Muscle Pathological Features and Extra-Muscle Involvement in Idiopathic Inammatory Myopathies With Anti-Mitochondrial Antibody

Backgroud Anti-mitochondrial antibodies (AMAs) can be detected in some idiopathic inammatory myopathy (IIM) patients. We aimed to investigate the clinical features of IIM patients with AMAs. Methods We retrospectively analysed consecutive 1,167 patients with IIM for AMAs-associated myositis and compared them to age- and sex-matched AMA-negative patients.


Background
Idiopathic in ammatory myopathy (IIM) is an autoimmune systemic disease that is characterised by skeletal muscle injury and multi-system involvement, including the skin, lung, heart, and oesophagus. Many autoantibodies, including myositis-speci c antibodies (MSAs) and myositis-related antibodies (MRAs), can be detected in the serum [1]. MSAs are closely related to distinct clinical characteristics [2,3]. Although MRAs have been detected in other connective tissue diseases or overlap syndrome of IIM and other connective tissue diseases, it has been reported that some MRAs were associated with unique features in IIM. Anti-Ku antibody isolated from the serum in IIM was associated with concomitant severe interstitial lung disease (ILD) and immune-mediated necrotising myopathy [4]. Serum anti-Ro-52 antibody in dermatomyositis (DM) could develop rapidly progressive ILD, which mainly presents with organising pneumonia on high-resolution computed tomography [5,6].
Anti-mitochondrial antibodies (AMAs) is biomarkers of primary biliary cholangitis (PBC). A recent study noted that AMAs can be detected in some IIM patients [7]. Previous studies have suggested that IIM patients with AMAs frequently have cardiac involvement [8][9][10][11]. However, most studies were case reports or smallsample studies, and studies with larger cohorts are needed.
To better understand the clinical features of AMAs-associated myositis, we retrospectively reviewed 29 patients with AMAs from 1,167 IIM patients in our centre. Baseline characteristics, clinical features, electrocardiography, echocardiography, laboratory, histopathological ndings and outcomes were reviewed in this study.

Study population
Clinical data were retrospectively analysed from consecutive 1,167 patients with IIM in the Department of Rheumatology at China-Japan Friendship Hospital from January 2008 to December 2019. The classi cation criterion of IIM was based on the 2017 European League against Rheumatism/American College of Rheumatology Classi cation Criteria for Adult Idiopathic In ammatory Myopathies [12]. The diagnosis of PBC can be established when two of the following three criteria are met [13]: (1) biochemical evidence of cholestasis based on alkaline phosphatase elevation; (2) presence of AMAs or other primary biliary cholangitis-speci c autoantibodies, including sp100 or gp210, if AMAs are negative; and (3) histologic evidence of non-suppurative destructive cholangitis and destruction of interlobular bile ducts. To study the clinical features of idiopathic in ammatory myopathy patients with AMAs, we randomly selected 116 idiopathic in ammatory myopathy patients with comparable age and sex as disease controls from the 1,167 idiopathic in ammatory myopathy patient populations. The study protocol was approved by the Ethics Committee of the China-Japan Friendship Hospital, and written informed consent was obtained from each participant. The study was conducted according to the Declaration of Helsinki 2000.

Clinical characteristics
We gathered detailed clinical records, including sex, age at onset, mode of onset, disease course (recorded monthly), muscle strength, rash (heliotrope rash, Gottron rash, periungual erythema, and "mechanic's hand"), ILD, and oesophageal lesions. Onset times less than 2 weeks and more than 3 months were de ned as acute onset and chronic onset, respectively; sub-acute onset was between these limits. Muscle strength was measured using the manual muscle test (MMT8) proposed by the International Myositis Outcome Assessment Collaborative Study (http://www.niehs.nih.gov/research/resources/ imacs/diseaseactivity/index.cfm). High-resolution lung computed tomography was performed to investigate radiographic abnormalities consistent with ILD. ILD was de ned as the existence of one of the following abnormalities: parenchymal micronodules and nodules, linear opacities, irregularity of the interfaces between the peripheral pleura and aerated lung parenchyma, ground-glass opacities, honeycombing, and traction bronchiectases or bronchiolectases. Dysphagia, nasal or gastroesophageal regurgitation, and aspiration pneumonia were identi ed as oesophageal lesions, and further examinations were conducted to evaluate oesophageal function, including oesophageal manometry, barium-swallow examination, or endoscopic examination, if necessary. Cardiac involvement was evaluated by electrocardiography (ECG) and echocardiography (UCG). Some patients underwent cardiac magnetic resonance imaging (CMRI). Serum maximum creatine kinase (CK) levels and IgG/IgA/IgM were recorded. Furthermore, the positive samples were tested using a commercially available ELISA assay (EUROIMMUN). Patients whose sera were positive based on both detections above were included in the study.

