de y r

Aim To determine the reproducibility and histological respectively]) and focal calcification (0·78[0·74–0·82] and correlation of qualitative intravascular ultrasound imaging of atheromatous coronary arteries using the recently pro-posed European Society of Cardiology classification of plaque composition in conditions approximating the clinical setting. Methods Atheromatous lesions (n=21), identified from 30 post-mortem human coronary arteries, were imaged using intravascular ultrasound in a pulsatile flow system which simulates coronary flow. Fifty sites (21minimum lumen area and 29distal reference sites) were selected indepen-dently by two observers from continuous video recordings. Atheromatous plaque was classified as echodense, echolu-cent, heterogenous or calcified by each observer and by one observer on separate occasions. Arterial specimens were histologically sectioned at these sites and similarly analysed by a third observer blinded to the intravascular ultrasound appearances. Results Overall inter- and intra-observer reproducibility for plaque-type (Kappa 0·87[0·80–0·94] and 0·89[0·85–0·93 0·88[0·84–0·92]) was high. Differences in site selection sig-nificantly influenced reproducibility particularly at refer-ence sites. Agreement for overall plaque type between intravascular ultrasound and histology occurred in 89% of sites (Kappa 0·73[0·69–0·77]). Specificity and positive predictive values for individual plaque types were greater than 90%. Conclusion Using modern intravascular ultrasound tech-nology in an in vitro system which approximates the clinical setting the proposed ESC classification of plaque compos-ition by intravascular ultrasound is reproducible and corre-lates well with histology. It should therefore perform reliably in diagnostic intravascular ultrasound examin-ations and in the guidance of percutaneous coronary interventions