Background
Malaria vector control in sub-Saharan Africa is based on the use of long-lasting insecticide-treated nets (LLINs) and indoor residual spraying (IRS), which both rely on the use of chemical insecticides. Malaria transmission and control has been greatly improved by these vector control tools [
1]. Since these vector control tools depends on the use of chemical insecticides, for them to remain effective, malaria vectors have to remain susceptible to these insecticides, among other factors. Unfortunately, currently used vector control interventions are dependent on a limited number of insecticides from four chemical classes: organochlorines, organophosphates, carbamates and pyrethroids. Among the four main classes of insecticides used for public health, pyrethroids are the only class of insecticides currently recommended by the World Health Organization (WHO) for use in LLINs [
2]. The accomplishments made in malaria control are being threatened by reports of insecticide resistance to all the major classes of insecticides used for public health across sub-Saharan Africa [
1]. The two main mechanisms responsible for insecticide resistance are target site insensitivity known as knock down resistance (
kdr) and metabolic resistance due to elevated levels of insecticide detoxifying enzymes [
3].
Unfortunately, despite the ongoing malaria vector control efforts in Tanzania, malaria continues to be a main public health problem with high mortality and morbidity [
4]. The major malaria vectors in Tanzania mainland are
Anopheles gambiae sensu stricto (s.s.) being indoor feeders and
An. gambiae arabiensis (referred to as
An. arabiensis) being more outdoor feeders [
5]. High LLINs coverage and IRS, has dramatically changed Tanzania vector population whereby the predominant indoor
An. gambiae s.s. is replaced by
An. arabiensis, which leads to residual outdoor transmission of malaria [
6].
In Government’s efforts to elevate the burden and mortality of malaria, through the National Malaria Control Programme (NMCP) it has in the past decade and a half increased LLINs through universal coverage and increased IRS coverage around the Lake Zone area that has highest malaria prevalence. The distribution of LLINs has targeted the most exposed group (pregnant women and children under 5 years of age), through discounted vouchers issued at antenatal clinics [
7,
8], and then by free LLINs delivery campaign in 2010 [
8], which was protracted to the general population through a universal coverage LLIN distribution campaign in 2011. In 2013, a school-based LLIN continuous distribution approach (School Net Program) was started in the country’s Southern zone [
8] and is still ongoing and expanding to Northern zones. IRS operations were initiated in Kagera region around the Lake Victoria in 2006, and extended to the most of districts around the Lake Zone area in 2011. IRS operations started with using a pyrethroid (lambdacyhalothrin), followed by a carbamate (bendiocarb) in 2009, and an organophosphate pirimiphos-methyl (Actellic 300SC) in 2014.
Previous studies conducted in Tanzania showed wide-spread
An. gambiae sensu lato (s.l.) resistance to pyrethroids [
9‐
13], and focal resistance to DDT and bendiocarb [
9,
10,
12,
14]. The major objective of this study was to continue monitoring the susceptibility status of malaria vectors to insecticides used for IRS (i.e. pirimiphos-methyl, DDT and bendiocarb) and in LLINs (permethrin and deltamethrin). This study also aimed to determine associated resistance mechanisms in sampled mosquito populations. Such information is required when planning future vector control efforts and strategies.
Discussion
The present study aimed at describing the current insecticide resistance status of
An. gambiae s.l. and the associated resistance mechanisms in Tanzania. The study confirms the previously reported widespread resistance to pyrethroids, as well as focal resistance to DDT and bendiocarb [
12,
15,
16,
21,
22]. Additionally, for the first time in Tanzania
An. gambiae s.l. resistance to pirimiphos-methyl is reported.
The sampled mosquitoes were identified as
An. arabiensis and
An. gambiae s.s. in all sites. Replacement of the traditional malaria vector
An. gambiae s.s. by the more exophagic
An. arabiensis throughout the country has previously been reported [
6,
14,
22]. The national campaign to distribute LLINs and maintaining universal coverage, would only mean dramatic reduction of malaria transmission in these areas. However, malaria remain to have high mortality and morbidity especially to children under 5 years of age and pregnant women [
1]. It is important to note that the shift of once used to be the dominant
An. gambiae s.s. to
An. arabiensis which is an outdoor feeder, may undermine confidence in LLINs and IRS. Studies have shown that the shift in these sibling species has contributed to the drastic drop in density of
An. gambiae s.s. relative to
An. arabiensis which has led to residual malaria transmission [
6]. This poses a potential challenge for vector control, as most available approaches and tools target the endophagic
An. gambiae s.s. If residual malaria transmission in Tanzania is to be controlled, research on vector control tools targeting the exophagic
An. arabiensis should be prioritized.
This study showed that pyrethroids resistance is wide spread and observed in 13 sites out of 20 that were surveyed. This concurs with previous reports of the ongoing annual detection and insecticide resistance monitoring programme in Tanzania [
15]. The link between pyrethroids resistance and
An. arabiensis, which now a predominant species in Tanzania, was confirmed by significant correlation between the two variables.
