Introduction

Idiopathic inflammatory myopathies (IIMs) are a set of chronic muscle-inflammation-related disorders that have conventionally been divided into three 3 subtypes: inclusion body myositis, dermatomyositis, and polymyositis. The three categories classify IIMs largely based on the level of skin and muscle engagement; however, they inadequately address the range of clinical symptoms [1]. Autoantibodies are characteristic of IIMs and are classified as myositis-specific or myositis-associated autoantibodies (MAAs). MAAs can coexist with other autoantibodies and are detected in various connective tissue disorders while MSAs are specific to individuals with myositis and rarely coexist with other autoantibodies [2].

MSAs have a significant correlation with clinical characteristics and may be used to classify phenotypic subgroups. Clinical pictures and autoantibody profiles in IIM patients have been determined in studies involving Caucasian, Chinese, Japanese, and Indian populations [2,3,4,5]. However, there were conflicting conclusions in these previous studies because of the different ethnicities and sample sizes. Consequently, we sought to better characterize MSA profiles in a comprehensive Chinese cohort of IIM individuals and to examine their probable connections with clinical parameters, to provide useful data for clinical outcomes.

Materials and methods

Study cohort and patients

From June 2017 to November 2021, we retrospectively evaluated the medical data of 515 IIM adult patients hospitalized in the Second affiliated hospital of Zhejiang University School of Medicine. All hospitalized individuals’ blood samples were evaluated for MSAs/MAAs by commercial line blot assay EUROLINE Autoimmune Inflammatory Myopathies 16 Ag (IgG), and the results were noted in full electronic medical records. The Bohan and Peter criteria were used to diagnose both probable and definite cases of IIM [6]. Clinically amyopathic dermatomyositis fulfilled the Sontheimer criteria [7]. Patients with a positive serological test to ARS antibodies, the presence of ILD defined by the high-resolution computed tomography, and/or probable or definite myositis according to Bohan and Peter criteria were defined as ASS. Our institute’s ethical committee granted permission (approval number: 2020–225). Given the retrospective nature of the investigation, informed consent was waived.

Clinical data

A systematic review of the IIM patients’ medical records was used to obtain demographic and clinical data. Clinical manifestations included cutaneous involvement, interstitial lung disease (ILD), myalgia, muscle weakness, dysphagia, and cancer. Cutaneous features of interest were Heliotrope rash, V sign, shawl sign, mechanic’s hands, Gottron’s sign, periungual erythema, and calcinosis. The diagnosis of ILD was made following high-resolution computed tomography results. Rapidly progressive ILD (RP-ILD) was defined as progressive hypoxemia and dyspnea within 3 months of the beginning of respiratory symptoms [8]. Concomitant malignancies were defined as malignancies that occurred within the time period starting 3 years prior to diagnosis and ending with the final follow-up.

Laboratory data included lactate dehydrogenase, aspartate transaminase, creatine kinase (CK), ferritin, C-reactive protein, lymphocyte subsets, serum alanine transaminase, and frequency of anti-Ro52 antibody. All clinical, laboratory, and radiology details recorded at the time of first evaluation were retrieved from the database.

Statistical analysis

Continuous data was summarized using medians and ranges, whereas counts and percentages represented categorical variables. The chi-square or Fisher’s exact test was employed to compare categorical variables and an independent-sample t-test was used to compare continuous variables with a normal distribution. The Mann–Whitney U test was employed to compare nonparametric continuous variables. We used univariate logistic regression to determine the correlations among MSAs and clinical factors. The Kaplan–Meier technique was utilized and employed to conduct the survival analysis, and the log-rank test was used to determine differences across subgroups. We used logistic regression to study independent risk variables for a poor prognosis. P > 0.05 was considered statistically significant. The Kaplan–Meier survival analysis was performed using GraphPad Prism 6.0 software, whereas all other statistical analyses were conducted by SPSS statistical software (IBM SPSS Version 25).

