Reflections from an old Queenslander: can rear and release strategies be the next great era of vector control?

Abstract

In this perspective, I discuss the great eras of vector control, centring on Aedes aegypti, the primary vector of dengue, Zika and several other viruses. Since the discovery and acceptance of the role of mosquitoes as vectors of disease agents, several significant strategies have been developed and deployed to control them and the diseases they transmit. Environmental management, insecticides and, to a lesser extent, biological control have emerged as great eras of vector control. In the past decade, the release of massive numbers of specifically modified mosquitoes that mate with wild populations has emerged as a significant new strategy to fight vector-borne diseases. These reared and released mosquitoes have been modified by the addition of a symbiont (e.g. Wolbachia bacteria), radiation or introduction of a genetic construct to either sterilize the wild mosquitoes they mate with, crashing the population, or to reduce the wild population's capacity to vector pathogens. Will these new rear and release strategies become the next great era of vector control? From my vantage point as a dengue control manager and researcher involved in two Wolbachia programmes, I will discuss the hurdles that rear and release programmes face to gain widespread acceptance and success.

1. Introduction: the great eras of vector control

As I sit on the deck of my Queenslander house, contemplating my approach for this perspective article, I note the familiar dance of male Aedes aegypti at my feet. Three males, flying figures of eight, hoping to be the first to intercept and mate with a female, coming to bite me. This is not an ordinary review article, but a reflection of my 20 years directing dengue control operations in Queensland, my 25 years researching surveillance and control of the dengue vector Ae. aegypti, and my 22 years living in an old Queenslander house. Watching, counting and contemplating this most unusual mosquito that lives in my house—Ae. aegypti, the cockroach of mosquitoes. From this background, I am compelled to describe the three great eras of mosquito control, focusing on my expertise on Ae. aegypti. (NB: both vaccination and public health also contributed significantly to decline in vector-borne diseases, but my focus here is on vector control.) And to contemplate whether the new biologically based, rear and release strategies could become the next great era of vector control.

My old Queenslander house (figure 1) encapsulates these eras, having been born of and witness to all major strategies of vector-based mosquito-borne disease control. Queenslander houses are elevated, a wooden house perched on stumps, open to the air. The original architectural design was to elevate the house to avoid the miasma, the foul air close to the ground, thought to cause illness in residents of tropical Queensland Australia [1]. When Louis Pasteur and later Robert Koch developed the germ theory of disease, the penny dropped. Walter Reed soon confirmed that the parasite causing yellow fever was carried and transmitted by mosquitoes. Environmental management emerged as the first great era of vector control. While old Queenslanders continued to be built, windows in some houses were screened, water-holding containers in yards were removed and rainwater tanks, the common source of domestic use water and Ae. aegypti, were screened or removed. After the Second World War, insecticides emerged as the second great era of vector control. Lambda-cyhalothrin, a residual synthetic pyrethroid, was sprayed under tables and beds, and inside wardrobes, to kill Ae. aegypti in my Queenslander during dengue outbreaks in 2003 and 2009. As concern grew about the environmental impact of pesticides, and development of physiological resistance to insecticides developed in mosquitoes, the search for ‘natural' methods of vector control gained traction. Biological control emerged as potentially the next great era of vector control. I conducted field bioassays of Bacillus thurengiensis var. israelensis (Bti) against Ae. aegypti under my deck, and witnessed infestation of trial buckets by the predatory mosquito Toxorhynchites. Finally, my old Queenslander has been home to multiple releases of Wolbachia-infected Ae. aegypti. The wMel strain of Wolbachia, shown to block transmission of dengue, Zika and chikungunya viruses by Ae. aegypti, was established in the local mosquitoes in 2013. Since this time, there have been no dengue outbreaks in my neighbourhood, nor in much of north Queensland [2]. Is this the dawning of the fourth great era of vector control?

Figure 1.

An old Queenslander style house. Aedes aegypti readily enter unscreened windows, and harbour indoors.