Muscle biopsy
Serial frozen sections were processed for routine haematoxylin and eosin and immunohistochemistry for anti-major histocompatibility complex class , CD4, CD8, CD45, CD68, and anti-membrane attack complex.

Follow-up study
Patients who had one or more of three abnormal ECG, UCG, and CMRI ndings were classi ed as patients with cardiac involvement. The follow-up period started from the initial treatment to death or the last investigation. Disease activities were evaluated by the physician's global assessment based on the investigator's composite assessment of disease activity on the constitutional, cutaneous, skeletal, gastrointestinal, pulmonary, and cardiac scales of the Myositis Disease Activity Assessment Tool (MDAAT). The assessments were rated using 10-cm visual analogue scale (VAS) scores (0-10) with higher scores indicating severe disease activity. Disease activity in each subject was estimated by the same physician who was blinded to other clinical information. A VAS score of ≤1 was de ned as disease remission.

Statistical analysis
The analysis was performed using PASW statistics version 18 (IBM, Armonk, NY). Data are expressed as means ± standard deviation or median (interquartile range). Comparisons between groups were performed using the independent-samples t-test for continuous variables and Fisher's exact test for categorical variables. The Mann-Whitney U test was applied to data with non-normal distributions. Kaplan-Meier survival analyses were performed in a follow-up study. Two-sided p-values of <0.05 were considered statistically signi cant.

AMAs prevalence in idiopathic in ammatory myopathy patients
AMAs were detected in 29 (2.5%) out of 1,167 patients with IIMs from the China-Japan Friendship Hospital between January 2008 and December 2019. Of 29 patients with AMAs, 14 patients had polymyositis (PM), 11 patients had dermatomyositis (DM), and 4 patients had amyopathic dermatomyositis. Eight patients were diagnosed with PBC, and 2 patients had associated systemic sclerosis and primary Sjögren's syndrome, respectively.

Clinical characteristics of IIM patients with AMAs
The demographic and clinical characteristics of the 29 IIM patients (7 men, 22 women) with AMAs are shown in Table 1. The age of the patients ranged from 25-83 years, with a mean of 51.0 ± 12.7 years. The median disease duration was 6 (2, 48) months. Eight patients had a sub-acute onset, while the remaining patients had a chronic onset. Sixteen patients had cutaneous manifestations including heliotrope rash, Gottron Papules and Gottron Sign, V-Neck Sign, Shawl Sign, periungual changes, Raynaud's phenomenon, mechanic's hands, and calcinosis cutis. Dysphagia was noted in 8 patients, and 23 patients had cardiac involvement. ILD was present in 16 patient. Twenty-ve patients presented with muscle weakness, and the mean MMT8 score was 68.9 ± 8.9. The median maximum serum CK level was 1,650 (667.5, 4616.75) IU/L. Ten patients were positive following testing with MSAs. These subjects included 3 patients who tested positive for anti-MDA5, 1 patient for anti-TIFγ 1 patient for anti-NXP2, 1 patient for anti-SRP, and 4 patients for anti-synthetase (anti-Jo-1, anti-PL-7, and anti-PL12). In the spectrum of antinuclear antibodies, anti-Ro52 was detected in 13 patients. Two and three patients tested positive for anti-SSa and anti-dsDNA, respectively. In addition, one patient was anti-PM-SCL10 antibody-positive and the other was anti-gp210 antibody-positive.
By comparing IIM patients with AMAs and disease controls, we found that the incidence of cardiac involvement and maximum serum CK were signi cantly higher in patients with AMAs than in disease controls (DC) (p<0.001 and p=0.026, respectively; Table 1).

