Chagas disease is a parasitic infection that does almost nothing for two or three decades and then, in a substantial minority of the people carrying it, destroys the heart. It affects something on the order of six to seven million people, overwhelmingly in Latin America but increasingly in North America and Europe through migration, and it is one of the most consistently under-recognised causes of heart failure in the world.
The shape of the disease is what makes it difficult. By the time the heart declares itself, the moment when antiparasitic treatment would have helped most has usually passed. That is the argument this review is built around, and the centerpiece makes it arithmetically.
Start here: what Chagas disease is
The infection is caused by the protozoan parasite Trypanosoma cruzi. It was described in 1909 by Carlos Chagas, in a piece of work unusual in medical history for identifying the parasite, the vector and the human disease more or less at once [1].
The main route of transmission is a blood-feeding insect: a triatomine bug, known in English as a kissing bug, of which the vector species were catalogued in the standard revision of the subfamily [2] and one of which, Rhodnius prolixus, has had its genome sequenced [3]. The transmission itself is peculiar. The bug does not inject the parasite the way a mosquito does; it defecates while feeding, and the parasite in the faeces enters through the bite wound or a mucous membrane when the site is rubbed. The parasite also maintains itself in a wide range of domestic and wild mammal reservoirs [4].
Three other routes matter clinically, and two of them have become proportionally more important as vector control has succeeded. The parasite passes from mother to child in pregnancy, it passes in transfused blood and transplanted organs, and it can be acquired orally from contaminated food or drink.
The clinical course has three parts.
The acute phase follows infection and is usually mild or unnoticed, which is precisely the problem: it is the phase in which the parasite is detectable in blood and in which treatment works best.
The indeterminate form follows, and it is the defining feature of this disease. People are seropositive, have no symptoms, and have a normal ECG and chest radiograph. They can remain that way for decades, and many remain that way for life. This state has been characterised clinically and epidemiologically since the 1980s [5] [6].
The chronic determinate forms appear in a minority, years to decades later. The cardiac form is the important one: a cardiomyopathy that combines conduction disease, ventricular arrhythmias, heart failure, apical aneurysm and thromboembolism [7] [8]. There is also a digestive form — megaesophagus and megacolon — driven by destruction of the enteric nervous system, in which the neurons of the myenteric plexus are lost [9] [10]. The pathology of both was described in detail long before the mechanisms were understood [11] [12].
Vector control programmes across the Southern Cone have been one of the genuine public-health successes of the last forty years, interrupting domestic transmission across large areas [13] [14] [15] [16], with the epidemiological picture updated repeatedly since [17] [18] [19]. What those programmes did not do is treat the millions already infected, which is why the burden now is dominated by people infected decades ago.
Pillar 1: measurement and diagnosis
The test depends on which phase you are in
In the acute phase the parasite itself is present in the blood in sufficient numbers to be found directly — by microscopy, or by molecular methods. This is also the phase in which almost nobody presents, because the illness is mild.
In the chronic phase the parasite is scarce in blood and the diagnosis is serological. This is a genuine and underappreciated inversion: the phase in which the disease is diagnosable by finding the organism is the phase in which nobody looks, and the phase in which people actually present requires an antibody test.
Why two serological tests
Guidelines generally require two serological tests using different methods or antigens to establish chronic infection. The reason is arithmetic rather than dogma. A systematic review and meta-analysis found pooled ELISA sensitivity of 97.7 percent and specificity of 96.3 percent, with commercial recombinant-antigen assays reaching 99.3 percent and 97.5 percent [20]. Those are good numbers, but a specificity of roughly 96 to 97 percent applied to a population where most people are uninfected produces a great many false positives; requiring agreement between two independent assays is what makes a positive result trustworthy.
The same review is unusually blunt about the molecular alternative. PCR sensitivity in chronic disease is likely between 50 and 90 percent, with specificity close to 100 percent, and the authors conclude that PCR should not be used in clinical practice for diagnosing chronic Chagas disease [20]. A negative PCR does not exclude infection, because in the chronic phase the parasite is often simply not in the blood sample. Earlier work established PCR's role as a research and monitoring tool rather than a diagnostic one [21] [22], and it retains a real place in specific settings — detecting reactivation after transplantation, where parasite loads rise [23] [24], and in monitoring congenital infection [25].
Staging the heart, which is what actually determines prognosis
Once infection is established, the clinical question is no longer "is this Chagas?" but "has it reached the heart, and how far?"
The electrocardiogram does most of the work, and it does it cheaply. A prospective study of a Brazilian community followed for seven years found that electrocardiography during the early asymptomatic stage distinguished people with potentially lethal cardiac lesions from those with a benign prognosis: mortality was high among infected adults with ventricular conduction defects, and low and roughly equal to seronegative neighbours among those with normal ECGs [26]. Right bundle branch block, particularly with left anterior fascicular block, is the characteristic pattern. ECG abnormalities in Chagas disease have since been reviewed systematically [27], followed longitudinally in older cohorts [28], and compared directly between seropositive and seronegative blood donors [29].