Why were these first three vector control strategies successful (see figure 2 for summary)? Environmental management made inherent sense. If Aedes vectors were coming from flooded containers and cisterns, one could simply get rid of the containers or screen the cisterns to deny the vector access. It was also a sustainable approach in that once the house was screened, and the rainwater tank removed, it stayed that way. Improved housing also fitted into the model of new technology, and scientific advances to improve our standard of living, especially health. In New Orleans, the site of many an outbreak of yellow fever, legislation was introduced to outlaw cisterns [3]. Rainwater tanks in Brisbane, Australia, were largely replaced by piped, reticulated water systems, resulting in a crash in Ae. aegypti populations and a cessation in dengue [4]. Source reduction became the mainstay of many urban Aedes programmes [5]. Clean-up campaigns targeting containers in yards such as tyres, tins and buckets were part of most Ae. aegypti control programmes, and are still emphasized to this day. And these measures had early success. Epidemics of yellow fever waned in much of the southern USA [6]; dengue outbreaks ceased in Brisbane [4]; and the Panama Canal, aided by reduced yellow fever transmission following window screening and source reduction programmes targeting Ae. aegypti (as well as malaria vectors), was completed [7].

Figure 2.

The great and not so great eras of vector control, with an emphasis on Aedes aegypti.

The dawning of the insecticide era was largely based on the early successes of dichloro-diphenyl-trichlorethane (DDT) during the Second World War. DDT was used to successfully kill lice and stop outbreaks of louse-borne typhus [8,9]. This insecticide was deemed a magic bullet: relatively cheap to produce, highly toxic to insects, and persistent, allowing for one-off treatment campaigns. Fred Soper used DDT in his campaign to eradicate Ae. aegypti from the Americas [10]. Teams of vector control staff conducted perifocal spraying—spraying containers and adjacent habitat such as walls with DDT to kill emerging and resting Ae. aegypti for sustained periods of time. This strategy was highly successful, and Ae. aegypti was purportedly eliminated throughout much of Central and South America [11]. Insecticides were also relatively inexpensive, could be applied in a variety of ways (from vehicles, planes, hand) to almost any habitat (indoors, outdoors, in water and vegetation) and in ever-expanding formulations (thermal fogs, ultra-low-volume sprays, residual granules, even bednets). Easy, cheap and effective. This, of course, created a sustainable business model of manufacturer, distributor and customer. Big money was made. And the industry was nimble. When problems of insecticide resistance began to appear with DDT as early as 1947 [12], new types and eventually classes of insecticides were created to fix the problem. To this day, insecticides remain the most dominant strategy used to control Ae. aegypti.

Despite these successes, insecticidal control of mosquitoes, especially Ae. aegypti, is fraught with issues. Physiological resistance to several classes of both larvicides and adulticides is widespread in Ae. aegypti [13]. Application methods, particularly outdoor space spraying, are generally ineffective for indoor resting Ae. aegypti [14]. Thus, dengue and other Aedes-borne viruses, continue to spread seemingly out of control. Finally, the emergence of the environmental movement, initiated in no small part by Rachel Carson's Silent Spring, which detailed the impacts of insecticides on the environment, created public mistrust in insecticide use [15].

