Introduction
Chagas’ disease is a neglected tropical infection caused by the protozoan parasite Trypanosoma cruzi (Chagas, Reference Chagas1909). Recent estimates indicate that 8 million people are infected with this parasite worldwide (WHO, 2017). This disease is closely related to poverty and is endemic in South and Central America where it is considered a public health problem with more than 10 000 deaths per year (WHO, 2017). However, due to the intense migration of T. cruzi-infected Latin Americans to Asia, Europe and Oceania, there has been an increase in the number of cases of Chagas’ disease in these non-endemic areas since the early 1990s with successive increases in the number of cases in later years (Schmunis, Reference Schmunis2007).
The natural route of infection of the obligate intracellular parasite T. cruzi occurs when a triatomine insect vector deposits infective metacyclic trypomastigotes with their feces and urine on the host's skin during blood meal (Guimarães-Pinto et al., Reference Guimarães-Pinto, Nascimento, Corrêa-Ferreira, Morro, Freire-de-Lima, Lopes, DosReis and Filardy2018). In addition to humans, T. cruzi infects a wide variety of domestic and wild mammals such as Carnivora, Chiroptera, Didelphidomorphia, Lagomorpha, Perissodactyla, Pilosa, Prieta and Rodentia (Añez et al., Reference Añez, Crisante and Soriano2009; Herrera, Reference Herrera2010), with dogs being the main domestic reservoir (Montenegro et al., Reference Montenegro, Jiménez, Dias and Zeledón2002). In addition to vector insects, transmission of parasites can also occur through non-vector pathways such as blood transfusions (Moraes-Souza and Ferreira-Silva, Reference Moraes-Souza and Ferreira-Silva2011), transplants of infected organs (Márquez et al., Reference Márquez, Crespo, Mir, Pérez-Sáez, Quintana, Barbosa and Pascual2013), vertical transmission (Barrios et al., Reference Barrios, Más, Giachetto, Basjmadjián, Rodríguez, Viera, Baroloco and Sayaguez2015), laboratory accidents (Dias, Reference Dias2006) and by the ingestion of food contaminated with the infective forms (trypomastigotes) of T. cruzi (Shikanai-Yasuda and Carvalho, Reference Shikanai-Yasuda and Carvalho2012; Domingues et al., Reference Domingues, Hardoim, Souza, Cardoso, Mendes, Previtalli-Silva, Abreu-Silva, Pelajo-Machado and Calabrese2015). Vector transmission is mainly mediated by insects of the genus Triatoma, Panstrongylus and Rhodnius (Hemiptera; Reduviidae) (Coura and Viñas, Reference Coura and Viñas2010).
Trypanosoma cruzi is a parasite of high genetic diversity, composed of a set of strains or isolates that circulate between insect vectors and mammalian hosts (Rassi et al., Reference Rassi, Rassi and Marin-Neto2010). Although controversial, this heterogeneity has been associated with the wide variability of clinical manifestations and the different profiles of morbidity and mortality of Chagas’ disease (Macedo et al., Reference Macedo, Machado, Oliveira and Pena2004; Manoel-Caetano and Silva, Reference Manoel-Caetano and Silva2007). Regarding the T. cruzi strains, the most recent classification describe at least six genetic lineages or discrete typing units (DTUs), named TcI to TcVI (Zingales et al., Reference Zingales, Andrade, Briones, Campbell, Chiari, Fernandes, Guhl, Lages-Silva, Macedo, Machado, Miles, Romanha, Sturm, Tibayrenc and Schijman2009; Zingales, Reference Zingales2018). TcI predominates in the wild transmission cycle, is less resistant to antiparasitic reference chemotherapy (benznidazole and nifurtimox), and is associated with the human disease occurring in the northern region of Latin America. TcII predominates in the domestic environment of all South America, presenting a higher resistance to antiparasitic chemotherapy and high pathogenicity (Di Noia et al., Reference Di Noia, Buscaglia, De Marchi, Almeida and Frasch2002; Freitas et al., Reference Freitas, Lages-Silva, Crema, Pena and Macedo2005; Botero et al., Reference Botero, Mejía and Triana2007). This lineage was initially subdivided into five units of discrete typologies characterized as IIa, IIb, IIc, IId and IIe (Brisse et al., Reference Brisse, Dujardin and Tibayrenc2000), but Zingales et al. (Reference Zingales, Andrade, Briones, Campbell, Chiari, Fernandes, Guhl, Lages-Silva, Macedo, Machado, Miles, Romanha, Sturm, Tibayrenc and Schijman2009) propound that TcII is no longer divided into five subgroups but each of those subgroups constitutes an independent DTU (TcII–VI). TcIII predominates in the wild environments of South America, with most cases affecting small mammals such as bats and quatis cases being reported in Brazil, more specifically in the Amazon (Lisboa et al., Reference Lisboa, das Chagas, Herrera and Jansen2009; Rocha et al., Reference Rocha, Roque, de Lima, Cheida, Lemos, de Azevedo, Arrais, Bilac, Herrera, Mourão and Jansen2013), and with only one chronic case found in humans (Abolis et al., Reference Abolis, De Araujo, Toledo, Fernandez and Gomes2011). Recent researches agree that TcI and TcII are two pure lineages and that TcV and TcVI have a hybrid origin with TcII and TcIII, while the evolution of TcIII and TcIV still unclear (Zingales, Reference Zingales2018).
