Ultrasound Morphology Classification of the Arterial Wall and Cardiovascular Events in a 6-Year Follow-up Study
A 6-year follow-up based on an arterial morphology classification defined with an ultrasound assessment of carotid and femoral artery bifurcations was conducted on 2322 asymptomatic subjects. Four morphology classes were considered. When 2000 subjects (86% of total subjects; 1124 males, 876 females) completed a 6-year follow-up, the study was terminated. At 6 years, no cardiovascular events were observed in subjects who were in class I (80.05% of the population sample) at inclusion; there were 69 events in classes II, III, and IV (19.95% of the population; incidence, 17.3%); 59 events, including the five deaths, occurred in classes III and IV (10.85% of the population), producing an event incidence of 27.2%. The increased event rate in classes II, III, and IV was significant (log-rank test; P<.05, P<.025, and P<.025, respectively). Thus, the arterial morphology classification identified 19.95% of the population (subjects in classes II, III, and IV) in which all events occurred. There was a higher (P<.05) rate of progression of altered arterial morphology in 6 years in classes III (26.5% of subjects progressed) and IV (41.9% progressed) than in classes I and II. The total number of cigarette-years was higher (P<.05) in classes II, III, and IV than in class I. In conclusion, the ultrasound-based arterial classification was useful in selecting from the population sample 80.05% of subjects (class I) who remained event-free for 6 years. All events occurred in class II, III, and IV subjects (19.95%), and all five deaths (0.25% of the population) occurred in classes III and IV (10.85% of the sample).
Reprint requests to Dr G. Belcaro, Via Vespucci 65, 65100 Pescara, Italy.
- Received June 22, 1995.
- Revision received January 19, 1996.
B-mode ultrasound is used increasingly as a noninvasive method to study subclinical atherosclerosis. Increased intima-media thickness (IMT) in the carotid artery is an indication of the presence of atherosclerosis involving the coronary arteries1 2 3 4 5 and appears to be related to the presence and severity of risk factors.3 6 7 8 9
It has recently been demonstrated that the British Regional Heart Study (BRHS) risk score—which is derived from the evaluation of risk factors such as smoking, systolic blood pressure, previous ischemic heart disease or diabetes mellitus, family history, and the presence of angina10 —is linearly related to the IMT of the common carotid artery (r=.56).6 Although the BRHS score can identify a population with a high incidence of events on follow-up, only 40% of the total events occur in this high-risk subgroup, while 60% occur in the rest of the population.6 Thus, risk factors, while important etiologically,11 12 13 cannot be considered an ideal method for screening a normal population of asymptomatic individuals for the identification of subjects at high risk of cardiovascular events.6 10
In addition to measurement of the IMT, which has mainly been used to study atherosclerosis progression and incidence in epidemiological studies,14 15 16 17 a morphological classification of arterial wall changes7 8 18 19 20 21 has been developed. This classification is based not only on IMT changes but also on the morphology of these changes and the presence and size of plaques at both carotid and femoral bifurcations (Table 1⇓).18 Half of the individuals with such plaques have a normal common carotid IMT,19 and one study (a 4-year follow-up) has indicated that the morphological classification may be a better indication of the risk of cardiovascular events than IMT measurement.20 21
The aim of the present study was to use the ultrasonic classification of arterial wall changes (Table 1⇑) to divide a population of asymptomatic individuals into subgroups on the basis of arterial morphology and to follow up the groups to determine the cardiovascular event rates associated with each class at inclusion into a 6-year follow-up.
A randomly selected population sample of 2322 healthy, asymptomatic subjects living in the Pescara and Chieti provinces (central-east Italy) was recruited in local hospital and university clinics and institutes from the nursing, technical, and administrative staff, employees, and their relatives. The recruitment period was 24 months, and the study was terminated when 2000 subjects completed a 6-year follow-up. Subjects were aged 30 through 70 years; all gave their informed consent to the study.
All subjects with clinically significant disease (requiring any treatment), a previous history of treatment and/or hospital admission for cardiovascular, renal, or metabolic problems (including obesity), or chronic or genetic diseases were excluded. Since the major aim of the study was to evaluate the occurrence of cardiovascular events in asymptomatic, healthy subjects, we did not include diabetic, hyperlipidemic, or hypertensive subjects to exclude the possible effects of these conditions (and the potential effect on any chronic drug treatment) on the development of cardiovascular events. Subjects with an ankle brachial index <1.0 were also excluded.
