The authors have declared that no competing interests exist.
The effectiveness of anti-parasite treatment with benznidazole in the chronic Chagas disease (ChD) remains uncertain. We evaluated, using data from the NIH-sponsored SaMi-Trop prospective cohort study, if previous treatment with benznidazole is associated with lower mortality, less advanced cardiac disease and lower parasitemia in patients with chronic ChD.
The study enrolled 1,959 ChD patients and abnormal electrocardiogram (ECG) from in 21 remote towns in Brazil. A total of 1,813 patients were evaluated at baseline and after two years of follow-up. Those who received at least one course of benznidazole were classified as treated group (TrG = 493) and those who were never treated as control group (CG = 1,320). The primary outcome was death after two-year follow-up; the secondary outcomes were presence at the baseline of major ChD-associated ECG abnormalities, NT-ProBNP levels suggestive of heart failure, and PCR positivity.
Mortality after two years was 6.3%; it was lower in the TrG (2.8%) than the CG (7.6%); adjusted OR: 0.37 (95%CI: 0.21;0.63). The ECG abnormalities typical for ChD and high age-adjusted NT-ProBNP levels suggestive of heart failure were lower in the TrG than the CG, OR: 0.35 [CI: 0.23;0.53]. The TrG had significantly lower rates of PCR positivity, OR: 0.35 [CI: 0.27;0.45].
Patients previously treated with benznidazole had significantly reduced parasitemia, a lower prevalence of markers of severe cardiomyopathy, and lower mortality after two years of follow-up. If used in the early phases, benznidazole treatment may improve clinical and parasitological outcomes in patients with chronic ChD.
ClinicalTrials.gov, Trial registration:
Chagas disease remains as one of the most neglected infectious diseases of the world, with millions of persons infected and dozens of thousands of deaths every year. The disease has a very long natural history and few options of etiologic treatment, including antiparasitic [
Chagas disease (ChD), caused by the protozoa
ChD has a very long natural history which begins that an acute phase, generally a benign and self-limited febrile disease in childhood [
Two drugs with proven anti-parasite efficacy that are used to treat ChD patients are available: nifurtimox and benznidazole; most studies have focused on the use of benznidazole due to its better tolerability, reduced toxicity and, possibly, enhanced efficacy [
Although persistence of low-level systemic infection, including cardiac
SaMi-Trop is a prospective cohort study with two-year follow-up to date (recently funded to continue follow-up), including one baseline visit and another visit at 24 months [
For this investigation, we used two epidemiological study designs. The first was a cross-sectional analysis using baseline SaMi-Trop data collected at recruitment time (2013–14) and the second one was a prospective cohort analysis using only death as an outcome collected at the end of two-year of follow-up (2015–16, see
The exposition was self-reported treatment with benznidazole, which occurred (or not) several years before the recruitment for the study. The primary outcome was death from any cause after two-year follow-up (2015–16). Secondary outcomes were all evaluated at baseline of SaMi-Trop study (2013–14) and included the presence of typical ChD-associated ECG abnormalities, age-adjusted brain-type natriuretic peptide (NT-ProBNP) levels suggestive of heart failure, and polymerase-chain-reaction (PCR) positivity for
At baseline, all consenting participants were interviewed, with detailed questions on the duration and clinical manifestations of ChD and previous use of benznidazole, and blood was collected for
An ECG was also performed at the baseline visit. Although a report of an abnormal ECG was considered for recruitment of patients using the telehealth database, 17% of these ECG did not fulfill diagnostic criteria of the presence of an ECG abnormality after standardized central reading. The resting 12-lead ECG was recorded using a PC-based ECG machine (TEB, São Paulo, Brazil). The ECG recordings were sent electronically to the Telehealth system and the ECG measurements were obtained through automated analysis using the University of Glasgow ECG analysis program (release 28.5, issued on January 2014) [
Death during the 2-year follow-up period was ascertained for enrolled participants who could not be contacted for their 2-year follow-up visit by interviewing family members and reviewing death certificate data of the Mortality Information System database for the State of Minas Gerais.
The simple size was calculated considering the minimal number of events per variable acceptable in a proportional hazards regression analysis of 10 events per variable. For a 2-year follow-up period and annual mortality rate of 5% in chronic Chagas cardiomyopathy (10% in 2 years), the calculated sample size was about 2000 participants [
ChD subjects who received at least one course of treatment with benznidazole were classified as belonging to the treated group (TrG = 493) and those who were never treated as control group (CG = 1,320). Patients who did not know whether they had been treated with benznidazole use were excluded from analysis (n = 146). In descriptive analyses, categorical variables were presented as numbers and percentages, continuous variables were summarized as means and standard deviations.
