Efficacy of Olyset® Duo, a permethrin and pyriproxyfen mixture net against wild pyrethroid-resistant Anopheles gambiae s.s. from Côte d’Ivoire: an experimental hut trial

Pyrethroid resistance in malaria vectors has spread across sub-Saharan Africa. Alternative tools and molecules are urgently needed for effective vector control. One of the most promising strategies to prevent or delay the development of resistance is to use at least two molecules having unrelated modes of action in combination in the same bed net. We evaluated in experimental huts in Côte d’Ivoire, a new polyethylene long-lasting insecticidal net (LN) product, Olyset® Duo, incorporating permethrin (PER) and pyriproxyfen (PPF), an insect growth regulator (IGR). PPF alone or in combination with permethrin had a significant impact on fertility (7–12% reduction relative to control) and no effect on fecundity of wild multi-resistant An. gambiae s.s. These results triggered crucial research questions on the behaviour of targeted mosquitoes around the LN. To maximize the sterilizing effect of PPF in the combination, there would be a need for a trade-off between the necessary contact time of the insect with PPF and the surface content of the pyrethroid insecticide that is bioavailable and induces excito-repellency.


Introduction
Insecticide-treated bed nets are currently the cornerstone of malaria transmission prevention in Africa. To date, only pyrethroids are used for bed net treatment because they meet the necessary safety criteria [28]. For almost a decade, mass distribution of long-lasting insecticidal net (LNs) has been scaled-up across sub-Saharan Africa with support from the Global Fund to Fight AIDS, Tuberculosis and Malaria [28]. Unfortunately, there are increasing reports of pyrethroid resistance associated with reduced vector mortality and drastic loss of personal protection that pyrethroid-treated nets confer to humans [1,3,15,18]. The operational impact of metabolic-based resistance and target site mutations is still largely overlooked and even controversial [22].
Pyrethroid-resistance is now widespread among malaria vector populations [20,21], and its rapid and significant increase could undermine malaria control [26]. Faced with this threat, alternative tools and molecules are urgently needed for effective vector control. One of the most promising strategies to prevent or delay the development of resistance is to use at least two molecules having unrelated modes of action in combination in the same bed net. The rationale is that mosquitoes resistant to one insecticide should theoretically be killed by the other component. Combination nets are likely to be more effective than standard nets in areas with resistant malaria vectors [2]. Therefore, the use of mixtures is considered as an important strategy in the Global Plan for Insecticide Resistance Management (GPIRM) in malaria vectors [26]. During the last few years, new LNs claimed as resistance breaking products have emerged, some of them impregnated with a combination of a pyrethroid with the synergist pyperonyl butoxide (PBO) and have shown promising results [3,14,19,23,24].
Sumitomo Chemical recently developed a new LN product, so-called Olyset Ò Duo, made of polyethylene incorporating permethrin (PER) and pyriproxyfen (PPF), an insect growth regulator (IGR). PPF is a juvenile hormone analogue (JHA) which inhibits metamorphosis and embryogenesis in several mosquitoes [5] and is already widely used as a larvicide [4]. The idea is that mosquitoes surviving contact with the net, because they are resistant to permethrin, will be sterilized by the PPF after contact with the mixture LN. Nets treated with PPF showed good performance in the laboratory by sterilizing and shortening the longevity of Anopheles gambiae females after tarsal contact [17]. In a twin paper, we also reported insecticidal and sterilizing performances through releaserecapture experiments run in experimental huts in Benin against both susceptible and resistant laboratory strains [6]. The resistant strain used in the later study was homozygous resistant for the kdr L1014F mutation. Promising results in terms of sterilizing effect have been observed among surviving resistant females. It was then necessary to investigate the performances of such LNs against wild free-flying multi-resistant [9,10] An. gambiae s.s. in experimental huts.
The analysis of fecundity and fertility considered pyrethroid-resistant An. gambiae females that blood-fed and survived contact with permethrin on the net surface. In order to obtain enough surviving blood-fed females and fulfil our objectives, a variable consisting of the number of 4 cm · 4 cm holes was introduced in the study: LNs containing 6, 30 or 150 holes were used in the release-recapture study [6]. The results of the study indicated that PPF alone drastically reduced fecundity by 98% and fertility by 93% in laboratory-reared An. gambiae having kdr mutation as a resistance mechanism. More importantly, we showed that a number of holes as low as 6 or 30 in Olyset Ò Duo drastically reduced the mean number of eggs of the pyrethroid-resistant surviving blood-fed females, while the same LN with larger number of holes (150) did not. It is suggested that the number of holes combined with the irritant effect of permethrin limited the contact between the females and the treated surface, reducing mosquito uptake of sufficient PPF.
The challenge of the Phase II trial was to collect enough surviving blood-fed females in the hut with the Olyset Ò Net and Olyset Ò Duo. The objective was to have a sufficient number of surviving blood-fed females without compromising the actual contact time mosquitoes spent probing through the PPFtreated net. This was done to avoid underestimating the LN protective effect. For this reason, the configuration with 30 holes was chosen to be tested in experimental hut in M'Bé, central Côte d'Ivoire against wild free-flying Anopheles gambiae mosquitoes resistant to most current public health insecticides [9]. Its insecticidal and protective efficacy was expressed in terms of deterrence, blood-feeding inhibition, induced exophily and induced mortality. Its sterilizing performance was expressed in terms of reduction of fecundity and fertility of the females that survived treatment exposure.

