Presence of Intraplaque Hemorrhage Stimulates Progression of Carotid Atherosclerotic Plaques
http://www.100md.com
《循环学杂志》
the Departments of Radiology (N.T., C.Y., B.C., T.S., M.S.F.), Pathology (R.S.),Medicine, Division of Medical Genetics (G.P.J.), University of Washington, Seattle
Department of Cardiology, Juntendo University School of Medicine, Tokyo, Japan (N.T.)
Mountain-Whisper-Light Statistical Consulting, Seattle, Wash (N.L.P.)
Department of Surgery, University of Washington, and Veterans Affairs Puget Sound Health Care System, Seattle, Wash (C.I., J.M., C.N., T.S.H.).
Abstract
Background— Previous studies suggest that erythrocyte membranes from intraplaque hemorrhage into the necrotic core are a source of free cholesterol and may become a driving force in the progression of atherosclerosis. We have shown that MRI can accurately identify carotid intraplaque hemorrhage and precisely measure plaque volume. We tested the hypothesis that hemorrhage into carotid atheroma stimulates plaque progression.
Methods and Results— Twenty-nine subjects (14 cases with intraplaque hemorrhage and 15 controls with comparably sized plaques without intraplaque hemorrhage at baseline) underwent serial carotid MRI examination with a multicontrast weighted protocol (T1, T2, proton density, and 3D time of flight) over a period of 18 months. The volumes of wall, lumen, lipid-rich necrotic core, calcification, and intraplaque hemorrhage were measured with a custom-designed image analysis tool. The percent change in wall volume (6.8% versus –0.15%; P=0.009) and lipid-rich necrotic core volume (28.4% versus –5.2%; P=0.001) was significantly higher in the hemorrhage group than in controls over the course of the study. Furthermore, those with intraplaque hemorrhage at baseline were much more likely to have new plaque hemorrhages at 18 months compared with controls (43% versus 0%; P=0.006).
Conclusions— Hemorrhage into the carotid atherosclerotic plaque accelerated plaque progression in an 18-month period. Repeated bleeding into the plaque may produce a stimulus for the progression of atherosclerosis by increasing lipid core and plaque volume and creating new destabilizing factors.
Key Words: magnetic resonance imaging ; carotid arteries ; hemorrhage ; atherosclerosis ; plaque
Introduction
The progression of atherosclerosis is a complex and often longstanding phenomenon. Many studies have shown that inflammation and the subsequent immune response contribute to atherogenesis. The recruitment of inflammatory cells in atherosclerotic lesions is a constitutive phenomenon throughout the process of plaque growth.1,2 Intraplaque hemorrhage is commonly observed in atherosclerotic plaques and is considered by some to be caused by rupture of plaque neovasculature. These microvessels are fragile because of lack of support by smooth muscle cells and focal discontinuity of the endothelial lining.3
In a recent histopathological study of coronary artery specimens, Kolodgie et al4 suggested that intraplaque hemorrhage may represent a potent atherogenic stimulus by contributing to the deposition of free cholesterol, macrophage infiltration, and enlargement of the necrotic core. Immunohistochemical staining with the antibody to glycophorin A, a protein specific to erythrocyte membranes, was strongly associated with the size of the necrotic core and degree of macrophage infiltration. They also noted that rabbit lesions with induced intramural hemorrhage had significantly greater lipid content than control lesions without hemorrhage. Other investigators have noted that erythrocyte membranes contain more free cholesterol than any other cell in the body.5
To test the hypothesis that intraplaque hemorrhage is a predisposing factor in progression of human atherosclerosis, an imaging tool is needed that can serially examine the morphological and compositional characteristics of the lesion over time. High-resolution MRI is ideally suited for this purpose because it is noninvasive and has proven capability for distinguishing tissue characteristics.6–9 Our laboratory recently demonstrated that multicontrast MRI can accurately detect the presence and age of carotid intraplaque hemorrhage.10 Furthermore, MRI is capable of precisely quantifying overall lesion size and plaque compositional features such as the lipid-rich necrotic core.11–13 Thus, this imaging tool provides a unique opportunity to examine the temporal relationship between baseline plaque characteristics and subsequent progression in lesion size and composition.
