Background/Purpose : Interferon (IFN) pathways are dysregulated in a subset of patients with systemic lupus erythematosus (SLE). IFN dysregulation likely contributes to SLE pathogenesis and influences the disease course. This study examined correlations between IFN transcriptional modular scores, biologically relevant soluble mediators, and SLE disease activity in patients with or without the IFN signature.

Methods : Adult SLE patients (n=49) meeting >/= 4 ACR classification criteria provided longitudinal samples at times of low (SLEDAI<6) and high (SLEDAI≥6) disease activity. Immune pathway activity was evaluated by modular transcriptional analysis of Illumina Beadchip Microarray gene expression data in 98 patient samples and 20 age, race and gender matched healthy controls. Plasma soluble mediators (n=23) and antinuclear antibodies (n=11) were assessed by multiplex-bead based assays and ELISAs. Non-parametric paired tests were used for intra-individual univariate analyses, conditional logistic regression for multivariate analyses.

Results : Within a given individual, IFN signatures remained active and did not change with disease activity. In patients with active IFN signatures (n=29), several soluble mediators were elevated during high compared to low disease activity. These included MIG (median 168 pg/mL vs. 100 pg/mL; p<0.01), IFNa (5.82 pg/mL vs. 3.69 pg/mL; p<0.01), MCP1 (240 pg/mL vs. 209 pg/mL; p<0.01), BLyS (1.88 ng/mL vs. 1.18 ng/mL, p<0.01), TNFRI (2.82 ng/mL vs. 1.98 ng/mL, p<0.05) and TNFRII (622 pg/mL vs. 552 pg/mL, p<0.01). Additionally, patients with active IFN signatures showed higher M4.11 plasma cell module scores and dsDNA concentrations during high vs. low disease activity (p<0.05). Periods of high disease activity in patients with active IFN signatures were best distinguished by M4.11 plasma cell module scores, BLyS, MIG and hemoglobin in multivariate logistic modeling (p<0.01).     

Conclusion: Although IFN activity correlates with disease severity in lupus populations, IFN module scores may not correlate significantly with intra-individual changes in disease activity. In contrast, plasma concentrations of IFN- and TNF-related soluble mediators significantly correlated with SLEDAI scores, suggesting that longitudinal monitoring of soluble mediators may be useful in individuals at increased risk of disease flares.