- Email: [email protected]
Assessment of arterial medial characteristics in human carotid arteries using integrated backscatter ultrasound and its histological implications
Atherosclerosis 180 (2005) 145–154
Assessment of arterial medial characteristics in human carotid arteries using integrated backscatter ultrasound and its histological implications Masanori Kawasaki a, ∗ , Yoko Ito a , Haruko Yokoyama a , Masazumi Arai a , Genzou Takemura a , Akira Hara b , Yoshiro Ichiki c , Hisato Takatsu a , Shinya Minatoguchi a , Hisayoshi Fujiwara a a
Regeneration and Advanced Medical Science, Graduate School of Medicine, Gifu University School of Medicine, Gifu, Japan b Department of Tumor Pathology, Gifu University School of Medicine, Gifu, Japan c Department of Dermatology, Gifu University School of Medicine, Gifu, Japan Received 6 July 2004; received in revised form 2 August 2004; accepted 12 November 2004
Abstract Recently, ultrasound tissue characterization of the carotid arteries with an integrated backscatter (IB) analysis was shown to identify a high-risk group of atherosclerosis. To clarify whether IB ultrasound is useful in assessing arterial sclerosis as well as stiffness β and whether IB values reflect the histological structure, we measured IB values of common carotid media in 52 subjects without coronary risk factors and in 10 patients with systemic sclerosis (SSc) in the clinical studies and 12 patients in the histological studies with a Philips Medical Systems Sonos 5500. IB values were correlated with age (r = 0.69, P < 0.0001), intima-media thickness (r = 0.72, P < 0.0001) and stiffness β (r = 0.80, P < 0.0001) in the control subjects. IB values and stiffness β in the SSc group were greater than in an age- and sex-matched control group (IB values: 9.6 ± 2.7 dB versus 16.1 ± 1.8 dB; stiffness β: 11.5 ± 4.5 versus 20.6 ± 5.6, P < 0.01). IB values of the media were correlated with the elastic fragmentation index (r = 0.63, P = 0.029) and the collagen fiber index (r = 0.59, P = 0.046). Measurements of IB values of carotid media are useful for non-invasively evaluating arterial sclerosis. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Arterial sclerosis; Integrated backscatter; Ultrasound; Systemic sclerosis
1. Introduction In general, atherosclerotic changes consist of two components: atherosis and sclerosis. According to a pathological study, these changes are recognized as thickening of intimamedia thickness (IMT) which is associated with structural atheromatous changes and decreased extensibility which is associated with functional sclerotic changes in elastic and collagen fibers. There are several ultrasound parameters to evaluate atherosclerosis, such as IMT, which is associated with atheromatous plaque formations and stiffness β, which are associated with decreased extensibility of the ∗ Corresponding author. Present address: Regeneration & Advanced Medical Science, Graduate School of Medicine, Gifu University, 1-1 Yanagido, Gifu 501-1194, Japan. Tel.: +81 58 230 6523; fax: +81 58 230 6524. E-mail address: [email protected] (M. Kawasaki).
0021-9150/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2004.11.018
arterial wall. IMT measurement is widely performed for the detection of atheromatous lesions and associated with age and coronary risk factors [1,2]. However, IMT is not always associated with the severity of arterial sclerosis in patients with hypertension [3]. This may be due to the degenerative changes in the medial smooth muscle cells and variably increased amount of elastin and collagen in hypertensive vessels [4]. The elasticity of major arteries is also affected by cardiovascular risk factors such as hypertension, hyperlipidemia, diabetes and aging. Stiffness β, which was defined by Hayashi et al. was found to be independent of blood pressure in the physiological range and associated with the severity of coronary atherosclerosis [5–7]. An increase in arterial stiffness has been reported as an early sign of atherosclerosis [8]. Therefore, it is very important to evaluate arterial sclerosis non-invasively.
146
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
We reported on the tissue characterization of arterial plaques in human carotid, femoral and coronary arteries in vivo using IB ultrasound [9,10]. These studies showed IB measurements accurately reflected the tissue characteristics of human carotid arterial plaques and IB values recorded in vivo were precisely correlated with the IB values obtained just after excision at autopsy or the IB values after fixation [9]. Moreover, it was reported that IB values of carotid arteries reflected the risk of atherosclerosis in patients with ischemic heart disease [11]. In addition, it was reported that systemic sclerosis (SSc), which is often accompanied by systemic arterial sclerosis, is associated with a higher rate of cadiovascular mortality [12]. Pathological studies showed that vascular fibrosis suggestive of hypertension was found in the systemic arterial tree of patients with SSc [13]. However, there have been no studies that evaluated the relationship among IB values, histological findings and arterial stiffness in the clinical setting. Thus, the purpose of the present study is to define whether IB measurement is useful in assessing arterial sclerosis as well as stiffness β. In addition, we wanted to evaluate the relationship between ante-mortem IB values and histological analysis of the arterial media that included the elastic fiber index, elastic fragmentation index, collagen fiber index and collagen fragmentation index.
2. Methods 2.1. Subjects We measured IB values of right common carotid arteries of 52 subjects without coronary risk factors and ischemic heart disease (aged 14–87 years, 25 males, and 27 females) and 10 SSc patients (aged 45–71 years, 10 females). In addition, 12 patients (aged 68–84 years, 10 males, and 2 females) who were in the terminal stage of various diseases had an IB evaluation ante-mortem and these values were compared to a histological analysis of elastic and collagen fiber in the carotid media. The causes of death in these 12 patients were pulmonary emphysema, pneumonia, old myocardial infarction, hypertrophic cardiomyopathy, cancer of the lung, tongue and larynx or pancreas. One patient had a history of myocardial infarction and another patient was diagnosed with a hypertrophic cardiomyopathy. Informed consent was obtained from patients or their relatives in all cases.
5–12 MHz transducer, setting the region-of-interest ROI (0.6 mm × 0.6 mm) focused on the arterial media at posterior wall in all studies. The software, built into this equipment, enabled the acquisition, storage and retrieval of a sequence of continuous IB images, forming a continuous loop digital recording of 2 s (60 frames in 2 s). IB values in the anterior wall were excluded from the analysis because of the influence of angle dependency, erratic diffraction and reverberation phenomena when a sectorarray transducer was used for the measurement. Quantitative ultrasonic backscatter analysis is angle dependent, and this may potentially be a limitation of quantitative ultrasonic diagnosis [14]. In addition, in the IB measurements of the anterior arterial walls, subcutaneous tissues existing between the transducer and the arterial wall may have caused erratic diffraction when the tissue was in the near-field of the transducer since the ultrasonic pressure field varies rapidly over short length scales [15]. Furthermore, the reverberation phenomena originating in the anterior arterial wall, skin and/or subcutaneous tissues are relatively large and variable in the carotid artery. Therefore, only the posterior wall (an angle span of 30◦ between −15◦ and +15◦ ) of the arteries was included in the analysis (Fig. 1C). Since the reverberation phenomena are thought to have an effect on the evaluation of IB values of the posterior arterial wall in the present studies, the reverberation phenomena should be excluded for precise comparison of the IB values. The reverberation phenomena influence both posterior wall and vascular lumen to the same degree. Therefore, we corrected the IB values of the posterior arterial wall by subtracting the IB values of the lumens with flowing blood just above the posterior wall, because the same method has been used in myocardial tissue characterization using IB ultrasound [16]. The IB values were determined by averaging the IB values from 10 sites continuously moving ROI on the posterior arterial site (Fig. 1A). ROI was set over the media without plaques in the ring-like IB images. We defined the presence of plaques as an area where IMT was more than 1.6 mm. At the same time, we measured IMT and stiffness β by conventional ultrasound in the same segment of the carotid arteries in which IB values was measured. These ultrasound determinations were based on the observance of three specialists who were ignorant of the histological study. 2.3. Integrated backscatter system presets and data acquisition
2.2. Study protocol For the clinical studies, transverse scans were performed between the middle of the right common carotid arteries and bifurcation of the external carotid artery and the internal carotid artery. Then, cross-sectional IB images were acquired at a site of 10 mm away from the bifurcation using an ultrasonic imaging system (Philips Medical Systems Sonos 5500). The arterial tissue was characterized with a
Off-line analysis of the two-dimensional IB images was performed by retrieving the previously stored data from the built-in optical disc drive in the system. In the present study, the time gain compensation was set at 0 dB and the lateral gain compensation at 50 dB at every measurement in the clinical study. The effective dynamic range and the resolution of this ultrasonic imaging system were 64 dB and 0.1 mm, respectively. At this setting, IB values of stainless steel at a distance
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
147
Fig. 1. Integrated backscatter (IB) images from the carotid artery (A) IB image from the subjects without coronary risk factors (Bar = 1 cm). An 11 pixels × 11 pixels (0.6 mm × 0.6 mm) rectangle shaped region-of-interest (ROI) was set in the media of the common carotid arteries. High echo density area revealed high IB values and vice versa in gray scale images using a Philips Medical Systems Sonos 5500 platform. (B) IB image from the patients with systemic sclerosis (SSc) (Bar = 1 cm). IB image from the patients with SSc showed relatively thin intima-media thickness but high IB values compared to those from the normal controls (arrow head). (C) Posterior arterial wall with an angle span of 30◦ between −15◦ and +15◦ were included from the IB evaluation because of the erratic diffraction from the anterior wall, the reverberation phenomena on anterior wall and the influence of angle dependency on the lateral wall. Also, IB values of the arterial lumen were measured just above posterior wall shown by *.