Although the relationship between genotype and parasitic phenotype, tropism and clinical manifestations remain poorly understood (Macedo and Pena, Reference Macedo and Pena1998; Vago et al., Reference Vago, andrade, Leite, Reis, Macedo, Adad, Tostes, Moreira, Filho and Pena2000; Prata, Reference Prata2001), all T. cruzi strains isolated from the natural environment have been shown to infect mammalian hosts (Yeo et al., Reference Yeo, Acosta, Llewellyn, Sánchez, Adamson, Miles, López, González, Patterson, Gaunt, Arias and Miles2005; Herrera, Reference Herrera2010). In vertebrate hosts, T. cruzi establishes a systemic infection and parasitism of multiple organs, especially the heart, intestines and oesophagus (Lana and Tafuri, Reference Lana, Tafuri, Neves, De Melo, Linardi and Vitor2016). Although the neurological changes associated with Chagas’ disease are often neglected, there is evidence that T. cruzi is able to parasite and induce inflammatory lesions in structures of the peripheral nervous system (PNS) (Marin-Neto et al., Reference Marin-Neto, Cunha-Neto, Maciel and Simões2007) and central nervous system (CNS) (Masocha and Kristensson, Reference Masocha and Kristensson2012; Pittella, Reference Pittella2013). The CNS involvement during the acute phase of Chagas’ disease can lead to meningitis, seizures, restlessness, continuous crying, insomnia and transient coma (Sangster and Dobson, Reference Sangster, Dobson and Lee2002; Storino et al., Reference Storino, Jörg and Auger2003). The consequences of chagasic meningoencephalitis that occur at the chronic phase consist of motor and sensory disorders, psychic alterations and cerebellar impairment (Sangster and Dobson, Reference Sangster, Dobson and Lee2002). In addition, electrophysiological changes were determined as a consequence of the deterioration of the cerebral cortical function in individuals with chronic Chagas’ disease (Prost et al., Reference Prost, Morikone, Polo and Bosch2000).
Currently, PNS alterations are better understood, and dysautonomia secondary to ganglia and nerve endings of the sympathetic and parasympathetic autonomic nervous system have been consistently implicated in the pathophysiology of cardiomyopathy and chagasic megasyndromes (Oliveira et al., Reference Oliveira, Ferreira, Lopes, Chiari, Camargos and Martinelli2017). However, tropism, distribution and changes induced by T. cruzi in different structures and organs of the CNS are poorly understood. Considering that the current evidence is flawed because it is based on fragmented data, it is difficult to understand the range of the CNS changes that develop throughout the infection with T. cruzi. Therefore, from a structured and systematized search, we evaluated the preclinical evidence regarding the impact of T. cruzi infection on the CNS. In addition to characterizing the infection models used, we established the relationship between the characteristics of T. cruzi strains and their tropism to the CNS and other tissues and organs susceptible to parasitism as well as the most frequent lesions incurred. Moreover, we have critically evaluated the scientific evidence regarding the methodological quality of the studies included in this systematic review.
Materials and methods
Literature search
A comprehensive bibliographic survey completed on 11/20/2017 at 7:30 PM was conducted in the PubMed/Medline databases (https://www.ncbi.nlm.nih.gov/pubmed) and Scopus (https://www.scopus.com/home.uri). Structured descriptors were used in search filters constructed for three domains: Chagas disease, nervous system and animal model (Table S1). The filters on the PubMed/Medline platform were constructed using a hierarchical distribution of the MESH terms. We used the same PubMed search strategy to search the Scopus platform; however, we used the filter for animal studies provided by the Scopus platform. The non-MeSH descriptors were characterized by the algorithm [TIAB], which was also used to retrieve recently published but non-indexed (in-process) studies. This systematic review was developed according to the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analysis; Moher et al., Reference Moher, Liberati, Tetzlaff and Altman2009), which is used as a guide for selection, screening and eligibility of studies (Fig. 1).
