Opinion: To reduce the global burden of human schistosomiasis, use ‘old fashioned’ snail control

Abstract

Control strategies to reduce human schistosomiasis have evolved from ‘snail picking’ campaigns, a century ago, to modern wide-scale human treatment campaigns, or preventive chemotherapy. Unfortunately, despite the rise in preventive chemotherapy campaigns, just as many people suffer from schistosomiasis today as did fifty years ago. Snail control can complement preventive chemotherapy by reducing the risk of transmission from snails to humans. Here, we present ideas for modernizing and scaling up snail control, including spatiotemporal targeting, environmental diagnostics, better molluscicides, new technologies (e.g. gene drive), and ‘outside the box’ strategies such as natural enemies, traps, and repellants. We conclude that, to achieve the World Health Assembly’s stated goal to eliminate schistosomiasis, it is time to give snail control another look.

Targeting snails is a key to success for schistosomiasis control

Soon after Japanese researchers resolved the schistosome life cycle and identified its snail hosts in 1913, Japan launched a ‘snail picking’ effort that offered children a 0.5-yen bounty per container of snails they collected and destroyed [1]. After seven years, Japan shifted from this labor-intensive (and ineffective) effort [1], to controlling snails by cementing irrigation canals, draining wetlands, and applying molluscicides. By 1994, this sustained snail control effort plus drug treatment of infected people, led to the eradication of schistosomiasis in Japan [2]. Other countries, such as Guadeloupe, Iran, Iraq, Lebanon, Martinique, Morocco, Oman, Puerto Rico, Saint Lucia, Saudi Arabia, Tunisia, and Venezuela, have also controlled or eliminated schistosomiasis using snail control [3] (Table 1). Brazil, China, Egypt, Indonesia, the Philippines, and Zanzibar have long used snail control alongside preventive chemotherapy and other strategies to suppress schistosomiasis prevalence, whereas countries that have not pursued snail control have been less successful [3]. Snail control appears to be a key intervention needed to achieve the World Health Assembly’s stated goal to eliminate schistosomiasis [3, 4] (Table 1).

Table 1.

Outcomes and control strategies of all national schistosomiasis control programs during the past century.^a