of 1–2 cm from the transducer were 50 dB, which is within the dynamic range of the system. 2.4. Measurement of stiffness parameter β Stiffness β was defined as: Ps Dd β = ln × Pd Ds − D d where ln is the natural logarithm, Ps the systolic pressure, Pd the diastolic pressure, Ds the lumen diameter at systole, and Dd the lumen diameter at diastole. The lumen diameter of the carotid artery was measured in the same segment in which IB was measured. Blood pressure was measured in the brachial artery immediately after the ultrasound examinations. 2.5. Histological study Common carotid arteries were excised at autopsy with careful attention to their rotational position and longitudinal position from the bifurcation of the external carotid artery and the internal carotid arteries. To mark the “rotational” position and the corresponding site of the IB measurements of the included segment, surgical needles were carefully inserted into the anterior wall at site 10 mm away from the bifurcation of the external carotid artery and the internal carotid arteries to be used as a reference point at autopsy. In these procedures, ring-like arterial specimens obtained at a similar level (within
1 mm) to the ultrasound study were fixed with 10% neutral buffered formalin and decalcified for 5 h. They were embedded with paraffin and cut into 4 m thick transverse sections perpendicular to the longitudinal axis of the artery. They were stained with hematoxylin-eosin, elastic van Gieson and Masson’s trichrome staining. Histological images were digitized on a square frame (300 pixels × 300 pixels) with an optical microscope at magnification 400× (Fig. 2A and B). The area that was stained brown by elastic van Gieson staining was selected (green) in the digitized image (Fig. 2C–F). Measurements were made at a resolution corresponding to 0.56 m/pixel. The elastic fiber index was determined as the ratio of the number of pixels that were stained brown by elastic van Gieson staining to the total number of pixels for each microscopic image (i). The collagen fiber index was defined as the ratio of the number of pixels that were stained blue by Masson’s trichrome staining to the total number of pixels for each microscopic image (ii). We classified any region of at least 2 m in width and at least 10 m in length as significantly large; regions smaller than that were classified as fragmented. The elastic fragmentation index was defined as the ratio of fragmented elastic fibers to total elastic fibers (iii). The collagen fragmentation index was defined as the ratio of fragmented collagen fibers to total collagen fibers (iv). Elastic fiber index (%) =
total elastic fiber total pixels of ROI
(i)
148
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
Fig. 2. (A) Region-of-interest (ROI) (300 pixels × 300 pixels) was set on the digitized image. (B) The ROI consisted of histological image. (C, D) Histological images stained by elastic van Gieson staining (Bar = 10 m). (E, F) Digitized images, in which elastic fibers were selected by the threshold of the digital image LUZEX F (Nireco, Kyoto, Japan). (E) Elastic fiber index: 59.3%, elastic fragmentation index: 57.7% and IB value: 9.1 dB. (F) Elastic fiber index: 18.9%, elastic fragmentation index: 79.3% and IB value: 14.4 dB.
total collagen fiber Collagen fiber index (%) = total pixels of ROI
(ii)
Elastic fragmentation index (%) significant large elastic fiber = 1− × 100 total elastic fiber Collagen fragmentation index (%) significant large collagen fiber = 1− × 100 total collagen fiber
(iii)
(iv)
Quantitative assessments were performed using a multipurpose color image processor, LUZEX F (Nireco, Kyoto, Japan) and fiber indices were determined by the software algorithms attached to the processor. These histological procedures were performed by two specialists who were ignorant of the ultrasound echo study. 2.6. Reproducibility of data We determined intraobserver variability of IB values of carotid arteries in 20 randomly selected recordings by the same observer and in another 20 randomly selected recordings by a second observer. In addition, we determined interobserver variability in 20 randomly selected recordings by the
independent observers and in another 20 randomly selected recordings by other independent observers. Intraobserver variability of IB values was 7.2 ± 4.8% and 6.5 ± 4.2%. Interobserver variability of IB values was 9.4 ± 5.8% and 7.4 ± 5.6%. Differences between measurements were plotted within two-standard deviation of the differences assessed by the method described by Bland and Altman [17]. The measurement sets, which showed the greatest variation (intraobserver variability: 7.2 ± 4.8% and interobserver variability: 9.4 ± 5.8%) are shown in Fig. 3A and B. Intraobserver and interobserver reproducibility of IB measurement were sufficient.
2.7. Statistical analyses Statistical analyses were performed using Stat View version 5.0 (SAS Institute Inc., Cray, NC, USA). Data were reported as mean ± standard deviation (S.D.). Testing for significant differences of each parameter between the age- and sex-matched control group and the SSc group was performed with a Mann–Whitney U-test. Correlations among IB values, age, IMT, stiffness β and diastolic diameter were tested for significance by Pearson’s correlation coefficient. Significance was tested for correlation coefficient values using a standard t-test. P < 0.05 was considered to be statistically significant.
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
149
Fig. 3. Intraobserver and interobserver reproducibility of integrated backscatter (IB) measurement and reliability of IB values compared to stiffness β assessed by the method of Bland and Altman [17]. Note that the differences between measurements were kept within two-standard deviation of each difference in each comparison.
3. Results IB values in the 52 subjects without coronary risk factors and ischemic heart disease were correlated with age (r = 0.69, P < 0.0001) (Fig. 4A), IMT (r = 0.72, P < 0.0001) (Fig. 4B) and stiffness β (r = 0.80, P < 0.0001) (Fig. 4C). Stiffness β was correlated with age (r = 0.77, P < 0.0001) (Fig. 4D). However, stiffness β was not correlated with age in the SSc group. Diastolic diameters of common carotid arteries were correlated with age in the control group (r = 0.57, P < 0.0001). However, they were not correlated with age in the SSc group. In addition, IMT were correlated with stiffness β in the control group (r = 0.72, P < 0.0001), though IMT were not correlated with stiffness β in the SSc group. IB values and stiffness β in the SSc group were greater than those in the age- and sex-matched control group (n = 10) (IB values: 9.6 ± 2.7 dB versus 16.1 ± 1.8 dB, P = 0.0005; stiffness β: 11.5 ± 4.5 versus 20.6 ± 5.6, P = 0.0012) (Fig. 5A and B). On the other hand, IMT in the SSc group was lower than in the age- and sex-matched control group (0.80 ± 0.19 mm versus 0.66 ± 0.88 mm, P = 0.0284) (Fig. 5C). Because the slope of the linear regression between IB values and stiffness β was 0.77 when the linear regression was plotted through the origin, we analyzed the reliability of IB measurements comparing IB values to 0.77 × stiffness β by the method of Bland and Altman [17] (Fig. 3C).
