43 American trypanosomosis (Chagas disease)

American trypanosomosis (more usually known as American trypanosomiasis) is due to infection with Trypanosoma cruzi (Protozoa, Kinetoplastidae). This is a widespread parasite of small mammals and marsupials throughout most of the Americas, roughly from the Great Lakes of North America (approx. 42°N) to southern Argentina (approx. 46°S). It is mainly transmitted by blood-sucking bugs of the subfamily Triatominae (Hemiptera, Reduviidae) which are widespread in the Americas, but rare in the Old World. Except in some research laboratories, and infected immigrants from Latin America, T.cruzi has not been reported from the Old World, although closely-related trypanosome species are commonly found in Old and New World bats.

Human infection with T.cruzi is generally known as Chagas disease, taking the name of Brasilian clinician Carlos Justiniano das Chagas who first described it from patients in central Brasil (Chagas 1909). Chagas isolated and described the parasite, correctly deduced most of its life-cycle and clinical symptoms associated with the infection, identified the insect vectors and some of the reservoir hosts, and also trialed initial attempts to control it. He was nominated at least twice for the Nobel prize in medicine (Coutinho and Dias 2000; Lewinsohn 2003).

Although difficult to treat, Chagas disease can be controlled by measures to halt transmission, primarily by eliminating domestic populations of the insect vectors, together with serological screening to avoid transmission by blood donation from infected donors. Since 1991, a series of multinational initiatives have used this approach to halt transmission over vast regions of the areas previously endemic for the human infection. Estimated prevalence of the human infection has declined from the 1990 estimate of 16–18 million people infected, to the current estimate of just over 7 million infected (OPS 2006; Schofield and Kabayo 2008). Prevalence is expected to decline further, and control strategies are now being adjusted to develop a sustainable system of disease surveillance, focal vector control, and specific treatment for any new cases (Schofield et al. 2006; WHO 2007). Guidance for diagnosis and treatment is also required for non-endemic countries, where recent years have seen increasing migration from Latin America such that cases of chronic Chagas disease have now been reported from amongst Latin American migrants in Europe, USA and Canada, and Japan, together with some congenital cases and transmission from infected blood donors and by organ transplant.

Chronic Chagas disease is the principal cause of cardiomyopathy in Latin America, causing severe debility in up to 30% of those infected (Table 43.1). In some areas—particularly south of the Amazon region—the chronic disease is also associated with severe intestinal lesions leading to disfunctional dilations of the intestinal tract known as megasyndromes (eg. megaoesophagus, megacolon, and less frequently, megastomach) and sometimes also with neurological disorders although these are now less frequently seen. At its peak in the mid-1980s, the prevalence of human infection was estimated at around 24 million (Walsh 1984), suggesting an incidence of over 800,000 new cases each year in the absence of control interventions (Hayes and Schofield 1990) with an aggregate mortality rate of over 55, 000 deaths per year. The early acute phase of infection can be fatal in up to 10% of cases, but more usually progresses to a chronic phase, initially asymptomatic, that leads to progressive tissue destruction, particularly of heart muscle, but also nervous tissue, mainly of the autonomous parasympathic innervation, that controls the dynamics of the hollow viscera. In 1993, the World Bank ranked Chagas disease as the most serious parasitic disease of the Americas in terms of its public health impact, with a disease burden (measured as DALYs—disability-adjusted life years) far exceeding even the combined burden of other parasitic diseases such as malaria, schistosomosis, and leishmaniosis (World Bank 1993). Since then, a series of multinational control initiatives has effectively halted transmission over vast areas of Latin America, such that current prevalence estimates have been reduced to around 7 million people infected, with a current incidence of little more than 50, 000 new cases each year (OPS 2006; Dias et al. 2002; Schofield et al. 2006). Prevalence and incidence can be expected to decline further, although interruption of the current control and surveillance programmes would probably lead to a resurgence of transmission.

The causative parasite of Chagas disease, Trypanosoma cruzi, is mainly transmitted to humans in the faeces of blood-sucking bugs of the subfamily Triatominae (Hemiptera, Reduviidae). As the insects feed, they may defaecate the remains of their previous blood meal onto the host skin; parasites in these faeces can survive as long as they are moist, and will readily transit the mucosae of mouth, eye or nose, and in some cases may also transit abraded skin. Direct oral-route transmission of T.cruzi can occur by ingestion of food or drink contaminated with infected bug faeces, or triturated infected bugs, or by consuming undercooked meat or blood products of infected mammals. T.cruzi can also be transmitted by blood donation or organ transplant from infected donors, and occasionally by transplacental passage from infected mothers to the foetus (1–10% of infected mothers). Control in Latin America is therefore based primarily on eliminating domestic populations of the insect vectors, together with serological screening of blood donors to avoid the risk of transfusional transmission, and improved maternal health-care with specific treatment of infected new-borns. Vector-borne transmission does not yet occur outside the Americas (Schofield et al. 2009) but with migration of people from Latin America to other countries, there have been increasing reports of transmission by blood transfusion, organ transplant, and congenital transmission, in regions previously considered non-endemic for Chagas disease—particularly in North America and parts of Europe.