Results

MSA prevalence in IIM patients

Among the 515 IIM patients enrolled in the study cohort between June 2017 and November 2021, 366 (80.6%) were female and 149 (19.4%) were male. The median disease time was 3 months (range: 0.3–480 months). MSAs were found in 88.2% of the IIM patients. The most prevalent MSAs were anti-MDA5 (n = 131; 25.4%), anti-Jo-1 (n = 78; 15.1%), and anti-EJ (n = 49; 9.5%). Meanwhile, 285 patients (55.3%) showed anti-Ro52 antibodies. In addition, 17 patients carried two kinds of MSAs in their sera. Table 1 shows the distribution of the MSAs.

Table 1 Myositis specific autoantibodies (MSAs) and myositis-associated autoantibodies profile of the 515 patients with idiopathic inflammatory myopathies
Full size table

Comparison of clinical characteristics in IIM patients with and without MSAs

Regarding demographic characteristics, there are no significant differences in sex, age at onset, or disease time among IIM patients with and without MSAs. For the clinical characteristics, MSA-negative individuals are more prone to exhibit muscle weakness (P = 0.031) compared to MSA-positive patients. MSA-positive patients had a higher frequency of anti-Ro52 antibody positivity (P = 0.033) and higher levels of ferritin (P = 0.013) and CRP (P = 0.032) in comparison to MSA-negative patients (Table 2).

Table 2 Clinical and laboratory features of patients with MSAs and without MSAs
Full size table

Association of MSAs with clinical features

We employed linear regression models to study the correlations among MSA subgroups and clinical characteristics in individuals who tested positive for a single autoantibody. Table 3 summarizes the clinical characteristics related to the various MSAs. Anti-tRNA synthetase antibodies (ASAs) were the usual MSAs. Anti-Jo-1 was present in 15.1% of patients (n = 78) and non-anti-Jo-1 ASAs (anti-PL-12, anti-PL-7, anti-EJ, anti-OJ) were communally observed in 25% of patients (n = 129). Due to the limitations of the commercial kits, other anti-aminoacyl tRNA synthetases, such as anti-KS, anti-Zo, and anti-Ha, were not detected. Anti-Jo-1 was strongly correlated with ILD, arthritis, Raynaud phenomenon, decreased lymphocyte counts, and the presence of anti-Ro52 antibody. However, anti-Jo-1 was negatively correlated with raised ferritin and cutaneous features, including Gottron’s sign, V sign, shawl sign, heliotrope rash, and periungual erythema. Non-anti-Jo-1 ASAs were significantly correlated similarly to anti-Jo-1.

Table 3 Clinical associations of myositis specific autoantibodies
Full size table

Anti-MDA5 was present in 131 patients (25.4%) and had strong associations with rash, ILD, RP-ILD, elevated ferritin, and decreased lymphocyte counts. In accordance with expectation, anti-MDA5 was negatively correlated with elevated CK (OR = 0.24; 95% CI: 0.14–0.39; P < 0.0001).

Anti-TIF1-γ was present in 43 patients (8.3%) and was strongly linked with rash. These patients typically presented with Heliotrope rash, V sign, and Shawl sign, and muscle weakness and had a higher risk of cancer (OR = 3.51; 95% CI: 1.59–7.71; P = 0.002). However, anti-TIF1-γ was negatively associated with ILD (OR = 0 0.03; 95% CI: 0.01–0.09; P < 0.0001) and less prone to present with mechanic’s hands (OR = 0.31; 95% CI: 0.12–0.80; P = 0.02).

Anti-SRP was present in 26 patients (5.0%) and was correlated with significantly higher median CK level (OR = 5.34; 95% CI: 1.95–14.64; P = 0.001). Patients with anti-SRP had decreased risks of rash and ILD (P < 0.01 or P < 0.05).

Anti-Mi-2 was present in 30 patients (5.8%) and individuals with anti-Mi-2 had a lower risk of ILD (OR = 0.17; 95% CI: 0.07–0.42; P < 0.0001).

Anti-NXP2 was present in 27 patients (5.2%). Anti-NXP2 was strongly associated with V sign, dysphagia, myalgia, and elevated CK, but negatively associated with ILD (OR = 0.18; 95% CI: 0.08–0.42; P < 0.0001).

Anti-SAE1 was present in 7 patients (1.4%). Given the small sample size, no statistical analyses were performed.