Country/ territory name	Control outcome	Successful prevalence reduction (%)	Successful population- at-risk reduction (%)	Molluscicides	Snail habitat reduction	Snail biological control	Human tx by MDA	Other controls	Details on other control measures	References
Iran	E	100	100	•	•	•		•	land reclamation, draining swamps, providing latrines	[100, 101]
Japan	E	100	100	•	•			•	human test and tx; cementing irrigation canals	[1, 2, 102, 103]
Jordan	E	100	100	•				•	low coverage; human test and tx	http://www.who.int/schistosomiasis/resources/EMRO_report_Schistosomiasis.pdf
Lebanon	E	100	100	•	•			•	cementing irrigation canals; improved water and sanitation	[101]
Martinique	E	100	100			•		•	improved water and sanitation	[104]
Mauritius	E	100	100	•					some enigmatic snail declines	[105]
Morocco	E	100	100	•	•			•	mobile envoys to support human test and tx	[21, 106–112]
Puerto Rico	E	100	100	•		•		•	human test and tx; improved water; health education	[113]
Tunisia	E	100	100	•				•	human test and tx	[114]
Saudi Arabia	N	99.8	ND	•	•		•	•	little MDA; land reclamation, cementing canals	[23, 115–120]
Indonesia	N	99.5	ND	•			•	•	little MDA; agro-engineering; sanitation improvement	[121]
Iraq	N	99.5	ND	•				•	targeted tx of schoolchildren in early years	[120, 122]
Egypt	N	99	ND	•	•		•	•	safe water supply and agricultural drainage projects	[101, 120, 123–128] http://documents.worldbank.org/curated/en/756051468036566715/pdf/444660PPAR0P0051520Box327410B01PUBLIC1.pdf
China	N	98.9	79	•	•		•	•	tx of cattle and buffalo; agricultural engineering and safe water	[129, 130]
Venezuela	N	98.6	ND	•	•			•	human test and tx; invasive competitor snails	[22, 131–133]
Philippines	N	98.3	ND	•	•		•	•	shift from early focus on snails and targeted tx to later MDA	[134, 135]
St. Lucia	N	98.2	84.3	•	•	•		•	site of early Rockefeller studies on. control of schistosomes	[136]
Guadeloupe	N	96	ND	•	•	•		•	bridges; canal engineering to reduce snail habitat	[133, 137]
Zanzibar	N	84.7	ND	•	•		•	•	recent initiation of elimination program	[138–140]
Laos	N	84.6	ND				•	•	limited health education and improved sanitation	[141–143]
Cambodia	N	83	90				•	•	limited health education and improved sanitation	[141, 144]
Brazil	N	80	69	•			•	•	early human test and tx; improved water and sanitation	[145–148]
Burundi	N	74.4	ND	•			•	•	focal mollusciciding and sanitation were implemented rarely	[149–151]
Ghana	N	73.9	ND				•		low coverage	[152]
Madagascar	N	73.8	ND	•			•	•	molluscicides used rarely; improved water and sanitation	[153, 154] http://www.who.int/neglected_diseases/preventive_chemotherapy/PCTNewsletter12_En.pdf?ua=1, http://digitalcollections.sit.edu/isp_collection/1675/, http://apps.who.int/medicinedocs/pdf/whozip48e/whozip48e.pdf
Rwanda	N	69.5	ND				•	•	mapping and training health personnel	[155]
Libya	N	66.7	ND	•	•			•	human test and tx	http://www.who.int/schistosomiasis/resources/EMRO_report_Schistosomiasis.pdf
Surinam	N	61.5	69.3	•	•			•	human test and tx; education	[156, 157] http://www.who.int/schistosomiasis/resources/PAHO_report_Schistosomiasis_carribean.pdf
Burkina Faso	N	61.2	ND				•			[29, 158, 159]
Tanzania	N	60	0	•			•	•	limited mollusciciding and human test and tx	[160, 161]
Uganda	N	55.4	ND				•			[162, 163]
Mali	N	51.8	ND				•			[41, 164]
Sierra Leone	N	51.4	ND				•			[165, 166]
Kenya	N	51	ND			•	•		limited biological control in early years with Louisiana Crayfish	[74, 167]
Niger	N	44	ND				•			[168–171]
Yemen	N	44	ND				•			[172]
Malawi	N	43.3	ND	•	•		•	•	limited experimental mollusciciding and biological control	[173–176]
Congo	N	41.7	ND	•			•	•	limited mollusciciding, education and health training	http://apps.who.int/medicinedocs/pdf/whozip48e/whozip48e.pdf
Togo	N	30.9	ND				•			[162, 177]
Mozambique	N	28.9	ND				•			http://www3.imperial.ac.uk/schisto/wherewework/mozambique/mozambiquestrategy
Zambia	N	26.6	ND				•			[178, 179]
Cameroon	N	16.7	ND				•			[180]
Swaziland	N	9.6	ND				•			[181]
Senegal	N	1	ND		•		•		limited trials for biological control with river prawns	[9, 182] http://apps.who.int/iris/bitstream/10665/65978/1/WHO_CDS_CPC_SIP_99.2.pdf
Oman	N	0.6	ND	•				•	low coverage of control; human test and tx; concrete reservoirs	http://www.who.int/schistosomiasis/resources/EMRO_report_Schistosomiasis.pdf
Benin	N	−2	ND				•	•	improved water and sanitation (incidental to development)	http://www.evaluategroup.com/Universal/View.aspx?type=Story&id=153562
Somalia	N	−24	NA				•		low coverage of control	http://www.who.int/neglected_diseases/preventive_chemotherapy/databank/en/.
Sudan	N	−29.7	47				•		low coverage of control	[183–187]
Central AfricanRepublic	N	−58	ND				•		low coverage of control	http://apps.who.int/iris/bitstream/10665/69740/1/WHO_CDS_NTD_2007.4_eng.pdf
DominicanRepublic	N	ND	ND	•	•				lacking quantitative data on success	[13, 157, 188]
Syria	N	ND	ND					•	early human test and tx; cementing canals; conflict hinders effort	http://www.who.int/schistosomiasis/resources/EMRO_report_Schistosomiasis.pdf

^aThe most successful programs have focused on widespread and early snail control activities. Countries and territories are ordered by most to least successful in terms of reducing schistosomiasis prevalence. More details and data are available at [3], http://schisto.stanford.edu and https://purl.stanford.edu/yt060bn1019. “E” = schistosomiasis eliminated from the country/territory, “N” = schistosomiasis is not yet eliminated, “ND” = no data, “tx” = treatment campaigns.

