Background/Purpose: Childhood-onset systemic lupus erythematosus (cSLE) is a heterogeneous disease with differing levels of disease activity and organ-specific disease manifestations in each individual.  In particular, lupus nephritis (LN) persists as a leading cause of morbidity.  We largely lack effective, non-invasive biomarkers to accurately detect LN activity and also to predict children at greatest risk for poor prognosis with LN.  S100 proteins are a diverse group of calcium-binding proteins that can promote inflammation and be elevated in varied etiologies of nephritis.  The purpose of this cross-sectional study was to evaluate five S100 proteins (S100A4, S100A6, S100A8/9 and S100A12) in both the serum and urine as potential biomarkers of global and renal-specific disease activity in a cohort of cSLE patients.   

Methods: Patients were selected for inclusion from an ongoing cSLE Clinical and Research Database based on available serum at an active disease visit, designated by a SLE Disease Activity Index 2000 (SLEDAI-2K) ≥ 8.  We then searched for an accompanying paired serum sample from a less active and ideally inactive visit if available.  Serum and urine from the active and less active visits were analyzed for protein levels of S100A4, A6, A8/9 and A12 using commercial ELISAs (BÜHLMANN MRP8/14 and Circulex S100A12/EN-RAGE, S100A4, S100A6).  Clinical characteristics were also collected on all patients for each visit, including demographics, standard lab values, organ specific disease activity and current medications.   

Results: Serum S100 levels did not differ significantly between either paired or independent active/inactive visits, although S100A4 levels trended toward significance.  Urine S100A4, A6 and A12 levels were elevated in cSLE patients with active LN, as defined by a renal SLEDAI ≥ 4, when compared to both cSLE patients with active disease but no LN and cSLE patients with inactive disease (refer to table for median S100 levels). Urine S100A4, A6 and A12 levels also differed by the degree of LN activity.  In the paired sample analysis, only urine S100A4 levels differed significantly in patients with active LN, increasing by a median value of 4.98 ng/mL from inactive to active visits.  

Conclusion: Higher levels of S100 proteins in the urine are associated with LN activity in our cSLE cohort.  In particular, S100A4 and A6 levels are much higher in the urine than in the serum, suggesting the possibility of localized production within the kidney and a potential role in LN pathogenesis.  S100 proteins could serve as novel biomarkers of LN; however, it will first be necessary to establish that elevation of urine S100 proteins is specific to LN.  Further studies are required to validate urine S100 levels as a marker of LN activity in a separate cSLE cohort and to study the possible role of S100 proteins in LN pathogenesis.