- Email: [email protected]
Ultrasound determination of total arterial wall thickness
Ultrasound determination of total arterial wall thickness Timothy C. Hodges, MD, Paul R. Detmer, PhD, David L. Dawson, MD, Robert O. Bergelin, MS, Kirk W. Beach, PhD, MD, Thomas S. Hatsukami, MD, Brenda K. Zierler, RN, BSN, RVT, Janette A. Isaacson, RVT, RCVT, and D. Eugene Strandness, Jr., MD, Seattle, Washington Purpose: Ultrasonic measurement techniques for determining intima-media thickness and total arterial wall thickness have been described. The intima-media thickness measurements are currently in use in large epidemiologic trials. Intima-media thickness does not evaluate extramedial atherosclerotic change and so may not fully reflect pathologic changes in the arterial wall. Methods: After we performed variability studies of B-mode image acquisition and measurement, we measured total wall thickness and intima-media thickness of the common carotid arteries in 60 adult subjects in three groups: a control group aged 20 to 29 years, a control group aged 60 to 79 years, and a claudication group aged 60 to 79 years. Measurements were made with B-mode ultrasound images. Results: No statistical difference between sexes was noted. A statistically significant (p ::5 0.05) increase in intima-media thickness and wall thickness was found with increasing age, and an additional increase was observed with clinically significant lower extremity arterial occlusive disease (p ::5 0.05). Image quality had an effect on measurement accuracy. Conclusions: The finding that the wall thickness of common carotid arteries is increased in those patients with clinically significant lower extremity disease supports the theory that atherosclerosis affects the arterial wall in a systemic fashion. Because intima-media thickness also increases across subject groups without change in its proportional contribution to the total arterial wall thickness, extramedial arterial changes also occur with aging and the development of atherosclerosis. We propose that because increases in wall thickness measurements of common carotid arteries follow intima-media thickness increases (but do not necessarily measure the same physiologic change) and the wall thickness method can be used in cases when the intima-media thickness cannot be measured, arterial wall thickness measurement may serve as an alternate or confirmatory test of peripheral artery atherosclerotic severity. (J VAse SURG 1994;19:745-53)
Two of the challenges to modern vascular medicine are the early detection and accurate staging of atherosclerosis. It is accepted that the use of clinical endpoints as the sole means of detection reflects only the effects of advanced disease. Arteriography can detect early atherosclerotic changes, but its cost, risk, and invasive nature preclude its use in patients as a From the Vascular Surgery Section, Department of Surgery, University of Washington Medical Center, Seattle, Wash. Supported by National Institutes of Health grant HL 42270-03 (NIH, Bethesda, Md.). Reprint requests: D. Eugene Strandness, Jr., MD, Professor and Chief, Division of Vascular Surgery, RF-25, University of Washington, Seattle, WA 98195. Copyright © 1994 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/94/$3.00 + 0 24/1/51295
screening test. Duplex scanning remains the most reliable 1,2 noninvasive modality for performing measurements of both blood flow velocity and arterial dimension. 3-5 One theory concerning the development of atherosclerotic lesions states that atherosclerosis is a systemic vascular disease beginning with a diffuse thickening of the arterial wall. If this thickening does indeed precede plaque formation, then it should occur at arterial sites not commonly affected by plaque, such as the common carotid arteries (CCA). Measurement ofCCA thickening may be useful in the detection of early arterial wall changes that may occur with modification of risk factors. Beach et al. 6 developed a wall thickness (WT) measurement strategy to determine the arterial lumen to vein lumen distance, a measurement that encom745
746 Hodges et ai.
passes the entire arterial wall. The measurements were confined to specific anatomic locations where the imaged artery and vein were adjacent and within the depth resolution of the ultrasound image. The vein wall thickness was assumed to be small, constant with respect to the artery, and unaffected by arterial atherosclerosis. Using this method, the authors demonstrated that the average superficial femoral artery and vein WT was significantly higher in subjects with diabetes who had peripheral arterial disease than in the control group. To test the hypothesis that arterial WT increases in a systemic pattern, the WT measurement method described by Beach et al. 6 has now been applied to a nondiabetic population of both healthy subjects and those with arterial occlusive disease. In a prospective manner we have determined the total arterial wall thickness and intima-media thickness (IMT) in the common carotid arteries of subjects aged 20 to 29 years and 60 to 79 years. This article presents total arterial wall thickness measurements and their correlation with age and disease state. METHODS We proposed to measure total arterial WT in defined subject groups. To validate any new measurement method, sources of variability should be characterized. To that end, before data were collected on subjects in the study, a variability study was performed to estimate the amount of uncertainty in the measurements contributed by the ultrasound technologist when acquiring the image. To eliminate the variability introduced by multiple subjects and focus only on that error contributed by the technologist, right and left CCA images were obtained from a single volunteer subject (not part of the study group) as follows: (1) Before data collection was begun, the anatomic site was explored to find an acceptable location for measurement. The location of the scanhead on the skin was marked with a line drawn perpendicular to the long axis of the vessel. The scanhead was removed from the skin. (2) The scanhead was placed on the marked location, and an image was acquired. (3) The scanhead was removed from the skin. (4) Steps 2 and 3 were repeated a total of 10 times before moving to the next site. The thickness measurement algorithm (described later) was then applied 10 times to each image. To reduce carry-over effects in making the measurement, the operator did all first measurements across 10 images of each site, then all second measurements, etc. A total of 100 (10 images x 10 measurements per image) measurements were made for each
JOURNAL OF VASCULAR SURGERY April 1994
anatomic site; 200 variability measurements were made in all. For the WT measurements, volunteer adult subjects were recruited from the Seattle metropolitan area. Sources for the control group included University of Washington coworkers, their friends and relatives, and volunteers from a local community health fair. Subjects with intermittent claudication were recruited from previous studies of atherosclerosis conducted in this laboratory. All subjects completed a medical history questionnaire. Anldebrachial pressure indexes were obtained. Duplex examinations were performed with an Acuson 128 imaging system (Acuson, Mountain View, Calif.). B-mode ultrasound images of bilateral CCA were obtained at end diastole by selecting electrocardiographic-synchronous images from the cineloop memory of the instrument with the subject in a 15-degree Trendelenberg position (to increase venous dilation) with a 7.0 MHz linear array transducer. Following the method described by Beach et al. 6 the scanhead was positioned so that the diameters of the artery and vein were aligned along the central scanline of the image and then rotated 90 degrees to bring the artery-vein interface into longitudinal view. Images were obtained with a linear, flat depth gain setting so that all image areas maintained the same gain profile. Overall gain was set so that echogenic "backscatter" from the arterial and venous blood pools was detectable, with the total dynamic range set to 60 dB. Analysis of these images and review of the medical history and physical examination findings allowed stratification of the subjects into the following categories: a young control group (aged 20 to 29 years), no subject in this age group had any evidence of arterial disease), an older control group (aged 60 to 79 years), with no evidence of arterial occlusive disease by history, physical examination, or Doppler examination of the carotid, brachial, or femoral arteries), and an older group (aged 60 to 79 years, with clinical disease and reduced ankle-brachial pressure indexes, Doppler waveform evidence of lower extremity atherosclerosis, and a history of intermittent claudication). No B-mode image information was used to separate groups. In addition, subjects with a smoking history of greater than 5 packs/year were excluded from the control groups. Those subjects who had undergone carotid endarterectomy (n = 8) had the operative side excluded from analysis. No subjects with diabetes are included in any of the study groups. Among the claudicator group, 70% had a history of hypertension, 50% had
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
Hodges et al.