Three measures of treatment effectiveness were calculated. The first was an unadjusted effect measure (difference of proportions, averages or odds ratio, depending on the outcome being studied) using an unpaired t-test or Fisher’s exact test. The second comparison estimated the treatment effect using linear or logistic regression, adjusted for sex, age, income, literacy, duration of known ChD, and hypertension as covariates. The third approach was a pairing method (Genetic Matching), an evolutionary algorithm that seeks to maximize the balance of the distributions of the treated units and controls [
In the paired analysis (Genetic Matching), for the continuous outcomes, the standard test of the Matching package was used by means of
We performed Kaplan-Meier survival analysis and used the log-rank test to compare time to death between treatment and control groups. Because the number of patients after two years of follow-up was small, we censored the follow-up after this time.
The study was conducted using data from the NIH-sponsored São Paulo-Minas Gerais Tropical Medicine Research Center (SaMi-Trop) study, established in a highly endemic region of Brazil [
This investigation was conducted on the 1,813 patients who were confirmed as being infected with
Characteristics | Valid number | Treatment |
Control |
P-value |
---|---|---|---|---|
Age—mean (SD) | 1813 | 54.1±11.5 | 59.4±12.5 | <0.001 |
Female sex—n (%) | 1813 | 337/493 (68.4) | 893/1320 (67.7) | 0.755 |
Illiterate—n (%) | 1807 | 149/493 (30.2) | 600/1314 (45.7) | <0.001 |
Family monthly income—US$ | 1306 | 370.87±203.04 | 417.39+209.56 | <0.001 |
Duration of ChD > 10 yr—n (%) | 1768 | 355/484 (73.3) | 728/1284 (56.7) | <0.001 |
Hypertension—n (%) | 1813 | 280/493 (56.8) | 852/1320 (64.5) | <0.001 |
Diabetes mellitus—n (%) | 1813 | 34/493 (6.9) | 133/1320 (10.1) | <0.001 |
Pacemaker—n (%) | 1792 | 24/493 (5.0) | 86/1299 (6.6) | 0.291 |
Medication—n (%) | ||||
Diuretic | 1790 | 195/487 (40.0) | 670/1303 (51.4) | <0.001 |
Digoxin | 1795 | 22/487 (4.5) | 104/1308 (8.0) | <0.001 |
Amiodarone | 1749 | 87/464 (18.8) | 324/1283 (25.3) | <0.001 |
Beta-blocker | 1791 | 41/486 (8.4) | 91/1305 (7.0) | 0.210 |
ACE inhibitor | 1790 | 128/487 (26.3) | 381/1303 (29.2) | 0.150 |
ARA | 1791 | 125/487 (25.7) | 373/1304 (28.6) | 0.153 |
ECG typical ChD—n (%) | 1779 | 233/491 (47.3) | 773/1288 (60.0) | <0.001 |
High age-adjusted NT-ProBNP n (%) | 1810 | 30/492 (6.1) | 175/1318 (13.3) | <0.001 |
PCR positivity—n (%) | 1813 | 82/493 (16.6) | 481/1320 (36.4) | <0.001 |
Mortality in 2-year follow-up | 1813 | 14/493 (2.8) | 100/1320 (7.6) | <0.001 |
ACE = Angiotensin converting enzyme; ARA = angiotensin receptor antagonists; ECG = Electrocardiogram; NT-ProBNP = N-terminal of the prohormone brain natriuretic peptide; PCR = Polymerase-chain-reaction.
The mean age was higher for the CG (CG: 59.1 years versus TrG: 54.1 years, P<0.01). Most of the patients were female and the majority in both groups self-reported having been diagnosed with ChD for more than 10 years (TrG: 73.3% versus CG: 58.4%) before being interviewed. In general, patients in the CG presented with worse clinical conditions, including hypertension, diabetes mellitus, ECG typical of Chagas disease, NT-ProBNP levels suggestive of heart failure and PCR positivity, as well as higher use of medications other than benznidazole (P<0.01), except Beta-blockers, ACE and ARA.
