An ecotoxicological view on malaria vector control with ivermectin-treated cattle

The study addresses persistent malaria transmission in sub-Saharan Africa driven in part by Anopheles mosquitoes that exhibit behavioral resistance to indoor insecticides and bed nets by feeding on animals. This zoophagic and opportunistic feeding sustains vector populations and maintains residual transmission. The research question is whether treating cattle with ivermectin can sustainably complement existing vector control by making livestock blood toxic to mosquitoes, while integrating environmental health considerations. Framed within a One Health perspective inspired by historical lessons from DDT use, the purpose is twofold: (1) evaluate the efficacy and durability of a long-acting ivermectin depot formulation to reduce mosquito survival and predicted field populations; and (2) assess environmental fate and ecotoxicological risks of excreted ivermectin residues to non-target dung and soil organisms, proposing locally adapted risk mitigation for West African agroecosystems. The importance lies in bridging effective vector control with environmental safety to achieve sustainable malaria management.

The paper situates its work within literature on residual malaria transmission, behavioral and physiological resistance of Anopheles to conventional insecticides, and the potential of endectocide-treated livestock (ETL) as a complementary tool. Prior studies have shown short protective windows (~2 weeks) for standard ivermectin formulations and interest in long-acting delivery to span transmission seasons. Extensive ecotoxicology literature over 30+ years documents non-target effects of ivermectin and related avermectins on dung and soil biota, with field evidence of reduced dung insect diversity and ecosystem functioning, especially under long-acting formulations (e.g., eprinomectin). However, most data derive from temperate/Mediterranean systems, with notable gaps for West African dung fauna and tropical soils. Environmental fate research indicates strong sorption and moderate persistence of ivermectin in soils and manure, with dissipation influenced by temperature, soil properties, and aeration, but limited data under Sahelian/Sudanian/Guinean climates. WHO and EMA guidance provides criteria for endectocides and frameworks for environmental risk assessment (PEC-based), but adaptation to West African practices is needed.

Design: A randomized controlled study with 24 male crossbred cattle (Soudanese Fulani Zebu × Baoulé Taurine) allocated to three arms (n=8 each; one spare per arm; analyses on 7/arm): (1) monthly injections of commercial ivermectin formulation IVOMEC-D at 0.4 mg kg−1 bw; (2) single long-acting depot ivermectin formulation IVM-BEPO at 2.4 mg kg−1 bw; (3) untreated control. Doses were doubled relative to previous work to increase efficacy/remanence. Study duration: 34 weeks.

Formulations and dosing: IVM-BEPO is an in situ-forming depot based on biodegradable copolymers (BEPO technology): 45% w/w copolymer (triblock PLA97-PEG5-PLA97 and diblock mPEG45-PLA130), 5% w/w ivermectin, 50% w/w DMSO. After subcutaneous injection, polymers precipitate forming a depot for sustained release (target ≥6 months). Sterile-filtered (0.2 μm) and injected with 16G needle; volumes adjusted to projected weight at 3 months (accounting for expected 10–25% weight gain). IVOMEC-D (commercial injectable) dosed 0.4 mg kg−1 monthly, subcutaneous with 18G needle; cattle weighed before each injection.

Animal husbandry and ethics: Conducted at CIRDES (Bobo-Dioulasso, Burkina Faso); ethics approval 15/CE-CIRDES/16-10-2018. Animals kept in net-protected stable, pre-treated with diminazene aceturate and albendazole (to avoid ivermectin contamination), monitored by veterinary staff. Diet: rice straw (5 kg) and cotton cake (1 kg) per animal daily; water and mineral block ad libitum.

Sampling and assays: Blood plasma and dung sampled at scheduled time points (including pre-treatment controls) to quantify ivermectin pharmacokinetics. Direct skin-feeding assays performed at 20 exposure instances (including one pre-treatment) using an A. coluzzii colony established from wild females captured in Kou Valley. Target ~40 females per cattle per timepoint (in 4 cages of 10); daily survival tracked up to 30 days. Overall, 26,637 females exposed; 94% fed successfully.

Environmental studies: Selected dung samples stored to study ivermectin dissipation under two regimes: internal (insectary; 26±2°C, 75±5% RH) and external semi-field (shaded outdoors; day 32.1–39.6°C; night 17.9–22.8°C). Storage durations: 30, 60, 90 days; vials covered with bed net. Soils from 30 arable fields in three villages (Sébédougou, Waly, Kari) sampled (0–15 cm composite samples), characterized (pH in CaCl2, EC, coarse fraction, particle size, C:N; Ctotal as Corg proxy). Sorption experiments conducted on six selected fine-soil fractions (<2 mm) to derive Kd and Koc and correlate with soil properties (Pearson r).

