Diagnostic Utility and Economic Impact of a Sequential MRI Protocol in Axial Spondyloarthritis: A Retrospective...

Background Diagnosis of axial spondyloarthritis (axSpA) is a diagnosis based on both clinical history and characteristic radiographic and MRI findings. MRI is increasingly used to diagnose non-radiographic axSpA. The main question is whether all patients need both MRI of the sacroiliac joint (SIJ) and the spine to achieve a diagnosis of axSpA. Here, we propose a sequential MRI SIJ followed by MRI spine only for those patients who do not have inflammatory lesions on MRI SIJ, to optimise MRI use and save resources. Methods This is an observational retrospective cohort study conducted in our tertiary centre specialist axSpA clinic. We included only patients with a diagnosis of axSpA with both MRI spine and SIJ, who were referred between January 2024 and November 2024 to our tertiary service. Radiology reports were then reviewed by clinicians and determined if the MRI SIJ imaging alone was sufficient for the diagnosis, or if the MRI whole spine imaging was necessary to achieve a diagnosis of axSpA. Cost minimization analysis was performed. Logistical regression was performed to identify potential clinical characteristics that may predict the need for both SIJ and spine MRI for diagnosis. Results A total of 62 patients were included in this analysis. Of these, 51 patients (82.3%) had changes on MRI SIJ alone that were diagnostic for active inflammatory lesions compatible with a diagnosis of axSpA, and only 11 patients (17.7%) had inflammatory lesions in the spine, hence would have needed further spinal MRI to identify inflammatory lesions in addition to SIJ MRI. A cost minimisation analysis of utilising a sequential scanning protocol, where patients would initially have an MRI SIJ that would be reviewed by a clinician to determine whether further MRI spine was needed, identified potential cost savings of £11,118 and 34.0 hours of active scanning time. Logistical regression identified that there was a trend that patients with inflammatory bowel disease (IBD) had a greater odds ratio of 3.80 (95%CI 0.537- 26.952, p = 0.181) of needing a spinal MRI in addition to SIJ MRI to achieve an axSpA diagnosis. A history of human leukocyte antigen (HLA)-B27 positivity, psoriasis, and uveitis was not associated with a greater likelihood of needing an MRI spine in addition to MRI SIJs to achieve axSpA diagnosis. Conclusion Our analysis has demonstrated that there are potential significant resource savings from a cost and MRI-scanning time perspective by the implementation of a sequential MRI scanning protocol in patients with a high clinical suspicion of axSpA. Additionally, we have identified IBD as a clinical variable that may potentially predict the need for further MRI spine imaging in addition to MRI SIJ imaging.