Clinical characteristics of Vietnamese patients with idiopathic inflammatory myopathies and autoantibodies to aminoacyl‐transfer RNA synthetases

To assess clinical phenotypes of anti‐aminoacyl‐transfer RNA synthetases (aaRS) autoantibodies in Vietnamese patients of Kinh ethnicity with idiopathic inflammatory myopathies (IIM).


| INTRODUC TI ON
Idiopathic inflammatory myopathies (IIM) collectively named myositis, represent a group of rare disorders characterized by dysfunction of skeletal muscle and inflammatory infiltrates in muscle tissue and frequent involvement of extra-muscular organs such as skin, joints, lung, heart and the gastrointestinal tract. Based on different clinical and histopathologic manifestations, patients with myositis are often subclassified into dermatomyositis (DM), polymyositis (PM), inclusion body myositis (IBM) and more recently also into the subset named immune-mediated necrotizing myopathy (IMNM). [1][2][3][4] Autoantibodies are common in IIM, present in up to 80% of patients with PM and DM, and less frequently in IBM. 5,6 Recently a number of new autoantibody specificities have been identified in patients with IIM. They can be classified as myositis associated autoantibodies (MAAs) including anti-Ro52, anti-La, anti-Ku, anti-PM-Scl and anti-U1RNP that can also be found in other autoimmune diseases, and in myositis-specific autoantibodies (MSAs). 7 The MSAs are not only specific for myositis but are also strongly associated with distinct clinical phenotypes. The most frequent MSA, the anti-Jo1 autoantibody, present in up to 30% of Caucasian patients with IIM, is associated with a clinical entity known as antisynthetase syndrome (myositis, interstitial lung disease [ILD], non-erosive arthropathy, mechanic´s hands, fever, and Raynaud´s phenomenon). 8  The different anti-aaRS autoantibodies have been reported with different clinical phenotypes in patients with different ethnicities.
Thus in Caucasian patients myositis is more frequent in anti-Jo1 positive compared to anti-PL-7 or anti-PL-12 positive patients who have a higher frequency of ILD, whereas in Japanese patients anti-PL-7 was more often associated with myositis compared to anti-PL-12. Patient survival was also conditioned by the anti-aminoacyl-transfer RNA synthetases (anti-ARS) specificity, and was significantly lower in patients with anti-PL-7/12 autoantibodies than in anti-Jo1 positive patients. [10][11][12] Thus different anti-aaRS autoantibodies may be associated with different clinical phenotypes in different ethnic populations. In this study we aimed to assess the clinical phenotype of anti-aaRS autoantibodies in Vietnamese patients of Kinh ethnicity with IIM.

| Patients
This is a descriptive study including all patients who were seen at the Rheumatology Department at Bach Mai Hospital, Hanoi, Vietnam, between March 2011 and December 2013. From a cohort of 151 patients of Kinh ethnicity with IIM who were subject to a cross-sectional study previously published, we selected the aaRS autoantibody positive (n = 23) patients for a thorough review regarding clinical and laboratory data. 13 The aaRS antibody positive patients were compared to 69 patients seronegative for MSAs and MAAs and to patients from the 3 largest groups with other MSAs namely anti-signal recognition particle (anti-SRP) (n = 17), antimelanoma differentiation-associated protein 5 (anti-MDA5) (n = 11) or anti-Mi-2 (n = 8) autoantibodies. IIM was defined by experienced rheumatologists as probable or definite PM/DM according to the Bohan and Peter criteria. 14 Antisynthetase syndrome (ASS) was defined as a positive aaRS autoantibody together with at least 1 of the following clinical manifestations: myositis, ILD, fever, Raynaud's phenomenon, arthritis, or mechanic's hands. 17 Pulmonary involvement was systematically investigated at time of study and in some cases at time of diagnosis.
Pulmonary function was tested according to the American Thoracic Society guidelines, using standard equipment. 18 High resolution computed tomography (HRCT) of the lungs without intravenous contrast during end inspiration was performed in all patients. A radiologist who is a specialist of respiratory disease evaluated the HRCT findings in a blinded fashion. ILD was defined as presence of either: (a) chest radiograph abnormalities indicative of fibrosis, and pulmonary function tests; forced expiratory volume in 1 second (FEV 1 ) < 80%, and forced vital capacity (FVC) < 80% (predicted), and total lung capacity (TLC) < 80% and/or >80% predicted FEV 1 / FVC; or (b) abnormal findings on HRCT scan, showing at least 1 of the following features: reticulation and fibrosis, traction bronchiectasis, honeycombing, ground-glass opacification. 19 In patients with symptoms indicating an infection, bronchoscopy and bronchoalveolar lavage was performed to rule out tuberculosis or other infections. In addition, skin test of tuberculosis and extended history for exposure to tuberculosis was performed.  compared to anti-aaRS antibody positive group. *P < .05 compared to anti-aaRS antibody positive group. δ P < .01 compared to antibody negative group. # P <.0005 compared to antibody negative group.
Muscle involvement was defined as abnormalities on electromyography, or in muscle biopsy. Sera from all individuals who gave consent were stored at −80°C until analyses were performed.