Flow diagram of search results, study screening and eligibility to define the articles to be included in the systematic review according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyzes; www.prisma-statement.org).
Data extraction and management
An independent researcher (E.V.) selected eligible studies following the analysis of their titles and abstracts. When in doubt, an arbitration was requested from other independent reviewers (R.V.G, M.M.S. and R.D.N.) to decide whether any given study met the eligibility criteria previously defined, likewise to discard subjectivity in the data collection and selection process, the information was extracted independently and analysed separately. Data from each study were extracted and tabulated using standard information such as: (i) characteristics of the publication (title, author, year and country where the study was performed); (ii) experimental model (animal species, gender, age, weight and the number of animals and of experimental groups); (iii) infection characteristics (nature of infection, T. cruzi strain, inoculation route, amount of inoculum and the phase of parasitemia); and (iv) morphological and functional outcomes associated with the CNS (diagnostic test, infected tissue and types of changes). Whenever we encountered difficulties in obtaining the full-text papers, we requested the authors by e-mail to provide a copy of the article. Subsequently, the data were compared and the conflicting information identified and corrected after discussion among the researchers.
Eligibility criteria
Only original studies published in English, Portuguese and Spanish that met the following eligibility criteria were selected: (i) studies with mammals infected experimentally or naturally with T. cruzi; (ii) studies with at least one control group infected with T. cruzi that was not submitted to any treatment; (iii) studies using naturally occurring and non-genetically engineered strains; (iv) studies with hosts that were not genetically modified and that did not present alterations resulting from other interventions; (v) studies describing CNS-related morphological and/or physiological outcomes; and (vi) full-text studies. Literature reviews, comments, notes, book chapters as well as non-indexed studies were excluded.
Analysis of methodological bias
Bias analysis was structured according to the characteristics described in the ARRIVE strategy (Kilkenny et al., Reference Kilkenny, Browne, Cuthill, Emerson and Altman2010). To this end, we used criteria based on brief descriptions of the essential characteristics of all studies using animal models, such as the theoretical background, research aim, analytical methods, statistical approach, sample calculations and research outcome. A table summarizes all relevant and applicable aspects considering the specificity and the aims of the systematic review. The individual adherence to the bias criteria and the general mean of adhesion are expressed as absolute values (n) and percentage (%) (Pereira et al., Reference Pereira, Greco, Moreira, Chagas, Caldas, Goncalves and Novaes2017).
Results
Inclusion of studies
Initial research resulted in 1125 studies, but 186 were excluded because they were duplicate studies. After reading the title and abstract, 707 irrelevant studies were excluded. After the remaining 232 articles were read in their entirety, another 200 articles were excluded including studies describing alterations in the PNS (n = 59), clinical studies (n = 33), in vitro (n = 18) and secondary studies (n = 14). Finally, 32 studies fully met the inclusion criteria and were included in the systematic review (Fig. 1).
Analysis of infection models
The 32 studies were conducted in seven different countries: Brazil (40.6%; n = 13), USA (25%; n = 8) and Argentina (12.5%; n = 4). The most used animal models were mice (90.6%; n = 29), horse, pig and guinea pig (3.1%; n = 1 each). The most used mouse lines were C3H (40.6%; n = 13), Swiss (25%; n = 8) and C57BL/6 (18.8%; n = 6). The most used T. cruzi isolates were: Colombian (25%; n = 8), Brazil (15.6%; n = 5), Y, RA and Tulahuén (9.4%; n = 3 each). The most frequent route of inoculation was intraperitoneal (68.8%; n = 22) followed by subcutaneous, intradermal and intravenous (6.3%; n = 2 each). The inoculation route was not reported in four studies (12.5%). Tests to confirm infection were not described in 11 articles (35.4%) (Table 1). Most of the studies evaluated acute infections (62.5%; n = 20). Acute and chronic infections were simultaneously reported in eight studies (25%), while exclusively chronic infections were evaluated in only four studies (12.5%) (Table 2).
