**2.2 Pathogenesis**

The first clinical report to suggest a relationship between nonspecific myocarditis and dilated cardiomyopathy in AIDS patients appeared in 1986 [Cohen et al., 1986]; 3 patients had clinical, echocardiographic, and pathologic findings of dilated cardiomyopathy and 2 of the patients had focal lymphocytic infiltration associated with myocyte necrosis. Subsequent reports suggested an association between focal nonspecific myocarditis at autopsy and clinical cardiomyopathy [Barbaro et al., 1998a, Reilly et al., 1988]. Numerous hypotheses have been suggested to account for the etiology of nonspecific myocarditis and cardiomyopathy observed in HIV-infected patients, including direct HIV-1 infection of myocardial cells or coinfection with other cardiotropic viruses, [Olson 2003] cytokine cardiotoxicity [Suffredini et al., 1989, Lahdevirta et al., 1988, Levine et al., 1990], postviral cardiac autoimmunity [Herskowitz et al., 1989, 1993], nutritional deficiencies [Olson 2003], and cardiotoxicity due to illicit drugs [Olson 2003] or pharmacologic agents [Herskowitz et al., 1992a, Olson 2003].

The histologic findings of a monoclonal or oligoclonal inflammatory cellular infiltrate suggest a viral or autoimmune cause for the myocarditis associated with HIV infection. Myocarditis is more likely in individuals with more profound immunosuppression because CD4 < 400 cells/µL is more frequently observed in patients with dilated cardiomyopathy [Barbaro et al., 1998b]. In this same series, inflammatory myocardial cellular infiltrates were predominantly CD3 lymphocytes in 12 patients and CD8 lymphocytes in 64 patients. A separate report [Herskowitz et al., 1992b] described 35 HIV-infected patients with global LV dysfunction who underwent endomyocardial biopsy in which active or borderline myocarditis was observed in 55% of patients. For those patients with biopsy-proven myocarditis, mean LV ejection fraction was 28% whereas for patients without myocarditis it was 48%. The cellular infiltrate was primarily composed of CD8 T lymphocytes.

Myocarditis in HIV Positive Patients 155

lending support to an autoimmune mechanism of disease. A study by Gu et al. [Gu et al., 1992] used monoclonal antibodies to HIV core proteins that reacted with myocyte antigens in 38 of 42 AIDS patients (and 11 of 28 non-AIDS patients), suggesting antibodies occurring in AIDS patients may react with antigenic epitopes of myocytes, thereby promoting

Multiple cytokines are suspected of having a role in the mediation of myocardial inflammation, myocyte necrosis and ventricular dysfunction in myocarditis, although specifics are incompletely understood in human disease [Liu et al., 2001]. The mononuclear cells characteristic of lymphocytic myocarditis, including the focal nonspecific myocarditis of AIDS, are a likely source of cytokines that promote inflammation and maintenance of immune response, which may lead to impaired contractile function and fibrosis. Matsumori et al. 1994 showed that patients with myocarditis have markedly increased concentrations of cytokines, including tumor necrosis factor alfa (TNF-α) and interleukin -1 and -6. In animal models of myocarditis, similar profiles of cytokine activation have been described and have been demonstrated directly in cardiac tissue [Yamada et al., 1994]. TNF-α has been demonstrated to be a negative inotrope [Suffredini et al., 1989] and is increased in patients with congestive heart failure [Sharma et al., 2000], AIDS [Odeh et al., 1990, Yamamoto et al., 1995], and myocarditis [Matsumori et al., 1994]. HIV may cause myocyte injury by an "innocent bystander destruction" mechanism, as may occur in AIDS-associated encephalitis [Ho et al., 1987]. However, whether these mechanisms operate in the myocarditis and

Three histological patterns of myocarditis have been described in patients with AIDS: - lymphocytic infiltration with myocyte necrosis [Anderson et al., 1988], which meets the


Clinical presentation of HIV associated myocarditis in symptomatic patients is generally similar to myocarditis due to other causes. The absence of symptoms and signs of heart disease does not however exclude cardiac involvement, as occurrence of sub-clinical cardiac abnormalities with possible fatal consequences in this population has been described [Kaminski et al., 1990]. Diagnosis requires the possibility of cardiac involvement to be constantly in mind and symptoms associated with myocarditis are varied, and relate either to the actual inflammation of the myocardium, or the weakness of the heart muscle that is secondary to the inflammation. Signs and symptoms of myocarditis include: **chest pain** (often described as "stabbing" in character); **congestive heart failure** (leading to edema, breathlessness and hepatic congestion); **palpitations** (due to arrhythmias); **sudden death** (in young adults, myocarditis causes up to 20% of all cases of sudden death); **fever** (especially when infectious); symptoms in infants and toddlers tend to be more non-specific with generalized malaise, poor appetite, abdominal pain, chronic cough. Later stages of the illness will present with respiratory symptoms with increased work of breathing and is

Since myocarditis is often due to a viral illness, many patients give a history of symptoms consistent with a recent viral infection, including fever, rash, diarrhea, joint pains, and


autoimmune-mediated heart muscle disease.

dilated cardiomyopathy of AIDS is unknown.

