S 30 BrHeartJ 1994;72 (Supplement):S 30-S 34 Role of in dilated cardiomyopathy Br J: first published as 10.1136/hrt.72.6_Suppl.S30 on 1 December 1994. Downloaded from A L P Caforio

Persistent viral infection of the myocardium disposition and environmental influences. and autoimmunity are two of the main patho- The genetic predisposition accounts for both genic hypotheses for dilated cardiomyopathy the fact that different autoimmune conditions (DCM). A unifying hypothesis also suggests may be associated in patients or in their family that an initial viral insult may trigger or pre- members and the common finding that single cipitate autoimmunity. Is it becoming clearer autoimmune diseases often run in families. whether DCM is a chronic viral disease, a The inheritance of susceptibility is usually post-infectious autoimmune process, or polygenic. Organ-specific autoimmune dis- genetically determined organ-specific auto- eases are commonly associated with specific immune disease? HLA class II ,' but it is not known When tolerance to "self' antigens is lost how the HLA system determines the pre- autoimmune disease results. It is charac- disposition to a specific disease. terised by the presence of circulating autoanti- Most organ-specific autoimmune diseases bodies, which are not necessarily pathogenic are chronic and apparently idiopathic. Organ but represent markers of continuing tissue and disease specific are found in damage.' In autoimmune disease that is not the affected patients. These antibodies are organ specific there are against also detected in family members, sometimes ubiquitous autoantigens and tissue damage is years before the disease develops, and they generalised. In organ-specific autoimmune identify symptom free relatives who are at disease one organ only is affected and the risk.' autoantibodies react with its unique auto- antigens. Heart muscle is the only organ affected in Is organ-specific auto'imunity involved DCM and thus to test the autoimmune in human acute ? hypothesis we need to know whether the crite- There is evidence that acute myocarditis is a ria of an organ-specific autoimmune disease continuing autoimmune disease. Many are fulfilled (table 1). Organ-specific autoim- patients have circulating cardiac autoanti- mune diseases occur as a result of genetic pre- bodies (table 2)28 and some have several features of classic organ-specific autoimmu- http://heart.bmj.com/ nity or of insulin dependent diabetes mellitus, Table 1 Features oforgan specific autoimmunity which is an atypical but well established * Middle aged women most frequently affected organ-specific autoimmune disease in which * Familial aggregation viruses may be implicated (table 3). Familial * Organ and disease specific circulating in patients and in unaffected family members aggregation of acute myocarditis has not been * Cell mediated autoimmunity impaired systematically investigated, but acute myo- * Associated autoimmune diseases in the same patient or in family members and DCM have been reported in on September 23, 2021 by guest. Protected copyright. * HLA association different members of the same family.9 Little * Mononuclear ceU.infiltration and abnormal HLA molecule expression in the target organ is known about the associations of acute * Disease induced in animal models after immunisation myocarditis with other autoimmune diseases, with relevant autoantigen except that giant cell myocarditis is associated HLA, human leucocyte . with myasthenia gravis ' and pernicious

