Long-term trends in Anopheles gambiae insecticide resistance in Côte d'Ivoire

Abstract

Background: Malaria control is heavily dependent on the use of insecticides that target adult mosquito vectors via insecticide treated nets (ITNs) or indoor residual spraying (IRS). Four classes of insecticide are approved for IRS but only pyrethroids are available for ITNs. The rapid rise in insecticide resistance in African malaria vectors has raised alarms about the sustainability of existing malaria control activities. This problem might be particularly acute in Côte d'Ivoire where resistance to all four insecticide classes has recently been recorded. Here we investigate temporal trends in insecticide resistance across the ecological zones of Côte d'Ivoire to determine whether apparent pan-African patterns of increasing resistance are detectable and consistent across insecticides and areas.

Methods: We combined data on insecticide resistance from a literature review, and bioassays conducted on field-caught Anopheles gambiae mosquitoes for the four WHO-approved insecticide classes for ITN/IRS. The data were then mapped using Geographical Information Systems (GIS) and the IR mapper tool to provide spatial and temporal distribution data on insecticide resistance in An. gambiae sensu lato from Côte d'Ivoire between 1993 and 2014.

Results: Bioassay mortality decreased over time for all insecticide classes, though with significant spatiotemporal variation, such that stronger declines were observed in the southern ecological zone for DDT and pyrethroids than in the central zone, but with an apparently opposite effect for the carbamate and organophosphate. Variation in relative abundance of the molecular forms, coupled with dramatic increase in kdr 1014F frequency in M forms (An. coluzzii) seems likely to be a contributory factor to these patterns. Although records of resistance across insecticide classes have become more common, the number of classes tested in studies has also increased, precluding a conclusion that multiple resistance has also increased.

Conclusion: Our analyses attempted synthesis of 22 years of bioassay data from Côte d'Ivoire, and despite a number of caveats and potentially confounding variables, suggest significant but spatially-variable temporal trends in insecticide resistance. In the light of such spatio-temporal dynamics, regular, systematic and spatially-expanded monitoring is warranted to provide accurate information on insecticide resistance for control programme management.

Figure 1

Map of Côte d’Ivoire showing the distribution of main ecological zones. Modified from “Ecoregions of Côte d’Ivoire”. Source: World Wildlife Fund. Encyclopaedia of earth: http://www.eoearth.org/view/article/151626/.

Figure 2

Distribution of An. gambiae M (black) and S (red) molecular forms across the four ecological zones of Côte d’Ivoire; each recording is shown by a separate point and mean frequency prevalence described in Plot (A). Mean kdr 1014F frequency across sample sites in each form in early samples (from 1998) and more recent samples (2004–2012); and sample sites (and total numbers genotyped) are shown at the top of the plot (B).

Figure 3

Temporal trends in mortality to (A) DDT; (B) deltamethrin and (C) permethrin in the four ecological zones of Côte d’Ivoire. Solid lines indicate significant correlations, dashed lines are non-significant. In all plots quantitative analysis was performed only on ecozones 1 and 2. In all analyses there was a significant effect of year and year × ecozone on mortality. In (C) open symbols show data for bioassays with a concentration of 1% and filled circles with the current WHO standard of 0.75%.

Figure 4

Temporal trends in mortality to (A) carbamates (propoxur and bendiocarb) and (B) organophosphate (fenitrothion, barring two open symbols for malathion) in the southern and central ecological zones of Côte d’Ivoire. Solid lines indicate significant correlations, dashed lines are non-significant. In both analyses there was a significant effect of year x ecozone on mortality.

Figure 5

Records of resistance to different insecticides classes in relation to the number of classes tested in the two decades (a) and (b) spanning 1993 to 2012. Inset key shows how colours and shapes of the collection site markers illustrate the number of insecticide classes to which the mosquitoes were resistant and the number tested, respectively.