- Top of page
- Material and methods
Chagas disease (ChD) or American trypanosomiasis is caused by the flagellate protozoan Trypanosoma cruzi. ChD, including vectors and reservoirs, extends from the southern United States to Argentina and Chile, affecting 21 countries [World Health Organization (WHO) 2002; Organización Panamericana de la Salud (OPAS) 2006]. Migration and urbanisation in Latin America have led to cases of the disease in non-endemic areas (Schmunis 2007).
In the Amazon region, ChD is an important emerging anthropozoonosis, with hundreds of cases reported in recent decades. Deforestation, substandard rural housing and harvesting in forests have increased human contact with peridomestic vectors and mammal reservoirs, increasing the cases of alleged sylvatic transmission of ChD (Brasil 2009). The Brazilian Ministry of Health reported that >90% of cases of acute ChD occurred in the Amazon, 7.9% of them in the state of Amazonas (Brasil 2012). In this region, ‘discrete typing units’ (DTUs) TcI and TcIV of T. cruzi predominate. TcI exists in a natural cycle involving vectors of the genus Rhodnius and wild and synanthropic didelphids (Marcili et al. 2009). TcI and TcIV are associated with acute human ChD cases in outbreaks triggered by oral transmission (Marcili et al. 2009; Valente et al. 2009; Monteiro et al. 2012).
Benznidazole (BZ) is the only drug available for specific treatment of ChD in Brazil (Anonymous 2005). However, natural resistance to BZ is common, and its effectiveness varies in different geographical areas (Andrade et al. 1985; Filardi & Brener 1987; Yun et al. 2009). Several studies have investigated the biological and clinical properties of T. cruzi, including the response to chemotherapeutic agents in mice inoculated with different strains (Andrade et al. 1985; Filardi & Brener 1987; Revollo et al. 1998; Toledo et al. 2003, 2004). Brener et al. (1976) first observed the heterogeneity of responses of T. cruzi strains to specific treatment. Andrade et al. (1985) demonstrated an association between resistance to treatment with BZ and nifurtimox (NFX) and the biological behaviour of the parasite. Filardi and Brener (1987) found that the susceptibility in vivo to these drugs of strains from different hosts and geographical origins varied widely, from 0% to 100% cure. Filardi and Brener (1987) and Toledo et al. (1997) reported natural resistance to drugs in strains isolated from both wild and domestic cycles.
Although these studies have made significant contributions, none used a representative set of T. cruzi strains from emerging ChD areas such as the Brazilian Amazon. Few studies have explored the genetic and biological framework of strains of T. cruzi from the western Brazilian Amazon, where ChD causes less morbidity and mortality than in the classic endemic areas, appearing mainly in the chronic latent form. The T. cruzi DTUs in this region that infect humans differ from T. cruzi DTUs in other regions of Brazil. Acute cases have been reported in the Amazon region since 1969 (Shaw et al. 1969), but to our knowledge, no data exist on in vivo susceptibility to chemotherapeutic agents for the strains circulating in this region. This study evaluated the BZ susceptibility of natural populations of T. cruzi from the state of Amazonas, belonging to DTUs TcI and TcIV, comparatively with strains TcI and TcII from the states of Paraná and Minas Gerais, traditional endemic areas.
- Top of page
- Material and methods
The clonal evolution model postulated for T. cruzi (Tibayrenc & Ayala 1988) predicts a correlation between the phylogenetic divergence of parasite clonal genotypes and their biological and medical properties. Many researchers have studied the association between T. cruzi subspecific genetic diversity and the parasite's biological properties, including susceptibility in vitro and in vivo to various chemotherapeutic agents such as BZ, NFX and itraconazole (ITZ) (Andrade & Magalhães 1997; Revollo et al. 1998; Toledo et al. 2003, 2004). In general, different DTUs exhibit statistically different, although somewhat overlapping, biological properties.
