Background Systemic sclerosis (SSc) is a complex multi-organ autoimmune disease that is caused by inflammation, vasculopathy and fibrosis. Clinical heterogeneity, unpredictable course, high mortality and resistance to treatment make physicians still frustrated. Recently, there are unmet needs of useful biomarkers for diagnosis, disease activity and severity of SSc. Metabolomics is expected to be a useful tool for the identification of biomarkers and new therapeutic targets. Several researchers have applied metabolomics to autoimmune diseases such as systemic lupus erythematosus, and rheumatoid arthritis.

Objectives To identify the biomarker candidates for the diagnosis of SSc using metabolomic analysis.

Methods Fifty-five patients (48 females (87%); mean age 59.30 ± 11.44 years; disease duration 6.92 ± 4.36 years; 11 diffuse cutaneous SSc, 44 limited cutaneous SSc) and thirty age, gender matched healthy controls (HCs) were enrolled. Serum samples after 8 h of fasting were stored at -80° and analysed using nuclear magnetic resonance (NMR)-based metabolomics.

Results Multivariate analysis showed metabolic differences between SSc and HCs using partial least squares discrimination analysis (PLS-DA: R²Y=0.654, Q²=0.482) and orthogonal partial least squares discrimination analysis (OPLS-DA: R²Y=0.83, Q²=0.674) (Figure 1). We identified nine discriminatory metabolites (p<0.05): isopropanol, lactate, 2-oxoisocaproate, glucose, and formate were increased and pyruvate, glutamate, methylguanidine, and methanol were decreased in SSc compared with those in the HCs. Using these metabolites for diagnosis of SSc, sensitivity was 96.36% and specificity was 80% by Leave-one-out analysis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

PLS-DA and OPLS-DA plots based on NMR analysis of systemic sclerosis (white dots) and healthy controls (black square). PLS-DA: partial least squares discrimination analysis, OPLS-DA: orthogonal partial least squares discrimination analysis.

Conclusions There are considerable differences in the serum metabolomic characteristics between SSc and HCs. We expect that metabolomic analysis can be a useful tool for identification of potential diagnostic biomarkers of SSc.

Disclosure of Interest None declared