Despite these many successes, the modern orthodoxy paints snail control as old fashioned, preferring to focus instead on preventive chemotherapy via mass drug administration (MDA) of praziquantel [5–7]. Praziquantel’s introduction in the late 1970s and early 1980s, and the release of its generic form in the 1990s, led the World Health Assembly to adopt, in 2001, preventive chemotherapy as the recommended global strategy for schistosomiasis reduction [7, 8] (http://apps.who.int/gb/archive/pdf_files/WHA54/ea54r19.pdf). This is in line with recent emphasis on integrated preventive chemotherapy (distributing drugs against various preventable diseases). But despite distributing millions of pills in recent decades, sub-Saharan Africa’s schistosomiasis problem is as serious now as it was before praziquantel’s discovery, in part because reinfection after treatment can thwart long-term control [3]. Given this disappointing outcome, the World Health Assembly’s 2012 resolution 65.21 advocates adding modernized snail control and other control methods to preventive chemotherapy in order to achieve schistosomiasis elimination (http://www.who.int/neglected_diseases/mediacentre/WHA_65.21_Eng.pdf).

Together, preventive chemotherapy and snail control techniques offer our best opportunity for schistosomiasis elimination – and current technology for snail control has come a long way from snail picking. Here we argue it is time to refocus on snail control in the fight against schistosomiasis. We discuss which strategies remain relevant, and propose what future snail control might look like.

Snails and the schistosome life cycle

The schistosome life cycle encompasses two transmission processes: human-to-snail transmission and snail-to-human transmission (Fig 1). Schistosome eggs from human urine or feces reach fresh water, where eggs hatch and release the miracidia larvae that infect freshwater snails. After completing asexual reproduction in the snails, the schistosomes then release free-swimming cercariae that penetrate human skin, eventually migrating to the portal or pelvic veins, depending on the schistosome species.

Figure 1. The Schistosoma sp. lifecycle and snail control strategies.

human to snail transmission occurs via free-living miracidia released from eggs in urine and feces; and snail to human transmission occurs through free-living cercariae that exit infected snails into the water, seeking new vertebrate hosts. Control strategies should combine (A) human drug treatment or preventive chemotherapy with praziquantel (PZQ) with (B-F) creative methods to control infected snails such as: (B) chemical molluscicides; (C) natural enemies; (D) habitat modification; (E) creative technologies such as gene drive; (F) traps or repellants and other out of the box strategies.

It is not easy to eradicate snails, but snail eradication is not necessary for the elimination of schistosomes. To break the schistosome life cycle, snail densities must be driven below a threshold where snail infection rates are lower than snail death rates [9]. Schistosome and snail compatibility is complex and there are many strain differences across the world [10, 11], but despite this complexity, the simple fact remains that where schistosome-susceptible snails have been reduced schistosomiasis has often been eliminated from large areas (even whole countries). In Japan, where schistosomiasis has been eliminated since the 1990s, the snail intermediate host, Oncomelania nosophora, persists to this day – although its abundance is low enough to merit a vulnerable ranking on the International Union for Conservation of Nature (IUCN) red list (http://www.iucnredlist.org) [12]. In Guadeloupe, where snail reductions interrupted schistosomiasis transmission (with few to no documented cases during the past several decades [13, 14]), Biomphalaria glabrata intermediate host snails are still present and still susceptible to infection, at least up to 2005 [15]. The recalcitrance of snails to eradication means that snail control must be deployed with other approaches to reduce the chance that infected humans will re-introduce the parasite.

‘Old-fashioned’ snail control has included chemical molluscicides, habitat modification, and biological control, but modern methods could add ‘outside the box’ strategies – including some under development or yet to be devised. Given that snail populations persist, these snail control interventions are best complemented with traditional, human-centric schistosomiasis control strategies, like human drug treatment (such as mass drug administration or targeted testing and treatment), water, sanitation, and hygiene infrastructure (WASH), or behavior modification through education. Snail control – or any environmental intervention that reduces schistosomiasis transmission and slows reinfection after treatment – should decrease the frequency at which preventive chemotherapy is required and thus would spare drugs, increase MDA efficacy, reduce costs, and improve scalability. Simply put, elimination is possible if human infections can be interrupted via preventive chemotherapy, and snail densities can be reduced (Fig 1).

