Introduction

Chagas disease (CD), caused by the protozoan parasite Trypanosoma cruzi, was listed in 2010 by the World Health Organization (WHO) [1] as one of the neglected tropical diseases (NTDs).

However, 6 to7 million people are estimated to be infected around the world [2]. Due to migratory flows in recent decades, CD has shifted from being confined to endemic areas in Latin America, to being a disease that can be diagnosed worldwide.

Traditionally, in endemic areas, infection occured mainly through vector-transmission. Successful results were obtained through the action taken by countries to control vector and transfusion transmission of the parasite. These initiatives led to substantial reductions in these ways of transmission [2]. Vertical (mother-to-child) transmision is the main transmission route in non-endemic countries [3,4,5].

In Europe, Spain is the country with the highest number of CD cases [6]. Spanish guidelines [7] recommend the screening for CD in pregnant women who come from endemic countries, in children born to infected mothers (ideally from birth) and of all children coming from endemic areas, regardless of their age. For the diagnosis of congenital CD in children born to infected mothers, these guidelines recommend preforming: Polymerase Chain-Reaction (PCR) and /or microhaematocrit detection at birth repetead in a second determination, and a serological determination at 9 months of age or later. Nowadays, although this screening is recommended and increasingly widespread it is not universal, so undiagnosed pediatric cases are still ocurring.

CD is usually asymptomatic during childhood, as well as in adulthood. The disease has an acute and a chronic phase. Following the acute phase, most infected people enter a prolonged asymptomatic phase of the disease (called “chronic indeterminate”). Late complications, classically associated with CD, occur in the chronic phase, “chronic determinate”, in approximately one third of the patients [8]. The outcome in children depends on the age at treatment initiation. Treatment is not only better tolerated in younger children but also the possibilities of cure are greater.

There are two approved drugs for CD: benznidazole (of choice) and nifurtimox (alternative treatment). Although there are no randomized clinical trials that compare benznidazole and nifurtimox, benznidazole is generally preferred due to its better tolerability and tissue penetration, as well as its possible higher effectiveness. [9]

There are few publications on pediatric CD series in non endemic regions [10,11]. Children with CD are especially vulnerable, due not only to the high pathogenic potential of the disease but also to their frequent involvement in challenging socioeconomic situations, including migration. [12]

The present study had two principal aims. Firstly, to describe clinical and epidemiological characteristics of CD in a pediatric population diagnosed in 10 hospitals in the Community of Madrid, Spain. Secondly, to describe the safety and effectiveness of the treatment on this pediatric population.

Patients and methods

Results

Fifty-one cases of pediatric CD were identified (59% male, 41% female). Twenty-six (50,9%) were born in Spain and twenty-five (49%) came from endemic areas. All the mothers had Latin American origins: 94% (n = 48) came from Bolivia, 2%(n = 1) from México, 2%(n = 1) from Honduras and 2%(n = 1) from Paraguay. None of them had received treatment for CD prior to their children’s diagnosis, despite that 5 (10%) were aware of the diagnosis previously.

The median age at diagnosis was 103.5 months (IQR: 1,4–167). They ranged in age from 0 to 17 years-old. Sixteen cases (31.4%) were diagnosed under one year of age, the median age of diagnosis in this group was 0.7 months (IQR 0.09–1.54). For those older than one year-old (n = 35, 68.6%), the median age of diagnosis was 11.08 years (IQR: 8.14–14.7). The results broken down by age group (under 1 year-old vs over one year of age, are detailed in Table 1)

Download:

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

Table 1. Comparative table regarding age (<1 year-old vs > 1 year-old).

https://doi.org/10.1371/journal.pntd.0010232.t001

Vertical transmission, as defined in the Patients and Methods section, was considered in 26 cases (50.9%). Of these Spanish-born cases, follow-up from birth was performed in 61.5% (16/26). The other 10 (38,4%) were missed oportunities of early diagnosis (diagnosed at more than one year of age). They all (10) were born before 2013, the year of publication of Spanish guidelines for the diagnosis and management of pregnant women and children with Chagas disease [7]. None of these ten patients had a history of travel to endemic areas nor previous blood transfussions either.Their diagnosis (10/26) were reached because they were referred when the mothers were diagnosed with Chagas disease.

In the remaining 25, the route of transmission was not determined, since the children were not born in Spain, and therefore whether infection occured through vertical or vectorial transmission could not be ascertained.

Follow-up

Eigtheen (18/51) patients (35,3%), were lost to follow-up: in 2 cases (4%) before starting treatment (because they returned to their country of origin), in 1 (2%) during treatment (as he moved to another Spanish autonomous community). The remaining 15 (29,3%) patients were lost after having completed the treatment.