Echocardiography adds ventricular function and regional abnormalities, with established diagnostic and prognostic value [30]; multimodality imaging recommendations set out how the modalities fit together [31].
Prognosis can be quantified. A risk score developed in 424 Brazilian outpatients and validated in an independent cohort uses six factors — NYHA class III or IV, cardiomegaly on radiography, left ventricular systolic dysfunction, non-sustained ventricular tachycardia, low QRS voltage and male sex — to separate patients into groups whose ten-year mortality was 10, 44 and 84 percent in development and 9, 37 and 85 percent in validation, with a C statistic of 0.84 and 0.81 [32]. That spread is the reason staging matters.
Screening where the yield is high
Screening the blood supply, screening pregnant women from endemic areas, and screening the infants of infected mothers are the three settings where finding asymptomatic infection changes what happens next. In the United States, screening of donated blood had identified 1,908 confirmed infections since 2007 as of one national analysis [33].
Centerpiece: a simple simulatable model of the silent decades
The question that decides how this disease should be managed is quantitative: if most infected people never develop heart disease, why treat, follow, and worry about the ones who feel entirely well?
The answer is that "most" and "a small annual rate" are not the same thing, and over the timescales involved the difference is enormous.
Model the indeterminate form as a single state that people leave, at a constant annual hazard, for a cardiomyopathy state. This is the simplest two-state continuous-time Markov chain, and it gives the fraction who have developed cardiomyopathy after t years as C(t) = 1 − exp(−λt), where λ is the annual progression rate. There is one parameter, and it has been measured.
A retrospective cohort study followed initially healthy Brazilian blood donors with a T. cruzi-seropositive index donation, alongside age-, sex- and period-matched seronegative donors, and re-examined all of them roughly a decade later with ECG and echocardiography adjudicated by blinded core laboratories. Among 499 seropositives, 120 (24 percent) had definite Chagas cardiomyopathy; among 488 seronegatives, 24 (5 percent) did — an incidence difference of 1.85 per 100 person-years attributable to the infection [34]. That attributable rate is λ, and because it is a difference it already has the background rate of non-Chagas cardiomyopathy subtracted out.
The model is checked against the cohort's own headline result. Given only the rate, it predicts a cumulative incidence of 17.7 percent at the study's 10.5-year mean follow-up; the study observed a 19.1 percent difference in cardiomyopathy prevalence between its arms — a number the model was not given. The agreement is within one and a half percentage points.
Run it forward and the teaching point appears. At 1.85 percent per year, 31 percent of an infected cohort has cardiomyopathy at 20 years, 43 percent at 30 years, and 52 percent at 40 years. A rate small enough to sound reassuring in any single year is, across the span of a life spent infected from childhood, the majority of the cohort. That is the entire argument for lifelong follow-up of people who feel well, and for treating before the heart is involved rather than after.
The right panel is the honest counterweight, and it uses the same source. Of the 120 seropositives with adjudicated cardiomyopathy, only 31 (26 percent) had an ejection fraction below 50 percent and only 11 (9 percent) were NYHA class II or higher — the authors' own summary was that disease was mild at diagnosis [34]. Applying those proportions to the cumulative curve gives, at 30 years, 43 percent of the cohort with cardiomyopathy but roughly 11 percent with a reduced ejection fraction and about 4 percent with limiting symptoms. Cardiomyopathy and heart failure are not the same claim, and conflating them overstates the disease in one direction while the compounding understates it in the other.
Three honest limits. A constant hazard is an approximation, and probably a poor one at fine time resolution: the seven-year community study found that those who developed cardiac lesions largely did so soon after infection, with conduction-defect incidence highest before age 20 [26], so the flat curve should be read as a decade-scale average rather than a claim about any particular year. The rate itself comes from blood donors, who are healthier than the general infected population, and from two Brazilian cities; progression varies with parasite lineage and geography [35] [36] [37]. And a two-state model has no death state, so it describes the cardiomyopathy that occurs among survivors rather than competing risks.
Pillar 2: treatment
The antiparasitic drugs, and the timing that decides whether they help
Two drugs have been in use for decades: benznidazole and nifurtimox. Both are old, both require long courses, and both have meaningful side effects — the chemotherapy of this disease has been critically reviewed more than once [38].
In early infection they work. A randomised trial of benznidazole in early T. cruzi infection established efficacy [39]. In a double-blind randomised trial in children aged six to twelve in the indeterminate phase, given benznidazole at 5 mg/kg/day for 60 days and followed for four years, 62 percent of treated children and no placebo-treated child became seronegative, and xenodiagnosis was positive in 4.7 percent of treated versus 51.2 percent of placebo children [40].
