Introduction
According to the World Health Organization (WHO), dracunculiasis is the first parasitic disease set for eradication (https://www.who.int/dracunculiasis/eradication/en/). Human dracunculiasis has always been considered a water-borne disease. Humans become infected when they drink water contaminated with the small intermediate host, Cyclops or water fleas (copepods), harbouring the infective larvae. After a prolonged prepatent period of approximately a year, the under-skin migrant female of Dracunculus medinensis (Guinea worm) causes, first, a blister and, then, an ulcer, mainly in the lower limbs. When an infected person enters water, through the ulcer, a milky fluid is exuded from the female uterus containing hundreds of thousands of 550–760-μm-long first-stage larvae (Muller, Reference Muller1971). These embryos remain alive for no longer than 72 h in the water. Around 12–14 days (at temperatures above 25°C) after their ingestion by fish, 240–608-μm-long larvae are already infective for humans (Muller, Reference Muller1971). According to WHO, this simple water-borne life cycle, the easy and unambiguous diagnosis and the largely accepted belief that there was no known animal reservoir are among the factors that caused dracunculiasis to be considered an eradicable disease.
Due to the successful Guinea worm eradication campaign, started in the 1980s, WHO aimed at its eradication by 1995, being later postponed to 2015 (WHO, 2007). Approximately 3.5 million people became annually infected in 21 countries of Africa and Asia in the 1980s. Meanwhile, the incidence of dracunculiasis has been reduced by more than 99.99% after the implementation of the eradication programme. Specifically, during 2018, only 28 human cases were reported, 17 in Chad, ten in South Sudan and one in Angola (CDC, 2019a). For the period 1 January–28 February 2019, three cases in Chad and one in Angola were reported (CDC, 2019b). Introducing filters for drinking water to avoid the ingestion of the intermediate host is one of the most essential prevention measures taken, together with the treatment of the water sources with the larvicide Temephos to kill the copepods, as well as education to prevent infected people from contaminating water resources (Hopkins et al., Reference Hopkins, Ruiz-Tiben, Eberhard, Weiss, Withers, Roy and Sienco2018a).
In spite of this successful campaign, recent events related to the D. medinensis epidemiology in certain African countries are apparently posing new challenges to its eradication (Dumiak, Reference Dumiak2018). In fact, WHO postponed Guinea worm disease eradication to 2020 (Galán-Puchades, Reference Galán-Puchades2017).
Guinea worm disease has been reported in writings from India, Greece and the Middle East since antiquity. Female worms have even been found in Egyptian mummies (Greenaway, Reference Greenaway2004). However, more than 3000 years later, two novel facts have emerged in its epidemiology. First, dracunculiasis would have, in addition to humans, other well-established definitive hosts, dogs, but also cats and olive baboons (Papio anubis) have been found infected (Thiele et al., Reference Thiele, Eberhard, Cotton, Durrant, Berg, Hamm and Ruiz-Tiben2018). Second, a new route of infection for humans, or any other definitive hosts, has been suggested: the ingestion of paratenic or transport hosts (mainly fish and frogs, which previously would have ingested infected copepods) harbouring the infective larvae (Cleveland et al., Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017; Hopkins et al., Reference Hopkins, Ruiz-Tiben, Eberhard, Weiss, Withers, Roy and Sienco2018a). This means that dracunculiasis, instead of being an exclusively water-borne anthroponosis, would also be a food-borne zoonosis.
Anthroponosis vs. zoonosis
Recent genetic results revealed that Guinea worms found in naturally infected humans, dogs, cats and baboons in African countries belong to the same population, thus reinforcing the reservoir role of these mammals (Thiele et al., Reference Thiele, Eberhard, Cotton, Durrant, Berg, Hamm and Ruiz-Tiben2018). Considering only Chad, 1040 dogs and 25 cats were affected in 2018, in comparison with the 17 human cases (CDC 2019a).