Characteristics of the studies evaluating the changes in the central nervous system following infection with T. cruzi
AR, Argentina; BR, Brazil; ES, Spain; GB, United Kingdom; PE, Peru; US, United States; VE, Venezuela; Ms, mouse; Hs, horse; Gp, guinea pig; Pg, pig; ♂, male; ♀, female; ?, uninformed; E, experimental; N, natural; §, strains OPS21, SP104, 13379, Gamba; II, strains P11, ESQUILO, CUICA, P209, SO34; #, strains SO3, NR, BUG2148, BUG2149, MN, SC43; Ip, intraperitoneal; Id, intradermal; Idp, intradermoplantar; Sc, subcutaneous; Iv, intravenous; In, intranasal; FBE, fresh blood examination; HC, haemoculture; MHCT, microhaematocrit centrifugue technique; PCR, polymerase chain reaction; ELISA, Enzyme-Linked ImmunoSorbent Assay; Para, parasitaemia; XD, xenodiagnostic; S, serology.
Changes in CNS tissues or organs during T. cruzi infection
+, Volpato et al., Reference Volpato, Sousa, D’Ávila, Galvão and Chiari2017; Meza et al., Reference Meza, Kaneshima, de Oliveira Silva, Gabriel, de Araújo, Gomes, Monteiro, Barbosa and de Ornelas Toledo2014; Alves et al., Reference Alves, Regasini, Funari, Young, Rimoldi, Bolzani, Silva, Albuquerque and Rosa2012; Minning et al., Reference Minning, Weatherly, Flibotte and Tarleton2011; Andrade et al., Reference Andrade, Galvão, Meirelles, Chiari, Pena and Macedo2010; Zingales et al., Reference Zingales, Andrade, Briones, Campbell, Chiari, Fernandes, Guhl, Lages-Silva, Macedo, Machado, Miles, Romanha, Sturm, Tibayrenc and Schijman2009; ++, Zingales et al., Reference Zingales, Miles, Campbell, Tibayrenc, Macedo, Teixeira, Schijman, Llewellyn, Lages-Silva, Machado, Andrade and Sturm2012; ?, uninformed; ↑, increase; *, (macrophages, CD8+ and CD4+); **, during the chronic phase inflammatory infiltrates were mild or non-existent; ††, macrophages (data not shown) and CD8+ and, to a lesser extent, CD4+ T cells; †, CD8+ predominant respect to CD4 +; ‡, prevalence of macrophages, mononuclear cells and microglia; §, strains OPS21, SP104, 13379, Gamba; II, strains P11, ESQUILO, CUICA, P209, SO34; #, strains SO3, NR, BUG2148, BUG2149, MN, SC43; ¶, composed of lymphocytes, macrophages and occasional polymorphonuclear cells; ***, the greater inflammatory foci, the smaller the decrease in the number of neurons; ACh, acetylcholine; (b), without alterations in the brain; LSSP-PCR, low-stringency single specific primer; PCR, polymerase chain reaction; RT-PCR, reverse transcription polymerase chain reaction; Histpat, histopathological; Imnhisq, immunohistochemistry; DAACh, dilation of cerebral arterioles with acetylcholine; Bαb, binding of α-bungarotoxin; IFAT, immunofluorescence antibody test; TBARS, thiobarbituric acid reactive substances species; EA, enzyme assay; Radlab, radiolabelled; BBB, blood–brain barrier; FN, fibronectina; LN, laminin; BV, blood vessels.
The most frequently used T. cruzi genotypes were: TcI (40.6%; n = 13), TcII (12.5%; n = 4), TcIV (3.1%; n = 1) and TcVI (12.5%; n = 4). Some studies used more than one genotype (18.8%; n = 6); however, four studies (12.5%) did not identify the genotype of the strains. Histopathological analyses were performed in 23 studies (71.9%), six studies used immunohistochemistry (18.8%), six used polymerase chain reaction (18.8%), and three did Western-blot analysis (9.4%). The CNS organs with the largest changes were brain (65.6%, n = 21), followed by the spinal cord (25%; n = 8) and cerebellum (15.6%; n = 5) (Fig. 2).
Schematic representation demonstrating the distribution of morphological changes and tropism of the different strains of T. cruzi in the CNS. *: inflammatory focus; ⚫: presence of T. cruzi; ▲: presence of anti-T. cruzi antibodies; ■: presence of T. cruzi antigens; ★: vasculopathy; +: tissue damage; ♦: oedema; : gliosis; : satellitosis. The predominant strains in each region are presented in square brackets [.
