In response to a broad governmental referral, the French High Council for Biotechnology has published an opinion on the use of genetically modified (GM) mosquitoes for vector control [1].

Emerging techniques of vector control were developed to overcome (i) the lack of therapies, preventive treatments and vaccines for most mosquito-borne diseases, and (ii) the limitations of existing vector control techniques (the situation is particularly critical regarding insecticides: in France, essentially only one insecticide is used against adult mosquitoes (deltamethrin), and its efficacy decreases due to resistance evolution in mosquito populations).

To date, only one GM mosquito-based technique has been developed to an operational level, Oxitec's RIDL technique, which seeks to reduce a mosquito population by repeated mass releases of sterilising transgenic males [2]. Two other techniques under development rely on CRISPR-based gene drive, seeking to spread a genetic trait in a wild population, either to eliminate the population by spreading sterility [3] or to make the target mosquitoes incapable of transmitting pathogens [4].

To identify the specific benefits and limitations of the different GM mosquito-based techniques, a cross-analysis of different vector control techniques was conducted with respect to possible objectives, efficacy and sustainability, technical constraints and risks to health and the environment. Consideration was given to both existing techniques (chemical, biological, physical, and environmental) and emerging techniques based on release of mosquitoes, whether GM (RIDL and the different gene drive techniques) or non-GM–irradiated (standard sterile insect technique (SIT)) or carrying Wolbachia¹ (incompatible insect technique (IIT) and spread of pathogen interference (PI) technique).

As specified by the referral, we considered the mosquito-borne diseases and vector species present across France, including overseas territories. The French territories being dispersed across the world, the most notable mosquito-borne diseases worldwide were considered, namely dengue, chikungunya, Zika, yellow fever, West Nile fever, for the viral diseases, and malaria and lymphatic filariasis for the parasitic diseases. We focused on the corresponding local vector species: mainly Aedes aegypti and Aedes albopictus, and species of Anopheles and Culex.

These vector species have very distinct features, not only in distribution and vector competence, but also in bio-ecology (reproduction modes, potential for survival, host preferences, peaks and sites of aggressiveness, invasive potential…). The vector systems themselves (the triad mosquito/pathogen/vertebrate host), as well as the diversity of situations encountered across the territories add another layer of complexity. This overall complexity must be understood and taken into account in order to design the most appropriate vector control strategy.

Cross-analysis of the different vector control techniques has been conducted in great detail and has made it possible to identify specific features and relative benefits and limitations of each of these techniques. Detailed results are developed in HCB's opinion (HCB, 2017).

At a more general level, we found:

– no divide between GM and non-GM techniques or between emerging and existing techniques (Fig. 1);

– shared characteristics within different sets of techniques, i.e. (i) techniques based on release of mosquitoes, (ii) population reduction techniques² vs. population modification techniques³, (iii) self-limiting techniques⁴ vs self-sustaining techniques⁵;

– complementarity of the techniques.

Lastly, we found that the benefits and limitations of these vector control techniques cannot be treated in a generic manner, but will depend on the target vector species, the intended objective, and the broader context (epidemiological, environmental and socio-economic context, including available human and financial resources).

Key highlights for each of these broad conclusion points are developed below.

Because they operate through mating between released mosquitoes and field mosquitoes, a key feature of all techniques based on mosquito release is an unprecedented specificity of action, confined to the released mosquito species and any interfertile (sub)species. This has the major benefit of minimizing the direct impact of vector control on health and the environment. It does, however, entail as many individual interventions as there are species of non-interfertile vector mosquitoes to be targeted on a given site.

Population reduction techniques, whether or not they use mosquito release, and whether or not the released mosquitoes are GM, have in common:

– an environmental impact associated with the reduction of target mosquito population density and depending on the target species’ role in the ecosystem. This impact varies according to, amongst other factors, whether the relevant species is autochthonous or invasive, whether its habitat is urban or natural, whether specialist predators exist, the extent to which the population is reduced (simple reduction, local elimination, or eradication of the species⁶), the duration of the effects of a technique (depending, amongst other things, on how isolated the treated area is), and the specificity of the technique (techniques involving mosquito release being the most specific);

– the potential for unintended replacement of the target population by the population of another vector species, which increases the more the target population is reduced and the more this reduction persists over time.

Population modification techniques, whether or not they use GM mosquitoes, have in common:

– less of an impact, in principle, with regard to environmental and health risks, since they should not affect the density of mosquito populations. An assessment of the risks associated with the induced modification is still necessary;

– persistence and varying invasiveness of the modifications induced, with the need to consider the evolution and long-term effects of the factors responsible for these modifications (Wolbachia, transgenes), including their potential for transfer to other species.

Self-limiting techniques, whether or not they make use of mosquito release, and whether or not the released mosquitoes are GM, have in common:

– the advantage of being controllable and adjustable in the light of monitoring data;

– the drawback of calling for demanding maintenance in the long-term.

Self-sustaining techniques, whether or not they use GM mosquitoes, have in common:

– the advantage of not calling for maintenance or large-scale infrastructure;

– the drawback of being fairly inflexible, or even without the possibility of control (e.g., of intended spread affecting a whole species).

Complementarity of existing and emerging vector control techniques is well illustrated in Fig. 2, which represents the efficiency of the techniques depending on target population density.

Fig. 2 illustrates that:

– the efficacy of conventional methods of vector control is independent of density beyond a certain density threshold of the target mosquito population. Below this threshold, it declines with density until it is nullified before it can lead to elimination;

– conversely, reduction techniques such as the sterile insect technique SIT and the derived techniques such as RIDL and IIT can only be effective below a certain density threshold of the target mosquito population (depending on the ratio of released males to wild males and on the competitiveness of the males released in comparison with wild males). Beneath this threshold, they are all the more effective when the density is lower, thus leading to local elimination (referred to as “Eradication” on the figure) of the population.

These different context-dependent efficacy profiles for the various vector control techniques mean that compatible, complementary techniques ought to be combined in an integrated vector control approach.

As of now, gene drive techniques are still under development. Additional research is required before considering any field application, including reducing development of resistance, developing knowledge and procedures for assessing the long-term effects of gene drive on ecosystems, and strategies for controlling the spread of gene drive.

Self-limiting sterile insect techniques (SIT, RIDL, IIT) could be tested step by step on a precautionary basis for the purpose of contributing to vector control in French territories, depending on the vectors concerned, in combination with the conventional techniques currently used for integrated vector management. If successful, employing IIT, SIT or the RIDL technique would in particular help reduce insecticide use. In addition to a lesser risk of exposure for humans and ecosystems, lower insecticide use owing to the use of techniques based on mosquito release would preserve insecticide efficacy by lessening pressure for selection of resistance. This would thus enable insecticide use to be reserved specifically for epidemics and public health emergencies.

More generally, the choice between different existing and emerging vector control techniques or combinations of techniques should be informed by the intended objective, by vector biology and behavior, and by the epidemiological, environmental, and socio-economic context.