Adverse outcomes in the TrG unmatched-control (unadjusted or adjusted by linear or logistic regression), and matched-control (genetic matching) analysis are presented in
Outcomes | Treatment |
Control |
Univariate | Multivariate |
Genetic Matching |
---|---|---|---|---|---|
OR |
OR (95% CI) | OR (95% CI) | |||
Mortality 2-year follow-up | 2.8 (14/493) | 7.6 (100/1320) | 0.36 (0.19, 0.63) | 0.42 (0.23, 0.76) | 0.37 (0.21, 0.63) |
High age-adjusted NT-ProBNP | 6.1 (30/492) | 13.3 (175/1318) | 0.42 (0.27, 0.64) | 0.41 (0.27, 0.63) | 0.41 (0.28, 0.60) |
ECG abnormalities typical ChD | 48.4 (233/481) | 60.0 (773/1288) | 0.63 (0.50, 0.78) | 0.67 (0.53, 0.84) | 0.64 (0.52, 0.79) |
ECG abnormalities typical ChD + |
5.0 (24/480) | 12.6 (162/1286) | 0.37 (0.22, 0.57) | 0.36 (0.23, 0.57) | 0.35 (0.23, 0.53) |
PCR positivity for |
16.6 (82/493) | 36.4 (481/1320) | 0.35 (0.26, 0.45) | 0.35 (0.26, 0.46) | 0.35 (0.27, 0.45) |
NT-ProBNP = N-terminal of the prohormone brain natriuretic peptide; ECG = Electrocardiogram; ChD = Chagas disease; PCR = Polymerase-chain-reaction.
* Adjusted by age, gender, literate, duration of ChD, family monthly income and hypertension.
†Odds ratio
The age-adjusted NT-proBNP levels were lower in the TrG (6.1%, 30/492) than in CG (13.3%, 175/1,318), indicating an adjusted odds ratio was 0.41 (95%CI: 0.28; 0.60) in favor of the TrG. The TrG had a lower frequency of patients with NT-proBNP levels, above age-adjusted reference values (
In this investigation, when comparing patients previously treated with benznidazole with similar controls, we observed a lower frequency of markers of ChD cardiomyopathy severity a reduction in parasitemia and a significantly lower risk of mortality over a 2 years of follow-up period. Our findings help clarify the controversy about the effectiveness of benznidazole in the treatment of ChD, especially in the early chronic phase of the disease. Although there is an urgent need for better therapies than can stop or delay the progression of ChD in millions of infected persons around the world, [
A significant reduction in parasitemia, as measured by PCR, in patients treated with benznidazole has been reported in previous studies [
The most important contribution of the present work is our demonstration of a marked clinical benefit from benznidazole, including a reduction in well-established markers of ChD severity, such as typical ECG abnormalities, high NT-proBNP levels or both, as well as mortality during the two years of follow-up after recruitment. The presence of typical or major ECG abnormalities is an important marker of risk in Chagas disease [
Our results are further reinforced by the effect of previous treatment on NT-proBNP levels, as those who were treated with benznidazole had a significantly lower frequency of high age-adjusted levels than those who were untreated, both when NT-proBNP levels were considered as a single outcome or in association with presence of typical ECG findings. Indeed, high levels of BNP or NT-proBNP are an accurate marker of the presence of left ventricular systolic dysfunction [
The SaMi-Trop study is the first to show that previous treatment with benznidazole is associated with a reduced risk of death in ChD. Viotti et. al observed a non-significant reduction in the rate of death in the treated group [
The present study had limitations. We do not have reliable information about the dose of trypanocidal drug or about the exact duration of treatment, as most patients used the drug more than 10 years before the study interview. In this manuscript as the duration of treatment was self-reported, we believe in memory bias, considering the long time between exposure and interview to collect precision information. However, because benznidazole usage is a very specific event in the life of a ChD patient, we believe that most patients accurately recalled the prior use of this medication as Yes/No. Moreover, we cannot assure that patients who received benznidazole had the same clinical profiles as those who did not receive treatment. If those treated had an average, milder disease than those who were untreated, our results could have been biased in favor of treatment effect. This limitation is inherent to the study design, although it is likely that, more than 10 years ago, when the treatment decisions were made, most patients would have had milder forms of the disease. Additionally, those with severe cardiopathy (at the time of the decision of treatment or not) probably did not survived the recruitment period, considering the ominous prognosis of severe Chagas cardiomyopathy. The imbalance in various features among the groups, typical of all observational studies, was at least partly corrected by the method of matching used in this study, as confirmed by the multivariate analysis.
In Brazil, the transmission of ChD by
Among the strengths of the study were the large sample of ChD patients living in an endemic area, in contrast to other studies that had smaller numbers of patients from specialized clinics in large urban centers. Outcomes were measured in a reliable manner that effectively reflects the clinical conditions of all patients, as well as their parasitological status. The Genetic Matching applied in the statistical analysis is used as an approximation to traditional clinical trial for inference of causality, and its results were confirmed by traditional regression methods. Problems of regional heterogeneity of
In conclusion, if used in the early phases of the disease, benznidazole treatment may result in better clinical and parasitological outcomes. Because there are millions of untreated ChD patients in the world and no new treatments are available for the foreseeable future, it is reasonable to consider treating all Chagas disease patients without advanced cardiopathy with benznidazole, especially those who are less than 50 years of age.
(PDF)