OUP user menu

Inflammatory myopathies associated with anti-mitochondrial antibodies

Meiko Hashimoto Maeda, Shoji Tsuji, Jun Shimizu
DOI: http://dx.doi.org/10.1093/brain/aws106 1767-1777 First published online: 4 May 2012

Summary

Anti-mitochondrial antibodies, the characteristic markers of primary biliary cirrhosis, have been detected in most patients with this disease. However, the prevalence of these antibodies in inflammatory myopathies and their clinical and histopathological significance has not been determined. Sera from 212 consecutive patients with inflammatory myopathies were screened for anti-mitochondrial antibodies by enzyme-linked immunosorbent assay. The clinical and histopathological features of anti-mitochondrial antibody-positive patients were analysed and statistically compared with those of anti-mitochondrial antibody-negative patients. Twenty-four patients positive for anti-mitochondrial antibodies (seven patients with and 17 patients without primary biliary cirrhosis) were identified (11.3%). Thirteen patients had a clinically chronic disease course of >12 months before their diagnosis at hospitals. Six of these 13 patients (four asymptomatic patients with increased creatine kinase levels and two patients with arrhythmia) had not been aware of muscle weakness, but all 13 patients had muscle atrophy at initial presentation. As complications, eight patients had cardiac involvement including arrhythmias (five patients with supraventricular tachycardia; two with ventricular tachycardia; and one patient with atrioventricular block), six patients had moderately decreased ejection fraction and six patients had decreased vital capacity, two of whom required respiratory support. Regarding muscle histopathological findings, in addition to inflammation, 13 patients had chronic myopathic changes and six had granulomatous lesions. Statistical analysis showed that the clinical features of a chronic disease course, cardiac involvement and muscle atrophy, and the histopathological features of chronic myopathic changes and granulomatous inflammation, were significantly more frequently observed in patients with anti-mitochondrial antibody-positive inflammatory myopathy than in patients who were negative for anti-mitochondrial antibodies. Except for cardiac involvement, which is more frequently observed in patients with primary biliary cirrhosis, no significant differences in clinical or histopathological features were found between patients with or without primary biliary cirrhosis. Our study revealed that inflammatory myopathies associated with anti-mitochondrial antibodies were frequently found in patients with the clinical features of a chronic disease course, muscle atrophy and cardiopulmonary involvement, and the characteristic histopathological feature of granulomatous inflammation. Our study suggests that inflammatory myopathies associated with anti-mitochondrial antibodies form a characteristic subgroup.

  • inflammatory myopathy
  • anti-mitochondrial antibodies
  • primary biliary cirrhosis
  • cardiac involvement
  • granulomatous inflammation

Introduction

Inflammatory myopathies are a heterogeneous group of autoimmune diseases characterized by progressive muscle weakness and skeletal muscle inflammation. Among them, cases of inflammatory myopathy associated with serum autoantibodies exhibit characteristic clinical features (Love et al., 1991) in accordance with the type of autoantibody. Due to the close association between autoantibodies and characteristic clinical features, these autoantibodies are thought to be important not only as markers of subgroups of inflammatory myopathies, but also as factors involved in the mechanism underlying their pathogenesis (Love et al., 1991; Greenberg and Amato, 2004).

Primary biliary cirrhosis (PBC) is a chronic inflammatory autoimmune disease that mainly targets the cholangiocytes of interlobular bile ducts in the liver. Histopathologically, the hallmark of the disease is a loss of biliary epithelial cells and small intrahepatic bile ducts with the portal infiltration of T cells, B cells, macrophages, eosinophils and natural killer cells (Hohenester et al., 2009). The serological hallmark of the disease is the presence of circulating anti-mitochondrial antibodies (AMAs), which are found in 95% of cases with PBC (Van de Water et al., 1988, 1989; Mutimer et al., 1989; Miyakawa et al., 2001). They act against members of the 2-oxoacid dehydrogenase complexes existing in the inner membrane of mitochondria. Among them, the major autoantigen is the E2-subunit of the pyruvate dehydrogenase complex (PDC-E2). The reactivity of AMAs against other 2-oxoacid dehydrogenase complexes, namely, 2-oxo glutarate dehydrogenase (OGDC-E2) and the branched-chain 2-oxo acid dehydrogenase (BCOADC-E2) is also found at a low frequency (Selmi et al., 2011).

It has been reported that AMAs have a specificity of 98% for PBC when analysed with healthy controls (van de Water et al., 1989); however, there are several case reports showing the associations of PBC or AMA positivity with other autoimmune diseases including systemic sclerosis, Sjögren’s syndrome, rheumatoid arthritis (Manthorpe et al., 1979; Berg et al., 1986; Skopouli et al., 1994) and sensory ataxic neuropathy (Charron et al., 1980; Illa et al., 1989; Dahlan et al., 2003; Talwalkar and Lindor, 2003). Thus, the prevalence and the significance of AMAs in autoimmune disease have not yet been studied systematically.

