**4.1 Dysautonomia**

As early as 1922 Carlos Chagas noted that the chronotropic response to atropine was altered in chagasic patients (Chagas C. & Vilella, 1922), but it was not until late 1950's that Köberle published his works showing impressive neuronal depopulation in microscopic sections obtained from the intercaval atrial strip in chagasic patients using a standardized technique of cardiac intramural neuronal counting developed by himself (Köberle, 1956a, 1956b). These findings led to the "neurogenic hypothesis" (Köberle, 1959), which explained all megas in Chagas' disease as a consequence of neuronal depletion.

Although many other authors claimed to have confirmed this finding (Mott & Hagstrom, 1965, Oliveira J. S., 1985), other authors called to attention about the criteria used to diagnose neuronal depletion because of the great variability in the number of neurons in autonomic ganglia (Rossi L., et al., 1994) and they also remarked that the only right criterion to establish neuronal depletion is the presence of proliferation of satellite cells, with the formation of Terplan's nodules, a characteristic lesion described as proliferating satellite cells which replace degenerating neurons, forming nodular structures. These lesions, once considered patognomonic, can be found in other cardiomyopathies (Rossi L., et al., 1994). The same author could not confirm the loss of neurons or denervation in CCC (Rossi L., 1988). Finally, it was demonstrated that, using Terplan's nodules as diagnostic criterion,

Pathogenesis and Pathology of Chagas' Chronic Myocarditis 133

of endothelial cells. This results in thrombin binding and platelet aggregation (Libby, et al., 1986). *T. cruzi* also produces tromboxane A2 (TXA2), specially during amastigote state (Ashton, et al., 2007), also favouring platelet aggregation and vascular spasm. Direct parasitism of endothelial cells by *T. cruzi* has also been demonstrated, and this causes the activation of the NF-B pathway increasing the expression af adhesion molecules (Huang, et al., 1999a), and secreting proinflammatory citokines (Tanowitz Herbert B., et al., 1992a) and

Endothelin-1 (ET-1) is another proposed pathogenic element. Elevated levels of mRNA for preproendothelin-1, endothelin converting enzyme and endothelin-1 were observed in the infected myocardium (Petkova Stefka B., et al., 2000), and elevated levels of ET-1 have been found in CCC patients (Salomone, et al., 2001). Mitogen-activated protein kinases and the transcription factor activator-protein-1 regulate the expression of endothelin-1, and both are shown to be increased in myocardium, interstitial cells and vascular and endocardial endothelial cells (Petkova S. B., et al., 2001). Besides, treatment with phosphoramidon, an inhibitor of endothelin converting enzyme, decreases heart size and severity of lesions in an

Inflammation also produces dysfunction of endothelial cells. Macrophages secrete TXA2 and platelet activating factor (PAF) act on endothelium causing vasoconstriction (Rossi M. A. & Carobrez, 1985). Endothelial cells infected *in vitro* with *T. cruzi* lose their antithrombotic properties in response to interleukin 1 (IL-1) (Bevilacqua, et al., 1984, Nachman, et al.,

It is remarkable that, although the data presented, endothelial function seems to be normal in CCC patients without heart failure, as measured by increases in blood flow in response to acetylcholine and sodium nitroprusside (Consolim-Colombo, et al., 2004). Further, in our concept, microvascular damage found in CCC, seems to be secondary to fibrosis and distortion of myocardial fiber arrangement by necrosis and chronic infiltrates, but as once

The most striking characteristic of CCC is enlargement of heart with variable degrees of dilatation of chambers (Andrade, 1985) (Figure 4A). In autopsy series, hearts were overweighted (Andrade, 1985, Baroldi, et al., 1997, Bestetti, et al., 1993, Lopes, et al., 1981) compared with indeterminate chagasic patients and non-chagasic subjects. Marked cardiomegalies reached up to 500 grams. Right ventricle (RV) and atrium (RA) were generally more compromised than left chambers, being RV the most dilated in one paper

A second remarkable feature is the thinning of the left ventricular apical wall, resulting in

Other lesions described are flattening of the papillary muscles and a marked subendocardial sclerosis, parietal and/or aneurismal thrombosis and fibrotic plaques in pericardium (Milei,

apical aneurysm, a very characteristic lesion in CCC (Figure 4B) (Moia, et al., 1955).

iNOS (Huang, et al., 1999b).

1986).

**5. Pathology** 

**5.1 Macroscopic features** 

experimental model of Chagas' disease (Jelicks, et al., 2002).

established, may contribute to the perpetuation of myocardial damage.

Pathological findings are described mostly according to our own findings.

(Laranja, et al., 1956) but RA was in other (Andrade, 1985).

et al., 1996a, Milei, et al., 1991b, Storino & Milei, 1994).

CCC patients with heart failure had more neuronal depletion than patients with dilated cardiomyopathy of other causes (Oliveira J. S., 1985). In our experience the neuroganglionic involvement was variable in autopsies of chagasic hearts (Milei, et al., 1991b).

