Background

Insecticide-treated nets (ITNs) remain the key preventive tool for sustainable malaria control in endemic communities [1, 2]. Historically, all ITNs contained a single class of pyrethroid insecticide. However, the rapid increase in resistance to this class of insecticide in African malaria vectors [3] has driven the development and deployment of new types of ITNs that contain a second active ingredient in addition to the pyrethroid [4], referred to as dual-active-ingredient insecticidal nets. ITNs containing both pyrethroids and piperonyl butoxide (PBO; an insecticide synergist that primarily acts by inhibiting the action of oxidase enzymes in mosquito vectors) have been shown to be effective against pyrethroid-resistant Anopheles gambiae sensu lato (An. gambiae s.l.) and Anopheles funestus mosquitoes [5]. Recent large cluster randomised trials in Uganda [6] and Tanzania [7] showed that the pyrethroid-PBO nets Olyset Plus (Sumitomo Chemical, Tokyo, Japan) and PermaNet 3.0 (Vestergaard Frandsen, Lausanne, Switzerland) confer better protection against malaria than pyrethroid-only nets, leading the WHO to recommend pyrethroid-PBO nets in areas where the main malaria vectors exhibit intermediate levels of resistance mediated by the monooxygenase-based pyrethroid resistance mechanism [4]. However, there is still uncertainty regarding whether pyrethroid-PBO ITNs will perform better than pyrethroid-only ITNs against mosquito populations with extremely high levels of pyrethroid resistance or when other mechanisms, such as glutathione S-transferases (GST)-based metabolic resistance, are driving the insecticide resistance phenotype [8].

Two new types of dual-active ingredient ITNs were recently evaluated in clinical trials in areas with pyrethroid-resistant vectors. The epidemiological efficacy of an ITN combining pyrethroids and an insect growth regulator (pyriproxyfen) was evaluated in Burkina Faso, Tanzania and Benin [9,10,11]. Whilst the Olyset Duo ITN (containing permethrin + pyriproxyfen; Sumitomo Chemical) showed moderate clinical protection against malaria when compared to pyrethroid-only nets in Burkina Faso [9], the Royal Guard ITN (RG; Disease Control Technologies, Greer, SC, USA), which contains alphacypermethrin + pyriproxyfen, was not superior to standard ITNs in trials in Benin or Tanzania [10, 11]. In contrast, the Interceptor G2 ITN (IG2; BASF AGRO B.V./BASF Nederland B.V., Arnhem, The Netherlands; BASF, Ludwigshafen, Germany), containing pyrethroid + the pyrrole class pro-insecticide chlorfenapyr, showed significantly increased protective efficacy against malaria in both the Benin and Tanzania trials [4]. Consequently, the WHO has recently recommended the use of pyrethroid-chlorfenapyr-based nets against resistant malaria vectors []. Extrapolating from the results of clinical trials to predict the entomological efficacy of these next-generation ITNs in different epidemiological and ecological settings requires an understanding of the local mosquito population responses to different ITNs. Given the resources and time required for large-scale randomised controlled trials, repeating these in all settings is not practical. Malaria transmission models that were parameterised with data from semi-field studies (experimental hut trials [EHTs]) have been shown to be useful proxies for predicting the epidemiological efficacy of pyrethroid-only and pyrethroid-pyrrole ITNs [12, 13].

We performed EHTs to evaluate the performance of dual-active-ingredient ITNs compared to standard pyrethroid-only nets against pyrethroid-resistant An. gambiae s.l. mosquitoes from Za-Kpota, southern Benin. We also assessed the sub-lethal impacts of these new types of ITNs on the mosquitoes’ reproductive success and longevity.

Methods

Study site

Experimental hut trials were carried out at the Tropical Infectious Diseases Research Centre (TIDRC) field station in Za-Kpota district (7°10′58.4ʺN, 2°17′15.3ʺE), southern Benin. TIDRC is in a subtropical climate zone with two wet seasons (March-July and October–November) and two dry seasons (December-March and August–September). Monthly rainfall during the periods when the experiments were performed (July to August 2021 and October to November 2021) ranged from 1000 mm to 1200 mm (Agence pour la Sécurité et la Navigation Aérienne [ASECNA] of Benin), the mean relative humidity (RH) was 71% ± 5% and daily temperatures ranged from 29 °C to 33 °C. The main malaria vectors in the study area are pyrethroid-resistant An. gambiae s.l. [14].