747
Fig. 1. Technique of wall thickness measurement. 2 x magnified B·mode image of arterial wall was presented to observer. Measurements were made on specified 5 mm segment by marking border of vein lumen, arterial lumen (inner and outer), and intima-medial border. With image scale factor, computer algorithm calculates wall thickness and intima-media thickness.
Table I. Subject description
Control (20-29 yrs) Control (60-79 yrs) Claudicator (60-79 yrs)
Women
Men
(no.)
(no.)
13 10 2
11
a history of coronary artery disease, 35% had a previous myocardial infarction, 30% had undergone previous coronary artery bypass grafting, and 45% gave a history of hyperlipidemia. A total of 60 subjects in these three groups form the basis for this report (Table I). The ultrasound images of the CCA were recorded on VHS videotape, digitized (PCVISIONplus Frame Grabber, Imaging Technology Inc., Woburn, Mass.), and stored on computer disk. The digitization process involved the selection of a region of the image encompassing a 2.5 em-wide wall complex segment of the ultrasound image at a site free of atherosclerotic plaque. The ultrasound dimensional scale factor and patient identification data were incorporated into the disk file. An observer trained in the evaluation of ultrasound images (11) analyzed these image regions using a computer-assisted protocol for manual measurement ofWT. The measurements were obtained by marking the leading and
7 17
Mean age (yrs)
Range (yrs)
27.0 66.0 69.5
22-29 60-74 61-79
trailing edges of the blood-lumen interface and the double-line region on the opposite wall (Fig. 1). It should be noted that because of anatomic relationships and the "trailing-edge effect" of ultrasonic imaging, IMT measurements are necessarily performed on the arterial wall opposite the WT measurement. With the ultrasound scale factor, the computer then calculates the marked dimension. An image grading system was applied to all images to allow classification of the images with respect to their overall quality and apparent resolution (Table II). The distribution of image grades for each subject group is given in Fig. 2. Examples of representative image grades are shown in Fig. 3. Presented data include only those images that could be evaluated with confidence (grades 1 to 4). After WT and IMT measurements were performed on all images and image quality grades were assigned, a single side was chosen at random to represent the CCA WT for each subject. If the selected side had an image that could
JOURNAL OF VASCULAR SURGERY April 1994
748 Hodges et aI.
25
20
15 • Number of Images
20·29 Control
o 60·79 Control 51 60· 79 Claudieator 10
2 Image Grade
Fig. 2. Image grade distribution. Specific criteria (Table II) were applied to determine quality of B-mode images. Poor images (grade 5) were excluded from analysis.
Table II. Image grade criteria Grade requirements
lmagegrade 1
2 3 4
5
Easy measurement with no doubt about point location or image structures. Excellent image quality. Very confident that measurement is correct. No doubt about image structures and most point locations. 1-2 points required judgment or interpolation. Very confident that measurement is correct. No doubt about image structures. Indistinct edges or 5-6 points required judgment or interpolation. Confident that measurement is correct. Measurement may contain error, or most points required judgment in placement. Image structure (double line) probable but not definite. Confident that measurement is a fair estimate of true value. Measurement doubtful. Uncertain of identity of image structures (double line). Difficult or essentially impossible to obtain measurement. Would consider measurement only a guess.
not be measured (i.e., grade 5), and the contralateral side could be measured, the contralateral side was substituted. This substitution occurred in five cases. Variability in the manual outlining technique was examined by comparing two trained observers over a range of image grades in the following ways. (1) 20 images were selected from all test subject groups at random to provide a range of image grades. These images compose the variability set. (2) The images were presented to each observer in random order within each set. (3) Total wall thickness and double-line measurements were performed on the set three times by each observer. Observers were blinded as to the image quality rating. In analyzing the results of this study, the following comparisons were made. (1) Image acquisition variability; (2) observer variability; (3) WT vs age and disease category; and (4) contribution ofIMT to total WT measurement. Statistical differences across
groups were evaluated with Kruskal-Wallis one-way rank analysis of variance with Dunn's adjustment for multiple comparisons. 7,8 RESULTS Ultrasound images consisted of common carotid arteries in the described age and disease groups. Doppler examination revealed no occluded common carotid arteries. Across-image acquisition variability for the single-subject study was assessed by computing a standard deviation of all first, second, third, etc., measurements, giving 10 standard deviations in total that were averaged. The WT standard deviation for the right common carotid artery was 0.05 mm, and for the left it was 0.08 mm. Observer WT measurement variability was determined by calculating a standard deviation of all measurements made. For WT measurements, intraobserver variability was 0.26 mm (SD), and interobserver variability was 0.28 mm (SD). For IMT measurements, intraobserver vari-
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
Hodges et at.
749
Fig. 3. Image grade examples: A, grade 1; B, grade 3; C, grade 5.
ability was 0.09 mm (SD), and interobserver variability was 0.15 mm (SD). Differences between sexes were sought by comparison within each group. For 20- to 29-year-old men, the mean CCA WT measurement was 1.51 mm ± 0.20 mm (SD), and for women in this age group the mean WT measurement was 1.43 mm ± 0.16 mm (SD) (p = 0.3235). In the 60- to 79-year-old control group, men had a mean CCA WT of 1.77 mm ± 0.20 mm (SD), while women measured 1.65 mm ± 0.11 mm (SD) (p = 0.1427). An insufficient number of women to allow statistical power was present in the group with disease. Although women in the control group had an average WT 5% to 7% lower than that of men in the control group, no statistically significant difference in wall thickness was noted, and the sexes were combined for all subsequent analyses. (After adjustment for body surface area (BSA) with linear regression, the WT differences between men and women were
reduced in both control groups, which suggests that the tendency for smaller WT measurements in women may be due to body habitus.) Comparison ofWT versus age and disease state is presented in Fig. 4. The median WT for young patients in the control group was 1.50 mm, for older patients in the control group 1.69 mm, and for those with symptomatic claudication 2.01 mm. Significance was demonstrated between control groups and between older patients in the control group and age-similarpatientswithclaudication(p S 0.05). To exclude the influence of sex as a confounding variable between older patients in the control group and patients with claudication, WT was compared (in a separate test) for men only in these two groups. Men with claudication had a larger WT measurement than did older men in the control group at the p = 0.06 level. If one subtracts the IMT measurement from the total WT measurement, the result represents the
JOURNAL OF VASeULAR SURGERY April 1994
750 Hodges et al.
p < 0.05
p < 0.05
4.00
3.00
wall Thickness (mm)
2.00
1.00
0.00
+-~~~~~--t--~~~~~-t-~~~--~-t-~~~~~--I
20-29 Control
60-79 Control
60-79 Claudicator
Fig. 4. Wall thickness by group (median [whisker] and intequartile range [box]). CCA total arterial wall thickness increases significantly with age and increases further with clinically significant lower extremity arterial disease.
