progressive chronic chagas heart disease ten years after ... - CiteSeerX

0 downloads 0 Views 248KB Size Report
Federal District, is an area free of T. cruzi transmission, in ..... cruzi infections can be transmitted by several species of re- .... The stercorarian trypanosomes.
Am. J. Trop. Med. Hyg., 63(3, 4), 2000, pp. 111–118 Copyright 䉷 2000 by The American Society of Tropical Medicine and Hygiene

PROGRESSIVE CHRONIC CHAGAS HEART DISEASE TEN YEARS AFTER TREATMENT WITH ANTI-TRYPANOSOMA CRUZI NITRODERIVATIVES ˜ ES-BARBOSA, LIANA LAURIA-PIRES, MARIA S. BRAGA, ANA C. VEXENAT, NADJAR NITZ, AUGUSTO SIMO DOUGLAS L. TINOCO, AND ANTONIO R. L. TEIXEIRA Chagas Disease Multidisciplinary Research Laboratory, Faculty of Medicine, University of Brası´lia, Federal District, Brazil

Abstract. A randomized ten-year follow-up study involving 91 Chagas patients and 41 uninfected controls was undertaken to determine the effectiveness of nitroderivative therapy. Anti-Trypanosoma cruzi antibodies were consistently lower one year after treatment than 10 years thereafter (P ⬍ 0.001). The blood of all treated and 93.7% of untreated Chagas patients yielded polymerase chain reaction (PCR) product from probes annealing to T. cruzi nuclear DNA, indicating active infection. Competitive PCR showed means ⫾ standard deviations of 20.1 ⫾ 22.6 T. cruzi/ml of blood from untreated and 13.8 ⫾ 14.9 from treated Chagas patients, but the differences between means were not statistically significant (P ⬎ 0.05). Electrocardiograms recorded a gamut of alterations several-fold more frequent in Chagas patients, regardless of treatment, than in uninfected controls (P ⬍ 0.001). These results show that nitroderivative therapy for T. cruzi infections is unsatisfactory and cannot be recommended since it fails to eradicate the parasite or change the progression of heart disease in chronic Chagas patients. sity of Brasilia Faculty Committee on the Conduct of Human Research approved the research protocol. Over the past ten years the physicians have provided health assistance to street cleaners in the program. Immunological assays. Each individual with an epidemiological history of having lived in huts infested with triatomine bugs, was subjected to immunological evaluation.18 Specific antibodies to the parasite antigens were detected by enzyme-linked immunosorbent assay (ELISA), hemagglutination (IH), and immunofluorescence (IF) tests for Trypanasoma cruzi infection. Chronic infection was disclosed by seropositivity, using the criterion that a patient yielding two out of three positive tests is T. cruzi infected. In this study, each positive test was validated in parallel serological assays in a panel of sera from patients with parasitological confirmation of the infection.18 A skin test against a T. cruzi subcellular fraction T12E antigen was performed in every subject enrolled in this study.19 An intradermal injection of 20 ␮g of protein in 100 ␮L of this antigen resulted in local inflammatory reaction, the diameters of which were registered after 48 hr. The area of induration was calculated according to the formula: a/2 ⫻ b/2 ⫻ ␲ ⫽ cm2. In a previous study it was shown that Chagas patients yielded indurated reactions ⬎ 1.5 cm2, whereas controls yielded erythematous areas ⬍ 1 cm2, at the site of T12E antigen injection.19 Xenodiagnosis. Parasitological demonstration of chronic T. cruzi infections was sought by xenodiagnosis. One diagnostic test consisted of 40 first-instar nymphs of the reduviid bug Dipetalogoster maximus, obtaining blood meals from one patient with positive immunological tests for the infection.20 Feces from each bug were examined twice (30 and 60 days post-exposure) for parasites. Definition of study groups. One hundred and thirty-two street cleaners who volunteered to participate in this study were included in three groups. Among 91 street cleaners with serological evidence of T. cruzi infection, there were 45 chronic Chagas patients that had received specific nitroderivative therapy. In this group, 20 patients (45%) completed full treatment for 60 days, 14 (31%) ingested the drug for at least 30 days, and 11 patients (24%) received the nitroderivative for at least 20 days. We matched by age and gender

INTRODUCTION

Chronic Chagas disease is the main cause of heart insufficiency, arrhythmias, and sudden death in regions of the American continent where Trypanosoma cruzi infections are endemic.1–4 Late reports have shown that sterol biosynthesis inhibitors developed for treatment of fungal diseases may induce parasitological cure in 70–90% of mice infected with T. cruzi, and in 53% of chronic Chagas patients.5–7 Nevertheless, nitroderivative compounds with anti-trypanosomal activity have been used to treat T. cruzi infections in humans.8–14 Benznidazole (N-benzyl-2-nitro-imidazoleacetamide) and nifurtimox (4-[5-nitro-3-methyl-furphurylideneamino]-tetrahydro-4-H-1, 4-thiazine-1-1-dioxide) curtail parasitemias and are indicated for treating acute Chagas disease.9–11 However, these drugs cannot be administered without the supervision of a physician because they provoke undesirable side effects.12,15–17 In addition, their efficacy for treating chronic Chagas disease has yet to be determined. In the past 10 years, we have provided medical care for street cleaners of the Brası´lia Refuse Department. Brası´lia, Federal District, is an area free of T. cruzi transmission, in which street cleaners are not exposed to vectors of Chagas disease. Since this study population had migrated from rural regions endemic for Chagas disease in Brazil at least ten years before the beginning of the program, every serologically positive patient enrolled in this study had the chronic infection. We found that 18.5% of this randomized street cleaner population had at least two out of three positive immunological tests for T. cruzi infection.18 Interestingly, a cohort of these Chagas patients had already been treated with the nifurtimox or benznidazole nitroderivatives one year before they were enrolled in this health assistance program. In this follow-up study, we report clinical and laboratory data showing slow progression of heart disease in treated, as well as untreated Chagas patients. Furthermore, we used a molecular method and detected active T. cruzi infections in both groups of patients, regardless of treatment. MATERIALS AND METHODS