Pathological features of the skeletal muscle
Muscle pathological results were available for 23 AMAs-positive patients, and were classi ed as 12 cases of immune-mediated necrotising myopathy (IMNM), 1 of DM, 1 of PM characteristics, 6 of non-speci c myositis and 3 normal/minimal lesions. As for 49 histopathological ndings in DC group, there were 15 IMNM, 10DM, 13 non-speci c myositis and 11 normal/minimal lesions. There were no signi cant differences in muscle pathological features between AMAs-positive group and DC group (Table   1).Signi cantly, comparison in AMAs-positive IIM patients show that 11 of 16 patients (68.7%) with isolated anti-AMAs presented IMNM pathological features. In contrast, only 1 of 7 patients (14.3%) with coexistence of anti-AMAs and SRP antibody showed pathological features of IMNM. The incidence of IMNM in patients with isolated anti-AMAs was signi cantly higher than that of patients with coexistence of anti-AMAs and MSA (11/16 vs.1/7, p=0.026) ( Table 2).Typical pathologic DM, pathologic PM, non-speci c myositis and normal pathologic performance in patients with anti-AMAs were shown in Fig 1.

Cardiac involvement in AMAs-positive patients
Abnormal auxiliary heart examinations were detected in 21 AMA-positive patients. Only 4 patients (patient 9, 13, 27, and 28) had palpitation. No patients present clinical symptom of heart failure. Therefore, most cases of cardiac involvement were subclinical. As shown in Table 3, 58.6% (17/29) and 59.3% (16/27) patients were identi ed with abnormal ECG and UCG ndings, respectively. Among abnormal ECG ndings, ventricular premature beats (8/29), ST-T changes (7/29), and atrial brillation (4/29) were the most common. Regarding abnormal UCG, there were 6 cases with left atrial enlargement and 10 cases with valve regurgitation. Pericardial effusions were detected in 7 patients, and most of them (6/7) were small. In addition, CMRI was performed in 4 patients. Late gadolinium-enhanced images in the left ventricular lateral wall were found in 3 cases. It is worth noting that patient No. 21 had an abnormal result on CMRI, although his ECG and UCG were normal. Thus, CMRI helped to identify subclinical cardiac involvement.

Treatment and follow-up study
All patients received initial prednisone therapy (1 mg/kg), and the mean initial dose was 50.8± 11.6 mg. Twenty-ve patients had additional immunotherapies including methotrexate, cyclophosphamide, cyclosporine, tacrolimus, mycophenolate mofetil, and azathioprine. Twenty-four patients were followed up for 6-84 (mean 26.9 ± 20.7) months. Two patients (No. 11 and 16) died within 3 years and 7 months of diagnosis, respectively. Patient No. 16 died of aspiration pneumonia, and the other died for unclear reasons.
Both deaths were complicated by cardiac involvement ( Table 3). The remaining 22 patients had good recovery, and the mean maintenance prednisone dose was 8.3 ± 6.2 mg at follow-up. There were 7 patients without heart involvement and 15 patients had heart involvement. Time to disease remission (VAS ≤1) and prednisone tapering less than 10 mg were compared between patients with cardiac involvement and those without cardiac involvement. The details of the combined immunotherapies are shown in Supplementary Table 1. There were no signi cant differences in the initial VAS score [5 (4, 6) and 5 (4, 5), P=0.399] and prednisone doses (47.3 ± 10.3 and 45.7±7.9 mg, P=0.718) between patients with cardiac involvement and patients without cardiac involvement. As shown in Fig. 3, patients without heart abnormalities took signi cantly less time for disease remission and prednisone tapering off than patients with heart abnormalities (P<0.001 and P=0.001, respectively).