Trypanosoma cruzi is a widespread parasite of small mammals and marsupials throughout much of the New World, roughly from the Great Lakes of North America (approx. 42°N) to southern Argentina (approx. 46°S). It is a kinetoplastid flagellate whose bloodstream forms are typified by a large kinetoplast at the posterior end. This readily differentiates it from other trypanosomes, especially T.rangeli which is also a common parasite of small mammals in Latin America and can transiently infect humans (but without significant pathology). As far as is known, all mammals and marsupials are susceptible to infection with T.cruzi, although birds, amphibians and reptiles are refractory. In marsupials and autochthonous mammals, T. cruzi appears to cause little if any pathology, and the infection in such animals tends to retain a relatively high parasitaemia. In humans however, and in mammals (such as dogs, cats, murine rodents) probably brought in recent times to the Americas, parasitaemia is reduced to subpatent levels after the initial acute phase, and the parasites mainly develop intracellularly causing pathology primarily through inflammation and progressive tissue disruption.

Since the pioneering studies of Miles et al. (1977), the phylogenetics of T.cruzi have been extensively studied, showing that different isolates of the parasite could be consistently grouped according to their isoenzymatic profiles, with zymodeme 1 (Z1) being the most widespread, zymodeme 2 (Z2) found mainly from human infections and domestic animals in the Southern Cone region (south of the Amazon basin), and zymodeme 3 (Z3) from Amazonian isolates (see: Miles and Cibulskis 1986). Subsequent studies led to a consensus of two main lineages denoted T.cruzi I and T.cruzi II (Anon,1999). Of these, cruzi I (Z1) appears to be the more widespread both in North and South America, and in silvatic habitats seems particularly associated with didelphid opossums. T.cruzi II (including Z2 and Z3) is genetically more variable (now subdivided into cruzi II a-e, and also into a series of numbered clones, according to genetic differences) and is mainly found in South America (south of Panama) (Table 43.2). A reclassification of T.cruzi into 6 discrete typing units (DTUs) has recently been proposed (Zingales et al. 2009), elevating cruzi II a-e to cruzi II-VI. In human infections, both cruzi I and cruzi II can be associated with chronic cardiomyopathy, but in general it is only forms of cruzi II that are also associated with the gastrointestinal lesions known as megasyndromes.

There is genetic and biogeographical evidence to suggest that the ancestral forms of T.cruzi developed originally as parasites of marsupials within the southern supercontinent, and in genetic terms, some of the closest extant relatives of T.cruzi are trypanosomes of kangaroos and wombats in Australia (Stevens et al. 2001). From these ancestral forms, the original form of T.cruzi may have developed as a parasite of didelphid opossums as the supercontinent divided to give rise to South America some 70mya, accompanying the didelphids along their subsequent spread northwards. This original form would probably have been directly transmitted between opossums via their anal gland secretions and/or urine (Jansen and Deane 1985; Olsen et al. 1964), and would be represented today by the form now referred to as T.cruzi I. The genetic divergence from cruzi I to cruzi II, now dated as 4-10mya (Brisse et al. 2003) may have resulted from the advent of triatomine insect vectors, providing the physical means to transmit the parasite to different mammalian hosts, such that the variability now seen in cruzi II probably represents a series of adaptations to these different host species (Schofield 2000) with recombination and genetic exchange also contributing to differentiation within this lineage (e.g. Yeo et al. 2005; Freitas et al. 2006).

In general terms, T.cruzi seems a rather non-specific parasite, in the sense that all mammals and marsupials seem to be susceptible to infection, and, once infected, will usually retain the infection for life. Birds, reptiles and amphibians however, are refractory to the infection as the parasite is killed by complement-mediated lysis in avian or reptilian blood. Amongst mammals, the infection tends to be most prevalent in those that construct nests or lodges which may be infested with vector species of Triatominae, so that larger mammals such as cattle and horses are less likely to be infected.

T.cruzi can also survive in a wide variety of invertebrate hosts, even artificially infected leeches and wax-moth larvae (see Schofield et al. 2009). It seems to survive, at least transiently, in most blood-sucking insects especially ticks and cimicid bedbugs, but tends not to develop further except in Triatominae, and no invertebrate except for triatomine bugs has ever been shown to have epidemiological significance as a vector of T.cruzi. In triatomine bugs, ingested bloodstream forms of the parasite (bloodstream trypomastigotes) multiply as epimastigotes in the midgut, transforming to infective (metacyclic) trypomastigotes in the hindgut (rectum) from where they can be expelled as the bug defaecates. T.cruzi does not survive in the haemocoele of the bugs, and so cannot penetrate the salivary glands and cannot be transmitted by the bug’s bite.

The course of T.cruzi infection in mammals is extremely variable throughout the infection, depending on type of host and parasite strain. In general, the course of human infection is marked by an initial acute phase, defined by patent parasitaemia, followed by a life-long chronic phase during which the parasites are mainly present inside the cells of various organs rather than in the bloodstream.

After infection, the parasites multiply locally at the portal of entry, sometimes causing a localized palpable lesion known as a ‘chagoma’. Where the portal of entry was the eye, this may result in palpable unilateral ocular oedema known as ‘Romañas sign’ following the work of Argentine clinician Dr Cecilio Romaña who showed its clear association with acute Chagas disease in humans. In such cases, trypanosomes may be present in the chagoma, and, in the case of Romañas sign, in the orbital tissues, especially the ocular muscles and lachrymal glands (so that some trypanosomes may also be shed in the tears). But Romañas sign can be readily distinguished from other unilateral ocular oedemas since it is persistent (a month or more), painless, and without indications of inflammation or haematoma (thus distinguishing it from conjunctivitis which is generally inflamed and not so persistent, a blow to the eye which is usually painful and bruised, or a wasp/bee sting which would be inflamed, painful, and of relatively short duration). In some regions, over 50% of acute T.cruzi infections show Romañas sign or detectable chagoma, although in other regions such signs are now rarely seen. It is possible that a high frequency of Romañas sign is indicative of an unusual parasite strain becoming locally dominant, such that frequency of the sign may decline with time as that strain becomes more adapted to humans.

