In vivo relaxometric and biomechanical mapping using MRI for radiocarpal joints of healthy and rheumatoid arthritis patients

Introduction:: Rheumatoid arthritis (RA) is a systematic inflammatory disorder that can occur in many joints, resulting in inflammation, swelling, and joint destruction. The wrist radiocarpal joint is one of the common joints prone to such diseases. In the late stage, the deformities of the joint structure can be visualized by radiographs. However, methods are still lacking for the detection of early joint degeneration, including damaged cartilage structure, as well as altered biomolecular composition and biomechanical responses. Noninvasive and imaging-based analysis targeting tissue structure-function would provide potential biomarkers to detect RA at an earlier stage where pharmacological interventions may be more effective^[1]. Here, along with the quantitative MRI (qMRI) techniques that could reveal the biochemical changes of radiocarpal cartilage in RA wrists, we also performed biomechanical assessment by employing a pneumatic device that applies ulnar-radial deviation to the wrist. The displacements and strain maps were acquired at a high frame rate (25 Hz) using spiral Displacement ENcoding with Stimulated Echoes (DENSE) MRI. All MR images were obtained on both healthy and RA human wrists in vivo.


Materials and Methods:: Patients with RA met American College of Rheumatology 2010 criteria and controls had no history of RA or joint swelling. In this case, a total of four RA subjects (1 male, 3 females, 28-60 years old) and six control participants (3 males, 3 females, 29-55 years old) were recruited. In the clinical MRI scanner (3T, Siemens Prisma^fit), the participants were posed in a superman-like position. The wrist was strapped on a custom-built pneumatic loading device and placed in the center of an MRI coil (Siemens Tx/Rx 15 Flare Coil). Each study contained two 30-minute sessions with a 20-minute break in between. The cartilage in the radioscaphoid and radiolunate joints were selected as regions of interest (ROIs). In the first session, 3D Doubled Echo Steady State (DESS) images were collected for morphological purposes and establish a common imaging plane for multi-contrast acquisitions. For biomechanical assessment, a cyclic radial deviation motion inducing maximum 5° wrist angle was applied to the hand (Fig. 1A). The time-course DENSE MR images were collected only during radial deviation. In the second session, DESS images and relaxometric qMRI measurements (i.e. T2-, T2*-, and T1ρ-weighted images) were obtained while the wrist was static, and the corresponding relaxivity maps were calculated. Within each ROIs, the displacement and strain maps were calculated from the DENSE phase data using custom software (MATLAB)^[2], and the relaxometry values were averaged. Here, t-tests were used to compare treatment groups. This study was approved by the Institutional Review Board at the University of Colorado.


Results, Conclusions, and Discussions:: From the DESS images, bone oedema was observed within the carpal bones of RA subjects, but not in control participants (Fig. 1B). We also observed noticeable synovial fluids in radioscaphoid joints in both groups. The relaxometry map exhibited spatial variation within each ROI (Fig. 1C). The radioscaphoid joints had significantly higher averaged T2 and T2* values in the RA group than the control group, while the radiolunate joints were not significantly different (P >0.05), as well as their T1ρ values (Fig. 1D). For the biomechanical assessment, we used absolute displacement, principal strains, and maximum shear strain for quantification. One control participant was excluded due to image quality issues. The displacement and strain maps showed a spatial variation within each ROI, and the averaged values gradually increased throughout the loading time. The absolute displacement values were higher in the control group than the RA group from frame 1 (t=0) to frame 15 (t=560ms) for radioscaphoid and radiolunate joints. The strain values in radioscaphoid joints were higher in the control group than the RA group, while the radiolunate joints had similar strain values between the two groups (Fig. 1E).

The significant difference of the T2 and T2* values between the two groups suggest that the radioscaphoid joints may be more likely or susceptible to developing RA compared to the radiolunate joints^[3]. The DENSE results showed that the control participants had higher displacement and strain values in radioscaphoid joints than the RA patients. This may indicate the joint destruction caused by RA increases the apparent tissue stiffness, and the biomechanical assessment can detect RA in wrists. However, the sample size is small, and the collection of more data will likely aid in confirming the trends. Meanwhile, due to the presence of synovial fluid and low in-plane resolution, the ROI selection is more prone to human error. Nonetheless, we believe that the qMRI metrics along with the DENSE MRI results together can be useful noninvasive biomarkers for identifying RA status within radiocarpal cartilage. This method can be translated to other adjacent joints to further investigate the RA status in the cartilage of the carpus.


Acknowledgements (Optional): : We thank funding from the AB Nexus 20-3138 at the University of Colorado.


References (Optional): : [1] Wilsdon TD, et al., Aust Prescr. 40(2):51-58, 2017.
[2] Lee, W, et al. Magn Reson Med 89.2: 694-709, 2023.
[3] Biswas, S., 2015. Persistent Wrist Pain. Eplasty, 15.