Periodic Reporting for period 3 - ARBODYNAMIC (Coupling dynamic population immunity profiles and host behaviours to arboviral spread)
Okres sprawozdawczy: 2021-04-01 do 2022-09-30
Viruses transmitted by mosquitoes (arboviruses) continue to cause an enormous burden of public health. This project studies the mechanisms that drive arbovirus transmission, using dengue virus in Thailand as a case study. All four serotypes of dengue virus have circulated endemically in Thailand for decades. We are using historic serotype distributions and measures of antigenicity from novel cartographic methods to reconstruct the changing immune profile of the population. Further, we use detailed data on the historic genetic make-up of viruses that circulated in the country by sequencing isolates from each dengue serotype for each year over 50 years from Bangkok and over 29 years from Kamphaeng Phet (a rural region in the country). We then use measures of human behaviours (mixing patterns and mobility) obtained from mobile phone and questionnaire data as well as mosquito distributions to develop an integrative analytical framework that can bring together these different datasets and explore the consistency of different hypotheses of viral spread. Finally, we compare our inferences with that from other arboviruses that are spread by the same vector using data and isolates from recent outbreaks of chikungunya and Zika. This project allows valuable insights into long-term drivers of virus spread with key implications for public health.
The work for this project is divided in two areas: (1) identifying the role of host immunity in driving viral emergence and (2) Mechanisms that drive arbovirus transmission across spatial scales. We have made good progress in both work packages. Since the start of the project, we have published ten papers linked to the grant and have presented work from the grant at three major international conferences.
In the first work package, we have concentrated on the development and use of antigenic maps, that capture the antigenic potential of different viruses. We have used the results of these antigenic maps to explore whether we can explain patterns of hospitalizations from dengue. This work has been presented at major conferences (Epidemics, EEID). We have separately used longitudinal data from dengue cohorts in Southeast Asia to capture the dynamics of immunity over time, including in response to a dengue vaccine. We showed that vaccines elucidated a lower antibody response that natural infection, that vaccine efficacy was concentrated to the first three years after vaccination and that antibodies were an excellent correlate of protection. We have also identified a correlate of protection for chikungunya virus.
In the second work package, we have used a combination of sequences and mathematical models to capture the spread of Zika virus in Thailand. We showed that Zika has been circulating endemically in Thailand for over 20 years, highlighting the potential long term consequences from the virus. We have also developed novel modelling approaches that uses genetic data to explain relative importance of human behaviour data (how we move), local environment (e.g. mosquito density), and population immunity in driving dengue spread. We were able to show that infected people move a lot less than the surrounding susceptible population; that most transmission is highly local but the occasional long distance transmission event leads to mixing of the virus after a year. These methods are now being developed for other pathogens.
Finally, we have also been applying some of the techniques we have been developing in the ARBODYNAMIC project to help in the SARS-CoV-2 response. We worked with the French government to understand the level of immunity in the population and the patterns of spread in the country. We also used population immunity estimates to capture the underlying probability of death by age. This work has all been published.
In the first work package, we have concentrated on the development and use of antigenic maps, that capture the antigenic potential of different viruses. We have used the results of these antigenic maps to explore whether we can explain patterns of hospitalizations from dengue. This work has been presented at major conferences (Epidemics, EEID). We have separately used longitudinal data from dengue cohorts in Southeast Asia to capture the dynamics of immunity over time, including in response to a dengue vaccine. We showed that vaccines elucidated a lower antibody response that natural infection, that vaccine efficacy was concentrated to the first three years after vaccination and that antibodies were an excellent correlate of protection. We have also identified a correlate of protection for chikungunya virus.
In the second work package, we have used a combination of sequences and mathematical models to capture the spread of Zika virus in Thailand. We showed that Zika has been circulating endemically in Thailand for over 20 years, highlighting the potential long term consequences from the virus. We have also developed novel modelling approaches that uses genetic data to explain relative importance of human behaviour data (how we move), local environment (e.g. mosquito density), and population immunity in driving dengue spread. We were able to show that infected people move a lot less than the surrounding susceptible population; that most transmission is highly local but the occasional long distance transmission event leads to mixing of the virus after a year. These methods are now being developed for other pathogens.
Finally, we have also been applying some of the techniques we have been developing in the ARBODYNAMIC project to help in the SARS-CoV-2 response. We worked with the French government to understand the level of immunity in the population and the patterns of spread in the country. We also used population immunity estimates to capture the underlying probability of death by age. This work has all been published.
This work presents significant advances in our understanding of how arboviruses move, interact with environment as well as the critical role of host immunity in determining who gets infected and who gets sick. We have also developed methods that can allow us to make unbiased estimates of viral flow even when the source of sequences we have available are very biased, both in time and in space. In the remainder of the project, we will continue to work from this foundation to obtain a more detailed understanding on the importance of the specific antigenic history of infection (ie which specific viruses you were infected with) in determining where viruses move. We will also develop easy-to-you software thar can facilitate the use of our methods by other researchers interested in pathogen spread.