The acute phase of infection generally lasts 6–8 weeks, and is defined by the presence of a patent parasitaemia (ie. parasites visible by microscopy of a fresh blood film). This phase may show few if any symptoms, although it is often accompanied by fever, usually low-grade and non-cyclic, together with malaise, headache, lymphadenopathy, and diffuse heart pains. In some cases there may be tachycardia, hepatosplenomegaly, subcutaneous nodules, or a diffuse rash resembling measles. The acute-phase can be life-threatening, with mortality in 5–10% of acute-phase patients—usually due to overwhelming myocarditis or meningoencephalitis. For those who survive, parasites are progressively cleared from the bloodstream by a strong immune response, such that cruzi-specific serology then tends to remain positive for the remainder of the infection. In immunocompromised patients, the acute phase can be prolonged and life-threatening. Similarly, immunosuppression of patients in the chronic phase, e.g. by HIV infection or immunosuppression during transplant, can lead to reactivation of the acute phase. During the chronic phase of infection, the parasites mainly survive as intracellular amastigote forms that multiply and transform within the host cells until the infected cell bursts, liberating trypomastigotes to infect neighbouring cells where the process is repeated. A few of these trypomastigotes will also escape to the peripheral bloodstream—insufficient to cause patent parasitaemia, but generally sufficient to infect a triatomine bug feeding from this infected host.

For most patients, the chronic phase of T.cruzi infection is asymptomatic, and the infection is often described as ‘indeterminate’. For many, this state will persist for life, without any need for supportive treatment or modifications of their general life-style. Nevertheless, a positive diagnosis, e.g. following screening of blood donors, can be extremely stressful because, at present, no specific treatment is offered to chronic adult patients, and there is no reliable way to predict whether or not the infection will remain asymptomatic in that patient. It is therefore important that all chronic cases, even if entirely asymptomatic, receive adequate counselling to explain possible progression of the infection, together with regular clinical check-up to monitor progress. For some, estimated at 20–40% in different studies, the chronic infection will lead to moderate or severe lesions, especially of the heart, often decades after the initial infection.

The early signs of cardiomyopathy are revealed by electrocardiography, usually mild arhythmias such as extra-systoles or first degree atrioventricular block. These can often be controlled using antiaryhthmic drugs, of which amiodarone is generally recommended and seems also to have some specific anti-trypanosome effects (Benaim et al. 2006). In some cases the aryhthmias will progress to complete right bundle branch block (often considered pathognomic for chronic symptomatic Chagas disease) with left anterior hemiblock and eventually complete atrioventricular block and consequent Stokes-Adams attacks. Repeated heart attacks, severe aryhthmias and emboli are signs of a poor prognosis, and sudden death due to cardiac arrest or ventricular fibrillation is well documented. More common however, is simple pump failure due to extensive pancarditis with muscle fibre destruction—the patient is lethargic, with cold extremities, and quite unable to work (and it is this debility that, in epidemiological terms, is mainly responsible for the high socio-economic cost of the disease). In extreme cases, destruction of the heart muscle is severe enough to lead to cardiac aneurism, especially at the apex, with the heart bursting on exercise.

Cardiomyopathy is the most common symptom of chronic symptomatic Chagas disease, seen throughout the endemic areas of Latin America. Informal polls of cardiologists in Mexico and Central America indicate that 20–25% of all cardiac pacemakers implanted are due to chronic Chagas disease, and this has been a major factor prompting large-scale initiatives designed to halt transmission of T.cruzi. In the southern part of the continent however, where strains of cruzi II predominate in human cases, chronic Chagas disease is also commonly associated with lesions of the intestinal tract collectively known as megasydromes. Most common is megaoesophagous—the oesophagous becomes dilated with solid residues due to aperistalsis and spasm of the cardiac sphincter. Patients have difficulties in swallowing, and require copious liquid to assist the passage of food. Also common in some areas, especially central Brasil and Bolivia, is megacolon, associated with persistent constipation requiring manual evacuation. More rarely, other parts of the digestive tract may be involved, including megaduodenum, megabladder, mega-gallbladder, and megastomach. In early cases, treatment by sphincter dilation has been widely used, while for later cases surgical resectioning can be successful (Dias and Coura 1997).

One of the least understood aspects of Chagas disease has been the mechanism of pathogenesis, and this has affected policies of transmission control and policies regarding specific treatment.

The strong immune response generated by immunocompetent people during the acute phase of infection, which then persists throughout the infection, has been recognized since the earliest studies of the disease, with the first immunological assays of infection being developed in 1914. However, this immunological response also led to a generalized assumption that reinfection would be both unlikely and of little relevance, because any newly-introduced parasites would meet an immune defence already maintained by the initial infection. The implication for policies of transmission control is, therefore, that even halting new transmission would do nothing for those already infected. In recent years however, this idea has become strongly challenged, first by clinical anecdote and epidemiological indications that the prognosis for chronic cases seemed to improve in areas where vector-borne transmission had been interrupted (Dias et al. 2002) and then by direct experimentation in murine models showing that the likelihood of chronic cardiopathy was significantly increased in mice that were repeatedly challenged, compared to mice that were infected only once (Bustamente et al. 2002; 2007). There is also evidence that the likelihood of transplacental (congenital) transmission by infected mothers to the foetus tends to decline when the mothers are no longer subject to repeated reinfections, e.g. following successful vector control interventions. It appears that both pathology, and likelihood of transmission, have a positive relationship with what might be termed ‘parasite load’.