Population. The clinical study was conducted at the Medical Offices of the Brası´lia Refuse Department. The Univer-

111

112

LAURIA-PIRES AND OTHERS

those treated patients with 46 untreated chronic Chagas disease patients. Treated and untreated chronic Chagas disease groups were matched with 41 uninfected control male street cleaners with negative immunological tests for T. cruzi infection. Eighty-nine percent of treated, 91% of untreated Chagas patients, and 92% of uninfected controls were 31 to 60 years of age. Electrocardiographic (ECG) and echocardiographic analyses. A standard 12-lead ECG was obtained for every patient on inclusion in the study and on other occasions, at the physician’s discretion. Single patient multiple ECG alterations were cumulatively counted. Holter-type 24 hr recordings were obtained on a single occasion. Also, two-dimensional doppler echocardiographs (Esatec Echocardiograph, model SIM-7000, Firenzi, Italy) were performed once with standard views recorded on 0.5-inch VHS videotape. Echocardiograms were considered positive if they showed 1) hypokinesis or akinesis of two or more contiguous segments, 2) abnormal wall motion, and 3) global hypokinesis and systolic disfunction.21,22 Each of the above evaluation studies of pathophysiological parameters of heart functions was interpreted by two experienced specialists blinded to the patient’s clinical status and outcome. DNA extraction and polymerase chain reaction (PCR). We drew 10 mL of peripheral blood and collected the buffy coat to extract DNA from each subject in the study groups.23 DNA was analyzed by PCR with specific primers for the constant kDNA minicircle region.23,24 and for highly repetitive sequence of nDNA of T. cruzi.25 The kDNA primer set (S35, 5⬘-ATAATGTACGGG(T/G)GAGATGC-3⬘ and S36, 5⬘-GGTTCGATTGGGGTTGGTG-3⬘) annealing to the constant minicircle region yields a 330 bp product and its catamer 660 bp.23 The nDNA primer set (PON1, 5⬘TGGCTTGGAGGAGTTATTGT-3⬘ and PON2, 5⬘-AGGAGTGACGGTTGATCAGT-3⬘) amplifies a 250 bp fragment.25 The amplification of parasitic template DNA with this nested set of primers would indicate an active T. cruzi infection. A DNA thermal cycler (MJ Research, Watertown, MA) was used for 30–32 cycles as follows: PON1/2, 94⬚C for 2 min, 58⬚C for 1 min, and 72⬚C for 1 min; S35/36, 94⬚C for 2 min, 64⬚C for 1 min, and 72⬚C for 1 min. The reactions were run with 100 ng of Chagas patient target template and with 50 pg of DNA from T. cruzi culture forms. Each reaction was done in 25 ␮L aliquots containing 2.5 U of Taq polymerase (Perkin Elmer, Cetus Norwalk), 0.2 mM dNTPs, 50 mM TRIS-HCl (pH 9.0 at 20⬚C), 1.5 mM MgCl2, 200 mM ammonium sulfate, and 5 ␮M each primer. A 10 uL aliquot from each PCR reaction was subjected to a 1% agarose gel electrophoresis. Negative controls (water and P388D1-IL1 DNA) and positive control (T. cruzi DNA) were always included to detect DNA contamination and ensure that the PCR worked efficiently. Hybridization. The PCR amplification products were transferred by capillarity to a nylon membrane. Prehybridization and hybridization were performed in 6⫻ SSC, 5⫻ Denhardt’s, 0.5% SDS and 100 ␮g/mL salmon DNA solution. The membranes were prehybridized for 4 hr, and were then hybridized for 12 hr with internal oligonucleotide sequences for different amplified fragments. For S35/36 products, the sequence was S67 5⬘-GGTTTTGGGAGGGG(CG)(G/C)-(T/G)TC-3⬘.23 For PON1/2 products, the sequence was

PON3 5⬘-CCGGCCTGTGTCTGCGGC-3⬘.25 These oligonucleotides were radiolabelled with [␥ 32P]-dATP (3000 CimMol) using the polynucleotide kinase method following the manufacturer’s recommendation (Life Technologies). After the membrane was washed once for 5 min with 1⫻ SSC and 0.1% SDS and twice for 30 min each with 0.2⫻ SSC and 0.1% SDS, it was autoradiographed for variable periods of time. Competition assays. Competitive PCR assays were conducted with S35/36 primers set to quantify the number of parasites in the blood samples that had previously yielded amplification products of expected size with T. cruzi kDNA and nDNA probes. These assays consisted of a mixture of an unknown quantity of template DNA (T. cruzi minicircle DNA) with serial dilutions of a known quantity of competitor DNA. This 280 bp competitor DNA fragment, with binding sites for S35/S36 primers, resulted from digestion of a 330 bp kDNA product of PCR.26 The competitor fragment, purified and cloned in pT7 Blue vector (Novogen, Madison, WI), yielded a PCR product distinguishable from the template (280 versus 330 bp). The determination of equivalency points was made by visual comparison in 1.5% wide range agarose/0.5% standard agarose gel. In PCR-positive samples, using the known amount of the competitor DNA in the reaction, the unknown quantity of the template was calculated, under the asssumption that there were 10,000 minicircles/parasite and thus 15 fg of 330 bp template/parasite.27 Statistical analyses. Results are presented as mean ⫾ standard deviation. Differences between ECG alterations were determined by Fisher’s exact test. Categorical data showing ECG alterations recorded in two occasions were compared by Mantel-Haenszel chi-square.28 P ⬍ 0.05 was considered significant, and P ⱕ 0.1 was considered indicative of a trend. The SAS program was used for computer analysis. RESULTS