Department of Table 2 Circulating autoantibodies in acute myocarditis Cardiological Sciences, St George's % positive Hospital Medical School, London and Antibody type Method AM OCD Healthy controls Reference Institute of Clinical Muscle specific: IFL Medicine, ASA 47* NT 25 2 3 Department of AMLA 41* NT 12 2 3 Cardiology, AFA 28* NT 6 2 3 University of Padua, IFA 32* NT 3 2 3 Padua, Italy Heart-reactive IFL 59* NT 0 4 Anti-laminin ELISA 73* NT 6 5 A L P Caforio Anti-mitochondrial: Correspondence to: M7 ELISA 13* 10 0 6 Dr Alida L P Caforio, ANT SPRIA 91*t 0 0 7 Honorary Lecturer, Anti-P receptor Bioassay 96*t 8 0 8 Department of Cardiological Sciences, St George's *P < 0 05 v controls; tP < 0-05 v controls with other cardiac disease. AFA, anti-fibrillary antibody; AM, acute myocarditis; Hospital Medical School, AMLA, anti-myolemmal antibody; ANT, adenine nucleotide translocator; ASA, anti-sarcolemmal antibody; ELISA, enzyme Cranmer Terrace, London linked immunosorbent assay; IWA, anti-interfibrillary; IFL, indirect immunofluorescence; NT, not tested; OCD, other cardiac dis- SW17 ORE. ease; SPRIA, indirect microsolid-phase radioimmunoassay. Role ofautoimmunity in dilated cardiomyopathy S 31

Table 3 A comparison between acute myocarditis and organ-specific autoimmunity affected mice. Two populations of antibodies Features AM Recent onset IDDM OSAI developed in these mice: one cross-reacted with skeletal and cardiac myosin and the other Br Heart J: first published as 10.1136/hrt.72.6_Suppl.S30 on 1 December 1994. Downloaded from Clinical presentation Acute Acute Chronic Sex M>F M>F F>M was specific for the cardiac isoform.'6 Age (yr) 20-40 1-19 40 The key experiment to prove the autoim- Familial aggregation Yes? Yes Yes Viral aetiology ? ? ? mune nature of the chronic form of myocardi- Organ specific antibody +ve (%) 40-100 80 80 tis was then performed: immunisation of A-J HLA association ? DR3-4 DR3-4-5 Abnormal cell mediated ? Yes Yes normal mice with cardiac myosin alone, with- Association with other AI disease Yes? Yes Yes out virus, induced histologically and immuno- Mononuclear cell infiltration Yes Yes Yes Abnormal HLA expression: logically identical disease.'7 Immunisation Target cell No Yes Yes with skeletal myosin did not produce Endothelium De novo Enhanced Enhanced Animal models Yes Yes Yes myocarditis, showing that myosin-induced myocarditis was a model for an organ-specific AM, acute myocarditis; IDDM, insulin dependent diabetes mellitus; OSAI, organ-specific autoimmunity. autoimmune heart disease. Humoral and cellular transfer experiments in this auto- immune model showed that the myocarditis process was transferable only by T cells and anaemia."I Important autoimmune features, was inhibited by monoclonal antibodies bind- such as the impairment of cell mediated ing to HLA class II molecules which dis- immunity and HLA association have not yet turbed antigen presentation to the T helper been investigated, but Herskowitz et al did not cells."' find abnormal expression of HLA on the tar- These studies showed that the chronic form get cell, the myocyte, in acute myocarditis."2 of murine myocarditis was an autoimmune This abnormal expression is not an absolute disease triggered by a virus in animals with a prerequisite for autoimmune conditions: in predisposing genetic background, rather than multiple sclerosis, for example, HIA class II a persistent viral infection. In addition, the is not expressed on neurons."3 Importantly, finding that autoimmune myocarditis was (see below) animal models of autoimmune produced after stimulation of the immune myocarditis have been produced. system with complete Freund's adjuvant and autoantigen (myosin) indicates not only that viral infection but also that other external Evidence for autoimmunity in pro-inflammatory stimuli could trigger auto- experimental murine myocarditis immunity in genetically predisposed animals A model for autoimmune myocarditis was and humans. found accidentally when workers trying to unravel the mechanisms of myocardial damage in experimental Coxsackie B virus- Humoral autoimmune phenomena in induced murine myocarditis found that only dilated cardiomyopathy certain strains of mice (A-J background) were Clinical and experimental observations indi-

