Gadolinium-free assessment of synovitis using diffusion tensor imaging. NMR in biomedicine Sandford, H. J., MacKay, J. W., Watkins, L. E., Gold, G. E., Kogan, F., Mazzoli, V. 2021: e4614

Abstract

The dynamic contrast-enhanced (DCE)-MRI parameter Ktrans can quantify the intensity of synovial inflammation (synovitis) in knees with osteoarthritis (OA), but requires the use of gadolinium-based contrast agent (GBCA). Diffusion tensor imaging (DTI) measures the diffusion of water molecules with parameters mean diffusivity (MD) and fractional anisotropy (FA), and has been proposed as a method to detect synovial inflammation without the use of GBCA. The purpose of this study is to (1) determine the ability of DTI to quantify the intensity of synovitis in OA by comparing MD and FA with our imaging gold standard Ktrans within the synovium and (2) compare DTI and DCE-MRI measures with the semi-quantitative grading of OA severity with the Kellgren-Lawrence (KL) and MRI Osteoarthritis Knee Score (MOAKS) systems, in order to assess the relationship between synovitis intensity and OA severity. Within the synovium, MD showed a significant positive correlation with Ktrans (r=0.79, p<0.001), while FA showed a significant negative correlation with Ktrans (r=-0.72, p=0.0026). These results show that DTI is able to quantify the intensity of synovitis within the whole synovium without the use of exogenous contrast agent. Additionally, MD, FA, and Ktrans values did not vary significantly when knees were separated by KL grade (p=0.15, p=0.32, p=0.41, respectively), while MD (r=0.60, p=0.018) and Ktrans (r=0.62, p=0.013) had a significant positive correlation and FA (r=-0.53, p=0.043) had a negative correlation with MOAKS. These comparisons indicate that quantitative measures of the intensity of synovitis may provide information in addition to morphological assessment to evaluate OA severity. Using DTI to quantify the intensity of synovitis without GBCA may be helpful to facilitate a broader clinical assessment of the severity of OA.

View details for DOI 10.1002/nbm.4614

View details for PubMedID 34549476