Chemical analysis: Plasma and freeze-dried/ground dung were extracted with acetonitrile after spiking with doramectin surrogate; ultrasound-assisted extraction; centrifugation; evaporation; reconstitution; filtration. HPLC-fluorescence after derivatization (N-methylimidazole/acetonitrile, triethylamine, trifluoroacetic anhydride, trifluoroacetic acid); RP-C18 column; excitation 364 nm/emission 463 nm. LOD/LOQ: plasma LOD 1.47 ng mL−1, LOQ 4.47 ng mL−1; dung LOD 5.05 ng g−1 dw, LOQ 15.31 ng g−1 dw. Recoveries (doramectin): plasma 106.5±21%; dung 103.2±19.7%. Sub-LOQ values imputed as 0.5×LOQ. Technical duplicates per sample.

Statistical analysis and modelling: Noncompartmental PK (PKNCA) to estimate Cmax, Tmax, half-life, AUC for plasma and dung (28 days for IVOMEC-D; 211 days for IVM-BEPO). Survival analyses: Kaplan–Meier and Cox proportional hazards mixed models (cattle as random effect), hazard ratios and 95% CIs per timepoint. Dose–response: multivariate four-parameter log-logistic model to estimate 7-, 10-, 13-day LC50/LC90. Population modelling combined: (1) GAM + one-compartment PK model for plasma dynamics; (2) Cox model linking plasma to mosquito mortality; (3) two-host deterministic SEI malaria transmission model with LLINs and ivermectin-treated cattle; (4) vector behaviour model (host preference, LLIN encounter mortality). Scenarios varied cattle:human ratio r, human preference a, LLIN coverage, yielding HBI. Environmental dissipation in stored dung modelled with first-order kinetics to estimate k, DT50, DT90. Exposure assessment included calculation of an example PECsoil_initial for long-acting dosing using EMA framework.

Pharmacokinetics:

• IVOMEC-D (0.4 mg kg−1 monthly): ivermectin detectable in plasma and dung up to 28 days post-injection; peak mean plasma concentration 122±35 ng mL−1 on DAI 7 after the sixth injection (day 160); dung peak 1,927±912 ng g−1 dw on DAI 2 after first injection.
• IVM-BEPO (2.4 mg kg−1 single depot): plasma concentrations above LOQ throughout 211 days (study duration); max plasma concentration 22.9±8.9 ng mL−1 at DAI 7 in first month and 32.7±6.2 ng mL−1 at DAI 204 overall; dung peak 530±327 ng g−1 dw at DAI 7. Strong linear relation between plasma and dung concentrations for IVM-BEPO: dung (ng g−1 dw) = 16.585 × plasma (ng mL−1), R2 = 0.83.

Mosquitocidal efficacy:

• Feeding success ~94% across arms (no difference; Wald χ2 = 1.70, P = 0.42). Survival tracked for 5,384 (control), 4,852 (IVOMEC-D), 5,400 (IVM-BEPO) females.
• IVM-BEPO significantly increased mortality vs control at all timepoints (all hazard ratios >1; P < 0.001). At most sampling dates, IVOMEC-D mortality similar to control (hazard ratio not significantly different from 1 on DAI 120, 204, 211).
• IVM-BEPO-fed mosquitoes rarely survived beyond 10 days (median EIP for P. falciparum), implying strong transmission-blocking unless infection precedes ivermectin ingestion by a few days.
• 10-day LC50 (plasma): IVOMEC-D 7.5±0.7 ng mL−1; IVM-BEPO 8.3±0.3 ng mL−1 (ns, t = −1.17, P = 0.23). 10-day LC90: IVOMEC-D 31.8±2.8 ng mL−1; IVM-BEPO 19.2±0.6 ng mL−1 (significant, t = 4.20, P < 0.001).

Population modelling:

• Combining ivermectin-treated cattle with LLINs predicted 25–>90% reductions in mosquito field populations, depending on cattle:human ratio and vector host preference. Long-acting depot consistently reduced infectious vectors by >50% for ≥6 months, peaking at ~95%. Monthly IVOMEC-D yielded cyclic reductions with peaks of 5–72%, with rebounds between doses.

