Purpose: To elaborate the best MR imaging protocol for studies in rheumatoid arthritis (RA) and to evaluate the sensitivity and interobserver agreement with respect to detection of bone erosions (MR and radiography) and grading of synovial membrane hypertrophy (MR imaging only).
Material and methods: MR imaging and conventional radiography of wrist and metacarpophalangeal (MCP) joints were performed in 41 RA patients and 3 healthy controls. The following pulse sequences were applied: T1-weighted spin-echo (T1-SE) with and without contrast enhancement, T2-SE, T2-turbo-SE, T1-2D-FLASH, T1-3D-FLASH, fat-saturated-T1-SE, STIR and 3D-DESS.
Results: Bone erosions were found by MR compared to radiography in 261 versus 85 bones of the wrist (ratio 3.1) and 59 versus 21 MCP joint quadrants (ratio 2.81). MR and radiography interobserver agreements were both approximately 90%. Likewise, MR scored synovial membrane hypertrophy in wrist and MCP joints with a high interobserver agreement. The most informative MR sequence appeared to be contrast-enhanced T1-SE MR, preferably with fat saturation. A STIR sequence or T2-weighted fat saturation sequence was useful in screening for joint disease.
Conclusion: The sensitivity of MR is superior to conventional radiography with respect to detection of bone erosions in wrist and MCP joints. The interobserver agreement for MR and radiography was similar. Thus, MR of wrist and finger joints may become a useful supplement to conventional radiography in the evaluation of RA patients in clinical trials and clinical practice.