When it comes to the association of PBC with inflammatory myopathies, there have been 23 case reports (Uhl et al., 1974; Benoist et al., 1977; Epstein et al., 1981; Willson, 1981; Tsuchiya et al., 1985; Kumazawa et al., 1987; Saitoh et al., 1988; Ueda et al., 1988; Milosevic and Adams, 1990; Yasuda et al., 1990; Mader et al., 1991; Harada et al., 1992; Varga et al., 1993; Boki and Dourakis, 1995; Simpson and Nickl, 1995; Nakasho et al., 1996; Ono et al., 1996; Tsai et al., 1996; Matsui et al., 2000; Kasuga et al., 2004; Tanaka et al., 2007; Honma et al., 2008), since the first report by Uhl et al. (1974). In these case reports, clinical features including a chronic progressive course (Tsuchiya et al., 1985; Milosevic and Adams, 1990; Harada et al., 1992; Matsui et al., 2000; Kasuga et al., 2004; Tanaka et al., 2007), cardiac involvement (Uhl et al., 1974; Saitoh et al., 1988; Harada et al., 1992; Varga et al., 1993; Tsai et al., 1996; Kasuga et al., 2004; Tanaka et al., 2007), respiratory muscle weakness (Varga et al., 1993; Matsui et al., 2000; Kasuga et al., 2004; Tanaka et al., 2007) and muscle atrophy (Tsuchiya et al., 1985; Kumazawa et al., 1987; Saitoh et al., 1988; Ueda et al., 1988; Varga et al., 1993; Tsai et al., 1996) were described. Although previous reports suggest some characteristic clinical features of inflammatory myopathies associated with PBC, because of the lack of large-scale and systemic clinical and histopathological studies, the prevalence of PBC in inflammatory myopathies is unknown, and the characteristic clinical and histopathological features of inflammatory myopathy-associated PBC have not yet been clarified.

In this study, we retrospectively reviewed 212 patients with inflammatory myopathies and found 24 patients with AMA-positive myositis (11.3%) (seven patients with and 17 patients without PBC). The analysis of clinical and histopathological features revealed that inflammatory myopathies associated with AMAs frequently include patients with a clinically chronic disease course, muscle atrophy, cardiopulmonary involvement and granulomatous inflammation, regardless of the presence or absence of PBC. Our study suggests that inflammatory myopathies associated with AMAs form a characteristic subgroup.

Patients and methods

For the screening of patients with AMA-positive myositis, 212 consecutive patients with inflammatory myopathies referred to our department between November 1999 and April 2009 and whose serum samples were available were included in this study. The diagnosis of inflammatory myopathy was based on the criteria proposed by Bohan and Peter (1975a, b); in addition, one or two muscle biopsy findings, namely, inflammatory changes with necrotic and/or regenerating fibres and major histocompatibility complex (MHC) class I expression on non-necrotic muscle fibres (Bohan and Peter, 1975a, b; Hoogendijk et al., 2004) and exclusion of muscular dystrophy by immunohistochemistry were required. For the exclusion of inclusion body myositis, the criteria proposed by Griggs et al. (1995) were used, and sarcoid myopathy was excluded on the basis of clinical data including roentgen findings of the chest (Iannuzzi et al., 2007) and the titre of the angiotensin I-converting enzyme in serum.

The clinical records of the patients were reviewed to obtain clinical information. The clinical criteria established by the American Association for the Study of Liver Diseases (Lindor et al., 2009) were used for the diagnosis of PBC. The disease duration before diagnosis was defined as the duration between the time of the initial awareness of the symptoms and the time of muscle biopsy for histopathological diagnosis. In the assessment of neurological or laboratory studies, the findings at the time of muscle biopsy were used. Muscle power was evaluated using the Medical Research Council scale.

For the detection of AMAs, serum samples taken before the initiation of therapy and stored at −80°C were used. AMA titre was determined by enzyme-linked immunosorbent assay using a MESACUP-2 Test Mitochondria M2 (AMA-M2) kit (Medical and Biological Laboratories) (Kadokawa et al., 2003). In this method, recombinant PDC-E2, BCOADC-E2 and OGDC-E2 antigens are used as coating antigens, and peroxidase-conjugated anti-human immunoglobulin polyclonal antibody is used as a conjugate antibody to enable the capture the immunoglobulin G, M and A class autoantibodies against AMAs. Briefly, 100 μl of a patient’s diluted serum was added to each well of a microtitre plate precoated with recombinant PDC-E2, BCOADC-E2 and OGDC-E2 antigens, and incubated at room temperature for 60 min to allow anti-PDC-E2, -BCOADC-E2 and -OGDC-E2 antibodies to react with immobilized antigens. After washing, a peroxidase-conjugated goat anti-human immunoglobulin polyclonal antibody was dispensed into each well of the plate and incubated at room temperature for 60 min. Following another washing step, a peroxidase substrate was mixed with a chromogen and incubated at room temperature for 30 min. An acid solution (H2SO4) was then added to each well to terminate the enzyme reaction. The colour development was measured in a microplate reader at a frequency of 450 nm. ‘Calibrator 1’ (serum of 0 index) and ‘Calibrator 2’ (serum of 100 index) were also tested similarly. The index value was calculated using the following formula: (absorbance value of test serum – absorbance value of Calibrator 1)/(absorbance value of Calibrator 2 – absorbance value of Calibrator 1) × 100. Using this method, an index value of >7, which was determined using 168 normal control serum samples, was considered to indicate positivity for the antigens of interest with 90% sensitivity and 98% specificity (Takemura et al., 2001; Kadokawa et al., 2003). Myositis-specific/related autoantibodies including anti-Jo-1, -SRP, -Mi-2, -PL-7 and -PM/Scl100 antibodies were detected by the dot-blot method using recombinant Jo-1, SRP, Mi-2, PL-7 and PM/Scl100 (Diarect AG).

