Background Spinal mobility, as assessed by the Bath Ankylosing Spondylitis Metrology Index, BASMI, and imaging findings have been reported to correlate on the group level. Treatment with anti-TNF leads to improvement of both spinal inflammation as assessed by magnetic resonance imaging (MRI) and total BASMI scores.
Objectives Using data from the golimumab (GLM) in ankylosing spondylitis (AS) study (GO-RAISE), we analysed the relationship between single components of BASMI and MRI scores of the corresponding spinal segments in anti-TNF-treated AS patients.
Methods Complete MR imaging sets and spinal mobility data were available for 91 pts who participated in GO-RAISE. The MRI scores for active (ASspiMRI-a) and chronic changes (ASspiMRI-c) of the cervical and lumbar spine were compared to BASMI values for the cervical (cervical rotation (CR) and tragus-to-wall (TTW)) and the lumbar (lumbar flexion (LF) and lateral lumbar flexion (LLF)) spine using the linear definition. Spearman correlation coefficients were calculated for baseline scores and for changes in both BASMI and ASspiMRI-a and -c measurements of patients treated with GLM or placebo (PBO) after 14 weeks and after 2 years of GLM therapy. Subanalyses were performed with regard to age.
Results At baseline, ASspiMRI-a scores of the cervical spine correlated with TTW (r=0.31, p=0.003) and CR (r=0.32, p=0.002) measurements, while ASspiMRI-a scores of the lumbar spine correlated with LF and LLF scores (both r=0.41, p<0.0001). In addition, ASspiMRI-c scores of the cervical spine correlated with TTW (r=0.46) and CR (r=0.45), both p<0.0001, while ASspiMRI-c scores of the lumbar spine correlated with LF (r=0.34, p=0.001) and LLF scores (r=0.41, p<0.0001). ASspiMRI-a scores correlated better in patients <40 years (TTW: r=0.31, p=0.04, LLF: r=0.42, p=0.005), while ASspiMRI-c scores correlated better in patients >40 years (TTW: r=0.35, p=0.015, LLF: r=0.48, p<0.001). In contrast, no significant correlations were found in change scores (data not shown). There was a negative correlation between MRI chronicity scores and lateral lumbar flexion at 2y: r=-0.26, p=0.037.
Conclusions Our data confirm earlier reports on patients with active AS which showed that both inflammation and structural changes contribute to impairments of spinal mobility. In addition, we demonstrate significant correlations of MRI findings with detailed spinal mobility measures before anti-TNF treatment was started. Inflammatory changes had greater impact on spinal mobility in younger patients, while structural changes had more influence on spinal mobility in older patients. The correlation of the observed changes in MRI scores and spinal mobility was significant but not high. This may be due to the different mixture of active and chronic changes in individual patients.
Disclosure of Interest X. Baraliakos Grant/research support: Janssen Research & Development, LLC, K.-G. Hermann Grant/research support: Janssen Research & Development, LLC, S. Xu Grant/research support: Janssen Research & Development, LLC, B. Hsu Grant/research support: Janssen Research & Development, LLC, J. Braun Grant/research support: Janssen Research & Development, LLC