Background Magnetic resonance imaging (MRI) is the gold standard for detection of inflammation in the sacroiliac joints (SIJ) of patients (pts) with axial spondyloarthritis (axSpA), while for chronic, structural changes (erosions, sclerosis and ankylosis) conventional radiographs (CR) and computed tomography (CT) are often preferred. The 3D volumetric interpolated breath-hold sequence (VIBE) is an MRI technique, easy to acquire in daily practice, that can visualize cartilage especially well because of its good contrast to synovial tissue.

Objectives To compare the ability of the VIBE technique to detect structural changes in comparison to CR and CT in SIJs of axSpA patients in relation to symptom duration and age.

Methods Complete sets of MRI (T1 and VIBE techniques), CT and CR of SIJs of 109 AS patients were available. Two readers evaluated all images independently, blinded to demographic data and in separate sessions for each technique. The assessment of lesions was performed based on SIJ-quadrants (SQ) to score erosions, sclerosis and ankylosis (SIJ-halves). Lesions were counted as positive if both readers were in agreement. Comparisons between MRI techniques were performed by Wilcoxon-test. Linear regression analysis was used to evaluate the influence of age and disease duration on the occurrence of different structural lesions by modeling the differences in the number of lesions in different imaging techniques as dependent variable.

Results The mean age ± standard deviation was 45.3±13.9 years (y), 55 pts (50.5%) were aged ≤45y, 67.9% male, 82.3% HLA-B27+, 58 pts (53.2%) had a disease duration ≤3y. Agreement for positive and negative findings between MRI and CT was generally high (>80% of SQs in all subgroups) and agreement between readers for all techniques and lesion types was excellent (ICC=0.979–0.997).

Overall, MRI detected significantly more SQ with erosions in pts ≤45y (n=134) and in pts with disease duration ≤3y (n=125) as compared to CT (n=91, p=0.002 and n=90, p=0.003, respectively) and in pts with age ≤45y (n=61, p<0.001) as compared to CR, while there were no differences between MRI and CT in pts. >45y or disease duration >3y. Linear regression analysis showed that MRI was superior in the detection of erosions in younger ages as compared to CT (B=0.032, p=0.001).

However, CT detected significantly more SIJ halves with ankylosis in all subgroups and more SQ with sclerosis in pts with disease duration ≤3y (n=64 vs. n=37, respectively, p=0.006), and it also detected more SQ with sclerosis in pts >45y (n=67 vs. n=38, p=0.001) and disease duration >3y (n=64 vs. n=40, p=0.003) as compared to MRI, while no differences were found in the assessment of ankylosis.

Conclusions MRI in the T1 and VIBE technique is more sensitive in the detection of erosions as compared to CT and CR in axSpA pts with short disease duration and younger age. This is due to its ability to identify structural damage in the SIJ cartilage that has not yet extended to the underlying bone. These differences are not found in pts with longer disease duration or older age. This data suggests a more prominent role for MRI also for the early detection of structural changes in the SIJ of axSpA pts.

Disclosure of Interest None declared