Longitudinal changes in Size and Composition of Carotid Artery plaques using Ultrasound: Adaptation and validation of methods

Abstract

Introduction: B-mode ultrasonography of the carotid arteries makes quantitative measurements of atherosclerotic plaque area and composition assessed as grey-scale median (GSM) possible. The purpose of this study was to set up a standardized ultrasound protocol to measure longitudinal changes in plaque area and composition and to determine the intra- and inter-observer variability of the measurements.Method: A total of 219 participants from the AGES Reykjavik Study (76±6 years old, 36% males) and 10 participants from the REFINE Reykjavik Study (60.5±5.5 years old, 70% males)underwent 2D B-mode ultrasound examination of the carotid arteries approximately 5 years apart for a longitudinal assessment of plaque area and composition. Standardized protocol was used to acquire comparable images from both visits. Ultrasound was performed bilaterally on the common carotid artery, internal carotid artery and the bifurcation. A modified version of the Artery Measurement System (v1.141) was used to measure plaque area and GSM values. Three sonographers participated in this study, intra and inter observer variability was based on re-reading ultrasound images from 25 subjects selected from the AGES Reykjavik cohort and the 10 subjects selected randomly from the REFINE Reykjavik cohort.Results: Intra sonographers variability for plaque area: For sonographer1 (AGES and REFINE groups): the coefficient of variation (CoV) was 12.10%, r=0.96 and 9.82%, r = 0.96 respectively with no statistically significant difference between observations (p=0.74, p=0.23). For sonographer2 (AGES group) the CoV was 18.63%, r=0.90, statistically significant difference between observations (p<.0001). For sonographer3 (REFINE group) the CoV was 16.03%, r=0.96, no statistically significant difference between observations (p=0.72). Intra sonographers variability for plaque GSM: For sonographer1 (AGES and REFINE groups) the CoV was 7.77%, r=0.90 and 6.84%, and r=0.91 respectively, with no statistically significant difference between observations (p=0.73, p=0.36). For sonographer2 (AGES group) the CoV was 8.04%, r=0.86, with no statistically significant difference between observations (p=0.08). For sonographer3 (REFINE group) the CoV was 6.97%, r=0.88, with no statistically significant difference between observations (p=0.20).Inter sonographer’s variability for plaque area: For sonographers (1 and 2) (AGES group) the CoV% was 23.29%, r=0.81, statistically significant difference between sonographers (p<.0001). For sonographers (1 and 3) (REFINE group) the CoV was 18.20%, r=0.91, statistically significant difference between sonographers (p=0.0023).Inter sonographer’s variability for plaque GSM: For sonographers (1 and 2) (AGES group) the CoV was 8.55%, r=0.87, with no statistical difference between sonographers (p=0.35). For sonographers (1 and 3) (REFINE group) the CoV was 7.45%, r=0.82, with no statistical difference between sonographers (p=0.44).Conclusion: This study shows that ultrasound can be used consistently for assessment of changes in plaque area and GSM over time. This can be achieved by proper training of ultrasound sonographers and applying and following a strict protocol introduced in this study.