extramedial (i.e., adventitial, connective tissue, and vein wall) contribution to total WT. A comparison of total WT, IMT, and the remaining tissue (WT - IMT, summarized as extramedial thickness) measurement is shown in Fig. 5. Note that as total wall thickness and IMT increase for each test group, the extramedial thickness increases in similar proportion. No statistical differences in IMT as a proportion of total WT were found.
DISCUSSION Several cellular mechanisms for atherosclerotic plaque formation have been proposed, among them abnormal intimal insudation oflipids, smooth muscle cellular hyperplasia, and activation of inflammatory cell mediator pathways.9 On a macroscopic level these theories describe a diffuse thickening of the arterial wall (or intima-media complex) before focal impingement on the arterial lumen by plaque. An important advance that may allow early intervention in this process has been the detection and quantification of subclinical disease with noninvasive methods. Interest in the use of ultrasound to detect and monitor early atherosclerotic changes in the arterial wall has been stimulated by Pignoli et al.,10 who described IMT measurement with B-mode ultrasound. The intima-media complex appears as a "double-line" pattern in the ultrasound image at the artery-blood interface. Pignoli et al. showed that the
measured dimensions of this ultrasonic pattern correlated well with pathologic measurement ofIMT,u This measurement technique has been used to determine the IMT of arterial walls in patients with hypercholesterolemia. 12 These patients have thickening of the intima-media complex in the common carotid artery when compared with the control group, suggesting that IMT may be used as a marker of subclinical arterial disease. Wofford et al. 13 have demonstrated a strong correlation between B-mode determination of carotid atherosclerotic severity and degree of coronary artery disease. Wendelhag, Wiklund, and Wikstrand14 have also confirmed that the IMT increases in patients with familial hypercholesterolemia when compared with age- and sex-matched control groups. Although there has been some debate as to the true nature of the double-line pattern/ 5017 the method described by Pignoli et al. 11 is currently being used in epidemiologic studies. 18 ,19 The importance of developing a standard method for arterial WT measurement by ultrasonic means was emphasized by Wikstrand and Wiklund20 in a recent summary. As will be described, however, there are limitations to the IMT measurement. The IMT is only a partial contributor to total arterial WT, and its measurement does not account for any observed change in the adventitia. Furthermore, it has been our experience that visualization of the double-line pattern is not always possible in a given patient or anatomic site. The pattern may be
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
Hodges et at.
751
2 1.8
.-
1.6 1.4 1.2 mm
55.90%
c:=:JIMT _
1
Extramedlal
--TotalWT 0.8 0.6
40.30%
0.4 0.2 0 20-29 Control
60-79 Control
60-79 Claudicator
Fig. 5. Wall thickness components. Extramedial wall tissue (adventitia, connective tissue, and vein wall) increases with age and disease state without alteration in its proportional contribution to total wall thickness.
obscured by overlying tissue, or it may not be visible because of an unfavorable ultrasound B-mode angle of incidence with respect to the deep arterial wall or as a consequence of poor image quality. A more easily obtained and complete measure of arterial wall dimensions would seem beneficial. Determination of total arterial WT by our method allows measurement under a wider range of conditions, because the measurement can be performed even if the IMT pattern cannot be resolved. A main advantage of measuring the total WT is that it allows assessment of all arterial wall component changes with disease. As demonstrated by Beach et al., 6 superficial femoral artery WT in subjects with diabetes who have arterial occlusive disease is significantly greater than in those without detectable disease. We have now examined a second arterial bed in subjects who do not have diabetes with similar results using this measurement technique. Potential sources of variability in ultrasonic imaging include image acquisition methods based on the device and sonographer, and image feature interpretation (intraobserver and interobserver) during the process of measurement. O'Leary et al. 21 examined the variability in the ultrasonic measurement of IMT in the common and internal carotid arteries and determined that image acquisition reproducibility was best for the CCA, whereas the lowest IMT measurement variability was obtained at that site. We have taken advantage of this finding for this
study. Salonen et al. 22 investigated this variability in CCA IMT measurements and found an interobserver coefficient of variance of 10.5% and an intraobserver coefficient of variance of5.4% to 5.8%. All images in this study were gathered by one sonographer using a single ultrasound machine, eliminating two potential sources of variability. From our variability data, sonographer image acquisition of the common carotid artery introduced a variability ofless than 0.1 mm to the total WT measurement. It should be noted that the WT measurement technique differs from the IMT in its approach. To obtain the full measure of the arterial wall, there must be a clear ultrasonic boundary on both the near and far walls. The arterial and venous blood/lumen interfaces provide the necessary acoustic impedance changes for visualization. While visualization of these structures is a more constant occurrence than the IMT "double-line" pattern, WT measurement does require the use of the ultrasonic "trailing edge" on the deep aspect of the wall. We are aware that the trailing edge does not define a specific anatomic boundary, but we do believe that it rrovides useful information. The trailing edge of an ultrasound pulse is determined by the electronic damping of the transducer, the decay time of the pulse modulated by the echo strength, and the presence of multiple echoes within the tissue under examination. The protocol for image acquisition in the present study requires that the image grey scale intensity setting be within