The most frequent lesions in the CNS were the presence of inflammatory foci (68.8%; n = 22), with a predominance of lymphocytic mononuclear infiltrate (15.6%; n = 5). The encephalon presented moderate-to-intense inflammation with a marked perivascular distribution. To a lesser extent, inflammatory foci were found in the meninges (9.4%, n = 3), choroid plexus (9.4%, n = 3) and nuclei at the base (6.3%; n = 2). In the spinal cord, inflammatory foci were found mainly associated with nerve roots (50%, n = 16) and meninges (50%, n = 16) (Table 2; Fig. 2).
The presence of amastigote nests, free trypomastigotes or indeterminate forms of T. cruzi in the CNS was reported in 53.1% of the studies (n = 17). The presence of amastigotes in the cytoplasm of glial cells (astrocytes, microglia, ependymocytes and oligodendrocytes) was observed in the organs or tissues with the highest presence of parasites (68.8%; n = 22). Pseudocysts with intra and extracellular amastigotes were also found in the nuclei of the base (12.5%, n = 4), cerebellum (12.5%, n = 4) and Purkinje cells (12.5%; n = 4). Amastigotes were found in the white matter of the spinal cord, intra and extracellular (9.4%; n = 3).
The presence of anti-T. cruzi antibodies was described in three studies (9.4%) and T. cruzi antigens in four studies (12.5%). Vasculopathies were reported in five studies (15.6%), gliosis in three (9.4%), satellitosis in two (6.3%), while tissue damage due to necrosis and oedema was described in the cerebrum and spinal cord in three studies each (9.4%) (Table 2, Fig. 2). Only 11 studies (34.4%) evaluated the parasitic load on the day the animals were sacrificed, ranging from 0 to 69.3 × 106 trypomastigotes.
The rare reports covering the chronic phase of Chagas’ disease indicated inflammatory foci ranging from light to intense (9.4%; n = 3), presence of T. cruzi nests (6.3%; n = 2), tissue damage as a result of autoimmune lesions (3.1%, n = 1), and neuron degeneration and necrosis (3.1%, n = 1) were the most frequent alterations. The most affected sites were the brain, the blood–brain barrier (BBB) and the spinal cord.
Bias analyses
The results regarding the bias analyses are shown in Table 3. An average of 55.0 ± 12.3 ARRIVE items were met by the original studies. In general, studies performed up to 15 years ago were those that presented the greatest deficiency in the methodological detail and the description of the results (Fig. 3). Only seven articles (21.9%) justified the animal model used. Approval of the ethics committee was reported in 13 studies (40.6%). Only two studies (6.25%) justified the size of the T. cruzi inoculum used. No study justified the route of administration. All studies (n = 32) indicated the animal species and the T. cruzi strain used. The sex, weight and age of the animals were described in 84.4% (n = 27), 25% (n = 8) and 75% (n = 24) of the studies, respectively. Calculation of the sample size was made explicit in only one study (3.1%). The detailed description of the statistical analyses used was reported in 43.8% of the studies (n = 14). Sixteen studies (50%) reported modifications to the experimental protocol by adverse events (Table 3).
Analysis of the methodological bias (quality of the report) for each study included in the systematic review according to the ARRIVE guidelines (www.nc3rs.org.uk/arrive-guidelines). The dotted line indicates the average quality score (%). The detailed bias analysis, stratified by domains and evaluated items, is presented in Table 3.
Bias analysis (ARRIVE) of studies with changes in the central nervous system during infection with T. cruzi
Unmarked cells indicate that the criteria were not filled
Discussion
Using a systematic screening, we observed that most of the studies investigating CNS changes caused by T. cruzi were conducted in developing countries, corroborating the idea that research efforts about this parasite are concentrated in countries where Chagas’ disease is endemic (Antinori et al., Reference Antinori, Galimberti, Bianco, Grande, Galli and Corbellino2017). In addition, the overall methodological quality score for this set of studies was limited. Since the bias analysis presented herein was structured following the basic requirements for the rational acquisition and interpretation of results, the limited quality of the evidence can be attributed to studies with low individual methodological scores (Zoltowski et al., Reference Zoltowski, Costa, Teixeira and Koller2014). These aspects point to an urgent need for more rigorous analysis and interpretation of the evidence considering all the critical elements that may undermine the validity of the studies. Interestingly, our results also showed a temporal influence on the bias variation because older studies presented poor descriptions of the experiments and only met a few criteria established by the bias analysis. Nevertheless, our findings show that there has been an improvement in the detail presented by the studies over the years, probably due to the development of new techniques and statistical methods as well as the increase in the availability of guidelines and regulatory strategies adopted to stimulate the preparation of clearer and shorter scientific reports.