susceptible to acute myocarditis. More cate that myocarditis is a likely candidate for http://heart.bmj.com/ chronic sequelae resembling DCM developed autoimmune pathogenesis. But are organ and in these strains.'4 Early on (7-10 days) in the disease specific circulating cardiac antibodies disease CD8 T cells were cytotoxic for virally detectable in patients with DCM? infected cells. Later (2-3 weeks) they were Several groups have consistently found able to kill cells that were not infected with cardiac antibodies in DCM (table 4),3-8 19-22 viruses, suggesting the generation of auto- but the organ and disease specificities of these immune clones.'4 15 In the late phase of antibody types have not been always evalu-

autoimmune myocarditis no infectious virus ated. For instance, using indirect immuno- on September 23, 2021 by guest. Protected copyright. was detectable (by conventional virus culture fluorescence (IFL) earlier studies identified and isolation methods) in the heart of the antibodies to sarcolemmal and myofibrillar antigens, but these were not strictly cardiac specific because they cross reacted with Table 4 Autoantibodies in dilated cardiomyopathy skeletal muscle.' In addition, it was not clear % Antibody positive whether these muscle specific antibodies were Healthy associated with DCM, because controls with Antibody type Method AM OCD controls Reference other cardiac disease and with Muscle specific: IFL ASA 10 NT 25 3 not caused by DCM were not evaluated. AMLA 9 NT 12 3 When the frequency of cardiac antibodies AFA 24* NT 6 3 IFA 41* NT 3 3 was recently reassessed using IFL on human Organ specific cardiac IFL 26*t 1 3 19 , organ and disease specinfc antibodies Heart reactive IFL 20* NT 0 4 Anti-lamiiin ELISA 78*t NT 6 5 of the IgG class were found in 26% of DCM Anti-mitochondrial patients from England.'9 These antibodies did M7 ELISA 31* 10 0 6 not stain human skeletal muscle and their ANT SPRIA 57*t 0 0 7 Anti-# receptor specificity to the heart was confirmed by Inhibiting LBI 30*t 8 4 20 relevant studies.'9 Interestingly, Inhibiting ELISA 31*t 0 12 21 absorption Stimulating Bioassay 95*t 8 0 8 another group reported that 20% of DCM Anti-a and B AMHC Immunoblot 46*t 8 0 22 patients from the United States had anti- *P < 0 05 v controls; tP < 0 05 v OCD. bodies that gave a diffuse cytoplasmic staining LBI, ligand binding inhibition; MHC, myosin heavy chain. See footnote to table 2 for other pattem on rat heart.4 The cardiac specificity abbreviations. of these antibodies was not investigated by S 32 Caforio

testing on skeletal muscle and by absorption Antibodies against two distinct mitochon- studies, but their IFL pattern on rat heart was drial antigens, the M7 antigen6 and the ade- indistinguishable from that given on human nine nucleotide translocator (ANT),7 have Br Heart J: first published as 10.1136/hrt.72.6_Suppl.S30 on 1 December 1994. Downloaded from heart by the organ-specific cardiac antibodies. also been detected in DCM sera. Mito- This suggests that the antibodies detected by chondrial antigens have generally been classi- IFL on human and on rat heart recognise the fied as non-organ specific,' thus their same organ-specific cardiac autoantigen(s). suggested involvement in a supposedly organ- Two of the autoantigens recognised by these specific autoimmune disease seemed novel. antibodies have already been identified as the a None the less, the heart-specificity of the M7 and /3 heavy chains of myosin; additional and antibodies was shown by absorption studies; as yet unknown antigens are also present.22 ANT antibodies were not tested. Studies The /3 chain is expressed in skeletal and car- should be performed to rule out a potential diac myosin and is therefore not specific to cross reactivity with skeletal muscle because it the heart. The a isoform is expressed solely is not known whether there are two distinct within the myocardium. Antibodies to this translocator isoforms for cardiac and skeletal molecule are truly organ-specific cardiac muscle. autoantibodies.2 Schultheiss et al found that experimentally The identification of a and /B heavy chains induced affinity-purified anti-translocator of myosin as relevant autoantigens in DCM antibodies cross reacted with calcium channel patients parallels what is seen in the experi- complex proteins of rat cardiac myocytes, mental model of autoimmune myocarditis enhanced