The correlation between stiffness β and IB values remained strong (r: from 0.80 to 0.80, P < 0.0001) when the data in the SSc group were added to the control group (Fig. 6B), although the correlation between stiffness β and IMT weakened (r: from 0.72 to 0.42, from P < 0.0001 to P = 0.0006) (Fig. 6A) when the data in the SSc group were added to the control group. In the histological study, IB values of the arterial media were correlated with elastic fragmentation index (r = 0.63, P = 0.029) (Fig. 7B) and collagen fiber index (r = 0.59, P = 0.046) (Fig. 7C). However, IB values of the arterial media were not correlated with the elastic fiber index or collagen fragmentation index (Fig. 7A and D). That is, IB values of the media increased in proportion to the degree of fragmentation of the elastic fiber. However, IB values were not related to the degree of fragmentation of the collagen fiber. With regards to the relative volume of elastic and collagen fibers, only the relative volume of collagen fibers was related to the IB values.
4. Discussion 4.1. IB values in healthy control subjects There was a significant correlation among the IB values, IMT and age in the subjects without coronary risk factors
150
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
Fig. 4. Integrated backscatter (IB) value as a function of each parameter in the ultrasound studies. (A) IB value as a function of age. (B) IB value as a function of intima-media thickness (IMT). (C) IB value as a function of stiffness β. (D) Stiffness β as a function of age. Closed circle: control subjects without coronary risk factors and ischemic heart disease; open circle: patients with systemic sclerosis (SSc). Correlation coefficients were calculated excluding the data of SSc patients in each comparison. Note that patients with SSc had relatively thin IMT but high IB values and stiffness parameter β compared with those of the subjects without coronary risk factors.
and ischemic heart disease. In addition, the IB values of the carotid media increased with age and the thickening of IMT (Fig. 4A and B). The fact that IB values were significantly correlated with stiffness β means that the IB values of carotid media indicate the stiffness of carotid arteries (Fig. 4C). Previous IB analysis demonstrated that the carotid IB values increased in arterial hypertension and had a positive relationship with age in the healthy population [3]. It suggests that IB analysis can be a useful non-invasive tool to detect arterial
sclerosis related to hypertension and to aging. The aging process modifies the structural and functional parameters of the arterial tree. These changes commonly include structural elongation of large arteries and vessel-wall thickening, which can be measured by conventional ultrasound study [18]. It was reported that stiffness β increased in proportion to age [19]. In the present study, stiffness β of the carotid artery, which is a functional parameter, increased with aging and the thickening of the arterial wall; and diastolic diameter
Fig. 5. Comparison of integrated backscatter (IB) values between patient with systemic sclerosis (SSc) and age, sex-matched subjects without coronary risk factors. (Error bars indicate ± one standard deviation, n = 10, respectively).
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
151
Fig. 6. (A) Stiffness β as a function of intima-media thickness (IMT) before and after adding data from patients with systemic sclerosis (SSc) to data from the subjects without coronary risk factors. (B) Stiffness β as a function of integrated backscatter (IB) values before and after adding data from SSc patients to data from subjects without coronary risk factors. Note that the correlation coefficient of IB values to stiffness β did not decrease after the two data sets were combined (from 0.80 to 0.80) despite the fact that the correlation coefficient of IMT to stiffness β decreased after the two data sets were combined (from 0.72 to 0.42). Closed circle: control subjects without coronary risk factors and ischemic heart disease; open circle: patients with SSc.
of common carotid arteries, which is a structural change, was significantly correlated with age (Fig. 4D). Concerning the relationship between structural and functional change, there was interesting research suggesting that age-associated lumen enlargement showed a possible advantage by maintaining normal distensibility at the site of muscular, medium-sized arteries such as radial arteries. However, lumen enlargement associated with the aging process did not compensate for the changes in the arterial elastic properties in elastic arteries such as carotid arteries [20]. That is, carotid arteries lose their distensibility and stiffen with age. Our study reinforces these previously reported results. 4.2. IB values in patients with SSc SSc is a multi-system disorder in which Raynaud’s phenomenon and microcirculatory abnormalities in small arteries are well recognized. However, not only microvascular abnormalities but also macrovascular abnormalities were reported in patients with SSc. It was reported that SSc is associated with a higher rate of cardiovascular mortality [12]. In the present study, IB values in patients with SSc had no correlation with age or IMT. That is, almost all carotid arteries of SSc patients revealed high IB values despite being younger and having a thin arterial wall.
The correlation coefficient of IB values to stiffness β did not decrease after adding the data from SSc patients to the data from subjects without coronary risk factors and ischemic heart disease (from 0.80 to 0.80) (Fig. 6B). In contrast, the correlation coefficient of IMT to stiffness β decreased after adding the data from SSc patients to the data in the subjects without coronary risk factors and ischemic heart disease (from 0.72 to 0.42) (Fig. 6A). This means that IB values are an optimal parameter to assess arterial sclerosis for diseased arteries. Concerning ultrasound studies involving patients with SSc, it was reported that macrovascular disease is more common in patients with SSc assessed by carotid scanning using conventional ultrasound [20]. In the present study, most of the patients with SSc had relatively thin IMT but high IB values and stiffness β compared with the control group (IB values: 9.6 ± 2.7 dB versus 16.1 ± 1.8 dB, P = 0.0005; stiffness β: 11.5 ± 4.5 versus 20.6 ± 5.6, P = 0.0012) (Fig. 5). Previous reports showed the absence of a relationship between carotid IB values and carotid IMT in patients with hypertension and chronic renal failure [3,21]. Interestingly, no relationship between IB values and IMT was found in the SSc patients in the present study. In patients with hypertension, chronic renal failure and SSc, the increased carotid IB value is probably related to the histological structural changes associated with these diseases. Therefore,
152
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
Fig. 7. Relationship between integrated backscatter (IB) values and each parameter in the histological studies (A) Relationship between IB values and elastic fiber index. (B) Relationship between IB values and elastic fragmentation index. (C) Relationship between IB values and collagen fiber index. (D) Relationship between IB values and collagen fragmentation index.
it is very important to elucidate what histological factor IB values mean. 4.3. Histological studies and IB values Generally, elastic fiber in arterial media plays an important role in maintaining elasticity of arteries. On the other hand, collagen fiber functions to maintain the strength of arteries. It is recognized that the structure of the aortic media is closely related to the biomechanical tangential or circumferential strain resulting from intraluminal hydrostatic pressure [1]. In addition, the histological structure of the aortic media influences the degree of arterial sclerosis. Thus, it is impossible to assess sclerosis by the measurement of IMT in the diseased arteries such as in hypertension and SSc. Therefore, IB measurements of arterial media are useful to assess arterial sclerosis. In the preliminary in vitro studies, IB values reflected the structural and biochemical composition of atherosclerotic lesion and could differentiate fibrofatty, fatty and calcification of arterial walls [22,23]. Urbani et al. indicated that the IB index in fatty sites differed from that of fibrous and calcified sites and intraluminal thrombotic sites differed from fibrous and calcified subsets in the intima [15]. Wada et al. reported that increases in the histological severity grade, which was
classified into three to four grades according to medial thickness, intimal thickness and destruction of the intimal elastic membrane were correlated with increased stiffness β [24]. We also reported that the tissue characterization of the different tissue types in the intima was possible in human carotid, femoral and coronary arteries in vivo using IB [10,11]. However, those reports did not describe the tissue characteristics of arterial media. There are few reports that analyzed the relationship between IB values of arterial media and its histological structure. It was reported that anisotropy of the direction and backscatter power is related to tissue type of arterial media in an ex vivo study [25]. In that study, the maximum powers of the IB signal from dense collagen was higher than from loose collagen. In the present study, IB values of the media were correlated with the collagen fiber index and these results support the previous findings [25]. On the other hand, fragmented elastic fibers, which are spatially oriented at random and of various sizes, reflect high backscattered ultrasound compared to elastic fibers which have a regular orientation [23]. However, there have been no reports of the relationship among IB values, histological findings and arterial stiffness in the clinical setting. The present study is the first in vivo demonstration of the relationship among IB values of carotid media, its histological characteristics and arterial sclerosis (Fig. 7).
