The present study shows that anti-p140 (18.4%), anti-p155/140 (16.3%), anti-Mi2 (14.3%), and anti-ARS (12.2%) antibodies are common autoantibodies in Korean patients with inflammatory myositis; the common prevalence of anti-ARS and anti-Mi2 antibodies is consistent with our previous result . Anti-p140, anti-p155/140, and anti-Mi2 antibodies were found exclusively in DM patients as previously reported [9–16], and the first two antibodies were found to be associated with distinctive clinical subsets of myositis; anti-p140 antibody with rapidly progressive ILD and anti-p155/140 antibody with cancer-associated myositis. Furthermore, anti-p155/140 antibody was found to be associated with poor survival in cancer-associated DM patients. Anti-ARS antibody was present in both PM and DM patients and found to be associated with stable or slowly progressive ILD.
The most distinguished findings of this study are the high prevalence of anti-p140 antibody in classic DM patients and the association of this antibody with rapidly progressive ILD independent of CADM. Myositis-associated rapidly progressive ILD has been reported in both classic DM and ADM/CADM patients [5–7, 22–25]. Although rapidly progressive ILD often occurs in patients with ADM/CADM [5–7], its true incidence remains unknown due to poor estimates of the proportions of ADM/CADM patients in populations. Previous Japanese studies have reported that anti-p140 antibody is mainly found in CADM rather than classic DM patients [9, 12, 13], but these studies enrolled few classic DM patients with rapidly progressive ILD to conclusively determine whether anti-p140 antibody is a risk factor of CADM or of rapidly progressive ILD. Recently, Sato et al reported that the RNA helicase encoded by melanoma differentiation-associated gene 5 (MDA5) is an autoantigen recognized by anti-p140 antibody (which currently is called "anti-CADM-140 antibody"), and went on to develop an enzyme-linked immunosorbent assay technique using the recombinant protein . According to their results, anti-p140 antibody is exclusively found in CADM patients with or without rapidly progressive ILD; no significant difference was found between the antibody titers of those with and without ILD. This finding suggests that anti-p140 antibody is primarily associated with the CADM phenotype than rapidly progressive ILD in Japanese patients. On the other hand, our study shows that anti-p140 antibody is one of the most common myositis autoantibodies in Korean patients with classic DM, and that it has a striking association with rapidly progressive ILD. Because no patient in the present study met the ADM/CADM criteria , these observations suggest that anti-p140 antibody is primarily associated with rapidly progressive ILD independently of CADM at least in Korean patients.
It has been reported that anti-p155 and/or anti-p155/140 antibodies are associated with cancer-associated myositis [10–12, 14–16], and our results support this contention. In addition, we were able to confirm the previously reported lower incidence of ILD among anti-p155/140 positive patients [10–12, 14, 15]. The poor cancer-related outcome of anti-p155/140 positive patients compared with anti-p155/140 negative patients is a novel finding. However, our study is adopting a small number of patients to conclude with a certainty. We are cautious to claim the prognostic value of this antibody in cancer-associated myositis until it is proven via multivariate analyses that appropriately adjust other prognostic factors. Further studies employing a larger number of cancer-associated myositis patients are warranted to confirm this result and to determine if anti-p155/140 antibody predates or follows the onset of cancer, if its titer correlates with cancer progression, or if certain types of cancers are more prone to develop anti-p155/140 antibody. Another interesting finding was that 3 out of 6 anti-ARS positive patients had cancer-associated myositis, which, however, was not statistically significant. In fact, cancer-associated myositis in the presence of other myositis autoantibodies than anti-p155/140 antibodies (such as anti-ARS or anti-Mi2) has been previously reported [14, 27, 28]. Although anti-p140 positive patients have been shown to have low prevalence of malignancy [9, 12], the presence of other myositis autoantibodies rather than anti-p155/140 antibodies does not seem to rule out the presence of cancer in inflammatory myositis patients.
The limitations of this study are as follows. First, 26 sera were dropped out during consecutive enrollment, which leaves the possibility of selection bias. However, the clinical implications of anti-p140 and anti-p155/140 antibodies observed in this study are unlikely to be affected, because there was only one patient with rapidly progressive ILD and none with malignancy among these 26 patients during 78 ± 59.9 months of follow-up (data not shown). Second, we did not further examine the specificities of myositis autoantibodies beyond immunoprecipitation. Therefore, anti-p140 and anti-p155/140 antibodies detected in the present study may not exactly represent anti-CADM-140 and anti-p155/140 antibodies that have previously been shown to recognize MDA5 [13, 26] and transcriptional intermediary factor 1-γ , respectively. However, high accuracy of immunoprecipitation to specifically detect anti-p140 or anti-p155/140 antibodies in inflammatory myositis patients has been previously reported . In addition, distinctive clinical features have been demonstrated in association with each myositis autoantibody defined by immunoprecipitation in our study, which is generally in parallel with the results of previous studies [9–16].