Background & Aims: Risk scores have been designed to predict the development of hepatocellular carcinoma (HCC) in treatment-naive patients with chronic hepatitis B (CHB).However, little is known about their predictive accuracy in HBeAg-negative patients in the grey zone (GZ).We aimed to develop a HBcrAg-based HCC risk score and explore whether it outperforms other risk scores in GZ patients.
Methods: Two retrospective cohorts of HBeAg-negative patients with American Association for the Study of Liver Diseases-defined GZ were established for derivation and validation (Taiwanese, N = 911; Japanese, N = 806).All of them were non-cirrhotic at baseline and remained treatment-naive during the follow-up.The primary endpoint was HCC development.
Results: In a median follow-up period of 15.5 years, 85 patients developed HCC in the derivation cohort.We found that age, sex, 5 piece Storage Set alanine aminotransferase, platelet count, and HBcrAg, but not HBV DNA levels, were independent predictors and a 20-point GZ-HCC score was developed accordingly.
The 10-year and 15-year area under the ROC curve (AUROC) ranged from 0.83 to 0.86, which outperformed the HBV DNA-based HCC risk scores, including REACH-B and GAG-HCC scores (AUROC ranging from 0.
66 to 0.74).The better performance was also validated in EASL- and Asian Pacific Association for the Study of the Liver-defined GZ patients.
These findings remained consistent in the validation cohort.Finally, the low-risk and high-risk GZ patients (stratified by a score of 8) had an HCC risk close to inactive CHB and immune-active CHB patients, respectively, in both cohorts.Conclusions: The HBcrAg-based GZ-HCC reset switch score predicts HCC better than other HBV DNA-based risk scores in GZ patients who are HBeAg-negative patients, which may help optimise their clinical management.
Impact and implications: We have developed a risk score based on HBcrAg, which has shown better predictive ability for HCC compared with other risk scores based on HBV DNA.Using a score of 8, GZ patients can be classified into low- and high-risk groups, which can guide follow up and early treatment, respectively.This validated risk score is a valuable tool for optimising the management of GZ patients who are HBeAg-negative.