DCM is a chronic form of heart disease characterized by right ventricular dila-tion and impaired contraction. The clini-cal spectrum varies from asymptomatic cardiomopathy to severe congestive heart failure. Patients may also display signs or symptoms of arrhythmias, systemic pulmonary vein congestion, triple gallup rhythms, and mitral or tricuspid regurgita-tion. Once the diagnosis is established, the outcome is poor, with a five-year mortal-ity of 46 percent. In 1985 in Minnesota, the prevalence was 36.5/100,000, and African Americans and males were associated with increased risk. In Sweden, the incidence was 10/100,000.
It is uncertain how many cases rep-resent progression from myocarditis to DCM since in general the endomyocar-dial biopsy is not routinely performed in DCM patients. Some reports have indi-cated that there is a relationship between the two, but biopsies in both diseases are needed to prove or disprove this concept.
In both myocarditis and DCM, there appears to be an association with prior Coxsackie virus infections. In children and infants with DCM, significantly increased titers of neutralizing antibody to the virus has been found. In addition, Coxsackie virus B–specific nucleic acid sequences have been found in the heart tissues of a small number of these patients. Furthermore, using poly-merase chain reaction (PCR) techniques, many children with suspected myocarditis were PCR positive for several viral agents in endomyocardial biopsies (twenty-six of thirty-eight samples from thirty-four patients), and the viral sequences were more often adenoviral than enteroviral.
As in other immune-mediated diseases of the heart, many investigators have now explored the role of autoantibodies to the heart as an explanation for the dis-ease. Using cardiac tissue from rats and humans (either frozen tissue or isolated myocytes), immunofluorescence stud-ies showed the antibody localized on the myocyte, giving a sarcolemmal, myolem-nal pattern or on the striations, producing a fibrillar pattern. Whether these two pat-terns represent different forms of the dis-ease is unclear because both patterns were seen in the sera of mice immunized with cardiac myosin. The specificity of these patterns is unclear because many normal sera gave the same pattern in one series (91 percent of patients vs. 35 percent of controls). Another series was more spe-cific where 59 percent of acute myocarditis patients where positive versus 0 percent of healthy controls.
Perhaps more sensitive has been the use of the Western immunoblot technique. Here 97 of 103 samples exhibited positive reactivity but no single pattern was seen to be unique in either group. However, myocarditis sera reacted more to myosin heavy chain, while cardiomyopathy sera reacted more to muscle actin. Again, many normal sera also reacted to these antigens but in lower titers. Further techniques such as enzyme-linked immunosorbent assay (ELISA), using purified cardiac myosin heavy chain material, may be needed for clinical evaluation of these antibodies.
Using a variety of well-defined cardiac cytoskeletal proteins (myosin, laminin, and β1-adrenergic receptors) as well as mitochondrial components such as ade-nosine nucleotide translocator (ANT) and branched-chain ketodehydrogenase (BCKD), ELISA studies have revealed that autoantibodies to ANT were elevated in 24/32 patients with DCM. Control sera were within normal limits.
These results show that many patients with myocarditis and with DCM develop autoantibodies to a number of cardiac constituents. Large-scale evaluation is necessary before one can conclude that the detection of any single antibody, or group of antibodies, is sufficiently sensitive and specific to replace the endomyocardial biopsy as a primary diagnostic tool.
None of these antibodies to cardiac anti-gens is known to play a direct pathogenetic role in the disease. However, the presence of antibodies to β1-adrenergic receptors in DCM and Chagas’ disease is highly suggestive of a direct pathogenetic effect, since the antigen is accessible on the surface of the myocardiocyte. β1-adrenergic recep-tor antibodies can induce apoptosis in iso-lated adult cardiomyocytes and antibodies activating the receptors are associated with reduced cardiac function in chronic heart failure. Antibodies to the mitochondrial antigens, ANT and BCKD, may also have adverse consequences on cardiac function. It is not clear, however, whether these anti-bodies have access to their target antigens in vivo.
Because of the possible autoimmune origin of myocarditis and DCM in humans, and the well-documented association of experimental myocarditis with MHC in mice (Rose et al. 1988) a number of stud-ies to determine the relationship with the human MHC (HLA) have been carried out. Several investigators have reported that DCM patients had an increased frequency of HLA-DR4 and a decreased frequency of HLA-DR6. The largest study to date recon-firmed these findings and also found that DR4-Dqw4 haplotype conferred height-ened risk of disease.
A predominance of myocarditis in males has been reported in a number of studies. The proportion of male patients is about 60 percent. In this respect, myocardi-tis differs from most autoimmune diseases, which predominantly affect females.
Myocarditis has both infectious and non-infectious causes. Acute myocarditis is associated with infections of many types, including bacterial, rickettsial, viral, mycotic, protozoan, and helminthic. As stated previously, several viruses have been implicated in this disease and, in some cases, multiple viruses may be detected in the heart. In Europe and North America, among the most common agents are the enteroviruses and the adenoviruses. It has been reported that Coxsackie virus group B infections were associated with at least half of the acute cases of myocarditits and by immunofluorescence, the Coxsackie virus B antigen was found in the myocardium of 30.9 percent of routine autopsy specimens of myocarditis. Serotype B3 is identified most frequently.
