Can routine clinical measures predict ultrasound-determined synovitis and remission in rheumatoid arthritis patients?

OBJECTIVES

The purpose of this study was to determine if routine clinical measures can predict the presence and severity of ultrasound synovitis in rheumatoid arthritis (RA) patients.

METHODS

Bilateral 1-5 MCP (metacarpopharangeal) and wrist joints were examined using power Doppler (PD) ultrasound (US). Correlations between PD scores and routine clinical measures of RA - swollen joint count (SJC), tender joint count, patient's global assessment (GA), physician's GA, CRP, ESR, MMP-3, RF and anti-CCP antibody - were determined and used to identify significant predictors of PD score. Clinical measures were then compared between two groups (patients with and without PD) and analysed using multiple logistic regression, to derive a model that predicted the absence of PD signals.

RESULTS

SJC was the most significant predictor of PD score (R2 = 0.4566, p value <0.0001), but was an inadequate predictor of PD signal remission. However, the combination of Steinbrocker's stage I or II (odds ratio [OR] 9.23, p=0.0049), SJC=0 in 1-5 MCP and wrist joints on both sides (OR 6.60, p=0.0039), and SDAI (or CDAI) remission (OR 5.06, p=0.0450) had a positive predictive value of 100%, predicting the absence of PD signals in all study patients meeting the 3 criteria.

CONCLUSIONS

PD score and absence of PD signals can be predicted using routine clinical measures. When used in combination, Steinbrocker's stage, SJC and SDAI (or CDAI) can estimate disease activity and identify patients likely to have synovitis and requiring US.