Three-dimensional fat-saturated T1-weighted Cartesian volumetric interpolated breath-hold examination (VIBE) for the diagnosis of aortitis in patients with suspected large vessel vasculitis: a comparative study with 18F-FDG PET applying fully integrated PET/MRI

Abstract

To evaluate the feasibility of T1-weighted (T1W) three-dimensional (3D) fat saturated Cartesian volumetric interpolated breath-hold examination (VIBE) magnetic resonance imaging (MRI) sequence for the diagnosis of aortitis in patients with suspected large vessel vasculitis (LVV) applying fully integrated 2-[18F]-fluoro-2-deoxy-d-glucose (18F-FDG) positron-emission tomography (PET)/MRI. Fourteen patients with aortitis and 14 patients with a negative study for aortitis using 18F-FDG PET as the standard of reference for the evaluation of inflammatory aortic involvement were included retrospectively. All patients were imaged at 3 T using T1W VIBE pre- and post-contrast. Four aortic segments were evaluated for image quality (IQ), diagnostic confidence (DC), and the degree of inflammatory activity (IA) using a Likert scale. Binomial and generalised estimating equation model tests were used to assess the diagnostic performance of T1W VIBE. Cohen's k was applied to test for interobserver reproducibility with respect to IA. Spearman's rank correlation coefficient was calculated to examine correlations between IQ, DC, IA, and PET results. On a patient- and segment-based analysis, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 85.7% and 59.8%, 100% and 100%, 100% and 100%, 87.5% and 68%, and 92.9% and 82.1%, respectively. IQ and DC were acceptable to good in all examinations and substantial interobserver agreement was observed for IA (Cohen's k = 0.69). IQ and DC as well as IA and 18F-FDG vessel wall uptake were significantly correlated (r=0.763 and 0.679, respectively; p<0.0001). T1W 3D fat saturated VIBE MRI allows diagnosis of aortitis and may aid in the management of patients with suspected LVV.