- Email: [email protected]
A description of two morphologic patterns of aortic fatty streaks, and a hypothesis of their pathogenesis
Atherosclerosis 141 (1998) 153 – 160
A description of two morphologic patterns of aortic fatty streaks, and a hypothesis of their pathogenesis Gregory D. Sloop a,*, Robert S. Perret b, Jody S. Brahney a, Margaret Oalmann a a b
Department of Pathology, Louisiana State Uni6ersity School of Medicine, 1901 Perdido, New Orleans, LA 70112, USA Department of Radiology, Louisiana State Uni6ersity School of Medicine, 1901 Perdido, New Orleans, LA 70112, USA Received 26 August 1997; received in revised form 27 May 1998; accepted 5 June 1998
Abstract Two morphologic patterns of fatty streak were identified on examination of 74 aortas from the Pathobiological Determinants of Atherosclerosis in Youth study. Pattern 1, which predominated in 78% of aortas, is characterized by broad bands of intense stain which extend to the proximal edge of ostia. Pattern 2, which predominated in 11%, is characterized by less intense staining which is concave to the associated ostium. Pattern 1 predominated in older subjects and smokers. Aging and smoking decrease arterial elasticity, thereby decreasing the volume and duration of retrograde blood flow in diastole. Doppler ultrasonography of the posterior intercostal arteries and aorta in 42 healthy subjects revealed that retrograde blood flow in late systole/early diastole is normal in subjects in the 15–34 age group. Transition from retrograde to antegrade flow was associated with transient blood stasis. This stasis should prolong the residence time of lipid-rich particles, enhancing diffusion into the vessel wall. A region of lower flow velocity was noted in the periostial region in all patients during diastole. The anatomic, hemodynamic, and risk factor data suggest that the morphology of fatty streaks is determined by interaction of retrograde with antegrade blood flow as modulated by arterial elasticity. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Atherosclerosis; Fatty streak; Hemodynamics; Aorta
1. Introduction Although fatty streaks cause no symptoms, they are of interest as a possible precursor to clinically-relevant raised atherosclerotic plaques. The colocalization of fatty streaks and raised lesions noted in the proximal coronary arteries [1] raises the possibility that one hemodynamic abnormality is involved in the development of both lesions. Until recently, there has been considerable controversy over what hemodynamic abnormality leads to the localization of atherosclerosis [2]. Recent in vitro [3], ex vivo [4], and in vivo [5,6] data suggest that atherosclerotic lesions develop in areas of low shear. This notion, however, is not new. William Boyd, in the 1961 edition of his classic pathology text, * Corresponding author. Tel.: +1-504-5686083; Fax: + 1-5045686037; e-mail: [email protected].
noted that ‘‘changes in blood flow, such as slowing or eddies, favor deposition of platelets. This agrees with the localization of plaques around areas where there are changes in the blood flow, such as bifurcations or origin of vessels’’ [7]. In addition to favoring deposition of platelets, areas of low shear should also allow a prolonged residence time for lipid-rich particles such as low-density lipoprotein, fostering time-dependent diffusion into the vessel wall. Thus, areas of prolonged residence time may be visualized by staining for lipid. Therefore, the authors examined Sudan IV stained aortas, using material collected and stained in the Pathobiological Determinants of Atherosclerosis in Youth study [8–15], and correlated the results with quantitative risk factor data. In order to determine the hemodynamic milieu around aortic ostia, vascular sonography was performed on subjects of comparable age.
0021-9150/98/$ - see front matter © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 9 8 ) 0 0 1 6 7 - 1
154
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
2. Methods Seventy-four aortas from individuals aged 15–42 who died suddenly, typically of trauma, were examined. Specimens were collected by the fourteen cooperating institutions of the Pathobiological Determinants of Atherosclerosis in Youth Research Group. Aortas were removed from a point 2 cm proximal to the ligamentum arteriosum to a point 2 cm distal to the iliac bifurcation. Branching arteries were severed close to the aortic wall, and adventitial fat was removed by sharp dissection. Aortas were opened on the dorsal surface along a line midway between the orifices of the intercostal and lumbar arteries. The intimal surface was then rinsed with Hanks’ modified balanced salt solution, and flattened with the adventitial surface downward. Aortas were then bisected longitudinally along a line on the ventral surface midway between the intercostal and lumbar ostia. The left half was covered with absorbent cotton and fixed in 10% neutral buffered formalin in a flat pan for 48 h. After shipping to a central laboratory, the left half of each aorta was stained with Sudan IV. Each aorta was then stored in 10% formalin in a clear plastic bag, allowing examination of the intimal surface. The intimal surface, including intercostal, subcostal, and lumbar ostia, was then examined for the presence of Sudan staining. For statistical analysis, subjects were grouped by predominant pattern of fatty streak. Risk factor data, including age, sex, total serum cholesterol, serum high density lipoprotein cholesterol (HDL-C), and serum thiocyanate, a surrogate marker of cigarette smoking, were collected as previously described [10–14]. Mean arterial pressure during life was estimated by determining the ratio of intimal thickness to outer diameter of small renal arteries [15]. Subjects were classed as definitely hypertensive, borderline hypertensive, or normal. For statistical analyses, mean arterial pressure was scored as follows: definite hypertension= 3, borderline hypertension= 2, and normal= 1. The statistical significance of differences between groups was determined by Student t-tests. Statistical analyses were performed on Microsoft Excel (Microsoft, Redmond, WA). All data are expressed as mean2 S.E.M. Vascular sonography of the aorta and its intercostal branches was performed on 43 healthy subjects, aged 23 – 44, using state of the art color Doppler ultrasound machines (Acuson XP-128, Mountain View, CA; ATL HDI-3000, Bothel, Washington, DC) with transducer frequencies varying between 3.5 and 7 MHz. The lower thoracic aorta was initially imaged in the supine position using gray scale and Doppler techniques. In the same setting, the subject was then turned prone and
using a left posterior paraspinal window, the posterior intercostal arteries were Doppler sampled roughly 2–3 cm distal to their origin from the aorta. The dorsal rami of the intercostal arteries were easily seen and not sampled. Respiratory maneuvers including deep inspiration, expiration, and Valsalva were performed on the initial ten subjects and demonstrated no affect on the Doppler waveform of the aorta or intercostal arteries. The subsequent 34 subjects were scanned in quiet respiration. None of the studied subjects had a history of hypertension. Blood pressure measurements were performed on all eight subjects over the age of 35 and all were normotensive.