There were also many theories offering different explanations of the main pathogenic mechanisms, most of which tended to assume that pathology was triggered by the parasites but not otherwise directly dependent on them. The painstaking studies of Köberle for example, who carefully studied patterns of denervation in chronic infections, were generally interpreted to suggest that during the chronic phase the incipient organic damage had been done, so that specific chemotherapy against the parasite would be of little relevance. Similarly, the idea that Chagas disease was primarily an autoimmune phenomenon, i.e. a response to self-antigens that have been mimicked by the parasite, also carried the implication that specific treatment during the chronic phase would be of little relevance, because the parasite had already triggered the autoimmune response and so played little role in the development of chronic pathology. Again however, these ideas have been seriously questioned by more recent work that indicates that persistence of the parasite, together with some ‘imbalance’ of the immune system which may include autoimmune components, is the primary requisite for triggering and maintaining the inflammatory processes responsible for the chronic lesions. And since a key component of this process seems to be related to cellular destruction caused by the parasites themselves, these studies have led to considerable re-interest in the idea of developing new drugs, and also in offering treatment to all those infected, chronic as well as acute, as a means to inhibit development of the chronic disease.

Only two drugs are currently available for treatment of Chagas disease, both introduced some 30 years ago. Nifurtimox, a nitrofuran developed by Bayer in 1967 and marketed as Lampit^®, acts by reduction of the nitro group to give nitro-anions that then react with molecular oxygen to produce toxic superoxide and peroxide radicals. Benznidazole, a nitroimidazol developed by Roche in 1972 and marketed as Rochagan^® or Radanil^®, appears to act differently, producing metabolites that react with macromolecules such as DNA, RNA, proteins and possibly some lipids. In both cases however, the antiparasitic activity is intimately linked with their inherent toxicity, such that adverse side effects during treatment are quite common, especially amongst older patients. Both drugs can induce malaise, headaches, and loss of concentration; nifurtimox frequently causes loss of appetite, weight-loss and in some cases anorexia, while benznidazole is more frequently associated with allergic dermatitis and in some cases peripheral neuritis. In a few cases, the reactions can be severe enough to require suspension of treatment, although intolerance to one drug is rarely associated with intolerance to the other.

Both nifurtimox and benznidazole were quickly shown to be effective in treating acute infections, with cure rates typically well above 80% (Table 43.3). However, because of side-effects and doubts about the likelihood of cure for chronic patients (see above) their potential for treating chronic infections was largely ignored. Only since the work of Viotti et al. (1994) has the benefit of specific treatment of chronic infections begun to be more widely accepted, and since the Pan American Health Organization (PAHO) expert committee meeting in 1998 (PAHO document OPS/HCP/HCT140/99) there is growing consensus that treatment should be given to all acute cases and asymptomatic chronic cases, with the aim of inhibiting the development of disease even if radical parasitological cure is not obtained. At present, most endemic countries accept the recommendation to treat all chronic infections in children below the age of 15–16 years, although current studies are suggesting that specific treatment should also be offered, at least to asymptomatic cases, up to 50 years of age (Viotti et al. 2006, 2009).

Due to lack of demand during the 1980s and 90s, the production of nifurtimox was abandoned in 1997, but restarted in 1999 in response to the usefulness of this drug in treating melarsoprol-resistant cases of gambiense sleeping sickness in West Africa. Through agreement with the World Health Organization (WHO), production has now been assured, at least until 2012, and nifurtimox is available free-of-charge through WHO Geneva and WHO-PAHO offices in Latin America. Under a technology-transfer agreement from Roche, benznidazole is now produced by Laboratorios Farmaceuticas do Estado de Pernambuco (LAFEPE) in Brasil, but can be made available at cost through WHO Geneva. Several newer compounds are currently in development for the treatment of T.cruzi infections, especially sterol inhibitors such as pozaconazol and ravuconazol originally developed as fungicides, as well as cistein-protease inhibitors (cruzipain), trypanothione inhibitors, and pyrophosphate inhibitors (Urbina and Docampo 2003) although none is expected to become available for public health use for several years.

The Triatominae are traditionally defined as a subfamily of the Reduviidae, characterized by their blood-sucking habit and associated morphological features (such as a slim three-segmented rostrum with articulated third segment). Adults tend to be large insects, typically around 2.5 cm long, although some silvatic species reach no more than 0.5cm while the largest species, Dipetalogaster maxima from Baja California, has adults up to 4.5 cm long. Triatominae are hemimetabolous (exopterygote) insects, such that their developmental cycle proceeds from eggs through 5 nymphal stages to the adults. All the nymphal stages and both sexes of adult are haematophagous and potentially capable of becoming infected and transmitting T.cruzi. However there is no transovarial transmission of T.cruzi, meaning that newly hatched nymphs will not become infected until their first blood meal from an infected host. As a result, the likelihood of them being infected increases with the number of blood meals taken, so that the infection rate tends to be low in the youngest nymphs, but often increases to above 50% in the adults. Triatominae generally require at least one replete blood meal during each nymphal stage (although a series of smaller meals can be sufficient). Replete feeding typically requires 10–20 minutes, during which time the bugs will usually engorge several times their unfed weight of blood, so that the blood meal of an average-sized adult bug will typically be around 0.4ml—although adult D.maxima have been recorded as engorging up to 4.5ml during a single feed. At temperatures between 20–30°C, most Triatominae will try to feed every 5–15 days, and can complete their egg-to-adult development in 5–6 months with adults then living a further

3–12 months, although some of the larger species will often take 12 months to complete their developmental cycle. However, feeding activity and development is slower at cooler temperatures, and generally halts at temperatures below 15°C.