Biological control was largely born of necessity in response to the development of resistance; both physiological resistance of mosquitoes to insecticides, and public resistance to insecticides and their environmental impact. The World Health Organization (WHO), especially through their new Tropical Disease Research (TDR) programme, invested heavily in the search for biological control methods [16]. Basic biological research turned up an array of pathogens that could kill mosquitoes, especially mosquito larvae. Protozoans, fungi, bacteria, viruses, nematodes, crustaceans, vertebrates, even other mosquitoes were found to parasitize or predate mosquitoes. The literature is replete with observational studies confirming ‘larvivorous’ behaviour of a variety of animals [17,18]. Unfortunately, the promise of a universal biological control strategy to manage mosquitoes, especially container mosquitoes such as Ae. aegypti, was not fulfilled. Most methods did not provide sustained, high levels of control; for example, releases of Toxorhynchites mosquitoes that specifically target container mosquitoes) only provided around 75% control, and were impractical to sustain [19,20]. The crustacean copepod Mesocyclops consumes large numbers of first instar Ae. aegypti. It was successfully used in a community-based control programme in Vietnam, where dengue outbreaks were dramatically reduced [21]. However, the strategy is largely niche-based, and relies upon mass rearing of the copepod, and seeding them into water storage tanks that are the predominant source of Ae. aegypti. Unfortunately, in most dengue endemic areas Ae. aegypti exploit a variety of smaller, ephemerally flooded containers that are not suited to copepod treatment. Many biological control agents have relatively complex life cycles, such as the fungus Coelomomyces that uses copepods as an intermediate host. Thus, they are difficult and costly to rear, and commercial production at scale was not forthcoming. It comes as no surprise that the biological control agent that has been the greatest commercial success, Bti, is basically a biopesticide. The bacteria is easily mass reared in industrial fermenters, and the toxic spores and crystals harvested and formulated. Bti is available in many formulations, and today is widely used as a larvicide against Ae. aegypti and Aedes albopictus. Finally, synthesized versions of insect hormones, especially juvenile hormones that control metamorphosis, have been used against Aedes larvae. While technically not a biological control agent, these ‘biorational' insecticides are perceived as a ‘green alternative' to traditional insecticides.

2. Rear and release: the basics for mosquitoes

Rear and release involves the intentional mass rearing and release of modified mosquitoes [22]. These modified mosquitoes mate with existing members of the wild population, causing the desired change. This may include inducing sterility, infecting with a microbe or introducing a specific genetic element. The goal is either to suppress the wild population, potentially to the point of local elimination, or reduce the ability of the target population to vector pathogens such as dengue viruses. While some authors include rear and release programmes within the suite of biological control strategies [18], I distinguish the two, as the strategic approach to rear and release is quite unique. The key aspect of rear and release programmes is that they require the repeated release of very large numbers of mosquitoes. Almost all mosquito control campaigns and programmes involve the immediate killing or removal of mosquitoes. Thus, the mass release of mosquitoes in rear and release programmes is strange, even counterintuitive, to the public and vector control managers. Public education, community engagement, political buy-in and media campaigns are a required part of a mosquito rear and release programme [23].

(a). And a brief history

Rear and release insect control strategies were, like the insecticide revolution, born in the post-war technological boom of the 1940s–1950s. Knipling recognized that the mass production and release of sterilized male insects could be used to suppress, even eliminate, pest populations [24]. Subsequently, the sterile insect technique (SIT) was launched and used to eliminate the screw worm fly from the southeastern USA. Laven first used the concept of male-induced sterility, via release of cytoplasmic incompatible males, to eradicate a local population of Culex pipiens fatigans in a Burmese village [25]. Attempts were also made to use SIT, based on irradiated males, to suppress populations of Culex and Ae. aegypti in India, but the programme was catastrophically undermined by media and negative public opinion [26]. Another indication of the problems to face rear and release programmes occurred in the SIT programme targeting Culex tarsalis in California. In this case, releases were unsuccessful, as the released mosquitoes consisted of an inbred colony strain that failed to effectively mate with wild females [27].

Thus, despite a few successes [28], rear and release failed to gain widespread use as an effective mosquito control strategy. After all, it was expensive, required significant infrastructure, and could lead to negative public opinion. Indeed, as the failed releases in India highlighted, ‘Failed release programmes have the consequence of not only wasting resources but threatening the success of future programmes by creating public mistrust or donor reluctance in supporting these types of technologies' [29]. In the light of these risks, and costs, insecticides were the most attractive strategy.

3. Beyond the pilot study: can rear and release strategies be operationalized to treat cities at scale?

Aedes-borne viruses (ABVs), especially those involving Ae. aegypti, are concentrated in urban areas of the tropics. Thus, large cities such as Bangkok and Singapore are subject to repeated outbreaks and occasional epidemics [30]. Furthermore, large cities are thought to act as a viral reservoir from which the virus evolves and emerges to spread to other urban areas [31]. Thus, an effective ABV control strategy must target these source cities, and be scalable in size and speed, to effectively treat the thousands of square kilometres with 1–2 generations of the mosquito/virus life cycle (effectively within a month).

