Authors

J Hamley¹; M Walker²; P Milton¹; M G Basáñez¹;  ¹ Imperial College London, UK;  ² Royal Veterinary College, UK

Discussion

Background Onchocerciasis (river blindness) has been proposed, by the World Health Organization’s 2012 roadmap on neglected tropical diseases, for elimination in selected African countries by 2020 (and in 80% of endemic countries by 2025). To investigate the feasibility of elimination with current interventions (mainly mass annual or semi-annual distribution of the microfilaricidal drug ivermectin), onchocerciasis transmission models have been developed and refined to understand the factors determining the ability of control programmes to achieve interruption and elimination of transmission. Factors thus far identified include: a) the level of initial (pre-control) endemicity; b) the magnitude of transmission intensity and biting rate by local simuliid vector species (particularly in the absence of vector control); c) the duration of the programme, the therapeutic coverage achieved and maintained, and d) the level and nature of treatment adherence (random vs. systematic non-compliance). However, when developing individual-based versions of the models, it must be taken into account that individuals in a host population frequently differ in their exposure to infection. The level of exposure heterogeneity plays a central role in the distribution of parasites amongst individuals, and this distribution, in turn, determines the levels of infection prevalence, infection intensity, and the overall contribution of parasite population regulatory processes to the stability and resilience of the infection to interventions. Large variation between hosts in the number of vector bites they receive will result in high levels of parasite overdispersion. At a given prevalence, a population of highly aggregated parasites will require increased levels of mass drug administration (MDA), or the deployment of other (including targeted) interventions, in comparison to those with low aggregation, to suppress transmission. 

Methods  We use an individual-based version of our model, EPIONCHO-IBM, which tracks individual humans and accounts for the age-structure of adult worms and microfilariae, allowing for senescence in parasite mortality and fecundity. In line with deterministic EPIONCHO, we account for various density-dependent processes in parasite establishment and vector survival. The goal of this work is to explore how existing data on the relationship between infection prevalence and vector biting rate can be used to incorporate exposure heterogeneity into EPIONCHO-IBM, and explore how this can influence trends in (parasitological, serological and entomological) markers of infection following long-term ivermectin treatment as well as the dynamics of recrudescence after treatment cessation.

Results and Conclusions

Individual host variation in exposure to vectors is a crucial determinant of epidemiological dynamics during and after the treatment phase of a programme. Analyses of prevalence-vector biting rate data indicate that strong heterogeneity is necessary to stabilise low levels of infection prevalence. Levels of worm overdispersion were in broad agreement with those of other studies using different types of (parasitological) data and more severe than values used by other models predicting the probability of elimination. Areas of low pre-intervention prevalence and high exposure heterogeneity may experience infection recrudescence faster than areas of high pre-intervention prevalence but low exposure heterogeneity, for a given MDA duration and coverage. There is a pressing need to identify, gather, and analyse data on distributions of infection in host populations and their determinants (e.g. proximity to vector breeding sites, behavioural, occupational, immunological factors), to better parameterise parasite transmission and control/elimination models.