Introduction

American trypanosomiasis was first described by Carlos Chagas in 1909 [1]. Currently, we refer to Trypanosoma cruzi infection when no evidence of organ damage is present, while Chagas disease (CD) is the term used for the illness that results from T. cruzi infection. Up to 21 Latin American countries and the United States are considered endemic for T. cruzi infection and it is estimated to affect 6–7 million people [2]. Of all endemic countries, Bolivia has by far the highest reported prevalence, currently estimated at 6.1% [2]. Bites by triatomine vectors still represent the main route of transmission of T. cruzi to humans, but transmission is also possible orally (i.e., by ingestion of contaminated juices and foods) [3], congenitally from mother to child, by blood transfusion or organ donations [3–8].

In less than 5% of cases, individuals develop an acute symptomatic form of the disease [4]. After the acute phase is over, and if the infection remains untreated, T. cruzi progressively sets in the tissues with a high tropism for the heart and digestive tract. However, 60–70% of infected patients will not ever develop clinically significant organ damage [4]. Cardiac involvement is the most common organ damage, known as Chagasic cardiomyopathy (CC), which accounts for 25–30% of patients suffering from T. cruzi infection. Conduction disorders and dilatation of cardiac structures are typical manifestations, whose clinical spectrum comprise from asymptomatic to fatal outcomes [4]. Digestive involvement of CD (DCD) is less common (10–20%) and includes dilatation and dysfunction of digestive structures -mainly esophagus and colon [4].

Serological tests are the technique of choice for diagnosis of chronic T. cruzi infection [5]. Screening for cardiac involvement is indicated even in asymptomatic patients. The 12 lead ECG is the recommended test to detect conduction disorders and arrhythmias and should be offered to all patients [6,7]. Structural cardiomyopathy is classically assessed by thorax radiography, but last consensus recommends echocardiogram as it has more sensitivity and specificity [7–9]. Barium enema and esophagogram are recommended to assess digestive abnormalities due to CD but should only be performed when digestive symptoms are present [10].

In the last two decades, the global distribution of T. cruzi infection has changed due to population migration from endemic Latin America countries to Europe and other non-endemic settings [11]. In the early 2000s there was a large migration of Bolivian population towards Europe, mainly to large urban areas in Spain, due to the economic crisis and immigration restrictions in place in the USA [12].

A 2015 metanalysis estimated an average prevalence of T. cruzi infection of 4.2% in Latin American migrants living in European countries and, currently, Spain is the European country with the highest estimated prevalence [13]. Moreover, new cases of T. cruzi infection can occur in these traditionally non-endemic countries by non-vector-borne transmission. Thus, although screening for T. cruzi in populations originating from high endemic areas in Latin America should recommended in non-endemic countries, most European countries do not include this approach in their national health policies [14]. In Spain, there is no common legislation or initiative for T. cruzi screening; yet, in Catalonia an official program to detect and treat congenital T. cruzi infection has been implemented since 2010 [15].

In the last twenty years, several descriptions about clinical and epidemiological features of migrants at risk of T. cruzi infection have been published, but usually focused on blood donors or pregnant women, and the sample sizes of such studies were not large enough to describe the epidemiological profile of this population [16–20]. Here we present the clinical and epidemiological profile of a large sample of Latin American migrants attending the International Health Service (IHS) of the Hospital Clínic of Barcelona, Spain, which is the reference center for tropical diseases in the region.

Materials and methods

Results

General demographic and epidemiological variables

A total of 2,820 individuals born in T. cruzi endemic countries were included in the 17 years study period. The number of patients tested for T. cruzi infection per year progressively increased until 2009, when there was a peak of 356 individuals. The mean age was 36.8 years (SD 11.0). The median (IQR) of the duration of residence in Spain until the first visit to the IHS was six years (3.00–10.00) [2,813]. Most individuals came from Bolivia (Table 1), and 93.2% (n = 2,275/2,441) of these where from Cochabamba, Santa Cruz, and Chuquisaca Departments, all of which are considered highly endemic areas for T. cruzi transmission (Table 1). The overlap between reported risk factors is represented in the Fig 1. The greatest overlap (n = 987/1,889 individuals, 52.2%) was between the three risk factors related to the vector (rural area, adobe house, and exposure to the vector), while 101 individuals (N = 1,889; 5.35%) reported also having a mother with T. cruzi infection, and 114 (N = 1,889; 6.0%) had been residing in a rural endemic area and in an adobe house.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. General demographic and epidemiological variables of participants (N = 2,820).

https://doi.org/10.1371/journal.pntd.0011330.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Venn diagram showing the overlap between risk factors of T. cruzi infection. Venn diagram showing the overlap between risk factors of T. cruzi infection recorded in the database.