transmembrane calcium current, DCM.'6--8 But what is the pathogenic rele- and produced calcium dependent cell lysis in vance of these findings? Until recently it was the absence of complement.26 They suggested thought unlikely that any intracellular that these effects may impair the function of autoantigen would be of primary importance the carrier protein, imbalance energy delivery in pathogenesis, but it is now clear that and demand within the cell, and lead to cell endogenous peptides derived from the intra- death. They also speculate that this may be an cellular processing of self cytoplasmic proteins important pathogenic mechanism of auto- are normally present in the cleft of surface immune myocardial damage in DCM and in HLA molecules.23'Thus intracellular autoanti- myocarditis. The presence of this novel mech- gens can be seen by the immune system and anism of antibody dependent cell lysis has under normal conditions tolerance is main- not, however, been shown using the anti- tained. In the presence of appropriate triggers bodies present in patients' sera. Specific con- (for example, viral infection or other inflam- firmation of binding to and a deleterious matory stimuli) and of genetic predisposition, effect on the calcium channel is now required. tolerance could break down and autoimmu- Several groups have demonstrated antibod- nity to these self intracellular autoantigens ies against the /3, adrenoceptor.8 20 21 Limas might ensue. Recent work with T cell and Limas detected these antibodies in hybridomas showed that residential antigen 30-40% of DCM patients by a binding inhibi-

presenting cells (APC) in the normal mouse tion assay.20 Antibody positive DCM sera http://heart.bmj.com/ heart process and express myosin/MHC com- induced sequestration and endocytosis of /3' plexes (MHC is the mouse equivalent of the receptors that were predominantly dependent HLA system). Expression of these myosin on /3 receptor kinase and selectively inhibited complexes was increased by the induction of isoproterenol-sensitive adenylate cyclase autoimmune myocarditis.24 Clearly there is activity.20 local turnover of myosin molecules by the Wallukat et al studied another functional local APCs in the absence of cardiac damage. index of the /,3 receptor antibody. Antibody This strongly accords with myosin being an positive DCM sera accelerated the beating of on September 23, 2021 by guest. Protected copyright. important cardiac autoantigen. The release or neonatal rat heart myocytes in vitro.8 The exposure of antigen may be a vital step in the effect was inhibited by propranolol, bisopro- disease process but only in those individuals lol, and metoprolol. It is well known that that are susceptible to the disease. It may be some patients with DCM benefit from treat- that part of this susceptibility to autoimmune ment with /3 blockade. Wallukat et al sug- myocarditis is conferred both by MHC and gested that this permanent humoral non-MHC genes.'4 stimulation of the 3,' receptor could account Although myosin is probably an important for the accelerated decline in function of the autoantigen, the work on the animal model of failing heart in some patients with DCM. This DCM suggests that it is unlikely that the car- hypothesis offers a rationale for the use of/PI diac myosin antibody is directly responsible antagonists in these patients. for cardiac damage. Passive transfer of highly There is a question mark over the organ concentrated autoantibody in the mouse does specificity of the anti-3,B receptor antibodies. not induce myocarditis in genetically suscepti- /3 Receptors are situated in several different ble recipients, and the disease is clearly T cell tissues and even the supposedly cardiac mediated.25 T cells are also more likely to be specific /3, receptors are found in low concen- the effector arm of the suggested autoimmune tration in the liver and the kidney. There does pathogenesis in patients. This does not under- not appear to be a cardiac specific isoform for mine the possible importance of the antibody the /3 receptor as there is for the a myosin to cardiac myosin as a disease marker or even heavy chain. In addition cross reactivity was or predictive marker, like the islet cell auto- reported between anti-/3, receptor antibodies antibodies in IDDM.I and anti-HIA antibodies.27 If this is con- Role ofautoimmunity in dilated cardiomyopathy S 33