In all cases, biopsied muscle samples were processed for routine histochemistry and immunohistochemistry for the MHC class I, MHC class II, CD4, CD8, CD45, CD68 and C5b-9 antigens (DAKO).

During the analysis of clinical and histopathological features, differences between AMA-positive and -negative patients with myositis were compared. The differences were also analysed between AMA-positive myositis patients with PBC and those without PBC.

For the analysis of the correlation between the index AMA titre and clinical features (disease duration before diagnosis, modified Rankin scale and Medical Research Council scale), a simple linear regression was carried out. A two-tailed Mann–Whitney test was performed for different sets of continuous data of clinical features and a Fisher’s exact test was used to compare categorical data. A P-value of <0.05 was considered significant.

Results

Patients with inflammatory myopathy and anti-mitochondrial antibodies

Twenty-four patients with AMA-positive myositis were identified among 212 consecutive patients with inflammatory myopathies referred to our department. Thus, the prevalence of AMAs in myositis was 11.3%. The clinical and histopathological findings of AMA-positive patients are shown in Table 1. Of the 24 patients, seven, including two with liver histopathological findings consistent with PBC, showed biochemical evidence of cholestasis with an increased alkaline phosphatase level and fulfilled the diagnostic criteria (Lindor et al., 2009) of PBC. Thus, these seven patients were diagnosed as having myositis with PBC and 17 patients were diagnosed as having myositis-associated AMAs without any clinical features of PBC. In the seven patients with PBC, PBC preceded the development of inflammatory myopathies in three patients, whereas PBC and inflammatory myopathies were diagnosed concurrently in four patients. With regard to coexisting diseases other than PBC, seven patients had inflammatory myopathies associated with collagen diseases: (i) two patients with PBC: Sjögren’s syndrome and systemic sclerosis, n = 1; ulcerative colitis, n = 1; (ii) 5 of 17 patients without PBC: systemic sclerosis, n = 2; systemic lupus erythematosus, n = 2; rheumatoid arthritis, n = 1; and (iii) three patients had inflammatory myopathies associated with malignancies: colon cancer, n = 1; tongue cancer, n = 1; stomach and colon cancer, n = 1, two within 1 year of admission and one concurrently.

