strategies. Methods: A Markov model was built to estimate the overall survival (OS) of sequential first- and second-line systemic treatments. Sequences starting with first-line ATEZO plus BEVA followed by 5 second-line treatments (sorafenib [SORA], lenvatinib [LENVA], regorafenib, cabozantinib, and ramucirumab) were compared. The probability of transition between states (initial treatment, cancer progression, and death) was derived from RCTs. Life-year gained (LYG) was the main outcome. Rates of severe adverse events (SAEs) (≥ grade 3) were calculated. The incremental safety-effectiveness ratio (ISER) was calculated as the difference in probability of SAEs divided by LYG between the 2 most effective sequences. Results: ATEZO plus BEVA followed by LENVA (median OS, 24 months) or SORA (median OS, 23 months) was the most effective sequence, producing a LYG of 0.50 and 0.42 year, respectively. ATEZO plus BEVA followed by SORA was the safest sequence (SAEs 63%). At a willingness-to-risk threshold of 10% of SAEs for LYG, ATEZO plus BEVA followed by second-line SORA was favored in 72% of cases, while at a threshold of 30% of SAEs for LYG, ATEZO plus BEVA followed by second-line LENVA was favored in 69% of cases. Conclusion: Our simulation model provides a strong rationale to support ongoing trials evaluating second-line tyrosine-kinase inhibitors after first-line ATEZO plus BEVA. Future evidence from ongoing RCTs and prospective real-world studies are needed to prove the net health benefit of sequential treatment options for advanced HCC.
ASJC Scopus subject areas