Sampling analysis and parasitological investigation. Performing this longitudinal study in the street cleaner population who had an epidemiological background of exposure to insect-vectors of the T. cruzi infection was successful because these employees have job stability. Absenteeism was avoided and long-term health assistance either at work or at home was secured. The statistical sampling analysis revealed adequate proportions between groups of Chagas patients, regardless of treatment, and of control subjects. After 10 years, 75.5% of treated, 78% of untreated Chagas patients, and all control subjects continued to participate in this program. Chagas disease patient drop-out was due to retirement with subsequent change of address, or death; 4 (9%) treated and 3 (6.5%) untreated Chagas disease patients died during the 10-year period of clinical observation. Neither retirement nor death was registered among the uninfected control group consisting of 41 street cleaners showing negative immunological tests for the infection. Xenodiagnoses that were performed in every Chagas patient, one year after treatment, demonstrated the presence of blood forms of T. cruzi in 3 (6.6%) treated patients and in 5 (10.9%) untreated Chagas patients. We used serum and

113

TREATMENT OF HUMAN T. CRUZI INFECTIONS

TABLE 1 Validation of clinical diagnosis of chronic Chagas disease with phenotypic and genotypic markers of Trypanosoma cruzi infection* Phenotype†

PCR genotype‡

Chagas patients Treated

Positive Negative

Untreated

Controls

Chagas patients

IF

IH

ELISA

IF

IH

ELISA

IF

IH

ELISA

Treated

Untreated

Controls

44 1

43 2

44 1

44 2

45 1

42 4

0 41

0 41

0 41

32 0

30 2

3 29

* Fisher’s exact test showed that differences in proportions of phenotypic and genotypic markers between Chagas patients, regardless of treatment, and uninfected controls are statistically significant (P ⬍ 0.001). † Trypanosoma cruzi antibodies detected by immunofluorescence (IF), hemagglutination (IH), and enzyme-linked immunosorbent assay (ELISA) in sera of 45 treated and 46 untreated Chagas disease patients. ‡ Polymerase chain reaction (PCR) product amplified with nested set of primers PON1/2 is considered evidence of active infection.

DNA from these eight patients showing consistently positive results to validate immunologic and PCR assays. Phenotype and genotype markers of the T. cruzi infections. There was a significant statistical difference in proportions of positive results from IF, IH, and ELISA tests between Chagas patients and uninfected controls (P ⬍ 0.001). These statistically significant serological differences were not observed between groups of treated and untreated Chagas patients (Table 1). Furthermore, quantitative results of IF and IH tests conducted one year after treatment showed mean serological titers that were not statistically different among treated or untreated Chagas disease patients. Interestingly, IF and IH tests conducted in parallel on sera of Chagas patients, regardless of treatment, showed antibody titers consistently lower one year after treatment than 10 years thereafter, and the differences between these means were statistically significant (P ⬍ 0.001). Moreover, mean antibody titers detected on a single occasion in treated and untreated groups of Chagas patients, either by IF or by IH tests, were not statistically different (data not shown). In this series, the diagnosis of Chagas disease was covalidated in every case by skin testing with the T12E T. cruzi antigen one year after treatment with nitroderivatives. An intradermal injection of 20 ␮g of parasite protein in 100 ␮L of saline in the forearm of Chagas patients resulted in an indurated inflammatory reaction. The mean area of induration among treated Chagas patients (8.05 ⫾ 6.9 cm2) was

FIGURE 1. (A.) Polymerase chain reaction (PCR) amplification of sequences of nDNA from Trypanosoma cruzi with specific PON1/ 2 primers. Lane 1, 100 base pair (bp) molecular weight marker; Lane 2, water (negative control); Lane 3, P388D1-IL1 DNA; Lanes 4 and 5, untreated and treated Chagas patient DNA, respectively; and Lane 6, T. cruzi DNA. (B.) Southern hybridization with an internal fragment of DNA complementary to the PCR amplification product. Note the 250 bp bands in Lanes 4, 5, and 6 (see Methods).

not statistically different from that obtained from untreated Chagas patients (8.13 ⫾ 5.8 cm2). The mean of skin reactions in the control group was 0.72 ⫾ 0.3 cm2. The differences between means of skin reactions elicited by T. cruzi T12E antigen in Chagas patients and among control street cleaners are statistically highly significant (P ⬍ 0.001). In addition, the serological diagnosis of chronic Chagas disease was covalidated by a panel of 94 standard sera from patients with parasitological demonstration of T. cruzi.20 These results were used further to covalidate results of PCR in DNA samples extracted from human blood and tested with T. cruzi nuclear DNA primers PON1/2. Every treated patient and 93.7% of untreated Chagas patients yielded a specific PCR amplification product with primers that annealed to T. cruzi nDNA (Figure 1A). This product hybridized with its complementary internal sequence (Figure 1B), thus confirming nested PCR specificity. Polymerase chain reaction products amplified with T. cruzi nuclear DNA primers yielded positive results in proportions statistically different between Chagas patients and uninfected controls (P ⬍ 0.001, Table 1). Results of this genotype marker of the T. cruzi infection correlate with the results of specific antibodies, and were not statistically different between treated and untreated Chagas patients. Altogether, these findings showed a consistent association of phenotypic and genotypic markers of T. cruzi infections in Chagas disease patients. Quantitation of parasitemia by competitive PCR. Quantification of parasitemia was then performed in samples of DNA from Chagas patients yielding a PCR amplification product of 250 bp with PON1/PON2 primers, and a specific 330 bp product with T. cruzi kDNA primers S35/S36. The competitive PCR assays showed T. cruzi kDNA bands at various points of equivalency with the competitor DNA bands, regardless of whether the template DNA originated from treated or untreated Chagas patients. The competitive PCR quantification of T. cruzi in treated and untreated Chagas patients is shown in Figure 2 and Table 2. Twenty-seven untreated Chagas patients showed bands whose intensity was compared to that of DNA from a known quantity of T. cruzi showing equivalency in a range of ⬍ 0.4 to 75 parasites/mL of blood. The average means of T. cruzi/mL in this group was 20.1 ⫾ 22.6. In addition, 24 treated Chagas patients showed bands whose equivalency ranged from ⬍ 0.5 to 3 ⫻ 104. After excluding the treated patient with a dangerously high parasitemia, the average mean in treated Chagas disease group was 13.8 ⫾ 14.9 parasites/mL of blood. The differ-