Dallas criteria;

**2.3 Clinical manifestation** 

often mistaken for asthma.

frequent fatigue.

Although it has been suspected that myocarditis and cardiomyopathy associated with HIV-1 infection may be caused by direct viral infection of myocytes, definite evidence for this is lacking. Difficulty in demonstration of a link among HIV-1 infection, myocarditis, and cardiomyopathy in AIDS is related, in part, to lack of a suitable in vivo model of the disease. Because of the limited host range of HIV-1 and the difficulty in handling nonhuman primates infected with simian immunodeficiency virus-1, little investigation has been reported of myocarditis or cardiomyopathy associated with AIDS in animal models [Lewis et al., 2000].

What is the potential for direct myocardial infection? HIV-1 invades T cells by attachment to a CD4 surface-membrane receptor. However, there are no CD4 receptors on myocyte surface membranes. It is possible that the virus gains access to myocytes by other mechanisms, although the evidence to support this concept is limited. It is also possible that injury to myocytes may facilitate entry of the HIV virion; Epstein-Barr virus (EBV) promotes entry of HIV into CD4 receptor negative cells, with subsequent replication [Olson 2003].

The presence of viral genomic material within the myocytes of HIV-infected patients with myocarditis and cardiomyopathy does not definitely establish viral infection as causal. Furthermore, the significance of the finding of viral transcripts within cells is uncertain because patients may or may not have LV dysfunction [Herskowitz et al., 1993]. HIV-1 genomic material reportedly was detected within the genome of myocardial cells, although typically the findings have been sparse and may not have represented myocyte infection because the HIV nucleic acid sequence may actually have been located in endothelial cells or macrophages [Grody et al., 1990, Lipshultz et al., 1990, Flomenbaum et al., 1989, Cenacchi et al., 1990].

In one study, in 58 of 63 patients with AIDS, LV dysfunction, and biopsy-proven nonspecific lymphocytic myocarditis, a positive hybridization signal was observed but staining was weak and affected myocytes were generally not surrounded by inflammatory cells [Barbaro et al., 1998a].

In adults with AIDS-associated myocarditis, non-HIV viruses or viral genomic material identified in myocardial tissue has included CMV, Coxsackie virus group B, and EBV [Barbaro et al., 1998b, Wu et al., 1992]. In an autopsy study of 32 children who died with advanced HIV disease, including 23 with histologic evidence of myocarditis, viral sequences detected by polymerase chain reaction included adenovirus in 6, CMV in 3, and both adenovirus and CMV in 2. No other viruses were detected by polymerase chain reaction, including HIV [Bowles et al., 1999].

A high proportion of HIV-seropositive patients with LV dysfunction have evidence of latent infection of myocytes with CMV immediate-early genes [Wu et al., 1992]. Although observation of the intranuclear inclusions of active, lytic CMV infection is unusual, it has been suggested that latent viral infection may promote enhanced major histocompatibility complex expression, thereby provoking immune-mediated injury typical of animal models of myocarditis [Wu et al., 1992].

Immune-mediated mechanisms other than direct myocardial viral infection may account for the cellular infiltrates and cardiomyopathy observed in AIDS patients. Studies performed by Herskowitz demonstrated circulating autoantibodies in 4 of 6 AIDS patients with cardiomyopathy, whereas AIDS patients without cardiomyopathy did not have these antibodies [Herskowitz et al., 1989]. In the patients with autoantibodies, antimyosin antibodies were identified. In these same individuals, no evidence of HIV-1 or other viruses was identified from myocardial biopsy specimens evaluated by in situ hybridization,

Although it has been suspected that myocarditis and cardiomyopathy associated with HIV-1 infection may be caused by direct viral infection of myocytes, definite evidence for this is lacking. Difficulty in demonstration of a link among HIV-1 infection, myocarditis, and cardiomyopathy in AIDS is related, in part, to lack of a suitable in vivo model of the disease. Because of the limited host range of HIV-1 and the difficulty in handling nonhuman primates infected with simian immunodeficiency virus-1, little investigation has been reported of myocarditis or cardiomyopathy associated with AIDS in animal models [Lewis

