POS0965 MAGNETIC RESONANCE ENTEROGRAPHY AS A SCREENING TOOL FOR AXIAL SPONDYLOARTHRITIS IN CROHN’S DISEASE: A PROSPECTIVE SINGLE-CENTER CROSS-SECTIONAL OBSERVATIONAL STUDY USING MRE SCREENING FOLLOWED BY CLINICAL ASSESSMENT (ProSpA-CD)

Abstract

BackgroundPatients with inflammatory bowel disease (IBD) have an excess burden of axial spondyloarthritis (axSpA) (1), but currently there are no reliable screening tools available. Magnetic resonance enterography (MRE) is an oral contrast MRI of the small bowel used to aid the diagnosis of Crohn’s disease (CD) (2). MRE also captures the sacroiliac joints (SIJs) but the radiological assessment of this is often overlooked. MRE scans therefore contain potentially valuable unreported data, which could be used to screen for axSpA in CD patients.ObjectivesTo determine: (i) the validity, sensitivity, specificity and area under the receiver operating characteristic (ROC) curve (AUC) of MRE as a screening tool for axSpA in CD, using dedicated axial MRI scans with clinical assessment as the gold standard; (ii) the proportion of patients with evidence of axSpA on MRE who fulfill the 2009 Assessment of Spondyloarthritis International Society (ASAS) criteria for axSpA; (iii) whether the presence of extra-intestinal manifestations (EIMs) or CD location can predict the occurrence of axSpA in CD.MethodsProSpA-CD was a pair of linked prospective observational studies conducted at a large secondary care university hospital in the UK. In ProSpA-CD-Screen, patients who had undergone MRE between 2015-2019 were assessed for eligibility. The consenting participants‘ MRE images were evaluated by a rheumatologist and a radiologist, using a scoring system for evidence of axSpA. Participants with evidence of axSpA and a control group of participants without evidence of axSpA were invited to participate in ProSpA-CD-Assess, in which the performance of MRE as a screening test was assessed against a gold standard of clinical assessment, serological testing and dedicated MRI of the spine and pelvis of each participant. ROC curve and logistic regression analyses were performed.ResultsA total of 1344 MRE scans were performed between 2015-2019, of which 501 eligible patients were identified. For ProSpA-CD-Screen, 259 participants were recruited. MRE SIJ abnormalities were identified in 94/259 participants and from these, 90 cases were recruited to ProSpA-CD-Assess. For the control group, 71/165 participants with normal SIJs were recruited to ProSpA-CD-Assess (total cases + controls = 161). A clinical diagnosis of axSpA was made in 41/161 (25.47%) participants, of which 25/161 (15.53%) fulfilled ASAS criteria. We constructed a ROC curve to assess test performance using different thresholds on the MRE scoring system. The optimum threshold corresponded to a sensitivity of 0.60 (95%CI 0.39, 0.79) and specificity of 0.85 (95%CI 0.78, 0.91) for MRE as a screening tool for axSpA (ASAS criteria) with an AUC value of 0.78 (95%CI 0.65, 0.86, p<0.05). We did not find a significant association between axSpA and CD location. Multivariable logistic regression demonstrated significant associations between axSpA (ASAS criteria) and inflammatory back pain (OR 11.03; 95%CI 2.53, 48.10; p<0.05), HLA B27 (OR 9.94; 95%CI 1.84, 53.87; p<0.05), history of dactylitis (OR 51.01; 95%CI 1.07, 2423.90; p<0.05) and plantar fasciitis (OR 9.08; 95%CI 1.30, 63.55; p<0.05).ConclusionWe have shown that MRE as a screening tool for axSpA in a cohort of CD patients has good specificity, but poor sensitivity, suggesting that its use as a screening tool is limited. There was no significant association between axSpA and CD location or CD EIMs.