114

LAURIA-PIRES AND OTHERS

TABLE 2 Competitive polymerase chain reaction (PCR) quantification of Trypanosoma cruzi in the blood of untreated and nitroderivative treated chronic Chagas disease patients Chagas patients

FIGURE 2. Quantitation of Trypanosoma cruzi DNA in a treated Chagas patient by competitive polymerase chain reaction (PCR). The assays consisted of mixing an unknown quantity of T. cruzi template DNA with serial dilutions of a known quantity of competitor DNA. The competitor annealing to S35/36 primers amplified products distinguishable from the parasite DNA (280 versus 330 base pairs [bp]). Lane 1, molecular weight standard; Lane 2, water (negative control); Lane 3 P388D1-IL1 DNA; Lane 4, T. cruzi DNA equivalent to 1 parasite; Lanes 5 to 8, 100 ng of DNA from a Chagas patient and various quantities of competitor DNA: Lane 5, 0.15 fg; Lane 6, 1.5 fg; Lane 7, 15 fg; and Lane 8, 150 fg. The point of equivalency between the 330 and the 280 bp bands was at 15 fg (Lane 7) of competitor DNA, which is approximately the amount of DNA in 1 parasite. The parasitemia was calculated considering that 10,000 minicircles/parasite contains 15 fg of the 330 bp/parasite.27

ences between means of treated and untreated Chagas patients were not statistically significant (P ⬎ 0.05). The Chagas patient population was further divided into three categories following profiles of parasitemias demonstrated by competitive PCR: 1) ⬍ 2 parasites/mL, low untreated (0.6 ⫾ 0.2) and treated (1.0 ⫾ 0.4); 2) ⬍ 10 parasites/mL, moderate untreated (3.5 ⫾ 1.3) and treated (4.9 ⫾ 2.2); and 3) ⬎ 10 parasites/mL, high untreated (39.4 ⫾ 18) and treated (30.1 ⫾ 10). The categorical statistical analysis showed that differences among means were not significant (Table 2). Clinical study. Twelve-lead ECG recordings were obtained from chronic Chagas disease patients one and 10 years after anti-T. cruzi chemotherapy, from untreated Chagas patients, and from uninfected control subjects on the same occasions. Ventricular premature beats, bundle branch blocks, intraventricular conduction disturbances, and changes in ventricular repolarization were recorded several-fold more frequently in groups of Chagas patients, whether or not they had been treated, than among uninfected control subjects (Figure 3). It should be stressed that the differences among these proportions were statistically significant (P ⬍ 0.001). Furthermore, ECG recordings made 10 years later revealed the same gamut of alterations in higher proportions in Chagas patients than in uninfected controls, and those differences in proportions were also statistically highly significant (P ⬍ 0.001). Interestingly, the Mantel-Haenszel statistical analysis of ECG alterations, recorded one and ten years thereafter, showed highly significant differences in treated (P ⬍ 0.001) as well as in untreated Chagas patients (P ⬍ 0.025) in relationship to uninfected controls. The cumulative recordings of ECG alterations are summarized in Table 3. One year after treatment, ECG recordings showed alterations in 79% (34 of 43) of treated and 67.4% (31 of 46) of untreated Chagas patients. Ten years thereafter the ECGs recorded showed such alterations in 80.5% (29 of 36) of treated Chagas patients, and in 75% (27 of 36) of untreated patients. In contrast, ECG alterations were recorded in 24.3% (10 of 41) and in 22.5% (9 of 40) of uninfected control subjects, respectively, one and ten years thereafter. The frequencies of ECG alterations in treated and untreated Chagas patients were not statistically dis-

T. cruzi/mL*

Untreated

Treated

Low Moderate High

0.6 ⫾ 0.2 (n ⫽ 6) 3.5 ⫾ 1.3 (n ⫽ 7) 39.4 ⫾ 18 (n ⫽ 14)

1.0 ⫾ 0.4 (n ⫽ 6) 4.9 ⫾ 2.2 (n ⫽ 7) 30.1 ⫾ 10 (n ⫽ 10)

* Parasitemias detected by competitive PCR were considered Low (⬍2 parasites/mL), Moderate (⬍10 parasites/mL), and High (⬎10 parasites/mL). The assays consisted of mixing an unknown quantity of T. cruzi template DNA with serial dilutions of a known quantity of competitor DNA, as described in Methods. Untreated versus Treated, P ⬎ 0.05 for Low, Moderate, and High parasitemias. One treated patient showing very high parasitemia (3 ⫻ 104 T. cruzi/mL) was not included in the statistical analysis.

tinct (P ⬎ 0.05). Moreover, those frequencies were subjected to general association test and revealed lack of statistical significance (P ⫽ 0.19). However, the frequency of ECG alterations between both groups of Chagas patients and in the uninfected control group was statistically highly significant (P ⬍ 0.001), regardless of whether the ECG recordings were taken one or ten years apart. Furthermore, an increasing frequency of ECG alterations was recorded in untreated Chagas patients, ranging from 67.4% to 75% over a ten-year period, not being statistically different (P ⫽ 0.5) from those registered among treated Chagas patients (79% to 80.5%), and among uninfected control subjects (24.3% to 22.5%), over the same time span. In this study, Holter-type 24 hr ECG and echocardiographic recordings were obtained on a single occasion, ten years after treatment. The Holter recordings revealed ventricular arrhythmias in 69.6% (23 of 33) of treated Chagas patients, in 58.3% (21 of 36) of untreated Chagas patients, and in 23.5% (8 of 34) of uninfected control subjects (Table 3). The differences between proportions observed in treated and untreated Chagas patients are not statistically significant (P ⬎ 0.5). However, the differences in proportions between the former groups and the control group are highly significant (P ⬍ 0.001). The echocardiogram revealed alterations in proportions that were not statistically different between Chagas patients and control subjects.