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
4.4. Clinical implications Recently, a few techniques for the assessment of arterial sclerosis have been developed using ultrasound. Elastograph is useful to evaluate arterial medial sclerosis in vivo as well as the measurement of IB value. This technique has significant potential for quantitatively mapping the time-dependent mechanical behavior of elastic media [26,27]. The feasibility of this non-invasive technique to assess vascular non-linear elastic properties was demonstrated in an ex vivo experiment and further supported by in vivo measurement [28]. In addition, pulse wave velocity (PWV) represents a useful integrated index of vascular status and hence cardiovascular risk. A previous study reported higher PWV in patients with coronary risk factors [29]. Progressive atheromatous disease, including declining aortic elastin quality and content, leads to increasing aortic stiffness, detectable as increasing PWV, which defines arterial aging [30]. Tomiyama et al. reported that increased high-sensitive C-reactive protein, which has been thought to be an important risk factor for ischemic heart disease, is related to increased brachial-ankle PWV [31]. PWV is an index, which indicates the average degree of atherosclerosis of the entire arterial tree. In addition, PWV is also influenced by atherosis in the intima, while the IB value of the media is influenced only by medial sclerosis. Thus, the IB value is clearly differentiated from PWV. The measurement of IB value of the carotid media is useful to evaluate arterial medial sclerosis in vivo as well as stiffness β. 4.5. Study limitations There were a few limitations in the present study. First, measurements of IB values of arterial media in the present studies did not reflect the entire arterial structure of other advanced atherosclerotic sites because IB values were obtained from a single discrete site in the carotid arteries. Second, precise measurements of the blood pressure in the common carotid arteries were difficult without an invasive method. Therefore, we substituted the pressure in brachial artery for the pressure in the common carotid artery because the same method has been used for calculating stiffness β [19,24]. This may hinder precise measurements of stiffness β of carotid arteries. Third, histological analysis of the patients with SSc could not be performed because there were no autopsies in the patients with SSc. Thus, we could not determine the relationship between IB values and the histological characteristics of the arterial media in the SSc patients. Finally, stiffness β has been shown to be associated with severity of coronary atherosclerosis [5,6]. Therefore, IB values may also be associated with severity of coronary atherosclerosis. However, because the number of patients in the study was small, a multi-variant analysis for comparing patients with and without SSc was not possible. In addition, because the present study focused on individuals without coronary risk factors and patients with SSc, analysis of the incidence of stroke
153
and coronary events was not possible. A large-scale longterm study, that evaluates the relationship between IB values, coronary risk factors and the incidence of stroke and coronary events, will be required in the future.
References [1] Hadjiisky P, Peyri N, Grosgogeat Y, et al. Tunica media changes in the spontaneously hypertensive rat (SHR). Atherosclerosis 1987;65:125–37. [2] Mannami T, Konishi M, Baba S, et al. Prevalence of asymptomatic carotid atherosclerotic lesions detected by high-resolution ultrasonography and its relation to cerdiovascular risk factors in the general population of a Japanese city. Stroke 1997;28:518–25. [3] Rossi M, Cupisti A, Perrone L, et al. Carotid ultrasound backscatter analysis in hypertensive and in healthy subjects. Ultrasound Med Biol 2002;28(9):1123–8. [4] Wolinsky H. Response of the rat aortic media to hypertension. Circ Res 1970;26:507–22. [5] Ebrahim S, Papacosta O, Whincup P, et al. Carotid plaque, intima media thickness, cardiovascular risk factors, and prevalent cardiovascular disease in men and women. The British regional heart study. Stroke 1999;30:841–50. [6] Hirai T, Sasayama S, Kawasaki T, et al. Stiffness of systemic arteries in patients with myocadial infarction. A noninvasive method to predict severity of coronary atherosclerosis. Circulation 1989;80:78–86. [7] Hayashi K, Handa H, Nagasaka S, et al. Stiffness and elastic behavior of human intracranial and exracranial arteries. J Biomech 1980;13:175–84. [8] Ferrier KE, Muhlmann MH, Baguet JP, et al. Intensive cholesterol reduction lowers blood pressure and large artery stiffness in isolated systolic hypertension. J Am Coll Cardiol 2002;39:1020–5. [9] Kawasaki M, Takatsu H, Noda T, et al. Non-invasive tissue characterization of human atherosclerotic lesions in carotid and femoral arteries by ultrasound integrated backscatter. Comparison between histology and integrated backscatter images before and after death. J Am Coll Cardiol 2001;38:486–92. [10] Kawasaki M, Takatsu H, Noda T, et al. In vivo quantitative tissue characterization of human coronary arterial plaque by use of integrated backscatter intravascular ultrasound and comparison with angioscopic findings. Circulation 2002;105:2487–92. [11] Takiuchi S, Akugi H, Honda K, et al. Quantitative ultrasonic tissue characterization can identify high-risk atherosclerotic alteration in human carotid arteries. Circulation 2000;102:766–70. [12] Veale DJ, Collidge TA, Belch JJF, et al. Increased prevalence of symptomatic macrovascular disease in systemic sclerosis. Ann Rheum Dis 1995;54:853–5. [13] D’Angelo WA, Fries JF, Masi AT, Shulman LE. Pathologic observations in systemic sclerosis (scleroderma). A study of fiftyeight autopsy cases and fifty-eight matched controls. Am J Med 1969;46:428–40. [14] Picano E, Landdini L, Distante A, Salvadori M, L’Abbate A. Angle dependence of ultrasonic backscatter in arterial tissues: a study in vitro. Circulation 1985;72:572–6. [15] Urbani MP, Picano E, Parenti G, et al. In vivo radiofrequency-based ultrasonic tissue characterization of the atherosclerotic plaque. Stroke 1993;24:1507–12. [16] Naito J, Masuyama T, Mano T, et al. Ultrasound myocardial tissue characterization in the patients with dilated cardiomyopathy: value in noninvasive assessment of myocardial fibrosis. Am Heart J 1996;131:115–21. [17] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10.
154
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
[18] Kelly R, Hayward C, Arolio A, O’Rourke M. Non-invasive determination of age-related changes in the human arterial pulse. Circulation 1989;80:1652–9. [19] Kawasaki T, Sasayama S, Yagi S, et al. Noninvasive assessment of the age related changes in stiffness of major branches of the human arteries. Cardiovasc Res 1987;21:678–87. [20] Ho M, Veale D, Eastmond C, et al. Macrovascular disease and systemic sclerosis. Ann Rheum Dis 2000;59:39–43. [21] Rossi M, Cupisti A, Santoro G, et al. Ultrasonic tissue characterization of the carotid artery in chronic renal failure patients. Nephron 2002;91(2):270–5. [22] Barziliai B, Shffitz JE, Miller JG, Sobel BE. Quantitative ultrasonic characterization of the nature of atherosclerotic plaques in human aorta. Circ Res 1987;60:459–63. [23] Picano E, Landini L, Lattanzi F, Salvadori M, Benassi A, L’Abbate A. Time domain echo pattern evaluation from normal and atherosclerotic arterial walls: a study in vitro. Circulation 1988;77: 654–9. [24] Wada T, Kodaira K, Fujishiro K, et al. Correlation of ultrasoundmeasured common carotid artery stiffness with pathological findings. Arterioscler Thromb Vasc Biol 1994;14:479–82.
[25] De Kroon MGM, van der Wal LF, Gussenhoven WJ, Rijsterborgh H, Bom N. Backscatter directivity and integrated backscatter power of arterial tissue. Int J Card Imaging 1991;6:265–75. [26] de Korte CL, van der Steen AF, Cespedes EI, Pasterkamp G. Intravascular ultrasound elastography in human arteries: initial experience in vitro. Ultrasound Med Biol 1998;24(3):401–8. [27] Righetti R, Ophir J, Srinivasan S, Krouskop TA. The feasibity of using elastography for imaging the Poisson’s ratio in porous media. Ultrasound Med Biol 2004;30:215–28. [28] Kim K, Weitzel WF, Rubin JM, Xie H, Chen X, O’Donnell M. Vascular intramural strain imaging using arterial pressure equalization. Ultrasound Med Biol 2004;30:717–61. [29] Cruickshank K, Riste L, Anderson SG, Wright JS, Dumn G, Gosling RG. Aortic pulse-wave velocity and its relationship to mortality in diabetes and glucose intolerance. Circulation 2002;106:2085–90. [30] van Popele N, Grobbee D, Bots ML, et al. Association between arterial stiffness and atherosclerosis: the Rotterdam study. Stroke 2001;32:454–60. [31] Tomiyama H, Arai T, Koji Y, et al. The relationship between highsensitive C-reactive protein and pulse wave velocity in healthy Japanese men. Atherosclerosis 2004;174:373–7.