| Human leukocyte antigen cell surface receptor (HLA-DR) typing
HLA-DR typing was performed by sequence-specific primer polymerase chain reaction assay (SSP-PCR; DR low-resolution kit; Olerup SSP AB).

| Ethics
This study complies with the Declaration of Helsinki, (as revised in Brazil 2013), and was approved by the local ethics committee at Hanoi Medical University (1720/IRB-HMU) and informed consent has been obtained from the subjects in this study. were considered to be statistically significant. The antisynthetase group was compared to the anti-Mi-2, the anti-SRP, the anti-MDA5 autoantibody positive groups and the seronegative groups respectively. Although correction for multiple testing would be appropriate for our study, we present raw P values, but consider possible type I errors when making the conclusions.   (Table 1). A higher frequency of arthritis compared to the seronegative group was also recorded in the anti-MDA5 positive patients (Table 1).

| RE SULTS
Next, we compared disease activity applying the MDAAT score of patients with anti-aaRS autoantibodies to the seronegative comparator group and to the patients with each of the other myositisspecific antibodies. The anti-aaRS autoantibody positive group had a higher disease activity in the domains skin and pulmonary disease compared to the seronegative group but had lower disease activity in skeletal disease compared to the anti-MDA5-positive patients ( Table 2). The anti-MDA5-positive patients also had a higher disease activity in the cutaneous disease compared to the seronegative group (Table 2). At time of investigation the mean CK level of patients with anti-aaRS autoantibodies was 3019 U/L and for the seronegative group 1565 U/L, but the difference was not statistically significant. Additionally the patients with anti-SRP autoantibodies or anti-Mi-2 had significantly higher serum levels of muscle enzymes compared to the seronegative patients (Table 1). Seventy-four percent of patients with anti-aaRS autoantibodies had raised CRP compared to 42% in the seronegative group.
Finally, we compared the cumulative presence of clinical manifestations between the 3 aaRS autoantibody positive groups, anti-Jo1, anti-EJ and anti-PL7 (Table 3). Pulmonary hypertension was present in 69%, 25% and 50% of the anti-Jo1, anti-PL7 and anti-EJ positive patients respectively and ILD in 46%. 75%, and 50% of the anti-Jo1, anti-PL7, and anti-EJ positive patients respectively. These differences were not statistically significant.
of myositis compared to anti-PL12, anti-KS and anti-OJ. In addition, ILD was found in nearly all Japanese patients with anti-aaRS autoantibodies, a strikingly higher frequency in comparison with our cohort.
Notably, pulmonary arterial hypertension (PAH) was equally frequent as ILD in our patients with anti-aaRS autoantibodies and was more often recorded in patients with anti-Jo-1 autoantibodies compared to patients with other aaRS autoantibodies although the difference was not statistically significant. The frequency of PAH in our cohort is higher compared to other populations where PAH was reported in 7.9%-14.8% of patients with anti-aaRS autoantibodies. 29,30 An explanation for the high frequency observed in our cohort may be because all patients were screened with echocardiography.   Abbreviations: IIM, idiopathic inflammatory myopathy.
*No significant differences were found between the groups.
TA B L E 4 Human leukocyte antigen cell surface receptor (HLA-DR)B1 alleles in Vietnamese patients with anti-aaRS autoantibodies, patients with anti-SRP, anti-Mi2 or anti-MDA5 autoantibodies and seronegative IIM genetic analyses, 5%-10%; Ingrid E. Lundberg was involved in study design, discussing the results and writing the manuscript, 10%-20%.