3. Results
3.1. Anatomic studies A total of 701 ostia from the 74 aortas were examined. Two patterns of fatty streak were identified. The most common (Pattern 1), involving 57.0% of ostia, was characterized by broad bands of intense staining which extended to the proximal edges of ostia (Fig. 1). Occasionally, the edge of a lesion was continuous with the line of the flow divider, and convex to that ostium. Typically, the ventral intima was free of staining in this pattern. Pattern 1 predominated in 78% of aorta (n= 58). The second pattern (Fig. 2), involving 13.6% of ostia, was characterized by less intense staining, the edge of which was roughly continuous with the line of the flow divider and concave to that ostium. These lesions spared a well-defined zone surrounding the proximal edges of ostia. Occasionally, staining extended to involve a considerable portion of the ventral intima. Thus, the area stained in Pattern 2 corresponded to the area spared in Pattern 1. Pattern 2 predominated in 11% of aorta (n =8). Both Patterns 1 and 2 were found in thoracic and abdominal regions. Both patterns were noted within a single aorta (Fig. 3). Six aortas had an equal number of ostia with each pattern of fatty streak (8%). An indeterminate pattern of Sudan staining was seen around 2.7% of ostia. No fatty streaks were noted around 26.7% of ostia. Two aortas had no sudanophilia associated with ostia.
3.2. Demographic and risk factor data Complete data were not available from every subject from whom aortas were obtained. Data regarding age, sex and race (n=70), serum total cholesterol (n=52), serum HDL-C (n= 45), serum thiocyanate (n=51), and blood pressure (n=44) were available. Fifty-two subjects were male, and eighteen were female. Thirty-
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
155
eight subjects were African – American, and thirty-two were white. Mean age of subjects from whom aortas were obtained was 25.1 91.2 years (range: 15–42 years). Mean serum total cholesterol was 1809 19 mg/ dl (range: 43–376 mg/dl). Mean serum HDL-C was 499 9 mg/dl (range: 3.5 – 176 mg/dl). Mean serum thiocyanate was 949 13.5 mmol/l (range: 26.5 – 209 mmol/l). No difference was noted between groups with a predominance of Patterns 1 or 2 fatty streaks in sex or race (P= 0.97 and 0.88, respectively). No difference was noted between these groups in serum total cholesterol (184 920 vs. 134 945 mg/dl, respectively, P= 0.12), serum HDL-C (50 910 vs. 39 9 8 mg/dl, respectively, P = 0.45), or mean arterial blood pressure (P = 0.84). Two definitely hypertensive subjects and all three borderline hypertensive subjects showed a predominance of Pattern 1. One definitely hypertensive subjects had a predominance of Pattern 2.
Fig. 2. Photograph of a previously bisected thoracic aorta, showing fatty streaks typical of Pattern 2. Note that the intima proximal to the intercostal ostia is unstained. However, relatively weak stain is widespread on the ventral intima.
Subjects in whom Pattern 1 predominated had significantly higher serum thiocyanate levels (100914 vs. 459 18 mmol/l, respectively, P= 0.01). Subjects in whom the dominant pattern of fatty streak was Pattern 1 were older than those with Pattern 2 (25.891.3 vs. 20.593.5 years, P= 0.004) The mean age of subjects with equal prevalence of both patterns of fatty streak was 23.59 2 years. This was not significantly different from the groups showing predominantly Patterns 1 or 2 (P= 0.48 and 0.29, respectively). No significant differences were noted in race, sex, or mean arterial blood pressure between this group and any other (data not shown). Insufficient data regarding total serum cholesterol, HDL-C, and thiocyanate was present in this group to perform statistics. Fig. 1. Photograph of a previously bisected aorta, showing thoracic and proximal abdominal regions. Fatty streaks are typical of Pattern 1, characterized by relatively intense stain, located proximal to ostia, most of which are located dorsally. Note that the ventral aorta is free of stain, except for the region proximal to the celiac and superior mesenteric arteries.
3.3. Vascular sonography data Retrograde blood flow during late systole/early diastole was noted in both the aorta (Fig. 4A) and posterior
156
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
intercostal arteries (Fig. 4B) in 27 of 31 subjects aged 15 – 34 (Table 1). A region of lower flow velocity was noted in the periostial region in all patients during diastole. The transition from retrograde flow in late systole/early diastole to antegrade flow was associated with a period of stasis. Both the amplitude and duration of retrograde flow varied between subjects. The prevalence of continuous antegrade flow throughout the cardiac cycle in the thoracic aorta (Fig. 5A) and posterior intercostals (Fig. 5B) increased with age, being present in all subjects over 36 years of age.