Each oval represents a risk factor. The numbers represent how many individuals have the corresponding risk factor to the area where it is located in the diagram.

https://doi.org/10.1371/journal.pntd.0011330.g001

Seropositivity and its associated factors

Among 2,820 participants, 1,382 participants were diagnosed of T. cruzi infection based on serology (Fig 2). Variables associated with T. cruzi infection were analyzed and are summarized in Table 2. Proportion of T. cruzi infection by country is detailed in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Flowchart of screened individuals.

Flowchart of screened participants of this study. Diagnosis of T. cruzi infection was based on the serology.

https://doi.org/10.1371/journal.pntd.0011330.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Variables associated with T. cruzi infection using a multivariable model.

https://doi.org/10.1371/journal.pntd.0011330.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Frequency of T. cruzi infection according to country of origin.

https://doi.org/10.1371/journal.pntd.0011330.t003

Organ damage in patients with T. cruzi infection: Chronic Chagas disease

Out of 1,382 patients with a confirmed positive result for T. cruzi, the first ECG of 662 patients was recorded in the study database. Among them, 79.0% (n = 523/662) had a normal result, while 21.0% (n = 139/662) had an altered result and 22 of them were found to have more than one disorder in the ECG. Among the 178 patients with an echocardiogram performed and recorded in the database, a total of 122 patients (68.5%) had a normal result, while 56 of them (31.5%) had an altered result. Cardiac dilatation was the most frequent disorder (n = 19; 10.67%): 12 of them with ventricular dilatation (6.7%), seven patients with atrial dilatation (3.9%); two of them having a four-chamber dilatation (1.1%). Other findings, not considered to be CD related were: 14 patients with mild valve insufficiency, four of them with more than one valvopathy (11 cases of tricuspid insufficiency, six of mitral insufficiency, and one case of aortic insufficiency), and six had left ventricular hypertrophy. One patient had a severe pericardial effusion due to a non-chagasic etiology. Out of 153 patients with registered ejection fraction, six had a left ventricle ejection fraction (LVEF) of less than 35%, two patients had a LVEF between 35 and 45% and 145 had more than 45%. Main findings in the ECG and echocardiogram are summarized in Table 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Main ECG and echocardiogram findings in patients with T. cruzi infection.

https://doi.org/10.1371/journal.pntd.0011330.t004

In sum, 148 participants were diagnosed with CC. Of them, 16 patients (10.8%) had a normal ECG but with altered echocardiogram, nine of whom due to diastolic dysfunction (56.3% of them).

An esophagogram was recorded in 66 patients. Of these, 89.4% had a normal test (n = 59) and seven patients had a gastroesophageal dysfunction. Barium enema was registered in 71 patients, of whom 30 (42.3%) had a normal test. Dolichocolon was present in 29 patients, colon dilatation in 11 patients, including two who met the megacolon criteria. Cardiac involvement was present in patients from Argentina (n = 2), Bolivia (n = 138), Brazil (n = 3), Colombia (n = 2), Ecuador (n = 1), Paraguay (n = 1) and Venezuela (n = 1), while it was not present in patients from other countries. Patients with digestive involvement were from Argentina (n = 1) and Bolivia (n = 22), while it was not present in patients from other countries. Regarding the frequency of cardiac or digestive involvement in the first evaluation, 148 patients (17.1%) were registered to have CC, while 23 patients (2.6%) were recorded as DCD and 703 patients (80.8%) did not have any evidence of organ damage. Due to missing data, 512 patients were excluded of this analysis.