Ecotoxicology and environmental fate:

• Predicted dung concentrations corresponding to mosquito 10-day LC50–LC90 (based on plasma–dung regression) were ~138–319 ng g−1 dw. Observed mean dung concentration across IVM-BEPO samples: 337±185 ng g−1 dw (n=86; mean plasma 19.1±9.2 ng mL−1), overlapping and exceeding EC50 values for many dung fly species and some dung beetles, indicating potential non-target risks.
• Soil sorption (six soils): Kd 55–123 mL g−1 (mean 84.5); Koc 6,630–10,870 mL g−1; sorption positively correlated with Corg and pH (r = 0.86, P < 0.05), indicating strong, largely irreversible sorption and low mobility.
• Dung dissipation under internal storage (first-order kinetics): including day 0 yields k = 0.00145 d−1 (adj. R2 = 0.059), DT50 ≈ 478 d, DT90 ≈ 1,588 d; using days 30–90 yields k = 0.00381 d−1 (adj. R2 = 0.968), DT50 ≈ 182 d, DT90 ≈ 604 d. Concentrations under internal storage decreased significantly over time (e.g., 30 vs 60 days P = 0.02; 30 vs 90 days P = 0.0002). External storage showed different patterns with potential relative enrichment after 30 days due to organic matter loss.
• Example PECsoil_initial for long-acting dosing (2.4 mg kg−1; 211-day pasture scenario) estimated at 4.86 μg kg−1, comparable to pasture PECs reported in literature.

Risk mitigation:

• Proposed RMMs include preventing entry into surface waters; corralling treated animals; maintaining untreated herd fractions (refugia); veterinary monitoring to minimize drug use; dung collection and storage to promote degradation (e.g., fosses fumières); and avoiding repeated application of residue-containing dung without understanding long-term effects.

The findings demonstrate that a single long-acting ivermectin depot injection in cattle maintains mosquitocidal plasma concentrations for at least six months and substantially reduces mosquito survival below the threshold required for Plasmodium falciparum development. Modelling suggests substantial reductions in infectious vector populations, particularly in settings with higher cattle density and animal-feeding tendencies, thus addressing residual transmission not captured by human-targeted interventions. Importantly, the study integrates ecotoxicological evidence, showing that excreted ivermectin at efficacious levels likely impacts non-target dung fauna and that ivermectin exhibits strong sorption and moderate persistence in West African soils and stored dung. This dual perspective underscores that endectocide-based vector control can be effective but requires environmental safeguards. The relevance to the field is twofold: it validates ETL with long-acting formulations as a promising complementary tool for malaria control and establishes an environmental risk framework, including exposure estimation and RMMs, to guide sustainable implementation in agroecosystems.

This work introduces a long-acting injectable ivermectin formulation for cattle that sustains mosquitocidal activity for over six months, meeting WHO efficacy criteria and potentially enhancing malaria control when combined with LLINs. Empirical survival assays and modelling predict substantial reductions in infectious vector populations. Concurrently, the study reveals plausible ecotoxicological risks to dung-associated arthropods and documents strong soil sorption and persistence of ivermectin in dung, emphasizing the need for environmental oversight. The main contributions are: (1) demonstration of durable ivermectin plasma levels and improved vector control potential via a depot formulation; (2) quantitative linkage between plasma and dung residues; (3) first sorption data for Burkina Faso soils with implications for mobility; (4) preliminary risk assessment for non-target fauna; and (5) pragmatic, locally adapted RMMs. Future research should focus on (i) degradation dynamics in manure pits (fosses fumières) and optimized composting, (ii) characterization of West African dung and soil entomofauna and field-based ecotoxicology, (iii) refining PECs and environmental fate under regional climates and farming practices, (iv) monitoring and managing ivermectin resistance in vectors and livestock parasites, and (v) assessing potential plant phytotoxicity under realistic field conditions.

Key limitations include reliance on laboratory survival assays and standard ecotoxicological tests that may not fully capture field complexity; limited data on West African dung fauna sensitivity and community-level effects; uncertainty in translating plasma–dung relationships across breeds, diets, and management systems; potential overestimation of soil sorption due to use of fine-soil fractions; short-term storage studies (≤90 days) with extrapolated dissipation kinetics; climatic variability not fully represented in fate studies; and modelling assumptions regarding vector behavior, host availability, and LLIN use. The study did not assess phytotoxicity in local crops or directly measure field-level ecological impacts under operational ETL deployments, and resistance evolution in mosquitoes was not empirically evaluated.