Insecticide susceptibility

Anopheles gambiae s.l. larvae and pupae were collected from breeding sites near the experimental hut station at Za-Kpota in July and September 2021 using previously described methods [15] and transported to the insectary of the Tropical Infectious Diseases Research Centre, University of Abomey-Calavi, where they were reared until the adult stage for bioassays. Mosquitoes were maintained under standard insectary conditions of 70 ± 8% RH and 27 ± 2 °C ambient temperature.

To assess the susceptibility of An. gambiae s.l. from Za-Kpota to the active ingredients in the ITNs to be tested, namely the Olyset Plus (permethrin + PBO), PermaNet 3.0 (deltamethrin + PBO) and Interceptor G2 (alpha-cypermethrin + chlorfenapyr) ITNs, we performed the WHO tube assay and bottle tests. WHO tube tests were carried out on 3- to 5-day-old F0 An. gambiae s.l. adults according to the WHO protocol [16]. In brief, mosquitoes were exposed to filter papers impregnated with 0.05% alpha-cypermethrin, 0.75% permethrin or 0.05% deltamethrin (pyrethroids). Non-impregnated filter paper was also used as the control. Mortality was recorded 24 h after exposure to each insecticide. The insecticide-treated filter papers were obtained from the WHO via the Liverpool School of Tropical Medicine (LSTM), and their quality was assessed against susceptible An. gambiae sensu stricto (An. gambiae s.s.) mosquitoes (Kisumu strain). A PBO synergist test was also performed: mosquitoes were pre-exposed to 4% PBO and then exposed to 0.05% deltamethrin [16].

Bottle bioassays were performed using the discriminating concentration of chlorfenapyr (100 µg/bottle) in accordance with the WHO protocol [16], and mortality was recorded at 72 h post-exposure. The Anopheles species among the An. gambiae s.l. complex used in the susceptibility tests were identified using An. gambiae species-specific PCR as either An. gambiae s.s., An. coluzzii or An. arabiensis [17].

Description of treatments used in the EHTs

Two rounds of EHTs were performed (36 collection nights per round) during the long rainy season from July to August 2021 and during the short rainy season from September to November 2021. Nine different treatments were assessed over both EHTs, with untreated nets used as a negative control. To test the impact of net age, aged nets (PermaNet 3.0; Interceptor G2; Interceptor) were collected from households in the Cascades region of Burkina Faso 2 years post-distribution. All of the nets were 180 cm long × 170 cm wide × 170 high; the chemical and fabric specifications of each of the nets are described in Table 1. Prior to use in the experimental huts all nets were aired in the shade for 24 h.

Table 1 Specifications of insecticide-treated net products assessed in experimental hut trials
Full size table

EHTs procedure

The experimental huts used for the open eaves’ experiments were typical of the West African model []. Briefly, the huts were made of concrete bricks, with a corrugated iron roof, ceilings lined with palm thatch and cement plaster on the inside. Each hut was elevated on concrete plinths surrounded by water-filled moats to prevent the entry of mosquito predators and equipped with veranda traps to capture exiting mosquitoes. Mosquitoes enter the hut through four 1-cm-wide window slits that are located on three sides of the hut. Prior to the experiment, an awareness session was held with potential study volunteers (sleepers), also to obtain their consent; only consenting volunteers were included. Treatments were allocated to the experimental huts on a weekly basis, and volunteer sleepers were rotated daily between huts using a randomised Latin square design [18]. To simulate a worn net, each net was deliberately holed with six 4 × 4-cm holes (2 holes on each side and 1 hole on each end) according to WHO protocol [18]. To attract free-flying mosquitoes, volunteer sleepers slept in the experimental huts between 20:00 hours and 05:30 hours. Each morning, alive and dead mosquitoes were collected from the different compartments inside the hut [19, 20]. All live female An. gambiae s.l. were given access to a 10% honey solution and delayed mortality was recorded after 72 h for all of the treatments.

Surviving blood-fed mosquitoes were placed individually in egg-laying chambers, consisting of plastic cups fitted with an untreated net and containing approximately 50 ml of dechlorinated tap water. The chambers were monitored daily for egg laying, and the number of eggs laid by each female mosquito was recorded. A pinch of larval food (TetraMin® Baby Fish Food; Tetra GmbH, Melle, Germany) was added to each chamber containing eggs, and the number of larvae hatching was recorded and monitored until adult emergence. All live mosquitoes collected in the experimental huts were monitored until death to assess their longevity. A sample of the collected female An. gambiae s.l. were subjected to molecular species identification. Briefly, genomic DNA of mosquitoes was extracted using the 2% cetyl trimethyl ammonium bromide (CTAB) method. Species of the An. gambiae complex in the sample set were identified using species-specific PCR [17].

