The efficacy and safety of lenvatinib in patients who did not meet the inclusion criteria of the phase 3 trial (REFLECT trial) in real-world practice in Japan: A nationwide multicenter study in Japan.

e15629 Background: Lenvatinib (LEN) has been approved as a single agent for patients with unresectable hepatocellular carcinoma (u-HCC) since Mar 2018 in Japan. We conducted a nationwide multicenter study and especially focused on the efficacy and safety in the patients who did not meet the inclusion criteria of the phase 3 trial (REFLECT trial). Methods: A total of 202 patients received LEN from March 2018 at 21 sites in Japan was registered. Tumour assessments in accordance with modified RECIST were done using dynamic CT or MRI within 4-8 weeks and every 6-8 weeks thereafter. Adverse events (AEs) were graded according to the CTCAE ver4.0. Results: In this study, 128 of 202 (63%) patients did not meet the inclusion criteria of the REFLECT trial. For the reasons, 69 patients received TKI therapy before LEN, 37 patients were Child-Pugh B, 6 patients showed main portal vein invasion,13 patients had HCC which occupied ≥ 50% of liver, and 22 patients were with lower platelet ( < 75 × 10⁹ per L) . Median observation time was 4.1 months in all patients, 22 patients died, and median overall survival (OS) was not reached. As the 2nd-line therapy (after sorafenib), 45 patients received LEN and 24 patients did as the 3rd-line (after regorafenib). The OS between TKI naïve and experienced patients was not significantly different, even though baseline albumin level and ALBI score were significantly lower in TKI experienced patients than TKI naïve (p = 0.001 and 0.005). Objective response rate (ORR) and disease control rate (DCR) of TKI naïve patients were 37% and 84%, while those of TKI experienced were 27% and 76%. Child-Pugh B patients (n = 36) had significantly shorter OS than Child-Pugh A (p = 0.01) and median OS in Child-Pugh B patients was 6.1 months. Drug discontinuation due to AEs and elevated aspartate aminotransferase were significantly higher in Child-Pugh B patients (p = 0.005 and 0.001). The OS, treatment duration and drug discontinuation were not significantly different between patients with or without lower platelets. ORR and DCR in patients with lower platelets (n = 22) were 18% and 63%. Conclusions: The response rate of LEN in patients who experienced TKI therapies was similar to that in TKI naïve patients. The median OS in Child-Pugh B patients was 6.1 months and elevated aspartate aminotransferase during LEN was frequently reported in such patients. The OS and treatment duration were not significantly different between patients with or without lower platelets at the cut off of < 75 × 10⁹ per L.