Background/Aims: Finite nucleos(t)ide analog (NA) therapy has been proposed as an alternative treatment strategy for chronic hepatitis B (CHB), but biomarkers for post-treatment monitoring are limited. We investigated whether measuring hepatitis B core-related antigen (HBcrAg) after NA cessation may stratify the risk of subsequent clinical relapse (CR). Methods: This retrospective multicenter analysis enrolled adults with CHB who were prospectively monitored after discontinuing entecavir or tenofovir with negative HBeAg and undetectable HBV DNA at the end of treatment (EOT). Patients with cirrhosis or malignancy were excluded. CR was defined as serum alanine aminotransferase > two times the upper limit of normal with recurrent viremia. We applied time-dependent Cox proportional hazard models to clarify the association between HBcrAg levels and subsequent CR. Results: The cohort included 203 patients (median age, 49.8 years; 76.8% male; 60.6% entecavir) who had been treated for a median of 36.9 months (interquartile range [IQR], 36.5–40.1). During a median post-treatment follow-up of 31.7 months (IQR, 16.7–67.1), CR occurred in 104 patients with a 5-year cumulative incidence of 54.8% (95% confidence interval [CI], 47.1–62.4%). Time-varying HBcrAg level was a significant risk factor for subsequent CR (adjusted hazard ratio [aHR], 1.53 per log U/mL; 95% CI, 1.12–2.08) with adjustment for EOT HBsAg, EOT anti-HBe, EOT HBcrAg and time-varying HBsAg. During follow-up, HBcrAg <1,000 U/mL predicted a lower risk of CR (aHR, 0.41; 95% CI, 0.21–0.81). Conclusions Dynamic measurement of HBcrAg after NA cessation is predictive of subsequent CR and may be useful to guide post-treatment monitoring.

Background/Aims: Despite the high efficacy of direct-acting antivirals (DAAs), approximately 1–3% of hepatitis C virus (HCV) patients fail to achieve a sustained virological response. We conducted a nationwide study to investigate risk factors associated with DAA treatment failure. Machine-learning algorithms have been applied to discriminate subjects who may fail to respond to DAA therapy. Methods: We analyzed the Taiwan HCV Registry Program database to explore predictors of DAA failure in HCV patients. Fifty-five host and virological features were assessed using multivariate logistic regression, decision tree, random forest, eXtreme Gradient Boosting (XGBoost), and artificial neural network. The primary outcome was undetectable HCV RNA at 12 weeks after the end of treatment. Results: The training (n=23,955) and validation (n=10,346) datasets had similar baseline demographics, with an overall DAA failure rate of 1.6% (n=538). Multivariate logistic regression analysis revealed that liver cirrhosis, hepatocellular carcinoma, poor DAA adherence, and higher hemoglobin A1c were significantly associated with virological failure. XGBoost outperformed the other algorithms and logistic regression models, with an area under the receiver operating characteristic curve of 1.000 in the training dataset and 0.803 in the validation dataset. The top five predictors of treatment failure were HCV RNA, body mass index, α-fetoprotein, platelets, and FIB-4 index. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the XGBoost model (cutoff value=0.5) were 99.5%, 69.7%, 99.9%, 97.4%, and 99.5%, respectively, for the entire dataset. Conclusions Machine learning algorithms effectively provide risk stratification for DAA failure and additional information on the factors associated with DAA failure.