Although apparently a novel fact, the existence of reservoirs (mainly dogs, but also cats) of D. medinensis has been known since antiquity (Adamson, Reference Adamson1988; Cairncross et al., Reference Cairncross, Muller and Zagaria2002). Concerning disease eradication campaigns, on the whole, it is known that the biggest challenges in eradication programmes usually occur in their final stages (Whitty, Reference Whitty2015). Therefore, there will always be the chance of reservoirs of the disease, which were not anticipated at the start of the eradication attempt. Reservoirs only become apparent when incident cases by the main route of transmission have been reduced (Whitty, Reference Whitty2015). This is precisely the case of dracunculiasis: at the final stage of Guinea worm eradication, the possibility of an unanticipated (mainly) canine host has been raised, just as human cases have dramatically decreased. However, dogs could have been anticipated as definitive hosts having had a look at the literature. Dogs have repeatedly been considered not only a reservoir of the disease but also the best experimental animal model for reproducing human dracunculiasis (Moorthy & Sweet, Reference Moorthy and Sweet1936a, Reference Moorthy and Sweet1938; Muller, Reference Muller1971; Adamson, Reference Adamson1988; Cairncross et al., Reference Cairncross, Muller and Zagaria2002; Tayeh et al., Reference Tayeh, Cairncross and Cox2017). Indeed, researchers have known for decades that dogs and other mammals can acquire D. medinensis. However, these cases were merely considered rare examples of infection, overlooking the reservoir potential of mammals living close to infected people. It was only recently that Guinea worm infection in dogs and cats was reconsidered – that is, it is not a novel host switch as previously believed (Callaway, Reference Callaway2016; Thiele et al., Reference Thiele, Eberhard, Cotton, Durrant, Berg, Hamm and Ruiz-Tiben2018). In fact, the prevalence of dracunculiasis in people and dogs in Old Bukhara, Uzbekistan, from 1926 to 1931 – presented in graph 5 in Cairncross et al. (Reference Cairncross, Muller and Zagaria2002) – showed the relevant Guinea worm incidence in dogs in that city. However, in Uzbekistan, dog prevalence was only a fraction of that in humans, whilst in Chad the canine cases make up the vast majority and, apparently, emerged after there had been no human cases for a decade. Nonetheless, the comparison between these two scenarios is difficult from an epidemiological point of view, considering the total absence of reliable information on dog dracunculiasis in Chad in the years in which humans showed the highest prevalence.
Water-borne vs. food-borne route of infection
The food-borne route of dracunculiasis infection is considered novel in Guinea worm epidemiology (Cleveland et al., Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017; Thiele et al., Reference Thiele, Eberhard, Cotton, Durrant, Berg, Hamm and Ruiz-Tiben2018). However, the presence of D. medinensis infective larvae in fish has probably been known since, at least, the last century (Moorthy, Reference Moorthy1935). Related to this transmission route, D. medinensis is on the verge of eradication due to the control measures exclusively aimed at the water-borne route. Therefore, food-borne transmission is probably only of secondary importance in the epidemiology of the disease, at least in humans. In fact, copepod-eating fish were used as biological control in step-wells in India with remarkable success. Dracunculiasis was eradicated even though people caught those fish for consumption when the water was low (Moorthy & Sweet, Reference Moorthy and Sweet1936b).
Then, why has this alternative route of infection emerged after almost 40 years of a successful eradication campaign? Food-borne transmission is the explanation scientists gave to the large number of parasitized dogs found in Chad villages along the Chari River, whilst only a few sporadic infections were detected in humans in the period 2012–2016 (Callaway, Reference Callaway2016). This situation has been explained by the possible Guinea worm transmission through the ingestion of aquatic paratenic or transport hosts such as frogs and fish (Chippaux & Massougbodji, Reference Chippaux and Massougbodji1991; Cleveland et al., Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017; Hopkins et al., Reference Hopkins, Ruiz-Tiben, Eberhard, Weiss, Withers, Roy and Sienco2018a). The existence of this food-borne route had already been suggested by Caincross et al (Reference Cairncross, Muller and Zagaria2002) to explain isolated autochthonous human cases in Japan, Korea and Indonesia by means of the ingestion of freshwater fish as paratenic hosts.