View this table:
Table 1

Baseline characteristics and treatment response for individual subjects

Patient No.123456789101112131415161718192021222324
Onset age (years)/sex54/F72/F49/F48/M54/F39/F59/M67/F68/M39/M44/F33/F34/F86/M65/F58/F67/F57/M64/F57/M46/F53/M54/M32/F
Disease duration before diagnosis (months)111224242424601122.5335661012172436486060
Initial symptomsMuCMuMuAMuMuCMuCMuMuMuMuMuCMuMuMuMuMuMuAMu
Clinical signs and symptoms
    Weakness (MRC)
    Neck (flexion)NE343444455434533NE5453425
    UE (proximal/distal)4/54/54/53/54/54/53/45/54/54/53/44/44/45/54/44/53/33/45/55/54/53/54/45/5
    LE (proximal/distal)5/54/54/54/54/54/53/34/54/54/53/44/54/54/53/45/53/34/54/54/54/54/44/54/5
    Muscle atrophy(−)(−)(+)(+)(+)NE(+)NENENENE(+)(+)NE(+)NENE(+)(+)NE(+)(+)(+)(+)
    Lordotic posture(−)(−)(−)(−)(+)(−)(+)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(+)(−)
    ArrhythmiaAF(−)Af, PSVTAV blockNSVT (PM)(−)AF (CA)(−)(−)(−)(−)(−)(−)(−)(−)Paf(−)(−)(−)(−)Af(−)CRBBB, PVC (CA)(−)
    Cardiomyopathiesa(+)(−)(−)(+)(+)(−)(+)(−)(−)(−)(−)(−)(−)(−)(−)(+)(−)(−)(−)(−)(−)(−)(+)(−)
    Restrictive ventilatory     impairment (vital     capacity)(−)(−)(+); NIPPV(−)(−)(−)(+) (45.8%); NIPPVNE(−)NENE(−)(−)NE(−)(+) (50.0%)(+) (68.9%)NE(−)(−)(−)(+) (56.0%)(+) (77.5%)NE
    Modified Rankin scale112212311111112232311221
PBC preceding (duration: months)(+) 84(−)(−)(−)(+) 84(+) 10(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)
Diagnosis type of PBCAsymp.Asymp.Asymp.Asymp.Asymp.Symp.Asymp.(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)
Associated disorders
    Collagen diseaseUlcerative colitisSS, SSc(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)SScSLESLESSc(−)(−)RA(−)(−)
    Malignant disease(−)(−)(−)(−)(−)(−)(−)(−)Colon Ca(−)Tongue Ca(−)(−)Stomach Ca, Colon Ca(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)
Laboratory data
    CK level (IU/l)108553348836206191381558232423015132400017704332692117056951471143258184423982809243325
    ESR (mm/h)9649104813671NE2261NE423322342367833919613868NE29
    CRP level (mg/dl)0.50.210.50.490.91.41.62.72.62.30.300.60.050.901.430.060.3<0.30.430.260
    ALP level(U/l)b1540*292*754*515*381*850*309*192*135*151*NENE220*NE253*36*NENE122**NE153*127*179**NE
    AMAs (index)95.5100.3114117.9119.525116.215.87.437.618.190.813.4109.436.78.612.82067.6124.357.827.985.185.3
    Anti-nuclear antibodiesX1280 (sp)RNPX40 (sp)X160 (sp)X40 (sp, homo)dsDNAX80 (spe)X5120 (sp, nucl), RNPX40 (sp)X2560 (sp), Scl70dsDNA, RNPX5120 (sp), Scl70, RNPX320 (sp, nucl)centromereX320 (sp, nucl), dsDNAdsDNA
    Other autoantibodiesSS-A, SS-BRFRFSS-A, RFRFRFRFSS-A, SS-B, RFRFRFRF
    Myositis-specific/related     autoantibodies(−)(−)(−)(−)(−)(−)(−)(−)(−)Jo-1(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)(−)
Liver biopsyCNSDCNENECNSDCNENENENENENENENENENENENENENENENENENENENE
Treatment response
    Treatment(−)(−)NICS(−)NIAZPNICSCSCSCSNINICSCSNINICSNICSCSCSNI
    Follow-up period     (months)9636NI8454NI12NI123312NINI690NINI36NI841212NI
    Modified Rankin scale     change1→11→1NI2→01→1NI3→2NI1→01→11→01→0NINI2→22→2NINI3→1NI1→12→22→2NI
    Worsening of arrhythmia(+)(−)NI(+)(+)NI(−)NI(−)(−)(−)(−)NINI(−)(−)NINI(−)NI(−)(−)(−)NI
    CK level (IU/l) after     treatmentNEw.n.l.NIw.n.l.w.n.l.NIw.n.l.NIw.n.l.w.n.l.w.n.l.w.n.l.NINIw.n.l.w.n.l.NINIw.n.l.NI934w.n.l.w.n.l.NI
Histopathological findings
    Endomysial fibrosis(−)(±)(−)(+)(++)(+)(++)(−)(−)(−)(−)(++)(+)(+)(+)(+)(+)(−)(++)(−)(+)(+)(++)(−)
    Inflammationc(+)e(±)(±)(+++)e(+++)p(+)p(+++)p,e(±)(++)p(−)(+++)p(++)e(+++)p, e(+++)p(+++)p(+++)p(+++)p(+)p(+++)p(+)p(+++)p(+++)p(+++)p(++)p
    MHC class Id(−)(+)f(++)di(+)f(+)f(−)(++)di(+)f(+++)di(+++)di(++)di(+)f(+++)di(++)f(−)(+++)di(+++)di(+++)di(++)di(+++)di(+)f(+++)di(+)f(++)di
    Granulomatous     inflammation(−)(−)(−)(+) CD4(+)(−)(+) CD4(−)(−)(−)(−)(−)(+) CD4(−)(−)(+)(−)(−)(−)(−)(−)(−)(+) CD4(−)
  • a Cardiomyopathy was defined when ejection fraction decreased by <50%.

  • b The normal ranges are 80–260 U/l* and 100–325 U/l**.

  • c Density of inflammatory cells: −, 0; ±, 1 to <5; +, 5 to <10; ++, 10 to <20; +++, 20 to <100; ++++, ≥100.

  • d MHC class I antigen on the sarcolemma: −, 0%; +, <50%; ++, 50 to <75%; +++, 75 to <100%.

  • A = arrhythmia; Af = atrial fibrillation; AF = atrial flutter; ALP = alkaline phosphatase; Asymp. = asymptomatic PBC; AV = atrioventricular; AZP = azathioprine; C = increased serum CK level; Ca = cancer; CA = catheter ablation; CD4 = CD4-positive T cell predominance in granulomatous lesions; CK = creatine kinase; CNSDC = chronic non-suppurative destructive cholangitis; CRBBB = complete right bundle branch block; CRP = C-reactive protein; CS = corticosteroids; di = diffuse; dsDNA = anti-double-stranded DNA antibody; e = endomysium; ESR = erythrocyte sedimentation rate; f = focal; F = female; homo = homogeneous staining pattern; LE = lower extremities; M = male; MRC = Medical Research Council; Mu = muscle symptoms (weakness or atrophy); NE = not examined; NI = no information; NIPPV = non-invasive positive-pressure ventilation; NSVT = non-sustained ventricular tachycardia; nucl = nucleolar staining pattern; p = perimysium; Paf = paroxysmal atrial fibrillation; PM = pacemaker; PSVT = paroxysmal supraventricular tachycardia; PVC = premature ventricular contraction; RA = rheumatoid arthritis; RF = rheumatoid factor; RNP = anti-RNP antibody; Scl70 = anti-Sci70 antibody; SLE = systemic lupus erythematosus; sp = speckled staining pattern; SS = Sjögren syndrome; SS-A = anti-SS-A antibody; SS-B = anti-SS-B antibody; SSc = systemic sclerosis; Symp. = symptomatic PBC; UE = upper extremities; w.n.l. = within normal limit.

Of the 24 myositis patients with AMA, nine were male and 15 were female. The average age at disease onset was 54 years (range 32–86 years).The average disease duration before diagnosis was 20 months (range 1–60 months) and the disease duration was >12 months in 13 patients (six patients with PBC and seven patients without PBC). The initial symptoms of the 24 patients were muscle weakness or atrophy in 18 patients, arrhythmia in two patients, and no subjective muscle symptoms except for an increased serum creatine kinase level found by chance during medical checkups in four patients.