4. Discussion This is the first description of two distinct morphologic patterns of fatty streaks. Previously, aortic fatty streaks have usually been described as linear lesions extending between and to the side of intercostal ostia, which usually spare the flow dividers [16]. This descrip-
Fig. 3. Photograph of a previously bisected thoracic aorta, showing both Pattern 1 and Pattern 2 fatty streaks in the same aorta. The four most proximal ostia contain Pattern 2 lesions. The five ostia distal to those show Pattern 1.
tion corresponds to the most common lesion, Pattern 1. To the authors’ knowledge, the only other investigator to identify another pattern is Stehbens, who noted that the lateral borders of some lesions curve back toward the next ostium distally, giving the lesions a slightly scalloped appearance [17]. The authors’ examination of previously published figures by Zinserling [18] and Holman et al. [19], reveals staining consistent with both of the patterns identified in the present work. Fatty streaks are thought to develop by diffusion of lipid rich particles into the vessel wall [20]. Because diffusion is a time dependent process, it should increase in areas of low shear and prolonged residence time. Thus, sudanophilia is probably a consequence of increased residence time due to local hemodynamic conditions in the vicinity of branching arteries. Indeed, all previous investigators have noted that the majority of sudanophilia in the aorta is on the dorsal surface in the vicinity of ostia [20], suggesting that ostia play some role in fatty streak development. In a column of flowing blood, variation in the velocity of flow exists across the column such that maximal velocity flow is present in the center of the column and minimal velocity at the periphery. In the central higher velocity portion of the column of blood, there is little variation in the range of velocities. At the periphery of the column near the side wall, friction causes a slowing of forward flow, producing a broad spectrum of velocities, so-called spectral broadening, that is easily apparent at Doppler sonography. This region of low shear and slower flow is accentuated at points where small branches originate from larger vessels, the so-called exit effect of the velocity profile of the retrograde wave [21]. These hemodynamic effects provide a permissive environment for prolonged residence time at the level of the intercostal ostia. The association of Pattern 1 with increased age and higher serum thiocyanate, a surrogate marker of cigarette smoking, suggests a possible mechanism for fatty streak morphogenesis. Both aging [22] and smoking [23,24] are associated with decreased arterial elasticity, which decreases the velocity and volume of retrograde blood flow [23]. The sonographic data document that retrograde blood flow is normal in this age group, and that turbulence is a constant feature in the periostial region. Thus, the authors postulate that the morphologic pattern of fatty streak is determined by the interaction of retrograde and antegrade blood flow as modulated by arterial elasticity. We propose that in Pattern 1, decreased volume and duration of retrograde blood flow (relative to those with Pattern 2) results in stasis in the proximal portion an aortic branch, which is the portion of the vessel furthest from the reflection site of antegrade waves at peripheral resistance vessels. The border of this area of stasis is delimited by aortic blood flow directed toward
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
157
Fig. 4. Doppler spectral tracings from a 31-year-old male demonstrating reversal of flow in late systole/early diastole. (A) Tracing of lower thoracic aorta. (B) Tracing of a posterior intercostal (segmental) artery.
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
158
Table 1 Prevalence of retrograde flow in late systole/early diastole by age group Age
n
Retrograde blood flow during late systole/early diastole (%)
15–22 23–30 31–36 37–44
3 27 6 7
100 89 67 0
the ostium. Thus, the margin of the zone of stasis and fatty streak is convex to the ostium (Fig. 6). In Pattern 2, greater arterial elasticity, and with it, volume and duration of retrograde blood flow, allows the propagation of retrograde blood flow throughout the entire length of the aortic branch, pushing the zone of stasis away from the ostium. Thus, the edge of these
fatty streaks is defined by flow away from the ostium, and is concave to that ostium (Fig. 7). Thus, the area stained in Pattern 1 is the area spared in Pattern 2. These results are significant for several reasons. First, they provide anatomic evidence implicating prolonged residence time in the pathogenesis of fatty streaks. Second, if the pattern of fatty streak in an individual changes from Pattern 2 to Pattern 1 with advancing age, then the location of lipid accumulation in the vessel wall must change over time. Thus, these data also provide anatomic evidence of lipid turnover in the aorta. The diameter of an ostium, angle of branching, diameter and length of the branch, microvascular tone, and blood viscosity all contribute to the resistance to blood flow, and thus may play a role in forming the hemodynamic milieu around ostia. Furthermore, pulse rate, which determines the duration of diastole, may
Fig. 5. Doppler spectral tracings from a 40-year-old female demonstrating continuous antegrade flow throughout the cardiac cycle, without flow reversal in late systole/early diastole. (A) The spectral tracing of the lower throacic aorta. (B) The spectral tracing of a posterior intercostal (segmental) artery.
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
159
Fig. 6. Detail of Fig. 1, showing direction of blood flow in late systole/early diastole postulated in pathogenesis of Pattern 1. Note that the edge of staining is continuous with the flow divider of the ostia, and convex to that ostia.
Fig. 7. Detail of Fig. 2, showing direction of late systolic/early diastolic blood flow postulated in pathogenesis of Pattern 2. Note that the edge of staining is concave to the adjacent ostia.
also play a role in fatty streak morphogenesis. Recent work is elucidating the hemodynamic effects of lipoproteins, suggesting new roles for these particles in the pathogenesis of fatty streaks. Low density lipoprotein increases and HDL decreases blood viscosity in areas of low shear [25], and thus may contribute to stasis in the transition from retrograde to antegrade flow. Modulation of permeability by shear should be considered in determining the localization of fatty streaks. Many investigators have shown in vitro that endothelial permeability increases as a function of flow or shear within minutes of exposure to steady laminar flow [26]. However, the relevance of these in vitro observations in conditions of steady flow to pulsatile hemodynamics in vivo is questionable. Davies et al. have shown that endothelial responses to steady and oscillating shear are clearly different [27]. Therefore, it is possible that increased diffusion facilitated by prolonged residence may be more important in determining localization of fatty streaks than changes in permeability.