Despite the methodological limitations, important elements in the experimental designs were correctly identified in our survey, contributing to the reliability and reproducibility of the studies, especially in the most recent reports. Data such as the animal model, sex, weight, parasite strain, route of administration and parasitaemia were consistently described. Our results show that murine models were most used in the investigations. A suitable selection of animal species and genetic background is crucial in investigations of parasitic diseases, since these factors are directly related to host resistance and susceptibility to the pathogen (Andrade et al., Reference Andrade, Machado, Chiari, Pena and Macedo2002; León et al., Reference León, Montilla, Vanegas, Castillo, Parra and Ramírez2017). In the present study, the presence of T. cruzi infection was associated with a high prevalence of T. cruzi infection. In addition to the similarity with humans, murine models are easier to handle, lodge and present low maintenance costs compared with other animal models. Our data also revealed that only a reduced number of studies used larger animals as models of Chagas’ disease, especially horses and pigs. Possibly this limitation was due to the low availability, high costs and problems to attain the necessary approval by the ethics committees.
Most studies used similar strains to induce T. cruzi infection. The selection of the parasitic strain is essential because they vary in infectivity, pathogenicity, tropism and virulence (Andrade et al., Reference Andrade, Machado, Chiari, Pena and Macedo2002; Manoel-Caetano and Silva, Reference Manoel-Caetano and Silva2007; León et al., Reference León, Montilla, Vanegas, Castillo, Parra and Ramírez2017). Most of the strains used in the studies analysed in the present work are known to present high virulence and pathogenicity. These data are in accordance with the main morphological findings presented in our results, with a predominance of moderate-to-intense inflammatory foci, and a high number of mononuclear and lymphocytic infiltrates. These elements are closely correlated with acute patterns of infection since the animals often die before developing chronic infection (Chatelain and Konar, Reference Chatelain and Konar2015). Because the strains of parasites used matched the phases of interest in Chagas’ disease, i.e. the acute phase, the studies analysed herein exhibited an important element of methodological consistency, with a positive effect on the validity of the description.
The most frequent morphological findings found in our review were foci of inflammatory infiltrates, predominantly of mononuclear cells, mainly lymphocytes (CD4 + T and CD8 + T), in the CNS during the acute phase of T. cruzi infection. The sites most frequently identified with inflammatory foci were perivascular spaces, meninges of the brain, and the nerve roots of the spinal cord. Considering that the CNS is thought to be an immunoprivileged site due to the presence of the BBB (Ziv et al., Reference Ziv, Ron, Butovsky, Landa, Sudai, Greenberg, Cohen, Kipnis and Schwartz2006), the development of inflammatory infiltrates in these regions only occurs in cases of intense infection, especially in cases of American or African trypanosomiasis (Galea et al., Reference Galea, Bechmann and Perry2007; Masocha and Kristensson, Reference Masocha and Kristensson2012). This may explain why the way T. cruzi manages to enter the CNS is poorly studied. Increased BBB permeability occurs when factors derived from pathogens (e.g. cysteine protease) are recognized by T lymphocytes. The activation of these lymphocytes leads to the production of cytokines (IFNα/β, IFNγ and TNFα), which diffuse into the CNS, thereby stimulating the brain endothelial cells to produce Activated Leucocyte Adhesion molecules (ALCAM, ICAM-1) and Vascular Cell Adhesion-1 molecules (VCAM-1) that favour cell migration. In addition, these cytokines also stimulate astrocytes to produce chemotactic cytokines such as CXCL10 that increase the permeability of the BBB, allowing the dissemination of flagellate forms of T. cruzi and also of lymphocytes that may contain within them the amastigote form of the parasite (Rocha et al., Reference Rocha, de Meneses, De Meneses, da Silva, Ferreira, Nishioka, Burgarelli, Almeida, Turcato, Metze and Lopes1994; Silva et al., Reference Silva, Pereira, Souza, Silva, Rocha and Lannes-Vieira2010; Masocha and Kristensson, Reference Masocha and Kristensson2012). In our study, the presence of amastigotes in the cytoplasm of basal, glial (astrocytes, microglia, ependymocytes and oligodendrocytes) and Purkinje cells, as well as in the cerebellum was observed in most studies, along with the foci of inflammatory lymphocytic infiltrates in the CNS. On the other hand, the presence of flagellate trypanosomes also stimulates the humoral response and consequently increases the permeability of the BBB (Masocha and Kristensson, Reference Masocha and Kristensson2012).