At the time of writing, 140 species of Triatominae are recognized, customarily grouped into 5 tribes with 17 genera (Schofield and Galvão 2009) (Table 43.4). Over half of these species have been shown to be naturally or experimentally infected with T.cruzi, and it is assumed that all may be capable of transmitting the parasite. In epidemiological terms however, relatively few of these species are of major significance, because most occupy silvatic habitats with little contact with humans. Of greatest epidemiological importance are those species that have adapted to live in close association with humans, especially those that colonize human dwellings. Of these, most important in the Southern Cone countries is Triatoma infestans, together with T.brasiliensis and Panstrongylus megistus in NE and central Brasil. In the Andean Pact countries, the most important is Rhodnius prolixus in Venezuela and parts of Colombia, together with R.ecuadoriensis in southern Ecuador and northern Peru. A derivative form of R. prolixus is also the most important domestic vector in parts of Central America, together with  T.dimidiata whose distribution also extends southwards into Colombia and Ecuador, and northwards into Mexico.

Over their wide distribution in the Americas, silvatic Triatominae can usually be found in almost any habitat offering shelter and ready access to a source of vertebrate blood. Habitats such as mature palm-tree crowns, opossum lodges, rodent burrows, rockpiles, and the more permanent nests of colonial parrots or dendrocolaptid birds (irrespective of the host currently occupying the nest) seem particularly favoured. In general terms, species of Rhodnius tend to be associated with palm-tree crowns, Psammolestes with birdnests, Panstrongylus with burrows (especially of ground dwelling edentates), and Triatoma with rockpiles and rodent burrows. Species of Rhodnius are mainly found in the Amazon region and neighbouring parts of the Venezuelan llanos and Brasilian caatinga and cerrado, broadly bounded by the distribution of the three main groups of Triatoma—the rubrofasciata group mainly of Central and North America, the dispar complex along the Andean mountains, and the infestans group of the caatinga-cerrado-chaco corridor of open vegetation to the east and south of the Amazon basin (Schofield and Galvão 2009).

Most populations of silvatic Triatominae tend to be relatively small, consisting of only a few adults and nymphs, which probably reflects some unreliability in their likelihood of encountering a suitable food source. Similarly, individual bugs from silvatic populations are often physically larger than conspecific individuals from domestic populations, which may be a reflection of their need to be able to store food for longer periods between relatively irregular blood meals. The generally low nutritional status of silvatic Triatominae is also indicated by the efficiency of the ‘Noireau trap’, a sticky trap containing a live mouse or similar small host, which is widely used to collect samples of bugs from silvatic habitats (e.g. Abad-Franch et al. 2000). By contrast, similar animal-baited traps have proved much less effective is sampling domestic triatomine populations, presumably because the ‘bait’ must compete with a much wider range of other blood meal sources in a domestic habitat.

Silvatic triatomine populations are increasingly studied, because of their potential importance in recolonizing rural houses, and also for the capacity of some species to enter houses without necessarily succeeding in forming new colonies there. Bugs that fly into houses, e.g. attracted by house lights, can be a nuisance due to their relatively painful bites, and also represent a risk of transmitting T.cruzi even in areas where domestic bug populations do not occur. In such cases, transmission may be direct, e.g. to dogs and cats that may eat an infected bug, or it may be indirect due to the bugs contaminating food or drink. In and around the Amazon region for example, there are now several records of ‘family microepidemics’ where outbreaks of acute Chagas disease have been attributed to infected adult bugs contaminating fruit juice, soup, or other foodstuffs, that have then been consumed by several family members and their guests (e.g. Coura et al. 2002; Valente and Valente 1999). However, there are several key aspects of the biology of silvatic bugs that remain poorly understood—particularly the factors that may trigger adult bugs to fly into houses, their mechanisms of orientation, and the factors that influence whether or not they will successfully establish a new domestic colony.

Light-trap collections of silvatic bugs, and direct studies of laboratory-reared bugs, indicate that adult flight occurs only when the bugs are relatively starved, that is to say, well-fed adult bugs usually do not fly (Lehane et al. 1992). From this, and from the few epidemiological cases that have been studied in detail, it is suggested that flight of silvatic bugs is mainly triggered by ecological events such as flood, drought or fire, that lead to host death or flight. Bugs then become hungry, and adults may fly—presumably in search of new blood meal sources. However, little is known about their orientation mechanisms. Although it is generally assumed that bugs fly into houses attracted by light at night, such a process cannot explain how bugs move from one silvatic habitat to another. New silvatic habitats do become colonized, but it is not clear how the bugs find the new silvatic habitats. Specific odour attractants have not been clearly demonstrated, and it may be that the silvatic bugs are responding mainly to the warmth (e.g. infrared) emitted by different vertebrates.