Looking back

Effective ‘old-fashioned’ snail control strategies have included chemical molluscicides, reducing snail habitat, and biological control (i.e., intentional or unintentional introductions of competitor snails or snail predators) and snail control has sometimes been combined with a number of other strategies including human mass drug administration (MDA), human testing and treatment campaigns, and engineering interventions (Table 1).

Success with chemical molluscicides

During the 20^th century, molluscicides were among the most commonly used snail control strategies by governments and public health agencies, but molluscicides fell out of favor as costs of the chemicals increased, and concurrently, the cost of praziquantel fell, beginning in the 1990s [3]. Although the environmental impacts associated with chemical applications limit their acceptability in some circumstances, molluscicides have been effective in controlling schistosomiasis [3, 4, 16]. Since the 1960’s, the most-used chemical has been niclosamide, a formulation with lethal effects for snails up to 24 hours after application and low lethal concentration (LC90) for snails, at <1ppm [4]. In theory, these low concentrations are non-toxic to vertebrates including fish and humans, but uneven dispersal can lead to fish kills and health concerns [17]. Some countries have imposed restrictions on the use of niclosamide in the environment due to health concerns and to concerns regarding its non-target effects [18]. However, it is interesting to note that niclosamide has been approved for many decades as an anthelminthic treatment in people and has recently been explored as an anti-cancer therapy and a treatment for Zika virus [19, 20].

Success with snail habitat modification

Snail habitat modification for schistosomiasis control has taken several forms – including vegetation removal, land reclamation (e.g. wetland drainage), cementing canals, and occasionally, hydrological interventions to increase or alter stream flow (Table 1). For example, these strategies have controlled snails in Japan, Morocco, Saudi Arabia and Venezuela [2, 21–23]. In contrast, habitat changes linked to dam construction, irrigation expansion, and other water-related changes have resulted in unintentional and sometimes dramatic schistosomiasis outbreaks [24, 25].

Success with biological control

Schistosome-transmitting snails have various natural enemies. Some crustaceans, birds and fishes eat them. Other snail species compete with them. Non-schistosome trematodes castrate them. Such natural enemies can regulate snail populations, but most enemies have limited natural ranges, and could have non-target effects where they are non-native. Biological control has a bad reputation for non-target effects – but this stems from a few examples where spectacular collateral impacts have accompanied ill-conceived strategies [26]. Biological control can be both safe and effective in a modern context, especially when native species that are natural enemies of pests are used [9, 27].

Many biological control strategies have been researched for schistosomiasis control (for example, introduction of predators, competitors, and parasites of snails), but few strategies have been used widely in practice. One exception is the widespread use of competitor snail species that are not competent hosts for schistosome infection in Caribbean countries such as Antigua, Guadeloupe, Martinique, Montserrat, Puerto Rico, and St. Lucia; non-competent snails were introduced and successfully displaced schistosome-competent intermediate host snails. Schistosomiasis control has been pursued through snail introductions with species such as: Pomacea glauca, Marisa cornuarietis, Melanoides tuberculata, or Tarebia granifera [3, 13]. Displacement can be long-lasting if competitor snail populations are self-sustaining [13, 28].

No one-size-fits-all solution

No single strategy will reduce schistosomiasis transmission everywhere. For example, past attempts at widespread biological control using snail competitors worked to eliminate schistosomiasis on some Caribbean islands but not others [28]; and mass drug administration using praziquantel has durably reduced schistosomiasis in some parts of Burkina Faso but not others [29]. What worked well in one place or time can be ineffective or inappropriate in another. Deploying multiple strategies may help to balance the control portfolio. In particular, snail control is likely to be synergistic with traditional drug distribution campaigns employed in preventive chemotherapy and other well-established interventions like WASH infrastructure improvements, education, and sustainable development.

Looking forward

Future snail control strategies should build on past successes while responding to changing conditions and incorporating modern technologies. History has shown that controlling complex life-cycle parasites, like Schistosoma spp., requires interrupting transmission from humans to intermediate hosts and vice versa. Embracing a synergistic approach might deliver lasting disease reductions beyond those achievable by focusing on any single aspect of transmission [9]. Public health, conservation and sustainable development goals could be aligned if health interventions capitalize on co-benefits – as has been suggested, for example, in recent studies that focus on complementing human drug treatment with species restoration (of snail predators) to reduce snails, control schistosomiasis transmission, alleviate poverty, and restore ecosystems [9, 30–32].

