Malaria is a preventable and treatable disease that infects millions of people globally every year. The World Health Organization (WHO) reported the global malaria burden for 2018 as an estimated 228 million cases associated with approximately 405 000 deaths . This constitutes a significant increase from 219 million cases reported for 2017 . Africa is the most malaria-burdened continent, accounting for 93% of the cases in 2018 . Following notable declines in global malaria indices between years 2000 and 2015, these gains have reached a plateau and malaria cases have once again increased although the mortality has declined [1–3]. South Africa has shown similar trends, with significant decreases over the past two decades . In the 1999/2000 malaria season, there were over 60 000 malaria cases that decreased to less than 13 000 in the 2013/2014 malaria season . However, a steep increase in malaria cases and associated mortality in recent years (2017/18) has revealed the fragile nature of control efforts and the ease with which malaria can resurge [2, 3]. This multi-year trend of stagnating malaria control is an indication that the current malaria control strategies are no longer adequate and new or supplementary measures must be developed.
Human malaria is caused by five Plasmodium parasites which are transmitted by females of certain Anopheles mosquitoes . In Africa, malaria vector species are mainly from two taxonomic clusters: the Anopheles gambiae complex and the Anopheles funestus group . The most significant species from the An. gambiae complex are An. arabiensis and An. gambiae, and An. funestus from the An. funestus group [8, 9]. However, a number of other, less efficient, secondary vectors are also capable of transmitting malaria . The An. funestus group is broadly distributed across Africa, with its major species, An. funestus, widely distributed over subtropical and tropical Africa where it breeds in permanent large water bodies with emergent vegetation [7, 11]. Anopheles funestus is highly anthropophilic (human biting) and exhibits endophilic (indoors) feeding and resting behaviours [7, 12]. The major malaria vector species from the An. gambiae complex, An. arabiensis and An. gambiae are widely distributed across Africa . Members of the An. gambiae complex prefer to breed in temporary bodies of water that are clean and shallow [14, 15]. While An. gambiae is highly anthropophilic , An. arabiensis is zoophagic, feeding readily on animals in most areas, particularly cattle [15–17]. Studies have shown that An. arabiensis prefers to feed outdoors even in areas where it mostly feeds on humans [18, 19]. Anopheles arabiensis is therefore less impacted by indoor vector control strategies . The different behaviours of the major malaria vectors makes it challenging to control malaria transmission . In South Africa, An. arabiensis is widely acknowledged as the main vector although other species such as An. merus, An. rivulorum and An. funestus become locally important .
The primary control of malaria for decades in most malaria-endemic regions of the world has been the implementation of indoor residual spraying (IRS) and long-lasting insecticide-treated bednets (LLIN) [22, 23]. LLIN’s primarily utilise pyrethroid insecticides while a range of pyrethroids, organophosphates, carbamates as well as the organochloride DDT are applied to internal walls and ceilings of housing structures during IRS programmes . Both pyrethroid insecticides and DDT have the same target on the voltage-gated sodium channel found on the mosquitoes’ neurons therefore vectors become resistant to both strategies . Furthermore, these resistance alleles have spread at a rapid rate throughout Africa, requiring urgent action to prevent an increase in malaria [24, 26]. The increasing resistance developed within various Anopheles species to these insecticides poses a major challenge to the effectiveness of these key vector control methods [20, 24, 27].
Aside from escalating insecticide resistance affecting the value of IRS and LLIN’s indoor interventions, additional challenges are emerging . The major malaria vector species historically preferred feeding indoors [29, 30]. However, some recent studies have shown a shift in behaviours of An. gambiae and An. funestus to feeding outdoors instead of indoors in some areas where IRS and LLIN are implemented [29, 31]. Zoophilic characteristics in some of the major vectors are also a challenge . Yet another challenge is a temporal shift in feeding behaviour with the malaria vector species biting in the early evenings and mornings when people are not under their protective nets [32, 33]. Given these limitations of IRS and LLIN to curb transmission, the challenge of residual malaria poses a serious hurdle in reaching malaria elimination objectives, and the number of malaria cases and deaths remains unacceptably high .
