AB0727 MAGNETIC RESONANCE IMAGING IN SYMPTOMATIC BACK PAIN IN INFLAMMATORY BOWEL DISEASE: STRUCTURAL LESIONS AND HLA-B27 IMPROVE THE DIAGNOSTIC ACCURACY IN AXIAL SPONDYLOARTHRITIS

Abstract

Background: Inflammatory bowel disease (IBD) related arthropathy may manifest as peripheral arthritis, dactylitis or enthesitis as well as inflammatory back pain (IBP) due to sacroiliac joint (SIJ) and/or spinal inflammation. HLA-B27 correlates with the presence of MRI determined SIJ bone marrow oedema (BMO) in axial spondyloarthritis (axSpA) and axial psoriatic arthritis (PsA) patients. The prevalence of HLA-B27 positivity is low in IBD and patients may already be on therapy that potentially modifies MRI spinal lesions at the time of imaging. Objectives: To evaluate the utility of MRI to aid the diagnosis of axSpA in IBD patients with IBP and to explore the relationship of MRI abnormalities with HLA-B27 status. Methods: Cross-sectional, retrospective audit of consecutive MRI scans of the SIJ and spine performed (2008-2018) in a large teaching hospital. All scans were requested in IBD patients presenting with IBP and clinical suspicion of axSpA. Demographic and clinical data were retrieved from the medical notes. Decision from the clinician whether the patient had axSpA related to the IBD was also recorded. MRI scans were scored by 2 readers using the semiquantitative Leeds Scoring System (BMO grade from 0 to 3)1. An overall score for inflammatory (sum of SIJ and Spine BMO scores) and structural lesions (sum of lesions per quadrant) was calculated. Results: MRI scans from 119 IBD (Crohn’s n=82, ulcerative colitis n=31, and undifferentiated IBD n=6) patients were available for analysis. 63.9% were female, mean age 38.7 years at time of MRI with mean age of IBP onset 36.3 years. The majority (n=65/83, 78.3%) were HLA-B27 negative (missing data n=36). Thirty subjects were receiving biologic therapy for IBD. A summary of MRI findings (SIJ and spine) is shown on Table 1. Degenerative disc disease was common in the cohort with 55/119 having at least one affected level. Other incidental findings included: hemangiomas (n=8), osteoporotic fractures (n=2) and myeloma (n=1). The total BMO MRI scores (Spine plus SIJ), SIJ only BMO score, erosions, sclerosis and fat scores were higher in the HLA-B27 positive group as seen in Figure 1, with only differences in the erosion (P=0.002) and sclerosis score (P=0.038) being statistically significant. Less active (BMO) and structural lesions were seen in the biologic treated group, reaching statistical significance in total BMO (P=0.041), erosions (P=0.039), fat (P=0.003) and sclerosis (P=0.013) scores. A clinical diagnosis of axial SpA was made in 42/119 (35%) cases. Of these, n=25/42 had BMO (SIJ or spine) and structural lesions. Structural lesions alone were evident in 13/42 of these cases. BMO lesions alone with no structural changes was found in 2 cases (grade 1). Conclusion: MRI findings of sacroiliitis and spinal involvement, defined as a combination of active and structural lesions, are common in IBD subjects with symptomatic back pain. The low prevalence of BMO seen in this cohort may represent partially treated disease. The addition of structural lesions to the current definitions of a positive MRI, may help with the diagnosis of axSpA. The presence of HLA-B27 positivity appears to define a more severe axial phenotype in IBD associated axSpA.