This work is limited by the small number of cases showing Pattern 2 and cases with equal prevalence of both patterns. Further work with larger study populations may elucidate weaker associations between risk factors such as mean arterial blood pressure and fatty streak pattern, as well as if those aortas with equal numbers of both patterns are truly intermediate in age between those with a predominance of Patterns 1 or 2. In summary, the authors have identified two distinct patterns of aortic fatty streak. The anatomic, hemodynamic, and risk factor data suggest that the morphology of fatty streaks is determined by the interaction of retrograde with antegrade blood flow as modulated by arterial elasticity.
Acknowledgements Thanks to Dr Donald A. Boudreau, Department of Pathology, Louisiana State University School of
160
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
Medicine in New Orleans, for his helpful discussions. This research is supported by grant HL-45720 from the NIH-NHLBI. [14]
References
[15]
[1] Montenegro MR, Eggen DA. Topography of atherosclerosis in the coronary arteries. Lab Investig 1968;18:126–33. [2] Woolf N. Pathology of atherosclerosis. Boston, MA: Butterworth Scientific, 1982:187. [3] Ku DN, Giddens DP. Pulsatile flow in a model carotid bifurcation. Arteriosclerosis 1983;3:31–9. [4] Asakura T, Karino T. Flow patterns and spatial distribution of atherosclerotic lesions in human coronary arteries. Circ Res 1990;66:1045 – 66. [5] Gibson CM, Diaz L, Kandarpa K, Sacks FM, Pasternak RC, Sandor T, Feldman C, Stone PH. Relation of vessel wall shear stress to atherosclerosis progression in human coronary arteries. Arterioscler Thromb 1993;13:310–5. [6] Sawchuck AP, Unthank JL, Davis TE, Dalsing MC. A prospective, in vivo study of the relationship between blood flow hemodynamics and atherosclerosis in a hyperlipidemic swine model. J Vasc Surg 1994;19:58–64. [7] Boyd W. A textbook of pathology. Seventh Edition. Philadelphia: Lea and Febiger, 1961:512. [8] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Strong JP. Natural history and risk factors for early human atherogenesis. Clin Chem 1995;41:134 – 9. [9] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Wissler RW. New insights into the pathogenesis of atherosclerosis as revealed by PDAY. Atherosclerosis 1994;108 (Suppl):3–20. [10] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Strong JP, Oalmann MC, Malcom GT, Newmann III WP, McMahan CA. Rationale, methodology, and selected risk factor findings. Cardiovasc Risk Factors 1992;2:22 – 30. [11] Wissler RW. USA multicenter study of the pathobiology of atherosclerosis in youth. Ann NY Acad Sci 1991;623:26– 39. [12] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Natural history of aortic and coronary atherosclerotic lesions in youth: findings from the PDAY study. Arterioscler Thromb 1993;13:1291–8. [13] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Strong JP, Oalmann MC, Malcom
.
[16] [17]
[18]
[19] [20]
[21]
[22] [23]
[24]
[25]
[26]
[27]
GT Jr. Atherosclerosis in youth: relationship of risk factors to arterial lesions. In: Lauer RM, Luepker RV, Filer LJ Jr, editors. Monograph on prevention of atherosclerosis and hypertension beginning in youth. Malvern, PA: Lea and Febiger, 1994:13–8. Hornick CA, Baker HN, Malcom GT, Newman WP, Roheim PS, Strong JP. Lipoproteins and apolipoproteins in postmortem serum. Mod Pathol 1988;2:480 – 4. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, McGill HC Jr, Strong JP, Tracy RE, McMahan CA, Oalmann MC. Relation of a postmortem renal index of hypertension to atherosclerosis in youth. Arterioscler Thromb Vasc Biol 1995;15:2222 – 8. Strong JP. Atherosclerotic lesions: natural history, risk factors and topography. Arch Pathol Lab Med 1992;116:1268–75. Stehbens WE. Atherosclerosis and degenerative diseases of blood vessels. In: Stehbens WE, Lie JT, editors. Vascular pathology. New York: Chapman and Hall, 1995:177. Zinserling WD. Unttersuchungen uber Atherosklerose. I. Uber die aortaverfettung bei kindern. Virchow Arch Path Anat 1925;255:677 – 705. Holman RL, McGill HC, Strong JP, Geer JC. The natural history of atherosclerosis. Am J Pathol 1958;34:209 – 35. Stary HC, Chandler AB, Glagov S, et al. A definition of initial, fatty streak, and intermediate lesions of atherosclerosis. Circulation 1994;89:2462 – 78. Burns PN. Hemodynamics. In: Taylor KJW, Burns PN, Wells PN, editors. Clinical application of Doppler ultrasound. New York: Raven Press, 1995:35 – 54. Nichols WW, O’Rourke MF. McDonald’s blood flow in arteries. Philadelphia: Lea and Febiger, 1990:402. Caro CG. Cigarette smoking causes acute changes in arterial wall mechanics and the pattern of arterial blood flow in healthy subjects: possible insight into mechanisms of atherogenesis. In: Diana JN, editor. Tobacco smoking and atherosclerosis (Ann Exp Med Biol, vol. 273). New York: Plenum Press, 1990:273–80. Levenson J, Simon AC, Cambien FA, Beretti C. Cigarette smoking and hypertension-factors independently associated with blood hyperviscosity and arterial rigidity. Arteriosclerosis 1987;7:572 – 7. Sloop GD, Garber DW. The effects of low-density lipoprotein and high-density lipoprotein on blood viscosity correlate with their association with risk of atherosclerosis in humans. Clin Sci 1997;92:473 – 9. Waters CM. Flow-induced modulation of the permeability of endothelial cells cultured on microcarrier beads. J Cell Physiol 1996;168:403 – 11. Davies PF, Dewey CF Jr, Bussolari SR, Gordon EJ, Gimbrone MA Jr. Influence of hemodynamic forces on vascular endothelial function. J Clin Invest 1984;73:1121 – 9.
