DEVELOPMENT OF A NEW COST-EFFECTIVENESS MODEL FOR CHRONIC HEPATITIS B VIRUS INFECTION BASED ON THE B-WELL TRIAL: VALIDATION OF HYBRID COHORT MODEL APPROACH
Author(s)
Anne-Claire Peultier, MSc, PhD1, Zsofia Kiss, MA, MSc2, Nathaniel Smith, PhD3, Frank van Hees, PhD4, Tiago Rua, PhD5, Lewis Michaelwaite, PhD6, Ashley Brown, MD7, Adeola Oliyide, MSc2, Alan Martin, PhD8, Dickens Theodore, PhD9, Afisi S. Ismaila, PhD10.
1GSK, Wavre, Belgium, 2GSK, London, United Kingdom, 3Maple Health Group, New York, NY, USA, 4Maple Health Group, Nijmegen, Netherlands, 5Maple Health Group, Porto, Portugal, 6Maple Health Group, Leeds, United Kingdom, 7Imperial College Healthcare NHS Trust, London, United Kingdom, 8GSK (at the time of study), London, United Kingdom, 9GSK, Durham, NC, USA, 10GSK, Collegeville, PA, USA.
1GSK, Wavre, Belgium, 2GSK, London, United Kingdom, 3Maple Health Group, New York, NY, USA, 4Maple Health Group, Nijmegen, Netherlands, 5Maple Health Group, Porto, Portugal, 6Maple Health Group, Leeds, United Kingdom, 7Imperial College Healthcare NHS Trust, London, United Kingdom, 8GSK (at the time of study), London, United Kingdom, 9GSK, Durham, NC, USA, 10GSK, Collegeville, PA, USA.
OBJECTIVES: Chronic hepatitis B virus (cHBV) infection remains a major global health burden. Nucleos(t)ide analogues (NAs) effectively suppress viral replication but rarely achieve sustained hepatitis B surface antigen (HBsAg) and HBV DNA loss, necessitating lifelong therapy. Bepirovirsen, a novel antisense oligonucleotide, is being developed as a finite treatment to reduce HBsAg and HBV DNA. We describe the modelling approach of a de novo cost-effectiveness model (CEM) comparing bepirovirsen+NAs versus NAs alone in adults with cHBV.
METHODS: A hybrid cohort-level CEM was constructed, comprising a decision tree based on B-Well trials (NCT05630807/NCT05630820) to capture short-term outcomes (including functional cure [FC] and HBsAg <100 IU/mL), linked to a state transition model simulating long-term progression to compensated cirrhosis (CC), decompensated cirrhosis, hepatocellular carcinoma (HCC), liver transplant, and death.
RESULTS: The analysis adopted a lifetime horizon from the UK National Health Service perspective with 3.5% annual discounting. FC was assumed to reduce risks of long-term outcomes, while achieving HBsAg <100 IU/mL status was linked to a higher probability of HBsAg loss informed by literature. Advanced-disease transitions were not stratified by HBeAg status, as it is not a predictor of disease progression. The model population comprised the anticipated label population of patients on stable NAs with baseline HBsAg ≤3,000 IU/mL, including an anticipated HBsAg <100 IU/mL subgroup outside B-Well eligibility. Efficacy for this subgroup was modelled conservatively by applying FC rates from the HBsAg <200 IU/mL subgroup and other short-term outcomes from the 100-1,000 IU/mL cohort. Key outcomes were life-years, quality-adjusted life-years, costs, incremental cost-effectiveness ratios, and incremental net monetary benefits. Deterministic, probabilistic, and scenario analyses explored parameter uncertainty. External validation showed modelled HCC incidence aligned with cohort data, supporting post-trial risk plausibility.
CONCLUSIONS: This robust hybrid framework provides a flexible structure to evaluate the health economic value of bepirovirsen in cHBV.
Funding: GSK (Study 307206).
METHODS: A hybrid cohort-level CEM was constructed, comprising a decision tree based on B-Well trials (NCT05630807/NCT05630820) to capture short-term outcomes (including functional cure [FC] and HBsAg <100 IU/mL), linked to a state transition model simulating long-term progression to compensated cirrhosis (CC), decompensated cirrhosis, hepatocellular carcinoma (HCC), liver transplant, and death.
RESULTS: The analysis adopted a lifetime horizon from the UK National Health Service perspective with 3.5% annual discounting. FC was assumed to reduce risks of long-term outcomes, while achieving HBsAg <100 IU/mL status was linked to a higher probability of HBsAg loss informed by literature. Advanced-disease transitions were not stratified by HBeAg status, as it is not a predictor of disease progression. The model population comprised the anticipated label population of patients on stable NAs with baseline HBsAg ≤3,000 IU/mL, including an anticipated HBsAg <100 IU/mL subgroup outside B-Well eligibility. Efficacy for this subgroup was modelled conservatively by applying FC rates from the HBsAg <200 IU/mL subgroup and other short-term outcomes from the 100-1,000 IU/mL cohort. Key outcomes were life-years, quality-adjusted life-years, costs, incremental cost-effectiveness ratios, and incremental net monetary benefits. Deterministic, probabilistic, and scenario analyses explored parameter uncertainty. External validation showed modelled HCC incidence aligned with cohort data, supporting post-trial risk plausibility.
CONCLUSIONS: This robust hybrid framework provides a flexible structure to evaluate the health economic value of bepirovirsen in cHBV.
Funding: GSK (Study 307206).
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
EE44
Topic
Economic Evaluation
Topic Subcategory
Trial-Based Economic Evaluation
Disease
Infectious Disease (non-vaccine)