The aim of this study was to test the hypothesis that intraplaque hemorrhage, as detected by high-resolution MRI, is associated with greater progression in both lipid-rich necrotic core and plaque volume over a period of 18 months.
Methods
Study Population
Cases and controls were selected from an ongoing prospective serial carotid MRI study referred to as PRIMARI. Subjects for PRIMARI were recruited from the diagnostic vascular ultrasound laboratories at the University of Washington Medical Center and Veterans Affairs Puget Sound Health Care System after informed consent was obtained. The study procedures and consent forms were reviewed and approved by each site’s institutional review board. Inclusion criteria for the PRIMARI study included (1) 50% to 79% carotid stenosis by duplex ultrasound examination and (2) asymptomatic status with regard to carotid artery disease within the 6 months before enrollment in PRIMARI. Exclusion criteria included (1) prior carotid endarterectomy (CEA) on the side of the index carotid artery, (2) prior neck irradiation, and (3) contraindication for MRI. Subjects undergo bilateral carotid artery MRI every 18 months. At the time of analysis, 142 subjects had at least 2 MRI scans (baseline and 18 months). Among the 142 patients (284 arteries) available, 66 arteries were excluded because of marginal image quality, insufficient plaque size, or absence of the common carotid bifurcation in the field of view. In this serial MRI study, the common carotid artery on the index side was used as an internal fiducial marker to match cross-sectional locations between MRI examinations. In some cases, the bifurcation on the contralateral, nonindex side occurred at a different axial level, outside the field of view. Four arteries with total occlusion were excluded, as were 11 arteries with MRI evidence of fibrous cap rupture or ulceration with mural thrombus formation. The latter were excluded to focus the analysis of this study on the association between intraplaque hemorrhage (presumably from rupture of plaque neovasculature) and plaque progression.
Images from the baseline MRI examination were first reviewed to identify cases and controls blinded to the follow-up MR studies. Among the 203 arteries remaining, 14 arteries (99 cross-sectional locations) demonstrated clear MRI evidence of early or recent intraplaque hemorrhage. Fifteen arteries (103 cross-sectional locations) with comparable baseline plaque volume that clearly had no evidence of early, recent, or chronic intraplaque hemorrhage were selected as controls. The baseline and the 18-month follow-up MRI examinations were used to compare cases and controls for progression in overall plaque volume and plaque constituents. Two patients underwent CEA at 2 months and 1 year after the second MRI examination, and detailed histological analysis was performed for comparison with the MRI findings with the use of a previously published protocol.14
Baseline Clinical and Laboratory Data
Before the baseline MRI examination, study subjects were asked to complete a detailed health questionnaire and physical examination to assess risk factors for atherosclerosis (hypertension, diabetes, smoking, family history), history of coronary artery and peripheral vascular disease, current medications, and body mass index. Lipid measurements were performed on fasting whole plasma. Standard enzymatic methods were used to determine levels of total cholesterol, triglycerides, and HDL cholesterol on an Abbott Spectrum analyzer.15,16 LDL cholesterol was calculated by the Friedewald equation.
MRI Protocol
Patients were imaged with a custom-designed phased-array surface coil in a 1.5-T GE Signa Scanner (Horizon EchoSpeed, version 5.8, GE Medical Systems). A standardized protocol was used to obtain 4 different contrast-weighted images (time of flight [TOF], T1, proton density [PD], and T2) of the carotid arteries centered at the common carotid bifurcation on the index side.8 MRI parameters were as follows: T1-weighted (T1W): double inversion recovery, black blood, 2D fast spin-echo, repetition time (TR)/effective echo time (TE)=800/10 ms, echo train length=8; PD-weighted (PDW) and T2-weighted (T2W): fast spin echo, cardiac gated, TR=3 or 4 cardiac R-R intervals, effective TE=20 ms for PDW and 40 ms for T2W, echo train length=6; and 3D TOF: TR/TE 23/3.8 ms, flip angle 25°. Fat suppression was used for T1W, PDW, and T2W images to reduce signals from subcutaneous fat. Images were obtained with a field of view of 13 to 16 cm, matrix size of 256, slice thickness of 2 mm, and 2 signal averages. Interslice spacing was 0 mm for T1W, PDW, and T2W and –1 mm for 3D TOF (1 mm overlapping between adjacent slices). The longitudinal coverage of each carotid artery was 24 mm (12 slices) for T1W, 30 mm (15 slices) for PDW and T2W, and 40 mm (40 slices) for 3D TOF. A zero-filled Fourier transform was used to reduce pixel size (0.25x0.25 to 0.31x0.31 mm2) and minimize partial-volume artifacts.
