Spine inflammatory changes in patients with ankylosing spondylitis assessed by magnetic resonance image

Abstract

Objective. To develop the optimal mode of spine evaluation with magnetic resonance image (MRl) in pts with ankylosing spondylitis (AS) and to study relationship between MR! signs of spinal inflammatory lesions (IL), spondylitis duration and clinical features of AS activity. Material and methods. MRl was performed in 36 pts (22 male, 14 female) fulfilling the modified NY criteria of AS. Median age of pis was 26 years (range 19 - 55), Median AS duration - 8 years (range 1,8 - 24). 34 (97%) pts were HLA-B27 positive. 21 (64%) pts had high AS activity - median BASDAI 40 (range 10 - 77). 92% of pts had inflammatory spine pain (VAS>20 mm) and 61% of pts had night pain. Median inflammatory pain duration had been defined separately for every part of the spine assessed by MRl. Median duration of axial pain was 36 months (range: 1-240). MR-scanning (Magnetom Symphony, Siemens, 1.5 T) was performed inTl, T2 and T2-FS (fat signal suppression) modes. IL scoring was done only in 29 pts evaluated in both sagittal and axial planes. We used two scoring methods: 1) individual IL score of the each spine element (vertebral bodies, processes, arches, zygapophyseai, costovertebral and costotransverse joints, ligaments), and 2) separate IL scoring in the vertebral bodies and posterior spinal elements in order yes/no”. Results. 50 MRl images of different parts of the spine (8 cervical, 30 thoracic and 12 lumbar) have been obtained in 36 pts. Spine IL were found in 35 pts. 26% of all IL were revealed in axial planes. 3 pts with short AS duration had IL only on axial slices (zygapophyseai lumbar joints, costotransverse joints, processes). IL were revealed more often in thoracic (average score: 7.1), than in lumbar (3.7) and cervical (2.1) spine. In most (26 from 29 pts, 90%) pts IL were found in painful parts of spine. There was no IL score difference between pts(n=12) with low (BASDAI 40; n=17) AS activity. Me and range were 4 (1.8-10.3) and 6 (4-16), respectively; p=0.35. There was also no difference in percent of images with IL between pts with short (Me: 4 months, range: 1-18; n= 10) and prolonged (Me: 54 months, range: 24-180; n=16) duration of spondylitis (100% and 94% of images, respectively). However, pts with early spondylitis had significantly more IL in posterior spinal structures than in vertebral bodies (92.3% and 23.1% images, respectively; p<0,001). Conclusion. Inflammatory MRl lesions are frequently observed in pts with active AS, more often in thoracic spine, and independently of spondylitis duration. Inflammatory MRl lesions in early spondylitis are revealed more often in posterior structures of spine. These results show the necessity to obtain MRl scans for early diagnosis of AS not only in sagittal but also in axial plane.