Cytokine and autoantibody clusters interaction in systemic lupus erythematosus

Abstract

Background: Evidence supports the existence of different subphenotypes in systemic lupus erythematosus (SLE) and the pivotal role of cytokines and autoantibodies, which interact in a highly complex network. Thus, understanding how these complex nonlinear processes are connected and observed in real-life settings is a major challenge. Cluster approaches may assist in the identification of these subphenotypes, which represent such a phenomenon, and may contribute to the development of personalized medicine. Therefore, the relationship between autoantibody and cytokine clusters in SLE was analyzed.

Methods: This was an exploratory study in which 67 consecutive women with established SLE were assessed. Clinical characteristics including disease activity, a 14-autoantibody profile, and a panel of 15 serum cytokines were measured simultaneously. Mixed-cluster methodology and bivariate analyses were used to define autoantibody and cytokine clusters and to identify associations between them and related variables.

Results: First, three clusters of autoantibodies were defined: (1) neutral, (2) antiphospholipid antibodies (APLA)-dominant, and (3) anti-dsDNA/ENA-dominant. Second, eight cytokines showed levels above the threshold thus making possible to find 4 clusters: (1) neutral, (2) chemotactic, (3) G-CSF dominant, and (4) IFNα/Pro-inflammatory. Furthermore, the disease activity was associated with cytokine clusters, which, in turn, were associated with autoantibody clusters. Finally, when all biomarkers were included, three clusters were found: (1) neutral, (2) chemotactic/APLA, and (3) IFN/dsDNA, which were also associated with disease activity.

Conclusion: These results support the existence of three SLE cytokine-autoantibody driven subphenotypes. They encourage the practice of personalized medicine, and support proof-of-concept studies.

Keywords: Anti-dsDNA antibodies; Antiphospholipid antibodies; Autoantibodies; Cluster analysis; Cytokines; Interferon alpha; Interleukin 12p40; Interleukin 8; Personalized medicine; Subphenotypes; Systemic lupus erythematosus; Taxonomy.

Fig. 1

Autoantibody clusters

Fig. 2

Cytokine clusters

Fig. 3

a Association between cytokine clusters and activity of disease (p = 0.022, by Kruskal–Wallis test). b Association between cytokine clusters and autoantibody clusters (p = 0.031 for the comparison between neutral and IFNα/Pro-inflammatory clusters, by Chi-square test)

Fig. 4

Integrative analysis. a Clusters of cytokines and autoantibodies. b Association between clusters and activity of disease (p = 0.036)