MRI Review and Criteria
Image quality was rated per artery for each contrast weighting on a 5-point scale (1=poor, 5=excellent) dependent on the overall signal-to-noise ratio and clarity of the vessel wall boundaries.17 Images with an image quality 2 were excluded from the study. After identifying cases and controls, 1 reviewer (N.T.), who was blinded to time points, examined all of the MR images.
Area measurements of the lumen, outer wall, and tissue components (lipid-rich necrotic core, calcium, and hemorrhage) were obtained with a custom-designed imaging analysis tool (QVAS).13 Wall area was calculated as the difference between outer wall and lumen area. The carotid bifurcation was used as the landmark for matching the 4 different contrast weightings of the 2 image sets at 2 time points. Only matched locations covered by both time points were used for analysis.
Areas of hemorrhage, lipid-rich necrotic core, and calcification were identified as described previously,10 with modification in terminology, in which the terms type I and type II rather than fresh and recent were used to describe the age of intraplaque hemorrhage. Briefly, fresh intraplaque hemorrhage (type I) appears as a hyperintense signal on T1W and TOF images and as an isointense signal on PD/T2W images compared with the signal of the adjacent sternocleidomastoid muscle. Recent hemorrhage (type II) is identified by a hyperintense signal on all 4 contrast weightings. The lipid-rich necrotic core without hemorrhage is identified by a hyperintense to isointense signal on T1W and PDW images, a varied signal on T2W images, and an isointense signal on TOF. Calcium is characterized by well-defined hypointense signal areas in all 4 contrast weightings. Figure 1 shows representative images of a lipid-rich necrotic core without hemorrhage, a necrotic core containing type I hemorrhage, and a necrotic core containing type II hemorrhage.
Histology Processing
After the second MRI scan, 2 patients underwent CEA. The specimens were fixed in formalin, decalcified, and embedded in paraffin. Samples were sectioned (10 μm thick) every 0.5 to 1.0 mm throughout the length of the specimen and stained (hematoxylin-eosin, Mallory’s trichrome). The slides were independently evaluated by a reviewer (M.S.F) who was unaware of the imaging results. Immunohistochemical staining of glycophorin A (1:500 dilution; Daco) and Mallory’s iron stain were performed.4,18 The age of hemorrhage was classified as previously described.10
Statistical Analysis
Data were expressed as mean±SD. Percent changes in measurements over time were calculated for individual subjects. The Student paired t test was used for comparison of means within the same group, and the Student unpaired t test was used for comparison between 2 groups. Linear regression was used to compare percent change between groups while controlling for potentially confounding variables. Categorical variables were presented as counts and percentages per category and compared by means of Fisher exact test. All calculations were made with SPSS 12.0 for Windows. Statistical significance was defined as a value of P<0.05. Volumes were measured including all slices available with American Heart Association grade 3. Volumes were then normalized for varying numbers of slices by adjustment to yield a volume for 10 slices: adjusted volume=observed volumex(10 slices/observed number of slices).
Results
Patient Characteristics
The mean observational periods of both groups were 17.8 months (range, 16.6 to 19.6 months) and 18.0 months (range, 16.4 to 19.5 months) for the hemorrhage and control groups, respectively. There was no significant difference in the observation period between the 2 groups (P=0.5).
