Résumé
Objective Recently, three distinct phenotypes of patients with Sjögren disease (SjD) have been described based on cluster analysis: B cell active with low symptoms (BALS), high systemic activity (HSA), and low systemic activity with high symptoms (LSAHS). We aimed to assess whether these clusters were associated with distinct biomarkers and the prognostic value of interferon (IFN) signature. Methods The Assessment of Systemic Signs and Evolution in Sjögren's Syndrome cohort is a 20‐year prospective cohort of patients with SjD. The following biomarkers were compared: IFN‐α2, IFN‐γ, CXCL10, CXCL13, BAFF, interleukin (IL)‐7, fms‐like tyrosine kinase 3 ligand, CCL19, and tumor necrosis factor receptor 2 (TNF‐RII). IFN signature was assessed using transcriptomic analysis. We then compared systemic and symptomatic evolution, and the risk of new immunosuppressant prescription and of lymphoma, according to the IFN signature across the three clusters. Results A total of 395 patients (94% female, median age 53 [interquartile range 43–63] years) were included. Higher levels of CXCL13, IL‐7, and TNF‐RII were found in the BALS and HSA clusters compared with the LSAHS cluster. A high IFN signature was mainly found in the BALS cluster (57%, vs 48% and 38% in the HSA and LSAHS clusters, respectively). This IFN signature was mainly driven by type I IFN, with higher levels of IFN‐α2. In the BALS cluster, a high IFN signature was associated with a higher risk of new immunosuppressant treatment (hazard ratio 9.38; 95% confidence interval 1.22–72.16). All lymphoma occurred in patients with high IFN signature. Conclusion The three SjD clusters displayed distinct expressions of IFN signature and markers of T and B cell activation, confirming distinct pathophysiologic mechanisms. High IFN signature could predict systemic evolution in the BALS cluster.