Background and Aims: The multikinase inhibitor cabozantinib has been approved for hepatocellular carcinoma (HCC) previously treated with sorafenib. We report safety and efficacy data of an international, multicenter, real-life cohort of patients with advanced HCC treated with cabozantinib. Methods: Patients with HCC who were treated with cabozantinib were retrospectively identified across 11 centers in Austria, Switzerland, and Germany. Patients’ characteristics, adverse events, duration of treatment and overall survival (OS) data were analyzed until April 1, 2020. Results: Eighty-eight patients from 11 centers were included. The predominant underlying liver diseases were NAFLD/NASH in 26 (30%) and hepatitis C infection in 21 (24%) patients. Seventy-eight patients (89%) were classified as Barcelona clinic liver cancer (BCLC) stage C. Sixty patients (68%) were Child-Pugh A, whereas 22 (25%) were Child-Pugh B, respectively. Cabozantinib was used as systemic second- and third-line or later treatment in 41 (47%) and 46 (52%) patients, respectively. The following best responses under cabozantinib were documented: partial response in 6 (7%), stable disease in 28 (32%), and progressive disease in 28 (32%) patients, respectively. Fifty-two patients (59%) died during follow-up. The median OS from start of cabozantinib treatment was 7.0 months in the entire cohort and 9.7 months in Child-Pugh A patients, while Child-Pugh B patients had a median OS of 3.4 months, respectively. Thirty-seven (42%) patients fulfilled the CELESTIAL inclusion and exclusion criteria, showing a median OS of 11.1 months. Most common adverse events were fatigue (15.6%) and diarrhea (15.6%). Conclusion: Cabozantinib treatment was effective, safe, and feasible in patients with advanced HCC in patients with compensated cirrhosis. Patients in the real-life setting had more advanced liver disease – in which 25% of patients were Child-Pugh B. However, OS in patients with Child-Pugh A cirrhosis was similar to that reported in the phase 3 trial (CELESTIAL).
