Susceptibility bioassays
We performed WHO susceptibility bioassays (23) during each trial using F1 progeny of Anopheles gambiae sensu lato (s.l.) collected from the experimental hut site to assess the susceptibility of the Covè vector population to the active ingredients in the ITNs. Mosquitoes were exposed in bottle bioassays to the discriminating concentrations (DCs) of alpha-cypermethrin (12.5 µg) and CFP (100 µg) and in tube tests to filter papers impregnated with the DC of deltamethrin (0.05%) to assess susceptibility to these insecticides. Further exposures were performed with various multiples of the DCs of alpha-cypermethrin and deltamethrin to assess pyrethroid resistance intensity. To assess synergism and the role of P450s in pyrethroid resistance, we also exposed mosquitoes to bottles coated with the DC of alpha-cypermethrin (12.5 µg) with pre-exposure to PBO (400 µg/bottle) and filter papers impregnated with the DC of deltamethrin (0.05%) with pre-exposure to PBO (4%/paper). To prepare test bottles, we prepared stock solutions for each insecticide and dose by dissolving technical grade insecticide in acetone. Bottles were then coated by introducing 1 ml of stock solution into bottles and rotating using a tube roller. Approximately 100 3–5 day-old mosquitoes were exposed to each insecticide and dose for 60 mins in four batches of 25. Concurrent exposures were performed with PBO alone, acetone-coated bottles and silicone oil-impregnated papers as controls. Knockdown was recorded at the end of exposure, after which mosquitoes were transferred to untreated containers, provided access to 10% (w/v) glucose solution and held at 27±2°C and 75±10% relative humidity (RH). Delayed mortality was recorded after 24 h for the alpha-cypermethrin and deltamethrin exposures and every 24 h up to 72 h for CFP.
Experimental hut trials
The overall aim of this study was to evaluate the entomological impact of combining pyrethroid-CFP and pyrethroid-PBO ITNs inside the same household. Experimental hut trials are standardised simulations of human-occupied housing recommended by WHO for the evaluation of indoor vector control interventions. Mosquitoes enter huts at night and interact ad-libitum with the human host and vector control intervention(s) contained inside. In the morning on each day of the trial, mosquitoes are then collected from each hut and scored for entomological outcomes correlated with epidemiological impact (24), notably mortality and blood-feeding. Compared to other test designs, experimental huts therefore, provide a more realistic representation of a household setting with a compact design allowing for natural interactions of wild, free-flying mosquitoes with two nets.
Study site and experimental huts
Both experimental hut trials were performed at CREC/LSHTM field station in Covè, southern Benin (7°14′N2°18′E). The huts are located in a vast area of rice irrigation that provides permanent and extensive mosquito breeding sites. An. coluzzii and An. gambiae sensu stricto (s.s.) occur sympatrically with the former predominating particularly in the rainy season. The Covè vector population exhibits a high frequency (≥90%) and intensity (200-fold) of pyrethroid and organochlorine resistance driven by the presence of the knockdown resistance (kdr) L1014F mutation and overexpression of P450s (25). It remains susceptible to other unrelated insecticides including CFP (26). Experimental huts used were of standard West African design but with larger dimensions (15 m3) to accommodate two sleeping areas. They were constructed from concrete bricks with cement-plastered walls, a corrugated iron roof and a polyethylene ceiling. Mosquitoes entered via four window slits with a 1 cm opening positioned on two sides of the hut. A wooden-framed veranda projected from the rear wall of each hut to capture exiting mosquitoes. Huts were surrounded by a water-filled moat to preclude mosquito predators.
Experimental hut treatments
We performed two experimental hut trials to evaluate the entomological impact of different types of pyrethroid-CFP ITN and pyrethroid-PBO ITN applied alone and in combination inside the same household. Both trials consisted of two components. The first component compared the efficacy of each ITN type applied singly while the second compared different combinations of ITNs applied together inside the same experimental hut.