firmed the incidence of the anti-fl1 receptor and assessment of more than one antibody antibodies in DCM will be considerably lower specificity in standard positive and negative and their projected importance much less. samples is likely to clarify this. Br Heart J: first published as 10.1136/hrt.72.6_Suppl.S30 on 1 December 1994. Downloaded from In other organ-specific autoimmune In addition to the presence of the antibody diseases antibodies in most cases are not markers, how many of the criteria for an cytotoxic; so attempts passively to transfer autoimmune disease are fulfilled in DCM? As DCM from human to genetically susceptible with myocarditis most features are present mice with one or more antibody would (table 5). Myocarditis resembles recently provide conclusive evidence for antibody diagnosed IDDM, whereas DCM resembles mediated pathogenesis. long-standing IDDM. These common fea- The argument that ANT, the M7 antigen, tures include: male predominance, HLA-DR4 and fl, receptor are important cardiac association,3' familial aggregation of cases autoantigens would also be strengthened if (20-25% of DCM cases)32, and a relative lack they can be shown to induce myocarditis in of mononuclear cell infiltrates in the target the animal model in the same way as myosin organ. Indeed most patients with DCM come does. to their physician with overt heart failure and a history going back several weeks, months, or even years. It is highly likely that such patients Is organ-specific autoimmunity involved are presenting at a late stage in the natural his- in dilated cardiomyopathy? tory of the disease, when the autoimmune About 30-40% of patients with DCM have damage has already occurred and fibrosis is all organ and disease specific autoantibodies. In that is seen on the cardiac biopsy specimen. these patients the disease may be claimed to Cell-mediated immunity seems to have be autoimmune. But what about the antibody been less extensively investigated than humoral negative patients? There are several concomi- immunity in patients with myocarditis or tant or alternative explanations. Firstly, DCM DCM and the results are controversial.'3-36 may be an aetiologically heterogeneous condi- Examination of T cell subsets has revealed tion; the absence of cardiac antibodies could isolated abnormalities of T helper/suppressor indicate that cell-mediated autoimmune cell ratios33 and of suppressor cell function.34 mechanisms are predominant, or that autoim- Natural killer (NK) cell activity has been munity is not involved in the antibody nega- shown to be reduced in DCM,35 but this tive patients, or both. finding is not cardiac specific. Leucocyte Secondly, because cardiac antibodies are migration inhibition and lymphocyte transfor- early markers of disease that become unde- mation assays did not show significant differ- tectable as the disease progresses, as seen in ences between DCM patients and controls, IDDM, they may not be found in some DCM but some DCM patients did show leucocyte patients.28 There is a report that cardiac migration inhibition with cardiac antigens, antibodies are more common in patients with suggesting that autoimmunity might be better exercise tolerance, as measured by oxy- involved in this subgroup.36

gen consumption during maximal symptom- Most of this work was completed before http://heart.bmj.com/ limited treadmill exercise testing.29 In another animal studies and human humoral work preliminary report, cardiac antibodies were established that the myosin heavy chain,22 the detected in asymptomatic DCM relatives with fl, receptor,8202' and the ANT7 were poten- early cardiac dysfunction.30 These data sup- tially important autoantigens. None of this port the possibility that the absence of cardiac previous cellular work has involved antigen antibodies, at least in some DCM patients, is specific targeting or T cell proliferation stud- related to disease progression. Clinical and ies. None studied patients with early disease,

