The baseline interferon signature predicts disease severity over the subsequent 5 years in systemic lupus erythematosus

• Published in: Arthritis research & therapy 2021-01, Vol.23 (1), p.29-29, Article 29
• Main Authors: Mai, Lloyd, Asaduzzaman, Arundip, Noamani, Babak, Fortin, Paul R, Gladman, Dafna D, Touma, Zahi, Urowitz, Murray B, Wither, Joan
• Format: Article
• Language: English
• Subjects: Arthritis; Biomarkers; Development and progression; Diagnosis; Disease; Disease course; Gene expression; Health aspects; Interferon; Longitudinal studies; Lupus; Properties; SLEDAI-2K; Systemic lupus erythematosus
• ISSN: 1478-6362; 1478-6354; 1478-6362
• DOI: 10.1186/s13075-021-02414-0

Type I interferons (IFNs) play an important role in the pathophysiology of systemic lupus erythematosus (SLE). While cross-sectional data suggest an association between IFN-induced gene expression and SLE disease activity, interest in this as a biomarker of flare has been tempered by a lack of fluctuation with disease activity in the majority of patients. This led us to question whether IFN-induced gene expression might instead be a biomarker of overall disease severity, with patients with high levels spending more time in an active disease state. Levels of five interferon-responsive genes were measured in the whole peripheral blood at baseline visit for 137 SLE patients subsequently followed for 5 years. Log transformed values were summed to yield a composite IFN5 score, and the correlation with various disease outcomes examined. Receiver operator characteristic analyses were performed for outcomes of interest. Kaplan-Meier curves were generated to compare the proportion of flare-free patients with high and low IFN5 scores over time. The baseline IFN5 score was positively correlated with the adjusted mean SLE disease activity index-2000, number of flares, adjusted mean prednisone dose, and number of new immunosuppressive medications over the subsequent 5 years. Optimal cut-offs for the IFN5 score were determined using Youden's index and predicted more severe outcomes with 57-67% accuracy. A high baseline IFN5 level was associated with a significantly increased risk of subsequent flare. Measurement of the type I IFN signature is a useful tool for predicting the subsequent disease activity course.