Assessment of arterial medial characteristics in human carotid arteries using integrated backscatter ultrasound and its histological implications Masanori Kawasaki a, ∗ , Yoko Ito a , Haruko Yokoyama a , Masazumi Arai a , Genzou Takemura a , Akira Hara b , Yoshiro Ichiki c , Hisato Takatsu a , Shinya Minatoguchi a , Hisayoshi Fujiwara a a
Regeneration and Advanced Medical Science, Graduate School of Medicine, Gifu University School of Medicine, Gifu, Japan b Department of Tumor Pathology, Gifu University School of Medicine, Gifu, Japan c Department of Dermatology, Gifu University School of Medicine, Gifu, Japan Received 6 July 2004; received in revised form 2 August 2004; accepted 12 November 2004
Abstract Recently, ultrasound tissue characterization of the carotid arteries with an integrated backscatter (IB) analysis was shown to identify a high-risk group of atherosclerosis. To clarify whether IB ultrasound is useful in assessing arterial sclerosis as well as stiffness β and whether IB values reflect the histological structure, we measured IB values of common carotid media in 52 subjects without coronary risk factors and in 10 patients with systemic sclerosis (SSc) in the clinical studies and 12 patients in the histological studies with a Philips Medical Systems Sonos 5500. IB values were correlated with age (r = 0.69, P < 0.0001), intima-media thickness (r = 0.72, P < 0.0001) and stiffness β (r = 0.80, P < 0.0001) in the control subjects. IB values and stiffness β in the SSc group were greater than in an age- and sex-matched control group (IB values: 9.6 ± 2.7 dB versus 16.1 ± 1.8 dB; stiffness β: 11.5 ± 4.5 versus 20.6 ± 5.6, P < 0.01). IB values of the media were correlated with the elastic fragmentation index (r = 0.63, P = 0.029) and the collagen fiber index (r = 0.59, P = 0.046). Measurements of IB values of carotid media are useful for non-invasively evaluating arterial sclerosis. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Arterial sclerosis; Integrated backscatter; Ultrasound; Systemic sclerosis
1. Introduction In general, atherosclerotic changes consist of two components: atherosis and sclerosis. According to a pathological study, these changes are recognized as thickening of intimamedia thickness (IMT) which is associated with structural atheromatous changes and decreased extensibility which is associated with functional sclerotic changes in elastic and collagen fibers. There are several ultrasound parameters to evaluate atherosclerosis, such as IMT, which is associated with atheromatous plaque formations and stiffness β, which are associated with decreased extensibility of the ∗ Corresponding author. Present address: Regeneration & Advanced Medical Science, Graduate School of Medicine, Gifu University, 1-1 Yanagido, Gifu 501-1194, Japan. Tel.: +81 58 230 6523; fax: +81 58 230 6524. E-mail address: [email protected] (M. Kawasaki).
0021-9150/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2004.11.018
arterial wall. IMT measurement is widely performed for the detection of atheromatous lesions and associated with age and coronary risk factors [1,2]. However, IMT is not always associated with the severity of arterial sclerosis in patients with hypertension [3]. This may be due to the degenerative changes in the medial smooth muscle cells and variably increased amount of elastin and collagen in hypertensive vessels [4]. The elasticity of major arteries is also affected by cardiovascular risk factors such as hypertension, hyperlipidemia, diabetes and aging. Stiffness β, which was defined by Hayashi et al. was found to be independent of blood pressure in the physiological range and associated with the severity of coronary atherosclerosis [5–7]. An increase in arterial stiffness has been reported as an early sign of atherosclerosis [8]. Therefore, it is very important to evaluate arterial sclerosis non-invasively.
146
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
We reported on the tissue characterization of arterial plaques in human carotid, femoral and coronary arteries in vivo using IB ultrasound [9,10]. These studies showed IB measurements accurately reflected the tissue characteristics of human carotid arterial plaques and IB values recorded in vivo were precisely correlated with the IB values obtained just after excision at autopsy or the IB values after fixation [9]. Moreover, it was reported that IB values of carotid arteries reflected the risk of atherosclerosis in patients with ischemic heart disease [11]. In addition, it was reported that systemic sclerosis (SSc), which is often accompanied by systemic arterial sclerosis, is associated with a higher rate of cadiovascular mortality [12]. Pathological studies showed that vascular fibrosis suggestive of hypertension was found in the systemic arterial tree of patients with SSc [13]. However, there have been no studies that evaluated the relationship among IB values, histological findings and arterial stiffness in the clinical setting. Thus, the purpose of the present study is to define whether IB measurement is useful in assessing arterial sclerosis as well as stiffness β. In addition, we wanted to evaluate the relationship between ante-mortem IB values and histological analysis of the arterial media that included the elastic fiber index, elastic fragmentation index, collagen fiber index and collagen fragmentation index.
2. Methods 2.1. Subjects We measured IB values of right common carotid arteries of 52 subjects without coronary risk factors and ischemic heart disease (aged 14–87 years, 25 males, and 27 females) and 10 SSc patients (aged 45–71 years, 10 females). In addition, 12 patients (aged 68–84 years, 10 males, and 2 females) who were in the terminal stage of various diseases had an IB evaluation ante-mortem and these values were compared to a histological analysis of elastic and collagen fiber in the carotid media. The causes of death in these 12 patients were pulmonary emphysema, pneumonia, old myocardial infarction, hypertrophic cardiomyopathy, cancer of the lung, tongue and larynx or pancreas. One patient had a history of myocardial infarction and another patient was diagnosed with a hypertrophic cardiomyopathy. Informed consent was obtained from patients or their relatives in all cases.
5–12 MHz transducer, setting the region-of-interest ROI (0.6 mm × 0.6 mm) focused on the arterial media at posterior wall in all studies. The software, built into this equipment, enabled the acquisition, storage and retrieval of a sequence of continuous IB images, forming a continuous loop digital recording of 2 s (60 frames in 2 s). IB values in the anterior wall were excluded from the analysis because of the influence of angle dependency, erratic diffraction and reverberation phenomena when a sectorarray transducer was used for the measurement. Quantitative ultrasonic backscatter analysis is angle dependent, and this may potentially be a limitation of quantitative ultrasonic diagnosis [14]. In addition, in the IB measurements of the anterior arterial walls, subcutaneous tissues existing between the transducer and the arterial wall may have caused erratic diffraction when the tissue was in the near-field of the transducer since the ultrasonic pressure field varies rapidly over short length scales [15]. Furthermore, the reverberation phenomena originating in the anterior arterial wall, skin and/or subcutaneous tissues are relatively large and variable in the carotid artery. Therefore, only the posterior wall (an angle span of 30◦ between −15◦ and +15◦ ) of the arteries was included in the analysis (Fig. 1C). Since the reverberation phenomena are thought to have an effect on the evaluation of IB values of the posterior arterial wall in the present studies, the reverberation phenomena should be excluded for precise comparison of the IB values. The reverberation phenomena influence both posterior wall and vascular lumen to the same degree. Therefore, we corrected the IB values of the posterior arterial wall by subtracting the IB values of the lumens with flowing blood just above the posterior wall, because the same method has been used in myocardial tissue characterization using IB ultrasound [16]. The IB values were determined by averaging the IB values from 10 sites continuously moving ROI on the posterior arterial site (Fig. 1A). ROI was set over the media without plaques in the ring-like IB images. We defined the presence of plaques as an area where IMT was more than 1.6 mm. At the same time, we measured IMT and stiffness β by conventional ultrasound in the same segment of the carotid arteries in which IB values was measured. These ultrasound determinations were based on the observance of three specialists who were ignorant of the histological study. 2.3. Integrated backscatter system presets and data acquisition
2.2. Study protocol For the clinical studies, transverse scans were performed between the middle of the right common carotid arteries and bifurcation of the external carotid artery and the internal carotid artery. Then, cross-sectional IB images were acquired at a site of 10 mm away from the bifurcation using an ultrasonic imaging system (Philips Medical Systems Sonos 5500). The arterial tissue was characterized with a
Off-line analysis of the two-dimensional IB images was performed by retrieving the previously stored data from the built-in optical disc drive in the system. In the present study, the time gain compensation was set at 0 dB and the lateral gain compensation at 50 dB at every measurement in the clinical study. The effective dynamic range and the resolution of this ultrasonic imaging system were 64 dB and 0.1 mm, respectively. At this setting, IB values of stainless steel at a distance
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
147
Fig. 1. Integrated backscatter (IB) images from the carotid artery (A) IB image from the subjects without coronary risk factors (Bar = 1 cm). An 11 pixels × 11 pixels (0.6 mm × 0.6 mm) rectangle shaped region-of-interest (ROI) was set in the media of the common carotid arteries. High echo density area revealed high IB values and vice versa in gray scale images using a Philips Medical Systems Sonos 5500 platform. (B) IB image from the patients with systemic sclerosis (SSc) (Bar = 1 cm). IB image from the patients with SSc showed relatively thin intima-media thickness but high IB values compared to those from the normal controls (arrow head). (C) Posterior arterial wall with an angle span of 30◦ between −15◦ and +15◦ were included from the IB evaluation because of the erratic diffraction from the anterior wall, the reverberation phenomena on anterior wall and the influence of angle dependency on the lateral wall. Also, IB values of the arterial lumen were measured just above posterior wall shown by *.
