Clinical heterogeneity and prognostic features of South Australian patients with anti-synthetase autoantibodies

Authors


  • Funding: None.

  • Conflict of interest: None.

Abstract

Aim:  To determine the clinical, serological and prognostic features of patients with autoantibodies against three aminoacyl-transfer RNA synthetases (ARS), namely Jo-1 (histidyl-tRNA synthetase), PL-7 (threonyl-tRNA synthetase) and PL-12 (alanyl-tRNA synthetase) in South Australia.

Methods:  Patients with autoantibodies against ARS detected by line immunoassay (anti-Jo1, anti-PL7, anti-PL12) or enzyme-linked immunosorbent assay (anti-Jo1) were identified from existing laboratory databases for the period 1994–2009. Demographic, clinical and serological data were obtained by retrospective review of patients' medical records and laboratory databases.

Results:  Forty-two patients with autoantibodies were identified (anti-Jo1 = 37, anti-PL7 = 4, anti-PL12 = 1). Females were more commonly affected than males (M : F = 12:30). Twenty-one patients had polymyositis (anti-Jo1 = 17, anti-PL7 = 4), seven dermatomyositis (anti-Jo1 = 6, anti-PL12 = 1), 10 overlap syndrome (anti-Jo1 = 10; lupus = 4, scleroderma = 3, Sjögren's syndrome = 2 and rheumatoid arthritis = 2) and four had interstitial lung disease (ILD) only (anti-Jo1 = 4). ILD was present in 69%, polyarthritis in 59% and positive anti-nuclear antibody (ANA) in 43% of patients. Concurrence of autoantibodies against Ro-52 with Jo-1 was seen in 12 patients. The mean follow-up period was 8.3 years (95% CI 5.8–10.8) with 12 deaths. Poor prognostic indicators were age of onset >60 years (P= 0.001), cancer (P= 0.002), negative ANA (P= 0.006) and negative autoantibodies to extractable nuclear antigens (P= 0.02).

Conclusion:  Patients with autoantibodies against ARS present with varied clinical features and occasionally with isolated lung involvement (amyopathic ILD). Older age of onset, malignancy and negative immunologic tests are predictors of poor prognosis. Concurrence of autoantibodies against Jo-1 and Ro-52 may reflect a coupling effect during generation of autoimmunity.

Introduction

Anti-synthetase syndrome (ASS) is characterized by the presence of autoantibodies against aminoacyl-transfer RNA synthetase (ARS) enzymes accompanied by a constellation of organ manifestations, including myositis, synovitis, interstitial lung disease (ILD) and skin rashes.1 The clinical heterogeneity of ASS emphasizs the importance of cross-speciality consultation in the diagnosis and management of this chronic immune-mediated condition. Anti-Jo1 or anti-histidyl-tRNA synthetase is the commonest myositis-specific autoantibody (MSA) found in 20–30% of patients with idiopathic inflammatory myopathy (IIM).1–6 To date autoantibodies against seven other ARS antigens – PL-7 (threonyl), PL-12 (alanyl), EJ (glycyl), OJ (isoleucyl), KS (asparaginyl), Ha (tyrosyl) and Zo (phenylalanyl) – have been described.7–19 Notably, there are minor clinical variations in ASS associated with the distinct synthetases like the presence of milder muscle disease in patients with anti-PL710 and the predominance of ILD in patients with anti-PL12, anti-KS and anti-OJ.12,15,17

However, questions regarding Jo-1 being the commonest antigenic target among ARS,1–6 concurrence of autoantibodies against ARS and Ro-522,6 and the varied clinical presentation of ASS remain unanswered.1 Theavailability of a commercial assay as recently reported20 to detect autoantibodies against Jo-1, PL-7 and PL-12 may lead to an increased identification and recognition of the ASS and facilitate further studies. This may help to uncover the pathogenic mechanisms involved and have implications in terms of management and prognostication. The aims of this study were to determine the clinical, serological and prognostic features of patients with autoantibodies against Jo-1, PL-7 and PL-12 in South Australia.

Materials and methods

Patients with autoantibodies directed against ARS as detected by line immunoassay (LIA (anti-Jo1, anti-PL7 and anti-PL12); Euroline myositis antigen profile assay, Euroimmun AG, Luebeck, Germany) or enzyme-linked immunosorbent assay (ELISA (anti-Jo1); RELISA, Diagnostic solutions) were identified from existing laboratory databases for the period 1994–2009. The Euroline myositis antigen profile LIA detects IgG antibodies against Mi-2, Ku, Pm-Scl, Jo-1, PL-7, PL-12 and Ro-52 and has been used in our laboratory since 2007. Stored serum samples from patients with IIM (myositis registry) were also analysed by the myositis antigen profile LIA and included in the study. Patients were recruited into the myositis registry by identifying positive muscle biopsies carried out at all public hospitals of South Australia, which are reported at a single centre. Demographic, clinical and serological features were obtained by retrospective review of patients' medical records and laboratory databases. Ethics approval was obtained from the Research Ethics Committee, Royal Adelaide Hospital. The diagnosis of autoimmune conditions was confirmed in accordance with the current classification criteria being used.21–25 Myositis was further defined by the presence of symmetrical muscle weakness, elevated creatine kinase (CK), myopathic changes on electromyography and a diagnostic muscle biopsy as definite (three criteria), probable (two criteria) and suspected (one criterion). The presence of distinctive skin changes was needed in addition to the above criteria for the diagnosis of dermatomyositis (DM). The diagnosis of ILD was considered to be present if interstitial infiltrates were seen on chest radiography or high-resolution computed tomography or if the disease was confirmed by histopathology. Poor prognosis was defined as death of the patient. The survival of patients was calculated from the date of diagnosis to the date of death or last follow up. The normal laboratory value for CK was 0–180 U/L, while a positive anti-nuclear antibody (ANA) and rheumatoid factor (RF) were defined as a titre above 1:40 dilution and more than 20 kIU/L respectively. Qualitative nailfold capillaroscopy was carried out by a single observer (PJRT), using a dissecting microscope as described before26 to assess the presence of capillary dilatation, dropouts, distortion or asymmetry, and microvascular haemorrhage in selected patients.

Data were entered on Excel worksheet and analysed by GraphPad Prism 5.0 software (GraphPad Software Inc., San Diego, CA, USA). The survival analysis was carried out by Log-rank analysis of the Kaplan–Meier survival curves and hazard ratios (HR) provided with 95% confidence intervals (CI).

Results

Forty-two patients with autoantibodies against ARS were identified (anti-Jo1 = 37, anti-PL7 = 4, anti-PL12 = 1). Twenty-five of these were from the myositis registry (n= 128). Females were more commonly affected than males (M : F = 12:30); however, the difference in mean age of onset of disease was not statistically significant (M = 56.7 years (95% CI 45–68.4); F = 49 years (95% CI 43.4–54.6); P= 0.35). Fourteen patients had definite polymyositis (anti-Jo1 = 12, anti-PL7 = 2), six had probable polymyositis (anti-Jo1 = 4, anti-PL7 = 2), one suspected polymyositis (anti-Jo1 = 1), five DM (anti-Jo1 = 5), two DM sine myositis (anti-Jo = 1, anti-PL12 = 1), 10 overlap syndrome (anti-Jo1 = 10; lupus = 4, scleroderma = 3, Sjögren's syndrome = 2 and rheumatoid arthritis = 2) and four had ILD only with no muscle weakness or elevated CK (anti-Jo1 = 4). The clinical and serological features (summarized according to the autoantibodies detected: anti-Jo1, anti-PL7 and anti-PL12 in Table 1) observed were as follows: fever (16, 38%), current or ex-smokers (17, 40%), Raynaud's phenomenon (13, 31%), mechanic's hands (9, 21%), muscle weakness (28, 67%), polyarthritis (25, 59%), deforming arthropathy (7, 17%), ILD (29, 69%), positive ANA (18, 43%) and positive RF (12, 29%). Seven patients had history of cancer: melanoma, 2; basal cell carcinoma, 1; colon cancer, 1; non-Hodgkin lymphoma, 1; thyroid cancer, 1; and squamous cell lung cancer, 1. A muscle biopsy was carried out in 23 patients (55%); four patients who had presented with ILD only (amyopathic ILD) did not undergo a muscle biopsy as they had normal muscle power and normal CK levels. Nailfold capillaroscopy abnormalities were detected in nine out of 11 (82%) patients studied. The morphological changes consisted of distortion of the capillary arcades, capillary dilatation, capillary dropouts and the presence of microhaemorrhages.

Table 1. Clinical and serological characteristics according to autoantibody profile in patients with anti-synthetase syndrome
CharacteristicsAnti-Jo1 (n= 37)Anti-PL7 (n= 4)Anti-PL12 (n= 1)Total (n= 42), n (%)
  • Normal laboratory values 0–180 U/L.

  • ‡Titre above 1:40 dilution.

  • §

    Titre above 20 kIU/L. ANA, anti-nuclear antibody; CK, creatine kinase; ILD, interstitial lung disease; RF, rheumatoid factor.

Female254130 (71)
Fever151016 (38)
Raynaud's phenomenon102113 (31)
History of cancer7007 (17)
Mechanics' hands7119 (21)
Muscle weakness253028 (67)
Polyarthritis241025 (59)
Deforming arthropathy7007 (17)
Abnormal nailfold capillaroscopy6219 (21)
ILD262129 (69)
Elevated CK263029 (69)
Positive ANA171018 (43)
Positive RF§92112 (29)
Muscle biopsy carried out193123 (55)
Lung biopsy carried out82010 (24)
Deceased111012 (29)

The concurrence of autoantibodies (summarized in Table 2) against Ro-52 with ARS was detected in 15 patients (15/42, 36%): Ro-52 with Jo-1 in 12 patients (12/37, 32%), Ro-52 with PL-7 in two (2/4, 50%) and Ro-52 with PL-12 in one patient (1/1, 100%) respectively. Autoantibodies against Ro-60 and La were detected mostly in patients with overlap syndrome (8/10 and 4/5 respectively, Table 2), but also in one patient with amyopathic ILD who had no other features of lupus or Sjögren's syndrome.

Table 2. Profile of autoantibodies detected in patients with anti-synthetase syndrome
DiagnosisDefinite PMProbable PMSuspected PMDMAmyopathic ILDOverlap syndromeTotal
  1. DM, dermatomyositis; ILD, interstitial lung disease; PM, polymyositis.

Number1461741042
Anti-Jo11241641037
Anti-PL72200004
Anti-PL120001001
Anti-Ro5252131315
Anti-Mi21000001
Anti-Ku1000012
Anti-PmScl0000011
Anti-Ro6010001810
Anti-La0000145
Anti-u1RNP0000011

The mean follow-up period was 8.3 years (95% CI 5.8–10.8) with 12 deaths (M : F = 3:9, ILD = 11). The cause of death was known in five patients and attributed to progression of ILD in two patients, community-acquired pneumonia in one patient and cancer (metastatic melanoma and squamous cell lung cancer respectively) in two patients. Treatment options used either alone or in combination included: prednisolone (41, 98%), methotrexate (17, 40%), azathioprine (11, 26%), cyclophosphamide (5, 12%), hydroxychloroquine (4, 10%), intravenous immunoglobulin (3, 7%), cyclosporine (2, 5%) and plasmapheresis (1, 2%). Poor prognostic indicators (summarized in Table 3) were age of onset >60 years (P= 0.001, HR 4.96; 95% CI 2.46–43.4) and cancer (P= 0.002, HR 4.47; 95% CI 2.61–61.60). Positive ANA (P= 0.006, HR 0.10; 95% CI 0.05–0.60) and presence of autoantibodies against other extractable nuclear antigens (ENA; P= 0.02, HR 0.22; 95% CI 0.07–0.81) were associated with better survival outcomes. Patients with ILD also had a poor prognosis, but this was statistically not significant (P= 0.06, HR 5.09; 95% CI 0.92–11.87).

Table 3. Prognostic indicators in patients with anti-synthetase syndrome
CharacteristicAlive (n= 30)Dead (n= 12) P-valueHazard ratios (95% CI)
  • *

    Statistically significant difference.

  • †Titre above 1:40 dilution.

  • Normal laboratory values 0–180 U/L. ANA, anti-nuclear antibody; CI, confidence interval; CK, creatine kinase; ENA, extractable nuclear antigen; F, female; ILD, interstitial lung disease; M, male.

Age of onset >60 years570.001*4.96 (2.46–43.4)
History of cancer160.002*4.47 (2.61–61.60)
Positive ANA13110.006*0.10 (0.05–0.60)
Presence of other ENAs1930.02*0.22 (0.07–0.81)
Presence of ILD18110.065.09 (0.92–11.87)
Gender M : F9:213:90.521.4 (0.40–6.02)
Fever970.182.19 (0.67–7.97)
Polyarthritis1960.080.39 (0.08–1.12)
Raynaud's phenomenon850.670.77 (0.22–2.66)
CK >1000 U/L1350.321.73 (0.51–7.85)
Elevated CK (>180 U/L)2270.740.84 (0.20–3.21)

Discussion

We have reviewed the clinical and laboratory features of South Australian patients with autoantibodies against ARS and report an array of interesting findings. Our study confirms the diverse spectrum of clinical manifestations associated with ASS. The prevalence of mechanic's hands (21%), deforming arthropathy (17%), ILD (69%), positive ANA (43%) and positive RF (29%) observed herein is comparable to that previously reported by Schmidt et al. in their review of 231 patients with anti-Jo1-associated ASS.27 However, the frequency of polyarthritis observed in three-fifths, Raynaud's phenomenon in one-third and myositis in two-thirds of our patients is lower compared with published literature, possibly explained by the varied nature of ASS1,27 and population selection bias of our cohort. Notably, more than half of our patients had a negative ANA test, a commonly used screening test for autoimmune conditions, which could provide false reassurance of the absence of autoimmunity and potentially delay the recognition of this syndrome in patients with subtle clinical manifestations. Furthermore, only antibodies against Jo-1(and not other ARS) are routinely sought on routine ENA testing. It could be speculated that earlier recognition, diagnosis and treatment in patients with positive autoantibodies accounts for their survival advantage as observed in our study. The new Euroline myositis antigen profile LIA detects antibodies against Jo-1, PL-7, PL-12, OJ and EJ and is available in our laboratory on specific request. Occasionally, these patients can present with isolated ILD with no muscle or joint symptoms (amyopathic ILD) as reported in our study and a recent publication.28 Consequently, patients with autoantibodies against non-Jo1 ARS may have previously been labelled as idiopathic ILD or inflammatory myopathies. In our experience, qualitative nailfold capillaroscopy assessment,26 which was abnormal in four-fifths of all patients studied (9/11, 82%), can be a useful diagnostic tool in evaluating these patients with subtle clinical presentations.

The co-occurrence of autoantibodies against Ro-52 and ARS as reported earlier2,6,29–31 was observed in one-third of our study patients. The concurrence of these autoantibodies is difficult to explain in the absence of demonstrable molecular cross-reactivity or mimicry between these antigens.29,31 However, it has been proposed that these antigens might be co-expressed in apoptosis or inflamed tissues31 and a coupling effect (pairing of autoantibody generation for Jo-1 and Ro-52 antigens) during generation of autoimmunity may explain their concordance. Jo-1 is the commonest antigenic target among the eight ARS identified so far and multiple autoantibodies against different ARS in a single patient have never been observed.1–6 This was also observed in our study with every patient having autoantibodies to only one ARS (Jo-1, PL-7 or PL-12), but autoantibodies against other ENAs (Ro-52, Ku, Mi-2, Pm-Scl) were present as reported earlier.2,6 This suggests that autoantibodies against ARS do not cross-react with each other and are mutually exclusive.

The Euroline myositis antigen profile LIA is a valuable tool in the detection of autoantibodies in patients with IIM. This assay is easier to carry out than the more demanding immune precipitation techniques. As reported recently, the Euroline myositis antigen profile LIA has a sensitivity and specificity of 45% and 62%, respectively, for the detection of any one autoantibody in a cohort of 153 patients with IIM and 77 disease controls.20 However, it should be understood that this assay has seven different antigens, each of which has its own sensitivity and specificity in IIM. Further studies are needed to validate this commercial assay before it can be recommended for routine laboratory use in the evaluation of patients with autoimmune diseases.

The management of patients with ASS is challenging and requires cross-speciality collaboration as these patients may present with isolated clinical manifestations, recruiting additional organ involvement as the disease progresses. Recent reports have shown promise in the use of rituximab in refractory cases32 and mycopheonlate mofetil in ILD.27 Patients with ASS have a poor prognosis compared with non-ARS-related inflammatory myositis owing to the increased frequency and severity of steroid resistant ILD.1,33 In our study, patients with ILD showed a trend towards poor prognosis (P= 0.06), failing to reach statistical significance because of small numbers. Patients with older age of onset, history of malignancy and negative ANA and ENA were associated with a poor prognosis in our study. Limitations of our study include retrospective assessment, missing data and inability to carry out LIA on all patients to assess for concordance with anti-Ro52 as there were no stored sera for 17 patients.

Conclusion

To our knowledge, we have described the clinical, serological and prognostic profile of the second largest cohort of patients with ASS,1 adding to the limited literature available. Clinicians should meticulously explore for subtle signs and request specific testing for autoantibodies against ARS (if available) when evaluating patients with undifferentiated autoimmune conditions involving the lungs, muscles and joints in isolation or in combination. Inter-speciality collaboration between pulmonologists and rheumatologists through combined clinics and commercially available assays to detect autoantibodies against ARS may lead to a better recognition and management of this syndrome.

Acknowledgement

V. Limaye has been a recipient of a research grant from Royal Adelaide Hospital for setting up the myositis registry.

Ancillary