Validation Study of Scores Predicting Hepatocellular Carcinoma Risk in Chronic Hepatitis B Patients Treated With Nucleos(t)ide Analogues

Abstract

Chronic hepatitis B virus (HBV) infection is a leading cause of hepatocellular carcinoma (HCC) worldwide. Nucleos(t)ide analogues (NAs) are widely used in chronically HBV-infected patients, but the risk of HCC still remains in NA-treated patients. In this study, we aimed to validate the HCC risk scores for HBV-infected patients treated with nucleos(t)ide analogues (NAs). Among a total of 360 chronically HBV-infected patients who were treated with NAs, 253 patients without a history of HCC were used to validate the PAGE-B, mPAGE-B, PAGED-B, APA-B, and aMAP scores, as well as a recently developed score, the FAL-1 score, which consists of the FIB-4 index and ALT at 1 year of NA. In this cohort, the cumulative incidence of HCC at 5, 10, and 15 years was 2.9%, 7.8% and 11.0%, respectively. Most scores significantly stratified the HCC incidence and, for the FAL-1 score, the cumulative incidence of HCC at 10 years was 0%, 11.3% and 17.2% for the score-0 (n = 91), score-1 (n = 129) and score-2 (n = 30) groups, respectively. Compared with the other scores, the FAL-1 score was shown to efficiently identify patients at very low risk of HCC. An analysis using both this validation and the previously reported derivation cohorts demonstrated the utility in patients with either HBV genotype B or C. In conclusion, the utility of the FAL-1 score was reproduced in this validation study as well as other scores. In particular, the FAL-1 score may be useful to efficiently identify patients with a low risk of HCC.