Abstract
This review describes the role that dogs play in Latin American countries where Chagas disease is endemic. Multiple studies determined the high frequency with which canine populations are infected with Trypanosoma cruzi. The infection prevalence of dogs is greater than that of humans and the presence of infected dogs in households is associated with a higher risk of human infection. Dog infectiousness to triatomine vectors is several-fold higher than that of humans, thereby underscoring their major role in the domestic transmission of T. cruzi. Insecticide spraying of houses is in most cases efficacious but the lack of sustainability hinders this vector-focused strategy. Multi-pronged approaches have been adopted to improve control measures but dog intervention was never included. Experimental evaluation of systemic insecticides or deltamethrin-impregnated collars suggested that dog intervention leading to triatomine killing could curb domestic transmission of T. cruzi. Larger field studies are required to determine its applicability and efficacy. However, the implementation of dog intervention could complement other control measures currently in place, mostly in periods when vector spraying has been interrupted.
Introduction
C
Dog Participation in the Natural Cycle of T. cruzi
Dogs are susceptible hosts and reservoirs of Tc in many countries of Latin America (Fujita et al. 1994, Crisante et al. 2006, Jimenez-Coello et al. 2008, Pineda et al. 2011, Lima et al. 2012, Enriquez et al. 2013, Bustamante et al. 2014, Alroy et al. 2015). Remarkably, studies in the Southwestern United States also demonstrated that Tc infection in dogs is highly prevalent in regions where human infection is virtually nonexistent (Kjos et al. 2008, Curtis-Robles et al. 2017), signifying the complex epidemiological variables associated with transmission. The efficiency with which dogs amplify the Tc cycle can be deducted from observations made in kennels from Texas infested by triatomine bugs. Dog infection prevalence by molecular methods (polymerase chain reaction [PCR]) was 17.4% while vector infection (Triatoma spp.) reached 80.6% prevalence. The blood meal analysis confirmed that vectors had fed only on dogs (Curtis-Robles et al. 2017).
Like humans, dogs must become infected through contact with vector feces harboring trypomastigotes or eventually by congenital infection. Also, ingestion of infected triatomines (which was shown experimentally in raccoons) cannot be ruled out despite the absence of published information (Roellig et al. 2009).
The socioeconomic and housing conditions in regions where triatomines are endemic determine the public health importance of the Tc natural cycle. In resource-limited areas there are no clear barriers impairing Tc circulation between the extra- peri-and intradomestic environments. Although some wild animals, such as opossums and various rodent species, constitute the link between extra- and peridomestic environments, dogs are the principal source of Tc for domiciliated triatomines (Roque et al. 2013, Gurtler and Cardinal 2015). Triatomine vectors are opportunistic feeders and analyses of host-feeding patterns of >150 studies identified a wide diversity of blood sources (Gurtler and Cardinal 2015).
The association of blood meal source and bug infection ( = infective blood meal index) showed that dogs, among other animals and humans, played a significant role as parasite source (Gurtler and Cardinal 2015). The evaluation of >1000 domestic Triatoma infestans showed that dogs had the highest infective blood meal index (49%) compared with cats (39%), humans (38%), and chicken (29%) (Gurtler and Cardinal 2015). Modeling of transmission dynamics suggested that the presence of animals in the domestic and peri-domestic milieus could have a zooprophylactic effect by diverting triatomines away from humans but could increase the triatomine population. Nevertheless, the authors stated that the potential decrease in human infection due to the dilution effect observed in the model requires further evaluation (Peterson et al. 2015).
Another modeling study explored the oral and congenital routes of infection and arrived at the conclusion that dogs are “the driving force of the infection cycle” (Coffield et al. 2013). In fact, several studies indicated that the presence of domestic animals, principally T. cruzi-infected dogs, increase the risk of Tc infection in humans (Gurtler et al. 1996, Crisante et al. 2006, Jimenez-Coello et al. 2010, Pineda et al. 2011, Enriquez et al. 2014).
The infection prevalence of stray and household dogs in Latin American countries underscores the importance of canine populations for the maintenance of active transmission cycles [reviewed in Gürtler and Cardinal (2015)]. A few examples of field studies performed throughout the Americas illustrate the epidemiological situation. A survey in the city of Mérida, Mexico showed that 17.3% (n = 243) of dogs were infected with Tc as determined by PCR (Jimenez-Coello et al. 2008). Subsequent serological surveys demonstrated that between 10.7% and 21.3% of the canine populations of the Yucatan peninsula were exposed to Tc (Lopez-Cespedes et al. 2013).