In houses, especially rural dwellings with cracked adobe or mud walls, domestic populations of Triatominae encounter a ready supply of blood meal sources in the form of the people and domestic animals occupying the house, together with some protection from climatic extremes and silvatic predators. As a result, they can build up very large populations, often with several thousand nymphs and adults in a single house—the ‘record’ is a rural house in Colombia dismantled to reveal a population of over 11,000 R.prolixus (Sandoval et al. 2000). The final population size is naturally regulated through a density-dependent mechanism mediated primarily by the bugs’ access to blood meals—in a sense, more blood = more bugs (Schofield 1980). Hungry bugs, resting in cracks and crevices of the house walls and roof during the day, emerge at night to feed on the sleeping occupants. As more bugs try to feed from a single host, that host–although still asleep–tends to become more restless, disturbing the bugs from completing their meal (Schofield et al. 1986). Nymphs with a smaller blood meal develop more slowly, and adult females with a smaller blood meal lay fewer eggs, so that the rate of recruitment from one stage to the next tends to decline, so reducing the overall blood demand. Conversely, as bug density declines, then each bug can access more blood, so increasing their rate of development and rate of egg-laying and increasing the rate of recruitment from one stage to the next. In this way, an established domestic bug population tends to be stable, not in numbers, but in total blood demand (in the sense that an older bug has a larger blood demand than a younger bug). Studies of T.infestans in houses in central Brasil, and of R.prolixus in houses in Venezuela, both indicate that the average blood intake of the bug populations is equivalent to about 2.5ml per person per night, typically representing about 25 bites per person per night (Rabinovich et al. 1979; Schofield 1981). For many people, especially those on a poor diet or with other blood-sucking parasites such as intestinal hookworms, this level of chronic blood loss probably contributes significantly to chronic iron-deficiency anaemia.

As Triatominae feed, they may defaecate the remains of their previous blood meal onto the host, and, if infected, this can pose a risk of T.cruzi transmission. This method of transmission is inefficient compared to direct injection of, say, malaria parasites by a feeding mosquito. Studies in Venezuela suggest that on average, each successful transmission event is associated with around 1,000 feeding contacts by infected R. prolixus (Rabinovich et al. 1995) although with 25 bites per night, this is readily achieved in a few months, and helps explain why the average age of new human infections tends to be in the first few years of life. The timing of defaecation depends on bug species, such that some (such as the aptly-named T.protracta) may defaecate up to an hour after having left their host, while others are more likely to defaecate while still in the act of feeding and so still in contact with the host. But for each species, defaecation also seems to be a density-dependent process, in the sense that in high density bug populations where each insect takes only a small blood meal, defaecation tends to be later than in low density populations where each bug is feeding towards complete engorgement (Kirk and Schofield 1987; Trumper and Gorla 1991). Paradoxically therefore, it is in low density bug populations (i.e. those recently established, that are still growing towards their maximum) where the risk of new transmission is highest, since each bug is more likely to take a full meal and so defaecate while feeding on the host. By contrast, in high density bug populations, where each bug takes a smaller meal, defaecation tends to be retarded and generally occurs when the fed bug has returned to its resting site. The walls of heavily infested houses are thus often streaked with the characteristic faecal smears of Triatominae—white streaks representing their uric acid, together with dark brown streaks representing the undigested part of haemoglobin.

Understanding these density-dependent processes in the development and feeding behaviour of domestic Triatominae carries an important moral for the vector control strategies, suggesting that poorly-executed or partial vector control interventions can be more dangerous than no control at all. Partial control, that does not eliminate the bug population, implies that the remaining bugs will now be able to feed to a maximum, defaecating while still in contact with their hosts, and so facilitating further transmission of T.cruzi.

The essence of Chagas disease control in Latin America is the elimination of domestic vector populations. Prior to the current multinational initiatives (see below) domestic vectors were estimated to account for well over 80% of all transmission (e.g. Schofield 1994) and adequate techniques for eliminating these populations had been available since the 1940s. In addition, there are further justifications for eliminating domestic Triatominae, beyond their role as Chagas disease vectors, because of the severe nuisance and chronic blood-loss associated with them.

Early attempts to control domestic Triatominae relied on the idea of improving rural houses, for example by plastering over the cracked walls where bugs were seen to be resting. By itself, house improvement is usually insufficient to eliminate a domestic triatomine population, but can reduce the likelihood that an un-infested house becomes recolonized by the bugs (Guillén et al. 1997). In the 1940s, attempts were also made with cyanide gas fumigation, petrol or diesel applied to house walls, and even flame-throwers, but such approaches proved impractical on an extended scale (Dias and Schofield 2004). Since then, a wide range of other techniques has been tested, including insecticides of various classes, insect growth regulators, chemosterilants, insect pathogens or predators, and various approaches at genetic control, e.g. by trying to select for heritable sub-sterility, and also by genetically-modified symbionts rendering the bugs refractory to infection with T.cruzi. Most have been rejected for large-scale use due to ineffectiveness or impracticability for large-scale interventions, except for indoor application of adequate insecticides, house improvement where economically feasible, and health education to promote community participation in reporting domestic infestations so they can be adequately sprayed.

Although DDT was quickly found to be ineffective against domestic Triatominae, other organochlorine insecticides such as gamma-BHC (also known as HCH or lindane) and dieldrin proved to be effective when applied at high doses (>500mg a.i./m²⁾ and these compounds remained the main products used for domestic triatomine control until being progressively replaced by synthetic pyrethroids during the 1980s. The alpha-cyano pyrethroids proved to have several advantages, being effective at much lower doses and so proving cost-effective in spite of their higher price per kilo (Table 43.5). They were also preferred by spraymen and householders as they were easier to transport and apply, and generally did not mark the sprayed walls.

