Detection of cholesterol crystals by optical coherence tomography.

Abstract

The aim of this study was to determine the ability of optical frequency domain imaging (OFDI)/optical coherence tomography (OCT) imaging systems to visualise the presence of cholesterol crystals (CCR) in human atherosclerotic coronary arteries. We performed ex vivo imaging of human coronary arteries by OFDI/OCT. A total of 559 cross-sectional images from 45 autopsy cases were co-registered with histology; 117 histologic sections showed presence of necrotic core with cholesterol clefts (CC). We modified a previously used OFDI/OCT definition for identification of CCRs which we now define as a linear and discrete high-intensity signal (bright area) within the plaque with sharp borders between it and adjacent low-/intermediate-intensity tissue. Additionally, the high-intensity signal is not a spot but a well-defined area distinguishing it from macrophages which lack sharp borders. OFDI/OCT imaging identified the presence of CCR in 30 of the 117 histologic sections. The sensitivity and specificity of OFDI/OCT for detection of CCR was 25.6% and 100.0%, respectively. By multivariate analysis, significant predictors to visualise CCR by OCT/OFDI were 1) an overlying fibrous plaque, and 2) the presence of stacked CC, defined as CC arranged one on top of another with >3 layers of CC. The prevalence of complicated plaques (i.e., plaque haemorrhage and late necrotic core) was significantly higher in detectable CCR by OFDI/OCT as compared to undetectable CCR. The presence of stacked CCs is required to detect CCR by OFDI/OCT. Detection of CCR by OCT/OFDI may help us to identify the late stages of atherosclerotic coronary plaque progression and improve risk stratification.