The long-term safety, public health impact, and cost-effectiveness of routine vaccination with a recombinant, live-attenuated dengue vaccine (Dengvaxia): a model comparison study

This is the final version of the article. Available from the publisher via the DOI in this record.Background: Large Phase III trials across Asia and Latin America have recently demonstrated the efficacy of a recombinant, live-attenuated dengue vaccine (Dengvaxia) over the first 25 mo following vaccination. Subsequent data collected in the longer-term follow-up phase, however, have raised concerns about a potential increase in hospitalization risk of subsequent dengue infections, in particular among young, dengue-naïve vaccinees. We here report predictions from eight independent modelling groups on the long-term safety, public health impact, and cost-effectiveness of routine vaccination with Dengvaxia in a range of transmission settings, as characterised by seroprevalence levels among 9-y-olds (SP9). These predictions were conducted for the World Health Organization to inform their recommendations on optimal use of this vaccine. Methods and Findings: The models adopted, with small variations, a parsimonious vaccine mode of action that was able to reproduce quantitative features of the observed trial data. The adopted mode of action assumed that vaccination, similarly to natural infection, induces transient, heterologous protection and, further, establishes a long-lasting immunogenic memory, which determines disease severity of subsequent infections. The default vaccination policy considered was routine vaccination of 9-y-old children in a three-dose schedule at 80% coverage. The outcomes examined were the impact of vaccination on infections, symptomatic dengue, hospitalised dengue, deaths, and cost-effectiveness over a 30-y postvaccination period. Case definitions were chosen in accordance with the Phase III trials. All models predicted that in settings with moderate to high dengue endemicity (SP9 ≥ 50%), the default vaccination policy would reduce the burden of dengue disease for the population by 6%–25% (all simulations: –3%–34%) and in high-transmission settings (SP9 ≥ 70%) by 13%–25% (all simulations: 10%– 34%). These endemicity levels are representative of the participating sites in both Phase III trials. In contrast, in settings with low transmission intensity (SP9 ≤ 30%), the models predicted that vaccination could lead to a substantial increase in hospitalisation because of dengue. Modelling reduced vaccine coverage or the addition of catch-up campaigns showed that the impact of vaccination scaled approximately linearly with the number of people vaccinated. In assessing the optimal age of vaccination, we found that targeting older children could increase the net benefit of vaccination in settings with moderate transmission intensity (SP9 = 50%). Overall, vaccination was predicted to be potentially cost-effective in most endemic settings if priced competitively. The results are based on the assumption that the vaccine acts similarly to natural infection. This assumption is consistent with the available trial results but cannot be directly validated in the absence of additional data. Furthermore, uncertainties remain regarding the level of protection provided against disease versus infection and the rate at which vaccine-induced protection declines. Conclusions: Dengvaxia has the potential to reduce the burden of dengue disease in areas of moderate to high dengue endemicity. However, the potential risks of vaccination in areas with limited exposure to dengue as well as the local costs and benefits of routine vaccination are important considerations for the inclusion of Dengvaxia into existing immunisation programmes. These results were important inputs into WHO global policy for use of this licensed dengue vaccinSF and MJ received funding from WHO and Gavi, the Vaccine Alliance, to conduct this work. LC is a paid employee at Sanofi Pasteur. GM and JK were funded by the University of Western Australia, with computing resources provided by the Pawsey Supercomputing Centre, which is funded by the Australian Government and the Government of Western Australia. MR is funded by a Royal Society University Research Fellowship. NF, ID and DJL received research funding from the UK Medical Research Council, the UK NIHR under the Health Protection Research Unit initiative, NIGMS under the MIDAS initiative, and the Bill and Melinda Gates Foundation. IRB and DATC were funded by MIDAS Center Grant NIH/NIGMS U54-GM088491 and the Bill and Melinda Gates Foundation. DATC was also supported by NIH/NIAID R01-AI114703. TJH, IL, and CABP were funded by a Dengue Vaccine Initiative Grant to IL, NIH/NIAID R37 AI32042. THJ, IL, and KK were funded by MIDAS Center Grant NIH/NIGMS 1135 U54 GM111274. All other authors have received no specific funding to conduct this work. The funders had no role in the study design, data analyses, decision to publish or preparation of the manuscript