Hepatic flares, their immune signatures, and ALT variability after nucleos(t)ide analogue cessation in HBeAg-negative hepatitis B
JHEP Reports, 2026
Holmberg M., Niehrs A., Dalgard O., Berhe N., Desalegn H., Aleman S., Weis N., Stenstad T., Heggelund L., Samuelsen E., Karlsen L., Lindahl K., Kleppa E., Winckelmann A., Brugger-Synnes P., Simonsen H., Svendsen J., Björkström N., Reikvam D., Johannessen A.
| Disease area | Application area | Sample type | Products |
|---|---|---|---|
Hepatology | Pathophysiology | Serum |  Olink Target 96 |
Abstract
Background and Aims
Hepatic flares frequently occur after nucleos(t)ide analogue (NA) cessation in patients with chronic hepatitis B (CHB) and can be beneficial (“good flares”) or harmful (“bad flares”). We characterised flares after NA cessation aiming to identify predictors and immunological correlates of good and bad flares.
Methods
This study was nested in the prospective Nuc-Stop study, in which 127 patients with e-antigen negative CHB discontinued NA treatment with a 36-month follow-up. Flares were defined as an ALT increase >2× the upper limit of normal or >2× baseline. Predictors of flares were identified by logistic regression. In 32 patients with flares without treatment restart, we compared clinical characteristics and soluble immune marker profiles of good flares (hepatitis B surface antigen (HBsAg) loss or >1 log10 decline or sustained virological control) and bad flares (neither HBsAg decline nor virological control).
Results
Flares occurred in 58.3% of patients. Age (per 1-year increment; adjusted odds ratio (aOR), 1.07; 95% confidence interval (CI) 1.02–1.12) and end-of-treatment HBsAg level (per 1 log10 IU/mL increment; aOR, 2.09; 95% CI 1.20–3.62) independently predicted flares. Good flares displayed less ALT variability after the initial spike than bad flares (standard deviation 9.7 vs. 22.7 U/L, p=0.002). Analysis of soluble immune markers confirmed distinct clusters for good and bad flares at end-of-treatment, with higher serum levels of specific proteins in good flares (e.g. IL-13, TRAIL) and bad flares (e.g. CXCL11, OPG, TNF).
Conclusions
Flares occurred in over half the patients after NA cessation and were associated with age and end-of-treatment HBsAg levels. ALT variability following the initial flare and soluble marker profiles might serve as prognostic factors and distinguish good from bad flares.