FIGURE 3. Progressive electrocardiographic alterations recorded on two occasions, ten years apart, in nitroderivative treated and untreated Chagas patients. VRC ⫽ ventricular premature contractions; BRB ⫽ bundle branch block; CBB ⫽ combined branch block; ICD ⫽ intraventricular conduction disturbance; VRC ⫽ ventricular repolarization change; NIL ⫽ absence of alteration. Note that bars representative of electrocardiogram (ECG) alterations, recorded on the second occasion, shifted to the right of the scale.

115

TREATMENT OF HUMAN T. CRUZI INFECTIONS

TABLE 3 Frequency of electrocardiogram (ECG), 24 hr Holter monitoring, and echocardiographic (Echo) alterations in treated and untreated Chagas disease patients and in uninfected control subjects over a ten-year period 1998 Recordings

1988 ECG

ECG

Holter

Echo

Chagas treated Chagas untreated Uninfected controls

34/43 (79.0%) 31/46 (67.4%) 10/41 (24.3%)

29/36 (80.5%) 27/36 (75.0%) 9/40 (22.5%)

23/33 (69.6%) 21/36 (58.3%) 8/34 (23.5%)

9/29 (31.0%) 10/35 (28.5%) 7/32 (21.8%)

* Number patients with alterations/number tested. Alterations of ECG, Holter, and Echocardiography were counted as indicated in the text.

DISCUSSION

Records in the files of Chagas disease patients in the study groups have shown that nitrofuran nifurtimox (10 mg/kg/ day) was prescribed for 62% (28 of 45) of Chagas patients. Nitroimidazole benznidazole (10 mg/kg/day) was prescribed to treat 38% (17 of 45) of the remaining Chagas patients. The files registered various side effects such as headache, anorexia, gastric discomfort, visual and mental disturbances, loss of libido, dermatitis, peripheral neuropathy, seizures, and weight loss. Probably as a consequence of the severe side effects, 35.7% of those patients took the dose of nifurtimox for less than 30 days. Benznidazole produced less frequent side effects and 70% of the patients took the drug dose for 60 days. A clinical and laboratory follow-up study of 45 Chagas patients subjected to treatment with trypanocidal nifurtimox or benznidazole nitroderivatives over a period of ten years, showed that specific chemotherapy for chronic T. cruzi infections should not be implemented. Certainly, it may be argued that the efficacy of the drug was underestimated because chemotherapy was not completed in some Chagas patients to whom nifurtimox was prescribed. However, Chagas patients who completed full treatment with benznidazole failed to achieve parasitological cure, as shown by consistently positive results of serology and PCR assays. Chagas patients were treated with trypanocidal nitroarenes because physicians thought that treatment of T. cruzi infections would prevent progression of severe ECG alterations they already presented, considered to represent a high risk of heart trouble and death. However, results after 10 years of clinical and laboratory observations showed that no benefit was achieved. Phenotypic and genotypic markers of T. cruzi strongly suggested that the chronic infections persisted after treatment, and progressive ECG alterations were recorded in treated, as well as untreated Chagas patients, the proportions not being statistically different. The controversial findings reported in the literature concerning the efficacy of Chagas disease treatment with either nifurtimox or benznidazole may be considered to be a consequence of a lack of systematic randomization and an adequate methodological approach for unravelling critical features of the chronic infections such as:9–14 1) use of a low sensitivity xenodiagnostic method to detect a low level of parasitemia in a chronic patient; 2) requirement of follow-up study for time periods longer than those during which the patients were under clinical and laboratory observations; and 3) lack of improvement of a single clinical finding which could not associate a clear benefit of chemotherapy. In this study, detection of persistent T. cruzi infections in

treated Chagas patients by highly sensitive PCR amplification of parasite DNA, which indicates persistence of the infection in every nitroderivative-treated patient, casts doubt on the efficacy of specific chemotherapy. Using this molecular method, we revealed levels of parasitemias that were not statistically different among treated and untreated Chagas patients. This finding and slow progressive clinical alterations observed in both groups of patients discussed below explain why treatment of chronic infection appeared to have no benefit. We conducted PCR with T. cruzi nuclear DNA specific set of primers to monitor drug efficacy in treatment of chronic infections because previous studies showed that sequences of kDNA minicircle of T. cruzi may integrate in the DNA of the host cell.29 In this study, PCR amplification products were obtained with nuclear DNA primers PON1/PON2 and template DNA derived from 100% of treated and from 93.7% of untreated Chagas patients, thus showing persistence of the infection, regardless of whether the patient was treated. Since the parasite nuclear DNA does not integrate in the host DNA, the presence of PCR amplification product with T. cruzi nuclear DNA PON1/PON2 was considered evidence of a living infection.29 Furthermore, the evidence of ongoing infections is in keeping with immunological test results, which were indistinguishable in treated and untreated groups of patients. Anti-T. cruzi antibodies that were present in treated as well as untreated Chagas patients showed consistently increasing IF and IH titers, probably a consequence of chronicity of the infections. Further, skin testing with a parasite subcellular antigen showed a delayed-type indurated reaction in every Chagas patient, regardless of whether they were treated. These findings show that humoral and cell-mediated immune responses cannot be distinguished, either qualitatively or quantitatively, in either treated or untreated Chagas patients. However, an acquired cell-mediated immunosuppression against the parasite antigen was described in the course of administration of benznidazole, and nitroderivative-related compounds are immunosuppressive drugs used to prevent homograft rejection.30,31 Chronic Chagas heart disease is a leading cause of arrhythmias and sudden death. ECG recordings showing ventricular premature contractions, single or combined bundle branch blocks, intraventricular conducting disturbances, and changes of ventricular repolarization were present in significantly higher proportions of Chagas patients, regardless of treatment, than in uninfected control subjects.1,2 ECG alterations were also recorded 10 years later in significantly higher proportions among Chagas patients than among age-