Since enteroviruses have been most often implicated in human myocardi-tis and DCM, animal models with these agents have been used to investigate the pathogenic mechanisms of these diseases. Although infections with Cox Sackie Virus B3 are relatively common in humans, the development of clinically significant car-diac disease in humans is uncommon, sug-gesting that differences in host response may be crucial. These differences may be related to genetic factors or may relate to virus-specific receptors on host tissue.
Thus, mice with a wide variety of genetic backgrounds have been used to study the viral host relationships related to human disease.
All strains of mice tested developed acute myocarditis starting two or three days after CB3 infection. The disease reached its peak on day 7 and gradually resolved, so that by twenty-one days after infection the heart was histologically normal. In a few strains of mice, however, the myocar-ditis persisted. Further studies resolved CB3-induced myocarditis into two distinct phases. The first phase occurred during the first week after infection and was charac-terized by focal necrosis of myocytes and an accompanying focal acute inflamma-tory response with a mixed-cell infiltrate, consisting of polymorphonuclear and mononuclear cells. The second phase of CB3-induced myocarditis becomes evident about nine days after infection and is fully manifested by fifteen to twenty-one days after infection. Histologically, the inflam-matory process was diffuse rather than focal and consisted mainly of mononuclear interstitial infiltrate. Little or no myocyte necrosis was evident at that time. Infectious virus could be cultured only during the first phase of disease; no virus was isolated after day 9. Heart-reactive autoantibodies were present in all strains that developed the second phase of myocarditis. Only cer-tain inbred strains of mice developed this secondary autoimmune myocarditis. Sus-ceptibility was determined primarily by non-H-2-background genes, although H-2-encoded differences influenced the sever-ity of the autoimmune response initiated by molecular mimicry between the virus and heart antigens. Available evidence suggests that the autoimmune response depends on virus-induced damage to the heart since it was observed that transgenic mice expressing IFN-γ in their pancreatic beta cells failed to develop CB3-induced myocarditis, even though the virus prolifer-ated in other sites. This work challenges the concept of molecular mimicry as the mech-anism initiating the autoimmune response and suggests that it is due to a “bystander effect.” The virus infections may serve as an adjuvant for the cardiac antigens liber-ated during the viral infection of the heart.
The presence of autoantibodies in the late phase of CB3-induced myocarditis pro-vided the opportunity to evaluate the patho-genic significance of the immune response. The first step was to characterize the car-diac antigen, and it was found that cardiac myosin heavy chain was the major antigen. Knowing this, Rose and colleagues demon-strated that purified mouse cardiac myosin injected with complete Freund’s adjuvant (CFA) produced lesions that resembled the late phase of CB3-induced myocardi-tis. Furthermore, only strains of mice that were genetically susceptible to the disease produced these lesions, whereas resistant strains did not. One can also induce the disease with peptides derived from the car-diac myosin molecule.
Cellular immunity also played a role in that autoreactive T cells were present in the lesions. Two types of cytotoxic T cells were discovered. One was specific for CB3-infected myocytes, and one was specific for uninfected myocytes. The latter reactivity was probably a result of the appearance of novel determinants expressed on the myo-cyte following altered myocyte metabolism. With regard to the myosin-specific T cells, mice depleted of CD4+ or CD8+ T cells failed to develop myocarditis after myosin immunization. Both humoral and cellular responses to cardiac myosin participate in the lesion as antibody plus complement is deposited in the hearts of these mice.
Recent investigations have focused on the critical role of the initial, innate immune response in switching from harmless autoimmunity to a pathogenic autoimmune response. Administration of pro-inflammatory cytokines, IL-1 or TNF-α, to genetically resistant B10A mice confers susceptibility to myocarditis after challenge with CB3 or murine myosin in CFA. However, antibody to TNF-α or IL-1 receptor antagonist prevents or delays the onset of myocarditis in genetically suscep-tible A/J mice, leading to the conclusion that these early inflammatory mediators are required for autoimmune myocarditis.
Myocarditis can be induced in mice by other viruses, including mouse cytomega-lovirus (CMV), a herpes virus. Sublethal mouse CMV infections of BALB/c mice induce inflammation of the heart similar to that seen in the CB3-induced disease in mice. Infectious virus was not detected in the heart after ten days, but antibodies to cardiac myosin were evident.
Until recently, the only treatments for myo-carditis and DCM have been supportive therapies, such as bed rest and treatment of heart failure, arrhythmias, and embolic events if present. In many centers, cardiac transplantation has become the eventual treatment of choice in patients with refrac-tory heart failure. The role of immunosup-pressive therapy in myocarditis remains controversial. Numerous reported studies on relatively small numbers of patients have generally found that, while some individuals respond well to immuno-suppression (prednisone and cyclospo-rine), others fail to respond or even have serious adverse reactions that preclude continued treatment. The major problem at present is the difficulty in distinguishing immune-mediated cardiac disease from infectious, genetic, or toxic forms of the disease. Obviously, until there are reliable biomarkers to distinguish autoimmune myocarditis/DCM, treatment cannot be rational or capable of evaluation.