Discussion
IIM is a heterogeneous disease, and one or more autoantibodies can be detected in serum. The recognition of MSA allows the understanding of the distinct clinical phenotypes of IIM. Meanwhile, MRA is an autoantibody that can be detected in IIM and other autoimmune diseases. Previous studies indicated that some MRAs, such as anti-Ku antibody and anti-Ro52 antibody, identi ed a distinct in ammatory myopathy phenotype.
AMAs, as biomarkers of PBC, have been reported in several case reports and cohorts in IIM [14][15][16][17]. Therefore, AMAs are identi ed as MRAs. To our knowledge, the current study is the largest cohort of AMAspositive IIM patients to date. The prevalence of AMAs was 2.5% in our IIM cohort, which differs from that of previous studies (0.006-11.5%) [14]. Similar to the previous results, the average age was 51 years, and the majority were women.
We did not nd a signi cant difference in rash, ILD or oesophageal lesions between AMAs-positive and AMAs-negative patients. Although the maximum serum CK level in the AMAs-positive group was signi cantly increased compared with the disease controls, the MMT8 score did not show a statistically signi cant difference.
Skeletal muscle pathology is crucial for diagnosis and disease evaluation. Previous study showed that AMAs were not associated with a speci c subgroup which may present as DM, PM and IMNM. The study by Maeda MH suggested that histopathological ndings more frequently showed variation in muscle bre size, endomysial brosis and granulomatous in ammation in AMAs-positive patients. In our cohort, 12/23(52.2%) cases showed IMNM-like pathological features. However, we did not found statistic differences in muscle pathological features of AMAs-positive IIM patients. One possible reason is small numbers of cases. Interestingly, IMNM was a distinguishing feature in isolated anti-AMAs compared to coexistence of anti-AMAs and MSA among AMAs-positive patients. As for patients with coexistence of anti-AMAs and MSA, pathological characteristics probably match with respective MSA.
The study by Albayda et al. suggested that cardiomyopathy and arrhythmia were present in 71.4% (5/7) of AMAs-positive patients [9]. Meiko et al reported that 33.3% (5/24) of AMAs-positive patients showed arrhythmias, and 25% (6/24) showed decreased ejection fraction [14]. In our cohort, results also showed a close relationship between AMAs in IIM and cardiac abnormalities. We found that 72.4% of AMAs-positive patients had at least one abnormal ECG, UCG, or CMRI nding, and the incidence was signi cantly higher than that of AMA-negative patients. However, only 4 patients had clinical symptoms, which indicated that most heart involvements were subclinical. Speci cally, heart involvement included arrhythmia, atrioventricular enlargement, abnormal heart failure, valve disorder, and pericardium.Furthermore, the frequency of abnormal ECG and UCG was signi cantly higher in AMAs-positive patients than in AMAsnegative patients.As AMAs are biomarkers of PBC, we investigated the effect of PBC on clinical features in AMAs-positive patients. We found that abnormal UCG was more common in patients without PBC than in patients with PBC. However, the study of Meiko et al showed that cardiac involvement was more likely to be present in patients with PBC [14]. The cause of the discrepancy may be due to different cohorts or small sample sizes.
A follow-up study indicated that most AMAs-positive patients improved with prednisone and immunosuppressant. Two deaths with abnormal ECG and UCG occurred in our cohort. In addition, we found that patients without heart abnormalities took less time to achieve disease remission and prednisone tapering below 10 mg.
This study has some limitations. First, the diagnosis of arrhythmia was based on ECG and not on 24-hour dynamic ECG (Holter), which can provide better information about heart rhythm. Second, only 4 patients underwent CMRI in our cohort, CMRI is helpful for detecting subclinical heart disease. Third, most data on prognosis were obtained by telephone follow-up survey; the results of auxiliary examinations such as ECG and UCG results were incomplete. Therefore, recovery from heart abnormality was not clari ed in this study.

Conclusions
IMNM was a major histopathological nding in patients with isolated anti-AMAs antibody. AMAs were signi cantly associated with cardiac involvement in IIM, and the cardiac involvement was mostly subclinical. PBC seemed to be a protective factor for abnormal UCG in AMAs-positive patients. A follow-up study indicated that disease activities may be less in patients without cardiac involvement. This work was supported by Science and Technology Commission Foundation of Beijing Z191100006619012 .
Authors' contributions LZ and XL contributed to the study conception, study design, data acquisition, data analysis, data interpretation, and drafting and revising the manuscript. HBY and YHX collected data and conducted the assessment of disease activity in the follow-up study. JPL conducted the random sampling of disease controls and participated in statistical work. QLP and GCW assisted in data interpretation and revised the manuscript. All authors read and approved the nal manuscript. Tables   Table 1 is not available with this version.