In established cardiomyopathy they do not change cardiac outcomes. This is the most important negative result in the field and it deserves stating precisely. The BENEFIT trial randomised 2,854 patients with Chagas cardiomyopathy to benznidazole or placebo for up to 80 days and followed them for a mean of 5.4 years. The primary composite outcome — death, resuscitated cardiac arrest, sustained ventricular tachycardia, device insertion, transplantation, new heart failure, stroke or other thromboembolism — occurred in 27.5 percent on benznidazole and 29.1 percent on placebo, a hazard ratio of 0.93 (95 percent confidence interval 0.81 to 1.07, P=0.31) [41].
What makes that trial genuinely informative rather than merely disappointing is that the drug demonstrably worked on the parasite and still did not help the heart. PCR conversion to negative was 66.2 percent with benznidazole versus 33.5 percent with placebo at end of treatment, and still 46.7 versus 33.1 percent at five years or more, all highly significant [41]. The parasite was being cleared. The cardiac damage proceeded anyway.
The reading that follows is the treat-early message, and it is a conclusion about timing rather than about the drug: once the myocardium has been remodelled, removing the parasite does not undo it.
Preventing transmission to the next generation
Congenital transmission is now a leading route of new infection where vector control has succeeded, and treating women before pregnancy prevents it. In a multicentre cohort of 354 chronically infected mother-child pairs, congenital infection occurred in 34 of 222 children of untreated mothers (15.3 percent) and in 0 of 132 children of previously treated women [42]. The same study found a protective effect on the mothers themselves: ECG changes compatible with chagasic cardiomyopathy were present in 15.2 percent of untreated women against 2.2 percent of those treated at 15 or older and none of those treated before 15 [42]. Treating congenitally infected infants is effective and can be monitored molecularly [25] [43].
Treating the heart
Once cardiomyopathy is established, management is cardiological: heart failure therapy, antiarrhythmic management, pacemakers and defibrillators for conduction disease and ventricular arrhythmia, anticoagulation for thromboembolic risk, and transplantation at the end — with the specific caveat that immunosuppression can reactivate the infection, which is why molecular monitoring has a role after transplant [23]. Guidance is set out in a scientific statement from the American Heart Association [7], a Brazilian cardiology guideline [44], and a World Heart Federation roadmap [45].
It is worth knowing that Chagas heart failure carries a worse prognosis than other causes. A comparison of contemporary outcomes found chagasic heart failure fared worse than other non-ischaemic and ischaemic cardiomyopathies [46], consistent with earlier risk stratification in a large Brazilian heart-failure cohort [47]. The disease also carries a distinct stroke risk, independent of other factors [48].
Pillar 3: what is unresolved
Better antiparasitic regimens. The current drugs are long, poorly tolerated and were not designed for this. Trials of alternatives and of shortened regimens have been run: posaconazole was compared with benznidazole in chronic disease [49], and the E1224 programme tested a new agent alongside benznidazole in adult indeterminate disease with several dosing regimens [50]. Neither displaced benznidazole. Shorter or better-tolerated benznidazole courses remain the most plausible near-term improvement, and drug discovery against trypanosomatids continues [51].
Who to treat in the indeterminate phase. Because BENEFIT was negative in established cardiomyopathy and the pediatric trials were positive in early infection, the contested ground is the middle: adults with the indeterminate form and decades of infection behind them. The compounding in the centerpiece is an argument for treating them; the absence of a trial demonstrating that it prevents cardiomyopathy in that group is the reason it remains debated.
Biomarkers of progression. Nothing currently identifies which seropositive person will progress. The immune response is incompletely understood, and has been described as an unsolved puzzle by people working on it [52] [53]. Parasite genetic diversity may contribute [35] [37].
The North American picture. T. cruzi is established in the United States in wildlife and in vectors, and autochthonous transmission occurs. Domestic transmission cycles have been documented in southern Texas [54], the vectors have been characterised biogeographically there [55], canine infection is widespread [56], sylvatic transmission has been modelled [57], and the ecology has been reviewed [58]. The overwhelming majority of US cases, however, are in people infected in Latin America: one analysis estimated 238,091 infections nationally as of 2012, with four states above 10,000 cases and a further seven above 5,000 [33]. The binding problem is not surveillance but care — access to diagnosis and treatment in the US has been analysed as a health-systems failure [59], and the same under-recognition applies in Europe [60] [61]. The global economic burden has been modelled and is substantial [62] [63].
Dig deeper in lmmol
Chagas cardiomyopathy is, in the end, a cause of heart failure — and an instructive one, because it is a heart failure whose cause is a treatable infection that has usually stopped being treatable by the time the heart fails [46]. The stroke risk it carries connects to stroke [48]. As a vector-borne parasitic disease of the tropics whose control has been driven by attacking the vector, it sits naturally beside malaria and dengue; and it shares with Lyme disease the awkward structure of an infection that is easy to cure early, hard to detect at that moment, and contested in what it leaves behind. For another disease where a long asymptomatic reservoir determines the epidemiology, see tuberculosis. The full collection is at health.