In general terms, the interventions begin by informing the householders and checking to see if the house has evidence of a current infestation. Several techniques are used for this, involving householder reports of bugs, and visual inspection of the interior house walls and furnishings (using torch and long blunt forceps) for evidence such as live bugs, bug eggs, eggshells, exuviae, and faecal streaks. In some cases, the search for live bugs is aided by spraying an aqueous dislodgant agent into cracks, usually a non-residual pyrethroid such as tetramethrin, which irritates the bugs making them more easy to see. With the aid of the householders, the house is then prepared for residual spraying, removing to the outside all children, animals, foodstuffs, kitchen utensils, bedding and clothes, and the furniture is moved to give access the walls behind. All internal surfaces are then sprayed with the recommended dose of pyrethroid (Table 43.5) including the furniture and inner roof space. Spraying is best done by well-trained operators, thus avoiding the possibility of missing areas that can serve as untreated refuges for the bugs. In most cases, peridomestic structures are also sprayed, including latrines, chicken coops and animal enclosures, since these can also harbour bugs.

A thorough spray with an adequate pyrethroid at the correct dose is usually sufficient to eliminate the domestic Triatominae, and can also give transient control of other domestic pests such as fleas, cockroaches, and scorpions. However, the sprayed house remains vulnerable to reinfestation by bugs coming from elsewhere. The technical response to this problem has been to try to increase the residual effect of the insecticides, for example by using longer-lasting formulations such as wettable powders (WP), suspension concentrates (SC), microencapsulated formulations (CS), or even polymerized formulations (often known as ‘insecticidal paints’). But the strategic response is to eliminate the sources of ‘reinvading’ bugs, i.e. to extend the geographical scale of the control programme to include all the infested houses that could provide source populations for bugs to re-infest those houses already treated. This implies very large-scale coverage, often reaching national or multinational scales of intervention in the case of the most widespread domestic vectors.

Following from a series of field trials, and large-scale campaigns in São Paulo (Wanderley 1994; Jacintho da Silva 1999) and in several states of Venezuela (Guevara de Sequeda et al. 1986) the government of Brasil in 1983 launched the first national campaign designed to eliminate all domestic populations of their main Chagas disease vector, Triatoma infestans. Although highly successful at the outset (Dias 1987) the campaign was interrupted in 1986 by re-emergence of Aedes aegypti in several coastal cities, with the associated threat of widespread dengue outbreaks. The national Chagas disease campaign essentially collapsed as field staff engaged in Chagas disease control were redeployed against urban Aedes.

Two key features of this Brasilian campaign guided development of what then became known as the Southern Cone Initiative (INCOSUR). The national campaign had shown the effectiveness of the control techniques and operational strategy, but also indicated that elimination of T.infestans would be unsustainable in Brasil if neighbouring countries did not carry out similar campaigns. Control interventions would also be unsustainable if interrupted by changing ministerial priorities. In essence the campaign would require a wider geographical scale together with continuous political commitment. The enlightened response was to develop a multinational initiative designed to cover the entire geographical distribution of T.infestans, some 6 million km² covering parts of seven countries (Argentina, Bolivia, Brazil, Chile, Paraguay, southern Peru, and Uruguay). This initiative, coordinated by the PAHO, was launched in 1991 following resolution of the Ministers of Health at their meeting in Brasilia, with the stated objective of interrupting the transmission of Chagas disease by elimination of T.infestans (and suppression of other domestic vectors in the same areas) and improved screening of blood donors to reduce the risk of transfusional transmission (Schofield and Dias 1999).

Although the INCOSUR objectives have not been entirely realized, substantial progress has been made. Through large-scale indoor spraying of infested premises, the distribution of T.infestans has been reduced from its predicted maximum of some 6.28 million km² (Gorla 2002) to less than 1 million km². Uruguay was formally declared free of Chagas disease transmission due to T.infestans in 1997, Chile in 1999, Brasil in 2006, together with several provinces and departments of Argentina and Paraguay. Some progress has also been made in parts of Bolivia and southern Peru, although much remains to be done in the Chaco region of Bolivia and northwestern Argentina. Concurrently with the vector control interventions, substantial progress has also been made on improved serological screening of blood donors, with all countries of the region now achieving close to 100% screening coverage.

The Southern Cone Initiative also encouraged similar initiatives elsewhere in the endemic regions of Latin America. By 1995, detailed planning discussions were being held for two further multinational initiatives, covering the Central American and Andean Pact regions (Schofield et al. 1996). Both were launched by resolution of the respective Ministers of Health in 1997, with the primary targets of eliminating the main domestic vector of these regions (Rhodnius prolixus), suppressing secondary vectors such as T.dimidiata and R.ecuadoriensis, and also improving serological screening of blood donors. With initial support from the government of Taiwan, followed by sustained support from the Japanese International Cooperation Agency (JICA), substantial progress has been made in the Central American initiative (IPCA), such that R.prolixus now seems to have been eliminated from El Salvador and Guatemala, and may be close to elimination from Nicaragua and Honduras, with substantial reductions also in the rate of domestic infestation with T.dimidiata in these countries (Yamagata and Nakagawa 2006). Both Central America and the Andean Pact countries have also made considerable improvements in blood donor screening, with coverage close to 100% in most of the region. However, in spite of strenuous efforts by University-based research teams, much remains to be done in terms of vector control interventions in the Andean Pact countries. In Mexico also, in spite of progress in epidemiological research and surveys, and some improvements in blood donor screening, much remains to be done in terms of elimination of the domestic vector populations in most of the endemic rural areas.