At its beginning, Guinea worm eradication (aimed at interrupting water-borne transmission) was exclusively addressed at humans. Since dogs, as well as cats and baboons, do not drink filtered water, transmission may effectively continue among these animals by the water-borne route, but not in humans as long as they continue to drink safe water. The possible presence of infected fish and frogs means that there must be infected copepods in the water. Therefore, dogs (or cats or baboons) can also become infected ingesting infected cyclops whilst drinking. However, it has been suggested that dogs (or other animals that ingest water by lapping) do probably not get infected when drinking water as their lapping scares away copepods and, in addition, infected copepods are sluggish and tend to sink to the bottom of the water source, whereas animals drink near the water surface (Muller, Reference Muller1971).
Dracunculiasis transmission depends on the characteristics of the water source which, in turn, has an important consequence for the design of the eradication programme (Cairncross et al., Reference Cairncross, Muller and Zagaria2002). It is known that dracunculiasis is not transmitted in flowing water sources such as rivers or streams. It has been suggested that transmission does not occur when there are fewer than one susceptible copepod per litre in the pond (Chippaux & Massougbodji, Reference Chippaux and Massougbodji1991). In dry areas in sub-Saharan West Africa, peak transmission occurs at the onset of the rainy season. Farmers are then planting their crops and drinking water is collected from rain-filled pools in the fields. Transmission decreases at the end of the rainy season when these ponds dry out. However, in the more humid savanna areas of West Africa, transmission occurs mainly during the dry season. In the rainy season, there are so many places with surface water that transmission is less likely to occur (https://www.who.int/water_sanitation_health/resources/vector324to336.pdf). In the dry season, the drinking water supply is limited to only a few village ponds where conditions for transmission are more favourable. Rivers, like the Chari River, recede in the dry season, and smaller ponds, accessible to reservoirs, appear (Hopkins et al., Reference Hopkins, Ruiz-Tiben, Eberhard, Weiss, Withers, Roy and Sienco2018a).
Infected cyclops become less and less active due to Guinea worm infection. Copepods become practically incapable of moving more than a few centimetres from the bottom of the water source when they already harbour the infective larvae (Onabamiro, Reference Onabamiro1954). In particular, copepods decrease their motility, so that their escaping ability is reduced, which, in turn, favours the completion of the Dracunculus life cycle. Copepods infected with other helminth larvae show the same modified behaviour that facilitates their ingestion (Pasternak et al., Reference Pasternak, Huntingford and Crompton1995; Pulkkinen et al., Reference Pulkkinen, Pasternak, Hasu and Valtonen2000; Franz & Kurtz, Reference Franz and Kurtz2002). When dogs (or cats, baboons, etc.) put their legs into the small ponds they, invariably, stir up the bottom with the probable result that sluggish and less elusive cyclops, containing infective Dracunculus larvae, will even be more easily ingested.
In spite of the heroic efforts made, the level of dog containment to prevent water contamination in Chad was 75% in 2018 (CDC, 2019a). That means that around 25% of infected dogs could continue to contaminate water resources. Considering the Guinea worm's potential reproductive rate, if it is accepted that a single human case can spread the infection to as many as 85 new humans (CDC, 2019a), one dog could spread the infection to as many as 85 new dogs (or humans or any other reservoir). In addition, unlike Mali, South Sudan or Ethiopia, with a chemical control coverage of water resources with Temephos of 90–100%, this coverage in Chad only reaches 24% (CDC, 2019b). Therefore, although traditional drinking water sources are currently not believed to be the primary source of infection (Eberhard et al., Reference Eberhard, Cleveland, Zirimwabagabo, Yabsley, Ouakou and Ruiz-Tiben2016; CDC, 2019b), the water-borne route of reservoir infection is absolutely feasible.
