Clinical Trials for Curing Chronic Hepatitis B: A Comprehensive Analysis of Trial Design Considerations and Their Association With Trial Performance.

BACKGROUND AND AIM: The development of novel agents with curative intent for chronic hepatitis B (CHB) has accelerated, but few candidates transition to phase III trials. We evaluated trial design features associated with registry-defined trial completion and phase transition, and synthesized end-of-treatment hepatitis B surface antigen (HBsAg) decline as an on-treatment antiviral signal.

METHODS

We systematically reviewed clinical trial registries and bibliographic databases for CHB cure trials. Associations between trial design characteristics and trial performance were analyzed using Cox regression (time to registry-defined completion), logistic regression (phase I to phase II transition), and random-effects meta-analysis of end-of-treatment HBsAg decline.

RESULTS

Larger sample size (HR = 0.819, 95% CI = 0.710-0.945) and longer treatment duration (HR = 0.897, 95% CI = 0.816-0.987) were associated with a lower likelihood of registry-defined completion in phase I trials. In phase II trials, larger sample size was associated with a lower likelihood of registry-defined completion (HR = 0.936, 95% CI = 0.879-0.997). Novel trial designs were not associated with faster completion. Among phase I trials with a definitive status, trials enrolling healthy volunteers were more often followed by phase II registration (adjusted OR = 2.82, 95% CI = 1.16-7.08). Meta-analysis showed that direct-acting antivirals were associated with marked end-of-treatment HBsAg decline (SMD = 1.28, 95% CI = 0.47-2.08).

CONCLUSIONS

Smaller and shorter early-phase trials were more likely to complete, whereas novel trial designs were not associated with faster completion. Direct-acting antivirals showed marked end-of-treatment HBsAg decline, an on-treatment signal rather than durable functional cure.