Background MRI is used as a diagnostic tool to detect active disease of the sacroiliac joint. T1 weighted and STIR images are generally used and ASAS definition is based on bone marrow edema (BME). However, imaging after intravenous administration of gadolinium (Gd) may improve the detection of BME, synovitis, capsulitis and/or enthesitis compared to STIR sequence1.
Objectives To investigate the additional value of T1 fatsat after Gd (T1/Gd), compared to T1 and STIR sequence in the detection of active lesions of the SIJ typical for spondyloarthritis (SpA) and to assess its influence on final MRI diagnosis based on the ASAS definition of active sacroiliitis.
Methods All patients included in the SpondyloArthritis Caught Early (SPACE)-project received MRI of the SIJ (MRI-SIJ). Inclusion criterion for this study was chronic back pain of short duration (≥3 months, ≤2 years, onset <45 years). Imaging was performed on 1.5T (Philips, Best, Netherlands). Acquired sequences were coronal oblique T1, STIR and T1/Gd. The parameters evaluated were BME, synovitis and capsulitis/enthesitis. Parameters were scored on STIR as well as T1/Gd sequence and compared in conjunction with unenhanced T1 images. A positive MRI was defined as the presence of BME on the STIR images according to the ASAS definition. Scoring was done by three blinded trained readers. If 2 out 3 readers stated positive, the MRI was considered as such.
Results In 127/157 patients included in SPACE a baseline MRI (with the use of Gd) was obtained and in 67 patients also a follow-up MRI after 3 months (with the use of GD) was obtained. In 22/127 patients (17.3%) active sacroiliitis was diagnosed according to the ASAS definition based on the STIR sequence. No additional BME was found on the T1/Gd. In 7 patients (5.5%) in addition to present BME, synovitis and/or capsulitis/enthesitis were found at baseline. All patients with capsulitis also showed synovitis. In 1 patient (0.8%) synovitis was an isolated finding. This patient did not fulfill the ASAS, ESSG, Amor or modified New York classification criteria. These findings did not change at follow up. The patients with a positive MRI, capsulitis or synovitis at follow-up were the same patients who show these signs at baseline (see table; * not estimated with STIR).
Conclusions STIR sequence by itself is sufficient to detect active sacroiliitis according to the ASAS definition. The additional presence of synovitis, capulitis/enthesitis observed with Gd, is seen in the presence of BME, except in one patient without clinical SpA. In line with the recommendations by ASAS, our data show that Gd is not needed in the MRI assessment of patients with SpA.
Hermann K. J Rheumatol 2005;32:2056–60
Disclosure of Interest None Declared