OPTICAL COHERENCE TOMOGRAPHY IN CHILDREN: DEVELOPMENT OF A PROTOCOL FOR THE STANDARDIZED QUANTITATIVE ANALYSIS OF CORONARY ARTERIES

Abstract

Optical coherence tomography (OCT) is a high-resolution intravascular imaging technique that has shown promise for the early detection of intimal thickening of coronary arteries in pediatric heart transplant recipients. The potential utility of OCT as a diagnostic and/or prognostic tool is currently unclear due to a lack of a standardized approach to analyse OCT images in this population. We established the OCT Core Imaging Laboratory at BC Children’s Hospital to develop a standardized protocol for the quantitative analysis of OCT images. Our study population included n=30 pediatric cardiac transplant recipients who had OCT imaging of the coronary arteries performed as part of standard of care at BC Children’s Hospital (St Jude Medical, now Abbott). We deemed a ‘pullback’ analyzable if ≥10mm of the segment was of good image quality (defined as ≥75% of cross-section visible without artifact). We evenly selected 10-20 cross-sectional frames ≥1 mm apart, avoiding bifurcations. Using the device-manufacturer’s software, we applied digital planimetry to trace the lumen-intima border, the internal elastic lamina, and the external elastic lamina for each cross-sectional frame. Cross-sectional areas (CSA, mm2) were derived for the lumen, the intima and the media; maximal intimal thickness (mm) was defined as the widest point between the circular markers for the intima and media. We calculated the intima/media (I/M) CSA ratio. For each pullback, we derived summary values from cross-sectional frames for relevant variables as means, medians, 95th percentile (p95) and maximal values. We performed inter-rater and intra-rater reliability analyses by intra-class correlation coefficients and Bland-Atman Analyses. Overall, 82% of analyzable pullbacks met inclusion criteria for our protocol (≥10-20 frames, ≥1 mm apart), resulting in 67 pullbacks from 28 patients analysed. The right coronary artery was most frequently imaged (58%), and more often met inclusion criteria for quantitative analyses (85% vs. 82% overall). Mean vs. median and max vs. p95 summary statistics yielded very similar values for I/M CSA and maximal intimal thickness (all ICC>0.98; difference < 3%), suggesting our protocol provides robust summary values. We registered excellent inter-rater and intra-rater reliability for I/M CSA and maximal intimal thickness (all ICC>0.95; difference < 5%). We provide a systematically developed and rigorously tested protocol for the quantitative analysis of OCT images in children. The novel application of OCT to a pediatric population requires expert image analysis and has the potential to improve the care of pediatric transplant patients who are at risk of structural coronary changes.