Regarding neurological findings, all 24 patients, except for one with diffuse muscle weakness (Patient 17), showed muscle weakness with proximal dominance. Thirteen patients showed muscle atrophy (four patients with PBC and nine patients without PBC) and three patients with 24 or 60 months of disease duration showed lordotic posture (Patients 5, 7 and 23). Regarding laboratory findings, serum creatine kinase levels were elevated in all patients ranging from 232 to 15132 IU/l (2322 ± 3121 I/U).

The associated autoantibodies were detected in 10 of 24 patients with myositis with AMAs: (i) in six of seven patients with PBC: anti-nuclear antibody, n = 4; rheumatoid factor, n = 3; anti-Sjögren’s syndrome-A antibody, n = 2; anti-Sjögren’s syndrome-B antibody, n = 1; (ii) in 14 of 17 patients without PBC: anti-nuclear antibody, n = 12; rheumatoid factor, n = 7; anti-Sjögren’s syndrome-A antibody, n = 1; anti-Sjögren’s syndrome-B antibody, n = 1.

Muscle CT images were assessed in 13 patients, and 10 patients showed muscle atrophy with proximal dominance, and three patients (Patients 5, 7 and 23) with lordotic posture showed atrophy with fatty changes in paravertebral muscles (Fig. 1).

Figure 1

Muscle CT images of AMA-positive patients with the symptoms of weakness in neck flexion and lordotic posture. Paravertebral muscles at the upper thoracic level showed muscle atrophy and fatty changes in Patient 5 (A), Patient 7 (B) and Patient 23 (C).

Regarding cardiac involvements, eight patients (five patients with PBC and three patients without PBC) showed arrhythmias (supraventricular tachycardia, n = 5; ventricular tachycardia, n = 2; atrioventricular block, n = 1), and six patients (four patients with PBC and two patients without PBC) showed a decreased ejection fraction (<50%). In the patients with arrhythmias, two (Patients 7 and 23) were treated by catheter ablation. With regard to respiratory involvements, six patients (two patients with PBC and four patients without PBC) had a vital capacity of <80%. Among them, two patients (Patients 3 and 7) required respiratory support.

Detailed information about the responses to treatments was available in 15 patients (observation period, 35.0 ± 34.8 months; range 3–96 months). Among them, 11 patients were treated with corticosteroids, one patient was treated with azathioprine and three patients refused treatment. In our series, bile acid therapy had been performed in only one patient (Patient 1, three years before the diagnosis of myositis). Thus, it was difficult to know whether bile acid therapy had some association with any change in muscle pathology/function.

Of the 12 patients treated, all except one patient, who was treated with low-dose corticosteroids in accordance with her wishes, showed normalization of creatine kinase levels within 3 months. Improvement of muscle power was observed in six patients and it remained the same in the others. One patient, whose index of AMAs was followed up after initiation of treatment, showed a mild decrease in index, from 85.1 to 21.0 in Patient 23, but two patients did not show a decrease.

In three non-treated patients (Patients 1, 2 and 5), despite no worsening of muscle weakness during the observation period (62.0 ± 30.8; range 36–96 months), two patients (Patients 1 and 5) developed arrhythmia, one of whom (Patient 5) required an implantable pacemaker.

Histopathological features of patients with inflammatory myopathy that have anti-mitochondrial antibodies

With regard to histopathological studies, in addition to the findings of inflammatory myopathies (necrotic and/or regenerating fibres, n = 23; inflammatory changes, n = 23; positive staining of the sarcolemma with MHC class I, n = 21; variation of muscle fibre size, n = 18), endomysial fibrosis, which suggests a chronic myopathic process, was found in 15 patients. No perifascicular atrophy was observed, and the invasion of non-necrotic muscle fibres by mononuclear cells was observed in one patient (Patient 22). Intriguingly, six patients (three patients with PBC and three patients without PBC) had granulomatous inflammatory lesions and four of them showed CD4-positive T cell predominance over CD8-positive T cells in the lesions (Fig. 2).

Figure 2

Muscle histopathology of AMA-positive patients with granulomatous inflammation (Patients 4, 5 and 7). Granulomatous inflammatory changes in the endomysial space are replacing muscle fibres in Patient 4 (A), Patient 5 (B) and Patient 7 (C). Marked fibre variation in size and increased volume of connective tissue in endomysium are also observed in Patient 7 (D). Surface marker analysis of infiltrating lymphocytes shows the predominance of CD4-positive lymphocytes compared with CD8-positive lymphocytes in Patient 7 (E and F). Scale bar = 100 μm.

Comparison of clinical and histopathological features between anti-mitochondrial antibody-positive and anti-mitochondrial antibody-negative patients

Table 2 shows a comparison of clinical and histopathological features between AMA-positive and AMA-negative patients, and between AMA-positive patients with myositis with and without PBC.