Resting blood pressure was measured phonographically in the right arm after 30 minutes of supine rest at the time of the ultrasound examination. A standard sphygmomanometer was used. The measurements were calculated to the nearest 1 mm Hg, and blood pressure was considered to be the mean of two recordings.15 22 Subjects were considered hypertensive and were excluded on the basis of (1) a history of clinical hypertension, (2) systolic blood pressure (average of four measurements) >160 mm Hg or mean diastolic blood pressure ≥95 mm Hg, or (3) use of antihypertensive medication. In doubtful cases, a 24-hour ambulatory monitoring system (Medilog) was used to diagnose hypertension.23
Information on smoking habits was obtained from a self-administered questionnaire. The total number of years of smoking was multiplied by the average number of cigarettes smoked daily. The product of this calculation was called cigarette-years.15
Risk Factor Assessment and Biochemical Analysis
Risk factor measurements were carried out at the baseline, inclusion examination and repeated every year to exclude diabetic, hyperlipidemic (cholesterol >5.2 mmol/L at three separate measurements), and hypertensive subjects.
Blood was drawn 2 weeks before or after the ultrasonographic scanning. The main lipoprotein fractions (VLDL, LDL, and HDL) were measured. Only subjects with blood lipid values within the normal range were included (total cholesterol <5.2 mmol/L; HDL cholesterol <0.9 mmol/L; triglycerides <213 mmol/L).
High-Resolution Ultrasound Evaluation of the Arterial Wall
Both carotid and femoral arterial bifurcations were studied with an ATL Ultramark 4 duplex scanner using a high-resolution, 7.5-MHz linear transducer. After localizing the carotid and femoral bifurcations by a transverse scan, the probe was rotated 90° to obtain and record a longitudinal image of both the anterior and posterior walls. The carotid artery was evaluated for a length of about 3 cm (1.5 cm proximal and 1.5 cm distal to the flow divider).18 20 21 The femoral artery was examined at the femoral bifurcation and scanned for a length of 3 cm (1.5 cm proximally and distally to the flow divider). By this technique, the three ultrasonic vessel wall layers (intima-media, adventitia, and periadventitia) were clearly visible. Technical ultrasound parameters (dynamic range, depth range, power, reject, edge, gray scale, and smooth) were maintained constant.
The previously described ultrasound-based morphological classification8 18 19 included five classes, corresponding to five scores7 8 18 19 ranging between 0 and 8 for each artery (plus a sixth class including symptomatic subjects). In the present study, the previously described classes I and II were merged into one class, as it was not possible to differentiate class I from II on the basis of histology and/or association with the subsequent incidence of cardiovascular events. Therefore, only four classes (scores 2 through 8) were used in the present study (see Table 1⇑). A plaque without hemodynamic disturbance (class III), as indicated in Table 1⇑, was arbitrarily defined as a localized IMT >2 mm, and was associated with increased ultrasound density (echogenicity).19
The subject's ultrasound score was obtained by adding the scores of the four arteries studied. During the first year of this study, all classifications were revised by studying VHS tape recordings of the scans after the patient had left. However, after the first 12 months, all class attributions were established during the scan directly from on-line screen evaluation, as no difference was found between on-line, real-time classification and VHS-derived classification.24
All operators performing the high-resolution scanning were physicians with at least 3 years of postgraduate clinical experience who were familiar with noninvasive investigations and had personally studied at least 3000 patients with vascular problems by using B-mode ultrasound, duplex scanning, and/or color duplex scanning before the beginning of this study.
Intraobserver and Interobserver Variability
Intraobserver variability was evaluated as described previously.24 One hundred subjects (20 from each of the original five classes, four arteries per patient) were examined on two different occasions within 10 to 15 days to estimate the intraobserver variability of ultrasound tape recording and class attribution. The two recordings and class attributions were performed by the same physician, who was unaware of the result of the first examination. The coefficient of variation was on average 5% (4 arteries incorrectly classified in the lower or higher class out of 80 arteries in each class).
The results from a study of interobserver variability of ultrasound classification have already been reported.19 24 At the beginning of this study, interobserver variability was comparable to the intraobserver variability or lower. The error in class attribution in 100 recordings (100 subjects) blindly read and classified by four different observers who were unaware of the identity of the subjects under evaluation was 2.75% (11 classification discrepancies of one arterial morphology class out of 400 recordings studied).
Plan of the Study and End Points
At inclusion, all subjects received a complete clinical cardiovascular assessment, including ankle pressure index and resting electrocardiogram. Each individual was placed in a class (I through IV) according to the worst artery (Table 1⇑) at inclusion.
End points were considered to be (1) the development of cardiovascular events and (2) the progression from the class at inclusion to the next, more advanced class.