Survival analysis

The median follow-up duration was 8.75 months (range: 1–48 months). During analysis, 55 of the 515 patients (10.7%) had died. Patients with anti-MDA5 had the highest mortality within the first year after onset (Fig. 1A and 1B). Comparisons of the Kaplan–Meier curves between the different MSA subgroups using the log-rank test revealed significant differences (P < 0.0001; Fig. 1A). Patients with anti-MDA5 had a poor prognosis, especially those with a double-positive for anti-MDA5 and anti-Ro52 (anti-MDA5 alone vs. combined anti-MDA5 and anti-Ro52: P = 0.03; Fig. 1C). Patients with ASAs had a better prognosis than patients without ASAs (P = 0.002; Fig. 1D); nonetheless, longevity among those who tested positive for MSA and those who tested negative for MSA showed no significant difference (data were not shown).

Fig. 1
figure 1

Kaplan–Meier survival curves for the patients who survived or died. Kaplan-Meier survival curves and log-rank test p-values generated for the patients based on MSA status. A Survival rates stratified according to the MSA status; B survival rates stratified according to anti-MDA5 antibody status; C survival rates in anti-MDA5 antibody positive IIM patients with and without anti-Ro52 antibody; D survival rates in IIM patients with and without anti-ARS antibody. Anti-ARS, anti-aminoacyl-transfer ribonucleic acid synthetase; anti-MDA5, anti-melanoma differentiation-associated gene 5; anti-NXP2, anti-nuclear matrix protein 2; anti-SAE, anti-small ubiquitin-like modifier activating enzyme; anti-Mi2:anti-SRP, anti-signal recognition particle; anti-TIF1-γ, anti-transcriptional intermediary factor 1- γ; MSA, myositis-specific autoantibody; IIM, idiopathic inflammatory myopathies

Full size image

Prognostic factors

A multivariable Cox regression analysis was carried out to recognize indicators for poor outcomes in patients with IIMs. Univariate analyses revealed that older age, male sex, heliotrope rash, shawl sign, V sign, Gottron’s sign, RP-ILD, anti-MDA5, and elevated serum ferritin were significantly linked with poor outcomes in IIM patients (P < 0.05 or P < 0.0001). These nine risk factors were added in a multivariate analysis to recognize prognostic factors. As shown in Table 4, male, old age, RP-ILD, and elevated serum ferritin were recognized as independent risk factors for the worst prognosis in IIMs patients. MSAs were not identified as predictive factors for the long-term prognosis (all P > 0.05).

Table 4 Multivariate Cox regression analysis of prognostic factors in patients with IIM
Full size table

Discussion

MSAs were detected in the majority (88.2%) of IIM patients in the present study, different from previous findings in Chinese and Caucasian cohorts [2, 9]. This may be related to the cohort size and different ethnicities. Additionally, different MSA detection methods and the potential environmental influences on the frequency of MSAs also should be considered. Furthermore, there were 17 patients who concurrently carried two MSA subsets even though the coexistence of two MSA subsets in the same patient is rare [10].

In accordance to previous studies [2, 9, 11], anti-Jo-1 was found to be the usual ASA in IIM individuals and detected in 15.1% of our present cohort. The non-anti-Jo-1 ASAs were not unrare, combining to a prevalence of 25% and computing a total prevalence of 62.3% for all ASAs. The prevalence of non-anti-Jo-1 ASAs was much higher than previously reported [12, 13]. This may partially be explained by differences in population. Meanwhile our cohort of patients was collected during hospitalization; therefore, the most severe cases may have been intercepted, and the higher prevalence of non-anti-Jo-1 ASAs may be the result of this selection. Antisynthetase syndrome is characterized by the presence of antiaminoacyl-RNA synthetase antibodies, inflammatory myositis, interstitial lung disease (ILD), arthritis, fever, mechanic’s hands, and Raynaud phenomenon (RP) [14,15,16,17,18,19]. In this research, we found that both anti-Jo-1 and non-anti-Jo-1 ASAs were strongly correlated with ILD, arthritis, and Raynaud phenomenon and coexist with anti-Ro52 antibody. However, non-anti-Jo-1-positive IIM individuals were more prone to mechanic’s hands than non-anti-Jo-1 ASA-positive patients (P = 0.03).