Two patients in the hemorrhage group underwent CEA during the course of the study. One of the CEA patients developed transient hemispheric ischemia in the distribution of the index carotid artery 3 weeks before surgery, and the other patient had asymptomatic progression of carotid stenosis to >80%. In the control group, 1 patient described an episode of amaurosis fugax on the side of the index carotid artery during the observational period. However, CEA was not performed because of prohibitive cardiac risk to surgery.
Percent Volume Change
The percent change in wall volume (6.8±7.9% versus –0.15±5.1%; P=0.009) and lipid-rich necrotic core volume (28.4±29.7% versus –5.2±17.3%; P=0.001) was significantly higher in the hemorrhage group than in the control group (Table 3). Lumen volume was reduced in the hemorrhage group compared with the control group (–8.5±12.2% versus 1.5±7.9%; P=0.001). Maximum and mean wall thickness increased significantly in the hemorrhage group versus the control group (8.3±11.3% versus –3.2±10.7%, P=0.009; 8.5±8.8% versus –1.6±4.7%, P=0.001). Figure 2 is an example of intraplaque hemorrhage, lumen narrowing, and wall increase over the 18-month follow-up period.
In the hemorrhage group, wall volume and lipid-rich necrotic core volume increased despite statin therapy. However, the percent change in volume was smaller in the subgroup treated with statins compared with those not receiving statin therapy (5.9±6.3% versus 8.4±10.9% for wall volume and 25.4±14.7% versus 33.9±48.7% for lipid-rich necrotic core volume, respectively), but these differences were not significant (P=0.6).
Hemorrhage Signal Intensity Change
Baseline examination of the 29 patients produced 80 MR slices (83 locations) of hemorrhage. Type I hemorrhage was present in 6 slices (7.5%), and type II hemorrhage was observed in 71 slices (88.8%). Three slices (3.8%) contained both type I and type II hemorrhages in the same slice level. Interestingly, at the 18-month examination, 78 locations (94.0%) demonstrated signal intensity, indicating that the hemorrhage stage remained unchanged. Eight locations (88.9%) of type I hemorrhage and 70 locations (94.5%) of type II hemorrhage remained constant. Figure 3 shows a representative case of type II hemorrhage that retained the same hyperintense signal in all 4 weightings over a period of 18 months. CEA was performed in this case 1 year after the second MRI examination, and the specimen was processed and serially sectioned. Histology examination demonstrated recent hemorrhage (type II) in the same regions as in the matched MR images (Figure 3C). Those areas also had abundant cholesterol crystals and showed extensive staining for glycophorin A within the necrotic core (Figure 3D, 3E).
New Hemorrhage
Six of the 14 patients (43%) with intraplaque hemorrhage at baseline developed intraplaque hemorrhage in new locations within the plaque during the 18-month study period, whereas no new incidents of hemorrhage were detected in the control group (P=0.006, Fisher exact test) (Table 4). In location-based analysis, new hemorrhage was observed in 12 sites (12.6%). Further evidence of repeated intraplaque hemorrhage is provided by the increase in volume of intraplaque hemorrhage over the 18-month period, from 146.6±118.9 to 181.5±161.8 mm3, but the increase was not statistically significant (P=0.06).
Arteries that exhibited evidence of repeated intraplaque hemorrhage were also associated with significant increase in percent change of mean wall thickness (14.8% versus 3.7%; P=0.013) as well as a significant decrease in percent change of lumen volume (–16.4% versus –2.5%; P=0.013) compared with plaques without evidence of repeated intraplaque hemorrhage. The change in wall and lipid-rich necrotic core volume absent the volume of new hemorrhage was still significantly greater than change in the control group (5.7% versus –0.2%, P=0.021; 24.0% versus –5.2%, P=0.002). In addition, a comparison of intraplaque hemorrhage patients without new hemorrhage during follow-up (n=8) with controls (n=15) also showed that the percent changes of wall and lipid-rich necrotic core volumes were greater than the changes of the control group (5.3% versus –0.2%, P=0.049; 19.0% versus –5.2%, P=0.03).
