Reproducibility of serial intravascular ultrasound measurements in patients with angiographically silent coronary artery disease after heart transplantation

Background Intracoronary ultrasound (ICUS) imaging is the most sensitive me thod for the early detection and serial evaluation of vasculopathy of trans plants. Both lack of agreement between observers and lack of agreement betw een serial, independent pullback procedures (repeatability), which can resu lt in a variable intraluminal catheter position may limit the reproducibili ty of ICUS measurements. Objective To evaluate the reproducibility of serial measurements of standar d linear and area cross-sectional coronary dimensions in patients with non- obstructive transplant vasculopathy. Methods We performed ICUS imaging of patients without angiographic evidence of obstructive epicardial coronary artery disease after heart transplantat ion. A 30 MHz phased-array transducer was used. Two independent pullbacks o f the left anterior descending coronary artery were performed and recorded on CD-ROM for off-line quantitative analysis of the most severely diseased site. Agreement of observers and repeatability of serial measurements were calculated by the use of linear regression analysis and Bland-Altman plots. Results Regarding agreement of observers, correlation coefficients for intr a-observer agreement ranged from r = 0.98 to r = 0.99; those for interobser ver agreement ranged from r = 0.87 to r = 0.98. Serial measurements of the identical coronary artery cross-section within independent catheter pullbac k procedures were possible for 104 of 112 target lesions (92.9%). Correlati on coefficients ranged from r = 0.91 to r = 0.97 (for lumen diameter r = 0. 91, for lumen area r = 0.93, for vessel diameter r = 0.91, for vessel area r = 0.97, for thickness of plaque r = 0.96 and for area of plaque 0.94). Th e mean difference of measurements was around zero for all parameters with S D from 0.13 to 0.4 mm for linear parameters and from 1.53 to 1.82 mm(2) for area parameters. Conclusion Serial intravascular ultrasound measurements are highly reproduc ible without any evidence of systematic error and a SD of differences of me asurements beyond the maximal spatial resolution of currently available int ravascular ultrasound catheters. Coron Artery Dis 11:555-562 (C) 2000 Lippi ncott Williams & Wilkins.