Stakeholder engagement and mathematical modelling to inform chikungunya vaccination programmes in endemic and epidemic settings

Introduction: S Despite vaccine availability and potential expansion to younger age groups, evidence gaps hinder effective vaccine implementation in endemic and epidemic settings. My PhD was conceptualised in alignment with the research needs of stakeholders, particularly the Global Chikungunya Vaccine Clinical Development Program (GCCDP) at International Vaccine Institute (IVI). While several chikungunya vaccine candidates were in the pipeline for introduction before starting my PhD, critical evidence gaps remained. These include: (1) unknown global burden of disease complicating disease prioritisation, (2) limited identification of high-risk areas and underlying immunity due to insufficient surveillance and diagnostic capacities, and (3) uncertainties regarding optimal age groups for outbreak response immunisation. Prior studies have estimated global chikungunya burden or evaluated vaccine impact but lack highspatial resolution predictions based on seroprevalence data and evaluation of age-specific vaccination strategies. The overall aim of my thesis is to engage global chikungunya stakeholders and conduct infectious disease modelling to support vaccine introduction in chikungunya affected countries by estimating chikungunya transmission intensities using synthesised age-stratified seroprevalence data, global health burden, and the potential impact of outbreak response immunisation strategies to inform vaccine use-case scenarios. My thesis address four research objectives: 1. Engage stakeholders at the global, regional, and national levels to assess evidence gaps to inform chikungunya vaccine introduction and identify research priorities that can be addressed through mathematical modelling. 2. Estimate comparable long-term average annual chikungunya force of infection and pooled prevalence of chronic disease and hospitalisation among confirmed chikungunya cases. 3. Predict global chikungunya force of infection at high spatial resolution (5x5 km) and estimate chikungunya burden at global, regional, and national levels. 4. Estimate the impact of age-specific chikungunya outbreak response immunisation strategies using two licensed vaccines under different vaccine protection mechanisms. Methods: To identify evidence gaps to inform in chikungunya vaccine introduction, interviews were conducted with global, national, and sub-national stakeholders from Latin America, Asia, and Africa. This was complemented by a review of the chikungunya vaccine pipeline to assess the challenges in the path to licensure, access, and uptake of chikungunya vaccines and a stakeholder meeting to discuss chikungunya disease burden, vaccine deployment strategies, regulatory challenges, and access issues. Chikungunya transmission patterns vary across settings. To enable comparison across these heterogenous settings, I estimated long-term average annual force of infection (FOI), defined as the average annual per capita risk of infection among susceptible individuals. I conducted a systematic literature review to collect age-stratified seroprevalence data. I fitted both constant and time-varying catalytic models to all survey locations and classified study locations as endemic or epidemic. To propagate uncertainty in inter-epidemic periods inferred from the time-varying models, I performed 1,000 simulations over a 100-year period, converting average inter-epidemic intervals into annual outbreak probabilities and simulating epidemic events using binomial sampling. To estimate proportion of chronic disease and hospitalisation among confirmed chikungunya cases, I conducted a meta-analysis using a random-effects model Using the estimated long-term average annual FOIs as ground-truth data, I predicted FOIs in globally affected and at-risk regions using a random forest model. I included climatic, socioeconomic, and ecological covariates highly correlated with chikungunya transmission to extrapolate FOIs. I estimated the global, regional, and national chikungunya burden under two scenarios: (1) focal scenario (lower-bound), limiting burden estimates to locations with reported transmission between 1927 to 2024, and (2) at-risk scenario (upper-bound), based on all environmentally suitable areas. To evaluate age-specific chikungunya outbreak response immunisation strategies, I constructed an age-structured transmission dynamic model calibrated with state-level Brazilian surveillance data from 2022. I assessed immunisation strategies targeting ages 1–11, 12–17, 18–59, and ≥60 years for both Ixchiq® and Vimkunya®. I calibrated reporting rates using predicted annual symptomatic burden and reported cases from Brazil to estimate epidemic size. I estimated vaccine impact by symptomatic cases, deaths, disability-adjusted life years (DALYs) averted, and the number needed to vaccinate (NNV), under disease-blocking only, and disease and infection blocking vaccine protection mechanism. (DALYs) averted, and the number needed to vaccinate (NNV), under disease-blocking only, and disease and infection blocking vaccine protection mechanism Results: Based on the stakeholder survey, review of chikungunya vaccine landscape, and stakeholder meeting, I identified three research questions (corresponding to objectives 3-5) that could be addressed by mathematical modelling in this thesis. I identified seroprevalence surveys from 76 locations in 38 countries through systematic literature review and classified 17 (22%) locations as endemic and 59 (78%) as epidemic settings. The global long-term average annual FOI was 0.007 (95% uncertainty interval [UI] 0.003–0.010) and varied from 0.0001 (0.00004–0.0002) to 0.113 (0.07–0.20). I estimated that 51% (95% confidence interval [CI] CI 45–58) of people with laboratory-confirmed symptomatic chikungunya had chronic disability after infection and 4% (3–5) were admitted to hospital following infection. The extrapolated global long-term average annual FOI was predicted at 0.012 (0.007, 0.019) for focal scenario and 0.013 (0.005, 0.03) for at-risk scenario in 103 countries. I estimated global chikungunya burden annual at 14.4 (11.0–17.8) million infections and 0.96 (0.56–1.6) million DALYs in the focal scenario, and 34.9 (26.7–43.1) million infections and 2.3 (1.4–3.8) million DALYs in the at-risk scenario. Ixchiq® and Vimkunya® had similar vaccine impacts. Across strategies, vaccinating children 1- 11 years yielded the lowest NNV for both vaccines, whereas vaccinating adults 18-59 years achieved the greatest absolute reduction in symptomatic cases, averting 17.2% (95% UI: 14.6– 29.9) of total symptomatic cases with Vimkunya® and 18.1% (15.6–30.8) with Ixchiq®, under disease and infection blocking mechanism. Vaccinating adults 18-59 years with Ixchiq® or Vimkunya® yielded similar efficiency, with NNVs to avert a DALY of 339 (39–3,412) and 361 (40–3,777) respectively, under disease and infection-blocking mechanism. Implications: Based on the stakeholder needs and evidence gaps that could be addressed through mathematical modelling, I estimated chikungunya force of infection, burden, and potential impact of outbreak response immunisation. These findings address key evidence gaps to inform chikungunya vaccine policy and programmatic decision-making. I inferred subnational heterogeneity in long-term average annual FOI and transmission dynamics and identified both endemic and epidemic settings across different countries. Brazil, Ethiopia, Malaysia, and India included both endemic and epidemic settings. Long-term average annual FOI was higher in epidemic settings than endemic settings. However, longterm cumulative incidence of chikungunya can be similar between large outbreaksin epidemic settings with a high FOI and endemic settings with a relatively low FOI. Median global FOI estimates were similar between focal and at-risk scenarios; however, the at-risk scenario yielded higher upper-bound FOI estimates and DALY burdens due to its broader assumed population at risk. While chikungunya transmission has high geographical uncertainty, high FOI is not limited to tropical regions and distributed across all continents. The estimates of chikungunya burden are useful for prioritisation of regions and target agegroups for chikungunya vaccine introduction. Under current licensure, vaccinating adolescents aged 12-17 years first, followed by 18-59 years are efficient strategies, with similar NNVs for both Ixchiq and Vimkunya. If eligibility expands to younger populations, vaccinating 1–11-year age group will have relatively higher efficiency.