The clinical utility of high resolution magnetic resonance imaging in the diagnosis of giant cell arteritis - A critically appraised topic

Background: Giant cell arteritis (GCA) is a relatively common form of systemic vasculitis, known for its predisposition to affect extracranial branches of the carotid artery and associated potential for causing visual loss and stroke. Neurologists need to be vigilant for this disorder, diagnose it early, and institute effective corticosteroid therapy. The differential diagnosis can be broad. Unfortunately, all clinical and laboratory features of GCA are limited by either low sensitivity or low specificity. Temporal artery biopsy remains the gold standard, but it has its own limitations. Noninvasive imaging techniques, like magnetic resonance imaging (MRI), may be capable of detecting the occurrence of GCA. Objective: How useful is high resolution MRI as a diagnostic test in establishing the diagnosis of GCA? Methods: We addressed the question through development of a structured critically appraised topic. Participants included consultant and resident neurologists, clinical epidemiologists, medical librarian, and clinical content experts in the field of neuroradiology, rheumatology, and vascular neurology. Participants started with a clinical scenario and a structured question, devised search strategies, located and compiled the best evidence, performed critical appraisals, synthesized the results, summarized the evidence, provided commentary, and declared bottom-line conclusions. Results: A single study which assessed the diagnostic value of MRI against a reference standard in GCA was appraised. For the MRI, the estimated sensitivity was 81% (95% CI 67-95), specificity was 97% (91-100), positive likelihood ratio (LR) was 26.6 (95% CI 3.8-184.8), negative LR was 0.20 (95% CI 0.10-0.41). The study exhibited several methodological weaknesses which interfered with its validity. Conclusions: The specificity and positive LR of high resolution MRI are sufficiently high that a positive MRI combined with other clinical and laboratory data consistent with GCA may be useful in diagnosing GCA. Given the relatively low sensitivity of the test, a negative MRI would not be sufficient to rule out the diagnosis of GCA.