A description of two morphologic patterns of aortic fatty streaks, and a hypothesis of their pathogenesis Gregory D. Sloop a,*, Robert S. Perret b, Jody S. Brahney a, Margaret Oalmann a a b
Department of Pathology, Louisiana State Uni6ersity School of Medicine, 1901 Perdido, New Orleans, LA 70112, USA Department of Radiology, Louisiana State Uni6ersity School of Medicine, 1901 Perdido, New Orleans, LA 70112, USA Received 26 August 1997; received in revised form 27 May 1998; accepted 5 June 1998
Abstract Two morphologic patterns of fatty streak were identified on examination of 74 aortas from the Pathobiological Determinants of Atherosclerosis in Youth study. Pattern 1, which predominated in 78% of aortas, is characterized by broad bands of intense stain which extend to the proximal edge of ostia. Pattern 2, which predominated in 11%, is characterized by less intense staining which is concave to the associated ostium. Pattern 1 predominated in older subjects and smokers. Aging and smoking decrease arterial elasticity, thereby decreasing the volume and duration of retrograde blood flow in diastole. Doppler ultrasonography of the posterior intercostal arteries and aorta in 42 healthy subjects revealed that retrograde blood flow in late systole/early diastole is normal in subjects in the 15–34 age group. Transition from retrograde to antegrade flow was associated with transient blood stasis. This stasis should prolong the residence time of lipid-rich particles, enhancing diffusion into the vessel wall. A region of lower flow velocity was noted in the periostial region in all patients during diastole. The anatomic, hemodynamic, and risk factor data suggest that the morphology of fatty streaks is determined by interaction of retrograde with antegrade blood flow as modulated by arterial elasticity. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Atherosclerosis; Fatty streak; Hemodynamics; Aorta
1. Introduction Although fatty streaks cause no symptoms, they are of interest as a possible precursor to clinically-relevant raised atherosclerotic plaques. The colocalization of fatty streaks and raised lesions noted in the proximal coronary arteries [1] raises the possibility that one hemodynamic abnormality is involved in the development of both lesions. Until recently, there has been considerable controversy over what hemodynamic abnormality leads to the localization of atherosclerosis [2]. Recent in vitro [3], ex vivo [4], and in vivo [5,6] data suggest that atherosclerotic lesions develop in areas of low shear. This notion, however, is not new. William Boyd, in the 1961 edition of his classic pathology text, * Corresponding author. Tel.: +1-504-5686083; Fax: + 1-5045686037; e-mail: [email protected].
noted that ‘‘changes in blood flow, such as slowing or eddies, favor deposition of platelets. This agrees with the localization of plaques around areas where there are changes in the blood flow, such as bifurcations or origin of vessels’’ [7]. In addition to favoring deposition of platelets, areas of low shear should also allow a prolonged residence time for lipid-rich particles such as low-density lipoprotein, fostering time-dependent diffusion into the vessel wall. Thus, areas of prolonged residence time may be visualized by staining for lipid. Therefore, the authors examined Sudan IV stained aortas, using material collected and stained in the Pathobiological Determinants of Atherosclerosis in Youth study [8–15], and correlated the results with quantitative risk factor data. In order to determine the hemodynamic milieu around aortic ostia, vascular sonography was performed on subjects of comparable age.
0021-9150/98/$ - see front matter © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 9 8 ) 0 0 1 6 7 - 1
154
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
2. Methods Seventy-four aortas from individuals aged 15–42 who died suddenly, typically of trauma, were examined. Specimens were collected by the fourteen cooperating institutions of the Pathobiological Determinants of Atherosclerosis in Youth Research Group. Aortas were removed from a point 2 cm proximal to the ligamentum arteriosum to a point 2 cm distal to the iliac bifurcation. Branching arteries were severed close to the aortic wall, and adventitial fat was removed by sharp dissection. Aortas were opened on the dorsal surface along a line midway between the orifices of the intercostal and lumbar arteries. The intimal surface was then rinsed with Hanks’ modified balanced salt solution, and flattened with the adventitial surface downward. Aortas were then bisected longitudinally along a line on the ventral surface midway between the intercostal and lumbar ostia. The left half was covered with absorbent cotton and fixed in 10% neutral buffered formalin in a flat pan for 48 h. After shipping to a central laboratory, the left half of each aorta was stained with Sudan IV. Each aorta was then stored in 10% formalin in a clear plastic bag, allowing examination of the intimal surface. The intimal surface, including intercostal, subcostal, and lumbar ostia, was then examined for the presence of Sudan staining. For statistical analysis, subjects were grouped by predominant pattern of fatty streak. Risk factor data, including age, sex, total serum cholesterol, serum high density lipoprotein cholesterol (HDL-C), and serum thiocyanate, a surrogate marker of cigarette smoking, were collected as previously described [10–14]. Mean arterial pressure during life was estimated by determining the ratio of intimal thickness to outer diameter of small renal arteries [15]. Subjects were classed as definitely hypertensive, borderline hypertensive, or normal. For statistical analyses, mean arterial pressure was scored as follows: definite hypertension= 3, borderline hypertension= 2, and normal= 1. The statistical significance of differences between groups was determined by Student t-tests. Statistical analyses were performed on Microsoft Excel (Microsoft, Redmond, WA). All data are expressed as mean2 S.E.M. Vascular sonography of the aorta and its intercostal branches was performed on 43 healthy subjects, aged 23 – 44, using state of the art color Doppler ultrasound machines (Acuson XP-128, Mountain View, CA; ATL HDI-3000, Bothel, Washington, DC) with transducer frequencies varying between 3.5 and 7 MHz. The lower thoracic aorta was initially imaged in the supine position using gray scale and Doppler techniques. In the same setting, the subject was then turned prone and
using a left posterior paraspinal window, the posterior intercostal arteries were Doppler sampled roughly 2–3 cm distal to their origin from the aorta. The dorsal rami of the intercostal arteries were easily seen and not sampled. Respiratory maneuvers including deep inspiration, expiration, and Valsalva were performed on the initial ten subjects and demonstrated no affect on the Doppler waveform of the aorta or intercostal arteries. The subsequent 34 subjects were scanned in quiet respiration. None of the studied subjects had a history of hypertension. Blood pressure measurements were performed on all eight subjects over the age of 35 and all were normotensive.