View this table:
Table 2

Comparison of clinical and histopathological features between AMA-positive patients versus AMA-negative patients with myositis and AMA-positive patients with PBC versus AMA-positive patients without PBC

Clinical and histopathological findingsAMA-positive myositis patients (N = 24)AMA-negative myositis patients (N = 188)Significance (AMA-positive versus AMA-negative)AMA-positive myositis patients with PBC (N = 7)AMA-positive myositis patients without PBC (N = 17)Significance (PBC+ versus PBC)
Sex (M:F)9:1558:130NS5:24:13NS
    Age at disease onset (years)54 ± 1355 ± 15NS54 ± 1054 ± 15NS
    Disease duration before diagnosis (months)20 ± 2012 ± 36P < 0.000526 ± 1717 ± 21NS
Clinical signs and symptoms, n/N (%)
    Severe limb muscle weakness (≤MRC3)7/24 (29)55/177 (32)NS2/7 (29)5/17 (29)NS
    Severe neck muscle weakness (≤MRC3)7/22 (32)51/137 (37)NS2/6 (33)5/16 (31)NS
    Myalgia7/23 (30)95/163 (58)P < 0.052/6 (33)5/17 (29)NS
    Muscle atrophy13/24 (54)40/188 (21)P < 0.0054/7 (57)9/17 (53)NS
    Skin rash2/24 (8)60/188 (32)P < 0.050/7 (0)2/17 (12)NS
    Cardiac involvementa8/24 (33)17/188 (9)P < 0.0055/7 (71)3/17 (18)P < 0.05
    Restrictive ventilatory impairmentb6/19 (32)41/126 (33)NS2/7 (29)4/12 (33)NS
    Dysphagia4/21 (19)30/121 (25)NS2/7 (29)2/14 (14)NS
Associated disorders, n/N (%)
    Collagen disease7/24 (29)43/188 (23)NS2/7 (29)5/17 (29)NS
    Malignancies3/24 (13)28/188 (15)NS0/7 (0)3/17 (18)NS
    Interstitial lung disease6/24 (25)75/188 (40)NS1/7 (14)5/17 (29)NS
Laboratory data
    CK level (U/l)2322 ± 31213160 ± 8627NS1183 ± 11262791 ± 3567NS
    ESR (mm/h)51 ± 2641 ± 30P < 0.0573 ± 2643 ± 20P < 0.05
    CRP level (mg/dl)0.8 ± 0.81.5 ± 4.1NS0.7 ± 0.40.8 ± 0.9NS
    AMAs (index)62.8 ± 43.22.4 ± 1.5P < 0.000198.3 ± 33.648.2 ± 38.5P < 0.01
Histopathological findings, n/N (%)
    Variation in muscle fibre size18/24 (75)75/188 (40)P < 0.0057/7 (100)11/17 (65)NS
    Disruption of myofibrillar architecture6/24 (25)72/188 (39)NS1/7 (14)5/17 (29)NS
    Internal nucleic9/24 (38)59/188 (32)NS5/7 (71)4/17 (24)NS
    Necrotic and/or regenerating fibres23/24 (96)168/188 (89)NS7/7 (100)16/17 (94)NS
    Endomysial fibrosis15/24 (63)46/188 (25)P < 0.015/7 (71)10/17 (59)NS
    Perifascicular atrophy0/24 (0)20/188 (11)NS0/7 (0)0/17 (0)
    Inflammatory cell infiltrationd23/24 (96)173/188 (92)NS7/7 (100)16/17 (94)NS
    Mononuclear cells invading non-necrotic muscle fibres1/24 (4)8/188 (4)NS0/7 (0)1/17 (6)NS
    Positive staining of the sarcolemma with MHC class Ι21/24 (88)174/188 (93)NS5/7 (71)16/17 (94)NS
    Granulomatous inflammation6/24 (25)11/188 (6)P < 0.013/7 (43)3/17 (18)NS
    CD4-positive lymphocyte predominance4/24 (17)10/139 (7)NS2/7 (29)2/17 (12)NS
  • a Cardiac involvement was defined as a condition when ejection fraction was decreased by <50% or presence of arrhythmia.

  • b Restrictive ventilatory involvement was defined as a condition when the vital capacity was <80%.

  • c Positive finding of internal nuclei was defined as a finding of >5% of muscle fibres associated with internal nuclei.

  • d Positive inflammatory change was defined as a change when more than five inflammatory cells infiltrated the perivascular or endomysium in the muscle tissues.

  • CK = creatine kinase; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; F = female; M = male; MRC = Medical Research Council; NS = not significant.

Regarding clinical features, the AMA-positive patients with myositis had a longer disease duration before diagnosis (P < 0.0005), more frequently showed muscle atrophy at the initial presentation (P < 0.005), showed cardiac involvement (P < 0.005) and less frequently showed skin rash (P < 0.05) than the AMA-negative patients. Regarding histopathological findings, the AMA-positive patients with myopathy more frequently showed variation in muscle fibre size (P < 0.005), endomysial fibrosis (P < 0.01) and granulomatous inflammation (P < 0.01) than the AMA-negative patients. In the AMA-positive patients with myositis, no significant difference was found between the patients with PBC and those without PBC, except that cardiac involvement was more frequently observed in patients with PBC.

Titre of anti-mitochondrial antibodies and clinical features

The indices of the 24 AMA-positive patients ranged from 7.4 to 124.3 (62.8 ± 43.2 index). In 13 patients with a disease duration of 12 months or longer, the index increased significantly (82.0 ± 38.6) in comparison with that of the 11 patients with a disease duration of <12 months (40.5 ± 38.8; P < 0.05).