Overall, the outcomes from the EHTs were: (i) deterrence (the proportional reduction of mosquito entry into huts with ITNs relative to huts with untreated nets; (ii) 72-h mortality rate (the number of mosquitoes dead after 72 h as a proportion of the total numbers entering the experimental huts with that treatment); (iii) exophily rate (estimated as the number of mosquitoes collected from the verandas as a proportion of all mosquitoes collected in the given experimental hut); (iv) blood-feeding rate (estimated as the proportion of mosquitoes collected that had blood fed for each experimental hut); (v) oviposition rate (estimated as the proportion of surviving blood-fed mosquitoes that laid eggs); (vi) fecundity (the mean number of laid eggs per surviving blood-fed mosquito); (vii) fertility (the mean number of emerged adults per surviving blood-fed mosquito); and (viii) longevity (number of days alive mosquitoes survive after collection from experimental huts).

Data analysis

Data were recorded in specifically designed forms, entered into Microsoft Excel (Microsoft Corp., Redmond, WA, USA) and analysed using R statistical software version 4.3.1 [21]. Susceptibility test results were interpreted following the WHO criteria [16]. Mosquito populations were considered to be susceptible to an insecticide if the mean mortality was ≥ 98% and resistant if the mean mortality was ≤ 90% mortality. Since no mortality was recorded in controls, Abbott’s formula was not necessary to correct the mortality rates.

Outcomes from the EHTs were analysed. Deterrence was modelled using a generalised linear mixed model (GLMM) with a log link and a negative binomial distribution. Proportional data were analysed using GLMMs with a logit link and a binomial distribution (mortality 72 h) or beta-binomial distribution (exophily rate, blood-feeding rate and oviposition rate). All GLMMs were performed using the lme4 R package [22]. Fecundity and fertility were described using descriptive statistics. All models were fitted considering sleepers, huts and weeks as random effects variables while treatment was considered as a fixed effect variable. For the subsequent analysis, mosquitoes from the two trials were pooled. Taking into account the variability that could be induced by the two rounds of EHTs performed, several GLMMs were run for each parameter, systematically including "round" as a fixed effect. The final models were selected without a round included in the model according to the Akaike information criterion (AIC). In addition, we evaluated model fit by performing a quantile test, uniformity test and dispersion test using the DHARMa R package [23].

Three levels of comparison were determined: (i) untreated nets (UTNs) versus ITNs; (ii) standard net PermaNet 2.0 (P2) versus ITNs; and (iii) ITNs versus aged ITNs. The effects of treatment, physiological status and collection location on mosquito survivorship were analysed using weighted Cox regression to generate unbiased averaged hazard ratios (HR) and their corresponding 95% confidence intervals (CIs) since the proportional hazard assumption was violated [24]. This analysis was performed using the Coxphw R package with date as a cluster [25].

Results

Susceptibility of Anopheles gambiae s.l. to insecticides at Za-Kpota

Pyrethroid resistance was observed in field An. gambiae mosquitoes subjected to susceptibility tests. Of the 700 An. gambiae s.l. adult female mosquitoes exposed, 72.9% and 27.1% were identified as An. coluzzii and An. gambiae s.s., respectively, from the June 2021 collections. Regarding the adults obtaining from the September 2021 larvae collections, mosquito populations comprised 71.4% An. coluzzii and 28.6% An. gambiae s.s. Following exposure to discriminating doses, the 24-h mortality rates of An. coluzzii and An. gambiae s.s. were 17.5% and 28.3%, respectively, for permethrin, and 62.9% and 56.8%, respectively, for deltamethrin (Table 2). Twenty-four-hour mortality rates of 90.9% (An. coluzzii) and 93.8% (An. gambiae s.s.) were recorded with discriminating doses of alpha-cypermethrin. There were no significant differences in 24-h mortality between An. gambiae s.s. and An. coluzzii (df = 1, χ2 = 0.43, P = 0.508). When pre-exposure to PBO was followed by exposure to the discriminating dose of deltamethrin, the mortality rate reached 100% for both An. coluzzii and An. gambiae s.s. Both An. coluzzii and An. gambiae s.s. were susceptible to chlorfenapyr, with 72-h mortality rates of 100% and 98.07%, respectively (Table 2). All tested insecticides induced 100% mortality rate in the laboratory-maintained susceptible An. gambiae s.s. Kisumu strain.