serological follow up studies of the antibody in whom antigen-specific cellular activation is on September 23, 2021 by guest. Protected copyright. positive patients, and relatives with cardiac more likely to be detected. Thus the central antibodies are needed. Finally, different role of cell-mediated immunity in experimental DCM patients can have antibodies with dif- DCM is now very clear whereas the potential ferent antigen specificities. Collaborative work role of the T lymphocyte in human DCM with exchange of sera by different laboratories warrants further investigation. Autoi-mmunity is involved in causing myocardial damage in about a third of patients with dilated cardio- myopathy. Table 5 A comparison between dilated cardiomyopathy and organ specific autoimmunity ALPC is supported by the Veneto region target project on Longstanding cardiomyopathies (Venice, Italy) and the National Research Features DCM IDDM OSAI Council target project FAT.MA (Rome, Italy). Clinical presentation Chronic Chronic Chronic Sex M>F M>F F>M 1 Bottazzo GF, Tood I, Mirakian R, Belfiore A, Pujol- Age 20-40 20-40 40 Borrell R. Organ-specific autoimmunity. A 1986 Familial aggregation Yes Yes Yes overview. Immunol Reviews 1986;94: 137-69. Viral aetiology 2 Maisch B, Deeg P, Liebau G, Kochsiek K. Diagnostic Organ specific antibody +ve (%) 30 30 80 relevance of humoral and cell-mediated immune re- HLA association DRA-5 DR3-4 DR3-4A5 actions in patients with viral myocarditis. Glin Exp Abnormal cell-mediated immunity? ? Yes Yes Immunol 1983;48:533-45. Association with other AI disease No? Yes Yes 3 Maisch B, Deeg P, Liebau G, Kochsiek K. Diagnostic Mononuclear cell infiltration No No Yes relevance of humoral and cytotoxic immune reactions in Abnormal HLA expression: primary and secondary dilated cardiomyopathy. Am Jf Target cell No Yes Yes Cardiol 1983;52:1071-8. Endothelium De novo Enhanced Enhanced 4 Neumann DA, Burek CL, Baughman KL, Rose NR, Animal models Yes Yes Yes Herskowitz A. Circulating heart-reactive antibodies in patients with myocarditis or cardiomyopathy. _J Am Coil DCM, dilated cardiomyopathy. See footnote to table 1 for other abbreviations. Cardiol 1990;16:839-46. S 34 Caforio

5 Wolff PG, Kuhl U, Schultheiss HP. Laminin distribution Andersson B, Vahlne A, et al. Mapping of a functional and autoantibodies to laminin in dilated cardiomyopathy autoimmune epitope on the fl-adrenergic receptor in and myocarditis. Am HeartJ 1989;117:1303-9. patients with idiopathic dilated cardiomyopathy. J Clin 6 Klein R, Maisch B, Kochsiek K, Berg PA. Demonstration Invest 1990;86: 1658-63. Br Heart J: first published as 10.1136/hrt.72.6_Suppl.S30 on 1 December 1994. Downloaded from of organ-specific antibodies against heart mitochondria 22 Caforio ALP, Grazzini M, Mann JM, Keeling PJ, Bottazzo (anti-M7) in sera from patients with some forms ofheart GF, McKenna Wj, Schiaffino S. Identification of a and disease. Clin Exp Immunol 1984;58:283-92. fl myosin heavy chain isoforms as major autoantigens 7 Schultheiss HP, Kuhl U, Schwimmbeck P, Strauer BE. in dilated cardiomyopathy. Circulation 1992;85: Biomolecular changes in dilated cardiomyopathy. In: 1734-42. Baroldi G, Camerini F, Goodwin JF, eds. Advances in 23 Braciale TJ, Braciale VL. Antigen presentation: structural cardiomyopathies. Berlin: Springer Verlag. 1990;221-34. themes and functional variations. Immunol Today 1991; 8 Wallukat G, Morwinski M, Kowal K, Forster A, Boewer 12:124-9. V, Wollenberger A. Antibodies against the fl-adrenergic 24 Smith SC, Allen PM. Expression of myosin-class II major receptor in human myocarditis and dilated cardiomy- histocompatibility complexes in the normal myocardium opathy: f-adrenergic agonism without desensitization. occurs before the induction of autoimmune myocarditis. EurHearty 1991;12(suppl D):178-81. Proc Nat Acad Sci 1992;89:9131-5. 