Magnetic resonance imaging of rheumatoid arthritis in metacarpophalangeal joints

Objective. To report the development of high-resolution targeted magnetic-resonance imaging (MRI) techniques (not using injections of contrast media) to investigate and monitor rheumatoid arthritis (RA) in the metacarpophalangeal (:MCP) joints. Design and patients. A total of 25 RA patients (age range 30-68 years) with varying degrees of disease severity ranging from early onset through active disease to the burntout stage, were imaged. (One patient subsequently underwent surgery and histological data was obtained.) A series of 10 control subjects were also studied - two for each 10-year age range. All the RA subjects were assessed for disease activity using standard clinical criteria and radiography as part of normal procedures. MRT was carried out using a targeted system and novel I radiofrequency coil. Images of the MCP were performed at very high resolution with 1.5 mm slice thickness and in-plane resolution 130 mu m. Standard gradient-echo (GE) sequences were used far anatomical imaging, multiple-echo GE sequences used to produce effective spin-spin relaxation time (T2*) maps and optimised binomial-pulse presaturation used in conjunction with a GE sequence to generate magnetization-transfer (MT) ratio maps. Results. High-quality high-resolution images of the MCP joints were obtained which highlighted normal anatomy and key features characterising the disease state (e.g. pannus, bone erosions, vascularity). Accurate measurements of T2* and MT with variations of +/-4% and +/-2% respectively were achieved. In active disease, variations in T2* and MTR;IT could be determined throughout areas of pannus, clearly demonstrating the heterogeneity of this erosive tissue. Pannus in MCP joints with active destruction was found to have high values of T2* varying from 25 ms to 40 ms with pockets up to 100 ms, whereas pannus present in chronic destruction, or burnt-out disease, had T2* values ranging from 21 to 29 ms. MT-active tissue was uniformly distributed in burnt-out disease, which was confirmed histologically in one case, compared with a more heterogeneous distribution in active disease. Conclusion. The MRI sequences and targeted system developed allow high-resolution studies of RA disease progression and activity. The data confirm the variable pattern of the disease and, in particular, heterogeneity of pannus.