Moreover, mononuclear cells and macrophages respond by recognizing circulating Pathogen-Associated Molecular Patterns (PAMPs), thereby producing proinflammatory cytokines, such as IL-1 and IL-6, which diffuse into the CNS and stimulate the production of mediators, such as prostaglandin E, that increase vascular permeability and consequently facilitate the entry of inflammatory cells into the CNS (Vitkovic et al., Reference Vitkovic, Konsman, Bockaert, Dantzer, Homburger and Jacque2000; Banks, Reference Banks2009; Chizzolini and Brembilla, Reference Chizzolini and Brembilla2009; Kawai and Akira, Reference Kawai and Akira2010; Guillamón-Vivancos et al., Reference Guillamón-Vivancos, Gómez-Pinedo and Matías-Guiu2015). All these alterations allow the installation of an inflammatory process that will be controlled by astrocytes, microglia and neurons (Galea et al., Reference Galea, Bechmann and Perry2007). However, the mechanisms underlying this control and which mediators are involved in inhibiting cell proliferation remain unclear. It is now known that regulatory T cells are also activated to control cell migration and consequently inflammation (Trajkovic et al., Reference Trajkovic, Vuckovic, Stosic-Grujicic, Miljkovic, Popadic, Markovic, Bumbasirevic, Backovic, Cvetkovic, Harhaji, Ramic and Stojkovic2004). However, in the case of T. cruzi infection, this modulation is not sufficient to prevent cell migration and consequently to limit the installation of acute inflammation in the tissue (Cabral-Piccin et al., Reference Cabral-Piccin, Guillermo, Vellozo, Filardy, Pereira-Marques, Rigoni, Pereira-Manfro, DosReis and Lopes2016).
The various clinical manifestations that occur throughout the development of Chagas’ disease are directly related to the genotype of the circulating parasites, the geographic origin and the cycles of wild and domestic transmission. This is because these variations in the populations determine the tropism to the tissues, the parasitaemia, and the pathogenesis in the vertebrate hosts during the acute and chronic phase of the disease (Andrade et al., Reference Andrade, Machado, Chiari, Pena and Macedo1999; Macedo et al., Reference Macedo, Machado, Oliveira and Pena2004; Magalhães-Santos et al., Reference Magalhães-Santos, Souza, Lima and Andrade2004). In our review, we observed that, after 50 inoculations with more than 20 different T. cruzi strains, those belonging to the TcI (ex Colombian), TcII (ex Y) (Galea et al., Reference Galea, Bechmann and Perry2007) and TcIV (ex AM05) (Meza et al., Reference Meza, Kaneshima, de Oliveira Silva, Gabriel, de Araújo, Gomes, Monteiro, Barbosa and de Ornelas Toledo2014) were those that presented histotropism for the CNS. The TcI and TcII strains can be found in other tissues (Andrade et al., Reference Andrade, Galvão, Meirelles, Chiari, Pena and Macedo2010; Galea et al., Reference Galea, Bechmann and Perry2007; Zingales et al., Reference Zingales, Miles, Campbell, Tibayrenc, Macedo, Teixeira, Schijman, Llewellyn, Lages-Silva, Machado, Andrade and Sturm2012), although the TcIV genotypes favour CNS tropism (Meza et al., Reference Meza, Kaneshima, de Oliveira Silva, Gabriel, de Araújo, Gomes, Monteiro, Barbosa and de Ornelas Toledo2014). This trend shows us the importance of knowing the genotype of T. cruzi to fully understand the manifestations and clinical evolution of the disease. Based on this tropism, it is possible to evaluate the need for new, more efficient and less toxic treatments according to the main infection sites of the parasite. The relationship between the parasite genotype and tropism may be relevant for the rational design of drugs capable of reaching the priority infection sites. However, there is a natural difficulty in the treatment of infections in the CNS, because the BBB is a highly selective component that limits the therapeutic distribution, making it difficult to use effective concentrations for parasitism in the nervous tissue without causing toxic effects to the organism. Due to this real difficulty, some groups are dedicated to the study and development of new drugs effective and with low side effect (Flores-Vieira and Barreira, Reference Flores-Vieira and Barreira1997; Flores-Vieira et al., Reference Flores-Vieira, Chimelli, Fernandes and Barreira1997; Jeganathan et al., Reference Jeganathan, Sanderson, Dogruel, Rodgers, Croft and Thomas2010; Perin et al., Reference Perin, Moreira da Silva, Fonseca, Cardoso, Mathias, Reis, Molina, Correa-Oliveira, Vieira and Carneiro2017).
