Impact of repeated annual community directed treatment with ivermectin on loiasis parasitological indicators in Cameroon: Implications for onchocerciasis and lymphatic filariasis elimination in areas co-endemic with Loa loa in Africa

Abstract

Background: Loiasis is a filarial infection endemic in the rainforest zone of west and central Africa particularly in Cameroon, Gabon, Republic of Congo, and Democratic Republic of the Congo. Repeated treatments with ivermectin have been delivered using the annual community directed treatment with ivermectin (CDTI) approach for several years to control onchocerciasis in some Loa loa-Onchocerca volvulus co-endemic areas. The impact of CDTI on loiasis parasitological indicators is not known. We, therefore, designed this cross sectional study to explore the effects of several rounds of CDTI on parasitological indicators of loiasis.

Methodology/principal findings: The study was conducted in the East, Northwest and Southwest 2 CDTI projects of Cameroon. Individuals who consented to participate were interviewed for ivermectin treatment history and enrolled for parasitological screening using thick smears. Ivermectin treatment history was correlated with loiasis prevalence/intensity. A total of 3,684 individuals were recruited from 36 communities of the 3 CDTI projects and 900 individuals from 9 villages in a non-CDTI district. In the East, loiasis prevalence was 29.3% (range = 24.2%-34.6%) in the non-CDTI district but 16.0% (3.3%-26.6%) in the CDTI district with 10 ivermectin rounds (there were no baseline data for the latter). In the Northwest and Southwest 2 districts, reductions from 30.5% to 17.9% (after 9 ivermectin rounds) but from 8.1% to 7.8% (not significantly different after 14 rounds) were registered post CDTI, respectively. Similar trends in infection intensity were observed in all sites. There was a negative relationship between adherence to ivermectin treatment and prevalence/intensity of infection in all sites. None of the children (aged 10-14 years) examined in the East CDTI project harboured high (8,000-30,000 mf/ml) or very high (>30,000 mf/ml) microfilarial loads. Individuals who had taken >5 ivermectin treatments were 2.1 times more likely to present with no microfilaraemia than those with less treatments.

Conclusion: In areas where onchocerciasis and loiasis are co-endemic, CDTI reduces the number of, and microfilaraemia in L. loa-infected individuals, and this, in turn, will help to prevent non-neurological and neurological complications post-ivermectin treatment among CDTI adherents.

Fig 1. Map showing the locations of the study sites (QGIS software version 2.0.1).

Fig 2. Pyramid of ages of males and females of the study population.

