A multicenter italian study to assess flow mediated dilation variability for clinical studies.

Abstract Objective: Endothelial dysfunction is associated to cardiovascular risk factors providing prognostic information when non-invasively studied by measurement of flow-mediated dilation (FMD). Despite the large effort to standardize the methodology, FMD evaluation is still characterized by problems of reproducibility and reliability. The aim of this multicentre study was to standardize the procedure for FMD assessment among different research Centres and evaluate FMD variability over time in healthy volunteers. Design and Methods: Seventy-five healthy subjects (aged 20-60 years) were recruited in 6 Italian Hypertension Units. FMD was assessed as dilation of the brachial artery secondary to 5 minutes wrist ischemia by trained operators using a clamp to held the ultrasound probe. Sequences of B-mode images of the brachial artery were VCR recorded for baseline FMD (time 0) and repeated 1 hour after maintaining the probe in the same position (time 1). A third sequence was obtained 1 month apart (time 2). Endothelium independent vasodilation to glyceril trinitrate (25 ug, sublingual) was also evaluated at time 1 and 2. FMD and response to GTN changes were measured blindly as percentage changes in brachial artery diameter by an automatic edge detection system at the coordinating Centre. The intra- (time 0 versus 1) and inter-session (time 0 versus 2) coefficients of variation were calculated for the 6 different research Centres and overall to assess FMD and GTN variability over time. Results: All recordings were suitable for analysis. FMD was 7.5±3.2% at time 0, 7.3±3.3% at time 1 and 7.4±2.9% at time 2. Overall, the intra-session FMD variability was 10.2± 12.1%, ranging from 7.1 to 10.9°/0 in the different Centres. Inter-session FMD variability was 13.0±8.9%, ranging from 12.2& to 13.8% in the different Centres. GTN response was 13.9±4.1% at time 1 and 12.8±4.8% at time 2. Overall inter-session variability of GTN response was 12.9±9.8%, ranging from 10.1% to 17.1%. Conclusions: This multicenter study shows that intersession FMD coefficient of variation (1 month apart) results to be similar to the; intra-session one (1 hour apart), which would represent the intrinsic variability in the endothelial response. Thus, a standardized procedure, including operator training, defined experimental settings and automatic brachial artery measurements, ensures an adequate FMD reproducibility over time and it can be routinely used for the assessment of endothelial function in clinical studies.