Histological findings from the excised carotid specimens provided further evidence of repeated intraplaque hemorrhage. Figure 4 shows extensive microvasculature with surrounding intraplaque hemorrhage of an earlier phase than that deep within the core. Mallory’s stain for iron was used on sections from each level of the 2 excised specimens to provide confirmation of red cell breakdown and erythrophagocytosis. Iron, in the form of inclusions, was predominantly found in macrophages in tissues peripheral to the necrotic core.
Discussion
To our knowledge, this study is one of the first to demonstrate an association between baseline plaque compositional characteristics, specifically intraplaque hemorrhage, and progression of human atherosclerosis in a prospective, serial fashion with the use of high-resolution MRI. In this study carotid plaques with evidence of early or recent intraplaque hemorrhage at baseline were significantly more likely to increase in overall wall volume and lipid-rich necrotic core volume and more likely to decrease in lumen size over an 18-month period. Furthermore, patients with intraplaque hemorrhage at baseline had a greater probability of repeated carotid intraplaque hemorrhage.
Findings from this study are consistent with the suggestion by Kolodgie et al4 that intraplaque hemorrhage may stimulate atherogenic activity by being a source of free cholesterol and macrophage activation. Other studies based on histopathological examination of excised plaques also suggest that intraplaque hemorrhage may be a powerful instigator of progression.19–21 Kockx et al20 reported that intraplaque microhemorrhages initiated phagocytosis of platelets and erythrocytes, leading to activation of inducible NO synthase expressing foam cells. They also found that clusters of macrophages located near microvessels appeared to have ingested red cell fragments. Yuan et al21 noted that erythrophagocytosis enhances the capacity of macrophages to oxidize LDL, leading to conversion of LDL to a form that promotes lipid accumulation in macrophages. This longitudinal, in vivo study corroborates these in vitro findings by showing not only an increase in the total volume of the atherosclerotic vessel wall (6.8±7.9%) but also an increase in the lipid-rich necrotic core (28.4±29.7%) among plaques with intraplaque hemorrhage at baseline. In addition, iron staining of the endarterectomy specimens in this study demonstrated that the iron deposits within macrophages were frequently observed in tissues peripheral to the necrotic core, suggesting that erythrophagocytosis might contribute to the formation of foam cells, as previously described.4,20
In contrast, the control group did not exhibit an increase in plaque volume (–0.15±5.1%) and showed a decrease in the lipid-rich necrotic core volume (–5.2+17.3%). These results suggest that plaques without hemorrhage are less active or perhaps that statin treatment (73.3% of control subjects) truly did abolish progression of arteriosclerosis, as previously described in studies by Corti et al22 and Nissen et al.23 However, because plaques in the hemorrhage group progressed rapidly despite treatment with statins (64.3% of subjects), this may indicate that the physical increase in plaque volume and cholesterol supply by repeated intraplaque hemorrhage may exceed the effect of statin therapy. On the other hand, statin-treated plaques in the hemorrhage group had a smaller percent change of wall (5.9±6.3% versus 8.4±10.9%; P=0.6) and lipid-rich necrotic core (25.4±14.7% versus 33.9±48.7%; P=0.6) volume, which suggests that statin therapy may suppress the progression of plaques even in the presence of hemorrhage. However, neither change in these volumes was statistically significant. Further study with a larger group of patients with intraplaque hemorrhage is needed for this aspect of the study.