Table 2 Insecticide susceptibility data recorded according to WHO and CDC bottle methods against field F0 female Anopheles gambiae sensu lato mosquitoes collected in July and September 2021
Full size table

Experimental hut results

Mosquito abundance and species identification

Overall, a total of 9353 mosquitoes comprising 593 An. gambiae s.l. (6.3%), 360 Culex spp. (3.8%) and 8400 Mansonia spp. (89.81%) were collected during the short rainy season trial. For the long rainy season trial, 4959 mosquitoes comprising 1751 An. gambiae s.l. (35.3%), 81 Culex spp. (1.63%) and 3127 Mansonia spp. (63.1%) were collected. Of the 2344 free-flying An. gambiae s.l. females collected during the 72 nights of collection (36 nights for each trial round), 786 were sampled randomly from both the ETHs and identified to species level by molecular techniques. The predominant species was An. gambiae s.s. (99.4%). Overall, the collected An. gambiae s.s. mosquitoes were mostly found as unfed mosquitoes (61.4%) (Figs. 1, 2).

Fig. 1
figure 1

Distribution of collected Anopheles gambiae sensu lato mosquitoes according to their physiological status in each treatment from the experimental hut trials during the short rainy season from September to November 2021. Histograms represent the number by collection day, and pie charts represent the overall proportions. a Untreated net, b Interceptor G2 net, c Interceptor G2 Aged net, d PermaNet 3.0 net, e PermaNet 3.0 Aged net, f PermaNet 3.0 net

Full size image
Fig. 2
figure 2

Distribution of collected Anopheles gambiae sensu lato mosquitoes according to their physiological status in each treatment from experimental hut trials during the long rainy season from July to August 2021. Bar charts represent the number by collection day, and pie charts represent the overall proportions. a Untreated net, b Interceptor net, c Interceptor Aged net, d Interceptor G2 net, e Royal Guard net, f Royal guard Used net

Full size image

Anopheles gambiae s.s. entry into and exiting from huts

None of the ITNs tested significantly deterred mosquito entry into the huts relative to the UTN (Table 3). A significantly higher exophily was recorded with the Interceptor G2 net (IG2) (odds ratio [OR] = 2.1, P < 0.0013) and Royal Guard net (OR = 1.6, P < 0.04) compared to that of the UTN (Table 4).

Table 3 Comparative odds ratio of deterrence by contrasting treatments
Full size table
Table 4 Comparative odds ratio of exophily rate by contrasting treatments
Full size table

Blood-feeding rates of wild An. gambiae s.s.

The proportion of blood-fed An. gambiae s.s. in the experimental huts with UTNs was 14.6%, compared to 0.5–19.5% blood-fed An. gambiae s.s. in the experimental huts with ITNs (Table 5). The proportion of blood-fed An. gambiae s.s. decreased significantly in the huts equipped with Royal Guard (RG; OR = 0.05, P < 0.001) nets compared to those equipped huts with UTN or the standard PermaNet 2.0 net (P2) (OR = 0.07, P = 0.0021). No significant inhibition of blood feeding was observed in huts equipped with Interceptor, Interceptor Aged, Interceptor G2 (IG2), Interceptor G2 Aged (IG2 Aged), PermaNet 3.0 (P3) and PermaNet 3.0 Aged (P3 Aged) nets compared those with UTNs or the standard net P2. There was no significant difference in blood feeding between huts with ITNs and those with aged ITNs in any pairwise comparison (Table 5).

Table 5 Blood-feeding rate by treatment and comparative odds ratio by contrasting treatments
Full size table

Seventy-two-hour mortality rates of wild An. gambiae s.s.