116

LAURIA-PIRES AND OTHERS

matched controls. A tendency to progression of ECG lesions was determined by statistical analysis showing a variable increment of those alterations recorded in a time-lapse period of 10 years. Furthermore, increment of those ECG alterations was statistically validated for both groups of treated and untreated Chagas disease patients.28 Chagas disease patients that received nitroderivatives demonstrated no improvement in any clinical parameter ten years after treatment. The ECG recordings showed more frequent alterations in the group of treated Chagas patients than in the uninfected control group. Those alterations showed progressive features; while their proportions in treated and untreated Chagas patients remained unchanged 10 years after treatment. Also, Holter-type 24 hr ECG confirmed those alterations in proportions that were significantly different between Chagas patients and uninfected controls, and lack of differences in proportions between Chagas patients, regardless of treatment.32 Furthermore, four treated and three untreated patients died over a period of 10 years of clinical observations. Effective chemotherapy of T. cruzi infections is needed because of the enormous burden chronic infections impose on human populations living in endemic areas where active insect-vector transmission of parasites occurs. Trypanosoma cruzi infections can be transmitted by several species of reduviid bugs, but those species of Triatominae with peridomestic habits are mainly responsible for the endemicity of Chagas disease.3 In this study, we observed that over onethird of the chronic Chagas disease population had high parasitemias (average 30.1 ⫾ 10 treated and 39.4 ⫾ 18, untreated); one Chagas patient presented with approximately 3 ⫻ 104 T. cruzi/mL of blood, demonstrated by molecular method. This patient can be considered a reservoir facilitating peridomestic insect-vector dissemination of the infection in endemic areas. This human case suggests that an efficient anti-trypanosomal agent that curtails the parasitemia would be of great assistance in controlling this endemic disease. Immune responses for this patient would ensure diagnosis and preclude dissemination of infection by blood donation.33,34 Furthermore, an effective chemotherapy for the infection is also highly desirable as it would help determine the role of the parasite in the pathogenesis of multiple pathological manifestations of Chagas disease.35–40 The administration of nitroderivatives in this series of Chagas patients produced severe side effects, such as peripheral neuropathy, encephalopathy, and seizures that precluded some patients taking the drug for 60 days. Under experimental conditions, apathy, ataxia, spastic tetraplegia with hyperreflexia of stretching reflexes, balance disorders, asymmetrical gait, and electroencephalographic alterations were described in mongrel dogs treated with doses of benznidazole ranging from 5 to 40 mg/kg/day.16,17 Histopathologically, those alterations correlated with various types of lesions affecting the meninges, cerebral cortex, hemisphere white matter and subcortical gray matter, brain stem, cerebellum, and the spinal cord.17 Determining long-term toxicity of nitroderivative chemotherapy requires monitoring for several years. Its chronic toxicity in terms of tumor growths should be measured epidemiologically, because nifurtimox and of benznidazole administered to rabbits resulted in lymphoma growth in 36.5%

of treated animals; untreated controls showed no growth.15,41–43 The potency of the drug, rather than the duration of its administration, was associated with lymphoma in female Swiss mice injected with 2-nitroimidazole, 5-nitroimidazole, or with a 5-nitrophuran derivative.43 Furthermore, the survival rate among 16 patients subjected to benznidazole, for treatment of Chagas disease reactivation, was 57% after 2 years, compared to 82.4% among 75 non-Chagas disease heart transplant patients.31 Interestingly, 37.5% of heart-transplanted Chagas patients developed malignant tumors after a mean follow-up of 25.3 ⫾ 2.1 months in contrast to only 2.7% of patients in the untreated control group after 34.6 ⫾ 3.6 months.31 Chemotherapy of chronic T. cruzi infections with nitroderivative compounds should not be recommended because observation results discussed here show: 1) severe side effects precluded patients older than 20 years of age from complying with full treatment prescribed by physicians; 2) nitroderivative therapy may release parasitemia as a consequence of drug-induced immunosuppression and parasite resistance, probably through the selection of highly virulent clones;30,44–46 3) lack of clinical findings showing a benefit of treatment with nitroderivatives; 4) lymphoproliferative tumors, such as those described in Chagas patients undergoing heart transplant and benznidazole therapy, can be produced experimentally in rabbits and mice upon injection of either nifurtimox or benznidazole at the dose used to treat human Chagas disease patients;15,31,41–43 and 5) eradication of the infection appears to be required, in view that decreasing parasitemias did not abrogate humoral and cell-mediated immune responses associated with autoimmunity and pathogenesis in Chagas disease patients, but this result cannot be achieved by administration of nitroderivatives.36–40 Acknowledgments: We thank Dr. Eduardo Freitas da Silva, Department of Statistics of the University of Brası´lia, for the statistical analysis and Dr. Wesley C. Van Voorhis, Division of Allergy and Infectious Diseases, SJ-10, Department of Medicine, University of Washington, for the competitor DNA. Financial support: Support was provided by FINEP-Financiadora de Estudos e Projetos, and from Fundac¸ a˜ o de Amparo a` Pesquisa do Distrito Federal, Brazil. Authors’ addresses: Liana Lauria-Pires, Maria S. Braga, Ana C. Vexenat, Nadjar Nitz, Augusto Simo˜ es-Barbosa, Douglas L. Tinoco, and Antonio R. L. Teixeira, Chagas’ Disease Multidisciplinary Research Laboratory, Faculty of Medicine, University of Brası´lia, Postcode 70.919-970. Brası´lia, DF, Brazil. Reprint requests: Antonio R. L. Teixeira, Chagas Disease Multidisciplinary Research Laboratory, Faculty of Medicine, University of Brası´lia; P.O. Box 04536. Postcode 70.919-970. Brası´lia, Federal District, Brazil. Fax: 55⫹61 273-4645. Tel: 55⫹61 349-4987. Email: [email protected]. REFERENCES

1. Campos de Carvalho AC, Masuda MO, Tanowitz HB, Wittner MW, Goldenberg C, Spray DC, 1994. Conduction defects and arrhythmias in Chagas disease. Possible role of gap junctions and humoral mechanisms. J Cardiov Electrophys 5: 686–698. 2. Maguirre JH, Hoff R, Sherlock I, Guimara˜ es AC, Sleigh AC, Ramos NB, Mott KE, Weller TH, 1987. Cardiac morbidity and mortality due to Chagas disease: prospective electrocardiographic study of a Brazilian community. Circulation 75: 1140–1145.