Discussion

Several studies performed in non-endemic countries have assessed the clinical and epidemiological profile of people at risk for T. cruzi infection [13,17–19,29,30]. However, most of these studies, especially in the USA [31–33], had either small sample sizes or focused on specific populations (e.g., blood donors or pregnant women). Furthermore, prevalence of T. cruzi infection in these studies was very low because most of the migrants who settled in North America originated from countries with low T. cruzi prevalence [31–33]. To our knowledge, this study describes one of the largest sample of patients at risk of T. cruzi infection in a non-endemic setting.

Trends in the influx of medical consultations are determined by Bolivian population migratory flows, since this population accounts for 86.6% of the participants. It is worth noting that patients from our study spent a median of six years in Spain before screening for T. cruzi infection was performed, which is longer than observed in other previous European studies [29,34]. This delay may be explained by barriers to healthcare, such as the population’s legal immigration and residential status, poor socio-economic conditions, and cultural barriers. It is estimated that a significant number of people living in Spain with T. cruzi infection are unaware of their condition. In fact, the index of underdiagnosed cases of T. cruzi infection is estimated to be of 71% [35] and, indeed, most of our participants had never been tested for T. cruzi infection prior to our assessment. The great predominance of women and working-age people from endemic countries reflects the demographic profile of Latin-American migrants who came to Europe in the early 2000’s, as job offers were usually as domestic workers, which was socially attributed to women [12].

Health promotion campaigns among the Bolivian community were decisive in defining the attendance trends in the medical attention seeking in the IHS of Hospital Clinic de Barcelona, as shown by several health promotion programs directed to communities from T. cruzi endemic countries [36]. Most patients came because they were advised by a family member or a friend, which reveals the great importance of the social capital in the knowledge and care of the disease.

This study presents higher proportion of T. cruzi infection among participants than other European studies [13]. However, a selection bias was present because the participants were attending a referral center to which sometimes they were directed after a positive test. For this reason, proportion of seropositivity from this study should never be used to estimate the prevalence of T. cruzi infection in the general Latin American population living in Spain, which was already estimated to be of 2.1% in a recent study [35]. Nevertheless, we must consider that Latin Americans in Barcelona come from highly endemic areas compared with those in other European cities [37]. In fact, most individuals were from the Bolivian departments of Cochabamba, Santa Cruz, and Chuquisaca, which are hyperendemic areas of T. cruzi infection [5].

Factors of potential vector exposure (adobe house, rural area, to have been in contact with the vector) were present in most patients, but there was a significant overlap among all the risk factors. Thus, many participants who had received a transfusion in an endemic country or whose mother had T. cruzi infection, have also had potential vector exposure.

Congenital and transfusion risks account for the possibility of new cases in Europe and justifies the implementation of screening programs for these risk groups as a priority in health public policies [14]. Moreover, screening for T. cruzi infection in all Latin American adults living in Europe has been shown to be cost-effective [37]. Additionally, young people infected with T. cruzi on an early chronic phase of T. cruzi infection seems to be the most benefited from an early diagnosis and anti-trypanosoma therapy [38,39]. Even more so in women of childbearing age, since treatment before pregnancy dramatically reduces the chances of congenital transmission [40,41].

Results from ECG assessment show RBBB as the most frequent and with the strongest association with T. cruzi infection, consistent with previously published studies [42]. LAFB was more frequent than sinus bradycardia, contrary to what was observed in previous studies [20,43]. Although the sensitivity of the ECG may be limited because it cannot evaluate function and morphology, it is an accessible and low-cost test capable of detecting potentially serious abnormalities. Indeed, more than 10% of patients with diagnosis of cardiac involvement had a normal ECG result and were diagnosed by echocardiogram. Similar data observing altered echocardiogram with normal ECG have been previously published [44,45]. Therefore, echocardiogram is considered a necessary test for this evaluation by most authors and guidelines [7–9], our findings pointing in the same direction. Typically, wall dilatation has been considered the main echocardiogram disorder in patients with T. cruzi infection, but diastolic dysfunction is suggested to be an earlier alteration of CC [26,46]. Our study shows diastolic dysfunction as the second most frequent disorder among patients with T. cruzi infection. Although this frequency is not very high compared with that of general population [47], its presence may be considered of Chagasic etiology, especially taking into account the young age of our patients. Presence of isolated mild tricuspid or mitral insufficiency without dilated ring should not be interpreted as a CC, as several studies have demonstrated a high prevalence (up to 80%) of mild regurgitation in healthy subjects [48] and, to our knowledge, there are no studies which associate it with disease progression. Regarding digestive involvement, dolichocolon was very frequent in patients with a barium enema performed. In agreement with previous work [10,49,50], dolichocolon was not considered as a digestive involvement for CD, thus the observed frequency of digestive involvement was relatively low. Furthermore, digestive involvement was only present in patients from southern countries of Latin America (Bolivia and Argentina). This would be consistent with the notion of tissue specific tropism of different strains of T. cruzi with characteristic geographic distribution [51], but the validity of this observation cannot be concluded since practically all the patients come from those same countries.

