ObjectiveTo analyze whether erosions, sclerosis, and squaring assessed on radiographs precede the development of syndesmophytes in patients with ankylosing spondylitis (AS).
MethodsPatients with AS from the Outcome in Ankylosing Spondylitis International Study (OASIS) cohort were followed up for 12 years, with radiographs obtained every 2 years. Two readers (reader 1 and reader 2) scored radiographs according to the modified Stoke Ankylosing Spondylitis Spine Score (mSASSS) and recorded abnormalities per vertebral corner. Progression from erosions, sclerosis, or squaring to (bridging) syndesmophytes was investigated using multilevel longitudinal (autoregressive and time-lagged) regression analysis. Interactions with reader and spinal region were investigated.
ResultsThe analysis included 211 patients (71% male, 85% HLA-B27 positive) with a mean SD age of 43 +/- 13 years and a symptom duration of 21 +/- 12 years. A total of 921 radiographs were included, with 20,509 (reader 1) and 20,568 (reader 2) vertebral corners evaluable. Erosions were found in 1% and 2.5%, sclerosis in 0.3% and 1.7%, squaring in 6.5% and 5.1%, and syndesmophytes in 25% and 27%, by reader 1 and reader 2, respectively. The odds ratio for a new syndesmophyte to occur within 2 years at the same vertebral corner if erosions, sclerosis, or squaring were present was 2.0 (95% confidence interval [95% CI] 1.7-2.3) for the cervical and lumbar spine together, 3.1 (95% CI 2.5-3.9) for the cervical spine, and 1.3 (95% CI 1.0-1.6) for the lumbar spine. When vertebral corners with erosions, sclerosis, and squaring were analyzed separately, this effect was statistically significant for erosions and for sclerosis, but not for squaring.
ConclusionErosions and sclerosis occur infrequently in patients with AS, but when they do occur, they precede the development of a new syndesmophyte, and are therefore important to assess.