The present study evaluated biventricular geometrical and functional ventricular changes in an experimental model of Chagas disease in Syrian hamsters over time in comparison with a Control group.
The findings revealed that parameters of longitudinal and circumferential myocardial deformation obtained by STE are significantly depressed early after the infection with T. cruzi. In contrast, LVESD and LVEF showed changes at an advanced time point of CD, when mortality is high. Besides, the present investigation indicates an early involvement of the right ventricle in animals infected with T. cruzi, before the overt dysfunction of LV.
Evolution of geometrical and biventricular functional parameters in the experimental animal model of CD in hamsters
Although the phases and forms of CD are difficult to be reproduced in rodent models of T. cruzi infection, in Syrian hamsters CD can be replicated through all clinical stages including the acute phase, the latent period (indeterminate), the chronic cardiomyopathy, with myocardial dysfunction caused by multifocal and diffuse myocarditis and interstitial fibrosis. The first month after infection in hamsters presents the acute phase of the disease, with a high mortality rate. In our study, we observed 27% of mortality at this point and this finding is similar to that reported in a previous study, which showed a mortality rate of 33% in the first month [16]. After the acute phase, survivors evolve to an apparent quiescent period, with no clinical manifestations neither evidence of overt myocardial systolic dysfunction, similarly to the indeterminate form [32] of chronic CD in humans. Prior studies using the same animal experimental model did not find LV myocardial impairment with conventional echocardiography until 6 months after the initial infection.[33] In our study, LVEF and LVESD were reduced in infected animals compared to controls at 6 months of disease evolution, suggesting that the latent phase of CD in this experimental animal model may be shorter than previously reported [15, 16, 34]. It is also plausible to assume that those differences could be explained, at least in part, by the different echocardiographic techniques applied [15], which, in our study, used a higher spatial resolution ultrasound system.
Other studies reported on late LV dilatation in hamsters infected with T. cruzi [15], In our study, LVEDD was not different in infected animals from controls until the end of experimentation time at 8 months. This finding was also observed in the investigation from Tanaka et al and may suggest not only LVEDD is a very late marker of CCC but also it could not represent LV dysfunction [33]. Our study was the first to assess two-dimensional derived LVEF by the area-length method in T. cruzi infected Syrian hamsters. Previous studies had already shown the feasibility of two-dimensional LVEF in mice and rat models of surgical myocardial overload, infarction, and LV dysfunction [35]. However, as CCC is typically a segmental disease [36], M-mode derived LVEF (Teichholz method) may not accurately estimate LV systolic dysfunction.
This was also the first study to prospectively evaluate RV function through the evolution of CD in Syrian hamsters. Our data showed RV is early compromised. These findings corroborate the results from studies in humans that have shown early changes of RV geometry and function even in the indeterminate form of CD [37, 38][39]. The damage to the RV imposed by CD was previously demonstrated in studies of RV myocardial tissue samples [40] and in a murine model of disease [41] but only in our investigation a sequential evolution of RV dysfunction has been described over time.
Speckle tracking echocardiography as a tool for early detection of myocardial dysfunction in an animal experimental model of CD.
This was the first study to use STE in Syrian hamsters infected with T. cruzi. Our results demonstrated that GLS and GCS were able to detect early myocardial damage in animals infected with the protozoan when compared with control animals at the same age. It is noteworthy that the change in these parameters occurred despite normal values of LVEF, as early as one month after infection, a moment recognized in this model as the acute phase of the CD when mortality rates are high.
There have been some studies with STE in patients with Chagas disease. However, the behavior of myocardial deformation through all stages of the disease in a sequential analysis had not been previously demonstrated. Although in CCC global LV deformation parameters and LVEF are reduced [25, 42], results are still conflicting about the early stages of the disease [24, 26, 43]. Few clinical studies evaluated LV myocardial deformation in patients with the indeterminate form of CD. One study showed that LV radial strain was reduced in 32 patients compared to controls, but statistical significance was achieved only at trend analysis [25]. Gomes et al did not encounter STE differences between controls and patients at the early stages of the disease, except for a small group of 7 patients who presented significant fibrosis diagnosed by cardiac magnetic resonance [24].
In contrast to the reduction of LV GLS and GCS observed in infected animals since the very early phase of the disease, GRS was not different between Chagas and control groups in our study. Radial strain interpretation has been challenging in several aspects. The wide variability of GRS indexes and also low reproducibility are recognized even in normal human hearts.[44] Also, when providing normal reference intervals of STE for the Syrian hamster model, GRS presented the lowest values of reproducibility [45]. Technically the lower reproducibility of GRS is comprehensible in any, human or small animal experimental images [46], based on the difficulty not to measure strain, but to track speckles in lateral lobes of bidimensional LV images. Measurements of radial strain are made from parasternal short-axis views, where segments in lateral lobes of the ultrasound beam do not have enough spatial resolution to provide adequate tracking [47], Additionally, radial deformation may be enhanced at the initial phases of myocardial damage, as a compensatory mechanism to the reduction of longitudinal and circumferential deformation [23, 48]. Thus, clinical use and interpretation of radial strain should be cautious. Lima et al, in concordance with our findings, also did not show differences in radial strain in Chagas patients at the early stages of the disease [49].
Measurements of myocardial deformation in this study strictly followed the latest published guidelines [48]. EKG signals were always recorded, end-systolic time was marked and strain peak was always measured as a systolic peak [47]. Definitions about what strain peak was measured were not always clarified in previous studies using STE in CD [25, 26]. In pathologies such as CCC, several LV conduction disturbances may be present and hence some LV segments can show post systolic peak deformations, which may not accurately represent patients’ myocardial systolic function [23].
Histological analysis
Corroborating previous findings from our laboratory [16, 50] histologic data presented in this study show an increase of both, myocardial inflammatory cells and interstitial fibrosis in T. cruzi infected animals compared to controls. The low extent of myocardial fibrosis (4%) corroborates this model as a representative of initial myocardial damage when studied until 8 months. Previous studies show fibrosis of 10% of the whole LV myocardial area when animals reach 10 months of infection [16]. Despite the lower incidence of fibrosis, GLS correlation to inflammatory cell count reaffirms this deformation parameter as an early marker of myocardial damage in Chagas disease´s physiopathology, even before significant fibrosis. Other mechanisms which could be present in Chagas cellular damage as myocardial disarray [51, 52] were not explored in this study.