Currently, many rear and release programmes are in the pilot stage. Most have confirmed the method works in the laboratory using colony mosquitoes held in small cages. The more promising methods have proven their worth in biosecure, semi-field cages [32,33]. And the few that have passed through regulatory hurdles and the prism of public opinion have been tested in small pilot studies in the field [34]. Finally, if the pilot trials are successful, public opinion favourable and financial backing secured, large-scale releases can go forward. To date, this is happening with Wolbachia, with the World Mosquito Program targeting Ae. aegypti in a population replacement strategy, and, to a lesser extent, population suppression using Wolbachia-infected males. Wolbachia-based population suppression field trials have been conducted against Ae. aegypti [35], Ae. albopictus [36,37] and Aedes polynesiensis [38].

Pilot releases of genetically modified male Ae. aegypti by Oxitec have occurred in several countries, including the Cayman Islands, Brazil and Malaysia. These pilot studies are time and space limited, and subject to stringent biosecurity protocols. Furthermore, they demand strong community and political support. Genetically modified mosquito (GMM) trials attracted considerable publicity, much of it negative [39,40]. This can derail open releases even if regulatory approvals have been given, much like what happened to Oxitec in Stock Island, Florida. Planned releases of GMM OX513a, despite being given regulatory approval by the US Food and Drug Administration, were cancelled due to ‘community backlash and logistics' only to have releases of Wolbachia-infected male Ae. aegypti subsequently released [41]. Strong community engagement, supplemented by community-driven advisory committees are vital to project success. Perhaps the key reason for Eliminate Dengue's (now World Mosquito Program) successful pilot and scaled releases of wMel-infected Ae. aegypti has been the focus, from the beginning, on community engagement and support.

4. Key questions from the vector control desk … and the polling booth

Managers of existing vector control programmes watch these developing rear and release programmes with great interest … and a bit of trepidation. Additionally, vector control staff interact with the public and are a source of information regarding mosquitoes. As such, we have several basic questions that we'd like project managers of rear and release programmes to consider (table 1).

Table 1.

Questions from a vector control manager … and a voting member of the public.

1. Is it safe? Are there any potential public health concerns during and after release? Is it safe to the environment? Can it be contained or reversed if required?

2. Is it effective? What is the evidence that it works? Does it reduce nuisance biting? Does it reduce disease risk? What is Plan B if it fails?

3. Is it approved? Does it have a formal risk assessment and regulatory approval? What does the community think? Why are you releasing it here?

4. Is it expensive? How does it compare with existing programme costs? Who will pay? Are we mortgaged to an expensive programme for years?

5. Will it eliminate jobs for me and my staff? Will vector control staff become obsolete? Can we be a part of the programme? Will we be upskilled? Is it sustainable?

6. What happens when releases are done? Will you simply pack up and leave? Will the strategy continue to work after you leave? Can my vector control team manage the programme down the road? Will you be there to help if it fails? What is your liability?

How best to communicate this information to the vector control team, and the public? I suggest reading the article by Jim Lavery's group, who chronicled the community engagement strategies used by Eliminate Dengue [23]. And, you should definitely read the Nature editorial ‘Oh, New Delhi; oh, Geneva' [26], which highlights how not to do it. In addition to traditional community information sessions, developing a community-led ‘project advisory committee, community reference group, etc.' is critical [42]. These groups disseminate information to key community and governmental leaders, and serve as a source of information from the public—a two-way communication strategy. I do think the GMM approach will be subject to intense public scrutiny and negative publicity, aided by the internet, especially social media. Hopefully, time will foster the gradual acceptance of GMM programmes, provided they do not falter in their execution.