Fig 3

A-C: Box and whiskers plots of L. loa microfilarial prevalence in the adult population of the East, Northwest and Southwest 2 CDTI project sites. For the East project, baseline data were not available, but data from a geographically close non-CDT district are shown A) East sites. Left plot presents the results for the non-CDTI communities (n = 710), with median mf prevalence of 31.30% (minimum = 24%; maximum = 35%, 95% CI = 29.05–29.55); the plot on the right corresponds to 10 years of CDTI (n = 1130), with median prevalence = 15.2% (min = 3%; max = 27%, 95% CI = 16.00–16.52). B) In the Northwest site (n = 1028, 9 years of CDTI), the median prevalence on the left (baseline plot) is 34.6% (min = 10%; max = 53%, 95% CI = 29.5–31.5%) at baseline; the plot on the right (n = 1089, 9 years of CDTI) depicts a median prevalence of 19.3% (min = 2.8%; max = 35%, 95% CI = 17.8–18.8%). C) In the Southwest 2 project, the median prevalence of the left (pre-ivermectin, n = 1458) plot is 7% (min = 2.2%; max = 16.8%, 95% CI = 7.9–8.3%). Despite 14 years of CDTI (n = 733), the median mf prevalence is 5.4% on the right plot (min = 2.2%; max = 18.8%, 95% CI = 7–7.7%). The p-values shown correspond to the appropriate tests to compare 2-sample proportions applied to the overall mf prevalence (see main text), D-E: Box and whiskers plots of the prevalence of individuals with high L. loa microfilarial loads (8,001–30,000mf/ml) in the adult population of the East, Northwest and Southwest 2 CDTI project sites. For the East project, baseline data were not available, but data from a geographically close non-CDT district are shown albeit not formally compared. G) East sites. Left plot presents the results for the non-CDTI communities, with median high load mf prevalence of 6.7% (minimum = 1.4%; maximum = 12.5%, 95% CI = 4.5–9.8%); the plot on the right corresponds to 10 years of CDTI, with median prevalence = 0.7% (min = 0%; max = 4%, 95% CI = 0.5–2.1%). H) In the Northwest site (9 years of CDTI), the median prevalence on the left (baseline plot) is 8.4% (min = 0%; max = 18.7%%, 95% CI = 4.2–12.2%) at baseline; the plot on the right (9 years of CDTI) depicts a median prevalence of 2.4% (min = 0.7%; max = 4.5%, 95% CI = 1.6–3.1%). I) In the Southwest 2 project, the median prevalence of the left (pre-ivermectin) plot is 0.7% (min = 0%; max = 2.8%, 95% CI = 0.4–1.4%). Despite 14 years of CDTI, the median mf prevalence is <0.00001% on the right plot (min = 0%; max = 3.1%, 95% CI = -0.2–1.8%). The p-values shown correspond to the appropriate tests to compare 2-sample proportions applied to the overall mf prevalence (see main text) and G-I Box and whiskers plots of the prevalence of individuals with very high L. loa microfilarial loads (>30,000mf/ml) in the adult population of the East, Northwest and Southwest 2 CDTI project sites. For the East project, baseline data were not available, but data from a geographically close non-CDTI district are shown albeit not formally compared. D) East sites. Left plot presents the results for the non-CDTI communities, with median very low load mf prevalence of <0.00001% (minimum = 0%; maximum = 3.3%, 95% CI = -0.1–1.7%); the plot on the right corresponds to 10 years of CDTI, with median prevalence = <0.00001% (min = 0%; max = 2.2%, 95% CI = 0.1–0.9%). E) In the Northwest site (9 years of CDTI), the median prevalence on the left (baseline plot) is 2.4% (min = 0%; max = 11%, 95% CI = 0.8–6.1%) at baseline; the plot on the right (9 years of CDTI) depicts a median prevalence of 0.4 (min = 0%; max = 2.8%, 95% CI = 0.01–1.6%). F) In the Southwest project, only the whiskers are plotted (median and interquartile values were equal to zero), with min = 0%; max = 1.9% for the baseline data, and with min = 0% and max = 1% after 14 years of CDTI. The p-values shown correspond to the appropriate tests to compare 2-sample proportions applied to the overall mf prevalence (see main text).

Fig 4. Relationship between Loa loa microfilarial prevalence and ivermectin intake in the East, Northwest and Southwest CDTI project sites.

A) In the East, an increase in the number of treatment rounds led to a decrease in L. loa mf prevalence although with a non significant difference. B) In the Northwest, L. loa mf prevalence inversely correlated with ivermectin treatment rounds with a very significant difference. C) The same relationship was depicted in the Southwest where a decrease in L. loa mf prevalence was related to an increase in number of ivermectin treatment rounds with a significant difference. The r_s, p values and sample sizes are presented on the figures.

Fig 5. Relationship between the percentage of individuals in different Loa loa microfilarial load classes and ivermectin intake in the three CDTI projects.

A) In the East, an increase in treatment adherence led to a decrease in the proportion of individuals with high and very high mf loads. B) in the Northwest, the same relationship was depicted and here no individual with >7 treatment rounds was found to have high or very high mf loads. C) in the Southwest, individuals with > 5 treatment rounds were not found to have high or very high mf loads. The sample sizes are indicated on the figures.