Ultrasound measurement of atherosclerosis.

Ultrasound Measurement of Atherosclerosis To the Editor: The recent review of intima-media thickness (IMT) measurements by Bots et al,1 while presenting cogent arguments for use of maximum versus mean carotid IMT for studies of interventions, failed to discuss important limitations of IMT. They claim as an important advantage of IMT, as opposed to morbidity and mortality as end-points, the considerable reduction in sample size and duration of study needed to show efficacy of new interventions; however they neglect to mention methodology that has significant advantages compared with IMT, namely the measurement of carotid plaque. IMT is very insensitive to change in plaque, because plaque grows along the carotid in the axis of flow 2.4 times faster than it thickens.2 Thus, measurement of plaque, as opposed to measurement of IMT, detects change with treatment much more readily. It is also very important to understand that ultrasound measurements of various aspects of atherosclerosis such as stenosis, plaque, or IMT assay biologically distinct phenomena. The main determinants of IMT are age and blood pressure; multiple regression with traditional risk factors gives an R of only 0.15 to 0.17 for IMT,3 compared with an R of 0.52 for carotid plaque area.4,5 This leads to important distinctions that must be made both for genetic studies of atherosclerosis and for studies of interventions aimed at atherosclerosis. These issues were discussed in a recent paper in Stroke.6 Total carotid plaque area is a stronger predictor of outcomes than IMT: patients in the top quartile of plaque area have 3.5 times the risk of stroke, death, or myocardial infarction when compared with patients in the lowest quartile,7 after adjustment for age, sex, blood pressure, cholesterol, smoking, diabetes, homocysteine, and treatment for lipids and blood pressure; in contrast, patients in the top quintile of IMT had a relative risk of 3.15 versus the lowest quintile, after adjustment for a smaller panel of risk factors.3 While IMT represents end-organ disease in the artery wall, it consists mainly of media and correlates poorly with coronary disease;8 it represents mainly hypertensive medial hypertrophy and correlates better with left ventricular mass than with coronary stenosis.9 Carotid plaque correlates better with coronary artery disease than does IMT.10,11 For the most successful study of carotid IMT, Bots et al provide sample size estimates ranging from 468 per group for a parallel clinical trial with an effect size of 30% over 2 years, to 30 per group for a 100% effect size over 3 years. This is far inferior to measurement of plaque: study of 2-D plaque area requires sample sizes of 150 per group for a 30% effect size over 2 years,2 and the study of 3-D plaque volume can show significant changes in 3 months in 20 patients per group with an effect size of 100% (presented at the AHA stroke meeting in February 2004). Instead of fussing over what is the best way to use IMT, a 1980s technology, it would be better to move to 3-D ultrasound measurement of plaque volume12 for evaluating effects of interventions aimed at atherosclerosis. For genetic studies, it is important to distinguish among noninvasive phenotypes, as they will be influenced differently by genetic factors affecting blood pressure, oxidative stress, lipids, and other factors affecting atherosclerosis.6 J. David Spence, MD, FRCPC, FAHA Director of Stroke Prevention and Atherosclerosis Research Centre Robarts Research Institute Prof. Neurology and Clinical Pharmacology University of Western Ontario