3. Results
3.1. Anatomic studies A total of 701 ostia from the 74 aortas were examined. Two patterns of fatty streak were identified. The most common (Pattern 1), involving 57.0% of ostia, was characterized by broad bands of intense staining which extended to the proximal edges of ostia (Fig. 1). Occasionally, the edge of a lesion was continuous with the line of the flow divider, and convex to that ostium. Typically, the ventral intima was free of staining in this pattern. Pattern 1 predominated in 78% of aorta (n= 58). The second pattern (Fig. 2), involving 13.6% of ostia, was characterized by less intense staining, the edge of which was roughly continuous with the line of the flow divider and concave to that ostium. These lesions spared a well-defined zone surrounding the proximal edges of ostia. Occasionally, staining extended to involve a considerable portion of the ventral intima. Thus, the area stained in Pattern 2 corresponded to the area spared in Pattern 1. Pattern 2 predominated in 11% of aorta (n =8). Both Patterns 1 and 2 were found in thoracic and abdominal regions. Both patterns were noted within a single aorta (Fig. 3). Six aortas had an equal number of ostia with each pattern of fatty streak (8%). An indeterminate pattern of Sudan staining was seen around 2.7% of ostia. No fatty streaks were noted around 26.7% of ostia. Two aortas had no sudanophilia associated with ostia.
3.2. Demographic and risk factor data Complete data were not available from every subject from whom aortas were obtained. Data regarding age, sex and race (n=70), serum total cholesterol (n=52), serum HDL-C (n= 45), serum thiocyanate (n=51), and blood pressure (n=44) were available. Fifty-two subjects were male, and eighteen were female. Thirty-
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
155
eight subjects were African – American, and thirty-two were white. Mean age of subjects from whom aortas were obtained was 25.1 91.2 years (range: 15–42 years). Mean serum total cholesterol was 1809 19 mg/ dl (range: 43–376 mg/dl). Mean serum HDL-C was 499 9 mg/dl (range: 3.5 – 176 mg/dl). Mean serum thiocyanate was 949 13.5 mmol/l (range: 26.5 – 209 mmol/l). No difference was noted between groups with a predominance of Patterns 1 or 2 fatty streaks in sex or race (P= 0.97 and 0.88, respectively). No difference was noted between these groups in serum total cholesterol (184 920 vs. 134 945 mg/dl, respectively, P= 0.12), serum HDL-C (50 910 vs. 39 9 8 mg/dl, respectively, P = 0.45), or mean arterial blood pressure (P = 0.84). Two definitely hypertensive subjects and all three borderline hypertensive subjects showed a predominance of Pattern 1. One definitely hypertensive subjects had a predominance of Pattern 2.
Fig. 2. Photograph of a previously bisected thoracic aorta, showing fatty streaks typical of Pattern 2. Note that the intima proximal to the intercostal ostia is unstained. However, relatively weak stain is widespread on the ventral intima.
Subjects in whom Pattern 1 predominated had significantly higher serum thiocyanate levels (100914 vs. 459 18 mmol/l, respectively, P= 0.01). Subjects in whom the dominant pattern of fatty streak was Pattern 1 were older than those with Pattern 2 (25.891.3 vs. 20.593.5 years, P= 0.004) The mean age of subjects with equal prevalence of both patterns of fatty streak was 23.59 2 years. This was not significantly different from the groups showing predominantly Patterns 1 or 2 (P= 0.48 and 0.29, respectively). No significant differences were noted in race, sex, or mean arterial blood pressure between this group and any other (data not shown). Insufficient data regarding total serum cholesterol, HDL-C, and thiocyanate was present in this group to perform statistics. Fig. 1. Photograph of a previously bisected aorta, showing thoracic and proximal abdominal regions. Fatty streaks are typical of Pattern 1, characterized by relatively intense stain, located proximal to ostia, most of which are located dorsally. Note that the ventral aorta is free of stain, except for the region proximal to the celiac and superior mesenteric arteries.
3.3. Vascular sonography data Retrograde blood flow during late systole/early diastole was noted in both the aorta (Fig. 4A) and posterior
156
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
intercostal arteries (Fig. 4B) in 27 of 31 subjects aged 15 – 34 (Table 1). A region of lower flow velocity was noted in the periostial region in all patients during diastole. The transition from retrograde flow in late systole/early diastole to antegrade flow was associated with a period of stasis. Both the amplitude and duration of retrograde flow varied between subjects. The prevalence of continuous antegrade flow throughout the cardiac cycle in the thoracic aorta (Fig. 5A) and posterior intercostals (Fig. 5B) increased with age, being present in all subjects over 36 years of age.