Schistosomiasis, today, is linked to poverty [33, 34] and the long time course required to reduce or eliminate schistosomiasis can erode interest by philanthropic organizations and individual donors [3]. Economic sustainability therefore remains a pressing concern for the future of schistosomiasis control.

For snail control, cost-effectiveness could benefit from strategic improvements such as: i) targeting control to where and when most transmission occurs to increase effectiveness while reducing coverage needs (e.g. considering hubs and hotspots of transmission in space and time), ii) using complementary natural enemies (e.g., predator ducks and their echinostome trematodes) that offer affordable win-win solutions that simultaneously reduce schistosomiasis and generate revenue or other co-benefits, iii), applying novel technologies to improve snail management and control (such as gene drive), iv) discovering molluscicide formulations that are less harmful and more sustainable, and finally, v) integrating snail control with other available tools, including preventive chemotherapy, education, and sanitation.

Understanding the landscape of schistosomiasis infection risk: ecological surveillance, network theory, and optimal control

Snail populations and their schistosome parasites can be dynamic and difficult to predict at the spatial and temporal scales relevant to control campaigns. Theory and empirical data from other disease systems indicate that strategic timing and spatial distribution of control effort improves the efficiency of control, but little schistosomiasis-specific research on this topic exists [35–39].

Although there are few empirical data on snails and their schistosome parasites, especially for Sub-Saharan Africa where most human schistosome infections occur today, the existing data suggest that schistosome-infected snails have aggregated distributions, so that infection risk is distributed in hotspots [40, 41]. A hotspot might be a particular water access site or village, with infection risk varying from village to village (across tens to hundreds of meters; e.g., [42–44]). Furthermore, water flow can move cercariae away from high densities of infected snails [45], making it harder to pinpoint the source of infection risk to humans.

Planning and assessing the success of snail control requires mapping and tracking snail abundance and infection prevalence, but the most common traditional snail sampling technique is timed snail counts (Box 1). Although useful for evaluating relative risk among sites or across time within a single study, the relative abundance method does not measure absolute risk, which is best expressed as infected snail density (combined with information on the density of cercariae emitted from snails through time, Box 1). The use of relative abundance snail sampling methods has been rationalized by invoking investigator safety, time constraints, and the need for simple, straightforward sampling designs when working in challenging field conditions. Absolute sampling using quadrats – that is, the kind of quantitative invertebrate sampling used in other aquatic habitats [46]–is time-consuming and logistically challenging, but yields a more useful, quantitative measure of snail abundance.

Box 3. Quantifying snails and their trematode infections.

Researchers define and measure human risk for schistosomiasis transmission in several ways: prevalence or density of infected snails assessed through snail surveys [189, 190], cercarial density measured via cercariometry or molecular detection in water, [191, 192], or density of infective cercariae derived through mouse exposure [193]. Because snail densities vary widely, the prevalence of infected snails is a poor way to estimate infection risk in humans. Infection rates in sentinel mice are the most direct way to measure risk in humans. However, mouse exposures are expensive and pose ethical concerns to some [192]. Next best is cercarial density, but filtering for cercariae is challenging because waters are often turbid, cercariae have short lifespans (hours), are small, and have soft bodies [194]. Newly available environmental DNA sampling still requires ground truthing and cannot distinguish cercariae (infective to humans) from miracidia (infective to snails). Therefore, snail sampling via timed searches (e.g. [9, 52, 59]) has been by far the most common way to measure risk in research studies and monitoring efforts for the last several decades. In a traditional search, trained technicians spend a set time (e.g., 15 minutes) searching for potential snail habitat at water access points, then agitate the substrate or vegetation with fine-mesh scoops (~ 1–2 mm mesh size, pictured) – and retrieve the scoops and pull out the snails [194] (Figure I). Collected snails are put in vials under bright light to shed cercariae, which can then be identified and used to estimate which snails are infected[194]. Such timed searches are quick and inexpensive, but by targeting snail habitat, the actual measure probably reflects snails density within snail habitat rather than overall snail density, perhaps explaining why many past studies conclude that infected snail density at a site does not correlate well with measures human infection rates [195, 196]. On the contrary, studies using systematic or random quadrat sampling (including dissecting snails to examine for trematode infections) have found more robust correlations between infected snail density and human infection rates [197]. Future work should aim to develop cost-effective and accurate ways to assess infection risk.