A field study in neighboring Guatemala established that 37% of dogs had antibodies against Tc. Furthermore, the analysis of triatomine blood meals determined that 17% had fed on this domestic animal (Bustamante et al. 2014). In Panama, 11.1% of dogs living in rural areas had been in contact with Tc, while 5.1% were infected with the nonpathogenic Trypanosoma rangeli (Pineda et al. 2011).
In South America, the high prevalence of Tc infection in canine populations was associated with both established or reemergent transmission foci. In Western Venezuela, the presence of infected dogs is vastly distributed with an overall seropositivity of 67.6%. As expected, the higher rates of dog seropositivity (84%) were related to households inhabited by seropositive persons (20%) and frequent indoor triatomine infestation (70%) (Crisante et al. 2006).
Serological surveys in the Caribbean region of Colombia established that 71.6% of dogs had anti-Tc antibodies, while children were infrequently infected (0.25%) (Cantillo-Barraza et al. 2015). In this country, dog participation in the Tc cycle varied between 0% and 41% according to the ecological regions where the surveys were carried out (e.g., inter-Andean valleys (Boyacá) or Oriental plains) (Turriago Gómez et al. 2008, Ramirez et al. 2013, Jaimes-Duenez et al. 2017). Several states in Northeastern Brazil have active sylvan and urban transmission cycles in which dog infection rates ranged between 21.9% and 29.4% (Lima et al. 2012, Rocha et al. 2013).
While no publications focusing on Tc infection of dogs in Ecuador are available, studies in Peru showed epidemiological trends similar to those of other Latin American countries, reporting 12.3% canine seroprevalence in peri-rural areas of Arequipa (Castillo-Neyra et al. 2015). The Gran Chaco region is the hot, semi-arid plain shared among Paraguay, Bolivia, and Argentina where Chagas disease is endemic (Gurtler et al. 2007). Field studies in Paraguay and Bolivia (Cochabamba city) reported that 36% and 23.5% of surveyed dogs had anti-Tc antibodies, respectively (Fujita et al. 1994, Medrano-Mercado et al. 2008). Even in countries where Tc transmission was interrupted, like Chile, dogs still are blood sources for triatomines (Mepraia spinolai) known to have high infection prevalence (Canals et al. 2001, Botto-Mahan et al. 2005).
In Argentina, multiple studies established the relevant participation of dogs in Tc endemic areas of that country (Gurtler et al. 1996, Lauricella et al. 1998, Gurtler et al. 2007, Gurtler and Cardinal 2015). In the Northwestern region, dogs had significantly higher infection rates (65.1%) than humans (34.2%). More importantly, dogs were 18-fold more infectious to T. infestans than humans (Gurtler et al. 1996). Dog malnutrition is common in endemic areas and this condition was associated with a 6.3-fold increased infectivity to vectors compared with infected well-nourished dogs from the same endemic area as determined by xenodiagnosis (Petersen et al. 2001). In that work, there was a trend of decreased cell-mediated immunity in malnourished dogs suggesting that a deficient innate immune response favored higher parasitemias and infectiousness.
The identification of truly infectious dogs is a potential problem that quantitative PCR could solve in absence of xenodiagnosis capabilities (Enriquez et al. 2014). On the other hand, serology (indirect immunofluorescence) cannot be used to identify infectious dogs since the antibody titers do not correlate with the infectivity to triatomines used for xenodiagnosis (Turriago Gómez et al. 2008).
Plausible Contribution of Dog Intervention to Curb Tc Transmission
A current map from Pan American Health Organization (PAHO) details the countries where putative interruption of vector transmission was achieved (Belize, Guatemala, Honduras, El Salvador, Nicaragua, Chile) and countries where transmission is still taking place either in one or more of their endemic areas. Remarkably, the latter is occurring in countries where Chagas disease control programs were implemented, underscoring the difficulty in achieving sustainability (WHO 2015).
Initial control efforts targeted vector elimination from human premises as the sole strategy to curb Tc transmission. This led to varied results ranging from sustained decrease of Tc incidence to modest, temporary success (Gurtler and Yadon 2015). Ultimately, the epidemiological complexity of Chagas disease forced governments to adopt a multi-pronged approach in which social participation, education, and housing improvement were part of the strategy added to the traditional insecticide spraying (Gurtler and Yadon 2015). However, dog intervention was never implemented as part of the multifaceted control strategies. It is clear that new alternatives, and consideration of their inherent benefits and drawbacks, need to be assessed (Table 1).