Anopheles arabiensis showed the highest levels of resistance to all classes of insecticide tested, as demonstrated in Arumeru site. There was no significant correlation between
An. gambiae s.s. and pyrethroids, this could be due to the limited sample size of the species as they are now being replaced by their sister sibling. The observed resistance to pyrethroids could be attributed to insecticide pressure created by the cumulative effect of insecticide compounds used on insecticide-treated nets [
21,
23] livestock pest control and in agriculture [
10]. The most common LLINs used in Tanzania are permethrin-impregnated Olyset® (Sumitomo) nets [
9]. Additionally, deltamethrin has been used in re-treatment of conventional bednets since the early 2000’s before the introduction of LLINs [
9]. These results are in line with previous studies, which reported expansion of pyrethroid resistance in the country [
10‐
14]. The strong correlation found in the pattern of phenotypic resistance between deltamethrin and permethrin across all sites, confirms these two insecticides have the same mode of action, hence resistance impact on the current pyrethroids dependant vector control tools may lead to operational failure. Even though the results reported in this study are of selected sites and from diagnostic tests and do not give the conclusive indication on the impact of insecticide resistance on the current control tools, however they are a first step in identifying the problem which in the future may deem detrimental to malaria control efforts in the country.
Pirimiphos-methyl (Actellic® 300CS, Syngenta) is the only organophosphate used in IRS in Tanzania since 2014. This study reports for the first time in three of 20 sites in Tanzania,
An. gambiae s.l. resistance to pirimiphos-methyl. The detection of pirimiphos-methyl resistance is of much concern, because this compound is being used as an alternative insecticide after mosquitoes became resistant to pyrethroids and carbamates [
10,
13,
14,
21]. Resistance to pirimiphos-methyl was detected in Geita and Muleba, two sites around Lake Victoria, and in Arumeru in the Arusha region. The rapid development of insecticide resistance has been observed in laboratory experiments [
24]. Use of the compound of same class as pirimiphos methyl as agrochemicals and for IRS is likely to have contributed to the observed rapid emergence of this insecticide resistance in malaria vectors [
25] in the reported sites. Most insecticides used in agriculture are of the same chemical classes, having the same targets and modes of action as those used for vector control [
25,
24]. In Tanzania, pirimiphos-methyl is used in several formulations to control agricultural pests in farms and in storage of agricultural produce such as cereals and legumes (Nkya, Unpublished data). Historically, Arumeru district has intensive agriculture with expansive use of insecticides of various classes, but a limited insecticide pressure from vector control activities (Nkya, unpublished data), and hence potentially accounts for the observed high level pirimiphos-methyl resistance; the insecticide pressure created by its use in agriculture [
25,
24] might have contributed to resistance developing in malaria vectors from this site. Reports of pirimiphos-methyl resistance in areas that rely on its use for malaria control could have implications on the effectiveness of this malaria vector control intervention. IRS has been implemented in Muleba district since 2006, and during the span of almost a decade, three classes of insecticides have been used for spraying, and now resistance to all three classes of insecticides has been documented [
14]. Preliminary 2016 results confirmed pirimiphos-methyl resistance in Arumeru, but pirimiphos-methyl susceptibility was seen in Geita/Muleba, suggesting the resistance detected at those sites in 2015 might not be stable.
In previous studies done in Muleba district, concur with the results of this study. Previously, it has been reported mosquitoes from Muleba to be resistant to bendiocarb, DDT, permethrin and deltamethrin [
14], which concurs with the reported resistance to the same insecticides. In terms of species distribution, the previous study in Muleba reported
An. gambiae s.s. as the predominant species [
14], however this study reports
An. arabiensis as the predominant species. The difference in species distribution between these studies might be due to the once ongoing IRS program in Muleba district, which targeted mostly indoor
An. gambiae s.s. and not outdoor
An. arabiensis. Over time this might have contributed to shift in specie composition that is reported in this study. This study did not report the results for
kdr mutation as they were inconclusive and at a very low frequency. This could be due to fact that the species sampled for this study was predominantly
An. arabiensis and an absence of
kdr mutation in this species had been reported in previous studies [
11].
The underlying resistance mechanisms reported in this study is predominantly metabolic resistance which is the mechanism mostly associated with
An. arabiensis [
11,
14]. Elevated levels of P
450 oxidases, NSEs and GSTs have been reported to be associated with insecticide resistance across all classes of insecticides [
10,
25‐
28]. In all of the analysed mosquitoes from all sites, only four sites had significantly elevated levels of GSTs, P
450 oxidase and NSEs as compared to the susceptible Kisumu strain. NSEs and GSTs, were detected in significant levels in mosquito populations from Arumeru. Elevation of NSEs and GSTs have been previously linked to organophosphate resistance [
10,
25‐
30], thus corroborating the findings of this study. In the three sites where vectors demonstrated pirimiphos-methyl resistance, they were also found to have elevated levels of GSTs, NSEs, acetylcholinesterase and mixed function oxidases. Moreover, elevated levels of these enzymes were also reported from other sites that did not have pirimiphos-methyl resistance, but showed resistance to pyrethroids, DDT and bendiocarb. The general assumption is that insecticide resistance give selection pressure to all insecticides with similar mode of action ineffective. This is not true when it comes to metabolic resistance mechanisms, as some P450 s enzymes show specificity for type I pyrethroids (such as permethrin) or typeII pyrethroids (such as deltamethrin) [
31,
32]. This further confirms that metabolic resistance may indeed be associated with resistance to different classes of insecticides. As the various insecticide classes used for malaria vector control have been frequently used for crop protection, this may explain why resistance to most insecticides develops so rapidly in malaria vectors in Tanzania.
Authors’ contributions
TEN and BK conducted mosquito sampling, bioassays and molecular work, analyzed results and drafted the manuscript. ZM and DM contributed to sample preparation and data analysis and drafting the manuscript. SM, GG, NK, MM, and RR contributed to drafting the manuscript. HJO and LML contributed to study coordination and reviewing the manuscript. WN K and SM conceived the study. WNK coordinated the study, participated in sample preparation and molecular work, analyzed results and wrote the manuscript. All authors read and approved the final manuscript.