The potential impact of moxidectin on onchocerciasis elimination in Africa: an economic evaluation based on the Phase II clinical trial data

Abstract

Background: Spurred by success in several foci, onchocerciasis control policy in Africa has shifted from morbidity control to elimination of infection. Clinical trials have demonstrated that moxidectin is substantially more efficacious than ivermectin in effecting sustained reductions in skin microfilarial load and, therefore, may accelerate progress towards elimination. We compare the potential cost-effectiveness of annual moxidectin with annual and biannual ivermectin treatment.

Methods: Data from the first clinical study of moxidectin were used to parameterise the onchocerciasis transmission model EPIONCHO to investigate, for different epidemiological and programmatic scenarios in African savannah settings, the number of years and in-country costs necessary to reach the operational thresholds for cessation of treatment, comparing annual and biannual ivermectin with annual moxidectin treatment.

Results: Annual moxidectin and biannual ivermectin treatment would achieve similar reductions in programme duration relative to annual ivermectin treatment. Unlike biannual ivermectin treatment, annual moxidectin treatment would not incur a considerable increase in programmatic costs and, therefore, would generate sizeable in-country cost savings (assuming the drug is donated). Furthermore, the impact of moxidectin, unlike ivermectin, was not substantively influenced by the timing of treatment relative to seasonal patterns of transmission.

Conclusions: Moxidectin is a promising new drug for the control and elimination of onchocerciasis. It has high programmatic value particularly when resource limitation prevents a biannual treatment strategy, or optimal timing of treatment relative to peak transmission season is not feasible.

Figure 1

The dynamic effect of a single dose of ivermectin (A) and moxidectin (B) on skin microfilarial load. The data points are derived from skin microfilarial loads (the mean of four microfilarial counts [17]) collected from (A) the 45 control participants (who took ivermectin) and (B) the 38 treated participants (who took moxidectin) as part of the Phase II clinical safety trial of moxidectin for the treatment of onchocerciasis [17]. The effect of a single dose of ivermectin previously fitted to microfilarial load data collated as part of a meta-analysis [20] is shown as the solid blue line in (A) (note that the microfilarial dynamics induced by ivermectin are not re-estimated here and hence provide a validation of the previous parameterization). The dynamical effect precipitated by moxidectin was fitted to the trial data on microfilarial loads from treated participants using the same approach as in [20] (described in the Supporting information (Additional file 1: Text S.2)) and is shown as the solid red line in (B). Error bars show the 95% confidence intervals which in some circumstances were narrower that the plotted data point and so are not discernible.

Figure 2

Comparison of annual and biannual community-directed treatment with ivermectin (aCDTI, bCDTI) versus annual community-directed treatment with moxidectin (aCDTM) in areas where control has not been previously implemented. The green, blue and red bars correspond to a pre-control endemicity level of 40%, 60%, and 80% microfilarial prevalence, respectively. The aCDTI, bCDTI and aCDTM strategies are indicated by solid, dashed and dotted bars respectively. Error bars represent the results of varying the provisional operational thresholds for treatment interruption followed by surveillance (pOTTIS) by adding or subtracting 0.5% (i.e. 0.9% or 1.9% microfilarial prevalence). Results shown assume a therapeutic coverage of 80%; a proportion of systematic non-compliers of 0.1%; perennial transmission, and a 7% per dose (cumulative) reduction in microfilarial production of female adult worms. A discount rate of 3% was applied to the costs. *pOTTIS (1.4% microfilarial prevalence) not attained within the 50-year time horizon and percentage of costs calculated based on costs of 50 years of aCDTI. † Costs do not include value of the (donated) drugs.

Figure 3

Impact of switching to biannual community-directed treatment with ivermectin (bCDTI) or annual community-directed treatment with moxidectin (aCDTM) at different stages of an ongoing annual CDTI (aCDTI) programme. The green, blue and red bars correspond to a pre-control endemicity level of 40%, 60%, and 80% microfilarial prevalence, respectively. The aCDTI, bCDTI and aCDTM strategies are indicated by solid, dashed and dotted bars, respectively. Error bars represent the results of varying the provisional operational thresholds for treatment interruption followed by surveillance (pOTTIS) by ± 0.5%. The number of additional years of treatment and the ratio of additional costs are considered from the point of switching to bCDTI or aCDTM (and not from the start of control). Modelling assumptions are as in the legend of Figure 2. *pOTTIS (1.4% microfilarial prevalence) not attained within the 50-year time horizon and percentage of costs calculated based on costs of 50 years of aCDTI. † Costs do not include value of the (donated) drugs.

Figure 4

The effect of coverage and non-compliance on programme duration under strategies of annual and biannual community-directed treatment with ivermectin (aCDTI, bCDTI) and annual community-directed treatment with moxidectin (aCDTM). Black bars represent the increment in programme duration caused by a decrease in therapeutic coverage from 80% to 60% of the total population. The green, blue and red bars correspond to a pre-control endemicity level of 40%, 60%, and 80% microfilarial prevalence, respectively. aCDTI, bCDTI and aCDTM are indicated by solid, dashed and dotted bars, respectively. Modelling assumptions are as in the legend of Figure 2. *Provisional operational threshold for treatment interruption followed by surveillance (pOTTIS <1.4% microfilarial prevalence) not attained within the 50-year time horizon.

Figure 5

The effect of the timing of treatment rounds relative to seasonal peaks in transmission by blackfly vectors on programme duration under strategies of annual community-directed treatment with ivermectin (aCDTI) and annual community-directed treatment with moxidectin (aCDTM). The time between treatment and the peak of transmission was varied. The x-axis represents the number of months between treatment at time 0 and the peak of transmission. Scenario 1: transmission only takes place during a five month long rainy season. Scenario 2: high levels of transmission during a five month long rainy season but continuing at a low level during the rest of the year (Table 1 and Additional file 1: Figure S1). The green, blue and red lines correspond to, respectively, a pre-control endemicity level of 40%, 60%, and 80% microfilarial prevalence. The years are those needed to reach the pOTTIS (<1.4% microfilarial prevalence). The lines show the mean microfilarial load per mg of skin after 15 years of treatment in those aged ≥ 20 years. Solid and dashed lines indicate aCDTI and aCDTM respectively. Modelling assumptions are as in the legend of Figure 2.