Regorafenib dosing in patients with advanced hepatocellular carcinoma (HCC) - a retrospective real-world experience

Background The multityrosine-kinase-inhibitor regorafenib is approved for the treatment of hepatocellular carcinoma (HCC) and colorectal cancer (CRC). A recent trial in patients with CRC showed that starting with lower dose than recommended led to better drug tolerability and outcome. Here, we evaluated the association between regorafenib starting dose, side effects and outcome in a retrospective real-world cohort of HCC patients.

Methods Patients with HCC treated at the Medical University of Vienna between 04/2015 and 02/2020 were studied.

Results Of 29 patients (27 male, 93 %) included, most patients received regorafenib as second line treatment (n = 22, 76 %) and all were pre-treated with sorafenib. Only in 9 patients (31 %) regorafenib was started with the full dose (160mg/d), resulting in a mean starting dose of 105 ± 39mg. Adverse events (AEs) were reported in 22 patients (76 %), which were severe (grade ≥ 3) in 6 patients (21 %). AEs included pain (n = 10, 34 %), hand-foot-skin reaction (n = 9, 31 %), fatigue (n = 8, 28 %), and diarrhoea (n = 8, 28 %). Interestingly, proportion of patients with AE was comparable between patients who were started with full vs. reduced dose (67 % vs. 70 %;p = 0.642). Dose reductions were required in 44 % started full dose, while dose could be escalated in 45 % of patients started lower. Final dose was 124 ± 42mg (full starting dose) and 116 ± 41mg (lower starting dose), respectively (p = 0.614). Median time on regorafenib was numerically longer in the lower starting dose group (3.1 (95 %CI: 1.9-4.4) vs. 4.3 (95 %CI: 2.6-6.0) months; p = 0.522). Furthermore, disease control rate (DCR = proportion of patients achieving stable disease, partial or complete response) was double in patients started on reduced dose (25 % vs. 53 %;p = 0.582).

Conclusions Starting with reduced regorafenib dose was associated with longer time on treatment and better disease control. The lack of statistical significance may be explained by low patient numbers. This finding needs further evaluation in a larger cohort.

Publication History

Article published online:
26 May 2020