Session Type: Abstract Submissions (ACR)
Background/Purpose: To develop methodology for quantification of spinal inflammation in AS using a volumetric MRI-based technique and to explore relationships between these findings and clinical measures of disease activity.
MRI scans were undertaken using the same scanning protocol on a 3T Philips Achieva system in 11 AS patients prior to and 12 weeks after commencing anti-TNF therapy. Scans were anonymized for patient identity and time points. They were initially scored using the Berlin method by 2 musculoskeletal radiologists (AL and KS) concurrently. Images were subsequently assessed using Dynamika software to perform volumetric analysis by 2 independent scorers (scorer 1-EP and scorer 2- AL). Areas in mm2, volumes in mm3 and intensity of corresponding lesions drawn by both scorers were also compared using Bland-Altman plots. The product of volume in mm3 and mean intensity was calculated. The volume in mm3 of lesions was compared to total Berlin scores as well as BASDAI scores using Spearman Rank correlation. Percentage change in volume, intensity and the product of volume and intensity was also compared to percentage change in BASDAI scores.
Results: Bland-Altman plots showed good inter-observer variability for corresponding areas in mm2, volume in mm3 and mean intensity of lesions but compared most favourably for maximum intensity.
Correlations were observed between total volume scores and total Berlin scores, more so at baseline (r=0.7, p = 0.01 for scorer 1, r=0.63, p=0.04 for scorer 2) than at follow-up (r=0.4, p value = 0.14). No correlation was seen between volumes and BASDAI scores. However, good correlations were seen between percentage change in BASDAI and percentage change in volume in mm3 (r=0.8, p=0.002 for scorer 1 and r=0.6, p=0.04 for scorer 2) as well as the product of volume and intensity (r=0.7 for both scorers, p=0.02 for scorer 1 and p=0.03 for scorer 2). A poor correlation was seen between percentage change in Berlin scores and percentage change in BASDAI (r=0.2, p=0.56).
Conclusion: Volumetric analysis of active bone oedema lesions compares well to semi-quantitative scoring using the Berlin method. Unlike semi-quantitative scoring, volumetric analysis also shows a good correlation to change in BASDAI.
P. C. Taylor,
J. V. Hajnal,
« Back to 2013 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/quantification-of-inflammatory-disease-of-spine-in-ankylosing-spondylitis-using-volumetric-magnetic-resonance-imaging-analysis/