What is the potential for direct myocardial infection? HIV-1 invades T cells by attachment to a CD4 surface-membrane receptor. However, there are no CD4 receptors on myocyte surface membranes. It is possible that the virus gains access to myocytes by other mechanisms, although the evidence to support this concept is limited. It is also possible that injury to myocytes may facilitate entry of the HIV virion; Epstein-Barr virus (EBV) promotes entry of HIV into CD4 receptor negative cells, with subsequent replication [Olson 2003]. The presence of viral genomic material within the myocytes of HIV-infected patients with myocarditis and cardiomyopathy does not definitely establish viral infection as causal. Furthermore, the significance of the finding of viral transcripts within cells is uncertain because patients may or may not have LV dysfunction [Herskowitz et al., 1993]. HIV-1 genomic material reportedly was detected within the genome of myocardial cells, although typically the findings have been sparse and may not have represented myocyte infection because the HIV nucleic acid sequence may actually have been located in endothelial cells or macrophages [Grody et al., 1990, Lipshultz et al., 1990, Flomenbaum et al., 1989, Cenacchi et

In one study, in 58 of 63 patients with AIDS, LV dysfunction, and biopsy-proven nonspecific lymphocytic myocarditis, a positive hybridization signal was observed but staining was weak and affected myocytes were generally not surrounded by inflammatory cells [Barbaro

In adults with AIDS-associated myocarditis, non-HIV viruses or viral genomic material identified in myocardial tissue has included CMV, Coxsackie virus group B, and EBV [Barbaro et al., 1998b, Wu et al., 1992]. In an autopsy study of 32 children who died with advanced HIV disease, including 23 with histologic evidence of myocarditis, viral sequences detected by polymerase chain reaction included adenovirus in 6, CMV in 3, and both adenovirus and CMV in 2. No other viruses were detected by polymerase chain reaction,

A high proportion of HIV-seropositive patients with LV dysfunction have evidence of latent infection of myocytes with CMV immediate-early genes [Wu et al., 1992]. Although observation of the intranuclear inclusions of active, lytic CMV infection is unusual, it has been suggested that latent viral infection may promote enhanced major histocompatibility complex expression, thereby provoking immune-mediated injury typical of animal models

Immune-mediated mechanisms other than direct myocardial viral infection may account for the cellular infiltrates and cardiomyopathy observed in AIDS patients. Studies performed by Herskowitz demonstrated circulating autoantibodies in 4 of 6 AIDS patients with cardiomyopathy, whereas AIDS patients without cardiomyopathy did not have these antibodies [Herskowitz et al., 1989]. In the patients with autoantibodies, antimyosin antibodies were identified. In these same individuals, no evidence of HIV-1 or other viruses was identified from myocardial biopsy specimens evaluated by in situ hybridization,

et al., 2000].

al., 1990].

et al., 1998a].

including HIV [Bowles et al., 1999].

of myocarditis [Wu et al., 1992].

lending support to an autoimmune mechanism of disease. A study by Gu et al. [Gu et al., 1992] used monoclonal antibodies to HIV core proteins that reacted with myocyte antigens in 38 of 42 AIDS patients (and 11 of 28 non-AIDS patients), suggesting antibodies occurring in AIDS patients may react with antigenic epitopes of myocytes, thereby promoting autoimmune-mediated heart muscle disease.

Multiple cytokines are suspected of having a role in the mediation of myocardial inflammation, myocyte necrosis and ventricular dysfunction in myocarditis, although specifics are incompletely understood in human disease [Liu et al., 2001]. The mononuclear cells characteristic of lymphocytic myocarditis, including the focal nonspecific myocarditis of AIDS, are a likely source of cytokines that promote inflammation and maintenance of immune response, which may lead to impaired contractile function and fibrosis. Matsumori et al. 1994 showed that patients with myocarditis have markedly increased concentrations of cytokines, including tumor necrosis factor alfa (TNF-α) and interleukin -1 and -6. In animal models of myocarditis, similar profiles of cytokine activation have been described and have been demonstrated directly in cardiac tissue [Yamada et al., 1994]. TNF-α has been demonstrated to be a negative inotrope [Suffredini et al., 1989] and is increased in patients with congestive heart failure [Sharma et al., 2000], AIDS [Odeh et al., 1990, Yamamoto et al., 1995], and myocarditis [Matsumori et al., 1994]. HIV may cause myocyte injury by an "innocent bystander destruction" mechanism, as may occur in AIDS-associated encephalitis [Ho et al., 1987]. However, whether these mechanisms operate in the myocarditis and dilated cardiomyopathy of AIDS is unknown.

Three histological patterns of myocarditis have been described in patients with AIDS:

