Statistics from Altmetric.com
We read with interest the article by Fathi et al.1 We agree with the main message indicating that interstitial lung disease is a common early clinical manifestation in patients with polymyositis (PM) and dermatomyositis (DM). The prevalence in their study was much higher than commonly reported. However, for the sake of objective information, some important methodological insufficiencies and disputable data in the article need to be highlighted.
The immunoserological profile of the patients with myositis studied seemed to be peculiar in light of previous reported data. The very unusual serological finding was the high frequency of anti-La antibodies. To our knowledge, it was in only one multicentre study by Brouwer et al,2 that anti-La antibodies were detected in 6% of patients with PM and 3% with DM, using an enzyme linked immunosorbent assay (ELISA) and western blot analysis. Otherwise, neither in earlier studies3,4 nor in more recent ones,5,6 has any anti-La activity been reported despite several different techniques used for the detection of antibodies against intracellular antigens.
Further, the authors reported positive anti-Ro antibodies in 44% of the patients with PM and in 25% of the patients with DM, which differs considerably from the data on the prevalence of anti-Ro in patients with pure idiopathic myositis described by others. Those studies, using counterimmunoelectrophoresis or an immunodiffusion technique, found antibodies against native Ro antigens in 0–14% of the examined sera,3–6 while relevant ELISAs showed antibodies to recombinant Ro60 in 0–6%5,6 of the tested sera, and with a higher frequency of 17–70%6–8 to Ro52, along with anti-Jo1 antibodies, constituting the so called antisynthetase syndrome.7,8 The prevalence of autoantibodies against La and Ro antigens in patients with PM and DM in the study by Fathi et al1 was thus not in agreement with the majority of published data and should have been discussed more extensively.
Unfortunately, the authors did not present any details of the antibody detection methods, which are of the utmost significance. Owing to many different to the detection of autoantibodies against intracellular antigens, several options are available, ranging from “in-house” methods to commercial kits. The choice may be crucial in considering those patients/cases where an autoantibody profile can substantially influence the final clinical decision. It has been suggested that in such situations two independent techniques should be employed to confirm the presence of specific autoantibodies in the patient serum,9 which is especially applicable to the family of antibodies against Ro antigens.
We thank Dr Rozman and coauthors for their interest in our study presented in the Annals and their thoughtful reflections on the autoantibody profile in patients with polymyositis and dermatomyositis.1 As Dr Rozman et al comment, we found that anti-SSA/Ro antibodies were detected in 35% (6/17) and anti-SSB/La antibodies in 12% (2/17) of our patients with poly- or dermatomyositis. We agree that these numbers are higher than expected in comparison with most previously published myositis cohorts.2,3 Notably, we had excluded patients with other defined rheumatic diseases.
The analyses for anti-SSA/Ro and anti-SSB/La antibody detection that were used in our study were performed with an enzyme linked immunosorbent assay (ELISA) containing both Ro52 and Ro60 recombinant antigen. These autoantibody tests were performed at the time of the diagnostic check up for each person as part of the routine procedure at the Karolinska University Hospital, Solna, Sweden. The laboratory procedures were performed at the Department of Clinical Immunology, which is ISO/IEEC certified for autoimmune analyses.
To further confirm the anti-SSA/Ro positive results we have retested the anti-SSA and anti-La positive sera with two other assays—immunodiffusion from Immunoconcepts and line immunoassay from Innogenetics. We could confirm the anti-SSA/Ro positivity from the ELISA tests with immunodiffusion in three of the six cases. The three additional cases were negative for anti-SSA using immunodiffusion, but were positive for Ro52 with the immunoblot technique. None of the six cases was positive for Ro60. Two cases had anti-La antibodies, both in addition to anti-SSA antibodies, using the ELISA test. The anti-La positivity was in one case confirmed by immunodiffusion, the other was not.
The discrepancy between the results in our patients and previously presented data is likely to depend on the different methods that have been used in different studies.2 The patient cohort presented in our study is small, which makes analyses from stratified patient groups into subgroups such as polymyositis or dermatomyositis uncertain. Both ELISA and the line immunoassay are generally more sensitive than immunodiffusion and counterimmunoelectrophoresis.4 Moreover, the sensitivity threshold of various immunodiffusion tests varies with the manufacturer’s content of specific antigens, like the Ro52 antigen. The assay used for immunodiffusion is, in our hands, less likely to detect Ro52 positive sera than other assays and, consequently, likely to oversee anti-SSA autoantibody reactivity if used alone. On the other hand, the higher sensitivity of ELISA and the line immunoassay may be at the cost of a weaker clinical specificity for the detection of anti-SSA antibodies associated with disease.
We still believe that our data are valid as similar results were achieved by both the ELISA and the line immunoassay using recombinant Ro52 antigen from different sources and even confirmed by immunodiffusion in three of the six cases. Another possible explanation for the lower frequency of anti-SSA antibodies in patients with polymyositis in older studies might be that patients with inclusion body myositis were not excluded, as autoantibodies seem to be less common in this disease entity.
Based upon the validating laboratory procedures presented above, we believe that anti-SSA/Ro and anti-SSB/La antibodies may be more common in pure polymyositis and dermatomyositis than previously recognised. This hypothesis remains to be tested in a larger cohort of patients with myositis, particularly with regard to Ro52 alone.