Histopathological analysis was the most used strategy to study morphological changes in the CNS during T. cruzi infection, most probably because it is a simple, fast and economical method when compared with electronic microscopy and immunohistochemistry analysis. The method allows the study of large sections of the tissue sample and provides a valuable diagnostic tool to examine the internal architecture of the infected tissues (Mescher, Reference Mescher2016). In addition, histopathological studies allow the identification of typical tissue responses that vary as the infection progresses from the acute to chronic or disseminated phases (Gupta et al., Reference Gupta, Bhalla, Khurana and Singh2009). The most great challenge for the real comprehension of the pathogenesis of the nervous clinical form of Chagas disease is the lack of association between the morphological/histopathological lesions and the clinical manifestations of patients. When histological changes observed in tissues have a direct relevant relationship with the clinical manifestations, and can thus provide complementary information to correctly identify some particular type of microorganism that may be causing of alteration in tissues (Woods and Walker, Reference Woods and Walker1996; Procop and Wilson, Reference Procop and Wilson2001). Therefore, the analysis of studies that report specific morphophysiological changes caused by parasites or a particular strain of the parasite may contribute to the association between tissue/physiological changes and the clinical picture manifested by individuals with parasitic diseases, which may help to make a diagnosis and treatment more efficient.
This review is the first to systematically compile the results of studies describing the changes caused by T. cruzi in the CNS. Our findings reinforce the importance of some analyses in the early stages of the diagnosis of Chagas’ disease, such as parasite load, since in some cases the surrounding parasites may not be detected, but may be causing progressive damage to organs such as the heart, oesophagus and colon (Gironès and Fresno, Reference Gironès and Fresno2003; Teixeira et al., Reference Teixeira, Nascimento and Sturm2006). This negative correlation is due to critical aspects of Chagas’ disease such as the genotype and the infecting strain of T. cruzi as well as the host's immunogenetics (Costa et al., Reference Costa, da Costa Rocha, Moreira, Menezes, Silva, Gollob and Dutra2009), which would dictate the final predictive parameters. Thus, the parasite's persistence mechanisms and the quality of the immune response may determine the extent of tissue damage (Gutierrez et al., Reference Gutierrez, Mineo, Pavanelli, Guedes and Silva2009). Based on this, we described herein the organs or tissues that can undergo alterations and the type of alterations, which may help an accurate description of the clinical picture associated with the disease. Although this study evaluated only animal models and does not necessarily accurately reflect human disease, it addresses clinically relevant issues, including tissue tropism, symptoms, immune response and treatments (Chatelain and Konar, Reference Chatelain and Konar2015), and therefore may have its results extrapolated to human chagasic patients.
The selection of the studies composing this review was based on widely accepted and recommended practices for systematic reviews. A relevant issue highlighted in our study is the bias of the publications. To detect this, we used the ARRIVE Guidelines (Kilkenny et al., Reference Kilkenny, Browne, Cuthill, Emerson and Altman2010), which allow to test the degree of reliability of the studies individually and later collectively. It allowed us to notice that various aspects related to the organization and description of the experiments were neglected, among them the lack of randomization and the absence of double-blind studies, mainly in studies performed more than 15 years ago. Our data suggest a low methodological rigor of the studies at the beginning of the research efforts involving T. cruzi. For this reason, a systematic review on this subject is important, since it indicates the shortcomings of the work already carried out and indicates that future work should be more careful to allow the reproducibility of the techniques and the quality of the results.
In conclusion, the present systematic review was able to compile studies that evaluated histopathological changes in the CNS during T. cruzi infection, in which the differential tropism of the TcI, TcII and TcIV and TcVI genotypes was evidenced by structures of the brain, cerebellum and spinal cord. Changes such as the intensity of the inflammatory foci and the number of nests of parasites were shown to be linked to the genetic diversity of the different strains of T. cruzi, geographic origin and cycles of wild and domestic transmission of the strains. Finally, we highlight how detailed knowledge about the various clinical conditions that may occur during Chagas’ disease are determinant not only to support the current knowledge about this disease but also as a facilitator of early and efficient diagnosis to guarantee an adequate treatment and a good quality of life for the individuals affected.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0031182019000210
Author ORCIDs
Rômulo Dias Novaes, 0000-0002-3186-5328; Reggiani Vilela Gonçalves, 0000-0002-5831-3590
Acknowledgments
We would like to thank the ‘Conselho Nacional de Desenvolvimento Científico e Tecnológico’ (CNPq) and ‘Fundação de Amparo à Pesquisa do Estado de Minas Gerais’ (FAPEMIG).
Financial support
This work received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflict of interest
None.
Ethical standards
Not applicable.