JOURNAL OF VASCULAR SURGERY
752 Hodges et aI.
the 60 dB dynamic range of the system without clipping the signal. The linear time gain compensation setting standardizes the trailing edge width across images. With these provisions the trailing edge will change reliably with the leading edge as arterial thickness increases. In the outlining protocol it was determined that image quality played a role in the ability of the observer to confidently locate the edges of the vessel walls. Some of the image quality differences observed may have been introduced in a systematic fashion, because the image appearance was not optimized during acquisition by the ultrasonographer but instead was fixed by the linear gain setting required by the Beach protocol. Without excluding the grade 5 images, intraobserver variability for total wall thickness was still as low as 0.26 mm, whereas interobserver variability was 0.28 mm. Because these values are less than the reported 0.3 to 0.5 mm depth resolution of the ultrasound device, we conclude that image acquisition and measurement performance by this method is both reliable and reproducible. In a related study Persson et al. 23 found that the intraobserver variability for IMT in the CCA was 0.08 ± 0.07 mm (N = 65), and interobserver variability for this measurement was 0.10 ± 0.10 mm. An interesting finding from this measurement technique is the lack of significant WT differences between sexes. Female WT tended to be smaller, and our sample size may have been too small to demonstrate a true difference, if one exists. This finding also points out that though male sex is an accepted risk factor for the development of atherosclerosis, its effect on wall thickness is probably smaller than is the effect of aging. It should be noted that there is considerable overlap of the group distributions of total wall thickness despite the fact that the group averages are statistically different from each other. This overlap means that the measurement of CCA total wall thickness is not an especially good predictor of age or of evidence of lower extremity arterial disease. In comparing the WT measurements between the young and older control groups, we found a statistically significant increase with age. This increase compares favorably with that noted by Poli et al.,12 who also demonstrated an increase in IMT with increasing age. This increase may relate to deposition of lipid, or it may reflect an adaptive response to hemodynamic stress over time. Of greater importance, perhaps, is the finding that WT increases with the presence of clinically manifest claudication, over and above the thickening effect of age alone. This
April 1994
finding supports the systemic atherogenesis theory, because a significant increase in arterial wall thickness was found in these patients at an anatomic site distant from their clinically evident lower extremity disease. Finally, it is of interest to note that the increase in WT is due not only to an increase in IMT but also to an increase in extramedial tissue thickness. Extramedial tissue refers to all tissue between the external elastic lamina of the arterial tunica media and the venous lumen. It is generally accepted that the vein wall and connective tissue are not affected by atherosclerosis. The changes in the adventitia associated with atherosclerosis have not been well described. Mitchell and Schwartz. 24 demonstrated an increase in adventitial cellular infiltration, which they related to plaque severity in several arterial beds. This cellular infiltrate consisted primarily of lymphocytes, and their relationship to plaque formation remains unknown. It is possible that the increased cellular infiltrate (or an associated inflammatory mediator response) contributes to adventitial thickening. In conclusion, the data indicate that total arterial WT increases as a function of age and lower extremity atherosclerotic disease state. We propose that this measurement technique may be useful in combination with IMT measurements as a means of determining arterial dimension in cases where the IMT may not be obtained. In addition, total WT measurement provides additional information about extramedial arterial change associated with atherosclerosis. REFERENCES 1. Ricotta JJ, Bryan FA, Bond MG, Kurtz A, O'Leary DR, Raines JK, Berson AS, Clouse ME, Calderon-Ortiz M, Toole JF, DeWeese JA, Smullens SN, Gustafson NF. Multicenter validation study of real-time (B-mode) ulttasound, arteriography, and pathologic examination. J VAse SURG 1987;6: 512-20. 2. Kohler TK, Nance DR, Cramer MM, Vandenberghe N, Sttandness DE Jr. Duplex scanning for diagnosis of aortoiliac and femoropopliteal disease: A prospective study. Circulation 1987;76:1074-80. 3. Goldberg B, Osttum B, Isard H. Ulttasonic aortography. JAMA 1966;198:353-8. 4. Blankenhorn DH, Chin HP, Stikwerda S, Bamberger J, Hestenes JD. Work in progress: Common carotid artery contours reconsttucted from three dimensions from parallel ulttasonic images. Radiology 1983;148:533-7. 5. Handa N, Matsumoto M, Maeda H, Hougaku H, Ogawa S, Fukunaga R, Yoneda S, Kimura K, Kamada T. Ultrasonic evaluation of early carotid atherosclerosis. Sttoke 1990;21: 1567-72. 6. Beach KW, Isaac CA, Phillips DJ, Sttandness DE Jr. An ulttasonic measurement of superficial femoral artery wall thickness. Ulttasound Med BioI 1989;15:723-8. 7. SPSSjPC Version 4.0. SPSS Inc. Chicago: 1990.
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
8. Hollander M, Wolfe DA. Nonparametric statistical methods. New York: John Wiley & Sons, 1973. 9. Glagov S, Zarins CK, Giddens DP, Davis HR Jr. Atherosclerosis: What is the nature of the plaque? In: Strandness DE Jr, Didisheim P, Clowes AW, Watson IT, eds. Vascular diseases: Current research and clinical applications. Harcourt Brace Jovanovich, 1987. 10. Pignoli P. Ultrasound B-mode imaging for arterial wall thickness measurement. Atherosclerosis Rev 1984;12:17784. 11. Pignoli P, Tremoli E, Poli A, Oreste P, Paoletti R. Intimal plus medial thickness of the arterial wall: A direct measurement with ultrasound imaging. Atherosclerosis 1986;74: 1399-406. 12. Poli A, Tremoli E, Colombo A, Sirtori M, Pignoli P, Paoletti P. Ultrasonic measurement of the common carotid artery wall thickness in hypercholesterolemic patients. Atherosclerosis 1988;70:253-61. 13. Wofford JL, Kahl F, Howard GR, McKinney WM, Toole JF, Crouse JR. Relation of extent of extracranial carotid artery atherosclerosis as measured by B-mode ultrasound to the extent of coronary atherosclerosis. Arterioscler Thromb 1991; 11:1786-94. 14. Wendelhag I, Wiklund 0, WikstrandJ. Arterial wall thickness in familial hypercholesterolemia. Atheroscler Thromb 1992; 12:70-7. 15. Nolsoe CP, Engel U, Karstrup S, Torp-Pederesen S, Garre K, Holm HH. The aortic wall: An in vitro study of the double-line pattern in high-resolution US. Radiology 1990; 175:387-90. 16. Fitzgerald P, Yock P. Letter to the editor. Radiology 1991;179:878-9.
Hodges et al.
