Background Tenosynovitis (TS) is a common manifestation of rheumatoid arthritis and postulated to be an early marker of disease but requires imaging to reliably detect its presence. The prevalence of tenosynovitis in patients with very early arthritis is unknown and there have been no studies on tenosynovitis as a predictor of the development of either rheumatoid arthritis (RA) or persistent disease.
Objectives 1) to establish the prevalence of tenosynovitis assessed by ultrasound in patients with unselected very early arthritis, 2) to explore whether the presence of tenosynovitis defined by ultrasound improves the prediction of RA or persistence compared with corresponding predictive algorithms, i.e. the Leiden and Visser score.
Methods We included 91 patients with clinically apparent synovitis of at least one joint and duration ≤3 months of any symptom attributable to inflammatory arthritis. Patients were followed prospectively for 18 months and underwent clinical, laboratory, radiographic and ultrasound assessments at baseline. We used ultrasound of hand, wrist, shoulder and ankle tendons to establish the prevalence of tenosynovitis detected by greyscale and power Doppler modalities. Ultrasound tenosynovitis variables were then analysed for their ability to improve sensitivity and specificity by calculating area under the curve (AUC) values, and comparing with predictive algorithms: the Leiden score for RA and the Visser score for persistent disease.
Results Out of the 91 patients, 39 patients developed very early rheumatoid arthritis (VERA), 17 developed very early non-rheumatoid arthritis (VENRA) and 35 had resolving disease at follow-up. All patient groups had evidence of tenosynovitis at baseline (92%, 71%, and 71% respectively).
Symmetrical PD involvement of wrist extensor tendons was more prevalent in VERA patients compared with patients with VENRA (36% vs. 6%, p<0.05) and resolving disease (11%) but there were no other significant differences in tenosynovitis distribution between VERA and VENRA groups. Tenosynovitis variables did not improve Leiden score AUC values for the prediction of RA. In contrast, tenosynovitis variables including wrist flexor and hand extensor tenosynovitis had good specificity (both 94%) for persistent disease. Tenosynovitis of the hands was more common among patients with persistent disease (VERA and VENRA) compared with patients with resolving disease (63% vs. 17%, p<0.001).
Compared to the Visser score alone, presence of hand tendon or wrist flexor tendon PD involvement improved area under the curve values for the prediction of persistent arthritis, independent of the Visser score (0.816 and 0.878, p<0.05). However ultrasound assessment of tenosynovitis conferred no additional benefit and was less useful than the additional predictive value gained from ultrasound greyscale or PD scanning for joint involvement.
Conclusions Tenosynovitis is common in very early arthritis and symmetrical wrist extensor tendinitis is significantly more common in very early RA. We have identified tenosynovitis ultrasound variables that are predictive for persistent disease. However these have inferior predictive value compared to ultrasound assessment of joints alone in very early disease.
Disclosure of Interest A. Filer Grant/Research support from: Cellzome and Pfizer, M. Cader: None Declared, A. Abhishek: None Declared, G. Allen: None Declared, C. Buckley Grant/Research support from: Wyeth, Cellzome, UCB and Pfizer, P. de Pablo: None Declared, K. Raza Grant/Research support from: Wyeth, Cellzome, UCB and Pfizer
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.