4. Discussion This is the first description of two distinct morphologic patterns of fatty streaks. Previously, aortic fatty streaks have usually been described as linear lesions extending between and to the side of intercostal ostia, which usually spare the flow dividers [16]. This descrip-
Fig. 3. Photograph of a previously bisected thoracic aorta, showing both Pattern 1 and Pattern 2 fatty streaks in the same aorta. The four most proximal ostia contain Pattern 2 lesions. The five ostia distal to those show Pattern 1.
tion corresponds to the most common lesion, Pattern 1. To the authors’ knowledge, the only other investigator to identify another pattern is Stehbens, who noted that the lateral borders of some lesions curve back toward the next ostium distally, giving the lesions a slightly scalloped appearance [17]. The authors’ examination of previously published figures by Zinserling [18] and Holman et al. [19], reveals staining consistent with both of the patterns identified in the present work. Fatty streaks are thought to develop by diffusion of lipid rich particles into the vessel wall [20]. Because diffusion is a time dependent process, it should increase in areas of low shear and prolonged residence time. Thus, sudanophilia is probably a consequence of increased residence time due to local hemodynamic conditions in the vicinity of branching arteries. Indeed, all previous investigators have noted that the majority of sudanophilia in the aorta is on the dorsal surface in the vicinity of ostia [20], suggesting that ostia play some role in fatty streak development. In a column of flowing blood, variation in the velocity of flow exists across the column such that maximal velocity flow is present in the center of the column and minimal velocity at the periphery. In the central higher velocity portion of the column of blood, there is little variation in the range of velocities. At the periphery of the column near the side wall, friction causes a slowing of forward flow, producing a broad spectrum of velocities, so-called spectral broadening, that is easily apparent at Doppler sonography. This region of low shear and slower flow is accentuated at points where small branches originate from larger vessels, the so-called exit effect of the velocity profile of the retrograde wave [21]. These hemodynamic effects provide a permissive environment for prolonged residence time at the level of the intercostal ostia. The association of Pattern 1 with increased age and higher serum thiocyanate, a surrogate marker of cigarette smoking, suggests a possible mechanism for fatty streak morphogenesis. Both aging [22] and smoking [23,24] are associated with decreased arterial elasticity, which decreases the velocity and volume of retrograde blood flow [23]. The sonographic data document that retrograde blood flow is normal in this age group, and that turbulence is a constant feature in the periostial region. Thus, the authors postulate that the morphologic pattern of fatty streak is determined by the interaction of retrograde and antegrade blood flow as modulated by arterial elasticity. We propose that in Pattern 1, decreased volume and duration of retrograde blood flow (relative to those with Pattern 2) results in stasis in the proximal portion an aortic branch, which is the portion of the vessel furthest from the reflection site of antegrade waves at peripheral resistance vessels. The border of this area of stasis is delimited by aortic blood flow directed toward
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
157
Fig. 4. Doppler spectral tracings from a 31-year-old male demonstrating reversal of flow in late systole/early diastole. (A) Tracing of lower thoracic aorta. (B) Tracing of a posterior intercostal (segmental) artery.
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
158
Table 1 Prevalence of retrograde flow in late systole/early diastole by age group Age
n
Retrograde blood flow during late systole/early diastole (%)
15–22 23–30 31–36 37–44
3 27 6 7
100 89 67 0
the ostium. Thus, the margin of the zone of stasis and fatty streak is convex to the ostium (Fig. 6). In Pattern 2, greater arterial elasticity, and with it, volume and duration of retrograde blood flow, allows the propagation of retrograde blood flow throughout the entire length of the aortic branch, pushing the zone of stasis away from the ostium. Thus, the edge of these
fatty streaks is defined by flow away from the ostium, and is concave to that ostium (Fig. 7). Thus, the area stained in Pattern 1 is the area spared in Pattern 2. These results are significant for several reasons. First, they provide anatomic evidence implicating prolonged residence time in the pathogenesis of fatty streaks. Second, if the pattern of fatty streak in an individual changes from Pattern 2 to Pattern 1 with advancing age, then the location of lipid accumulation in the vessel wall must change over time. Thus, these data also provide anatomic evidence of lipid turnover in the aorta. The diameter of an ostium, angle of branching, diameter and length of the branch, microvascular tone, and blood viscosity all contribute to the resistance to blood flow, and thus may play a role in forming the hemodynamic milieu around ostia. Furthermore, pulse rate, which determines the duration of diastole, may
Fig. 5. Doppler spectral tracings from a 40-year-old female demonstrating continuous antegrade flow throughout the cardiac cycle, without flow reversal in late systole/early diastole. (A) The spectral tracing of the lower throacic aorta. (B) The spectral tracing of a posterior intercostal (segmental) artery.
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
159
Fig. 6. Detail of Fig. 1, showing direction of blood flow in late systole/early diastole postulated in pathogenesis of Pattern 1. Note that the edge of staining is continuous with the flow divider of the ostia, and convex to that ostia.
Fig. 7. Detail of Fig. 2, showing direction of late systolic/early diastolic blood flow postulated in pathogenesis of Pattern 2. Note that the edge of staining is concave to the adjacent ostia.
also play a role in fatty streak morphogenesis. Recent work is elucidating the hemodynamic effects of lipoproteins, suggesting new roles for these particles in the pathogenesis of fatty streaks. Low density lipoprotein increases and HDL decreases blood viscosity in areas of low shear [25], and thus may contribute to stasis in the transition from retrograde to antegrade flow. Modulation of permeability by shear should be considered in determining the localization of fatty streaks. Many investigators have shown in vitro that endothelial permeability increases as a function of flow or shear within minutes of exposure to steady laminar flow [26]. However, the relevance of these in vitro observations in conditions of steady flow to pulsatile hemodynamics in vivo is questionable. Davies et al. have shown that endothelial responses to steady and oscillating shear are clearly different [27]. Therefore, it is possible that increased diffusion facilitated by prolonged residence may be more important in determining localization of fatty streaks than changes in permeability.
This work is limited by the small number of cases showing Pattern 2 and cases with equal prevalence of both patterns. Further work with larger study populations may elucidate weaker associations between risk factors such as mean arterial blood pressure and fatty streak pattern, as well as if those aortas with equal numbers of both patterns are truly intermediate in age between those with a predominance of Patterns 1 or 2. In summary, the authors have identified two distinct patterns of aortic fatty streak. The anatomic, hemodynamic, and risk factor data suggest that the morphology of fatty streaks is determined by the interaction of retrograde with antegrade blood flow as modulated by arterial elasticity.
