Background/Purpose: Recent achievements in Magnetic Resonance Imaging (MRI) have been the gradient-echo-based T1-delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) focussing on the detection of proteoglycan content in cartilage; moreover, the multi-echo and spin-echo T2-mapping has been developed for the assessment of cartilage hydration and collagen microstructure. Current MRI-studies suggest a link between dGEMRIC and T2-mapping and early changes of cartilage in inflammatory diseases. However, their diagnostic value in inflammatory diseases has not yet been fully clarified..To investigate the relation between morphological and biochemical alterations in the cartilage of patients with Rheumatoid Arthritis (RA) by high-resolution 3Tesla-MRI.

Methods: 29 RA-patients received a 3Tesla-MRI scan of the 2nd and 3rd metacarpophalangeal (MCP) joint of the dominant hand. T2-mapping and dGEMRIC were performed with 2 small-diameter surface coils designed for high-resolution imaging of cartilage (0.26×0.26x105mm voxel size; Verio Siemens Healthcare, Erlangen, Germany). T2 and T1 relaxation times were obtained via a region-of-interest (ROI) evaluation. MCP heads and bases were scored semiquantitatively for synovitis, bone marrow edema (BME) and bone erosion (BE) using the RA MRI scoring (RAMRIS) system; joint space and cartilage thickness were measured perpendicular to the joint plane in all 3 joint regions (fig. 1a).

Results: Inter- and intraobserver agreement was good (details see fig. 1b). For correlation analysis mean values of real and total joint spaces (RJS, TJS), RAMRIS-subscores, and ROIs of dGEMRIC- and T2-evaluations were used; image 1a shows details of source data acquisition, image 1c data of correlation analysis. Interestingly, early changes of cartilage such as BME and synovitis in RAMRIS as compared to dGEMRIC were correlated negative (p=0.029; p=0.003); likewise, BME and synovitis showed positive correlation (p=0.013; p=0.015) in T2. In contrast, periarticular changes occurring later in the course of the disease such as BE did not correlate with dGEMRIC (p=0.704) and weakly with T2 (p=0.026). All joint space subanalyses of MCP3 despite RJS correlated with dGEMRIC and T2 mainly in medial region (TJS p=0.017; TCT p=0.020; CT p=0.018). Additionally, in dGEMRIC MCP2 and RJS correlated in ulnar side (p=0.001).

Conclusion: High-resolution MRI using dGEMRIC and T2-mapping enables meticulous detection of very early inflammatory changes in cartilage which are known to precede RA-typical periarticular bone damage. This is important both for early discovery of those damages, adequate therapy decisions and therapy monitoring; moreover, it may also have an impact on the development of anti-inflammatory drugs in the future. Additional studies on mechanical influences on cartilage are needed to further evaluate the mechanisms of joint space alterations.