Background Ankle synovitis occurs in over a third of early inflammatory arthritis patients and may be present in isolation. However, the clinical outcomes associated with ankle involvement in this population are not clearly defined.
Objectives The objectives of this study were to a) compare the prevalence of ankle synovitis on clinical examination, and of ankle synovitis or tenosynovitis on ultrasound; and b) to examine the diagnostic outcomes of patients with clinical ankle synovitis in a very early arthritis (VEA) cohort.
Methods We used data from the Birmingham VEA cohort (symptom duration ≤3 month, synovitis of ≥1 joint at presentation) to obtain information on age, sex, rheumatoid factor (RF), anti-CCP antibody (ACPA), baseline 68-joint swollen joint count (SJC), ultrasound ankle synovitis/tenosynovitis at baseline, and diagnostic outcome during 18 month follow up. Clinical ankle synovitis at baseline was defined as ankle joint swelling on examination. Ultrasound ankle synovitis and tenosynovitis were defined as any grey scale synovitis or effusion in the ankle, and any grey scale synovitis in the tendons anterior, medial or lateral to the ankle. We estimated the prevalence of ankle involvement as defined by clinical and ultrasound examination. Kappa statistic was used to examine their agreement. Relative risk (RR) and 95% confidence interval (95% CI) were used to estimate whether clinical ankle synovitis at baseline predicts diagnostic outcomes in patients with VEA, adjusting for age, sex, RF or ACPA positivity, and 68-joint SJC.
Results The cohort included 324 patients, 52% women (mean age 51 years). 85 of the 324 patients had 38-joint musculoskeletal ultrasound (MUS). Clinical ankle synovitis was present in 103 (32%). Of the 85 patients who had MUS, 70 (82%) had ultrasound synovitis/tenosynovitis at any ankle. Of these, 35% ankles were clinically swollen. There was poor agreement between MUS and clinical examination (k =0.12 at each ankle). Patients with clinical ankle synovitis were younger, and had greater SJC (mean (S.D) age 47 (18) years, and median (inter-quartile range (IQR)) SJC 3 (2-7)) compared with those who did not have clinical ankle synovitis (mean (S.D.) age 52 (17) years and median (IQR) SJC 3 (1-6); p<0.05, <0.01 respectively). Twenty-five patients presented with isolated ankle synovitis. Their outcomes were: unclassified arthritis (UA) 12 (48%); seronegative spondarthritis 3 (12%); rheumatoid arthritis (RA) 2 (8%); definite acute sarcoid arthritis i.e. erythema nodosum and inflammatory arthritis with intra-thoracic involvement 2 (8%), possible acute sarcoid arthritis i.e. erythema nodosum and inflammatory arthritis 4 (16%), gout 1 (4%), and SLE 1 (4%). Patients with clinical synovitis at any ankle were more likely to develop either definite or possible acute sarcoid arthritis (aRR 5.68, 95% CI 1.22-26.44). This was statistically significant for bilateral but not for unilateral ankle synovitis (aRR, 95%CI 10.15, 1.13-90.89; and 2.14, 0.36- 12.63 respectively). Among those presenting with oligoarthritis, ankle swelling increased the risk of UA over RA (aRR 1.81, 95%CI 1.31-2.49).
Conclusions Ankle involvement is more frequent in VEA than is clinically apparent and occurs in younger patients. Clinically apparent ankle synovitis at presentation significantly increases the risk of acute sarcoid arthritis, and of UA in those with oligoarthritis.
Disclosure of Interest None Declared