Trial 1 evaluated the impact of combining the alpha-cypermethrin-based ITNs; Interceptor® G2 (BASF) and DuraNet® Plus (Shobikaa Impex). Comparison was also made to a combination of Interceptor® G2 and Interceptor®(BASF), an alpha-cypermethrin based pyrethroid-only ITN, to control for confounding effects of the pyrethroid. The following treatments were tested in trial 1:
- Untreated net (control I)
- Interceptor®
- DuraNet® Plus
- Interceptor® G2
- Untreated net + Untreated net (control II)
- Interceptor® + Interceptor®
- DuraNet® Plus + DuraNet® Plus
- Interceptor® G2 + Interceptor® G2
- Interceptor® G2 + Interceptor®
- Interceptor® G2 + DuraNet® Plus
Trial 2 evaluated the impact of combining the deltamethrin-based ITNs; PermaNet® Dual and PermaNet® 3.0 (Vestergaard Sàrl). Interceptor® G2 was also included as a single net treatment arm in trial 2 to compare its performance to PermaNet® Dual. The following treatments were tested in trial 2:
- Untreated net (control I)
- PermaNet® 3.0
- Interceptor® G2
- PermaNet® Dual
- Untreated net + Untreated net (control II)
- PermaNet® 3.0 + PermaNet® 3.0
- PermaNet® Dual + PermaNet® Dual
- PermaNet® Dual + PermaNet® 3.0
ITN characteristics and preparation
- Interceptor® is a WHO-prequalified pyrethroid-only ITN made of polyester filaments coated with 5 g/kg of alpha-cypermethrin.
- DuraNet® Plus is a WHO-prequalified pyrethroid-PBO ITN made of polyethylene monofilament incorporated with 6 g/kg of alpha-cypermethrin and 2.2 g/kg of PBO.
- Interceptor® G2 is a WHO-prequalified pyrethroid-CFP ITN made of polyester filaments coated with 2.4 g/kg of alpha-cypermethrin and 4.8 g/kg of CFP.
- PermaNet® 3.0 is a WHO-prequalified pyrethroid-PBO ITN which consists of polyester side panels coated with 2.1 g/kg of deltamethrin and a polyethylene roof panel incorporated with 4 g/kg of deltamethrin and 25 g/kg of PBO.
- PermaNet® Dual is a WHO-prequalified pyrethroid-CFP ITN made of polyester filaments coated with 2.1 g/kg of deltamethrin and 5 g/kg of CFP.
We selected 6 replicate nets for each type per treatment arm for the experimental hut trials and rotated them within treatments daily. Bed nets were erected over sleeping areas inside huts by tying the corners of the roof panel to nails positioned on the uppermost sides of hut walls. In huts containing single nets, nets were hung in the centre of the room while in huts containing two nets, the hut room was divided vertically into two equally sized sleeping areas with nets hung on either side. All nets were given 6 holes each measuring 4 x 4 cm to mimic wear-and-tear from routine use.
Treatment and sleeper rotation
In both trials, we rotated treatments between experimental huts weekly according to randomised Latin square designs (LSDs) to mitigate bias due to differences in positional attractiveness of experimental huts. For net combination arms, we also rotated nets between sleeping areas inside huts to reduce bias due to mosquito entry point preference. Human volunteers were recruited to sleep in experimental huts between 21:00 and 06:00 to attract wild, free-flying mosquitoes. Volunteers were randomly assigned to sleep alone in single net treatments or as pairs in double net treatments. Single sleepers and sleeper pairs were rotated according to separate LSDs to mitigate bias due to individual attractiveness to mosquitoes.
