Supporting Drug Development for Neglected Tropical Diseases Using Mathematical Modeling.

Drug-based interventions are at the heart of global efforts to reach elimination as a public health problem (trachoma, soil-transmitted helminthiases, schistosomiasis, lymphatic filariasis) or elimination of transmission (onchocerciasis) for 5 of the most prevalent neglected tropical diseases tackled via the World Health Organization preventive chemotherapy strategy. While for some of these diseases there is optimism that currently available drugs will be sufficient to achieve the proposed elimination goals, for others-particularly onchocerciasis-there is a growing consensus that novel therapeutic options will be needed. Since in this area no high return of investment is possible, minimizing wasted money and resources is essential. Here, we use illustrative results to show how mathematical modeling can guide the drug development pathway, yielding resource-saving and efficiency payoffs, from the refinement of target product profiles and intended context of use to the design of clinical trials.

Structural Uncertainty in Onchocerciasis Transmission Models Influences the Estimation of Elimination Thresholds and Selection of Age Groups for Seromonitoring.

BACKGROUND: The World Health Organization recommends monitoring Onchocerca volvulus Ov16 serology in children aged <10 years for stopping mass ivermectin administration. Transmission models can help to identify the most informative age groups for serological monitoring and investigate the discriminatory power of serology-based elimination thresholds. Model predictions depend on assumed age-exposure patterns and transmission efficiency at low infection levels. METHODS: The individual-based transmission model, EPIONCHO-IBM, was used to assess (1) the most informative age groups for serological monitoring using receiver operating characteristic curves for different elimination thresholds under various age-dependent exposure assumptions, including those of ONCHOSIM (another widely used model), and (2) the influence of within-human density-dependent parasite establishment (included in EPIONCHO-IBM but not ONCHOSIM) on positive predictive values for different serological thresholds. RESULTS: When assuming EPIONCHO-IBM exposure patterns, children aged <10 years are the most informative for seromonitoring; when assuming ONCHOSIM exposure patterns, 5-14 year olds are the most informative (as published elsewhere). Omitting density-dependent parasite establishment results in more lenient seroprevalence thresholds, even for higher baseline infection prevalence and shorter treatment durations. CONCLUSIONS: Selecting appropriate seromonitoring age groups depends critically on age-dependent exposure patterns. The role of density dependence on elimination thresholds largely explains differing EPIONCHO-IBM and ONCHOSIM elimination predictions.

Can mass drug administration of moxidectin accelerate onchocerciasis elimination in Africa?

Epidemiological and modelling studies suggest that elimination of Onchocerca volvulus transmission (EoT) throughout Africa may not be achievable with annual mass drug administration (MDA) of ivermectin alone, particularly in areas of high endemicity and vector density. Single-dose Phase II and III clinical trials demonstrated moxidectin's superiority over ivermectin for prolonged clearance of O. volvulus microfilariae. We used the stochastic, individual-based EPIONCHO-IBM model to compare the probabilities of reaching EoT between ivermectin and moxidectin MDA for a range of endemicity levels (30 to 70% baseline microfilarial prevalence), treatment frequencies (annual and biannual) and therapeutic coverage/adherence values (65 and 80% of total population, with, respectively, 5 and 1% of systematic non-adherence). EPIONCHO-IBM's projections indicate that biannual (six-monthly) moxidectin MDA can reduce by half the number of years necessary to achieve EoT in mesoendemic areas and might be the only strategy that can achieve EoT in hyperendemic areas. Data needed to improve modelling projections include (i) the effect of repeated annual and biannual moxidectin treatment; (ii) inter- and intra-individual variation in response to successive treatments with moxidectin or ivermectin; (iii) the effect of moxidectin and ivermectin treatment on L3 development into adult worms; and (iv) patterns of adherence to moxidectin and ivermectin MDA. This article is part of the theme issue 'Challenges in the fight against neglected tropical diseases: a decade from the London Declaration on NTDs'.

Demographic patterns of human antibody levels to Simulium damnosum s.l. saliva in onchocerciasis-endemic areas: An indicator of exposure to vector bites.