Acknowledgements Thanks to Dr Donald A. Boudreau, Department of Pathology, Louisiana State University School of
160
G.D. Sloop et al. / Atherosclerosis 141 (1998) 153–160
Medicine in New Orleans, for his helpful discussions. This research is supported by grant HL-45720 from the NIH-NHLBI. [14]
References
[15]
[1] Montenegro MR, Eggen DA. Topography of atherosclerosis in the coronary arteries. Lab Investig 1968;18:126–33. [2] Woolf N. Pathology of atherosclerosis. Boston, MA: Butterworth Scientific, 1982:187. [3] Ku DN, Giddens DP. Pulsatile flow in a model carotid bifurcation. Arteriosclerosis 1983;3:31–9. [4] Asakura T, Karino T. Flow patterns and spatial distribution of atherosclerotic lesions in human coronary arteries. Circ Res 1990;66:1045 – 66. [5] Gibson CM, Diaz L, Kandarpa K, Sacks FM, Pasternak RC, Sandor T, Feldman C, Stone PH. Relation of vessel wall shear stress to atherosclerosis progression in human coronary arteries. Arterioscler Thromb 1993;13:310–5. [6] Sawchuck AP, Unthank JL, Davis TE, Dalsing MC. A prospective, in vivo study of the relationship between blood flow hemodynamics and atherosclerosis in a hyperlipidemic swine model. J Vasc Surg 1994;19:58–64. [7] Boyd W. A textbook of pathology. Seventh Edition. Philadelphia: Lea and Febiger, 1961:512. [8] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Strong JP. Natural history and risk factors for early human atherogenesis. Clin Chem 1995;41:134 – 9. [9] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Wissler RW. New insights into the pathogenesis of atherosclerosis as revealed by PDAY. Atherosclerosis 1994;108 (Suppl):3–20. [10] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Strong JP, Oalmann MC, Malcom GT, Newmann III WP, McMahan CA. Rationale, methodology, and selected risk factor findings. Cardiovasc Risk Factors 1992;2:22 – 30. [11] Wissler RW. USA multicenter study of the pathobiology of atherosclerosis in youth. Ann NY Acad Sci 1991;623:26– 39. [12] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Natural history of aortic and coronary atherosclerotic lesions in youth: findings from the PDAY study. Arterioscler Thromb 1993;13:1291–8. [13] Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, Strong JP, Oalmann MC, Malcom
.
[16] [17]
[18]
[19] [20]
[21]
[22] [23]
[24]
[25]
[26]
[27]
GT Jr. Atherosclerosis in youth: relationship of risk factors to arterial lesions. In: Lauer RM, Luepker RV, Filer LJ Jr, editors. Monograph on prevention of atherosclerosis and hypertension beginning in youth. Malvern, PA: Lea and Febiger, 1994:13–8. Hornick CA, Baker HN, Malcom GT, Newman WP, Roheim PS, Strong JP. Lipoproteins and apolipoproteins in postmortem serum. Mod Pathol 1988;2:480 – 4. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group, McGill HC Jr, Strong JP, Tracy RE, McMahan CA, Oalmann MC. Relation of a postmortem renal index of hypertension to atherosclerosis in youth. Arterioscler Thromb Vasc Biol 1995;15:2222 – 8. Strong JP. Atherosclerotic lesions: natural history, risk factors and topography. Arch Pathol Lab Med 1992;116:1268–75. Stehbens WE. Atherosclerosis and degenerative diseases of blood vessels. In: Stehbens WE, Lie JT, editors. Vascular pathology. New York: Chapman and Hall, 1995:177. Zinserling WD. Unttersuchungen uber Atherosklerose. I. Uber die aortaverfettung bei kindern. Virchow Arch Path Anat 1925;255:677 – 705. Holman RL, McGill HC, Strong JP, Geer JC. The natural history of atherosclerosis. Am J Pathol 1958;34:209 – 35. Stary HC, Chandler AB, Glagov S, et al. A definition of initial, fatty streak, and intermediate lesions of atherosclerosis. Circulation 1994;89:2462 – 78. Burns PN. Hemodynamics. In: Taylor KJW, Burns PN, Wells PN, editors. Clinical application of Doppler ultrasound. New York: Raven Press, 1995:35 – 54. Nichols WW, O’Rourke MF. McDonald’s blood flow in arteries. Philadelphia: Lea and Febiger, 1990:402. Caro CG. Cigarette smoking causes acute changes in arterial wall mechanics and the pattern of arterial blood flow in healthy subjects: possible insight into mechanisms of atherogenesis. In: Diana JN, editor. Tobacco smoking and atherosclerosis (Ann Exp Med Biol, vol. 273). New York: Plenum Press, 1990:273–80. Levenson J, Simon AC, Cambien FA, Beretti C. Cigarette smoking and hypertension-factors independently associated with blood hyperviscosity and arterial rigidity. Arteriosclerosis 1987;7:572 – 7. Sloop GD, Garber DW. The effects of low-density lipoprotein and high-density lipoprotein on blood viscosity correlate with their association with risk of atherosclerosis in humans. Clin Sci 1997;92:473 – 9. Waters CM. Flow-induced modulation of the permeability of endothelial cells cultured on microcarrier beads. J Cell Physiol 1996;168:403 – 11. Davies PF, Dewey CF Jr, Bussolari SR, Gordon EJ, Gimbrone MA Jr. Influence of hemodynamic forces on vascular endothelial function. J Clin Invest 1984;73:1121 – 9.