5. Hurdles to successful implementation at scale

Even if community and regulatory approvals are provided, and pilot trials are successful, many hurdles to implementation at scale remain (table 2). Research into technological fixes for mass rearing, releasing and even monitoring is under way. There are two huge expenses for rear and release programmes, excluding communications: the production and release of massive numbers of mosquitoes, and monitoring. Any method that can reduce the release numbers will result in significant savings. Thus, integration with existing vector control programmes or citizen science programmes that involve source reduction to reduce wild populations, and thus required release numbers, should pay dividends. Using the public to literally become the rear and release machine, by hatching egg strips and rearing larvae in disposable paper containers, helped WMP release and establish Wolbachia-infected Ae. aegypti at scale in Queensland [42]. Finally, the released material must be fit relative to the wild population. Initial releases of wMel-infected Ae. aegypti in Rio de Janeiro, Brazil, failed as the wild background population was resistant to the synthetic pyrethroids used by residents to kill mosquitoes, but the released mosquitoes were not. Backcrossing to introduce the resistant alleles into the released populations finally resulted in wMel establishment [43].

Table 2.

Major hurdles to overcome for successful rear and release.

1. Mass rearing:

a. Development of appropriate strain of vector and IIT/SIT agent, obtaining regulatory approval for its release.

b. Development of effective mass rearing methods, especially for population suppression methods that require high release ratio methods.

c. High-fidelity sex sorting especially for population suppression methods that use males infected with Wolbachia.

d. Maintenance of high fitness release material. Need assays and SOPs to maintain and measure in released material. Is pesticide resistance an issue? Do you need a large rearing cage? Genetic analysis of strain versus wild material.

e. Infrastructure—low or high? Can a rearing facility be located close to release area? Can a mobile rearing facility be used?

2. Mass releases:

a. Development of effective mass release methods, especially for population suppression methods that require high-release-ratio methods.

b. Cost effective, efficient, with minimal impact of released mosquito fitness.

c. Integration of technology: drones, GIS mapping, automated release vehicles.

d. Community involvement: from passive cooperation, to active involvement. Citizen science potential exists.

3. Surveillance:

a. Rapid inexpensive way to monitor wild populations.

b. Low or no power traps.

c. Use of technology to remotely monitor traps.

d. Simple assays to identify the agent or transgene from wild material.

e. Can citizen science be employed to save costs and engage the public?

4. Community engagement:

a. Effective methods to obtain consent over large areas.

b. Effective methods to respond to crises and controversy; and to quickly and effectively address community concerns throughout the life of the project.

c. Effective means, and roles, for community involvement/citizen science.

d. How to manage social media, the press and politics.

e. How to build public confidence and acceptance of GMMs.

5. Governance beyond community borders.

a. How to obtain regulatory approval for release of non-endemic vector and/or symbiont strains, especially for Wolbachia-based releases.

b. How to obtain approvals for open field release of GMMs.

c. Can GMMs, and the transgenes, be contained?

d. Will the released material cross borders? Is this a problem?

e. How will the programme affect existing vector control programmes? How can it be integrated with them?

f. Can you develop a sustainable business or funding model?

g. Pilots first … no massive funding of large scale field trials before pilot success.

6. The big picture:

a. How do we measure impact? Vector? Virus? Agent?

b. What does success look like?

c. What do we do if it fails?

d. More importantly, what do we do if it works? Long-term management plan.

e. The long game—will other vectors emerge? The empty niche issue. Will evolution render the method ineffective? How fast?

f. Will success build complacency … and kill funding?

Mosquito monitoring could benefit from technology that can identify and tally captured target insects and telemetry results for remote access. Citizen science is also an alternative [44]. The Zika Mozzie Seeker programme in Queensland relies on citizens to obtain Aedes eggs using ovitraps [45]. The ovistrips are then posted to a laboratory for PCR analysis. A programme in Spain uses smartphone photographs to monitor Ae. albopictus [46]; could smartphones be employed as a proxy for relative mosquito abundance? With Ae. aegypti being an indoor-harbouring, human-biting mosquito, it is the best bet.