The 72-h mortality of An. gambiae s.s. was 44.3% in the experimental huts with UTNs and ranged from 56.4% to 90.1% in the experimental huts with ITNs. Among the ITNs, the lowest 72-h mortality rate occurred in huts equipped with the standard P2 net (56.04%) and Interceptor net (59.41%) (Table 6). The highest killing effect was recorded in the hut with the IG2 net (OR = 8.9, P < 0.001) compared to the hut with UTNs (Table 6). Similarly, the IG2 net (OR = 5.9, P < 0.001) induced the higher mortality rate when compared to the P2 net. Mortality was also higher in the huts with the P3 Aged net (OR = 2.5, P = 0.01) and RG net (OR = 2.4, P = 0.0011) compared those equipped with the P2 net (Table 6). Surprisingly, the Interceptor Aged net (OR = 2.25, P = 0.011) induced significantly higher mortality than the Interceptor net; conversely, the IG2 Aged net (OR = 0.38, P = 0.006) induced significantly lower mortality than the IG2 net.

Table 6 The 72-h mortality rate by treatment and comparative odds ratio by contrasting treatments
Full size table

Effect of exposure to the ITNs on oviposition, fecundity, and fertility

Only a low number of blood-fed mosquitoes were collected from EHTs and available for follow-up analysis (n = 60 for UTN and n = 105 for all ITNs combined). Oviposition was reduced in mosquitoes collected from huts with ITNs (of any type) compared to the control hut with UTNs (Table 7). No significant difference was observed between oviposition rates in huts with the standard P2 net and the other ITNs (Table 7). Generally, the mean number of eggs laid (fecundity) and the number of adults that emerged (fertility) from surviving blood-fed An. gambiae s.s. mosquitoes were lower in the huts with ITNs than in those with UTNs (Table 7). The mean number of eggs laid per female was 44.8 (95% CI 37.7–51.8) for An. gambiae s.s. mosquitoes collected from huts with UTNs. Lower mean numbers of eggs laid were observed for mosquitoes collected from huts with the IG2 (33.3; 95% CI 21.4–45.8]), P3 (33.4; 95% CI 27.3–39.4) and IG2 Aged (28.8; 95% CI 18.2–39.4]) nets. The mean number of An. gambiae s.s. adults that emerged was 44.7 (95% CI 37.6–51.8) in mosquitoes collected in huts with UTNs, with lower mean numbers of adults emerged for huts equipped with P3 (23.9; 95% CI 18.0–29.8) and Interceptor (18.3; 95% CI 5.6–31.1) nets (Table 7).

Table 7 Oviposition, fecundity and fertility in blood-fed female Anopheles gambiae collected from the experimental hut trials
Full size table

Effect of exposure to the ITNs on longevity

A strong negative effect was observed on the longevity of An. gambiae s.s. after exposure to all of the ITNs. The median survival time of female An. gambiae s.s. mosquitoes collected from huts with UTNs was 9 days, and varied between 1 and 3 days for An. gambiae s.s. exposed to ITNs (Fig. 3). A significantly higher risk of death was observed with all ITNs compared to UTN (Fig. 4). The highest risk was recorded in An. gambiae s.s. exposed to the IG2 (HR = 6.5, 95% CI 5.6–7.6, P < 0.001) and P3 (HR = 4.4, 95% CI 3.4–5.7, P < 0.001) nets (Fig. 4). Compared to mosquitoes collected in the veranda, the risk of death was significantly higher in those collected in the room (HR = 1.8, 95% CI 1.6–1.9, P < 0.001). The risk of death was also higher in gravid mosquitoes (HR = 1.1, 95% CI 1–1.2, P = 0.04) compared to unfed mosquitoes. The risk of death was significantly lower in fed mosquitoes (HR = 0.73, 95% CI 0.6–8.9, P < 0.001) than in unfed mosquitoes (Fig. 4).

Fig. 3
figure 3

Kaplan–Meier curves describing survival rates as a function of time in Anopheles gambiae sensu lato mosquitoes collected in experimental hut trials. Vertical dotted lines indicate median survival time. Day 0 corresponds to mosquito collection days in experimental huts. P2, P3, PermaNet 2.0, 3.0, respectively; IG2, Interceptor G2; RG, Royal Guard; UTN, untreated net; see also Table 1

Full size image
Fig. 4
figure 4

Comparative hazard ratio of death according to the treatments, collection location and physiological status in Anopheles gambiae. Error bar represents the 95% confidence interval, P2, P3, PermaNet 2.0, 3.0, respectively; IG2, Interceptor G2; RG, Royal Guard; UTN, untreated net; see also Table 1

Full size image

Discussion

African malaria vectors populations display a high level of pyrethroid resistance [3]. Country-wide surveys over the last decade in Benin have reported that pyrethroid resistance is widespread in malaria vector populations [26, 27]. Investigating tools that can complement or replace existing ones are therefore necessary to strengthen resistance management plans [28]. This study provides key information on the performance of dual-active-ingredient ITNs (Interceptor G2, PermaNet 3.0 and Royal Guard) and their sub-lethal impacts against a pyrethroid-resistant population of An. gambiae s.s. from Za-Kpota in southern Benin using experimental hut trials.