Materials and methods
Experimental hut trial

Study area
The Phase II trial was carried out in experimental huts located in M'bé (Côte d'Ivoire). The M'bé valley is a huge rice-growing area situated 30 km to the North of Bouaké (5.209963 W and 7. 970241 N) in Central Côte d'Ivoire. Bouaké is a transitional zone characterized by wet savannah. This region has one rainy season (April to October) and an average annual rainfall of 1200 mm and an average annual temperature of 25.8°C. The mosquito population in the area is composed of An. gambiae s.s., An. funestus, Culex sp. and Mansonia sp. Both An. coluzzii and An. gambiae s.s. of An. gambiae s.s. co-exist in sympatry but An. coluzzii is largely dominant (almost 99% of the population) [9]. The An. gambiae s.s. population is resistant to organochlorides, pyrethroids and carbamates with an allelic frequency of the L104F kdr mutation around 30% and the presence of metabolic resistance mechanisms [9].

Design of huts
The huts were made from concrete bricks, with a corrugated iron roof, a ceiling of thick polyethylene sheeting and a concrete base surrounded by a water-filled channel to prevent entry of ants [25]. Mosquito access was via four window slits constructed from pieces of metal, fixed at an angle to create a funnel with a 1 cm wide gap. Mosquitoes fly upward to enter through the gap and downwards to exit, this precludes or greatly limits exodus though the aperture enabling the majority of entering mosquitoes to be accounted for. A single veranda trap made of polyethylene sheeting and screening mesh measuring 2 m long, 1.5 m wide and 1.5 m high, projects from the back wall of each hut. Movement of mosquitoes between hut and veranda is unimpeded during the night.

Study design
The following treatment arms were tested: Before evaluation in experimental huts, the nets (including control) were deliberately holed with 30 holes each of 16 cm 2 (4 cm · 4 cm) to simulate a worn net. Nets were washed three times according to the WHO Pesticide Evaluation Scheme (WHOPES) standard washing methods 1 week before the trial [27].
Adult volunteer sleepers spent six nights per week under the mosquito nets in the experimental huts. They were rotated randomly among huts each night of the study. They entered the hut at dusk and slept until dawn. Once every week, the huts were thoroughly cleaned and aired to avoid contamination.
Each morning, dead and living mosquitoes were collected from the floor of the huts, the veranda traps and inside the nets. Mosquitoes were morphologically identified to species using taxonomic keys and were scored by location as dead or live and as fed or unfed. Live mosquitoes were placed in small netted plastic cups and supplied with 10% honey solution to assess delayed mortality after 24 h.
The primary outcomes were the ones usually measured in experimental hut trial: -deterrence by the treatments (i.e. the reduction in the number of mosquitoes in huts with LNs relative to the control untreated net); -induced exophily (IE) (i.e. the proportion of mosquitoes found in exit traps in huts with LNs relative to the total collected in hut with untreated net); -blood-feeding inhibition (BFI) (i.e. the reduction in bloodfeeding of mosquitoes in huts with LNs relative to the hut with control untreated net); -immediate and delayed mortality (i.e. the proportion of dead mosquitoes at the time of collection in the morning and after 24 h holding, corrected from untreated control mortality).

Reporting of adverse events
The volunteers were asked to report any adverse events associated with use of nets and a provision for medical care was made.