BACKGROUND: In onchocerciasis endemic areas in Africa, heterogenous biting rates by blackfly vectors on humans are assumed to partially explain age- and sex-dependent infection patterns with Onchocerca volvulus. To underpin these assumptions and further improve predictions made by onchocerciasis transmission models, demographic patterns in antibody responses to salivary antigens of Simulium damnosum s.l. are evaluated as a measure of blackfly exposure. METHODOLOGY/PRINCIPAL FINDINGS: Recently developed IgG and IgM anti-saliva immunoassays for S. damnosum s.l. were applied to blood samples collected from residents in four onchocerciasis endemic villages in Ghana. Demographic patterns in antibody levels according to village, sex and age were explored by fitting generalized linear models. Antibody levels varied between villages but showed consistent patterns with age and sex. Both IgG and IgM responses declined with increasing age. IgG responses were generally lower in males than in females and exhibited a steeper decline in adult males than in adult females. No sex-specific difference was observed in IgM responses. CONCLUSIONS/SIGNIFICANCE: The decline in age-specific antibody patterns suggested development of immunotolerance or desensitization to blackfly saliva antigen in response to persistent exposure. The variation between sexes, and between adults and youngsters may reflect differences in behaviour influencing cumulative exposure. These measures of antibody acquisition and decay could be incorporated into onchocerciasis transmission models towards informing onchocerciasis control, elimination, and surveillance.

What does the COVID-19 pandemic mean for the next decade of onchocerciasis control and elimination?

BACKGROUND: Mass drug administration (MDA) of ivermectin for onchocerciasis has been disrupted by the coronavirus disease 2019 (COVID-19) pandemic. Mathematical modelling can help predict how missed/delayed MDA will affect short-term epidemiological trends and elimination prospects by 2030. METHODS: Two onchocerciasis transmission models (EPIONCHO-IBM and ONCHOSIM) are used to simulate microfilarial prevalence trends, elimination probabilities and age profiles of Onchocerca volvulus microfilarial prevalence and intensity for different treatment histories and transmission settings, assuming no interruption, a 1-y (2020) interruption or a 2-y (2020-2021) interruption. Biannual MDA or increased coverage upon MDA resumption are investigated as remedial strategies. RESULTS: Programmes with shorter MDA histories and settings with high pre-intervention endemicity will be the most affected. Biannual MDA is more effective than increasing coverage for mitigating COVID-19's impact on MDA. Programmes that had already switched to biannual MDA should be minimally affected. In high-transmission settings with short treatment history, a 2-y interruption could lead to increased microfilarial load in children (EPIONCHO-IBM) and adults (ONCHOSIM). CONCLUSIONS: Programmes with shorter (annual MDA) treatment histories should be prioritised for remedial biannual MDA. Increases in microfilarial load could have short- and long-term morbidity and mortality repercussions. These results can guide decision-making to mitigate the impact of COVID-19 on onchocerciasis elimination.

Moxidectin: an oral treatment for human onchocerciasis.

INTRODUCTION: Moxidectin is a milbemycin endectocide recently approved for the treatment of human onchocerciasis. Onchocerciasis, earmarked for elimination of transmission, is a filarial infection endemic in Africa, Yemen, and the Amazonian focus straddling Venezuela and Brazil. Concerns over whether the predominant treatment strategy (yearly mass drug administration (MDA) of ivermectin) is sufficient to achieve elimination in all endemic foci have refocussed attention upon alternative treatments. Moxidectin's stronger and longer microfilarial suppression compared to ivermectin in both phase II and III clinical trials indicates its potential as a novel powerful drug for onchocerciasis elimination. AREAS COVERED: This work summarizes the chemistry and pharmacology of moxidectin, reviews the phase II and III clinical trials evidence on tolerability, safety, and efficacy of moxidectin versus ivermectin, and discusses the implications of moxidectin's current regulatory status. EXPERT OPINION: Moxidectin's superior clinical performance has the potential to substantially reduce times to elimination compared to ivermectin. If donated, moxidectin could mitigate the additional programmatic costs of biannual ivermectin distribution because, unlike other alternatives, it can use the existing community-directed treatment infrastructure. A pediatric indication (for children <12 years) and determination of its usefulness in onchocerciasis-loiasis co-endemic areas will greatly help fulfill the potential of moxidectin for the treatment and elimination of onchocerciasis.

Designing antifilarial drug trials using clinical trial simulators.