A low mortality rate to pyrethroid insecticides (permethrin and deltamethrin) was recorded for An. gambiae s.s. mosquitoes collected from Za-Kpota using the WHO susceptibility test. This result confirms the presence of resistance to those insecticides, which has also been reported in the neighbouring localities of Za-Kpota in southern Benin [26, 27]. A synergist assay with 4% PBO and deltamethrin was carried out in the Za-Kpota mosquito population, with the results revealing a potential involvement of cytochrome P450 genes in observed phenotypic resistance. This indicates that the pyrethroid resistance observed might also be driven by metabolic resistance mechanisms in the An. gambiae s.s. population of Za-Kpota.

The 72-h mortality rates were considered in the EHTs. The pyrethroid-PBO-based net (PermaNet 3.0 [P3]) displayed higher toxicity against field free-flying pyrethroid-resistant An. gambiae s.s. when compared to the pyrethroid-only based net (PermaNet 2.0 [P2]). The PermaNet 3.0 net is treated with a high concentration of deltamethrin on the side panels (2.1 g/kg) and with 4.0 g/kg deltamethrin + PBO (25 g/kg) on the roof panel. Thus, the high toxicity observed with PermaNet 3.0 could be due to the amount of active ingredients on this net. Given the well-known synergistic effect of PBO on pyrethroid resistance [29] and the data from the WHO susceptibility test, the difference in the killing effect recorded may be due to the partial or complete restoration of susceptibility to deltamethrin, thus increasing the efficacy of PermaNet 3.0 over PermaNet 2.0. The same trend was observed between Olyset and Olyset Plus [30], confirming the improved protective role of PBO nets on malaria prevalence in areas where the resistance phenotype is mainly driven by cytochrome P450 genes [8,9,10,11,12].

Nevertheless, PermaNet 3.0 does not provide total personal protection (induced mortality of < 80%) [29, 31]. These results highlight the possible presence of resistance mechanisms other than metabolic resistance mediated by monooxygenases in the vector population. They also highlight concerns regarding the efficacy of PermaNet 3.0, which have been reported in several previous studies in southern Benin [32, 33], raising questions regarding the true performance of PermaNet 3.0 against resistant malaria vectors where mechanisms such as GST-based metabolic resistance occur [8, 34, 35]. Such concerns have led to more attention being paid to the new dual-active-ingredient ITNs (Interceptor G2 [IG2] and Royal Guard [RG]), each of which contain two products with different modes of action. Interestingly, in the present study, a higher mortality rate was recorded with Interceptor G2 and Royal Guard nets against pyrethroid-resistant An. gambiae s.s. mosquitoes than with the standard pyrethroid-only net PermaNet 2.0.

The highest mortality was observed with the Interceptor G2 net, possible due to the combined toxic effects of alpha-cypermethrin and chlorfenapyr. Chlorfenapyr is activated when the N-ethoxymethyl group is removed by oxidation mediated by some cytochrome P450 enzymes, producing the toxic metabolite tralopyril [36]. Tralopyril disrupts the proton gradient across mitochondrial membranes and impairs ATP production (oxidative phosphorylation), leading to cell death [37, 38]. As demonstrated in this study, cytochrome P450 enzymes are involved in pyrethroid resistance in the An. gambiae s.s. population at Za-Kpota. Thus, overexpression of these enzymes may enhance chlorfenapyr activation [39,40,41], possibly resulting in increased toxicity of nets with insecticide combinations of pyrethroids and chlorfenapyr. A similar observation has been pointed out in several experimental hut trials [42,43,44].