More than 80% of patients were not found to have CC or DCD in the first evaluation. This form of the disease is classically known as "indeterminate form" of chronic T. cruzi infection [4] or rather "chronic phase without demonstrable pathology" [52]. Most patients were young, so a number of them will develop some form of organ damage in the following years at a rate of 1–2% per year [53]. Hence, age could explain the lower prevalence of CC and DCD in studies from non-endemic countries compared with those from endemic countries [29]. The lack of clinical and biological markers of progression limits a more accurate follow-up in this big group of patients. The future availability of such tools should be highly relevant for patients’ clinical management and healthcare routes [54].

One limitation of this study is that comorbidities were not specifically addressed. We acknowledge that screening for comorbidities will become increasingly important with the aging of Latin American migrants, but we posit that the impact on the overall findings of this study would probably be small due to the mean and median age of the participants included in our study. Also recall bias could be present in the answers to the epidemiological questionnaires. As the participants were recruited in a regional reference center a selection bias was present. Observer bias could also be present as electrocardiogram is an observer-dependent test, and the questionnaires were filled in by different physicians. This bias was probably reduced as diagnostic criteria of electrocardiographic disorders were established in detail. We were able to include the status of the patient at the moment of the screening and at the moment of the initial diagnosis of T. cruzi infection, but not their follow-up, that is why the number of ECG and echocardiogram recorded for the analyses was little compared with total sample size. Despite this discordance, the absolute number of tests registered and analyzed was also high and allowed us to extract enough data for the analysis and discussion. It would be interesting to analyze the longitudinal evolution of this cohort in terms of the prognostic value of the alterations found in this first evaluation.

Conclusions

Currently, T. cruzi infection is distributed throughout Europe due to migratory flows from Latin America of the last two decades. Big Spanish cities have received the largest population at risk. Although arrivals from endemic countries have decreased in the last years, we still observe many individuals at risk who have not been previously tested. Moreover, since universal testing of pregnant women is not widely implemented, children with T. cruzi infection may be born and remain undiagnosed. In this study, individuals from endemic countries settled in Barcelona were mostly working-age women who came between 2004 and 2007. Migrants with a similar epidemiological profile, especially of Bolivian origin or descent, should be the target for a national or European screening program. Thus, health promotion programs and an adequate information to the patient in the consultation are essential to carry out the screening program.

In concurrence with other studies, previous contact with the vector was the most frequent and strongly associated risk factor with T. cruzi infection among those related with vector exposure.

The primary assessment in patients with chronic T. cruzi infection should include an ECG and echocardiogram. RBBB and LAFB were the most typical ECG disorders among T. cruzi infection patients in our series, while wall dilatation and diastolic dysfunction were the most frequent findings in the echocardiogram. In our series, as well as in other previous studies, a relevant proportion of patients with CC were diagnosed by echocardiogram having a normal ECG, which reinforces the echocardiogram as an essential test in the assessment of CC. Although dolichocolon was the most frequent finding in barium enema, it should not be interpreted as a digestive involvement of CD. It is expected that prevalence of CC and DCD will gradually increase in Europe as migrant populations settled in these countries age.

We are gradually learning more about T. cruzi infection, but there are still important knowledge gaps regarding this infection, especially about the lack of clinical or biological prognostic factors. Further studies will be necessary for a better understanding of its epidemiological and clinical behaviour.

Supporting information

Acknowledgments

We first would like to thank all the participants of this study. We also thank the Hospital Clínic de Barcelona and ISGlobal, as well as the nursing team, the medical team, and the staff who have helped in this project.