Mosquito collections and processing
Each morning, volunteers collected mosquitoes from the different hut compartments (under the net, room, veranda) and deposited them in labelled plastic cups. Mosquito collections were then transferred to the field laboratory for morphological identification and scoring of immediate mortality and blood-feeding. Surviving, female An. gambiae s.l. were provided access to 10% (w/v) glucose solution and delayed mortality was recorded every 24 h up to 72 h after collection for all treatments. Mosquito collections were performed 6 days per week and on the 7th day, huts were cleaned to prevent contamination before the next rotation cycle. In both trials, mosquito collections continued for two full treatment rotations equating to 20 weeks for trial 1 between October, 2021 and March, 2022 and 16 weeks for trial 2 between May and October, 2022.
Experimental hut trial outcome measures
The efficacy of the experimental hut treatments was expressed in terms of the following outcome measures:
- Hut entry – number of mosquitoes collected in experimental huts
- Deterrence (%) – reduction in the number of mosquitoes collected in the treated hut relative to the untreated control hut. Calculated as follows:
Where Tu is the number of mosquitoes collected in the untreated control hut and Tt is the number of mosquitoes collected in the treated hut.
- Exophily (%) – exiting rates due to potential irritant effects of a treatment expressed as the proportion of mosquitoes collected in the veranda
- Inside net (%) – proportion of mosquitoes collected inside net
- Blood-feeding (%) – proportion of blood-fed mosquitoes
- Blood-feeding inhibition (%) – proportional reduction in blood-feeding in the treated hut relative to the untreated control hut. Calculated as follows:
Where Bfu is the proportion of blood-fed mosquitoes in the untreated control hut and Bft is the proportion of blood-fed mosquitoes in the treated hut.
- Personal protection (%) – reduction in the number of blood-fed mosquitoes in the treated hut relative to the untreated control hut. Calculated as follows:
Where Bu is the number of blood-fed mosquitoes in the untreated control hut and Bt is the number of blood-fed mosquitoes in the treated hut.
- Delayed mortality (%) – proportion of dead mosquitoes observed every 24 h up to 72 h after collection
- Overall killing effect (%) – number of mosquitoes killed in the treated hut relative to the number collected in the untreated control hut. Calculated as follows:
Where Kt is the number of dead mosquitoes in the treated hut, Ku is the number of dead mosquitoes in the untreated control hut and Tu is the number of mosquitoes collected in the untreated control hut.
Chlorfenapyr and PBO interaction bioassays
The entomological performance of pyrethroid-CFP ITNs when combined with pyrethroid-PBO ITNs inside experimental huts may also be influenced by the behaviour of the wild vector population. We therefore, performed bottle bioassays to assess the impact of PBO pre-exposure on CFP toxicity. By eliminating confounding effects associated with mosquito behaviour, these bioassays were used to investigate interactions between CFP and PBO and thus help explain findings from the experimental hut trials. The bioassays were performed with the pyrethroid-resistant An. gambiae s.l. Covè strain which are F1 progeny of wild mosquitoes collected from the experimental hut site. Test bottles were coated with the DC of PBO (400 µg), and 0.25x (25 µg), 0.5 (50 µg), 0.75x (75 µg) and 1x (100 µg the DC of CFP as previously described. Cohorts of approximately 150, 3–5 day-old mosquitoes were subsequently exposed to each dose for 60 mins with and without pre-exposure to the discriminating dose of PBO (400 µg) in 6 replicates of 25. Mosquitoes were then transferred to untreated containers with access to 10% (w/v) glucose solution and held at 27±2°C and 75±10% RH. Knockdown was recorded at the end of exposure and delayed mortality every 24 h up to 72 h.
Preparation of net pieces for bioassays and chemical analysis
In each trial, 5 net pieces (one from each panel) measuring 30 x 30 cm were cut from 1 new, unused net and 2 nets used in experimental huts for each ITN type for laboratory bioassays and chemical analysis. Because of the mosaic design of PermaNet® 3.0, two additional net pieces were cut from the roof panel to provide 7 pieces in total and ensure appropriate representation of PBO-incorporated pieces as per WHO specifications (27). Net pieces were wrapped in labelled aluminium foil and stored at 30°C before and during use in supplementary tunnel tests. Following use in tunnel tests, net pieces were stored at 4°C before being sent for chemical analysis.
