Background Power Doppler (PD) ultrasound is used to assess joint inflammation in rheumatoid arthritis (RA). The extent of PD signals is influenced by several factors such as technical settings, room temperature and other variables.
Objectives The aim of this study was to examine the relevance of the position of hands (extended versus neutral position) for PD ultrasound results in RA using a 2D semi-quantitative and a 3D quantitative analysis method.
Methods Cross-sectional study of 42 consecutive RA patients with active disease according to clinical (≥1 tender and/or swollen joint) and ultrasound examination (≥1 joint with PD grade ≥1). Ultrasound examination was performed at dorsal sites of wrists and MCPs (bilaterally) by 1 (out of 2) investigators blinded to clinical results. An Esaote MyLab Twice ultrasound device was used with an 18-MHz linear transducer (2D-US) and a 4–13 MHz volumetric three-dimensional ultrasound probe (3D-US). For 2D-US, PD abnormalities were subjectively graded from 0 to 3. For 3D-US, the probe was placed at MCPs in a longitudinal direction over the extensor tendon and the joint space and at wrist over the lunate bone. Automatized image acquisition was conducted and the percentage of PD pixels within the region of interest (ROI, 100 slides, total width 8.5 mm) was calculated using the integrated 3D software.
Ultrasound examination was first performed with patients' hands in a neutral position (i.e. slight flexion of the fingers - only the wrist, thumb and finger tips touching the table), subsequently in an extended position (i.e. all palm and inner finger area touching the table).
Results Out of the 42 RA patients, 66.7% were female. Median age was 57.5 (range 19–86) years. Patients had a median of 6 (0–23) swollen and 8 (0–45) tender joints, median SDAI was 22.1 (5.1–48.3).
In neutral position, a positive 2D-US PD result was found in 192/420 (45.7%) joints. A score of 1, 2 and 3 was observed in 42 (21.9% of all PD positive joints), 139 (72.4%) and 11 (5.7%) joints, respectively. 3D-US PD findings correlated well with 2D results, with median pixel values being 0.4 (range 1.0–1.9), 1.1 (0.2–7.4) and 2.3 (0.3–7.2) in joints with 2D-US scores of 1,2 and 3, respectively.
Changing the position from neutral to extended resulted in the disappearance of PD signals in 55/192 joints (28.7%, p<0.001, assessed with 2D-US). In detail, 29/42 joints with a PD score 1 in neutral position became PD negative in extended position (minus 69%, p<0.001). Out of those 139 joints with a PD score 2 in neutral position, 26 (18.7%) and 78 (56.1%) changed to scores of 0 and 1 in extended position, respectively. A PD score 3 was found to change to score of 1 and 2 at 4 (36.4% out of 11) and 5 (45.5%) joints, respectively.
Median 3D-US count was reduced from 0.9 (range 0.1–7.4) in neutral position to 0.2 (0.0–1.7, p<0.001) in extended position.
Inter-reader agreements for 2D-US and 3D-US results were good [ICC 0.82 (95%CI 0.69 to 0.90)] and [ICC 0.75 (0.27–0.93)] as was the intra-reader reliability for 3D-US [ICC 0.79 (0.38–0.94)].
Conclusions A standardized approach for the ultrasound examination of RA patient is essential to produce reliable results. Our study demonstrates a significant reduction of PD scores by extended vs. neutral position of hands.
Disclosure of Interest None declared