Modelling the health and economic impacts of M72/AS01E vaccination and BCG-revaccination: Estimates for South Africa

Background: Tuberculosis remains a major public health problem in South Africa, with an estimated 300,000 cases and 55,000 deaths in 2021. New tuberculosis vaccines could play an important role in reducing this burden. Phase IIb trials have suggested efficacy of the M72/AS01E vaccine candidate and BCG-revaccination. The potential population impact of these vaccines is unknown. Methods: We used an age-stratified transmission model of tuberculosis, calibrated to epidemiological data from South Africa, to estimate the potential health and economic impact of M72/AS01E vaccination and BCGrevaccination. We simulated M72/AS01E vaccination scenarios over the period 2030-2050 and BCGrevaccination scenarios over the period 2025-2050. We explored a range of product characteristics and delivery strategies. We calculated reductions in tuberculosis cases and deaths and costs and cost-effectiveness from health-system and societal perspectives. Findings: M72/AS01E vaccination may have a larger impact than BCG-revaccination, averting approximately 80% more cases and deaths by 2050. Both vaccines were found to be cost-effective or cost saving (compared to no new vaccine) across a range of vaccine characteristics and delivery strategies from both the health system and societal perspective. The impact of M72/AS01E is dependent on the assumed efficacy of the vaccine in uninfected individuals. Extending BCG-revaccination to HIV-infected individuals on ART increased health impact by approximately 15%, but increased health system costs by approximately 70%. Interpretation: Our results show that M72/AS01E vaccination or BCG-revaccination could be cost-effective in South Africa. However, there is considerable uncertainty in the estimated impact and costs due to uncertainty in vaccine characteristics and the choice of delivery strategy. Funding: This work was funded by the Bill & Melinda Gates Foundation (INV-001754). This work used the Cirrus UK National Tier-2 HPC Service at EPCC (https://www.cirrus.ac.uk) funded by the University of Edinburgh and EPSRC (EP/P020267/1).