Background Spondylarthritis (SPA) spectrum refers to a variety of skeletal and extra-articular inflammatory manifestations that may combine differently among patients.
Objectives To examine whether SPA disease manifestations at baseline would combine according to an ordered or random pattern in patients with early inflammatory low back pain (IBP) and suggestive of SPA.
Methods Baseline clinical and demographic characteristics, as well as imaging (i.e. pelvic X-rays, pelvic MRI and peripheral entheses ultrasound) and biological data (including HLA-B27) of patients included in the French multicenter DESIR cohort of early (<3 yrs duration) IBP suggestive of SPA were analysed by multiple correspondence analysis in order to graphically assess the association between the studied manifestations. Several methods of cluster analysis using those variables were then performed to identify subgroups of patients with similar characteristics.
Results Seven hundred and eight patients were included into the Cohort. Among the 700 analysed patients cluster analysis allowed us to classify 688 patients in 3 major groups (table 1). Cluster A contained a majority of men, with predominant isolated axial manifestations, and higher frequency of uveitis and positivity of HLA-B27. Cluster B predominantly consisted of women, with higher frequency of peripheral signs (including vascularisation at entheses by power Doppler ultrasound (PDUS)), psoriasis and younger onset of disease symptoms. Cluster C was composed by patients with predominant axial symptoms, equal distribution of sex, highest presence of pelvic X-rays and MRI positivity, and higher prevalence of HLA-B27 positivity.
Conclusions Cluster analysis of SPA manifestations among patients with early IBP allowed us to clearly identify at baseline 3 different groups of clinical phenotypes. Ongoing follow up will allow to determine whether these clusters may correspond to different severity patterns.
Funding The DESIR cohort is supported by an unrestricted grant from PFIZER France
Disclosure of Interest None Declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.