of 1–2 cm from the transducer were 50 dB, which is within the dynamic range of the system. 2.4. Measurement of stiffness parameter β Stiffness β was defined as: Ps Dd β = ln × Pd Ds − D d where ln is the natural logarithm, Ps the systolic pressure, Pd the diastolic pressure, Ds the lumen diameter at systole, and Dd the lumen diameter at diastole. The lumen diameter of the carotid artery was measured in the same segment in which IB was measured. Blood pressure was measured in the brachial artery immediately after the ultrasound examinations. 2.5. Histological study Common carotid arteries were excised at autopsy with careful attention to their rotational position and longitudinal position from the bifurcation of the external carotid artery and the internal carotid arteries. To mark the “rotational” position and the corresponding site of the IB measurements of the included segment, surgical needles were carefully inserted into the anterior wall at site 10 mm away from the bifurcation of the external carotid artery and the internal carotid arteries to be used as a reference point at autopsy. In these procedures, ring-like arterial specimens obtained at a similar level (within
1 mm) to the ultrasound study were fixed with 10% neutral buffered formalin and decalcified for 5 h. They were embedded with paraffin and cut into 4 m thick transverse sections perpendicular to the longitudinal axis of the artery. They were stained with hematoxylin-eosin, elastic van Gieson and Masson’s trichrome staining. Histological images were digitized on a square frame (300 pixels × 300 pixels) with an optical microscope at magnification 400× (Fig. 2A and B). The area that was stained brown by elastic van Gieson staining was selected (green) in the digitized image (Fig. 2C–F). Measurements were made at a resolution corresponding to 0.56 m/pixel. The elastic fiber index was determined as the ratio of the number of pixels that were stained brown by elastic van Gieson staining to the total number of pixels for each microscopic image (i). The collagen fiber index was defined as the ratio of the number of pixels that were stained blue by Masson’s trichrome staining to the total number of pixels for each microscopic image (ii). We classified any region of at least 2 m in width and at least 10 m in length as significantly large; regions smaller than that were classified as fragmented. The elastic fragmentation index was defined as the ratio of fragmented elastic fibers to total elastic fibers (iii). The collagen fragmentation index was defined as the ratio of fragmented collagen fibers to total collagen fibers (iv). Elastic fiber index (%) =
total elastic fiber total pixels of ROI
(i)
148
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
Fig. 2. (A) Region-of-interest (ROI) (300 pixels × 300 pixels) was set on the digitized image. (B) The ROI consisted of histological image. (C, D) Histological images stained by elastic van Gieson staining (Bar = 10 m). (E, F) Digitized images, in which elastic fibers were selected by the threshold of the digital image LUZEX F (Nireco, Kyoto, Japan). (E) Elastic fiber index: 59.3%, elastic fragmentation index: 57.7% and IB value: 9.1 dB. (F) Elastic fiber index: 18.9%, elastic fragmentation index: 79.3% and IB value: 14.4 dB.
total collagen fiber Collagen fiber index (%) = total pixels of ROI
(ii)
Elastic fragmentation index (%) significant large elastic fiber = 1− × 100 total elastic fiber Collagen fragmentation index (%) significant large collagen fiber = 1− × 100 total collagen fiber
(iii)
(iv)
Quantitative assessments were performed using a multipurpose color image processor, LUZEX F (Nireco, Kyoto, Japan) and fiber indices were determined by the software algorithms attached to the processor. These histological procedures were performed by two specialists who were ignorant of the ultrasound echo study. 2.6. Reproducibility of data We determined intraobserver variability of IB values of carotid arteries in 20 randomly selected recordings by the same observer and in another 20 randomly selected recordings by a second observer. In addition, we determined interobserver variability in 20 randomly selected recordings by the
independent observers and in another 20 randomly selected recordings by other independent observers. Intraobserver variability of IB values was 7.2 ± 4.8% and 6.5 ± 4.2%. Interobserver variability of IB values was 9.4 ± 5.8% and 7.4 ± 5.6%. Differences between measurements were plotted within two-standard deviation of the differences assessed by the method described by Bland and Altman [17]. The measurement sets, which showed the greatest variation (intraobserver variability: 7.2 ± 4.8% and interobserver variability: 9.4 ± 5.8%) are shown in Fig. 3A and B. Intraobserver and interobserver reproducibility of IB measurement were sufficient.
2.7. Statistical analyses Statistical analyses were performed using Stat View version 5.0 (SAS Institute Inc., Cray, NC, USA). Data were reported as mean ± standard deviation (S.D.). Testing for significant differences of each parameter between the age- and sex-matched control group and the SSc group was performed with a Mann–Whitney U-test. Correlations among IB values, age, IMT, stiffness β and diastolic diameter were tested for significance by Pearson’s correlation coefficient. Significance was tested for correlation coefficient values using a standard t-test. P < 0.05 was considered to be statistically significant.
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
149
Fig. 3. Intraobserver and interobserver reproducibility of integrated backscatter (IB) measurement and reliability of IB values compared to stiffness β assessed by the method of Bland and Altman [17]. Note that the differences between measurements were kept within two-standard deviation of each difference in each comparison.
3. Results IB values in the 52 subjects without coronary risk factors and ischemic heart disease were correlated with age (r = 0.69, P < 0.0001) (Fig. 4A), IMT (r = 0.72, P < 0.0001) (Fig. 4B) and stiffness β (r = 0.80, P < 0.0001) (Fig. 4C). Stiffness β was correlated with age (r = 0.77, P < 0.0001) (Fig. 4D). However, stiffness β was not correlated with age in the SSc group. Diastolic diameters of common carotid arteries were correlated with age in the control group (r = 0.57, P < 0.0001). However, they were not correlated with age in the SSc group. In addition, IMT were correlated with stiffness β in the control group (r = 0.72, P < 0.0001), though IMT were not correlated with stiffness β in the SSc group. IB values and stiffness β in the SSc group were greater than those in the age- and sex-matched control group (n = 10) (IB values: 9.6 ± 2.7 dB versus 16.1 ± 1.8 dB, P = 0.0005; stiffness β: 11.5 ± 4.5 versus 20.6 ± 5.6, P = 0.0012) (Fig. 5A and B). On the other hand, IMT in the SSc group was lower than in the age- and sex-matched control group (0.80 ± 0.19 mm versus 0.66 ± 0.88 mm, P = 0.0284) (Fig. 5C). Because the slope of the linear regression between IB values and stiffness β was 0.77 when the linear regression was plotted through the origin, we analyzed the reliability of IB measurements comparing IB values to 0.77 × stiffness β by the method of Bland and Altman [17] (Fig. 3C).
The correlation between stiffness β and IB values remained strong (r: from 0.80 to 0.80, P < 0.0001) when the data in the SSc group were added to the control group (Fig. 6B), although the correlation between stiffness β and IMT weakened (r: from 0.72 to 0.42, from P < 0.0001 to P = 0.0006) (Fig. 6A) when the data in the SSc group were added to the control group. In the histological study, IB values of the arterial media were correlated with elastic fragmentation index (r = 0.63, P = 0.029) (Fig. 7B) and collagen fiber index (r = 0.59, P = 0.046) (Fig. 7C). However, IB values of the arterial media were not correlated with the elastic fiber index or collagen fragmentation index (Fig. 7A and D). That is, IB values of the media increased in proportion to the degree of fragmentation of the elastic fiber. However, IB values were not related to the degree of fragmentation of the collagen fiber. With regards to the relative volume of elastic and collagen fibers, only the relative volume of collagen fibers was related to the IB values.