753
17. Mercuri M. Letter to the editor. Radiology 1991;179:879. 18. Heiss G, Sharett AR, Barnes, Chambliss LE, Szklo M, Alzola C, ARIC Investigators. Carotid atherosclerosis measured by B-mode ultrasound in populations: Associations with cardiovascular risk factors in the ARIC study. Am J Epidemiol 1991;134:250-6. 19. Bond MG, Strickland HL, Wilmoth SK, Safrit A, Phillips R, Swstak L, for the MIDAS Research Group. Interventional clinical trials using noninvasive ultrasound end points: The multicenter isradipine/diuretic atherosclerosis study. J Cardiovasc Pharmacol 1990;15(suppl 1):S30-S33. 20. Wikstrand 1, Wiklund o. Frontiers in cardiovascular science: Quantitative measurements of atherosclerotic manifestations in humans. Arterioscler Thromb 1992;12:114-9. 21. O'Leary DH, Polak JF, Wolfson SK, Bond MG, Bommer W, Sheth S, Psaty BM, Sharrett AR, Manolio TA, on behalf of the CHS Collaborative Research Group. Use of sonography to evaluate carotid atherosclerosis in the elderly: The cardiovascular health study. Stroke 1991;22:1155-63. 22. Salonen R, Haapanen A, Salonen IT. Measurement of intima-media thickness of common carotid arteries with high-resolution B-mode ultrasonography: Inter- and intraobserver variability. Ultrasound Med Bioi 1991;17:225-30. 23. Persson 1, Stavenow L, Wikstrand J, Israelsson B, Formgren 1, Berglund G. Noninvasive quantification of atherosclerotic lesions. Arterioscler Thromb 1992;12:261-6. 24. Mitchell JRA, Schwartz CT. Arterial disease. Philadelphia: FA Davis Company, 1965:35-43.
Submitted April 7, 1993; accepted Aug. 21, 1993.
Two of the challenges to modern vascular medicine are the early detection and accurate staging of atherosclerosis. It is accepted that the use of clinical endpoints as the sole means of detection reflects only the effects of advanced disease. Arteriography can detect early atherosclerotic changes, but its cost, risk, and invasive nature preclude its use in patients as a From the Vascular Surgery Section, Department of Surgery, University of Washington Medical Center, Seattle, Wash. Supported by National Institutes of Health grant HL 42270-03 (NIH, Bethesda, Md.). Reprint requests: D. Eugene Strandness, Jr., MD, Professor and Chief, Division of Vascular Surgery, RF-25, University of Washington, Seattle, WA 98195. Copyright © 1994 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/94/$3.00 + 0 24/1/51295
screening test. Duplex scanning remains the most reliable 1,2 noninvasive modality for performing measurements of both blood flow velocity and arterial dimension. 3-5 One theory concerning the development of atherosclerotic lesions states that atherosclerosis is a systemic vascular disease beginning with a diffuse thickening of the arterial wall. If this thickening does indeed precede plaque formation, then it should occur at arterial sites not commonly affected by plaque, such as the common carotid arteries (CCA). Measurement ofCCA thickening may be useful in the detection of early arterial wall changes that may occur with modification of risk factors. Beach et al. 6 developed a wall thickness (WT) measurement strategy to determine the arterial lumen to vein lumen distance, a measurement that encom745
746 Hodges et ai.
passes the entire arterial wall. The measurements were confined to specific anatomic locations where the imaged artery and vein were adjacent and within the depth resolution of the ultrasound image. The vein wall thickness was assumed to be small, constant with respect to the artery, and unaffected by arterial atherosclerosis. Using this method, the authors demonstrated that the average superficial femoral artery and vein WT was significantly higher in subjects with diabetes who had peripheral arterial disease than in the control group. To test the hypothesis that arterial WT increases in a systemic pattern, the WT measurement method described by Beach et al. 6 has now been applied to a nondiabetic population of both healthy subjects and those with arterial occlusive disease. In a prospective manner we have determined the total arterial wall thickness and intima-media thickness (IMT) in the common carotid arteries of subjects aged 20 to 29 years and 60 to 79 years. This article presents total arterial wall thickness measurements and their correlation with age and disease state. METHODS We proposed to measure total arterial WT in defined subject groups. To validate any new measurement method, sources of variability should be characterized. To that end, before data were collected on subjects in the study, a variability study was performed to estimate the amount of uncertainty in the measurements contributed by the ultrasound technologist when acquiring the image. To eliminate the variability introduced by multiple subjects and focus only on that error contributed by the technologist, right and left CCA images were obtained from a single volunteer subject (not part of the study group) as follows: (1) Before data collection was begun, the anatomic site was explored to find an acceptable location for measurement. The location of the scanhead on the skin was marked with a line drawn perpendicular to the long axis of the vessel. The scanhead was removed from the skin. (2) The scanhead was placed on the marked location, and an image was acquired. (3) The scanhead was removed from the skin. (4) Steps 2 and 3 were repeated a total of 10 times before moving to the next site. The thickness measurement algorithm (described later) was then applied 10 times to each image. To reduce carry-over effects in making the measurement, the operator did all first measurements across 10 images of each site, then all second measurements, etc. A total of 100 (10 images x 10 measurements per image) measurements were made for each
JOURNAL OF VASCULAR SURGERY April 1994
anatomic site; 200 variability measurements were made in all. For the WT measurements, volunteer adult subjects were recruited from the Seattle metropolitan area. Sources for the control group included University of Washington coworkers, their friends and relatives, and volunteers from a local community health fair. Subjects with intermittent claudication were recruited from previous studies of atherosclerosis conducted in this laboratory. All subjects completed a medical history questionnaire. Anldebrachial pressure indexes were obtained. Duplex examinations were performed with an Acuson 128 imaging system (Acuson, Mountain View, Calif.). B-mode ultrasound images of bilateral CCA were obtained at end diastole by selecting electrocardiographic-synchronous images from the cineloop memory of the instrument with the subject in a 15-degree Trendelenberg position (to increase venous dilation) with a 7.0 MHz linear array transducer. Following the method described by Beach et al. 6 the scanhead was positioned so that the diameters of the artery and vein were aligned along the central scanline of the image and then rotated 90 degrees to bring the artery-vein interface into longitudinal view. Images were obtained with a linear, flat depth gain setting so that all image areas maintained the same gain profile. Overall gain was set so that echogenic "backscatter" from the arterial and venous blood pools was detectable, with the total dynamic range set to 60 dB. Analysis of these images and review of the medical history and physical examination findings allowed stratification of the subjects into the following categories: a young control group (aged 20 to 29 years), no subject in this age group had any evidence of arterial disease), an older control group (aged 60 to 79 years), with no evidence of arterial occlusive disease by history, physical examination, or Doppler examination of the carotid, brachial, or femoral arteries), and an older group (aged 60 to 79 years, with clinical disease and reduced ankle-brachial pressure indexes, Doppler waveform evidence of lower extremity atherosclerosis, and a history of intermittent claudication). No B-mode image information was used to separate groups. In addition, subjects with a smoking history of greater than 5 packs/year were excluded from the control groups. Those subjects who had undergone carotid endarterectomy (n = 8) had the operative side excluded from analysis. No subjects with diabetes are included in any of the study groups. Among the claudicator group, 70% had a history of hypertension, 50% had
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
Hodges et al.
747
Fig. 1. Technique of wall thickness measurement. 2 x magnified B·mode image of arterial wall was presented to observer. Measurements were made on specified 5 mm segment by marking border of vein lumen, arterial lumen (inner and outer), and intima-medial border. With image scale factor, computer algorithm calculates wall thickness and intima-media thickness.