Exit points were death from causes other than cardiovascular events and loss to follow-up.
All subjects developing signs and symptoms due to atherosclerosis were considered to be in a further, more advanced class (defined as class V in the present study and equivalent to class VI in previous studies20 21 25 ). In this study of asymptomatic subjects, class V subjects were not included.
Cardiovascular Events and Deaths
A cardiovascular event was defined as the development of cardiovascular signs and symptoms and complications related to atherosclerosis. Clinically relevant events in this study were considered to be those requiring hospital admission and receiving a confirmed diagnosis. All subjects who experienced events meeting these criteria required drug treatment after the episode and were excluded from the follow-up study. Unconfirmed diagnoses and minor events requiring neither hospital admission nor treatment were not considered as end point cardiovascular events, and subjects in this category continued to be followed up. Deaths documented as being of cardiovascular origin related to atherosclerosis were considered as cardiovascular end points. The passage to a symptomatic status by definition advanced subjects to the highest class (V).
The arterial scanning was repeated every year. Considering the highest classification (worst of four arteries) as the patient's class, the rate of progression in 6 years was defined as the percentage of subjects progressing to the next, more advanced class.
All statistics were analyzed using SPSS for Windows. The Mann-Whitney U test was used for comparison among the groups, and 95% confidence intervals were also calculated. A trend in the contingency table test was used to analyze the differences between the ultrasound score distribution and the occurrence of clinical cardiovascular events. Percentages of smokers (never, past, or current) were compared by a χ2 test.
The log-rank test was used to test the difference in the event-free cumulative survival rates in each class within the 6-year follow-up. Two-sided probability values <.05 were regarded as statistically significant.
A total of 2437 asymptomatic subjects (aged 30 through 70 years) were contacted and seen for possible inclusion in this study. However, 115 either declined or were considered unsuitable according to the exclusion criteria, so that 2322 were admitted to the study. Details of anthropometric data of the study population are shown in Table 2⇓. The distribution of subjects according to the ultrasonic arterial classes at inclusion, the incidence of cardiovascular events, and progression of ultrasonic arterial changes during the follow-up period are summarized in Table 3⇓. The study was terminated when 2000 (86%; 1124 males, 876 females) of the 2322 subjects (1311 males and 1011 females) completed the 6-year follow-up.
At 6 years, cardiovascular events had not been observed in any subjects in class I, which constituted 80.05% of the study population. The incidence of cardiovascular events was 5.5% in class II, 18.4% in class III, and 42% in class IV, numbering a total of 69 events, corresponding to an event incidence of 17.3% for classes II through IV. However, 59 of the events, including the five deaths, occurred in classes III and IV (10.85% of the population), producing an event incidence of 27.2% in those two classes (Table 4⇓). Details of the 69 cumulative events are shown in Table 5⇓. The cumulative event-free survival rate of the different classes is shown in the Figure⇓. The decreased event-free rate in groups II, III, and IV in comparison with class I was significant (log-rank test; P<.05, P<.025, and P<.025, respectively).
A total of 106 subjects progressed to a higher class within 6 years (Table 3⇑). The more advanced the initial classification at inclusion, the higher was the incidence of progression (from a 7% rate of progression in class I to 41.9% in class IV; P<.02 between classes I+II and classes III+IV).
The total number of cigarette-years was significantly higher (P<.05) in subjects in classes II, III, and IV than in those in class I. Also, the average number of cigarette-years in class II plus III and IV was higher than the average number of cigarette-years in class I (P<.05).
The average ultrasound score in each class at inclusion is shown in Table 3⇑. Individuals in class II had scores between 7 and 14; those in class III, between 14 and 20; and those in class IV, >22. Events occurred only in subjects with ultrasound scores >12 at inclusion. Therefore, both the worst arterial class in the single individual and the best ultrasound score (including four arteries) at inclusion appear to be related to the occurrence of cardiovascular events in the following 6 years.
According to these findings, the ultrasound score appears to be closely related to the classification but gives a more global idea of the involvement of the four arteries (and possibly of all arteries). Both class at inclusion, based on the worst artery, and ultrasound score were related to the occurrence of cardiovascular events.
Subclinical atherosclerotic lesions are observed in the arterial wall even in young adults and children, although the prevalence of subclinical atheromas is generally related to age.17 19 21 It has been observed that ≈10% of men have ultrasonographic signs of early atherosclerosis at the age of 40, while such lesions are observed in over 80% of subjects by 60 years of age.17 The clinical significance of early atherosclerotic lesions in predicting cardiovascular events due to atherosclerosis is still unclear.