Acute ChD caused by strains belonging to the DTU TcIV (formerly TcIIa; Brisse et al. 2000) has been described by Marcili et al. (2009). Monteiro et al. (2012) found that this DTU is the major aetiologic agent responsible for acute cases of the disease in the western Brazilian Amazon, including outbreaks of oral transmission. To our knowledge, ours is the first study focusing on the susceptibility of strains belonging to DTU TcIV derived from the Brazilian Amazon, to the specific chemotherapeutic agent BZ.
This in vivo study of the BZ susceptibility of strains of T. cruzi obtained from different sources in Amazonas and Paraná and belonging to DTUs TcI, TcII and TcIV showed a wide variation in cure rates, ranging from 27.3% to 100%. This result confirms previous studies that found rates varying from 0% to 100% in the in vivo susceptibility to BZ and NFX for strains of T. cruzi originating from different Brazilian states and Latin American countries (Andrade et al. 1985; Filardi & Brener 1987). Similar variation in susceptibility to BZ was observed among strains from different transmission cycles (Toledo et al. 1997) and clones of parasites belonging to different genotypes (Toledo et al. 2003).
In this study, we found that some Amazon strains were resistant to the specific chemotherapy, in agreement with Andrade et al. (1985) and Filardi and Brener (1987) who observed strains with natural resistance to the drug, even among people without previous exposure.
A correlation between susceptibility to chemotherapeutic agents and phenotypic and genotypic features was also described, respectively, by Andrade et al. (1985) and Toledo et al. (2003). However, the current study showed no clear association between susceptibility to BZ and the DTU of the strains, with no predominance of strains with a particular pattern of BZ resistance within the DTUs, despite the apparently greater BZ susceptibility of TcI strains from Amazonas than TcI strains from Paraná.
Unlike previous research, the use of PCR, a more sensitive technique for monitoring the cure of treated mice (Miyamoto et al. 2006, 2008), probably improved the efficiency in detecting therapeutic failure. This methodological difference may explain the divergent results concerning the correlation between genetic diversity of T. cruzi and susceptibility to BZ.
Evidence is increasing that T. cruzi DTUs are highly diverse, and this heterogeneity may be highly informative in epidemiological terms (Llewellyn et al. 2009). Although DTU TcI is the major cause of resurgent human disease in northern South America, it also occurs in sylvatic triatomine vectors and mammalian reservoirs throughout the continent. The great genetic variability of TcI isolates from different geographical regions in Colombia was sufficient to propose the existence of at least four haplotypes associated with distinct transmission cycles of the parasite (Herrera et al. 2007; Falla et al. 2009). The six TcI strains of this study were classified as R (1), SI (3) and S (2) to BZ, and mice inoculated with this DTU showed an overall cure rate of 62.5%, an intermediate level of resistance. These data are at variance with other studies that associated TcI with greater resistance to BZ (Andrade et al. 1985; Revollo et al.1998; Toledo et al. 2003). However, sylvatic TcI populations are extraordinarily genetically diverse and several genotypes exist within TcI (Llewellyn et al. 2009; Cura et al. 2010). Likewise, intra-DTU genetic variability was demonstrated for TcIV strains from Amazonas with those originating from SIRN Municipality containing several haplotypes (Monteiro et al. 2012). These genotypes/haplotypes may differ in susceptibility to drugs. Studying two TcI genotypes (genotypes 19 and 20; Tibayrenc & Ayala 1988), Toledo et al. (2003) found that all clones belonging to genotype 20 were resistant to ITZ and BZ, while genotype 19 clones varied in susceptibility to BZ.
The survival curves for TcI (Figure 4) showed that the most strains belonging to this DTU never developed patent parasitaemia, allowing us to conclude that FBE is an inadequate parameter for cure monitoring for Amazonian TcI strains and explaining the absence of a predominant susceptibility pattern.