TREATMENT OF HUMAN T. CRUZI INFECTIONS

3. Teixeira ARL, 1987. The stercorarian trypanosomes. Soulsby EJL, ed. Immune Responses in Parasitic Infections: Immunology, Immunopathology and Immunoprophylaxis. Boca Raton: CRC, 125–145. 4. Moncayo A, 1993. Chagas disease-tropical disease research progress 1991–1992. Eleventh Program Report of the UNDP/ World Bank/WhO Special Program for Research and Training of Tropical Diseases (TDR). World Health Organization, Geneva, 67–75. 5. Urbina JA, 1997. Lipid biosynthesis pathways as chemotherapeutic targets in kinetoplast parasites. Parasitol 114: 591– 599. 6. Urbina JA, Payares G, Molina J, Sanoja C, Liendo A, Lazardi K, Piras MM, Piras R, Perez N, Winckey P, Riley JF, 1996. Cure of short- and long-term experimental Chagas disease using D0870. Science 273: 969–971. 7. Apt W, Aguilera AA, Arribada A, Perez C, Miranda C, Sanchez G, Zulantay I, Corte´ s P, Rodriguez J, Juri D, 1998. Treatment of chronic Chagas disease with itraconazole and allopurinol. Am J Trop Med Hyg 59: 133–138. 8. Castro SL, 1993. The challenge of Chagas disease chemotherapy: an update of drugs assayed against Trypanosoma cruzi. Acta Tropica 53: 83–88. 9. Parada H, Carrasco HA, Anez N, Fuenmayor C, Inglessis I, 1997. Cardiac involvement is a constant finding in acute Chagas disease: a clinical, parasitological and histopathological study. Internatl J Cardiol 60: 49–54. 10. Andrade AL, Zicker F, Oliveira RM, 1996. Randomised trial of efficacy of benznidazole in treatment of early Trypanosoma cruzi infection. Lancet 348: 1407–1413. 11. Estani SS, Segura EL, Ruiz AM, Velasquez E, Porcel BM, Yampotis C, 1998. Efficacy of chemotherapy with benznidazole in children in the indeterminate phase of Chagas disease. Am J Trop Med Hyg 59: 526–529. 12. Coura JR, Abreu LL, Willcox HPF, Petana W, 1997. Estudo comparativo controlado com emprego de benznidazole, nifurtimox e placebo, na forma croˆ nica da doenc¸ a de Chagas, em uma a´ rea endeˆ mica com transmissa˜ o interrompida. 1. Avaliac¸ a˜ o preliminar. Rev Soc Bras Med Trop 30: 139–144. 13. Viotti R, Vigliano C, Armenti H, Segura E, 1994. Treatment of chronic Chagas disease with benznidazole: clinical and serological evolution of patients with long term follow-up. Am Heart Journal 127: 151–162. 14. Ferreira HO, 1990. Tratamento da forma indeterminada da doenc¸ a de Chagas com nifurtimox e benzonidazol. Rev Soc Bras Med Trop 23: 209–211. 15. Teixeira ARL, Cunha-Neto E, Rizzo LV, Silva R, 1990. Trypanocidal nitroarene treatment of experimental Trypanosoma cruzi infections does not prevent progression of chronic-phase heart lesions in rabbits. J Infec Dis 162: 1420. 16. Flores-Vieira CL, Antunes Barreira A, 1997. Experimental benznidazole encephalopathy: I. clinical-neurological alterations. J Neurol Sci 150: 3–11. 17. Flores-Vieira CL, Chimelli L, Franca Fernandes RM, Antunes Barreira A, 1997. Experimental benznidazole encephalopathy: II. Electroencephalographic and morphological alterations. J Neurol Sci 150: 13–25. 18. Tinoco DL, Garcia MP, Lauria-Pires L, Santana JM, Teixeira ARL, 1996. The use of four immunologic assays for diagnosis of Chagas disease in street-sweepers of the Federal District, Brazil. Rev Soc Bras Med Trop 29: 33–40. 19. Teixeira ARL, Figueiredo F, Rezende Filho J, 1995. Delayed hypersensitivity to Trypanosoma cruzi antigens. II. The use of the skin testing with the T12E antigen for diagnosis of Chagas disease. Rev Soc Bras Med Trop 28: 259–265. 20. Teixeira ARL, Figueiredo F, Rezende Fiho J, Macedo V, 1983. Chagas disease: a clinical, parasitological, immunological, and pathological study in rabbits. Am J Trop Med. Hyg 32: 258–272. 21. Kontos MC, Arrowood JA, Jesse RL, Ornato JP, Walter HP, Tatum JL, Nixon JV, 1998. Comparison between 2-dimensional echocardiogram and myocardial perfusion imaging in the emergency department in patients with possible myocardial ischemia. Am Heart Journal 136: 724–733.