For the patients who completed the treatment and continued on follow-up (n = 33): 24 (47%) continue to be monitored (as the serology has not yet been negative), while 9 (17.6%) have been deemed cured and discharged.

Cure (negative serology in at least two determinations) was observed in 80% of patients diagnosed when they were younger than one year-old (8/10), compared to 4.3% of those older than one (1/23), being this difference statistically significant (p <0.0001).

From the 24 patients who had positive serology at the last follow-up visit: 11 (45.8%) had decreasing titers (only 1 of them started treatment within the first year of life) and 13 (54.16%) did not show a decrease in antibody titers (12 of them older than one year-old at diagnosis).

The patient flow chart of the study is detailed in Fig 1

Download:

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

Fig 1. Patient flow chart of the study.

https://doi.org/10.1371/journal.pntd.0010232.g001

For all patients who had a positive PCR, it became negative after treatment. The mean time at which the PCR was negative (after the end of treatment) was 4.5 months (Range: 0.9–15.9). The average follow-up time of cured patients (time until negative serology) was 20.1 months (Range 8–57).

The detailed information of the 33 patients who completed the treatment and continued follow-up is in Table 2

Download:

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

Table 2. Age at the start of treatment, time of follow up and outcome in the 33 patients that continued follow-up.

https://doi.org/10.1371/journal.pntd.0010232.t002

The average follow-up time of this 33 patients was 57.8 months (Range 8–168), median time 36 months (IQR 18.5–72).

Discussion

To our knowledge, we present one of the largest pediatric CD series diagnosed, treated and followed in a non-endemic area, regardless of their origin. As in other published series [10,11], Bolivia was the most frequent country of origin of the cases or of their mothers in our study.

As main findings of our study, we highlight the need to perform screening programs of CD to allow early diagnosis and treatment, as most patients were asymptomatic. It is also important to highlight the significantly higher cure rate in early treated chidren (younger than one year-old), with lower incidence of adverse events to treatment also for this group. The high rate of loss to follow-up in this vulnerable population is also noteworthy.

It is important to carry out children screening from birth when the mothers have a previous positive serology (33.5% of patients in our series) [7], as well as the need to study the previous children of these mothers (in our series, 43% of patients were so diagnosed). Despite this, of the patients in the series born in Spain, 38,4% had not been followed from birth, maybe because this follow-up recommendations (including PCR and /or microhaematocrit detection at birth and 1 month later, and a serological determination at 9 months of age or later) were not published in our country until 2013[7], and all of them were born before this year. Although there were already recommendations before, there was perhaps less awareness. Even today this screening is not done systematically in all hospitals, even though it is becoming more widespread.

Furthermore, a good strategy to avoid disease underdiagnosis would be to screen people from endemic areas in primary care, particularly women of childbearing age. It is of paramount importance to treat the disease prior to pregnancy, since it might reduce or even prevent the rate of congenital transmission [7,16,17]. In our work, in 10% of the cases, even though the mothers knew the diagnosis before gestation, they had not been treated.

Another group to which special attention should be paid are immigrant population and internationally adopted children from Latin America [18]. 7% of the patients in our case series were diagnosed by screening from the intercountry adoption consultation.

CD is usually asymptomatic when there is vertical transmission, or other transmission route during childhood. In our case, 98% of the patients did not present suggestive symptoms of the disease at diagnosis

For patients with congenital infection, presenting with symptoms at birth, these may be variable and usually involve several organs and systems, as in the case of the only symptomatic patient in our series [15].

In other published series of pediatric-age patients in non-endemic areas, most cases have also been asymptomatic [10,11,19].

In their work, Fumadó et al [10] describe a series of 22 cases, 5 of congenital transmission, all asymptomatic, and mostly from Bolivian origin. Rodriguez-Guerineau [11] et al, report 45 cases of Chagas disease diagnosed or treated in Spain and Switzerland, with only one symptomatic case, with gastrointestinal complications in the chronic phase of the disease (megaesophagus). In our series there were no cases of megaesophagus or megacolon.

In their recent paper, Simón et al [19] describe 40 cases of Chagas disease in pediatric population, and they reported that only 3 of 12 children in the acute phase presented with signs and symptoms of Chagas disease: low weight at birth, hepatosplenomegaly, respiratory distress, myocarditis, anaemia and jaundice.

Due to the usual asymptomatic course, it is especially important to develop screening programs for the early diagnosis of the disease.

During the first year of life, a positive serological test can result from the placental transfer of maternal antibodies. Therefore, the diagnosis should be based on methods for direct parasite detection (PCR or blood smear examination-microhaematocrit), to be able to start early treatment if a positive result is obtained. In addition, these tests are more sensitive at this period because the acute infection phase presents with greater parasitaemia [14]. For the patients diagnosed within this age range of our series, the diagnosis was carried out in 94% by parasitological tests.