Alternatives for Dog Intervention in Endemic Areas of Chagas Disease
Several experimental vaccines based on live parasites (T. rangeli), recombinant proteins, or DNA are in the pipeline and have the potential to partially protect dogs and reduce Tc transmission (Quijano-Hernandez et al. 2013, Beaumier et al. 2016). However, there are no reports of large-scale field trials in dogs aimed at establishing its efficacy. Experience from canine visceral leishmaniasis (another vector-borne disease) suggests that dog culling to remove the domestic reservoir host is not a sensible approach. This method faces social unacceptability and epidemiological analyses yielded controversial results, many of which indicate no impact on disease transmission (Otranto and Dantas-Torres 2013).
For Chagas disease, less controversial approaches involving dog interventions to decrease the risk of Tc transmission were recently evaluated. These interventions fall into the “altruistic” category since animals are not protected against the vectors but decrease the risk of transmission to other hosts. Deltamethrin impregnated collars (DMIC) originally developed to protect dogs from sand fly bites in visceral leishmaniasis endemic areas were evaluated against Chagas disease vectors (Killick-Kendrick et al. 1997, Reithinger et al. 2005, Reithinger et al. 2006).
A field trial using experimental mud-thatched huts in an endemic area of Argentina evaluated the impact on triatomine bugs of DMIC in a small number of dogs (n = 7) compared to uncollared controls (n = 3). Over a period of ≈6 months, colony-reared uninfected bugs showed reduced feeding success (odds ratio [OR] = 0.40, p < 0.001), lower survival (OR = 0.15, p < 0.001), and lower fecundity (OR = 0.64, p < 0.001) (Reithinger et al. 2006). There was no information on the speed with which T. infestans died after feeding but results suggest that it could have been a slow, progressive process. Furthermore, it needs to be determined whether DMIC has some degree of repellent effect on bugs that could divert feeding toward humans, thereby increasing the risk of Tc infection.
The potential increase in pyrethroid resistance of triatomine bugs, the cost of the commercial product, and the predicted high frequency of DMIC losses requiring replacement could hamper the efficacy of this approach in control programs (Reithinger et al. 2004, Gurevitz et al. 2012, Echeverria et al. 2018). Fipronil, a spot-on insecticide commonly used to combat fleas, ticks, and lice infestations in dogs was evaluated against T. infestans with disappointing results (Gurtler et al. 2009, Amelotti et al. 2012). In addition to its negligible activity against T. infestans, the need for monthly applications makes this approach impractical and expensive.
Encouraging results were obtained with the experimental utilization of fluralaner (isoxazoline family), a drug marketed to combat fleas and ticks in pets. Fluralaner is a potent antagonist of γ-aminobutyric acid and
However, the sustained lethal activity of fluralaner on the bugs could significantly impair Tc transmission in households. The fact that T. infestans showed good feeding rates on treated dogs suggested that the risk of bug diversion to humans is low. If dogs tend to eat triatomine bugs, the use of insecticide treatment would be inefficient to halt transmission to other dogs within the same household. Nevertheless, transmission to humans would still decrease because any bug fed on insecticide-treated dogs would be rapidly killed or unable to feed a second time before dying.
Sustainability is a major hurdle for vector control programs mostly due to limited economic resources or political will (Gurtler 2009, Gurtler and Yadon 2015). In addition, triatomine bionomics is complex and encompasses sylvatic niches unreachable by vector control programs. T. infestans, initially described as a “domestic” species exclusively inhabiting households, was found in extradomestic environments with increasing frequency (Breniere et al. 2013). On the contrary, certain “extradomestic” Rhodnius species have shown domiciliation tendencies (Fitzpatrick et al. 2008, Brito et al. 2017). Nevertheless, the intradomicile feeding on treated dogs should lead to triatomine mortality, thereby impairing its participation in both the domestic and sylvatic Tc cycles.
Inside houses, triatomine bugs tend to feed more frequently on dogs than humans, possibly because of their greater accessibility during host seeking activity (Gurtler et al. 1996, Gurtler and Cardinal 2015). We believe that dog treatment using long-lasting systemic insecticides will transform the principal domestic reservoir into a biological control weapon. Using this strategy, the risk of Tc transmission could significantly diminish. In addition, it would prevent triatomine recolonization during periods lacking vector control efforts.
Published information on dog intervention underscores the urgent need of further evaluating this alternative in single and multi-pronged studies. Despite the sensible reduction of Chagas disease in Latin America, regions of active transmission still exist. Therefore, it is time to redesign strategies to complement vector control, making them more efficacious, enduring, and attainable.
Footnotes
Author Disclosure Statement
No conflicting financial interests exist.
