The specificity and limitations of sacroiliac joint magnetic resonance imaging in the diagnosis of axial spondyloarthritis in patients with chronic low back pain

Abstract

Objective: To evaluate the specificity and limitations of sacroiliac joint magnetic resonance imaging (MRI) in the diagnosis of axial spondyloarthritis (SpA)in patients with chronic low back pain. Methods: We retrospectively analyzed clinical data of 390 patients with chronic low back pain in Department of Rheumatology, the PLA General Hospital from January 2013 to December 2015, including clinical manifestations, laboratory examinations and MRI data of sacroiliac joints. Results: There were 238 men and 152 women recruited. A total of 326 cases were diagnosed as axial SpA, including 216 men and 110 women with mean age (27.10±8.64) years and mean duration (7.64±3.50) months. Among these 326 patients, 243 (74.5%) were HLA-B27 positive. The other 64 patients were considered as diagnoses rather than SpA (non-SpA), consisting of 22 men and 42 women with mean age (31.29±7.76) years and mean duration (5.75±2.90)months. Non-SpA group had 10 (15.6%) patients with HLA-B27 positive. There were 68.1% and 65.0% SpA patients showing bone marrow edema and bone erosion of sacroiliac joint in MRI imaging respectively. Although there were 25.0% non-SpA patients with bone marrow edema and 7.8% with bone erosion in MRI of sacroiliac joint, the scores of bone marrow edema 0.00(0.00, 0.75) and bone erosion [0.00(0.00, 0.00)] were significantly lower compared with those in axial SpA group [bone marrow edema scores 2.00(0.00, 4.00), bone erosion scores 1.00(0.00, 3.00); P<0.05]. The scores of fat infiltration [1.00(0.00, 4.25), 1.00(0.00, 4.00)] and bone sclerosis [0.00(0.00, 1.00), 0.00(0.00, 1.75)] were not statistically different between two groups. Diagnostic sensitivity of bone marrow edema and bone erosion for axial SpA were 56.4% and 64.1% respectively, specificity were 93.8% and 92.2% respectively. The positive predictive value of bone marrow edema and bone erosion for axial SpA were 9.09 and 8.21, negative predictive value were 0.46 and 0.38.Diagnositic sensitivity of fatty infiltration and bone sclerosis for axial SpA group were 29.1% and 57.7%, specificity were 64.1% and 46.9%.The positive predictive value of fatty infiltration and bone sclerosis for axial SpA were 0.81 and 1.08, negative predictive value were 1.10 and 0.90. Conclusion: Sacroiliac joint MRI is a valuable method to diagnose axial SpA in patients with chronic low back pain. Yet it still has some limitations. Clinical presentations and spinal MRI would be helpful in some patients.