Modelling the public health impact of second-generation malaria vaccines
File(s)
Author(s)
Thompson, Hayley Adelaide
Type
Thesis
Abstract
Despite significant progress in the control and elimination of malaria over the past two decades, the global burden remains high. The COVID-19 pandemic has seen malaria cases and deaths increase markedly over 2019 resulting in 241 million malaria cases and 627,000 malaria deaths worldwide in 2020, an increase of 14 million cases and 69,000 deaths. Around 47,000 of these additional deaths were linked to pandemic-related disruptions in the provision of malaria prevention, diagnosis and treatment. The need for a highly efficacious childhood malaria vaccine has never felt more pertinent and in 2021, after 30 years of research and development, the World Health Organization recommended the first ever childhood vaccine against P. falciparum malaria, RTS,S/AS01E (RTS,S) for widespread use.
The development of RTS,S, its deployment and continued evaluation has facilitated the synthesis of knowledge and data from across a wide range of different disciplines involved in malaria vaccine research. This depth of data has enhanced the development of mathematical modelling frameworks that combine immunological insights with epidemiological transmission models to address public health questions. These frameworks have formed a core part of the evaluation and policy recommendations surrounding RTS,S.
The work presented in this thesis builds upon these modelling frameworks to provide insights into the potential impact of alternative RTS,S vaccination approaches. The two RTS,S approaches examined in this thesis are a delayed-fractional primary series and a seasonally targeted vaccination schedule, both of which have demonstrated promising efficacy in human challenge studies and field trials respectively.
Drawing on data from the delayed-fractional RTS,S human challenge study I used a Bayesian framework to investigate immunological correlates of vaccine induced protection. I estimate that improvements to the quality, measured as antibody avidity, and not the quantity, measured as antibody titre, of the vaccine induced antibody response is critical to the increased efficacy against infection observed with this schedule.
Next, I utilised data from seasonal malaria chemoprevention and seasonal RTS,S vaccination clinical trials to fit and validate an updated efficacy profile of the drug combination Sulfadoxine-pyrimethamine plus amodiaquine (SP+AQ) used for seasonal malaria chemoprevention using a Bayesian survival analysis framework. This approach enabled me to capture uncertainty in the protection provided by seasonal malaria chemoprevention over time. I then use this updated efficacy profile along with the existing RTS,S vaccine efficacy profile to replicate trial cohorts in a transmission model in order to validate the intervention models against clinical trial data. I found that the existing RTS,S model underestimated the protection provided by the seasonal vaccination schedule and explored several biologically motivated alterations to the model that brought results in line with those of the trial. These results combined with the trial reported antibody data suggest that efficacy improvements with this regime were not driven by increases in antibody quantity. Further model results suggest that when vaccination and chemoprevention were combined this resulted in potential synergistic interactions that enhanced the efficacy of SP+AQ in particular. This work resulted therefore in several updated versions of RTS,S and SP+AQ efficacy models that capture the current uncertainty in intervention effects.
Finally extending these updated efficacy models from the validation exercise, I used a model of malaria transmission to investigate the long-term public health impact of novel RTS,S vaccination schedules compared to the original age-based RTS,S dosing schedule in seasonal settings. I considered the impact both in the presence and absence of seasonal malaria chemoprevention. I examined impact by degree of seasonality, transmission intensity and by wider health system and operational factors. RTS,S vaccination in seasonal malaria transmission settings could be a valuable additional tool to existing seasonal interventions, with seasonal delivery maximising impact relative to an age-based approach. Decisions surrounding deployment strategies of RTS,S in such settings will need to consider the local and regional variations in seasonality, current levels of other interventions and potential achievable RTS,S coverage.
The development of RTS,S, its deployment and continued evaluation has facilitated the synthesis of knowledge and data from across a wide range of different disciplines involved in malaria vaccine research. This depth of data has enhanced the development of mathematical modelling frameworks that combine immunological insights with epidemiological transmission models to address public health questions. These frameworks have formed a core part of the evaluation and policy recommendations surrounding RTS,S.
The work presented in this thesis builds upon these modelling frameworks to provide insights into the potential impact of alternative RTS,S vaccination approaches. The two RTS,S approaches examined in this thesis are a delayed-fractional primary series and a seasonally targeted vaccination schedule, both of which have demonstrated promising efficacy in human challenge studies and field trials respectively.
Drawing on data from the delayed-fractional RTS,S human challenge study I used a Bayesian framework to investigate immunological correlates of vaccine induced protection. I estimate that improvements to the quality, measured as antibody avidity, and not the quantity, measured as antibody titre, of the vaccine induced antibody response is critical to the increased efficacy against infection observed with this schedule.
Next, I utilised data from seasonal malaria chemoprevention and seasonal RTS,S vaccination clinical trials to fit and validate an updated efficacy profile of the drug combination Sulfadoxine-pyrimethamine plus amodiaquine (SP+AQ) used for seasonal malaria chemoprevention using a Bayesian survival analysis framework. This approach enabled me to capture uncertainty in the protection provided by seasonal malaria chemoprevention over time. I then use this updated efficacy profile along with the existing RTS,S vaccine efficacy profile to replicate trial cohorts in a transmission model in order to validate the intervention models against clinical trial data. I found that the existing RTS,S model underestimated the protection provided by the seasonal vaccination schedule and explored several biologically motivated alterations to the model that brought results in line with those of the trial. These results combined with the trial reported antibody data suggest that efficacy improvements with this regime were not driven by increases in antibody quantity. Further model results suggest that when vaccination and chemoprevention were combined this resulted in potential synergistic interactions that enhanced the efficacy of SP+AQ in particular. This work resulted therefore in several updated versions of RTS,S and SP+AQ efficacy models that capture the current uncertainty in intervention effects.
Finally extending these updated efficacy models from the validation exercise, I used a model of malaria transmission to investigate the long-term public health impact of novel RTS,S vaccination schedules compared to the original age-based RTS,S dosing schedule in seasonal settings. I considered the impact both in the presence and absence of seasonal malaria chemoprevention. I examined impact by degree of seasonality, transmission intensity and by wider health system and operational factors. RTS,S vaccination in seasonal malaria transmission settings could be a valuable additional tool to existing seasonal interventions, with seasonal delivery maximising impact relative to an age-based approach. Decisions surrounding deployment strategies of RTS,S in such settings will need to consider the local and regional variations in seasonality, current levels of other interventions and potential achievable RTS,S coverage.
Version
Open Access
Date Issued
2022-05
Date Awarded
2022-09
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Ghani, Azra
Sponsor
Medical Research Council (Great Britain)
Grant Number
G98669
Publisher Department
School of Public Health
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)