E073 The use of magnetic resonance imaging to assess disease activity in axial spondyloarthritis- data from a tertiary rheumatology centre

Abstract

Abstract Background/Aims The role of MRI of the spine and sacroiliac joints to aid the diagnosis of axial spondyloarthritis (axSpA) is now well established. However, its utility in disease monitoring, in particular assessment of treatment non-response is unclear with limited and conflicting data on the relationship between inflammatory changes suggestive of disease activity on MRI and outcome measures such as ASDAS-CRP and BASDAI. Although the most recent ASAS/EULAR guidelines recommend against routine use of MRI in the assessment of treatment response in axSpA, this is a pragmatic approach often undertaken in NHS clinics with substantial cost implications. We aimed to explore the current use of MRI to assess disease activity in axSpA in a large tertiary centre, in particular its utility in exploring treatment non-response and its impact on treatment switching decision. Methods As part of a service evaluation audit, MRIs performed under the “inflammatory back pain protocol” at Leeds Teaching Hospitals Trust between December 2021 and May 2023 were identified from radiology request data. Patients with axial spondyloarthritis (nr-axSpA, r-axSpA or axial psoriatic spondyloarthritis [axPsA]), were identified. Demographic data, baseline MRI findings, current medications, repeat MRI findings for ongoing axial symptoms, outcome measures and inflammatory markers were extracted from the clinical notes. Data on MRI findings consistent with inflammatory or structural changes related to axSpA or alternative findings, were extracted from the radiologist’s report. Results Overall, 111 patients had an MRI to assess disease activity, 93 had axSpA (49 r-axSpA, 44 nr-axSpA) and 18 axPsA. Patients were predominantly male (58.5%) and HLA-B27 positive (65.8%), 25.2% and 27.9% current smokers and ex-smokers respectively. A history of psoriasis was documented in 25.2% and peripheral arthritis in 42.3%. Baseline MRI was available on 76 patients (68.9%), of which 61 (80.3%) had spinal or sacroiliac inflammation as per the radiologist report. In total, 79 patients were on bDMARD therapy, with TNFi the commonest class (n = 59, 52.7%). Repeat MRI demonstrated active inflammation in the spine or sacroiliac joints in 58.6% (n = 65) as per the radiologist report. Patients with active inflammation in the repeat scan were more likely to have their bDMARD switched or treatment escalated (54% vs 36%). Other important findings included spinal stenosis (n = 4), disc disease (n = 15), vertebral fractures (n = 2) and potential malignancy (n = 1). Conclusion In our practice, MRI is of utility in the monitoring and assessment of non-response in axSpA in two-thirds of cases, aiding treatment decision making by leading to a switch of bDMARD in half of patients and excluding alternative pathologies, such as vertebral fractures and disc disease which may also affect components of disease activity scores. Further research exploring association of MR findings with clinical outcomes, including economic evaluation of routine MRI to assess disease activity in axSpA is warranted. Disclosure J. Weddell: None. R. Shah: None. P. Robinson: None. A. Barr: None. C. Vandevelde: None. J. Freeston: None. D. McGonagle: None. H. Marzo-Ortega: None.