Supplementary tunnel tests
The inappropriateness of cone bioassays for evaluating the efficacy of pyrethroid-CFP ITNs is well-documented (28). We, therefore, performed tunnel tests with two net pieces randomly selected from those cut from ITNs before and after the hut trials to provide supplementary ITN efficacy data. The tunnel tests were performed with the susceptible An. gambiae s.s. Kisumu strain to assess the pyrethroid component of the ITNs and the pyrethroid-resistant An. gambiae s.l. Covè to assess the additional impact of the CFP and PBO components.
- An. gambiae s.s. Kisumu strain is an insecticide-susceptible reference strain originated from Kisumu, western Kenya.
- An. gambiae s.l. Covè strain are F1 progeny of mosquitoes collected from the experimental hut site in Covè, southern Benin. This strain is highly resistant to pyrethroids and organochlorines but susceptible to other insecticide classes including CFP. Resistance is mediated by the kdr L1014F mutation and overexpression of cytochrome P450 monooxygenases (25).
Tunnel tests are an experimental chamber used to evaluate the efficacy of ITNs which simulate the behavioural interactions that occur between free-flying mosquitoes and nets during host-seeking. The design consists of a square glass tunnel divided at one third its length by a wooden frame fitted with a net piece. In the shorter section of the tunnel, a guinea pig bait was held in an open-meshed cage while in the longer section, approximately 100, 5-8 day old mosquitoes were released at dusk and left overnight. Net pieces were given 9 holes measuring 1 cm in diameter to facilitate entry into the baited chamber. In the morning, mosquitoes were collected from the tunnel and scored for immediate mortality and blood-feeding. Live mosquitoes were transferred to labelled plastic cups, provided access to 10% (w/v) glucose solution and held at 27±2°C and 75±10% RH. Delayed mortality was recorded every 24 h up to 72 after exposure. Mosquitoes were concurrently exposed to untreated net pieces as a negative control.
Chemical analysis of ITNs
All net pieces cut from ITNs of each type before and after the experimental hut trials were sent to accredited analytical laboratories to confirm that the chemical contents of all ITNs fell within WHO tolerance thresholds (±25%) before use in experimental huts and assess how this changed after the hut trial. The methods used for analysis of chemical content have been described in a previous publication (29). The results confirmed that the chemical contents of the ITNs were within WHO tolerance thresholds (±25%) before the hut trials except for the PBO content in DuraNet® Plus pieces (+36%). Chemical content of all active ingredients changed slightly after the hut trials but was still within the ±25% threshold. We provide detailed chemical analysis results as supplementary information (Table S1).
Data analysis
For experimental hut trial data, we analysed differences between treatments for proportional binary outcomes (mortality, blood-feeding, exophily) using blocked logistic regression and differences between count outcomes (entry) using negative binomial regression. We fitted separate models for each outcome and adjusted for variation associated with the different huts, sleepers/sleeper pairs and weeks of the trial. These analyses were performed in Stata version 17. Insecticide resistance bioassay data was interpreted according to WHO criteria (23) while interaction bioassay and tunnel test results were plotted on graphs to visualise differences between treatments for key outcomes.
Ethical considerations
We received approval for the conduct of the trials from the Research Ethics Committees of the Benin Ministry of Health (N°133, 17/11/2021) and the London School of Hygiene & Tropical Medicine (LSHTM) (Ref: 26429). All participants provided written informed consent prior to the study onset. We offered participants a free course of chemoprophylaxis spanning the duration of the study and up to 4 weeks following its completion to mitigate malaria risk. A stand-by nurse was also available to assess any febrile illness or adverse reactions to the test items. LSHTM Animal Welfare Ethics Review Board issued approval for the use of guinea pigs for mosquito blood-feeding and tunnel tests (Ref: 2020-01A). We maintained guinea pig colonies according to protocols developed in line with international guidelines governing the use of animals for scientific research.