LAUSR

Nucleos(t)ide analogs (NAs) cannot completely suppress the risk of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B (CHB). This study aimed to identify the risk factors for HCC development in naïve CHB patients treated with current NA. Patients receiving NA (n = 905) were recruited retrospectively from the 17 hospitals of the Japanese Red Cross Liver Study Group. All treatment‐naïve patients had been receiving current NA continuously for more than 1 year until the end of the follow‐up. We analyzed the accuracy of predictive risk score using the area under receiver operating characteristic curve. The albumin–bilirubin (ALBI) score was significantly improved by NA therapy (−0.171 ± 0.396; p < 0.001 at Week 48). A total of 72 (8.0%) patients developed HCC over a median follow‐up of 6.2 (1.03–15.7) years. An independent predictive factor of HCC development was older age, cirrhosis, lower platelet counts at baseline and ALBI score, and alpha‐fetoprotein (AFP) at 1 year after NA therapy according to multivariate analysis. The accuracy was assessed using the PAGE‐B, mPAGE‐B, aMAP, APA‐B, and REAL‐B scores that included these factors. Discrimination was generally acceptable for these models. aMAP and REAL‐B demonstrated high discrimination with 0.866/0.862 and 0.833/0.859 for 3‐ and 5‐year prediction from the status of 1 year after NA therapy, respectively. Baseline age and platelet count, as well as ALBI and AFP one year after NA, were useful for stratifying carcinogenesis risk. The aMAP and REAL‐B scores were validated with high accuracy in Japanese CHB patients.