Table I. Subject description
Control (20-29 yrs) Control (60-79 yrs) Claudicator (60-79 yrs)
Women
Men
(no.)
(no.)
13 10 2
11
a history of coronary artery disease, 35% had a previous myocardial infarction, 30% had undergone previous coronary artery bypass grafting, and 45% gave a history of hyperlipidemia. A total of 60 subjects in these three groups form the basis for this report (Table I). The ultrasound images of the CCA were recorded on VHS videotape, digitized (PCVISIONplus Frame Grabber, Imaging Technology Inc., Woburn, Mass.), and stored on computer disk. The digitization process involved the selection of a region of the image encompassing a 2.5 em-wide wall complex segment of the ultrasound image at a site free of atherosclerotic plaque. The ultrasound dimensional scale factor and patient identification data were incorporated into the disk file. An observer trained in the evaluation of ultrasound images (11) analyzed these image regions using a computer-assisted protocol for manual measurement ofWT. The measurements were obtained by marking the leading and
7 17
Mean age (yrs)
Range (yrs)
27.0 66.0 69.5
22-29 60-74 61-79
trailing edges of the blood-lumen interface and the double-line region on the opposite wall (Fig. 1). It should be noted that because of anatomic relationships and the "trailing-edge effect" of ultrasonic imaging, IMT measurements are necessarily performed on the arterial wall opposite the WT measurement. With the ultrasound scale factor, the computer then calculates the marked dimension. An image grading system was applied to all images to allow classification of the images with respect to their overall quality and apparent resolution (Table II). The distribution of image grades for each subject group is given in Fig. 2. Examples of representative image grades are shown in Fig. 3. Presented data include only those images that could be evaluated with confidence (grades 1 to 4). After WT and IMT measurements were performed on all images and image quality grades were assigned, a single side was chosen at random to represent the CCA WT for each subject. If the selected side had an image that could
JOURNAL OF VASCULAR SURGERY April 1994
748 Hodges et aI.
25
20
15 • Number of Images
20·29 Control
o 60·79 Control 51 60· 79 Claudieator 10
2 Image Grade
Fig. 2. Image grade distribution. Specific criteria (Table II) were applied to determine quality of B-mode images. Poor images (grade 5) were excluded from analysis.
Table II. Image grade criteria Grade requirements
lmagegrade 1
2 3 4
5
Easy measurement with no doubt about point location or image structures. Excellent image quality. Very confident that measurement is correct. No doubt about image structures and most point locations. 1-2 points required judgment or interpolation. Very confident that measurement is correct. No doubt about image structures. Indistinct edges or 5-6 points required judgment or interpolation. Confident that measurement is correct. Measurement may contain error, or most points required judgment in placement. Image structure (double line) probable but not definite. Confident that measurement is a fair estimate of true value. Measurement doubtful. Uncertain of identity of image structures (double line). Difficult or essentially impossible to obtain measurement. Would consider measurement only a guess.
not be measured (i.e., grade 5), and the contralateral side could be measured, the contralateral side was substituted. This substitution occurred in five cases. Variability in the manual outlining technique was examined by comparing two trained observers over a range of image grades in the following ways. (1) 20 images were selected from all test subject groups at random to provide a range of image grades. These images compose the variability set. (2) The images were presented to each observer in random order within each set. (3) Total wall thickness and double-line measurements were performed on the set three times by each observer. Observers were blinded as to the image quality rating. In analyzing the results of this study, the following comparisons were made. (1) Image acquisition variability; (2) observer variability; (3) WT vs age and disease category; and (4) contribution ofIMT to total WT measurement. Statistical differences across
groups were evaluated with Kruskal-Wallis one-way rank analysis of variance with Dunn's adjustment for multiple comparisons. 7,8 RESULTS Ultrasound images consisted of common carotid arteries in the described age and disease groups. Doppler examination revealed no occluded common carotid arteries. Across-image acquisition variability for the single-subject study was assessed by computing a standard deviation of all first, second, third, etc., measurements, giving 10 standard deviations in total that were averaged. The WT standard deviation for the right common carotid artery was 0.05 mm, and for the left it was 0.08 mm. Observer WT measurement variability was determined by calculating a standard deviation of all measurements made. For WT measurements, intraobserver variability was 0.26 mm (SD), and interobserver variability was 0.28 mm (SD). For IMT measurements, intraobserver vari-
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
Hodges et at.
749
Fig. 3. Image grade examples: A, grade 1; B, grade 3; C, grade 5.
ability was 0.09 mm (SD), and interobserver variability was 0.15 mm (SD). Differences between sexes were sought by comparison within each group. For 20- to 29-year-old men, the mean CCA WT measurement was 1.51 mm ± 0.20 mm (SD), and for women in this age group the mean WT measurement was 1.43 mm ± 0.16 mm (SD) (p = 0.3235). In the 60- to 79-year-old control group, men had a mean CCA WT of 1.77 mm ± 0.20 mm (SD), while women measured 1.65 mm ± 0.11 mm (SD) (p = 0.1427). An insufficient number of women to allow statistical power was present in the group with disease. Although women in the control group had an average WT 5% to 7% lower than that of men in the control group, no statistically significant difference in wall thickness was noted, and the sexes were combined for all subsequent analyses. (After adjustment for body surface area (BSA) with linear regression, the WT differences between men and women were
reduced in both control groups, which suggests that the tendency for smaller WT measurements in women may be due to body habitus.) Comparison ofWT versus age and disease state is presented in Fig. 4. The median WT for young patients in the control group was 1.50 mm, for older patients in the control group 1.69 mm, and for those with symptomatic claudication 2.01 mm. Significance was demonstrated between control groups and between older patients in the control group and age-similarpatientswithclaudication(p S 0.05). To exclude the influence of sex as a confounding variable between older patients in the control group and patients with claudication, WT was compared (in a separate test) for men only in these two groups. Men with claudication had a larger WT measurement than did older men in the control group at the p = 0.06 level. If one subtracts the IMT measurement from the total WT measurement, the result represents the
JOURNAL OF VASeULAR SURGERY April 1994
750 Hodges et al.
p < 0.05
p < 0.05
4.00
3.00
wall Thickness (mm)
2.00
1.00
0.00
+-~~~~~--t--~~~~~-t-~~~--~-t-~~~~~--I
20-29 Control
60-79 Control
60-79 Claudicator
Fig. 4. Wall thickness by group (median [whisker] and intequartile range [box]). CCA total arterial wall thickness increases significantly with age and increases further with clinically significant lower extremity arterial disease.
extramedial (i.e., adventitial, connective tissue, and vein wall) contribution to total WT. A comparison of total WT, IMT, and the remaining tissue (WT - IMT, summarized as extramedial thickness) measurement is shown in Fig. 5. Note that as total wall thickness and IMT increase for each test group, the extramedial thickness increases in similar proportion. No statistical differences in IMT as a proportion of total WT were found.