In recent years, advances in molecular genetic techniques, spatial statistics, and GIS mapping have made it possible to examine schistosomiasis transmission risk at fine-grained spatial scales [198]. These technologies, coupled with more robust and spatially quantitative snail sampling techniques – borrowed from ecological studies on freshwater invertebrates (e.g. [46, 197]) – could improve prediction capabilities for schistosomiasis transmission.

Figure I.

Two different snail scoops designed and deployed to sample snails in schistosomiasis transmission sites in Senegal. Image courtesy of The Upstream Alliance (http://wwwtheupstreamalliance.org), under the terms of the Creative Commons Attribution License CC BY 2.0.

In addition to improved methodologies to assess snail abundance and to sample transmission stages, species distribution models (habitat suitability models), environmental DNA, network models, and optimal control theory might improve current snail sampling efforts. Some indirect sampling methods might become cost-effective with additional refinement. For example, species distribution modeling [47, 48] encompasses various methods to correlate species occurrences to underlying habitat variables, such as temperature, rainfall, vegetation cover, etc. This technique could help generate maps that predict schistosomiasis transmission hotspots using readily available data, like land features and environmental variables [49]. For example, recent reviews [50, 51] concluded that spatial risk profiling for schistosomiasis using remotely sensed data is an under-used strategy in schistosomiasis research and control. Species distribution modeling might be particularly effective where strong seasons lead to dramatic snail-habitat ephemerality that is easily mapped, as in Burkina Faso and Cote D’Ivoire [52, 53]. These models still require ground truthing using environmental data for training and validation. An alternative indirect approach is to use environmental DNA (eDNA) to track snail density or parasite presence by detecting genetic material directly from water, soil, or other environmental samples without evidence of their biological source [54, 55]. The eDNA technique also requires more refinement and validation [54], especially before it can be calibrated for quantitative assessments. Furthermore, because schistosome eDNA might arise from DNA in living or dead miracidia or living or dead infectious cercariae, it might be hard to translate an eDNA signal to infection risk.

Schistosomiasis transmission maps onto where people work, live, and travel. Understanding the spatial and seasonal connectivity among snail and human populations (e.g. through network modeling, which tracks populations and their interconnections) could indicate critical links where control would be most effective. For example, targeting snail control based on identification of villages that are important hubs of transmission could reduce costs and improve scalability [37].

In Senegal, network models including human mobility – tracked through mobile phone records – predicted schistosomiasis prevalence better than models assuming homogenous mixing of people across cities and villages [37]. Ciddio et al. showed how a network model tracking human mobility and water-mediated snail and cercarial dispersal could be used to target environmental interventions to reduce human exposure and contamination risks [56].

In addition to network modeling, there is little published work on how to apply optimal control theory to neglected tropical diseases, including schistosomiasis. Yet, this approach, which is often used in optimization problems from engineering and economics [57] and more recently from biology and epidemiology [58], could provide a platform to tackle schistosomiasis transmission control, considering a complex landscape of competing costs and benefits [52, 59]. By incorporating economic considerations in the form of a cost function and considering control strategies that can vary continuously through time along an optimal path (rather than an “either or” or a “one size fits all” approach), these models could offer insight needed for ecosystem-specific decision-making on complex trade-offs in health, economic, and environmental factors influencing the management and control of schistosomiasis.

Future molluscicide formulations

New molluscicides (or new niclosamide formulations) that are safer, more effective, more specific, or that disperse more evenly would be beneficial in the fight against schistosomiasis. For example, some promising research areas include: slow-release niclosamide formulations [16], extracts from molluscicidal plants such as endod and others [4, 16, 60], and surfactant formulations that help disperse niclosamide or other molluscicides more evenly, delivering snail-killing efficacy with less opportunity for accumulating unsafe concentrations. Although some of these strategies have been investigated at small scales for many decades (e.g. molluscicidal plants), the investment of time, energy, and funding has not yet been sufficient to allow scale-up [61]. Understanding the spatiotemporal heterogeneity in snail and trematode abundance, as discussed above, could contribute to better targeting of molluscicide applications in space and time, and improve safety, efficacy, and cost-effectiveness for this historically successful, chemical-based snail control strategy.

Gene drive technologies for snail control

We might soon engineer snail hosts with new genetic properties similar to gene drive engineered malaria-resistant Anopheles gambiae mosquitos [62]. In 2016, a CRISPR-Cas9-based gene drive was used to insert genes conferring sterility to female A. gambiae mosquitos, revealing the potential for gene drive technologies to reduce malaria transmission [63]. Despite the fitness costs to the mosquitos that result from sterility-inducing genes, the gene drive system successfully increased the allele frequency of these genes in lab-reared populations over six generations.