Murta et al. (1998) also found no correlation between the response to specific chemotherapy with BZ and NFX and the genetics of the parasite for the DTUs TcI and TcII (formerly Z1 and Z2, respectively; Miles et al. 1977). However, the association between sensitivity to drugs and DTU TcVI (formerly zymodeme B; Romanha et al. 1979) was confirmed, and all TcVI strains evaluated to date are sensitive to treatment with BZ and NFX.
The different levels of susceptibility to BZ for strains from Paraná concord with previous findings in study that used a larger number of strains (Toledo et al. 1997). Two strains in the current study, PR2259 (TcII) and PR150 (TcI), were previously studied, and the BZ susceptibility of PR2259, re-isolated 10 years later from the same patient, was confirmed. However, PR150 was previously considered resistant (0% cure), but was assessed as IS (cure rate of 57.1%) in this study. In addition to its increased susceptibility to BZ, PR150 changed its biological behaviour in mice, displaying lower infectivity and parasitaemia from subpatent to patent, suggesting population selection due to manipulation of a polyclonal strain. Parasite subpopulations with different genotypes may be present in low percentages in the T. cruzi strain and could be selected after a long-term vertebrate host–T. cruzi interaction (Veloso et al. 2005), by successive blood passages or maintenance in acellular culture, influencing their susceptibility to BZ (Veloso et al. 2001; Caldas et al. 2008). The genetic heterogeneity of the DTU TcI suggests that the existence of correlations between genetic and biological properties should be investigated at the sub-DTU level (Revollo et al. 1998; Toledo et al. 2002, 2003; Cura et al. 2010).
In this study, there were no significant differences in cure rates when strains were grouped according to DTU, geographical origin or host, except that the TcI strains from Amazonas showed higher cure rates than TcI strains from Paraná. Specific treatment with BZ triggered a significant reduction in the parasitological, molecular or serological parameters for mice inoculated with all DTUs. However, a more significant reduction was recorded for TcIV, indicating that the treatment can be more or less beneficial for the host depending on the DTU. Parasitemia levels were up to 10 times lower in mice inoculated with T. cruzi strains from Amazonas in comparison with strains from Paraná (Reis et al. 2012), which may have affected these results, because the drug dose was the same for all animals. TcI and TcIV strains are prevalent in the Amazon, and these further findings in mice justify the use of BZ in the treatment of ChD patients in this region.
It is difficult to correlate the results for mice inoculated with strains from the Amazon region and patient outcomes, due to the scarcity of published data on the effectiveness of ChD treatment in this region. Although there is little evidence on the effect of specific treatment in Amazonian patients (Pinto et al. 2009, 2010; Valente et al. 2009), the literature suggests a correlation between treatment response in patients, and results of experimental chemotherapy in mice infected with the same strains (Filardi & Brener 1987; Andrade et al. 1992; Toledo et al. 2004).
In this study, TcIV strains isolated from acute Chagasic patients and from Rhodnius robustus in Amazonas varied in susceptibility to BZ in vivo, comprising sensitive, partially sensitive and resistant strains. Resistance to BZ was observed among T. cruzi strains from Amazonas, which were natural parasite populations without previous exposure to this drug. The cure rate of about 60%, the apparent greater sensitivity to BZ of TcI strains from Amazonas in relation to TcI strains from Paraná and the significant reductions in parameters of mice inoculated with Amazonian strains undergoing treatment with BZ indicate that this drug is appropriate to treat human patients from this region. However, more effective treatments for ChD need to be found.
In conclusion, the data presented here do not fit the predictions of the clonal evolution model of T. cruzi. The lack of association between the classification of the strains in DTUs and their in vivo susceptibility to BZ could be explained by the considerably higher genetic variability of T. cruzi strains from the sylvatic than from the domestic environment. Our results also warn of the necessity to take into account the lower phylogenetic subdivisions of this species (genotypes and haplotypes) in studies that seek to evaluate the correlation between genetic diversity and biological properties in T. cruzi.