Background Progression of joint damage in RA occurs primarily in joints that are clinically swollen, and repair is seen only in joints with no clinical swelling.1;2 On the other hand, joints exhibiting US but not clinical swelling may progress radiologically despite clinical remission.3 This discrepancy requires elucidation to understand the value of clinical compared with US joint assessment, since except for histology, there is no true gold standard for US joint activity.
Objectives To evaluate the differences in number of clinically and US active joints in RA, with special regards to the impact of US definitions of activity.
Methods We performed US imaging of 22 joints of the hands of RA patients in clinical remission (CDAI≤2.8). Each joint was assessed for grey scale synovial hypertrophy (GSSH; 0=no hypertrophy; 1=hypertrophy) and power Doppler (PD) signal (0= no hyperaemia, 1= mild, 2=moderate, 3=marked). We investigated the sensitivity and specificity of clinically swollen joints (SJC) for presence of US changes and repeated this analysis using different cut points and combinations of US definitions. We further assessed changes of CDAI if clinical SJC were replaced by US active joints.
Results Of the 58 patients in clinical remission, 96.6% had GSSH and 93.1% had positive PD signals. The sensitivity and specificity of clinical SJC for any US abnormality (GSSH>0 or PD>0) was 1.7% and 100%, respectively. When we used more stringent definitions of US activity (GSSH>0 & PD=3), these values changed to 22.2% and 99%. Replacing clinical by US counts in the calculation of CDAI, the US CDAI values approached the clinical CDAI with increasing stringency of the definition used (Fig.). In fact, the most stringent definition (GSSH>0 & PD=3) revealed identical average values as seen in the clinical CDAI.
Conclusions Low PD signals increase the numbers of patients classified as having active joints and lead to a poor sensitivity of clinical assessment in comparison to US assessment. When the PD signal is combined with assessment of GSSH, or the strength of the PD signal is considered, the differences decrease, particularly when looking at their impact on composite indices. The relevance of this finding will have to be tested in relation to radiographic progression of joint damage.
Klarenbeek NB et al. ARD 2010; 69(12):2107-2113.
Lukas C et al. ARD 2010; 69(5):851-855.
Brown AK et al. Arthritis Rheum 2008; 58(10):2958-2967.
Disclosure of Interest None Declared