Findings from this study also suggest the importance of repeated hemorrhage into the plaque as an additional mechanism of progression. New hemorrhages occurred in at least 43% of patients with previous intraplaque hemorrhage, whereas no new hemorrhages appeared in the control group. Plaques with MRI evidence of repeated intraplaque hemorrhage were associated with a significant increase in mean wall thickness (14.8% versus 3.7%; P=0.013) and decrease in lumen volume (–16.4% versus –2.5%; P=0.013) compared with plaques without evidence of repeated intraplaque hemorrhage. This difference suggests that repeated intraplaque hemorrhage may be a marker for plaque instability and may be a contributing factor toward a more rapid progression of atherosclerosis. The disruption of thin-walled microvessels with discontinuous endothelium is believed to be a cause of intraplaque hemorrhage. These fragile microvessels are implicated in repeated intraplaque capillary rupture.3,20 Indeed, close histological examination of the necrotic core of the excised CEA specimens showed extensive microvasculature with surrounding intraplaque hemorrhage of an earlier phase than that deep within the core (Figure 4). On the other hand, the change in wall and lipid-rich necrotic core volume absent the volume of new hemorrhage was still significantly greater than the change of the control group. Furthermore, intraplaque hemorrhage patients without new hemorrhage during follow-up also had greater change in wall and lipid-rich necrotic core volumes than the controls. These data suggest that increase in wall and lipid-rich necrotic core volume may result not only from the influx of new hemorrhage but also from other mechanisms, possibly from increase in plaque cholesterol and activation of macrophages, as suggested by Kolodgie et al.4
Our previously published work established the criteria for the identification of the stages of intraplaque hemorrhage with the use of multicontrast MRI.10,12 The concept of the change in signal intensity of intraplaque hemorrhage was based on the detection and staging of cerebral hemorrhage with the use of multicontrast MR images. The stages of hemorrhage are based on red blood cell content and the oxygenation of hemoglobin. The criteria used to classify early subacute cerebral hemorrhage were applied to the carotid intraplaque hemorrhage and given the term fresh. Morphologically, this includes the presence of intact red blood cells, macrophage and inflammatory cell infiltrate, and evidence of early thrombus formation, which by MRI appears hyperintense on TOF and T1W images. The second category, termed recent, used the late subacute criteria from brain hemorrhage, which includes hemorrhagic debris, intact and degenerating red blood cells, and the increased interstitial fluid associated with inflammation and early organization. These latter components combine to create hyperintense signal patterns in all contrast weightings. Our use of the terms type I and type II rather than fresh and recent, respectively, is based on an evolving understanding of the distinct time course of intraplaque hemorrhage in the carotid artery. This study demonstrates that in the majority (94%) of locations with intraplaque hemorrhage at baseline, the signal intensity associated with hemorrhage remained unchanged at the 18-month time point. MRI studies of intracranial hemorrhage demonstrate a rapid evolution of signal intensity changes within days after the hemorrhagic event that are related to the oxidative state of hemoglobin and the eventual lysis of red blood cells.24 Erythrocyte lysis and hemoglobin oxygenation are heavily dependent on local factors such as oxygen tension and pH.25,26 Our results suggest that atherosclerotic intraplaque hemorrhage may have a longer course for change due to inherent differences in the tissue environment within the necrotic core. Intraplaque hemorrhage may attract macrophages that eventually become trapped within the core and are unable to survive.4 Our histological assessment showed that macrophages with iron inclusions were found in fibrous areas around the core. The infrequency of macrophage iron-laden cells in the core indicates that the degradation of hemorrhage may be slow because of reduced activation of macrophages. Alternatively, repeated hemorrhage into the same plaque region may account for the absence of expected signal intensity changes over the 18-month interval between scans. Because our study included a relatively small number of type I hemorrhage patients at baseline, further analysis using more patients will be needed to solidify our understanding of the signal intensity change of type I hemorrhage. Additionally, and to establish the time course of the signal intensity changes of intraplaque hemorrhage, a more frequent schedule of MRI examinations after the intraplaque hemorrhage event is needed.
Finally, findings from this study also provide insight into the remodeling of the carotid atherosclerosis. Lumen stenosis progressed more rapidly in the hemorrhage group than in the control group. Percent change of lumen volume was –8.5±12.2% for the hemorrhage group and 1.5±7.9% for control subjects (P=0.014). However, in comparison, the outer wall volume increased by only a small amount among cases compared with controls (1.5±6.2% versus 0.2±46%, respectively; P=NS), suggesting that positive remodeling played little role in the increase in wall volume associated with intraplaque hemorrhage. Both of these trends are consistent with previous histopathological studies indicating that intraplaque hemorrhage has a significant correlation primarily with increased luminal narrowing.27
As demonstrated in this study, the clinical implications of serial MR examinations are significant. MRI can give information not only about the change in lumen size and plaque volume but also about plaque components,11 inflammation,28 and fibrous cap status17 noninvasively. Findings from these studies may provide new insight into the pathophysiology of atherosclerosis and permit direct assessment of the effect of pharmacological therapy.