According to pioneer neurogenic hypothesis (Köberle, 1959), early and irreversible damage to the parasympathetic system during acute phase of the disease causes a cathecolaminergic cardiomyopathy, but this point of view has been debated and evidence is contradictory. Functional test performed in CCC patients demonstrated impaired parasympathetic heart rate regulation (metaraminol, phenylephrine and atropine intravenous injections, facial immersion, Valsalva maneuver, head-up and head-down tilt tests, respiratory sinus arrhythmia, handgrip, graded dynamic exercise, and spectral analysis of Holter recordings) (Amorim, et al., 1968, Amorim, et al., 1973, Gallo, et al., 1975, Guzzetti, et al., 1991, Junqueira Junior, et al., 1985, Manço, et al., 1969, Marin-Neto, et al., 1975, Sousa, et al., 1987). However, a careful analysis of these data showed that many patients had normal autonomic function and most patients had heart failure, that could explain autonomic dysfunction *per se* (Davila, et al., 1998).

On the other hand, the study of indeterminate patients has shown conflicting results. While some authors could demonstrate impaired autonomic function (Molina, et al., 2006, Vasconcelos & Junqueira, 2009) others could demonstrate that autonomic function was normal in patients without myocardial damage and that abnormalities in autonomic dysfunction was proportional to heart dysfunction, leading these authors to propose that these abnormalities arise as a compensating mechanism for the progressive left ventricular dilatation (Davila, et al., 1991, Davila Spinetti, et al., 1999). These findings led to a new "neurogenic theory", which considers autonomic dysfunction as secondary to ventricular dilatation and hemodynamic alterations, but once installed, acts synergistically with parasitism and inflammation to cause further myocardial damage (Davila, et al., 2004).

#### **4.2 Microvascular damage**

Microcirculation abnormalities in CCC have been firstly pointed out by Jorg as an angiographic anarchy due to capillary loss (Jörg, 1974) and furtherly demonstrated in experimental models as well as in clinical practice (Rossi M. A., et al., 2010).

Many investigators have found abnormal myocardial perfusion using isonitrile-99mtechnetium (Castro R., et al., 1988) and thallium-201 (Hagar & Rahimtoola, 1991, Marin-Neto, et al., 1992) scintigraphy in chagasic patients with normal epicardial coronary arteries. Furthermore, the progression of left ventricular systolic dysfunction is associated with both, the presence of reversible perfusion defects and the increase in perfusion defects at rest (Hiss, et al., 2009, Schwartz & Wexler, 2009). Anatomopathological studies in humans also provided evidence of microvascular damage in CCC. In late 1950's first reports showing collapse of arterioles and intimal proliferation (Torres, 1960) caught the attention of investigators. Also, microthrombi have been described (Rossi M. A., et al., 1984). As said, in endomyocardial biopsies thickening of capillary basement membranes was also found (Milei, et al., 1992b).

Additional evidence of microvascular damage was obtained from experimental models. Vascular constriction, microaneurysms, dilatation and proliferation of microvessels has been demonstrated (Factor & Sonnenblick, 1982, Morris, et al., 1989, Tanowitz H. B., et al., 1996, Tanowitz Herbert B., et al., 1992b).

Many factors have been advocated in the genesis of these lesions. First, the parasite itself. It was shown that *T. cruzi* produces a neuraminidase that removes sialic acid from de surface of endothelial cells. This results in thrombin binding and platelet aggregation (Libby, et al., 1986). *T. cruzi* also produces tromboxane A2 (TXA2), specially during amastigote state (Ashton, et al., 2007), also favouring platelet aggregation and vascular spasm. Direct parasitism of endothelial cells by *T. cruzi* has also been demonstrated, and this causes the activation of the NF-B pathway increasing the expression af adhesion molecules (Huang, et al., 1999a), and secreting proinflammatory citokines (Tanowitz Herbert B., et al., 1992a) and iNOS (Huang, et al., 1999b).

Endothelin-1 (ET-1) is another proposed pathogenic element. Elevated levels of mRNA for preproendothelin-1, endothelin converting enzyme and endothelin-1 were observed in the infected myocardium (Petkova Stefka B., et al., 2000), and elevated levels of ET-1 have been found in CCC patients (Salomone, et al., 2001). Mitogen-activated protein kinases and the transcription factor activator-protein-1 regulate the expression of endothelin-1, and both are shown to be increased in myocardium, interstitial cells and vascular and endocardial endothelial cells (Petkova S. B., et al., 2001). Besides, treatment with phosphoramidon, an inhibitor of endothelin converting enzyme, decreases heart size and severity of lesions in an experimental model of Chagas' disease (Jelicks, et al., 2002).

Inflammation also produces dysfunction of endothelial cells. Macrophages secrete TXA2 and platelet activating factor (PAF) act on endothelium causing vasoconstriction (Rossi M. A. & Carobrez, 1985). Endothelial cells infected *in vitro* with *T. cruzi* lose their antithrombotic properties in response to interleukin 1 (IL-1) (Bevilacqua, et al., 1984, Nachman, et al., 1986).

It is remarkable that, although the data presented, endothelial function seems to be normal in CCC patients without heart failure, as measured by increases in blood flow in response to acetylcholine and sodium nitroprusside (Consolim-Colombo, et al., 2004). Further, in our concept, microvascular damage found in CCC, seems to be secondary to fibrosis and distortion of myocardial fiber arrangement by necrosis and chronic infiltrates, but as once established, may contribute to the perpetuation of myocardial damage.