In all cases in which PCR was performed on infected patients younger than one year-old, no false positives were identified (a positive serological test was confirmed after 9 months of age), in keeping with other studies [5]

For patients older than one year-old, like for adults, serology continues to be the most frequent method for diagnostic purposes.

Treatment during pediatric age is well tolerated and effective in 50–70% of cases, therefore it can decrease the risk of complications and future vertical transmission [17]. Benznidazole is the treatment of choice for CD (first treatment option in 98% of treated patients in our series), with nifurtimox being a normal second-line therapeutic option [7]. Benznidazole achieves cure for up to 100% of children with congenital infection treated during the first year of life and 76% in patients with acute infection, decreasing to 9–15% in adult patients with chronic disease [9]. Nifurtimox has cure rates of around 86% in children and 7–8% in adults [9].

Recently, a paediatric formulation of nifurtimox has been approved and it can probably will be used as a first line treatment as well in future.

The negativization of serology, and therefore cure in our series, is clearly greater in the group of patients diagnosed and treated within the first year of their life, in accordance with other studies (Rodriguez Guerineau et al, Simon et al). We highlight the low cure rate in those over than one year of age, probably due to the high median age at diagnosis (11.08 years). When the infection is not diagnosed in time, children enter the chronic phase of CD and they require a longer follow-up in time, in order to achieve cure criteria.

The main limitation in evaluating treatment response for CD (in patients older than 1 year, as in adults) stems from the need for a long-term follow-up (years to decades) to demonstrate seroreversion from positive antibodies to a negative serology with conventional tests. Alternative early markers of cure have been suggested, such as decrease of total anti T.cruzi antibody titers or ELISA anti F2/3[20], but nowadays the only criteria of CD cure is seroreversion [9].

In our case, of the 33 patients who continued follow-up, although only 9 patients (27,2%) were considered cured due to negative serology, 11 cases (33.3%) had decreasing antibody titers and will probably end up becoming negative when the follow up period will be prolonged.

Our results are similar to those of the series by Rodriguez-Guerineau et al, where they describe that the only factor associated with the cure rate was age at diagnosis and treatment (p = 0.008), finding a cure rate of 100% in children under one year, while it was only 4.2% in those over this age (median age: 4 years).

In other pediatric series published in endemic areas [21] the results of seroconversion were variable (from 5 to 90%), for different reasons (different parasite lineages, varying age groups and varying times of post-treatment follow-up and different times since acute phase of the disease).

There is better tolerance to treatment (lower AE rate) in children than in adults [22,23], specially when administered in patients younger than 7 years-old. In a study [24] conducted on 107 patients with CD treated with benznidazole, AE were described in 41% of cases, 77.3% of wich were in children older than 7 years-old.

In our series of patients, AE due to benznidazole were reported in almost 50% of cases, more frequent in children older than one year (78.2% of patients with benznidazole AE), with a median age of 11.4 years. Among the 32 patients older tan one year treated with benznidazole, 18 (56.25%) had adverse events whereas in the 15 under one year-old, 5(33,3%) did.

Our study presents some limitations, firstly due to its retrospective nature and because it presents a significant percentage of losses to follow-up (35%), higher than in other series in non-endemic areas [9], although this information is provided in few publications under the conditions of our study. We believe that this figure is largely attributable to the characteristics of the studied population (immigrant population that sometimes has difficulties in going to medical appointments or returns to their country of origin) [25]. This high rate of losses to follow-up is of concern and efforts should be made to improve it.

Another limitation could be the variability in the serological techniques among the different hospitals. Moreover, some hospitals changed the technique over time. Therefore, the ability to detect antibodies could be different and furthermore the new serological techniques detect lower levels of antibodies. This reinforces the need to explore new cure markers of this disease, more accurate than conventional serology.

The mean follow-up time has been heterogeneous depending on the patients year of diagnosis, specially for those with no cure criteria. Although in all there has been a minimum follow-up time, until at least two serologies have been performed. This represents another limitation of the study and may condition the results interpretation since probably some patients with declining serology titers end up fullfiling cure criteria by prolonging follow-up.

However, it provides data of one of the largest series of pediatric CD patients in non-endemic areas, reinforcing the importance of early diagnosis and treatment, to achieve higher cure rates, particularly in children under one year. We also highlight, the considerable rate of treatment AE in the pediatric population older than one year of our series, higher than in other studies [19].

Prospective studies with longer follow-up period, to determinate more accurately the characteristics of treatment response in pediatric age and future complications in adulthood are needed.