Atherosclerosis progression studies using angiography have determined the incidence of advanced atherosclerotic lesions and plaques and their correlation with risk factors,26 but studies evaluating progression in early phases of atherosclerosis cannot be based on repeated angiography because small arterial lesions may not be visible on angiograms. Also, population studies using angiography are not feasible, considering costs and risks.17
Ultrasonographic assessment has been used successfully in the study of the prevalence of extracoronary atherosclerosis and its progression.11 12 13 17 20 21 The ultrasonographic assessment of atherosclerosis is practical, reliable, and cost effective for population studies and is also suitable for monitoring atherosclerosis progression or regression.2 3 6 7 8 14 24 Ultrasound imaging of the common carotid artery and IMT measurement have been used in progression studies and in studies comparing IMT and risk factors, but subclinical atherosclerosis appears much more often at bifurcations (ie, carotid or femoral) than in the common carotid artery.20 21 In previous studies20 21 it was observed that 65% of asymptomatic subjects in classes II, III, and IV had a normal (according to age) common carotid IMT (<1 mm). A normal common carotid IMT was observed in 74% of subjects in class III, in 54% of subjects in class IV, and in 44% of subjects in class V. It thus may be possible to have a normal common carotid IMT and still have bifurcation plaques that may in a short period (1 or 2 years) cause signs and symptoms.7 20 21 Furthermore, 30% of subjects with a normal carotid artery have femoral atherosclerotic lesions. Therefore, the assessment of both femoral and carotid bifurcations offers a global view of the atherosclerotic status of the subject15 and is possibly a better and earlier indicator of subclinical atherosclerosis8 18 21 than common carotid IMT measurements.
There is clearly a need for a simple, noninvasive test to screen asymptomatic population subjects at high risk of cardiovascular events in the next few years.20 25 According to the results of our study, the ultrasound classification of subclinical atherosclerotic wall changes separates asymptomatic subjects into groups at different degrees of risk for cardiovascular events. Previous studies8 19 20 21 evaluating ultrasound arterial morphology have indicated that arterial wall morphology is predictive of the presence of silent ischemic coronary disease (electrocardiographic effort tests revealed a progressive incidence of silent coronary ischemic disease from 0% in class I to 45% in class V). In addition, subjects in classes IV and V have a significantly increased risk of cardiovascular events in the upcoming years and an increased rate of progression of atherosclerosis.
A low-cost, noninvasive population screening method for subclinical atherosclerosis must isolate from the population subjects at high risk of cardiovascular events.27 Screening for blood lipids is important but indicates only a limited factor in a multifactorial disease.28 Screening for atherosclerosis causing ischemic heart disease by measuring apoB alone or with apoA-I and apo(a) also appears to be a poor method to discriminate subjects at high risk who may benefit from drug treatment or lifestyle changes.29 Combinations of tests (ie, electrocardiogram, cardiac ultrasound, carotid artery wall thickness and stenosis, decreased ankle brachial blood pressure, assessment with a Rose questionnaire for angina or intermittent claudication) are expensive and more useful to assess the prevalence of advanced atherosclerosis than to screen the population with early subclinical atherosclerosis at risk of future events.30 The ankle-arm Doppler index has been used as a marker of atherosclerosis. This simple screening test has suggested that even minor decrease in the index indicates an increased risk of cardiovascular disease.31 However, this test is more useful for advanced subclinical atherosclerosis, since in the early subclinical phases of atherosclerosis (ie, classes I to IV in the present classification), the ankle-arm index is normal (>1.0) in 93% of subjects.21
The average cost of each carotid-femoral scan20 21 was 12 European Currency Units (equivalent to $15 US) per patient (including equipment, training, and staff costs), which may indicate a role for the ultrasound arterial classification as a screening technique in selecting from the asymptomatic population those prone to cardiovascular events or faster atherosclerosis progression.25
In conclusion, this study indicates that the prevalence of future cardiovascular events appears to be related to arterial morphology classification and ultrasound arterial score. Thus, arterial morphology classification may be considered as a low-cost screening method to assess early, subclinical atherosclerosis. The ultrasound arterial score is closely related to the classification but gives a more global view of the involvement of the four arteries (and possibly of all arteries). Both the class at inclusion, based on the worst artery, and the score at inclusion are related to the occurrence of events. In some studies,25 the ultrasound score (which represents an average of the arteriosclerotic involvement of four arteries) appears to have a better correlation with cardiovascular events and the occurrence of occult coronary disease than the classification of the worst artery. The main application of the ultrasound score is in its relationship with age (ie, for normal subjects around the age of 50, the average score has been defined in different populations. An increase in score over the average may indicate a more advanced arteriosclerotic involvement than expected for that age and suggest a more careful evaluation of risk factors). This method may be useful in early selection from the general population of subjects likely to suffer from future cardiovascular events and who need a more careful follow-up and possibly secondary prevention.
