Macroscopic and microscopic features of synovial membrane inflammation in the osteoarthritic knee: correlating magnetic resonance imaging findings with disease severity

Arthritis Rheum. 2005 Nov;52(11):3492-501. doi: 10.1002/art.21373.


Objective: To determine the magnetic resonance imaging (MRI), macroscopic, and microscopic characteristics of synovial membrane inflammation, to study the relationship between disease severity and the degree of synovial inflammation on MRI and on macroscopic and microscopic examination, and to look for colocalization of chondral lesions and synovial inflammation.

Methods: Thirty-nine patients with knee osteoarthritis (OA) were classified into 2 groups according to the severity of cartilage lesions as revealed by chondroscopy. Group 1 (n = 14) had mild cartilage lesion(s) without exposure of subchondral bone. Group 2 (n = 25) had severe cartilage lesion(s) with focal or diffuse exposure of subchondral bone. Synovitis was evaluated on T1-weighted MRI sequences according to the degree of synovial thickening on a 4-point scale (ranging from 0 to 3) in 5 regions of interest. Synovial membrane was macroscopically scored, and biopsies were performed on the 5 preselected sites for histologic scoring.

Results: The mean +/- SD synovial thickening score on MRI was 1.55 +/- 0.90, with no significant difference between groups 1 and 2. Intra- and interobserver reproducibility of the total synovial score was excellent, and interobserver reproducibility of the MRI grade was good. Synovitis was diffuse and associated with chondral lesions only in the medial femorotibial compartment (r = 0.49, P = 0.001). The degree of synovial thickening on MRI correlated with qualitative macroscopic analysis (r(s) = 0.58, P < 0.001) and with microscopic features (synovial lining cells [r(s) = 0.23, P < 0.007], surface fibrin deposition [r(s) = 0.12, P < 0.01], fibrosis [r(s) = 0.31, P < 0.006], edema [r(s) = 0.17, P = 0.07], congestion [r(s) = 0.30, P < 0.005], and infiltration [r(s) = 0.46, P < 0.0001]). Fibrin and infiltration parameters were more severe in end-stage disease (P = 0.009 and P = 0.02, respectively).

Conclusion: Synovitis may be present from the onset of OA and may be evaluated on MRI. MRI evaluation of synovitis could be used to classify OA patients in clinical trials and could help to identify those who could benefit from synovium-targeted therapy.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Arthroscopy / methods
  • Cartilage, Articular / pathology
  • Female
  • Gadolinium DTPA
  • Humans
  • Knee Joint / pathology*
  • Knee Joint / physiopathology
  • Magnetic Resonance Imaging / methods*
  • Male
  • Middle Aged
  • Osteoarthritis, Knee / pathology*
  • Osteoarthritis, Knee / physiopathology
  • Synovial Membrane / pathology*
  • Synovitis / pathology*


  • Gadolinium DTPA