DISCUSSION Several cellular mechanisms for atherosclerotic plaque formation have been proposed, among them abnormal intimal insudation oflipids, smooth muscle cellular hyperplasia, and activation of inflammatory cell mediator pathways.9 On a macroscopic level these theories describe a diffuse thickening of the arterial wall (or intima-media complex) before focal impingement on the arterial lumen by plaque. An important advance that may allow early intervention in this process has been the detection and quantification of subclinical disease with noninvasive methods. Interest in the use of ultrasound to detect and monitor early atherosclerotic changes in the arterial wall has been stimulated by Pignoli et al.,10 who described IMT measurement with B-mode ultrasound. The intima-media complex appears as a "double-line" pattern in the ultrasound image at the artery-blood interface. Pignoli et al. showed that the
measured dimensions of this ultrasonic pattern correlated well with pathologic measurement ofIMT,u This measurement technique has been used to determine the IMT of arterial walls in patients with hypercholesterolemia. 12 These patients have thickening of the intima-media complex in the common carotid artery when compared with the control group, suggesting that IMT may be used as a marker of subclinical arterial disease. Wofford et al. 13 have demonstrated a strong correlation between B-mode determination of carotid atherosclerotic severity and degree of coronary artery disease. Wendelhag, Wiklund, and Wikstrand14 have also confirmed that the IMT increases in patients with familial hypercholesterolemia when compared with age- and sex-matched control groups. Although there has been some debate as to the true nature of the double-line pattern/ 5017 the method described by Pignoli et al. 11 is currently being used in epidemiologic studies. 18 ,19 The importance of developing a standard method for arterial WT measurement by ultrasonic means was emphasized by Wikstrand and Wiklund20 in a recent summary. As will be described, however, there are limitations to the IMT measurement. The IMT is only a partial contributor to total arterial WT, and its measurement does not account for any observed change in the adventitia. Furthermore, it has been our experience that visualization of the double-line pattern is not always possible in a given patient or anatomic site. The pattern may be
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
Hodges et at.
751
2 1.8
.-
1.6 1.4 1.2 mm
55.90%
c:=:JIMT _
1
Extramedlal
--TotalWT 0.8 0.6
40.30%
0.4 0.2 0 20-29 Control
60-79 Control
60-79 Claudicator
Fig. 5. Wall thickness components. Extramedial wall tissue (adventitia, connective tissue, and vein wall) increases with age and disease state without alteration in its proportional contribution to total wall thickness.
obscured by overlying tissue, or it may not be visible because of an unfavorable ultrasound B-mode angle of incidence with respect to the deep arterial wall or as a consequence of poor image quality. A more easily obtained and complete measure of arterial wall dimensions would seem beneficial. Determination of total arterial WT by our method allows measurement under a wider range of conditions, because the measurement can be performed even if the IMT pattern cannot be resolved. A main advantage of measuring the total WT is that it allows assessment of all arterial wall component changes with disease. As demonstrated by Beach et al., 6 superficial femoral artery WT in subjects with diabetes who have arterial occlusive disease is significantly greater than in those without detectable disease. We have now examined a second arterial bed in subjects who do not have diabetes with similar results using this measurement technique. Potential sources of variability in ultrasonic imaging include image acquisition methods based on the device and sonographer, and image feature interpretation (intraobserver and interobserver) during the process of measurement. O'Leary et al. 21 examined the variability in the ultrasonic measurement of IMT in the common and internal carotid arteries and determined that image acquisition reproducibility was best for the CCA, whereas the lowest IMT measurement variability was obtained at that site. We have taken advantage of this finding for this
study. Salonen et al. 22 investigated this variability in CCA IMT measurements and found an interobserver coefficient of variance of 10.5% and an intraobserver coefficient of variance of5.4% to 5.8%. All images in this study were gathered by one sonographer using a single ultrasound machine, eliminating two potential sources of variability. From our variability data, sonographer image acquisition of the common carotid artery introduced a variability ofless than 0.1 mm to the total WT measurement. It should be noted that the WT measurement technique differs from the IMT in its approach. To obtain the full measure of the arterial wall, there must be a clear ultrasonic boundary on both the near and far walls. The arterial and venous blood/lumen interfaces provide the necessary acoustic impedance changes for visualization. While visualization of these structures is a more constant occurrence than the IMT "double-line" pattern, WT measurement does require the use of the ultrasonic "trailing edge" on the deep aspect of the wall. We are aware that the trailing edge does not define a specific anatomic boundary, but we do believe that it rrovides useful information. The trailing edge of an ultrasound pulse is determined by the electronic damping of the transducer, the decay time of the pulse modulated by the echo strength, and the presence of multiple echoes within the tissue under examination. The protocol for image acquisition in the present study requires that the image grey scale intensity setting be within