The CRISPR-cas9 gene drive system deserves to be explored as an avenue to schistosomiasis control. Some barriers to employing this technology have already been surmounted: genes that confer schistosome resistance have been identified in wild snail populations [15]; the Biomphalaria glabrata genome has been sequenced [64] and CRISPR-cas9 gene editing has been carried out in a marine gastropod [65]. However, a caveat is that Biomphalaria and Bulinus spp. snails (but not Onchomelania spp.), are hermaphroditic and can self-fertilize, making gene drive systems for population suppression more challenging, because drives intended to suppress population growth might lead to compensation by the wild-type snails in the form of more asexual reproduction (selfing) [66]. Gene drives that confer schistosome resistance are an alternative strategy, but seem limited in application given that existing resistance genes do not spread to fixation in host snails [67] (presumably due to associated fitness costs of resistance in uninfected snails). Though it is often implied to be highly precise, CRISPR-cas9 gene editing can produce off-target mutations with unpredictable effects so more work is required to ensure safety of releasing gene-edited snails into the wild [68]. Ethical limitations and methodological hurdles notwithstanding [69], the potential for this new gene drive technology to revolutionize control for human disease, including schistosomiasis and other vector-borne and environmentally transmitted diseases, is tantalizing, so long as safety, efficacy, and implementation constraints can be surmounted.

Thinking outside the box: traps, repellants, and natural enemies

Attempts to trap and kill snails or schistosomes emitted from snails, or repel them from humans, have not yet been applied widely in practice, but such ‘outside the box’ strategies could prove useful if new technologies make them more effective, feasible, or scalable. For instance, snails are attracted to lettuce homogenates (specifically, the amino acids glutamate and proline [70]) and wheat germ cereal [71] which could be used to bait traps. Snails can be repelled by molluscicides [71], artificial shade [72], and topical lipid formulations of N,N-Diethyl-meta-toluamide (DEET) applied to exposed skin [73].

Snail predators – particularly crustaceans, fish and birds – have been effective at reducing snail populations in the past, warranting more research to develop and scale-up the use of snail predators for disease control. For example, Louisiana crayfish (Procambarus clarkii) introduced to Kenya and Egypt can reduce snail abundance and therefore human schistosomiasis transmission [3, 74, 75]. More recently, native river prawns have been proposed as snail control agents in their native coastal ranges, where human-driven environmental change (e.g. dam building) has reduced prawn numbers [24, 32]. Dams are associated with greater increases in human schistosomiasis risk within river prawn native ranges than outside them, suggesting that prawns might have once controlled snail populations [24]. Indeed, reintroducing native river prawns (Macrobrachium volenhovenii) into Senegalese water access points – where they had been present before the nearby Diama Dam was built [32] – resulted in a reduction in snail density and human schistosomiasis re-infection rates [9]. In theory, prawn ladders designed to help juvenile prawns surmount dams could help restore river prawn migration pathways [76]. Other crustaceans might suppress snails and thus schistosomiasis transmission. For example, the Malaysian river prawn, Macrobrachium rosenbergii – in the same genus as the African river prawn – also eats schistosome-hosting snails [77]; unlike the African-native, M. rosenbergii is domesticated, and could therefore be deployed as a biological control agent in managed landscapes [9, 31, 77].

Some fish eat snails [78, 79]. The observation that fish might control snails has inspired efforts to use fish as a biological control tool, with mixed results [80]. However, one snail-eating cichlid, Trematocranus placodon, has shown promise [78], as has the African catfish, Clarias gariepinus [81].

With respect to birds, non-native, domestic ducks reduce snail density in Zimbabwean ponds, but present many logistical challenges – including high costs for duck breeding, maintenance, and protection against poaching [82]. Another role for birds might be in the trematodes they carry. Non-schistosome trematodes that use birds as final hosts, such as Ribeiroia guadeloupensis, castrate host snails and outcompete schistosomes inside infected snails [83], and other similar trematode species have been investigated for similar applications [84].