Fecundity and fertility assessment
Alive blood-fed females collected from all compartments of the huts were counted. All the blood-fed females per hut were put by batch of up to 20 females in cardboard cups (450 mL). We previously filled the bottom of the cardboard cups with 1 cm high layer of wetted cotton, covered with a filter paper disc to allow females to lay their eggs. Females in cardboard cups were maintained with honey solution at 28°C and 80% RH for 5 days before checking mortality. A picture of each filter paper was taken to count the eggs using egg counter software [13]. All eggs from the control and treated batches were immersed in water. After 6 days, the number of larvae was checked to determine the hatching rates. Because pyriproxyfen acts by sterilizing the adult female mosquito, the impact of the treatments on the reproduction of surviving blood-fed mosquitoes was investigated by detecting whether there was a reduction in fecundity (number of eggs per female) and fertility (proportion of laid eggs hatching) of these mosquitoes compared to the control.

LN bioassays
Standard WHO cone bioassays were used to determine bioefficacy of LNs against a susceptible laboratory-reared Anopheles gambiae Kisumu strain [27]. Four nets (one per treatment arm) were bioassayed at time 0 (i.e. the day before the first washing). Bioassays were carried out for a second time after washings and then for a third time at the end of the field trial with nets used in the huts. For each net, five cones were placed on the five sections of the net (roof and four sides). Ten unfed Kisumu females, 2-3 days old, were introduced per cone and exposed for 3 min to each. Knockdown (KD) was checked 60 min after exposure and mortality was recorded 24 h after exposure. Mosquitoes exposed to untreated nets were used as controls and Abbott's adjustment applied if mortality was >5% for the controls.

Statistical analysis
The software ''R'' was used for the statistical analyses [20]. The proportion of mosquitoes that exited early, successfully blood-fed or died was analysed using a logistic regression model. The ''brglm'' function from the brglm package was used for the analysis [11]. It enables fitting of binomialresponse regression models using the bias-reduction method developed in Firth [7]. These procedures return estimates with improved frequentist properties (bias, mean squared error) that are always finite even in cases where the maximum likelihood estimates are infinite (data separation). The number of collected mosquitoes entering the huts was analysed using negative binomial regression. Mortality and KD rates from WHO cone bioassays were compared between each net using the v 2 test. For statistical testing the level of significance was set at 5%.

Ethical considerations
Ethical approval was obtained from the Ministry of Health and Fight against AIDS in Côte d'Ivoire through the National Ethics Committee. Adult volunteers were recruited among the inhabitants of the villages close to the site. Volunteer sleepers were recruited after obtaining informed written consent. Medical supervision was provided throughout the trial and for an additional one month after the end of the study by a qualified medical doctor. Confirmed malaria cases were treated with ''artesunate + amodiaquine'', according to national policies, and all the volunteer sleepers were vaccinated against yellow fever after enrolment.

Results
Bio-efficacy of the treated nets (WHO cone bioassays) Table 1 shows the bio-efficacy of each comparison arm in terms of KD effect and mortality before washing, after three washes and after the field trial. Before washing, KD rates recorded with the An. gambiae s.s. susceptible Kisumu strain were 100% for both Olyset Ò Net and Olyset Ò Duo. Olyset Ò Duo caused 100% mortality whereas Olyset Ò Net induced significantly lower mortality (22% mortality; p < 0.0001). The control and PPF nets did not induce any KD effect or mortality.
After washing and prior to the field trial, the control and PPF nets did not induce any KD effect or mortality. With Olyset Ò Net, KD and mortality decreased significantly (p < 0.05) from before washing. With Olyset Ò Duo, only mortality decreased significantly (p < 0.001) after washing.
After the field trial, KD rate increased significantly (p < 0.05) for Olyset Ò Net, but remained stable for Olyset Ò Duo (p > 0.05). Mortality increased significantly (p < 0.05) for Olyset Ò Duo and Olyset Ò Net (p < 0.05) from the level observed just after washing. The PPF-treated nets induced neither KD effect nor mortality (Table 1). Overall, Olyset Ò Duo outperformed the other treatments before and after the field trial.