4. Discussion 4.1. IB values in healthy control subjects There was a significant correlation among the IB values, IMT and age in the subjects without coronary risk factors
150
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
Fig. 4. Integrated backscatter (IB) value as a function of each parameter in the ultrasound studies. (A) IB value as a function of age. (B) IB value as a function of intima-media thickness (IMT). (C) IB value as a function of stiffness β. (D) Stiffness β as a function of age. Closed circle: control subjects without coronary risk factors and ischemic heart disease; open circle: patients with systemic sclerosis (SSc). Correlation coefficients were calculated excluding the data of SSc patients in each comparison. Note that patients with SSc had relatively thin IMT but high IB values and stiffness parameter β compared with those of the subjects without coronary risk factors.
and ischemic heart disease. In addition, the IB values of the carotid media increased with age and the thickening of IMT (Fig. 4A and B). The fact that IB values were significantly correlated with stiffness β means that the IB values of carotid media indicate the stiffness of carotid arteries (Fig. 4C). Previous IB analysis demonstrated that the carotid IB values increased in arterial hypertension and had a positive relationship with age in the healthy population [3]. It suggests that IB analysis can be a useful non-invasive tool to detect arterial
sclerosis related to hypertension and to aging. The aging process modifies the structural and functional parameters of the arterial tree. These changes commonly include structural elongation of large arteries and vessel-wall thickening, which can be measured by conventional ultrasound study [18]. It was reported that stiffness β increased in proportion to age [19]. In the present study, stiffness β of the carotid artery, which is a functional parameter, increased with aging and the thickening of the arterial wall; and diastolic diameter
Fig. 5. Comparison of integrated backscatter (IB) values between patient with systemic sclerosis (SSc) and age, sex-matched subjects without coronary risk factors. (Error bars indicate ± one standard deviation, n = 10, respectively).
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
151
Fig. 6. (A) Stiffness β as a function of intima-media thickness (IMT) before and after adding data from patients with systemic sclerosis (SSc) to data from the subjects without coronary risk factors. (B) Stiffness β as a function of integrated backscatter (IB) values before and after adding data from SSc patients to data from subjects without coronary risk factors. Note that the correlation coefficient of IB values to stiffness β did not decrease after the two data sets were combined (from 0.80 to 0.80) despite the fact that the correlation coefficient of IMT to stiffness β decreased after the two data sets were combined (from 0.72 to 0.42). Closed circle: control subjects without coronary risk factors and ischemic heart disease; open circle: patients with SSc.
of common carotid arteries, which is a structural change, was significantly correlated with age (Fig. 4D). Concerning the relationship between structural and functional change, there was interesting research suggesting that age-associated lumen enlargement showed a possible advantage by maintaining normal distensibility at the site of muscular, medium-sized arteries such as radial arteries. However, lumen enlargement associated with the aging process did not compensate for the changes in the arterial elastic properties in elastic arteries such as carotid arteries [20]. That is, carotid arteries lose their distensibility and stiffen with age. Our study reinforces these previously reported results. 4.2. IB values in patients with SSc SSc is a multi-system disorder in which Raynaud’s phenomenon and microcirculatory abnormalities in small arteries are well recognized. However, not only microvascular abnormalities but also macrovascular abnormalities were reported in patients with SSc. It was reported that SSc is associated with a higher rate of cardiovascular mortality [12]. In the present study, IB values in patients with SSc had no correlation with age or IMT. That is, almost all carotid arteries of SSc patients revealed high IB values despite being younger and having a thin arterial wall.
The correlation coefficient of IB values to stiffness β did not decrease after adding the data from SSc patients to the data from subjects without coronary risk factors and ischemic heart disease (from 0.80 to 0.80) (Fig. 6B). In contrast, the correlation coefficient of IMT to stiffness β decreased after adding the data from SSc patients to the data in the subjects without coronary risk factors and ischemic heart disease (from 0.72 to 0.42) (Fig. 6A). This means that IB values are an optimal parameter to assess arterial sclerosis for diseased arteries. Concerning ultrasound studies involving patients with SSc, it was reported that macrovascular disease is more common in patients with SSc assessed by carotid scanning using conventional ultrasound [20]. In the present study, most of the patients with SSc had relatively thin IMT but high IB values and stiffness β compared with the control group (IB values: 9.6 ± 2.7 dB versus 16.1 ± 1.8 dB, P = 0.0005; stiffness β: 11.5 ± 4.5 versus 20.6 ± 5.6, P = 0.0012) (Fig. 5). Previous reports showed the absence of a relationship between carotid IB values and carotid IMT in patients with hypertension and chronic renal failure [3,21]. Interestingly, no relationship between IB values and IMT was found in the SSc patients in the present study. In patients with hypertension, chronic renal failure and SSc, the increased carotid IB value is probably related to the histological structural changes associated with these diseases. Therefore,
152
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
Fig. 7. Relationship between integrated backscatter (IB) values and each parameter in the histological studies (A) Relationship between IB values and elastic fiber index. (B) Relationship between IB values and elastic fragmentation index. (C) Relationship between IB values and collagen fiber index. (D) Relationship between IB values and collagen fragmentation index.
it is very important to elucidate what histological factor IB values mean. 4.3. Histological studies and IB values Generally, elastic fiber in arterial media plays an important role in maintaining elasticity of arteries. On the other hand, collagen fiber functions to maintain the strength of arteries. It is recognized that the structure of the aortic media is closely related to the biomechanical tangential or circumferential strain resulting from intraluminal hydrostatic pressure [1]. In addition, the histological structure of the aortic media influences the degree of arterial sclerosis. Thus, it is impossible to assess sclerosis by the measurement of IMT in the diseased arteries such as in hypertension and SSc. Therefore, IB measurements of arterial media are useful to assess arterial sclerosis. In the preliminary in vitro studies, IB values reflected the structural and biochemical composition of atherosclerotic lesion and could differentiate fibrofatty, fatty and calcification of arterial walls [22,23]. Urbani et al. indicated that the IB index in fatty sites differed from that of fibrous and calcified sites and intraluminal thrombotic sites differed from fibrous and calcified subsets in the intima [15]. Wada et al. reported that increases in the histological severity grade, which was
classified into three to four grades according to medial thickness, intimal thickness and destruction of the intimal elastic membrane were correlated with increased stiffness β [24]. We also reported that the tissue characterization of the different tissue types in the intima was possible in human carotid, femoral and coronary arteries in vivo using IB [10,11]. However, those reports did not describe the tissue characteristics of arterial media. There are few reports that analyzed the relationship between IB values of arterial media and its histological structure. It was reported that anisotropy of the direction and backscatter power is related to tissue type of arterial media in an ex vivo study [25]. In that study, the maximum powers of the IB signal from dense collagen was higher than from loose collagen. In the present study, IB values of the media were correlated with the collagen fiber index and these results support the previous findings [25]. On the other hand, fragmented elastic fibers, which are spatially oriented at random and of various sizes, reflect high backscattered ultrasound compared to elastic fibers which have a regular orientation [23]. However, there have been no reports of the relationship among IB values, histological findings and arterial stiffness in the clinical setting. The present study is the first in vivo demonstration of the relationship among IB values of carotid media, its histological characteristics and arterial sclerosis (Fig. 7).
