Loisel Léa

Abstract:

Intra-vector viral dynamics (IVD) is a key step in the transmission of arboviruses between vectors and hosts. The objective of this thesis was to study its variability and its impact on vector-borne transmission at the population level. The development of a compartmental mechanistic model of IVD, together with its inference from data derived from vector competence experiments, made it possible to characterize the variability of IVD for chikungunya (CHIKV), dengue (DENV), Zika, West Nile and Rift Valley fever viruses. This led to questioning the usual assumption of an exponential
distribution of the extrinsic incubation period (EIP), showing that more flexible distributions, in
particular the beta distribution, better capture the observed dynamics. The introduction of a
mechanism allowing exit from the transmitting state also challenged the assumption of a persistent state, under certain experimental conditions for CHIKV. Coupling the model with a DENV transmission model made it possible to assess the impact of the EIP distribution on epidemic dynamics in humans. This distribution mainly influences the timing and intensity of epidemics rather than their overall size, with later, lower but more prolonged peaks when the distribution is informed by experimental data and not assumed to be exponential. The impact of this distribution varies with the epidemiological context. The analysis of inference quality highlighted the potential use of the IVD model to guide the choice of the number and timing of observation days in vector competence
experiments

Keywords:

within-vector viral dynamic, arboviruses, mechanistic modelling, mosquitoes, transmission,
inference