Two clusters of systemic lupus erythematosus patients with muscle involvement in a Chinese cohort

Objectives This study aimed to depict the clinical features, including myositis specific or associated antibody (MSA/MAA) profile of systemic lupus erythematosus (SLE) patients with muscle involvement in a Chinese cohort. Methods We retrospectively studied a cohort of 1696 SLE inpatients and screened for concurrent myositis features from January 2013 to June 2021. Propensity score matching was applied to enroll controls without myositis features from our cohort. Demographic, clinical and laboratory variables were collected. MSA/MAA panels containing 16 autoantibodies (TIF1-gamma, MDA5, NXP2, Mi-2 alpha/beta, SAE1, Jo-1, PL-7, PL-12, EJ, OJ, SRP, HMGCR, cN-1A, PM-Scl75/100, Ku and Ro52) were tested by line-blotting assay. Binary logistic regression and K-means clustering were applied. Results Forty-one of 1696 (2.42%) SLE patients in our SLE inpatient cohort showed features of myositis. Binary logistic regression revealed that new-onset SLE (odds ratio [OR] = 4.77, 95% CI = 1.10-20.57), interstitial lung disease (ILD) (OR = 10.07, 95% CI = 1.65-61.51), positive anti-U1RNP antibody (OR = 4.38, 95% CI = 1.08-17.75), and Raynaud's phenomenon (OR = 7.94, 95% CI = 1.41-44.69) were associated with muscle involvement. Except for anti-Ro52 (50%), anti-Ku antibody (38.2%) was the next frequently detected MSA/MAA in the panel, followed by anti-NXP2 antibody (11.8%). It was noteworthy that multiple MSA/MAAs (>= 2, excluding anti-Ro52) coexisted in 9 patients. Patients with myositis features were clustered into 2 subgroups. Cluster 1 was characterized by anti-Ku or anti-Ro52 with high SLE Disease Activity Index, whereas cluster 2 presented with anti-U1RNP, Raynaud's phenomenon and pulmonary arterial hypertension resembling mixed connective tissue disease. Conclusion In our Chinese SLE inpatient cohort, muscle involvement was infrequent. Nevertheless, distinct features in these SLE patients deserve further study.