Lymphatic filariasis and onchocerciasis are neglected tropical diseases (NTDs) targeted for elimination by mass (antifilarial) drug administration. These drugs are predominantly active against the microfilarial progeny of adult worms. New drugs or combinations are needed to improve patient therapy and to enhance the effectiveness of interventions in persistent hotspots of transmission. Several therapies and regimens are currently in (pre-)clinical testing. Clinical trial simulators (CTSs) project patient outcomes to inform the design of clinical trials but have not been widely applied to NTDs, where their resource-saving payoffs could be highly beneficial. We demonstrate the utility of CTSs using our individual-based onchocerciasis transmission model (EPIONCHO-IBM) that projects trial outcomes of a hypothetical macrofilaricidal drug. We identify key design decisions that influence the power of clinical trials, including participant eligibility criteria and post-treatment follow-up times for measuring infection indicators. We discuss how CTSs help to inform target product profiles.

State-level tracking of COVID-19 in the United States.

As of 1st June 2020, the US Centres for Disease Control and Prevention reported 104,232 confirmed or probable COVID-19-related deaths in the US. This was more than twice the number of deaths reported in the next most severely impacted country. We jointly model the US epidemic at the state-level, using publicly available death data within a Bayesian hierarchical semi-mechanistic framework. For each state, we estimate the number of individuals that have been infected, the number of individuals that are currently infectious and the time-varying reproduction number (the average number of secondary infections caused by an infected person). We use changes in mobility to capture the impact that non-pharmaceutical interventions and other behaviour changes have on the rate of transmission of SARS-CoV-2. We estimate that Rt was only below one in 23 states on 1st June. We also estimate that 3.7% [3.4%-4.0%] of the total population of the US had been infected, with wide variation between states, and approximately 0.01% of the population was infectious. We demonstrate good 3 week model forecasts of deaths with low error and good coverage of our credible intervals.

Spatial analysis made easy with linear regression and kernels.

Kernel methods are a popular technique for extending linear models to handle non-linear spatial problems via a mapping to an implicit, high-dimensional feature space. While kernel methods are computationally cheaper than an explicit feature mapping, they are still subject to cubic cost on the number of points. Given only a few thousand locations, this computational cost rapidly outstrips the currently available computational power. This paper aims to provide an overview of kernel methods from first-principals (with a focus on ridge regression) and progress to a review of random Fourier features (RFF), a method that enables the scaling of kernel methods to big datasets. We show how the RFF method is capable of approximating the full kernel matrix, providing a significant computational speed-up for a negligible cost to accuracy and can be incorporated into many existing spatial methods using only a few lines of code. We give an example of the implementation of RFFs on a simulated spatial data set to illustrate these properties. Lastly, we summarise the main issues with RFFs and highlight some of the advanced techniques aimed at alleviating them. At each stage, the associated R code is provided.

Modelling exposure heterogeneity and density dependence in onchocerciasis using a novel individual-based transmission model, EPIONCHO-IBM: Implications for elimination and data needs.

BACKGROUND: Density dependence in helminth establishment and heterogeneity in exposure to infection are known to drive resilience to interventions based on mass drug administration (MDA). However, the interaction between these processes is poorly understood. We developed a novel individual-based model for onchocerciasis transmission, EPIONCHO-IBM, which accounts for both processes. We fit the model to pre-intervention epidemiological data and explore parasite dynamics during MDA with ivermectin. METHODOLOGY/PRINCIPAL FINDINGS: Density dependence and heterogeneity in exposure to blackfly (vector) bites were estimated by fitting the model to matched pre-intervention microfilarial prevalence, microfilarial intensity and vector biting rate data from savannah areas of Cameroon and Côte d'Ivoire/Burkina Faso using Latin hypercube sampling. Transmission dynamics during 25 years of annual and biannual ivermectin MDA were investigated. Density dependence in parasite establishment within humans was estimated for different levels of (fixed) exposure heterogeneity to understand how parametric uncertainty may influence treatment dynamics. Stronger overdispersion in exposure to blackfly bites results in the estimation of stronger density-dependent parasite establishment within humans, consequently increasing resilience to MDA. For all levels of exposure heterogeneity tested, the model predicts a departure from the functional forms for density dependence assumed in the deterministic version of the model. CONCLUSIONS/SIGNIFICANCE: This is the first, stochastic model of onchocerciasis, that accounts for and estimates density-dependent parasite establishment in humans alongside exposure heterogeneity. Capturing the interaction between these processes is fundamental to our understanding of resilience to MDA interventions. Given that uncertainty in these processes results in very different treatment dynamics, collecting data on exposure heterogeneity would be essential for improving model predictions during MDA. We discuss possible ways in which such data may be collected as well as the importance of better understanding the effects of immunological responses on establishing parasites prior to and during ivermectin treatment.