The Relationship of Arthroscopic Findings in RA to Histology, Radiographs and serum C-Reactive Protein levels


Orr C, McGarry T, Young F, Fearon U, Veale DJ.


Centre for Arthritis and Rheumatic Diseases, Dublin Academic Medical Centre, University College Dublin, Ireland.


The utility of synovial biopsy has been confirmed in increasing our understanding of the pathogenesis of RA, evaluating new treatments and identifying potential therapeutic targets.[1, 2] However, no scoring system for the assessment of synovitis at knee arthroscopy has been validated against the histological grade of inflammation observed in the synovial biopsies retrieved.

Furthermore, the power that arthroscopy may have in identifying patients with active inflammation despite normal CRPs has not been reported. Finally, we currently understand only 32% of the variance in factors that predict joint destruction,[3] and page 45 macroscopic findings at arthroscopy may present an additional opportunity in assessing those most at risk of this disease course.


To validate synovitis scores at arthroscopy with histology scores, CRP levels and erosive disease on radiographs.


141 patients with RA were recruited to undergo arthroscopy, and serum CRP levels were measured at the same time. The most recent set of hands and feet radiographs were assessed for the presence or absence of erosions. A macroscopic score of synovitis, graded at 5 unit intervals between 0-100, was recorded by the operator at each arthroscopy. Synovitis scores were analysed using Pearson’s test for correlation, with categorical data for histology findings (no inflammation, mild inflammation, and moderate-severe inflammation).The same test was used to determine if there was a correlation between synovitis scores and CRP levels. The Chi-square test was employed to test for a relationship between categorical synovitis scores (4 quartiles), and the presence or absence of erosions.


A correlation was observed between synovitis scores and histology findings (p≤0.002, r=0.3).

There was no correlation with synovitis scores and CRP levels. 49 (34.8%) patients had normal CRP levels (0-5mg/l), with 29 (59.2%) having synovitis scores >50%.

An association was also observed with higher synovitis scores and the presence of erosions (p≤0.02).


Synovitis can be reliably assessed by scores at arthroscopy, which correlating with histology. Arthroscopy has the power to identify patients with synovitis, where CRP levels are normal, favouring the concept that not all RA phenotypes manifest elevated CRP levels during active disease. Furthermore, those with high synovitis scores are more likely to have erosions.

Click to access the login or register cheese