9 O'Connell J, Fowles RE, Robinson JA, Subramanian R, 25 Neu N, Ploier B, Ofner C. Cardiac myosin induced Henkin RE, Gunnar RM. Clinical and pathologic myocarditis heart antibodies are not involved in the findings of myocarditis in two families with dilated induction of the disease. . Immunol 1990;145:4094-100. cardiomyopathy. Am HeartJ 1983;107:127-35. 26 Schultheiss HP, Ulrich G, Janda I, Kuhl U, Morad M. 10 Burke JS, Medline NM, Katz A. Giant cell myocarditis Antibody mediated enhancement ofcalcium permeability and myositis: associated with thymoma and myasthenia in cardiac myocytes. JExp Med 1988;168:2105-19. gravis. Arch Pathol 1969;88:359-66. 27 Sterin-Borda LA, Cremaschi G, Pascual J, Genaro A, 11 Kloin JE. Pernicious anemia and giant cell myocarditis: Borda E. Ailoimmune IgG binds and modulates cardiac new association. AmJMed 1985;78:355-60. fl-adrenergic receptors. Clin Exp Immunol 1984;58: 12 Herskowitz A, Ahmed-Ansari A, Neumann DA, 223-8. Beschorner WE, Rose NR, Soule WM, et al. Induction 28 Bottazzo GF, Gorsuch AN, Dean BM, Cudworth AG, of major histocompatibility antigens within the Doniach D. Complement-fixing islet-cell antibodies in myocardium of patients with active myocarditis. J Am type I diabetes; possible monitors of active f-cell Coll Cardiol 1990;15:624-32. damage. Lancet 1980;i:1668-72. 13 Wucherpfennig KW, Weiner HL, Hafler DA. T-cell recog- 29 Caforio ALP, Keeling PJ, Bent S, Bottazzo GF, McKenna nition of myelin basic protein. Immunol Today 1991; WJ. Cardiac autoantibodies in dilated cardiomyopathy: 12:277-82. relation to clinical features and exercise capacity [abstr]. 14 Lodge PA, Herzum M, Huber SA. Coxsackie B-3. Acute Circulation 1992;1750. and chronic forms of the disease by different 30 Caforio ALP, Keeling PJ, Zachara E, Mestroni L, iumunopathogenic mechanisms. Am J Pathol 1987;128: Camerini F, Mann JM, Bottazzo GF, McKenna WI. 455-63. Evidence from family studies for autoimmunity in 15 Huber SA, Lodge PA. Coxsackie B-3 myocarditis in dilated cardiomyopathy. Lancet 1994;344:773-7. Balb/c mice. Evidence for autoimmunity to myocyte 31 Anderson JL, Carlquist JF, Lutz JR, DeWitt CW, antigens. AmJYPathol 1984;116:21-30. Hammond EH. HLA A, B, and DR typing in idiopathic 16 Neu N, Beisel KW, Traystman MD, Rose NR, Craig SW. dilated cardiomyopathy: a search for immune response Autoantibodies specific for the cardiac myosin isoform factors. Am I Cardiol 1984;53:1326-30. are found in mice susceptible to Coxsackie B3-induced 32 Michels W, Moll PP, Miller FA, Tajik JA, Chu JS, myocarditis. J Immunol 1987;138:2488-92. Driscoll DJ, et al. The frequency of familial dilated car- 17 Neu N, Rose NR, Beisel KW, Herskowitz A, Gurri-Glass diomyopathy in a series of patients with idiopathic G, Craig SW. Cardiac myosin induces myocarditis in dilated cardiomyopathy. N EnglJ3 Med 1992;326:77-82. genetically predisposed mice. J Immunol 1987;139: 33 Sanderson JE, Koech D, Iha D, Ofiembo H. T lymphocyte 3630-6. subsets in idiopathic dilated cardiomyopathy. Am J 18 Smith SC, Allen PM. Myosin-induced myocarditis is a T Cardiol 1985-55:755-8. cell-mediated disease. J Immunol 199 1;147:2141-7. 34 Eckstein R, Mempel W, Bolte HD. Reduced suppressor 19 Caforio ALP, Bonifacio E, Stewart JT, Neglia D, Parodi cell activity in congestive cardiomyopathy and in 0, Bottazo GF, McKenna WI. Novel organ-specific myocarditis. Circulation 1982;6:1224-9. circulating cardiac autoantibodies in dilated cardio- 35 Anderson JL, Carlquist JF, Hammond EH. Deficient myopathy. JAm Coil Cardiol 1990;15:1527-34. natural killer cell activity in patients with dilated cardio- 20 Limas CJ, Limas C. fl-receptor antibodies and genetics in myopathy. Lancet 1982;ii:1 124-7. dilated cardiomyopathy. Eur HeartY 1991;12(suppl D): 36 Lowry PL, Thompson RA, Littler WA. Cellular immunity 175-7. in congestive cardiomyopathy. Hypersensitivity to 21 Magnusson Y, Marullo S, Hoyer S, Waagstein F, cardiac antigens. Br HeartY 1985;53:400-4. http://heart.bmj.com/ on September 23, 2021 by guest. Protected copyright.