Anti-MDA5 was found in 25.4% of IIM patients in the present cohort. This prevalence was similar to previous findings in a Japanese cohort [3], but higher than that in European and Indian populations [2, 5]. Anti-MDA5 was associated with RP-ILD, many types of rash, and periungual erythema in align with previously reported findings [2]. We further demonstrated that anti-MDA5 was correlated with elevated ferritin and decreased lymphocyte counts in our cohort, which may explain why some anti-MDA5-positive IIM patients are prone to present with macrophage activation syndrome (MAS) and pneumocystis jirovecii pneumonia infection [20, 21]. However, we lacked adequate data to further examine the claimed link between anti-MDA5 and cutaneous ulceration [22].

Anti-TIF1-γ showed a lower frequency in our sample than in another Chinese group previously studied [9]. Anti-TIF1-γ antibodies were shown to be highly positively related to malignancy, muscle weakness, and refractory skin rash, but negatively associated with ILD. Anti-TIF1-antibodies have a well-established connection with cancer [23]. The findings in the present cohort were consistent with previously described characteristics in anti-TIF1-γ-positive patients (cancer: OR = 3.51, P = 0.002; ILD: OR = 0.03, P < 0.0001).

Besides anti-MDA5 and anti-TIF1-γ, anti-NXP2, anti-Mi-2, and anti-SAE1 are three MSAs that have been conventionally considered dermatomyositis autoantibodies because of their relations with cutaneous features [18]. Anti-NXP2 antibodies were previously related to dysphagia, malignancy, and calcinosis [24, 25]. We had insufficient data to investigate reported associations between Anti-NXP2 and malignancy or calcinosis, which is a limitation of our study. Anti-Mi-2 was also reported to be correlated with malignancy [9]. However, in the present study, anti-Mi-2 was only negatively associated with ILD because of the small sample size. Only seven IIM patients in our cohort tested positive for anti-SAE1, and hence, the sample size was insufficient for further statistical analysis.

Autoantibodies against SRP are considered to be associated with immune-mediated necrotizing myositis. Our population did not produce an association of anti-SRP with severe muscle weakness, dysphagia, and cardiac involvement found in a previous report [26]. Again, the results could be explained by the relatively small sample size.

Anti-Ro52 antibody, which is considered to have a lack of specificity and observed in many connective tissue disorders, was the most common MAA detected in our IIM patients. A previous study identified the presence of anti-Ro52 antibody as an independent risk factor for ILD in dermatomyositis and found that anti-Ro52-positive dermatomyositis patients had a worse prognosis [27]. We did not obtain the same conclusion in this research. Regarding the associations between MSAs and prognosis, previous reports described that the presence of ASAs predicted favorable outcomes [28] and anti-MDA5-positive patients had rapid disease development, with characteristics of short disease course and high mortality [19, 29]. The present data confirmed that survival rates decreased more rapidly in anti-MDA5-positive individuals during the initial follow-up stage, with anti-MDA5-positive individuals having the highest mortality within the first year after onset. Meanwhile, patients with the anti-synthetase syndrome had better long-term outcomes than patients without the anti-synthetase syndrome. Patients with IIM who tested positive for both anti-Ro52 and anti-MDA5 had poorer outcomes than those who tested positive for anti-MDA5 alone, which is consistent with previous study [30]. This data implies that if both anti-MDA5 and anti-Ro52 antibodies are present, they need to be very cautious about the possibility of serious disease onset. However, anti-MDA5 was not an independent predictor of the worst prognosis in the multivariate analysis, presumably because of the short follow-up period and small sample size.

Conclusions

MSAs were detected in around 90% of IIM patients and were classified into separate clinical subgroups. Non-anti-Jo-1 ASA and anti-Jo-1 patients had comparable features. Anti-TIF1-γ was associated with cancer in adult patients. Anti-Ro52 and anti-MDA5 positivity in patients with IIMs resulted in a poorer prognosis. Male, old age, RP-ILD, and elevated serum ferritin were independently associated with a poor outcome in IIM patients.