For the Amazon region, encompassing parts of nine countries, the epidemiological situation is quite different. Long considered to be ‘non-endemic’ for Chagas disease, in spite of the presence of a very wide range of silvatic vectors, the Amazonian Chagas Initiative (AMCHA) focuses on epidemiological surveillance rather than vector control (Guhl and Schofield 2004). Vector-borne transmission of Chagas disease in the Amazon region is mainly due to silvatic bugs occasionally entering houses and causing oral-route transmission by contamination of food or drink. Domestic colonization is relatively rare (except for some populations of T.maculata on the northwestern fringes of the region) so that there is little scope for the type of vector control interventions widely used in the other multinational initiatives. Instead, the emphasis is on early detection and treatment of any new cases that occur. This is carried out through the primary health care networks, and also by the networks of malaria slide microscopists now trained to identify T.cruzi (as well as Plasmodium) in febrile patients. There is also increasing research on Amazonian vector species, which can help to guide the surveillance initiative to areas where additional transmission may be expected (e.g. Abad-Franch et al. 2009).

To complete the suite of multinational initiatives against Chagas disease, 2007 saw the launch of a further initiative focusing on the problem of Chagas disease in the so-called ‘non-endemic countries’. At the time of writing, this initiative encompasses the USA, Canada, Japan, much of Europe, and Australia—all countries that have received considerable inflows of immigrants from Latin America, particularly over the last decade. The focus is on health education (clinicians as well as patients), treatment and counselling of chronic cases, serological screening (or deferral) of blood and organ donors from Latin America, and also maternal health care with serological screening during pregnancy and checks for possible transplacental transmission. The initiative, coordinated by WHO Geneva, is much-needed because of the paucity of experience with Chagas disease in the ‘non-endemic’ countries, and procedures, both for diagnosis and treatment, currently vary enormously between these countries. In 2007, the USA Food and Drug Administration (FDA) recommended screening all blood donors (regardless of origin) resulting in substantial expense for executive agencies such as the American Red Cross; in the UK, by contrast, potential donors of Latin American origin are deferred, but routine serological screening is not yet carried out.

Throughout the twentieth century, the main focus of Chagas disease control has been on elimination of the domestic vector populations, and improved serological screening of blood donors. Much has been achieved in both fields, as shown by the marked decline in prevalence estimates, steadily reducing seropositivity rates in blood donors, and also by declining public awareness of the disease and its vectors in previously endemic areas. But abundant domestic vector populations remain in parts of Bolivia, Peru, northwestern Argentina, northeastern Brasil, Colombia, Ecuador, Venezuela, Costa Rica, and Mexico, and the technical knowledge and experience about how to eliminate them will count for little without the requisite organization and political commitment. Similarly, although most of the endemic countries have achieved substantial improvements in serological screening of blood donors, with coverage close to 100% in most areas, the quality of the screening remains variable, partly due to the range of different diagnostic approaches used, partly due to different management practices, and also to the different levels of quality control by external reference laboratories.

To a large extent, vector control operations over the last 15 years have concentrated on eliminating domestic species that appeared to have been accidentally spread from their presumed origins in association with human migration or trade. T.infestans, for example, appears to have been spread from its origins in the Andean valleys of Bolivia in association with human migrations (Schofield 1988; Panzera et al. 2004), while R.prolixus appears to have spread in Central America following an escape of laboratory-reared bugs originating from Venezuela (Dujardin et al. 1998; Zeledón 2004). The idea of targeting such populations, apart from their vectorial importance, carries the politically-convenient hope that once elimination has been achieved then that situation will be sustained without further intervention. But this is unlikely in practice, and takes no account of local silvatic or peridomestic species that may also invade and colonize rural houses. Instead, greater emphasis is now being placed on the idea of eliminating any domestic population of Triatominae, of whatever species, and developing a sustainable system of surveillance through which any future house infestations can be detected and treated, and any new cases of infection can also be diagnosed and treated—but with the implicit assumption that these surveillance activities must be continuous, carried out through an annual routine (Schofield et al. 2006).

The operational approach currently being developed relies on surveillance of school-age children (Ponce and Schofield 2004). The idea is to make an annual check of children to ask if they have seen any bugs (using life-size photographs of the local vector species) and also to make a rapid serological diagnosis of any new infections amongst the children (using one of the current generation of immunochromatographic rapid tests for T.cruzi antigens in peripheral blood). Such periodic health checks can be integrated within similar approaches to child health, such as vaccination programmes and head-lice control. Diagnoses of seropositive children are then confirmed by a quantitative ELISA test, with specific treatment and follow-up, and the cases can be taken as evidence of possible house infestation requiring a check and possible focal intervention by vector control specialists. Similarly, absence of infection and absence of visual sightings of the bugs can be taken as evidence of no current house infestation in the catchment area of that school. But the most important aspect of this approach is to establish it as an annual routine, accepting that in Latin America, triatomine bugs can and will enter houses, and can cause occasional transmission of T.cruzi. For the future however, no one should be obliged to live with domestic colonies of Triatominae, and any new cases of infection can and should be diagnosed and treated.

Preparation of this chapter has benefited from international collaboration through the ECLAT network. I also thank Dr Alejandro Luquetti (Universidade de Goiania, Brasil) and Dr Pedro Albajar-Viñas (WHO, Geneva) for guidance on clinical aspects.