The application of MR imaging in the detection of hip involvement in patients with ankylosing spondylitis

Objective: To investigate the changes in hip MR imaging, evaluate the frequency of hip involvement and compare the value of clinical symptoms, radiographs, and MR imaging in the detection of hip involvement in patients with ankylosing spondylitis (AS). Methods: Anteroposterior radiographs of the pelvis, MR imaging of the hip and clinical evaluation were undertaken in 58 patients with definite AS. All patients were followed up 3 years. Annual radiographs and clinical evaluation were carried out. The imaging data were independently assessed by two experienced radiologists who were blinded to patient identity and clinical characteristics. Based on the Bath Ankylosing Spondylitis Radiology Hip Index (BASRI-hip) scoring system, BASRI-hip scores >= 2 were defined as radiological hip involvement. On MR imaging, both acute and chronic inflammatory changes were considered positive signs for hip involvement. Symptomatic hip involvement was defined as current or past pain or limitation of the hip movement. The statistical analysis was performed using the chi(2) test for comparison of sensitivity among clinical symptoms, radiographs, and MR imaging in the detection of hip involvement and the Student's t-test for comparison of disease duration between with and without hip involvement. A P value <0.05 was considered to be statistically significant. For interpreting MRI and radiographs, the percentage of agreement between the two assessors and the kappa coefficients were calculated. Results: On MR imaging, positive changes were detected in 86 (74.1%) hips among 116 hips in all 58 patients. Joint effusion was observed in 73(62.9%) hips; 23 out of 27 patients who underwent fat-saturated contrast-enhanced T-1-weighted sequences had abnormal synovial enhancement in bilateral hips. The other abnormal MR findings included subchondral bone marrow edema in 35 (30.2%) hips, enthesitis in 22(19.0%) hips, fatty accumulation of the bone marrow in 28(24.1%) hips, bone erosive destruction in 32 (27.6%) hips, and joint-space narrowing in 4 (3.4%) hips. Based on the BASRI-hip scoring system, 68, 24, 18,6 and 0 hips had no, suspicious, mild, moderate or severe damage on conventional radiographs of the pelvis, respectively. Thirty-five hips in 20 patients had current or past pain or limitation. The proportion of hip involvement according to MR imaging, radiographs, and clinical symptoms was 74.1% (861116), 20.7% (241116), and 30.2% (351116), respectively. MR imaging yielded higher values than radiographs and clinical symptoms in the detection of hip involvement in patients with AS (chi(2) = 66.45 and 44.93, P <0.05). Interreader reliability for interpretation of findings was acceptable for both MRI and radiographs. During follow-up, radiological hip involvement were found in 10 hips with BASRI-hip scores <= 1 at baseline and clinical symptoms appeared in 15 sides of the original asymptomatic hip. On baseline MR imaging, inflammatory changes were seen in all hips which appeared symptoms and/or radiological involvement both at baseline and during follow-up. Conclusion: The proportion of hip involvement is much higher than that suggested by radiographic changes and clinical symptoms. MR imaging is superior to conventional radiographs and clinical symptoms in the detection of hip involvement. Joint effusion and synovial enhancement caused by synovitis are the commonest hip findings on MR imaging in patients with AS. (C) 2013 Elsevier Ireland Ltd. All rights reserved.