Contrast medium in power Doppler ultrasound for assessment of synovial vascularity: comparison with arthroscopy

OBJECTIVE: To evaluate the reliability of contrast-unenhanced power Doppler (CUPD) and contrast-enhanced power Doppler (CEPD) ultrasound (US) assessment of synovial vascularity of knee joint synovitis by prospective comparison with the "gold standard," arthroscopy. METHODS: A total of 18 knees of 17 patients with refractory rheumatoid and psoriatic knee joint synovitis were examined by US. Recognition of PD synovial vessel flow and its spatial arrangement in relation to the pannus/cartilage interface (P/CI) or fluid/synovium interface (F/SI) were studied by CUPD- and CEPD-US after a single intravenous bolus of galactosel palmitic acid (Levovist). Arthroscopy video recordings were reanalyzed by computer image analysis to assess synovial vascular marking. CUPD and CEPD flow signal scores were compared with each other and with corresponding vascular marking scores. Using villous vascular marking as reference, CUPD and CEPD sensitivity and specificity were measured. Interobserver variability was evaluated. RESULTS: Compared with the unenhanced PD method, contrast administration increased the PD flow signal score in 13/18 knees (72.2%), allowing increased detection of F/SI PD flow signal configuration (p < 0.018) and of the coexistence of P/CI and F/SI PD imaging (p < 0.0078). With arthroscopy as reference, contrast-enhanced PD was found to be more useful than the unenhanced method, showing more reproducible PD signal scores (p = 0.05 vs p = nonsignificant), as well as higher sensitivity (80% vs 30%), but lower specificity (62% vs 87%), in the recognition of increased vascularity of synovial villi. Interobserver agreement was 100%. CONCLUSION: The prospective comparison with arthroscopy showed the reliability of the CEPD method in synovial vessel recognition and its potential clinical usefulness in assessment of knee joint synovitis.