Computer-assisted validation of the synovitis score

• Published in: Virchows Archiv : an international journal of pathology 2008-06, Vol.452 (6), p.667-673
• Main Authors: Morawietz, Lars, Schaeper, Frank, Schroeder, Joerg H., Gansukh, Tserenchunt, Baasanjav, Nachin, Krukemeyer, Manfred G., Gehrke, Thorsten, Krenn, Veit
• Format: Article
• Language: English
• Subjects: Arthritis, Psoriatic - pathology; Arthritis, Rheumatoid - pathology; Biological and medical sciences; Chronic Disease; Correlation coefficient; Diseases of the osteoarticular system; Humans; Image Processing, Computer-Assisted; Investigative techniques, diagnostic techniques (general aspects); Medical sciences; Medicine; Medicine & Public Health; Miscellaneous. Osteoarticular involvement in other diseases; Observer Variation; Original Article; Osteoarthritis; Osteoarthritis - pathology; Pathology; Pathology, Clinical - methods; Pathology. Cytology. Biochemistry. Spectrometry. Miscellaneous investigative techniques; Rheumatoid arthritis; Synovial Membrane - pathology; Synovitis - classification; Synovitis - pathology
• ISSN: 0945-6317; 1432-2307
• DOI: 10.1007/s00428-008-0587-8

Histopathological examination of synovial specimens can contribute to the diagnosis of chronic joint diseases. A so-called synovitis score has been introduced as a standardised grading system, based on the semi-quantitative evaluation of the three determining features of chronic synovitis: enlargement of synovial lining, density of synovial stroma and inflammatory infiltrate, giving a score between 0 and 9. The present study examines the reliability of this procedure by comparison with exact measurements using computer-assisted image analysis (CAIA). Seventy-one synovial specimens from patients with osteoarthritis (OA, n  = 22), psoriatic arthritis (PsA, n  = 7), rheumatoid arthritis (RA, n  = 35) and from a control group (Co, n  = 7) were evaluated using both the synovitis score and CAIA. The measurements were transformed to semi-quantitative values analogous to the synovitis score. The differences between the transformed CAIA scores and the pathologist’s scores were 0 or +/−1 in 40 cases, whereas in 31 cases the difference was greater than 1 (correlation coefficient r  = 0.725). The CAIA scores differed significantly between Co and RA cases ( p  = 0.000) as well as between OA and RA ( p  = 0.000). We conclude that the synovitis score was validated by CAIA and can be regarded a reliable grading system that contributes to the diagnostic procedure of chronic joint inflammation.