117

22. Bestetti RB, Dalbo CMR, Freitas CO, 1994. Non-invasive predictors of mortality for patients with Chagas heart disease: a multiple step wise logistic regression study. Cardiology 84: 261–267. 23. Sturm NR, Degrave W, Morel CM, Simpson L, 1984. Sensitive detection and schizodeme classification of Trypanosoma cruzi by amplification of kinetoplast minicircle DNA sequences: use in the diagnosis of Chagas disease. Mol Biochem Parasitol 33: 205–214. 24. Avila HA, Sigman DS, Coohen LM, Millikan RC, Simpson L, 1991. Polymerase chain reaction amplification of Trypanosoma cruzi kinetoplast minicircle DNA isolated from whole blood lysates: diagnosis of chronic Chagas disease. Mol Biochem Parasitol 48: 211–222. 25. Requena JM, Jimenez-Ruiz A, Soto RM, Lopez MC, Alonso C, 1992. Characterization of a highly repeated interspersed DNA sequence of Trypanosoma cruzi. Its potential use in diagnosis and strain classification. Mol Biochem Parasitol 51: 271–280. 26. Degrave W, Fragoso SP, Britto C, Heurvrrswyn GZ, Kidane AB, Cardoso U, Mueller L, Simpson L, Morel CM, 1988. Peculiar sequence organization of kinetoplast DNA minicircles from Trypanosoma cruzi. Mol Biochem Parasitol 27: 63–70. 27. Centurion-Lara A, Barret L, Van Voorhis WC, 1994. Quantitation of parasitemia by competitive polymerase chain reaction. Amplification of parasite kDNA minicircles during chronic infection with Trypanosoma cruzi. J Infec Dis 170: 1334– 1339. 28. Agresti A, 1996. Introduction to Categorical Data Analysis. New York: John Wiley & Sons, 60–64. 29. Teixeira ARL, Argan˜ araz ER, Freitas Jr LH, Lacava ZGM, Santana JM, Luna H, 1994. Possible integration of Trypanosoma cruzi kDNA minicircles into host cell genome by infection. Mut Res 305: 197–209. 30. Teixeira ARL, Jabur E, Souto Maior IC, Solorzano E, 1983. Suppression of cell mediated immune response during administration of benznidazole. Rev Soc Bras Med Trop 16: 11–22. 31. Bocchi EA, Higuchi ML, Vieira ML, Stolf N, Bellotti G, Fiorelli A, Uip D, Jatene A, Pileggi F, 1998. Higher incidence of malignant neoplasms after heart transplantation for treatment of chronic Chagas heart disease. J Heart Lung Transplant 17: 399–405. 32. Maguirre JH, Ramos NB, Santana OO, Almeida EC, Guimaraes AC, 1981. Comparac¸ a˜ o do eletrocardiograma convencional com o eletrocardiograma dinaˆ mico na avaliac¸ a˜ o das arritmias na doenc¸ a de Chagas. Arq Bras Cardiol 37: 82–83. 33. Kerndt PR, Waskin HA, Kirchhoff LV, Steurer F, Waterman SH, Nelson JM, Gellert GA, Shulman IA, 1991. Prevalence of antibody to Trypanosoma cruzi among blood donors in Los Angeles, California. Transfusion 31: 814–818. 34. Kirchhoff LV, 1992. Perspective on Chagas disease in Latin America and United States. Walker DH, ed. Global Medicine: Current Status and Directions for the Twenty-First Century. New York: Springer Verlag, 219–226. 35. Teixeira ARL, Teixeira G, Maceˆ do V, Prata A, 1978. Trypanosoma cruzi-sensitized T-lymphocyte mediated 51Cr release from human heart cells in Chagas disease. Am J Trop Med Hyg 27: 1097–1107. 36. Cunha-Neto E, Rizzo LV, Teixeira ARL, 1989. Autoimmunity in Chagas disease: specific inhibition of reactivity of CD4⫹ T cells against myosin in mice chronically infected with Trypanosoma cruzi. Infect Immun 57: 2640–2644. 37. Ribeiro dos Santos R, Rossi MA, Laus JL, Santana Silva J, Savino W, Mengel J, 1992. Anti-Cd4 abrogates rejection and reestablishes long-term tolerance to syngeneic newborn hearts grafted in mice chronically infected with Trypanosoma cruzi. J Exp Med 175: 29–39. 38. Cunha-Neto E, Duranti M, Gruber A, Zingales B, Messias I, Stolf N, Belloti G, Patarroyo ME, Pileggi F, Kalil JE, 1995. Autoimmunity in Chagas disease: biological relevance of a cardiac myosin-specific epitope cross-reactive to an immunodominant Trypanosoma cruzi antigen. Proc Natl Acad Sci USA 92: 3541–3545.

118

LAURIA-PIRES AND OTHERS

39. Higuchi ML, Ries MM, Aiello VD, Benvenutti LA, Gutierrez PS, Bellotti G, Pileggi F, 1997. Association of an increase in CD8⫹ T cells with the presence of Trypanosoma cruzi antigens in chronic, human, Chagasic myocarditis. Am J Trop Med Hyg 56: 485–489. 40. Tarleton RL, Zhang L, Downs MO, 1997. Autoimmune rejection of neonatal heart transplants in experimental Chagas disease is a parasite-specific response to infected host tissue. Proc Natl Acad Sci USA 94: 3932–3937. 41. Teixeira ARL, Cordoba JC, Souto-Maior IC, Solorzano E, 1990. Chagas disease: Lymphoma growth in rabbits treated with benznidazole. Am J Trop Med Hyg 43: 146–158. 42. Teixeira ARL, Silva R, Cunha-Neto E, Santana JM, Rizzo LV, 1990. Malignant, non-Hodgkin’s lymphomas in Trypanosoma

43. 44. 45.

46.

cruzi-infected rabbits treated with nitroarenes. J Comp Path 103: 37–48. Teixeira ARL, Calixto MA, Teixeira ML, 1994. Chagas disease: carcinogenic activity of the antitrypanosomal nitroarenes in mice. Mut Res 305: 189–196. Buchner FS, Wilson AJ, White TC, Van Voorhis WC, 1998. Induction of resistance to azole drugs in Trypanosoma cruzi. Antimicrob Agents Chemother 42: 3245–3250. Murta SM, Gazzinelli RT, Brener Z, Romanha AJ, 1998. Molecular characterization of susceptible and naturally resistant strains of Trypanosoma cruzi to benznidazole and nifurtimox. Mol Biochem Parasitol 93: 203–214. Murta SM, Romanha AJ, 1998. In vivo selection of a population of Trypanosoma cruzi and clones resistant to benznidazole. Parasitol 116: 165–171.