This study has been partially supported with annual grants by the MURST (Italian Ministry of Scientific and Technological Research).
Salonen JT, Salonen R. Ultrasound B-mode imaging in observational studies of atherosclerotic progression. Circulation. 1993;87(suppl II):II-56-II-65.
Pignoli P, Tremoli E, Poli A, Oreste P, Paoletti R. Intimal plus medial thickness of the arterial wall: a direct measurement with ultrasound imaging. Circulation. 1986;74:1399-1406.
Poli A, Tremoli E, Colombo A, Sirtori M, Pignoli P, Paoletti R. Ultrasonographic measurement of the common carotid artery wall thickness in hypercholesterolemic patients: a new model for the quantitation and follow-up of preclinical atherosclerosis in living human subjects. Atherosclerosis. 1988;70:253-261.
Margitic SE, Bond MG, Crouse JR, Furberg CD, Probstfield JL. Progression and regression of carotid atherosclerosis in clinical trials. Arterioscler Thromb. 1991;11:443-451.
Family Heart Study Group. British Family Heart Study: its design and method, and prevalence of cardiovascular risk factors. Br J Gen Pract. 1994;44:62-67.
WHO Expert Committee. Prevention of Coronary Heart Disease. Geneva, Switzerland: World Health Organization; 1982.
WHO Expert Committee. Community Prevention and Control of Cardiovascular Diseases. Geneva, Switzerland: World Health Organization; 1986.
WHO Expert Committee. Appropriate Diagnostic Technology in the Management of Cardiovascular Diseases. Geneva, Switzerland: World Health Organization; 1988.
Wendelhag I, Wiklund O, Wikstrand J. Atherosclerotic changes in the femoral and carotid arteries in familial hypercholesterolemia: ultrasonographic assessment of intima-media thickness and plaque occurrence. Arterioscler Thromb. 1993;13:1404-1411.
Wendelhag I, Wiklund O, Wikstrand J. Arterial wall thickness in familial hypercholesterolemia: ultrasound measurement of intima-media thickness in the common carotid artery. Arterioscler Thromb. 1992;12:70-77.
Belcaro G, Fisher CM, Veller MG, Renton S, Geroulakos G, Rawashami G, Chong LC, Laurora G, Cesarone MR, De Sanctis MT, Incandela L, Barsotti A, Nicolaides AN. Screening asymptomatic subjects with subclinical arteriosclerotic lesions with arterial ultrasonic biopsy: the PAP study. Vasa. 1993;22:232-240.
Wikstrand J, Wiklund O. Frontiers in cardiovascular science: quantitative measurements of atherosclerotic manifestations in humans. Arterioscler Thromb. 1992:12:114-119.
Cesarone MR, De Sanctis MT, Laurora G, Ambrosoli L, Marelli C, Belcaro G. Effects of trandolapril on 24-h ambulatory blood pressure in patients with mild-to-moderate essential hypertension. J Cardiovasc Pharmacol. 1994;23(suppl 4):S65-S72.
Belcaro G, Veller M, Nicolaides AN, eds. Vascular Screening. Turin, Italy: Edizioni Minerva Medica; 1995.
Pearson TA. Coronary arteriography in the study of the epidemiology of coronary artery disease. Epidemiol Rev. 1984;6:140-147.
Blankenhorn DH, Hodis HN. Arterial imaging and atherosclerosis reversal. Arterioscler Thromb. 1994;14:177-192.
Hautecouverture M, Huet D, Bonnet V. Lipid balance in an arteritic patient and therapeutic implications. J Mal Vasc. 1992;147:326-328.
Kuller L, Borhani N, Furberg C, Gardin J, Manolio T, O'Leary D, Psaty B, Robbins J. Prevalence of subclinical atherosclerosis and cardiovascular disease and association with risk factors in the Cardiovascular Health Study. Am J Epidemiol. 1994;139:1164-1179.
Newman AB, Siscovick DS, Manolio TA, Polak J, Fried LP, Borhani NO, Wolfson SK. Ankle-arm index as a marker of atherosclerosis in the Cardiovascular Health Study: Cardiovascular Health Study (CHS) Collaborative Research Group. Circulation. 1993;88:837-845.