The use of cattle-administered endectocides is a promising strategy for outdoor vector control that could complement IRS and LLINs [35, 36]. Several endectocide drugs are effective against a wide range of both endo- and ectoparasitic nematodes and arthropods in humans and cattle [20, 35]. These drugs include ivermectin, eprinomectin, fipronil and diflubenzuron . Ivermectin was the first endectocide to be used in humans and continues to be used to treat river blindness through mass drug administration (MDA) [36, 38]. Ivermectin is a lipophilic drug belonging to the avermectin class of macrocyclic compounds  and is also used to treat onchocerciasis, strongyloidiasis, lymphatic filariasis, scabies and head lice [20, 38]. Endectocides are also of veterinary importance as they are used to control parasites in animals such as cattle and goats .
Endectocides utilize a different mode of action against insects to that of IRS and LLIN  and can thus complement traditional control measures. In the vector, ivermectin primarily targets the glutamate gated chloride channels, which are neurotransmission inhibitors through their 16-membered macrocyclic lactone [20, 45–47]. Binding to the channels leads to an influx in chloride ions leading to neuromuscular junction and hyperpolarization . In contrast, fipronil is a phenylpyrazole compound that works by blocking the GABA-gated ion channels, which are also in the central nervous system of arthropods [48–50]. Both ivermectin and fipronil result in flaccid paralysis and eventually death in the target parasites [20, 49]. These chloride gated iron channels are not present in vertebrates, therefore, ivermectin and fipronil are non-toxic to humans and livestock . Ivermectin MDA in humans is a promising malaria control tool that targets mosquitoes with control-avoidance biting behaviours and those that have developed physiological insecticide resistance [35, 47]. Several studies have shown that ivermectin-treated human blood decreases survival, feeding frequency, blood meal digestion and fecundity of mosquitoes [52–54]. Ivermectin (brand name Mectizan®) has been used successfully in humans for river blindness since 1987 and does not have toxic side effects at recommended doses [55, 56]. Several studies have shown that ivermectin and several other endectocides applied to cattle or other livestock also decrease the survival and fecundity of malaria vector mosquitoes [16, 35, 51, 57]. Fipronil has been approved for use on domestic animals in many countries and is used to control arthropods such as ticks, cockroaches and fleas . In addition, fipronil has been used in cattle to control leishmaniasis vectors [59, 60]. Similarly, fipronil is effective against all life stages of Anopheles mosquitoes . However, field studies on its use against mosquitoes are limited.
The effectiveness of endectocides is linked with their pharmacokinetics, which vary across different species . The route of administration also has a significant effect on the pharmacokinetics of endectocides. In previous studies, ivermectin pharmacokinetics studies were conducted in cattle to compare subcutaneous and oral routes of administration [16, 42]. Ivermectin injected subcutaneously in cattle was effective against An. arabiensis mosquitoes for longer than oral or topical treatment . Higher ivermectin plasma concentrations were produced with subcutaneous treatment than with oral administration. High concentrations produce an enhanced systemic availability which results in higher efficacy against the targeted parasites . For formulations typically used in recent years, the maximum concentration of subcutaneously injected ivermectin was reached at day 1 while the minimum was reached after 25 days . Similarly, fipronil injected in cattle reached its maximum and minimum concentrations rapidly within 24 hours . In a study where the pour-on fipronil was administered in cattle to investigate its effect against ticks, the mean plasma concentration values over time varied and maximum concentration was 73.7 g/L reached after 2.5 days . The pour-on fipronil concentration reached its half-life at day 19 and decreased slowly until minimal level at day 40 . Topical treatment results in exposure to environmental degradation such as mechanical removal by rain (Cid et al. 2016). Factors such as body weight, nutrition type and physiological status also lead to variation in drug concentrations within individuals of the same species .
Currently there are no studies investigating the impact of cattle-administered endectocides on mosquitoes in South Africa. Studies that have conducted this type of research in other countries have mostly focused on ivermectin only. The present study included an additional potential endectocide, fipronil. This study also considered the pharmacokinetics profiles of the two drugs and conducted feeding trials at various points including at the minimum and maximum concentrations. The aim of this study was to investigate the effectiveness of two endectocides, namely: ivermectin and fipronil for control of An. arabiensis in South Africa. The specific objectives were: 1) to demonstrate that ivermectin and fipronil reduce adult survival and fecundity of An. arabiensis, 2) to compare the efficacy of ivermectin against that of fipronil and 3) to assess the duration efficacy of each endectocide. We predicted that endectocides-treatment would result in a significant increased mortality of An. arabiensis and a reduction in the egg production. We also predicted that injected ivermectin would be more effective than pour-on fipronil. We further predicted that the efficacy of both endectocides would last for a month based on the manufacturer’s instructions on their duration effect against other parasites and data from previous related studies.