Erosive arthritis in systemic lupus erythematosus: application of cluster analysis

Objectives: In the present study, we applied cluster analysis (CA) in SLE patients with joint involvement to identify which disease subset most commonly develops erosive damage. Methods: We collected clinical and laboratory data of SLE patients with a clinical history of joint involvement (arthritis/arthralgia). Ultrasonographic assessment was performed at level of MCPs and PIPs joints, to identify erosive arthritis, defined as the presence of erosions in at least one joint. Moreover, we detected RF, ACPA anti-CarP, and Dkk1 serum levels. We applied an unsupervised hierarchical CA to identify the aggregation of patients into different subgroups sharing common characteristics in terms of clinical and laboratory phenotypes. Results: CA included 112 SLE patients (M/F 6/106; median age 45 years, IQR 17; median disease duration 96 months, IQR 165). Arthritis was observed in 82 patients (73.2%) and inflammatory arthralgia in 30 (26.8%). US-detected erosive arthritis was observed in 29 patients (25.9%). CA on clinical and laboratory features allowed the identification of four main clusters: in particular erosive arthritis was located in a cluster including renal and neuropsychiatric involvement, serositis, positivity for ACPA, anti-Carp, anti-Sm, anti-RNP, detectable levels of Dkk1. Conclusions: The application of CA made it possible to better characterise SLE phenotype including erosive arthritis. In particular, feature-driven CA leads to the identification of a more aggressive disease, due to a common pathogenic mechanism.