This study also showed that while the Interceptor G2 Aged (IG2 Aged) net induced a high mortality compared to the PermaNet 2.0, it induced significantly lower mortality compared to the Interceptor G2 net (IG2). This could be due to a decrease in the insecticide content of ITNs in operational use over time, as has been demonstrated in several studies [45, 46]. In contrast, Martin et al. [47] demonstrated that after 20 washes (supposed to have mimicked a 36-month-old field net), the Interceptor G2 washed net did not induce a significantly different mortality from the Interceptor G2 unwashed net, confirming the wash resistance of this net. This contrast reveals that the decrease in insecticide content of insecticide-treated nets over time would not be proportional to its reduction after washing. Future new net durability studies should focus more on aged nets, which more accurately capture ageing in the field due to different environmental conditions (temperature, humidity).

In addition to the high toxicity displayed by the three dual-active-ingredient ITNs (Interceptor G2, Royal Guard and PermaNet 3.0), these nets also had irritant properties that enabled them to inhibit blood feeding by pyrethroid-resistant An. gambiae s.s. mosquitoes from Za-Kpota. However, only the Royal Guard net significantly reduced the blood feeding compared to the standard net PermaNet 2.0. Of the ITNs evaluated in this study, the Royal Guard net was treated with the highest concentration of alpha-cypermethrin (216 mg/m2). The significant reduction in blood feeding induced by the Royal Guard net could be due to its high alphacypermethrin concentration. The latter could trigger an avoidance behaviour in the vectors [48], affecting their ability to take a blood meal. As a key parameter influencing malaria transmission potential [49], blood-feeding behaviour observed with new nets could benefit malaria vector control interventions [42,43,44]. This result confirms that high-dose pyrethroids would continue to play a valuable role in blood-feeding inhibition and personal protection [50].

In addition to the standard parameters such as mortality and blood feeding, the sublethal exposure effect could provide complementary information to better appreciate the performance of vector control strategies. These could include reductions in longevity, development rates, feeding, oviposition, fertility, fecundity and changes in sex ratio or behaviour [51]. Several studies have shown that insecticide resistance mechanisms can negatively affect the reproductive fitness and longevity of Anopheles mosquitoes [52,53,54]. In the present study, we analysed the sub-lethal impact of ITN exposure in collected An. gambiae s.s. mosquitoes using oviposition, fecundity, fertility and longevity in which a risk of death has been estimated. We found no significant difference in oviposition rate in collected blood-fed An. gambiae s.s. between standard net PermaNet 2.0 and the other ITNs.

Exposure to the dual-active-ingredient nets (Interceptor G2, Interceptor G2 Aged, PermaNet 3.0 and PermaNet 3.0 Aged) reduced fecundity compared to exposure to the standard net PermaNet 2.0. This reduced fecundity may be due to the synergistic action of alphacypermethrin, PBO or chlorfenapyr on fecundity in resistant mosquitoes, possibly as a result of a resource-based trade-off between fecundity and survival [55]. When mosquitoes are exposed to dual-active-ingredient nets, overuse of energy for survival could impact resource availability for fecundity. Such regulation could indicate a high cost of adaptation linked to insecticide resistance [56] which has been described in previous studies [57, 58]. The reduction in fecundity could translate, from an epidemiological point of view, into a decrease in vector density and, hence, a reduction in transmission [59].

In addition, reducing vector longevity is one of the objectives of current insecticide-based malaria control programmes and insecticide resistance management strategies. In this study, the risk of death was high with PermaNet 3.0, Interceptor G2 and Royal Guard nets, with the highest risk of death recorded with the Interceptor G2 net. This result shows once more the efficacy of these nets and, in particular, the combined toxic effects of alpha-cypermethrin and chlorfenapyr. Chlorfenapyr has a delayed impact [60] due to its mode of activation, so it could persist for several days after exposure, thus disrupting several life-history traits of the mosquito, including longevity. The present findings indicate that new nets in the community are likely to impede the developmental cycle of vectors and thus contribute to reducing malaria transmission. These sub-lethal data generated in this study, when added to available data, will undoubtedly contribute to the improvement of parameterised malaria transmission models to predict the long-term efficacy of new ITNs.

Conclusions

Novel, suitable alternative insecticides that can complement pyrethroids and improve control of pyrethroid-resistant malaria vectors are urgently needed to maintain ITNs as a means of malaria control. The dual-active-ingredient nets Royal Guard and Interceptor G2 have shown improved entomological efficacy compared to the standard net PermaNet 2.0 and appear promising for effectively controlling insecticide-resistant malaria vectors. This study provides entomological evidence of the efficacy of these new control tools in a region of southern Benin and suggests that they can be used in this area to control malaria-resistant vectors.