Insecticidal efficacy of treatments during the field trial
Mosquito collection in experimental huts was carried out over 24 nights between October 8 and November 3, 2012. In total, 5056 An. gambiae s.l. and 3171 other mosquitoes were caught. Of the other mosquitoes caught, 96.6% were Mansonia spp., 1.4% were Culex spp., 0.4% were Aedes spp. and 1.8% were Anopheles species other than An. gambiae s.s. The results in terms of deterrence, induced exophily, blood-feeding inhibition and induced mortality are summarized in Table 2.
During the trial, 1399 An. gambiae s.l. (i.e. a mean number of 58.3 per night) were caught in the control hut. We did not observe any deterrent effect by any treatment compared to the control (Table 2; p > 0.05).
Under the untreated control net, 482 An. gambiae s.l. were blood-fed; this corresponds to 20.1 bites per person per night. Blood-feeding rates of An. gambiae were not significantly reduced by any of the treatments, relative to control ( Table 2). Olyset Ò Net produced a mortality rate of An. gambiae s.l. similar to that of the untreated net (7.86% vs. 8.74%; p > 0.05). The huts with PPF-treated net and Olyset Ò Duo recorded significantly higher mortality of An. gambiae s.l. (12.5% and 14.7%, respectively) than the untreated net (p < 0.001). Mortality recorded with Olyset Ò Duo was significantly higher than that with the Olyset Ò Net (p < 0.01).

Other Culicidae
The mean number of Culicidae other than An. gambiae s.s. collected in the untreated hut was 34 per night. The only significant deterrence of these Culicidae entry rates to huts was with Olyset Ò Net (49.9%; p < 0.001) ( Table 2). None of the treatments repelled greater number of mosquitoes to the veranda than did the control untreated net. The proportion of other Culicidae blood-feeding under the untreated net was relatively low (19.6%). Nevertheless, significantly fewer mosquitoes blood-fed under the Olyset Ò Net (2.2%) and Olyset Ò Duo (1.2%), leading to BFI rates of 88.8% and 94.1%, respectively. The PPF-treated net by itself procured no protection against bites of these mosquitoes. Mortality rates of other Culicidae were 67.6% with Olyset Ò Net and 83.6% with Olyset Ò Duo, consistently higher than for An. gambiae s.l. (8.7% and 14.73%, respectively). The PPF-treated net killed negligible proportions of Culicidae (13.59%), similar to control (12.35%).

Adverse effects
No complaint or adverse effects (symptoms or disorders) in any treatment arm were reported by the sleepers during the experimental trial.

Fecundity and fertility of surviving blood-fed females
During the study of fecundity and fertility, we scored 15,802 eggs laid by 1886 wild surviving blood-fed An. gambiae females and 8407 larvae. The results of fecundity and fertility are summarized in Table 3.
The mean number of eggs/female ranged from 5 to 11 for all LN treatments, with no evidence that the LNs reduced fecundity of the wild pyrethroid-resistant An. gambiae s.l., relative to fecundity of females in the control hut (p > 0.05).
The hatching rates were 55% from the hut with the untreated net, 51% in the PPF-treated net hut, 48% in the Olyset Ò Duo and 56% in hut with the Olyset Ò Net. The impact for the PPF and Olyset Duo was significant (7-12% reduction in fertility; p < 0.01).

