Using ALBI score at the start of sorafenib treatment to predict regorafenib treatment candidates in patients with hepatocellular carcinoma

Background: Although sorafenib-regorafenib sequential therapy improves the prognosis of patients with hepatocellular carcinoma (HCC), many patients abandon sequential therapy due to worsening hepatic reserve function. Thus, it is important to clarify which patients can be treated using regorafenib. The albumin-bilirubin score is a good biomarker for hepatic reserve function. The aim of this study was to determine whether patient albumin-bilirubin scores at the start of sorafenib treatment could be used to identify candidates for subsequent regorafenib therapy. Methods: This is a retrospective cohort study. From 2009 to 2017, 267 hepatocellular carcinoma patients treated with sorafenib were enrolled. After sorafenib therapy, 138 progressive disease patients were analyzed. The patients were divided in two groups: (i) regorafenib candidate group (Child-Pugh class A, Eastern Cooperative Oncology Group Performance Status <= 1, and maintained sorafenib tolerance); and (ii) regorafenib non-candidate group. The primary endpoint was the albumin-bilirubin score. We assessed retrospectively whether albumin-bilirubin scores were useful for predicting regorafenib treatment regimen candidacy. Results: For the 138 analyzed patients, the median overall survival duration was 15.6 months in the regorafenib candidate group and 6.8 months in the regorafenib non-candidate group (P < 0.01). Using univariate analysis, etiology, aspartate aminotransferase >= 40 IU/L, prothrombin time >= 85% and albumin-bilirubin score <-2.53 at the start of sorafenib treatment were identified as predictors. Using multivariate analysis, albumin-bilirubin score <-2.53 was the only significant predictor. Conclusions: Based on the multivariate analysis results, albumin-bilirubin score at the start of sorafenib therapy is a useful marker for identifying candidate patients for starting regorafenib therapy.