JOURNAL OF VASCULAR SURGERY
752 Hodges et aI.
the 60 dB dynamic range of the system without clipping the signal. The linear time gain compensation setting standardizes the trailing edge width across images. With these provisions the trailing edge will change reliably with the leading edge as arterial thickness increases. In the outlining protocol it was determined that image quality played a role in the ability of the observer to confidently locate the edges of the vessel walls. Some of the image quality differences observed may have been introduced in a systematic fashion, because the image appearance was not optimized during acquisition by the ultrasonographer but instead was fixed by the linear gain setting required by the Beach protocol. Without excluding the grade 5 images, intraobserver variability for total wall thickness was still as low as 0.26 mm, whereas interobserver variability was 0.28 mm. Because these values are less than the reported 0.3 to 0.5 mm depth resolution of the ultrasound device, we conclude that image acquisition and measurement performance by this method is both reliable and reproducible. In a related study Persson et al. 23 found that the intraobserver variability for IMT in the CCA was 0.08 ± 0.07 mm (N = 65), and interobserver variability for this measurement was 0.10 ± 0.10 mm. An interesting finding from this measurement technique is the lack of significant WT differences between sexes. Female WT tended to be smaller, and our sample size may have been too small to demonstrate a true difference, if one exists. This finding also points out that though male sex is an accepted risk factor for the development of atherosclerosis, its effect on wall thickness is probably smaller than is the effect of aging. It should be noted that there is considerable overlap of the group distributions of total wall thickness despite the fact that the group averages are statistically different from each other. This overlap means that the measurement of CCA total wall thickness is not an especially good predictor of age or of evidence of lower extremity arterial disease. In comparing the WT measurements between the young and older control groups, we found a statistically significant increase with age. This increase compares favorably with that noted by Poli et al.,12 who also demonstrated an increase in IMT with increasing age. This increase may relate to deposition of lipid, or it may reflect an adaptive response to hemodynamic stress over time. Of greater importance, perhaps, is the finding that WT increases with the presence of clinically manifest claudication, over and above the thickening effect of age alone. This
April 1994
finding supports the systemic atherogenesis theory, because a significant increase in arterial wall thickness was found in these patients at an anatomic site distant from their clinically evident lower extremity disease. Finally, it is of interest to note that the increase in WT is due not only to an increase in IMT but also to an increase in extramedial tissue thickness. Extramedial tissue refers to all tissue between the external elastic lamina of the arterial tunica media and the venous lumen. It is generally accepted that the vein wall and connective tissue are not affected by atherosclerosis. The changes in the adventitia associated with atherosclerosis have not been well described. Mitchell and Schwartz. 24 demonstrated an increase in adventitial cellular infiltration, which they related to plaque severity in several arterial beds. This cellular infiltrate consisted primarily of lymphocytes, and their relationship to plaque formation remains unknown. It is possible that the increased cellular infiltrate (or an associated inflammatory mediator response) contributes to adventitial thickening. In conclusion, the data indicate that total arterial WT increases as a function of age and lower extremity atherosclerotic disease state. We propose that this measurement technique may be useful in combination with IMT measurements as a means of determining arterial dimension in cases where the IMT may not be obtained. In addition, total WT measurement provides additional information about extramedial arterial change associated with atherosclerosis. REFERENCES 1. Ricotta JJ, Bryan FA, Bond MG, Kurtz A, O'Leary DR, Raines JK, Berson AS, Clouse ME, Calderon-Ortiz M, Toole JF, DeWeese JA, Smullens SN, Gustafson NF. Multicenter validation study of real-time (B-mode) ulttasound, arteriography, and pathologic examination. J VAse SURG 1987;6: 512-20. 2. Kohler TK, Nance DR, Cramer MM, Vandenberghe N, Sttandness DE Jr. Duplex scanning for diagnosis of aortoiliac and femoropopliteal disease: A prospective study. Circulation 1987;76:1074-80. 3. Goldberg B, Osttum B, Isard H. Ulttasonic aortography. JAMA 1966;198:353-8. 4. Blankenhorn DH, Chin HP, Stikwerda S, Bamberger J, Hestenes JD. Work in progress: Common carotid artery contours reconsttucted from three dimensions from parallel ulttasonic images. Radiology 1983;148:533-7. 5. Handa N, Matsumoto M, Maeda H, Hougaku H, Ogawa S, Fukunaga R, Yoneda S, Kimura K, Kamada T. Ultrasonic evaluation of early carotid atherosclerosis. Sttoke 1990;21: 1567-72. 6. Beach KW, Isaac CA, Phillips DJ, Sttandness DE Jr. An ulttasonic measurement of superficial femoral artery wall thickness. Ulttasound Med BioI 1989;15:723-8. 7. SPSSjPC Version 4.0. SPSS Inc. Chicago: 1990.
JOURNAL OF VASCULAR SURGERY Volume 19, Number 4
8. Hollander M, Wolfe DA. Nonparametric statistical methods. New York: John Wiley & Sons, 1973. 9. Glagov S, Zarins CK, Giddens DP, Davis HR Jr. Atherosclerosis: What is the nature of the plaque? In: Strandness DE Jr, Didisheim P, Clowes AW, Watson IT, eds. Vascular diseases: Current research and clinical applications. Harcourt Brace Jovanovich, 1987. 10. Pignoli P. Ultrasound B-mode imaging for arterial wall thickness measurement. Atherosclerosis Rev 1984;12:17784. 11. Pignoli P, Tremoli E, Poli A, Oreste P, Paoletti R. Intimal plus medial thickness of the arterial wall: A direct measurement with ultrasound imaging. Atherosclerosis 1986;74: 1399-406. 12. Poli A, Tremoli E, Colombo A, Sirtori M, Pignoli P, Paoletti P. Ultrasonic measurement of the common carotid artery wall thickness in hypercholesterolemic patients. Atherosclerosis 1988;70:253-61. 13. Wofford JL, Kahl F, Howard GR, McKinney WM, Toole JF, Crouse JR. Relation of extent of extracranial carotid artery atherosclerosis as measured by B-mode ultrasound to the extent of coronary atherosclerosis. Arterioscler Thromb 1991; 11:1786-94. 14. Wendelhag I, Wiklund 0, WikstrandJ. Arterial wall thickness in familial hypercholesterolemia. Atheroscler Thromb 1992; 12:70-7. 15. Nolsoe CP, Engel U, Karstrup S, Torp-Pederesen S, Garre K, Holm HH. The aortic wall: An in vitro study of the double-line pattern in high-resolution US. Radiology 1990; 175:387-90. 16. Fitzgerald P, Yock P. Letter to the editor. Radiology 1991;179:878-9.
Hodges et al.
753
17. Mercuri M. Letter to the editor. Radiology 1991;179:879. 18. Heiss G, Sharett AR, Barnes, Chambliss LE, Szklo M, Alzola C, ARIC Investigators. Carotid atherosclerosis measured by B-mode ultrasound in populations: Associations with cardiovascular risk factors in the ARIC study. Am J Epidemiol 1991;134:250-6. 19. Bond MG, Strickland HL, Wilmoth SK, Safrit A, Phillips R, Swstak L, for the MIDAS Research Group. Interventional clinical trials using noninvasive ultrasound end points: The multicenter isradipine/diuretic atherosclerosis study. J Cardiovasc Pharmacol 1990;15(suppl 1):S30-S33. 20. Wikstrand 1, Wiklund o. Frontiers in cardiovascular science: Quantitative measurements of atherosclerotic manifestations in humans. Arterioscler Thromb 1992;12:114-9. 21. O'Leary DH, Polak JF, Wolfson SK, Bond MG, Bommer W, Sheth S, Psaty BM, Sharrett AR, Manolio TA, on behalf of the CHS Collaborative Research Group. Use of sonography to evaluate carotid atherosclerosis in the elderly: The cardiovascular health study. Stroke 1991;22:1155-63. 22. Salonen R, Haapanen A, Salonen IT. Measurement of intima-media thickness of common carotid arteries with high-resolution B-mode ultrasonography: Inter- and intraobserver variability. Ultrasound Med Bioi 1991;17:225-30. 23. Persson 1, Stavenow L, Wikstrand J, Israelsson B, Formgren 1, Berglund G. Noninvasive quantification of atherosclerotic lesions. Arterioscler Thromb 1992;12:261-6. 24. Mitchell JRA, Schwartz CT. Arterial disease. Philadelphia: FA Davis Company, 1965:35-43.
Submitted April 7, 1993; accepted Aug. 21, 1993.