A comparison of clinical and histopathological features between the patients who have high (>80) and low (≤80) indices of AMAs showed that the patients with high indices tend to have cardiomyopathies, arrhythmias and granulomatous inflammation more frequently, but differences between the two groups were not significant. However, five out of six patients with cardiomyopathies, six out of eight patients with arrhythmia and four out of six patients with the histopathological finding of granulomatous inflammation had high indices of AMAs.

In the correlation between the index of AMAs and clinical features, although the disease duration before diagnosis significantly correlated with the index of AMAs (P < 0.05), disease severity (modified Rankin and Medical Research Council scales) did not (Fig. 3). There was no correlation between the index of AMAs and severity or activity of PBC.

Figure 3

Correlation between titre of AMAs and disease duration before diagnosis. The index of AMA-M2 significantly correlated with the disease duration before diagnosis (P < 0.05).

Discussion

The association of PBC with inflammatory myopathies has been reported mainly as case reports, and comprehensive studies of the prevalence and the clinical and histopathological features of patients with inflammatory myopathies and AMAs or PBC have not been conducted thus far. In this study, we retrospectively reviewed 212 patients with inflammatory myopathies, and found 24 patients with AMAs (11.3%) and seven patients with PBC (3.3%). By analysing the clinical and histopathological features, we found that AMA-positive patients with inflammatory myopathies frequently include patients with a clinically chronic disease course, muscle atrophy, cardiac involvement and granulomatous inflammation regardless of the presence or absence of PBC. Although the comparative analysis did not show a significant difference, in our 24 patients with AMAs, two patients required respiratory support because of restrictive ventilator impairment, which is unusual for typical cases of inflammatory myopathies.

Considering the characteristic clinical features and the significant correlation between the index of AMAs and the disease duration before diagnosis, we believe that these suggest the importance of AMAs not only as markers but also as factors involved in pathogenic mechanisms. Of the 24 AMA-positive patients, despite the characteristic features of the whole group, some patients showed a subacute or acute clinical course, no muscle atrophy, and no association with cardiopulmonary involvements, indicating some variation in clinical features. Further study should be carried out to clarify the roles of AMAs in the pathogenesis of AMA-positive myositis.

Presently, the mechanisms underlying PBC pathogenesis remain unknown. In the initiation of the mechanism underlying PBC pathogenesis, the mimotopes of the vulnerable epitope of the PDC-E2 autoantigen are considered to become the autoantigens in PBC (Lleo et al., 2008). In the production of the mimotopes, the modification of native proteins following the exposure to infectious microorganisms, environmental xenobiotics/chemical compounds or apoptotic biliary epithelial cells has been suggested (Selmi et al., 2011). In the presence of altered regulation of self-tolerance and genetic background, it is suggested that AMAs are produced by PDC-E2 autoantigen-specific B cells and PDC-E2 autoantigen-specific T cells present in sera (Selmi et al., 2011). Regarding the pathogenesis of liver damage in PBC, it has been revealed that the autoreactive CD4-positive and CD8-positive T cells infiltrating the liver in PBC recognize PDC-E2 (Hohenester et al., 2009), which supports the hypothesis that T cell responses contribute to bile duct injury in PBC.

On the other hand, although AMAs are highly specific for PBC, the pathogenic role for them in this disease is uncertain since in contrast to other autoimmune diseases, PBC responds poorly to immunosuppressive agents and changes in autoantibody titre do not seem to correlate with disease severity (Selmi et al., 2011).

Recently, it has been suggested that the pathogenic immune attack in PBC may be directed not only against the proteins of the 2-oxo acid dehydrogenase family (M2 antigen) but also against other antigens that become exposed during apoptosis and proliferation of biliary epithelial cells. Among these antigens, nuclear antigens, neuroendocrine compartments such as the acetylcholine receptor muscarinic M3 receptor, the α1 adrenergic receptor and proteins of the Bcl-2 family have been suggested (Berg, 2011).

Indeed, it has been known that antinuclear antibodies are detected in ∼ 50% of serum samples from patients with PBC (Selmi et al., 2011). Among them, an autoantibody against glycoprotein 201 was reported to correlate with severity of PBC.

Furthermore it was recently shown that sera from patients with PBC have functionally active anti-acetylcholine receptor muscarinic M3 autoantibodies and the author suggested the possibility of receptor desensitization due to repeated interactions of the receptor with the autoantibodies (Berg, 2011). Nicotinic but not muscarinic anti-acetylcholine receptor autoantibodies were detected in AMA-positive patients with PBC in early studies (Sundewall et al., 1987; Kyriatsoulis et al., 1988). Considering a previous report describing the relationships between autoimmunity against the β1-adrenergic receptor autoantibody or muscarinic acetylcholine receptor and dilated cardiomyopathy (Jahns et al., 2004), in addition to autoreactive CD4-positive and CD8-positive T cells, autoantibodies against antigens other than AMAs may have some pathogenic role in patients with PBC. Further studies are necessary to reveal the mechanism of skeletal muscle damage in patients with PBC.

In our series, eight patients showed myositis associated with arrhythmia. Among them, two patients required treatment by catheter ablation and one patient required an implantable pacemaker. Furthermore, in three patients who refused any treatment, two developed arrhythmia, one of whom required an implantable pacemaker.

