Abstract
The OMERACT Magnetic Resonance Imaging (MRI) Task Force has developed and evolved the psoriatic arthritis MRI score (PsAMRIS) over the last few years, and at OMERACT 10, presented longitudinal evaluation by multiple readers, using PsA datasets obtained from extremity MRI magnets. Further evaluation of this score will require more PsA imaging datasets. As well, due to improved image resolution since the development of the original rheumatoid arthritis MRI scoring system (RAMRIS), the Task Force has worked on semiquantitative assessment of joint space narrowing, and developed a reliable method as a potential RAMRIS addendum, although responsiveness will need to be evaluated. One of the strengths of MRI is the ability to detect subclinical synovitis, so the group worked on obtaining low disease activity/clinical remission datasets from a number of international centers and presented cross-sectional findings. Subsequent longitudinal evaluation of this unique resource will be a major continuing focus for the group.
In rheumatoid arthritis (RA) clinical trials, magnetic resonance imaging (MRI) is now frequently used as an outcome measure. Given the difficulties with demonstrating structural radiographic change with modern trial designs, there is increasing need for MRI to provide a more sensitive tool for assessing outcomes in proof of concept and dose-finding Phase II and III trials. The Outcome Measures in Rheumatology Clinical Trials (OMERACT) MRI Task Force has been instrumental in supporting MRI advances, with its RA MRI score (RAMRIS, evaluating bone erosions, bone edema, and synovitis) becoming the standard outcome for trials involving MRI. After OMERACT 9 and feedback received from audience participation at that meeting, the group focused on developing some novel areas in RA and improving the performance of the psoriatic arthritis MRI score (PsAMRIS)1.
Joint Space Narrowing Assessment in RA
There is a long history of assessing damage in RA with radiographic erosion and joint space narrowing (JSN, a surrogate measure of cartilage loss) scores integral to the various modifications of Sharp. Evaluation of JSN was left out in the early phase of developing the RAMRIS (late 1990s), because image quality at this time was insufficient to allow acceptable levels of interreader reliability2. At OMERACT 9, it was suggested to revisit the assessment of JSN, and that this may improve the regulatory view of the usefulness of MRI. We subsequently devised a number of exercises to iteratively develop a reliable JSN score, and these were presented at OMERACT 103.
Understanding the “MRI Acceptable Disease Activity State” in RA
Previous reports have shown a high frequency of MRI signs of inflammation in patients with RA in clinical remission, and a recent study showed that MRI findings in such patients are predictive of progressive radiographic joint destruction4,5. This suggests that modern imaging is critical for future remission criteria. However, it is not yet clarified if there is an amount of MRI inflammation (synovitis and/or bone edema) below which patients will not show progressive joint destruction. We therefore undertook the initial steps to explore whether a “MRI acceptable disease activity state” can be defined. By gathering datasets with MRI and clinical information from a number of collaborating centers internationally, we have initiated an OMERACT dataset in which to explore the importance of so-called “subclinical” inflammation. After the considerable effort to gather these datasets of patients in low disease activity [Disease Activity Score 28 (DAS28) < 3.2] and clinical remission, an initial exercise was undertaken involving a cross-sectional analysis to determine if the extent of subclinical disease was as prevalent as previously reported in both low disease activity states and clinical remission6.
Ongoing Development of a PsAMRIS
The use of MRI in clinical trials of other inflammatory arthritides has also grown. Although the MRI features of peripheral joint pathology in PsA have been described (including synovitis, tenosynovitis, entheseal abnormalities, bone erosions and proliferations, and periarticular inflammation), until now there has been no well-accepted MRI scoring system for outcome assessment. Our group has worked on this over the previous 3 years. At OMERACT 9, we provided MRI definitions of important pathologies in peripheral PsA and suggestions concerning appropriate MRI sequences for use in PsA hands7. We also proposed a preliminary OMERACT PsAMRIS for evaluation of inflammatory and destructive changes in PsA hands, developed through an iterative process as was used for the RAMRIS. Following on from work and suggestions for development, at OMERACT 10 we were able to present the results of longitudinal analyses of a reliable PsAMRIS8.
Future Research Directions
Clearly, the areas discussed above all require further work in order to meet the standards of the OMERACT filter. The JSN score will require evaluation in longitudinal cohorts to determine its sensitivity to change and its advantages over radiographic JSN evaluation in clinical trials, although it is increasingly likely to be used in studies with MRI acquisition where inflammatory measures (synovitis and bone edema) are the key outcomes. Understanding the role of persistent imaging inflammation in RA patients achieving low clinical levels of disease activity remains a critical question for clinical practice and trials, especially in times of target-driven treatment. The compiling of an OMERACT database with imaging and clinical data provides a valuable resource for exploring definitions of acceptable disease activity levels, and audience feedback at OMERACT 10 supported ongoing analyses in this area, especially on longitudinal data. This will be a major focus for our group in the next couple of years. As well, some novel areas for research were suggested, including, given the extensive experience of the group in small-joint evaluation, development of MRI scoring for hand osteoarthritis.
