Validation of the albumin‐bilirubin grade‐based integrated model as a predictor for sorafenib‐failed hepatocellular carcinoma

BACKGROUND & AIMS: Sorafenib is the standard treatment for advanced hepatocellular carcinoma (HCC) but is challenging after treatment failure. Appropriate criteria for enrolling patients into second-line trials are still limited. In this study, we aimed to establish more objective criteria based on Albumin-Bilirubin (ALBI) grade to select patients with better post-progression survival (PPS) for second-line treatment. METHODS: Consecutive 404 advanced HCC patients receiving sorafenib were retrospectively enrolled. All patients were in Child-Pugh class A and BCLC stage C with either portal vein invasion or extrahepatic metastasis at the beginning of sorafenib treatment. Radiological evaluation based on mRECIST criteria and clinical assessments with compliance were performed on schedule. RESULTS: During the median follow-up period of 5.8 months, 310 patients developed progressive disease (PD) and 350 deaths occurred. The PD patients were randomized into derivation and validation cohorts by a 1:1 ratio. The independent predictors of poor PPS in derivation cohort were ALBI grade 3 at PD (hazard ratio [HR]=3.24, P = .002), new extrahepatic lesions (NEH) (HR=1.75, P = .011), and early PD within 4 months (HR=1.88, P = .037). ALBI-PD criteria were proposed by incorporating these three risk factors. In the validation cohort, PPS could be independently predicted by presence of early PD, NEH as well as ALBI grade 3 at PD. Patients within ALBI-PD criteria had significant longer median PPS than those beyond it even in Child-Pugh A (9.7 vs 4.9 months, P = .005) subpopulations. CONCLUSIONS: The ALBI-PD criteria can differentiate PPS and stratify the patients with advanced HCC for the second-line trials or salvage therapy.