Spinal mobility in the cervical and lumbar spine correlates with magnetic resonance imaging findings for inflammatory and structural changes in patients with active ankylosing spondylitis.

Objectives: We aimed to assess relationships between single Bath Ankylosing Spondylitis Metrology Index (BASMI) components and corresponding spinal segment magnetic resonance images (MRI) in anti-tumour-necrosis-factor-treated AS patients.

Methods: Using available MRI and BASMI data from the GO-RAISE trial (n=91 patients), MRI scores for active inflammatory (ASspiMRI-a) and chronic structural (ASspiMRI-c) changes in cervical and lumbar spine segments were compared with BASMI cervical (cervical-rotation [CR] angle, tragus-to-wall [TTW] distance) and lumbar (lumbar flexion [LF], lateral-lumbar-flexion [LLF]) spine component scores (linear definition). Generalised linear models were employed to assess relationships between BASMI components and ASspiMRI-a/ASspiMRI-c measurements at baseline and for week-14 (golimumab/placebo groups) and week-104 (all golimumab-treated) change scores.

Results: Baseline lumbar ASspiMRI-a scores correlated with LF and LLF (β=0.231 and 0.238, respectively; both p<0.01), while this was less prominent for ASspiMRI-c scores and LLF (β=0.142, p=0.04). A significant but weak correlation was found between changes from baseline to week 104 in cervical spine ASspiMRI-c score and TTW distance among all treated patients (β=0.161, p=0.003).

Conclusions: Detailed assessments indicated baseline spinal mobility impairment in patients with active AS correlated weakly with MRI-detected lumbar spinal inflammation; correlations with chronic, structural damage/changes were very weak. Improved, less variable MRI and spinal metrology assessments are needed for future clinical research.