The presence of D. medinensis larvae in the intestines of experimentally infected fish was already reported more than 80 years ago by Moorthy (Reference Moorthy1935). In his experiments, as well as in those carried out by Cleveland et al. (Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017), free infective larvae were observed in the intestines of fish immediately after death. However, it is not known for how long these larvae remain alive and/or infective in dead fish. Considering that larvae do not penetrate fish intestines, dogs could become infected by means of the ingestion of entire raw fish, or their intestines, only shortly after their capture (Cleveland et al., Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017). Even after several surveys, D. medinensis larvae have not been found in fish tissues in nature so far (Cleveland et al., Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017, Reference Cleveland, Eberhard and Thompson2019). This result could be related to the findings of Moorthy & Sweet (Reference Moorthy and Sweet1936b). According to their experimental studies, Guinea worm larvae ingested by fish appeared still alive in faeces. This fact suggests the larvae could act as parasites in transit through the gastrointestinal tract of the fish, explaining the low likelihood of finding larvae in the intestines of fish in nature, which, in turn, would make it difficult for dogs to become infected by this route.
To date, there has been no research conducted on the presence of D. medinensis larvae in the intestines of fish in Chad. Cleveland et al. (Reference Cleveland, Eberhard and Thompson2019) recently necropsied 234 fish from Chad and removed their gastrointestinal tracts. Unfortunately, the authors did not facilitate any information on the possible presence of D. medinensis larvae in those intestines. However, eight D. medinensis larvae were found in the tissues of five naturally infected frogs captured in Chad (Eberhard et al., Reference Eberhard, Cleveland, Zirimwabagabo, Yabsley, Ouakou and Ruiz-Tiben2016; Cleveland et al., Reference Cleveland, Eberhard and Thompson2019). Yet, it is hard to believe that the level of infection detected in dogs in Chad is the result of the repeated captures and ingestion of infected frogs by dogs in nature.
Dogs and rewards
The finding of a large number of infected dogs in Chad was considered a ‘mysterious epidemic’ (sic) (Callaway, Reference Callaway2016). Presumably, the total absence of control measures aimed at dogs (or at other potential reservoirs) in the past, together with the stimulating reward in cash given to those who find parasitized dogs, have given rise to the current dracunculiasis scenario in Chad (Galán-Puchades, Reference Galán-Puchades2016). In addition to Chad, Ethiopia and Mali have also reported infected dogs after the implementation of the reward for finding parasitized dogs, although to a lesser extent, probably due to the fact that the population in those countries is not yet fully aware of the offered reward (Hopkins et al., Reference Hopkins, Ruiz-Tiben, Weiss, Roy, Zingeser and Guagliardo2018b). The reward for the detection of dog and human cases has also been implemented in South Sudan recently, but not yet in Angola, although in this latter country, in which there were no previous cases, rewards have already been proposed due to the recent appearance of two human cases (CDC, 2019a, b). Rumours of 34 additional human as well as three animal cases have been revealed in that country (CDC, 2019b).
Therefore, when the same efforts that have been made in Chad for the detection of infected dogs are implemented in other countries (if possible) in order to reach a similar awareness of the dog (reservoir) problem, it will be possible to ascertain the degree of peculiarity of dracunculiasis epidemiology in that country.
Concluding remarks
The successful campaign for the eradication of Guinea worm disease, exclusively focused on the water-borne route, means that the food-borne transmission route would not be of epidemiological importance. Dracunculus medinensis larvae are not encapsulated, not hardy and are highly susceptible to desiccation; in addition, larvae exhibit a short life span in the fish intestinal tract (Cleveland et al., Reference Cleveland, Eberhard, Thompson, Smith, Zirimwabagabo, Bringolf and Yabsley2017). Consequently, the probability that fish-eating mammals, humans or reservoirs, become infected by means of this route is, apparently, rather low and may hardly serve as an explanation for such a high number of infected dogs.
If the water-borne route of infection continues to be considered the main route of transmission, dracunculiasis control should be aimed mainly at copepod control, in addition to humans continuing to drink filtered water. Thus, potential paratenic/transport hosts are much less likely to harbour Guinea worm larvae since this hypothetical second food-borne route actually depends on the existence of infected copepods. In addition, this vector control, in turn, prevents reservoirs, which do not drink filtered water, from becoming infected by the copepods.
The eradication of any human disease becomes a vastly more difficult enterprise if there are animal reservoirs of infection. Guinea worm eradication is facing its major challenge at the end of the campaign – that is, the control of the disease in reservoirs, at exactly the point at which social and political interests tend to wane (Whitty, Reference Whitty2015).
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflicts of interest
None.