Competition with other species can suppress snails or schistosomes. Past schistosomiasis control strategies have been successful in using competitor snails and this strategy could be revisited for deployment in modern schistosomiasis hotspots (see Looking Back section). In addition, schistosome species are outcompeted in their snail hosts by other trematode species that produce rediae – jawed reproductive structures that can kill sporocysts [85]. Indeed, many echinostome species including Echinostoma spp.[86, 87] as well as Exorchis sp. [88], Cotylurus lutzi [89], paramphistomoids [90] and others have been investigated for this purpose. However, other trematode species might facilitate schistosome infection, possibly by reducing the host’s immune defenses; evidence for this comes from Calicophoron microbothrium [91] and Zygocotyle lunata [92]. Such differences must be well understood before deploying trematodes as natural enemies. ‘Decoy hosts’ – non-competent snails and other aquatic organisms, such as fish and amphibians – absorb schistosome miracidia without becoming infected, potentially diverting miracidia from competent snail hosts and reducing infected snail prevalence. Though this effect has been observed in laboratory [93] and meso-cosm experiments [94] its success in scaled up control programs has not yet been demonstrated. The parasites’ free living stages also have predators that consume them directly (such as Chaetogaster spp., filter feeders, and small fish [95]); the use of trematode predators in schistosomiasis biological control is beyond the scope of this paper but remains an interesting and relatively unexplored alternative strategy that may – in some instances – complement snail control for schistosomiasis reduction.

Concluding Remarks

“Without snails, there can be no schistosomiasis.” This quote, from the World Health Organization Working Group on Schistosomiasis in 2005 (http://apps.who.int/iris/bitstream/10665/69482/1/TDR_SWG_07_eng.pdf) represents a necessary but insufficient assessment. Indeed, where the snail intermediate hosts for schistosome parasites cannot persist, there is no opportunity for schistosomiasis transmission, but even where snails and schistosomes co-exist, schistosome transmission might not be successful. Therefore more ecological research on schistosome-hosting snails and the conditions permissive to schistosome transmission seems warranted.

For the past century, snail control has been successful in reducing schistosomiasis transmission in many countries, but has fallen out of favor in the last few decades. Here, we have discussed how both new and old fashioned snail control technologies can be used to reduce the risk of schistosome transmission from snails to humans, but many questions remain unanswered (see Outstanding Questions box). We presented some ideas for modernizing, improving, and scaling up snail control, such as spatial targeting, temporal targeting, gene drive technologies, affordable environmental diagnostics, and outside the box strategies such as traps, repellants, natural enemies, and decoys. The goal of snail control is to reduce transmission. This can be maximized by better synergy between mass drug administration and environmental interventions that affordably slow human reinfection after treatment. A synergistic approach spares drugs and likely improves efficacy, cost effectiveness, and scalability.

Most of the two and a half billion dollars disbursed each year to treat and control neglected tropical diseases [96, 97] is directed toward mass drug administration. Although treatment has been effective, control has not, because there is not enough praziquantel to reach all 800 million people at risk today and drugs, alone, cannot address the environmental components of transmission [98, 99]. Coupling drug delivery with snail control has proven effective in the past, and seems the most cost effective option for the future global fight against schistosomiasis.

Trends Box.

• Despite a rise in the global effort towards preventive chemotherapy, just as many people suffer from schistosomiasis today as did fifty years ago

• Snail control can complement medical treatment, especially where transmission is endemic and reinfection after treatment is a common occurrence

• It is time to give snail control another look

• Modernizing snail control is a priority and might benefit from more research on spatiotemporal targeting, environmental diagnostics, better molluscicides, new technologies, and ‘outside the box’ strategies such as natural enemies, traps, and repellants

Outstanding questions.

1. Can network analysis of schistosomiasis transmission reveal hotspots and hubs to target for more efficient snail control?

2. At what spatial scale does schistosomiasis transmission occur? Can understanding transmission improve control (i.e., the spatial extent that must be treated to protect humans using a given water body).

3. Might CRISPR-cas9 gene editing and gene drive technologies be a safe and effective way to reduce schistosomiasis-infected snails?

4. Are molluscicides outdated or are there future formulations that could deliver successful snail control with fewer non-target effects?

5. How can natural enemies, repellants, traps, and decoys be used for snail control?

6. What is the most efficient and synergistic use of preventive chemotherapy and environmental controls, including snail control, in the global fight against schistosomiasis?

7. Can environmental DNA (eDNA) technology be used to indirectly track snail or schistosome presence and distribution in the environment?

8. Can optimal control theory contribute to an improved strategy for schistosomiasis elimination?