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
4.4. Clinical implications Recently, a few techniques for the assessment of arterial sclerosis have been developed using ultrasound. Elastograph is useful to evaluate arterial medial sclerosis in vivo as well as the measurement of IB value. This technique has significant potential for quantitatively mapping the time-dependent mechanical behavior of elastic media [26,27]. The feasibility of this non-invasive technique to assess vascular non-linear elastic properties was demonstrated in an ex vivo experiment and further supported by in vivo measurement [28]. In addition, pulse wave velocity (PWV) represents a useful integrated index of vascular status and hence cardiovascular risk. A previous study reported higher PWV in patients with coronary risk factors [29]. Progressive atheromatous disease, including declining aortic elastin quality and content, leads to increasing aortic stiffness, detectable as increasing PWV, which defines arterial aging [30]. Tomiyama et al. reported that increased high-sensitive C-reactive protein, which has been thought to be an important risk factor for ischemic heart disease, is related to increased brachial-ankle PWV [31]. PWV is an index, which indicates the average degree of atherosclerosis of the entire arterial tree. In addition, PWV is also influenced by atherosis in the intima, while the IB value of the media is influenced only by medial sclerosis. Thus, the IB value is clearly differentiated from PWV. The measurement of IB value of the carotid media is useful to evaluate arterial medial sclerosis in vivo as well as stiffness β. 4.5. Study limitations There were a few limitations in the present study. First, measurements of IB values of arterial media in the present studies did not reflect the entire arterial structure of other advanced atherosclerotic sites because IB values were obtained from a single discrete site in the carotid arteries. Second, precise measurements of the blood pressure in the common carotid arteries were difficult without an invasive method. Therefore, we substituted the pressure in brachial artery for the pressure in the common carotid artery because the same method has been used for calculating stiffness β [19,24]. This may hinder precise measurements of stiffness β of carotid arteries. Third, histological analysis of the patients with SSc could not be performed because there were no autopsies in the patients with SSc. Thus, we could not determine the relationship between IB values and the histological characteristics of the arterial media in the SSc patients. Finally, stiffness β has been shown to be associated with severity of coronary atherosclerosis [5,6]. Therefore, IB values may also be associated with severity of coronary atherosclerosis. However, because the number of patients in the study was small, a multi-variant analysis for comparing patients with and without SSc was not possible. In addition, because the present study focused on individuals without coronary risk factors and patients with SSc, analysis of the incidence of stroke
153
and coronary events was not possible. A large-scale longterm study, that evaluates the relationship between IB values, coronary risk factors and the incidence of stroke and coronary events, will be required in the future.
References [1] Hadjiisky P, Peyri N, Grosgogeat Y, et al. Tunica media changes in the spontaneously hypertensive rat (SHR). Atherosclerosis 1987;65:125–37. [2] Mannami T, Konishi M, Baba S, et al. Prevalence of asymptomatic carotid atherosclerotic lesions detected by high-resolution ultrasonography and its relation to cerdiovascular risk factors in the general population of a Japanese city. Stroke 1997;28:518–25. [3] Rossi M, Cupisti A, Perrone L, et al. Carotid ultrasound backscatter analysis in hypertensive and in healthy subjects. Ultrasound Med Biol 2002;28(9):1123–8. [4] Wolinsky H. Response of the rat aortic media to hypertension. Circ Res 1970;26:507–22. [5] Ebrahim S, Papacosta O, Whincup P, et al. Carotid plaque, intima media thickness, cardiovascular risk factors, and prevalent cardiovascular disease in men and women. The British regional heart study. Stroke 1999;30:841–50. [6] Hirai T, Sasayama S, Kawasaki T, et al. Stiffness of systemic arteries in patients with myocadial infarction. A noninvasive method to predict severity of coronary atherosclerosis. Circulation 1989;80:78–86. [7] Hayashi K, Handa H, Nagasaka S, et al. Stiffness and elastic behavior of human intracranial and exracranial arteries. J Biomech 1980;13:175–84. [8] Ferrier KE, Muhlmann MH, Baguet JP, et al. Intensive cholesterol reduction lowers blood pressure and large artery stiffness in isolated systolic hypertension. J Am Coll Cardiol 2002;39:1020–5. [9] Kawasaki M, Takatsu H, Noda T, et al. Non-invasive tissue characterization of human atherosclerotic lesions in carotid and femoral arteries by ultrasound integrated backscatter. Comparison between histology and integrated backscatter images before and after death. J Am Coll Cardiol 2001;38:486–92. [10] Kawasaki M, Takatsu H, Noda T, et al. In vivo quantitative tissue characterization of human coronary arterial plaque by use of integrated backscatter intravascular ultrasound and comparison with angioscopic findings. Circulation 2002;105:2487–92. [11] Takiuchi S, Akugi H, Honda K, et al. Quantitative ultrasonic tissue characterization can identify high-risk atherosclerotic alteration in human carotid arteries. Circulation 2000;102:766–70. [12] Veale DJ, Collidge TA, Belch JJF, et al. Increased prevalence of symptomatic macrovascular disease in systemic sclerosis. Ann Rheum Dis 1995;54:853–5. [13] D’Angelo WA, Fries JF, Masi AT, Shulman LE. Pathologic observations in systemic sclerosis (scleroderma). A study of fiftyeight autopsy cases and fifty-eight matched controls. Am J Med 1969;46:428–40. [14] Picano E, Landdini L, Distante A, Salvadori M, L’Abbate A. Angle dependence of ultrasonic backscatter in arterial tissues: a study in vitro. Circulation 1985;72:572–6. [15] Urbani MP, Picano E, Parenti G, et al. In vivo radiofrequency-based ultrasonic tissue characterization of the atherosclerotic plaque. Stroke 1993;24:1507–12. [16] Naito J, Masuyama T, Mano T, et al. Ultrasound myocardial tissue characterization in the patients with dilated cardiomyopathy: value in noninvasive assessment of myocardial fibrosis. Am Heart J 1996;131:115–21. [17] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10.
154
M. Kawasaki et al. / Atherosclerosis 180 (2005) 145–154
[18] Kelly R, Hayward C, Arolio A, O’Rourke M. Non-invasive determination of age-related changes in the human arterial pulse. Circulation 1989;80:1652–9. [19] Kawasaki T, Sasayama S, Yagi S, et al. Noninvasive assessment of the age related changes in stiffness of major branches of the human arteries. Cardiovasc Res 1987;21:678–87. [20] Ho M, Veale D, Eastmond C, et al. Macrovascular disease and systemic sclerosis. Ann Rheum Dis 2000;59:39–43. [21] Rossi M, Cupisti A, Santoro G, et al. Ultrasonic tissue characterization of the carotid artery in chronic renal failure patients. Nephron 2002;91(2):270–5. [22] Barziliai B, Shffitz JE, Miller JG, Sobel BE. Quantitative ultrasonic characterization of the nature of atherosclerotic plaques in human aorta. Circ Res 1987;60:459–63. [23] Picano E, Landini L, Lattanzi F, Salvadori M, Benassi A, L’Abbate A. Time domain echo pattern evaluation from normal and atherosclerotic arterial walls: a study in vitro. Circulation 1988;77: 654–9. [24] Wada T, Kodaira K, Fujishiro K, et al. Correlation of ultrasoundmeasured common carotid artery stiffness with pathological findings. Arterioscler Thromb Vasc Biol 1994;14:479–82.
[25] De Kroon MGM, van der Wal LF, Gussenhoven WJ, Rijsterborgh H, Bom N. Backscatter directivity and integrated backscatter power of arterial tissue. Int J Card Imaging 1991;6:265–75. [26] de Korte CL, van der Steen AF, Cespedes EI, Pasterkamp G. Intravascular ultrasound elastography in human arteries: initial experience in vitro. Ultrasound Med Biol 1998;24(3):401–8. [27] Righetti R, Ophir J, Srinivasan S, Krouskop TA. The feasibity of using elastography for imaging the Poisson’s ratio in porous media. Ultrasound Med Biol 2004;30:215–28. [28] Kim K, Weitzel WF, Rubin JM, Xie H, Chen X, O’Donnell M. Vascular intramural strain imaging using arterial pressure equalization. Ultrasound Med Biol 2004;30:717–61. [29] Cruickshank K, Riste L, Anderson SG, Wright JS, Dumn G, Gosling RG. Aortic pulse-wave velocity and its relationship to mortality in diabetes and glucose intolerance. Circulation 2002;106:2085–90. [30] van Popele N, Grobbee D, Bots ML, et al. Association between arterial stiffness and atherosclerosis: the Rotterdam study. Stroke 2001;32:454–60. [31] Tomiyama H, Arai T, Koji Y, et al. The relationship between highsensitive C-reactive protein and pulse wave velocity in healthy Japanese men. Atherosclerosis 2004;174:373–7.