Abstract
OBJECTIVES: lupus nephritis (LN) is a severe form of systemic lupus erythematosus (SLE) with renal complications. Current diagnosis is based on invasive renal biopsy and serum antibodies and complement levels that are not specific enough. The current study aims to identify new biomarker candidates for non-invasive diagnosis of LN and explore the pathogenic mechanisms that contribute to renal injury. MATERIALS AND METHODS: A metabolomics approach using 1H-nuclear magnetic resonance (1H-NMR), was developed for comparison of urine metabolic profile of 14 LN patients, 10 SLE patients, and 11 healthy controls (HCs). Differential biomarker candidates were identified by using multivariate modeling, and their diagnostic accuracy was evaluated by receiver operating characteristic analysis (ROC). RESULTS: Three metabolites were common in differentiating all three groups including beta-alanine, 2,2-dimethylsucssinic acid, and 3,4-Dihydroxyphenylacetaldehyde and suggested as a diagnostic panel for LN with AUC of 0.89, sensitivity of 81 %, and specificity of 100 %. Complementary analyses on pathways indicated that nicotinate and nicotinamide metabolism is the most important perturbed pathway in LN. CONCLUSION: Metabolomics is a useful tool for identification of biomarkers with the ability to diagnose LN patients and predict perturbed pathways responsible for renal injury.