Influence of initial axial involvement on the phenotypic evolution of axial spondyloarthritis: 10-year data from the DESIR cohort.

OBJECTIVES: To investigate whether the initial location of inflammatory axial pain in axial spondyloarthritis (axSpA) is associated with the development of peripheral or extra-rheumatological manifestations over a 10-year period. METHODS: Data were drawn from the 10-year prospective DESIR (Devenir des Spondylarthropathies Indifférenciées Récentes; French Cohort of Undifferentiated Spondyloarthritis) cohort of axSpA at disease onset. Patients with available data on the initial location of inflammatory axial pain were included. The main outcomes were the occurrence of extra-axial manifestations (arthritis, dactylitis, enthesitis) and extra-rheumatological manifestations (psoriasis, uveitis, inflammatory bowel disease). Patients were categorised according to the initial location of inflammatory axial pain as cervicothoracic, lumbosacral or diffuse (involving both regions). Associations between inflammatory axial pain location and extra-axial or extra-rheumatological manifestations were assessed using multivariate Cox models, adjusted for age, sex, exposure to conventional synthetic and biological disease-modifying antirheumatic drugs (as time-dependent covariates) and the presence of these manifestations at baseline. FRENCH COHORT OF UNDIFFERENTIATED SPONDYLOARTHRITIS : cohort of axSpA at disease onset. Patients with available data on the initial location of inflammatory axial pain were included. The main outcomes were the occurrence of extra-axial manifestations (arthritis, dactylitis, enthesitis) and extra-rheumatological manifestations (psoriasis, uveitis, inflammatory bowel disease). Patients were categorised according to the initial location of inflammatory axial pain as cervicothoracic, lumbosacral or diffuse (involving both regions). Associations between inflammatory axial pain location and extra-axial or extra-rheumatological manifestations were assessed using multivariate Cox models, adjusted for age, sex, exposure to conventional synthetic and biological disease-modifying antirheumatic drugs (as time-dependent covariates) and the presence of these manifestations at baseline. RESULTS: Among 662 included patients, 94 (14.2%) had cervicothoracic, 466 (70.3%) had lumbosacral and 102 (15.5%) had diffuse involvement at baseline. Patients with diffuse involvement were older, had a lower education level, a lower prevalence of prior dactylitis and higher modified Stoke Ankylosing Spondylitis Spinal Score compared with other groups. Over 10 years, diffuse involvement was independently associated with an increased risk of developing arthritis (HR 1.61; 95% CI 1.15 to 2.25) compared with lumbosacral involvement. No other significant associations were observed. CONCLUSION: Diffuse initial axial involvement identifies a subgroup of patients with axSpA at increased risk of developing peripheral arthritis. This finding has practical implications for clinicians in early risk stratification, patient counselling and interpretation of therapeutic trial outcomes.