Treatment of hepatocellular carcinoma with immune checkpoint inhibitors and applicability of first-line atezolizumab/bevacizumab in a real-life setting

Immune checkpoint inhibitors (ICIs) are the new frontier for the treatment of advanced hepatocellular carcinoma (HCC). Since the first trial with tremelimumab, a cytotoxic T-lymphocyte-associated protein 4 inhibitor, increasing evidence has confirmed that these drugs can significantly extend the survival of patients with advanced hepatocellular carcinoma (HCC). As a matter of fact, the overall survival and objective response rates reported in patients with advanced HCC treated with ICIs are the highest ever reported in the second-line setting and, most recently, the combination of the anti-programmed death ligand protein-1 atezolizumab with bevacizumab—an anti-vascular endothelial growth factor monoclonal antibody—demonstrated superiority to sorafenib in a Phase III randomized clinical trial. Therefore, this regimen has been approved in several countries as first-line treatment for advanced HCC and is soon expected to be widely used in clinical practice. However, despite the promising results of trials exploring ICIs alone or in combination with other agents, there are still some critical issues to deal with to optimize the prognosis of advanced HCC patients. For instance, the actual proportion of patients who are deemed eligible for ICIs in the real-life ranges from 10% to 20% in the first-line setting, and is even lower in the second-line scenario. Moreover, long-term data regarding the safety of ICIs in the population of patients with cirrhosis and impaired liver function are lacking. Lastly, no biomarkers have been identified to predict response, and thus to help clinicians to individually tailor treatment. This review aimed to summarize the state of the art immunotherapy in HCC and, by analyzing a large, multicenter cohort of Italian patients with HCC, to assess the potential applicability of the combination of atezolizumab/bevacizumab in the real-life setting.