In previous case reports, as far as we searched, cardiac involvement in myositis in the presence of AMAs has been described in eight patients (Uhl et al., 1974; Saitoh et al., 1988; Harada et al., 1992; Varga et al., 1993; Tsai et al., 1996; Kasuga et al., 2004; Tanaka et al., 2007). Among them, five patients had a chronic disease course (2 years, n = 2; 5 years, n = 3), and all eight patients, including one patient who required an implantable pacemaker, showed arrhythmia (supraventricular tachycardia, n = 5; ventricular tachycardia, n = 2; atrioventricular block, n = 1). Seven patients, including three patients who received treatment for dilated cardiomyopathies, had a decreased ejection fraction. Interestingly, six patients with cardiac involvement and PBC have been diagnosed as having asymptomatic PBC.

From the view-point of cardiovascular system functions, abnormal autonomic nervous system regulation in patients with PBC was reported (Selmi et al., 2011), and other reports show that autonomic dysfunction in PBC is associated with an increased cardiac mortality risk in non-liver chronic disease states (Neubauer et al., 1997; Jackson et al., 2007). Furthermore, a recent study showed impaired cardiovascular function in PBC using impedance cardiography and magnetic resonance methodologies (Jones et al., 2010). Considering the above previous reports, in addition to our present results, it is possible that the frequent association of cardiac complications is a characteristic feature in inflammatory myopathies with PBC. Further study should be carried out to clarify the association between arrhythmia and AMAs, and myositis patients with AMA should be followed up carefully for cardiac complications, especially arrhythmia.

From the diagnostic view-point of myopathies, it should be noted that the 10 patients in our series showed a clinically chronic course with muscle atrophy, the findings of which are common to those of muscular dystrophy. In patients with muscle atrophy, three had paraspinal muscle involvement with lordotic posture, which is an atypical feature of inflammatory myopathies. Since myositis patients with AMA respond to treatment, as a diagnostic approach to chronic myopathies, AMAs should be evaluated particularly in patients with chronic myopathies and muscle atrophy associated with or without lordortic posture, or cardiopulmonary involvement. Moreover, the patients presenting PBC associated with AMAs with increased serum creatine kinase levels should be evaluated if they show muscle involvement or cardiomyopathy.

It has been reported that granulomatous inflammation with bile duct injury is a characteristic liver histopathological change in PBC (Ludwig et al., 1978). Thus, it is interesting that six of the patients with a clinically chronic disease course showed granulomatous inflammation in muscle histopathology. In association with PBC, granulomatous extrahepatic lesions have been described in other organs such as the skin (Kishor et al., 2008) and lungs (Fagan et al., 1983). The presence of granulomatous inflammation in the patients with such characteristic clinical features also suggests that a pathogenic mechanism may be related to that of PBC. Further study should be carried out to clarify the exact background mechanism underlying this association.

AMAs in serum are highly sensitive and specific for PBC; they are detected in nearly 95% of patients with PBC, with specificity close to 100% when tested with recombinant antigens (Selmi et al., 2011). In our study, we used an enzyme-linked immunosorbent assay for the detection of immunoglobulin G, M or A class antibodies against at least one of the 2-OADC enzymes. When performed in accordance with the manufacturer’s protocol (Kadokawa et al., 2003), the specificity of the detection method is 98%. In our series, seven patients were diagnosed as having inflammatory myopathies with PBC and 17 patients were diagnosed as having inflammatory myopathies with AMAs without any clinical features of PBC. In comparison of the two groups, associated autoimmune diseases and autoantibodies were observed in both groups, and no significant difference in clinical features was found between the groups except for cardiac involvement being more frequently observed in patients with PBC than in those without PBC (Table 2).

Although AMAs serve as highly sensitive markers for the diagnosis of PBC, AMAs can frequently be detected in patients with other diseases, such as primary systemic sclerosis, Sjögren’s syndrome, rheumatoid arthritis and autoimmune hepatitis (Hu et al., 2010). It has also been reported that other autoantibodies associated with PBC are rheumatoid factors (70%), anti-smooth muscle antibodies (66%), anti-thyroid (anti-microsomal, anti-thyroglobulin) antibodies (40%) and anti-nuclear antibodies (35–50%) (Talwalkar and Lindor, 2003; Selmi et al., 2011). It has also been reported that AMA-positive individuals, even those without signs of cholestasis or liver inflammation, are very likely to develop PBC (Metcalf et al., 1996; Hohenester et al., 2009). It is not clear whether the AMA-positive patients without PBC in our series will develop PBC several years later. However, considering that inflammatory myopathies associated with AMAs are frequently observed in patients with a clinically chronic disease course, muscle atrophy, cardiopulmonary involvement and granulomatous inflammation regardless of the presence or absence of PBC, we believe that inflammatory myopathies associated with AMAs form a characteristic subgroup.

Further study, including the study about the roles of autoantibodies against antigens other than AMAs in PBC, should be carried out to clarify the exact background mechanism underlying this association.

Funding

The Health and Labour Sciences Research Grant on Intractable Diseases from the Ministry of Health, Labour and Welfare of Japan; Intramural Research Grant (23-5) for Neurological and Psychiatric Disorders of National Center of Neurology and Psychiatry (NCNP).

Abbreviations
AMA
anti-mitochondrial antibody
BCOADC
branched-chain 2-oxo acid dehydrogenase
E2
E2-subunit
OGDC
2-oxo glutarate dehydrogenase
PBC
primary biliary cirrhosis
PDC
pyruvate dehydrogenase complex

References

View Abstract