6. What if it works? Thoughts on a post-release strategy

My colleague, Bill Petrie, once headed the Cayman Islands Mosquito Research and Control Unit during the days of Oxitec's rear and release programme using releases of GMM male Ae. aegypti (strain OX513a) as a release of insects carrying a dominant lethal (RIDL) gene approach to suppress natural populations. He told me that he was asked at a conference. I asked him, ‘What do you do if it does not work?' He replied: ‘That's easy, you simply stop releasing. The harder question is “What do you do if it works?”’ And that's the rub. Most rear and release programmes are awash in the heady days of pre-release preparation (community engagement, population monitoring, testing of mass rearing and releases methods) and the frantic days of prolonged releases. Indeed, for all population suppression programmes (SIT, IIT, RIDL etc.) that require truly mass rearing this is a 24/7 exercise. Then, at some point, the end is reached. Maybe it's based on the end of the wet season, or the achievement of a vector population target, or a regulatory defined endpoint. For the WMP programme, it may be several consecutive weeks of > 50% Wolbachia infection, with the knowledge that CI will drive the symbiont to fixation with no further releases [42]. For IIT/SIT suppression, it may be reaching and maintaining almost 100% suppression. As deploying Wolbachia-infected mosquitoes effectively drives genes into populations [29], it's unsurprising that the two major strategies of programmes utilizing gene drive systems are also population replacement or suppression [47], with similar end points to the Wolbachia-based methodologies.

However, little thought has been published on what happens after releases [48]. For Wolbachia-based population replacement, releases stop but monitoring continues, albeit at a slower pace. The Wolbachia infection works silently in the background, blocking dengue transmission and maintaining high levels of infection. But eyes should be on the lookout for consequences, such as appearance of viral strains that break through virus blocking, or the relative decline of vector populations, such as what happened with wMelPop Wolbachia in Queensland [49]. For population suppression programmes (IIT, SIT, even GMM) simply folding up the tent and leaving town can result in rapid re-invasion and rebound of mosquito populations [50], much to the dismay of the public. Perhaps small-scale releases, coupled with traditional vector control, could sustain the achievements. For GMM releases, it is more complicated. No doubt searches for evidence that the genes have spread within and beyond the target population will be required, even within non-target species [51]. But for any of these rear and release programmes, a post-release plan is needed and should be developed as part of the risk and regulatory package. For large-scale releases, epi-endpoint goals will require monitoring of ABV transmission, probably for several years [52]. And funders take note: please do not fund extremely expensive epi-endpoint field trials until pilot trials prove convincing. Please see the testing pathways developed by the World Health Organization [51] and covered by James et al. [47].

7. Final thoughts

The need for novel, effective vector-borne disease control is critical, especially for ABV. New and re-emerging viruses continue to plague the tropics and, enabled by climate change and people's propensity for international travel, are emerging in temperate areas such as the USA and Europe. Insecticidal control, that grand great era of vector control, is increasingly fragile due to insecticide resistance, environmental concerns and a general greening of the public. New technology, both on the molecular/biological sphere, and in automation and artificial intelligence, offers solutions with rear and release strategies at the forefront. Finally, as rear and release strategies are geographically focused, targeting significant urban areas in the case of Ae. aegypti, many communities will be missed. Here, traditional vector control will still have to play the leading role.

I sit on the deck of my Queenslander, confident that the Ae. aegypti biting my ankle is no threat, effectively vaccinated against dengue by Wolbachia. Can novel rear and release strategies gain the initial successes, and thus public and political confidence, to advance to the next great era of vector control?

Data accessibility

This article has no additional data.

Authors' contributions

S.A.R. conceived and wrote the paper as a personal perspective. K.R. provided editorial input, helped write the paper and submitted the manuscript. S.A.R. and K.M.S. gave final approval for publication.

Competing interests

We declare that we have no competing interest.

Funding

S.A.R. has received research funding from Monash University as part of the Eliminate Dengue Wolbachia programme and from Verily Life Sciences. K.M.S. has been funded by Verily Life Sciences. S.A.R. is employed by Monash University as part of the World Mosquito Program starting 1 May 2019.