Discussion
We studied the insecticidal and sterilizing effect of a new LN impregnated with a mixture of permethrin and PPF against wild pyrethroid-resistant An. gambiae s.s. in Côte d'Ivoire. Bioassay on Olyset Ò Duo showed greater insecticidal efficacy than Olyset Ò Net and PPF-treated net against the standard susceptible An. gambiae s.s. Kisumu strain, reaching the WHO-PES criteria (>95% KD or >80% mortality) after three washes and after the field trial. The fact that Olyset Ò Duo and Olyset Ò Net showed greater insecticidal efficacy after the field trial might be due to their different regeneration times and bleed rates [16]. The apparently poor performance of Olyset Ò Net suggests that, in these conditions, this net required more time for full regeneration after the washing process. Gimnig et al. [8] showed that after washing, this net must be heated to 60°C to restore bioefficacy, but a previous study [25] indicated possible full regeneration within 2 weeks at 30°C and 80% RH. Nevertheless because the incorporation technology is different between LNs impregnated with one chemical (permethrin) and LNs impregnated with two chemicals (permethrin + PPF), the comparison will always suffer a bias. The results of such comparisons about KD and mortality dynamics must be interpreted carefully.
The Phase II trial was run in an area with strong resistance to insecticides and particularly permethrin where An. gambiae s.l. vectors bear both kdr-w and metabolic resistance mechanisms [9]. The results found in this study confirmed that the high resistance reported in this area had a strong impact on the efficacy of Olyset Ò Net both in terms of personal protection through blood-feeding inhibition and mortality of An. gambiae s.s. While the Olyset Ò Net was highly efficacious against other Culicidae mosquitoes, it performed poorly against wild multiresistant An. gambiae s.s., highlighting the crucial need for alternative tools for malaria vector control in areas of pyrethroid resistance. Emphasis is being put on use of the LN Olyset Ò Duo that combines two molecules having unrelated modes of action for improved impact and potential management of permethrin resistance.
Olyset Ò Duo induced significantly higher BFI and mortality than Olyset Ò Net in M'Bé. This result was confirmed by the higher mortality rates in bioassays and release-recapture trial with Olyset Ò Duo than Olyset Ò Net [6] and was similar to that found by Ngufor et al. [16]. The PPF did not show any insecticidal activity when used alone both in bioassays and release-recapture, but it induced mortality (5% corrected from the control) against the multi-resistant An. gambiae s.s. population of M'Bé. Ohashi et al. [17] have already shown that exposure to pyriproxyfen-treated netting shortened the longevity of An. gambiae s.s. It might be interesting to investigate the interaction between permethrin and PPF to assess any relative contribution of PPF to the insecticidal effect of Olyset Ò Duo. Olyset Ò Duo having a higher permethrin bleed rate than Olyset Net -(J. Lucas, pers. comm.) may be a confounding factor that may render difficult the study of the interaction between permethrin and PPF.
As expected, there were a large number of blood-fed females surviving all treatments. Nevertheless, the egg laying rates were almost 10-fold lower in the control than during the release-recapture trial [6]. Indeed, wild An. gambiae s.s. females had difficultly laid eggs under laboratory conditions compared to well-established An. gambiae s.s. strains in insectary. In this previous experiment [6], PPF alone or Olyset Duo containing 30 holes each drastically impacted fecundity of laboratory-reared An. gambiae s.s. having solely kdr as a resistance mechanism. In the current trial, the same LNs with the same amount of holes did not impact fecundity of the wild pyrethroid-resistant An. gambiae s.s. that in addition to kdr have metabolic resistance. Whether these additional metabolic resistances alone or in association with kdr impact negatively PPF has yet to be investigated.
The PPF activity or the Olyset Ò Duo effects on the hatching rate (7-12% reduction) indicated that wild mosquitoes had some contact with the PPF active ingredient. Nevertheless, it is impossible to quantify this contact duration and the active ingredient quantity picked up by wild mosquitoes. We could hypothesize that the active ingredient quantity was enough to impact fertility but not fecundity.
It would be extremely interesting to investigate in laboratory conditions the relationship between the forced tarsal contact time with a PPF-treated net and the impact on fecundity and fertility. These baseline data might allow us to better elucidate these observations in the field.
The behaviour of wild An. gambiae s.s. is different than those of An. gambiae s.s. strains reared in insectary [12]. Such behavioural differences might explain the difference in effect during the release-recapture experiment [6] and the current trial. Moreover using an IGR, such as PPF, on bed net is new; it is therefore important to update the evaluation criteria in order to allow fair evaluation of such new LN. To reach this goal, we suggest investigating the dynamics of the PPF activity in relation to the actual tarsal contact time mosquitoes spent on the net. Behavioural studies are also crucial to better develop this new strategy, especially the behaviour of aggressive An. gambiae mosquitoes in the presence of a torn net.

Conclusion
In previous experiments [6], PPF alone or Olyset Duo containing 30 holes each drastically impacted fecundity and fertility in laboratory-reared An. gambiae having a kdr resistance mechanism. The current trial provided evidence that the same LNs with same amount of holes in them significantly impacted on fertility but did not impact on fecundity of the wild pyrethroid resistant An. gambiae that, in addition to kdr, have metabolic resistance. More trials of this kind should be conducted in other areas with other type of pyrethroid resistance for comparison. Moreover the difficulties experienced to correctly investigate the IGR performances highlighted the urgent need to deeply investigate both IGR intrinsic activity and mosquito behaviour in order to better understand the results observed in natural conditions. Whether these additional metabolic resistances alone or in association with kdr impact negatively PPF efficacy has yet to be investigated.

Conflict of interest
This work was supported financially by Sumitomo Chemical. Funders participated in the study design and the decision to publish, but they have no role in data collection, analysis and preparation of the manuscript. The authors declare that they have no competing interests.