Disturbances in both humoral and cellular immunity have been described in patients with dilated cardiomyopathy and have been implicated in the initiation and progression of this disease entity. In particular, a variety of autoantibodies against cardiac cellular proteins have been identified in dilated cardiomyopathy, 1 2 3 4 5 6 7 and the list includes G-protein–linked receptors (such as the β 1-adrenergic and muscarinic cholinergic receptors), myosin, mitochondrial proteins (such as the adenine nucleotide translocator and keto-α-acid dehydrogenase), actin, tubulin, heat shock proteins, and the sarcoplasmic reticulum ATPase. The list is likely to expand as more putative autoantigens are tested. The pathophysiological relevance of these autoantibodies, however, is far from clear. After all, low titers of autoantibodies are found in normal subjects and are part of the immunologic repertoire. Interpretation of the findings is further complicated by the fact that dilated cardiomyopathy is most likely nosologically heterogeneous, and immune mechanisms may be important in only a subset of patients with this disease. Furthermore, observations are routinely made in patients with established disease, and inferences about the mechanisms by which myocardial injury was initiated must, by necessity, be based on indirect and circumstantial evidence. These reservations notwithstanding, three possible mechanisms through which autoantibodies can participate in the pathophysiology of dilated cardiomyopathy may be distinguished.

1. Initiation of cardiac injury. Evidence that autoantibodies can directly damage the myocardium and initiate the sequence of events that lead to dilated cardiomyopathy comes exclusively from experimental models. For example, in genetically susceptible strains of mice, injection of anti-myosin antibodies leads to autoimmune myocarditis morphologically similar to that induced by myosin injections. 8 It should be noted, however, that autoimmune myocarditis associated with high titers of anti-myosin antibodies can also be induced by injection of cardiac myosin, although this experimental model is thought to be mediated by T lymphocytes and not by the anti-myosin antibodies. 9 It is probable that development of autoantibodies is, by itself, insufficient to cause disease unless other, as yet undetermined, genetically based changes also take place. Two recent studies suggest that administration in rabbits of two other putative autoantigens, the muscarinic cholinergic 10 and the β 1-adrenergic receptors, 11 can lead to functional and morphological changes resembling cardiomyopathy. Although these are provocative findings, there is no corresponding evidence that autoantibodies can play an initiating role in human dilated cardiomyopathy.