Limitations of Study
At baseline, the lipid-rich necrotic core volume was larger in the hemorrhage group than in the controls, which was anticipated, given that intraplaque hemorrhage is typically admixed within the necrotic core of the plaque. Of note, however, when we controlled for baseline volume of lipid-rich necrotic core and baseline wall volume, the association of intraplaque hemorrhage with increased plaque volume was still statistically significant (P=0.009). It is important to remember that however intriguing these results may be, this study focused only on the carotid artery, and therefore extrapolation to other vascular beds is not possible at this time.
Also of note, 56 (19.7%) of the 284 arteries available for analysis were excluded because of marginal image quality. New technological developments in hardware and image acquisition techniques should improve overall image quality and reduce the number of excluded cases.
Conclusions
This in vivo MRI progression study demonstrates that hemorrhage into the carotid atherosclerotic plaque is significantly associated with progression in wall and lipid-rich necrotic core size and a decrease in lumen volume. Furthermore, lesions that had intraplaque hemorrhage at baseline had a greater probability of repeated intraplaque hemorrhage at 18 months. We propose that repeated hemorrhage and the resulting consequences of cholesterol released from trapped erythrocytes and activated macrophages may be important factors in the rapid progression of carotid atherosclerosis.
Acknowledgments
This study was supported by National Institutes of Health grants R01 HL61851, R01 HL073401, and HL67406.
References
Libby P. Inflammation in atherosclerosis. Nature. 2002; 420: 868–874.
de Boer OJ, van der Wal AC, Teeling P, Becker AE. Leucocyte recruitment in rupture prone regions of lipid-rich plaques: a prominent role for neovascularization; Cardiovasc Res. 1999; 41: 443–449.
Virmani R, Narula J, Farb A. When neoangiogenesis ricochets. Am Heart J. 1998; 136: 937–939.
Kolodgie FD, Gold HK, Burke AP, Fowler DR, Kruth HS, Weber DK, Farb A, Guerrero LJ, Hayase M, Kutys R, Narula J, Finn AV, Virmani R. Intraplaque hemorrhage and progression of coronary atheroma. N Engl J Med. 2003; 349: 2316–2325.
Yeagle PL. Cholesterol and the cell membrane. Biochim Biophys Acta. 1985; 822: 267–287.
Toussaint JF, LaMuraglia GM, Southern JF, Fuster V, Kantor HL. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo. Circulation. 1996; 94: 932–938.
Yuan C, Mitsumori LM, Ferguson MS, Polissar NL, Echelard D, Ortiz G, Small R, Davies JW, Kerwin WS, Hatsukami TS. In vivo accuracy of multispectral magnetic resonance imaging for identifying lipid-rich necrotic cores and intraplaque hemorrhage in advanced human carotid plaques. Circulation. 2001; 104: 2051–2056.
Cai JM, Hatsukami TS, Ferguson MS, Small R, Polissar NL, Yuan C. Classification of human carotid atherosclerotic lesions with in vivo multicontrast magnetic resonance imaging. Circulation. 2002; 106: 1368–1373.
Fayad ZA. MR imaging for the noninvasive assessment of atherothrombotic plaques. Magn Reson Imaging Clin N Am. 2003; 11: 101–113.
Chu B, Kampschulte A, Ferguson MS, Kerwin WS, Yarnykh VL, O’Brien KD, Polissar NL, Hatsukami TS, Yuan C. Hemorrhage in the atherosclerotic carotid plaque: a high-resolution MRI study. Stroke. 2004; 35: 1079–1084.
Saam T, Ferguson MS, Yarnykh VL, Takaya N, Xu D, Polissar NL, Hatsukami TS, Yuan C. Quantitative evaluation of carotid plaque composition by in vivo MRI. Arterioscler Thromb Vasc Biol. 2005; 25: 234–239.
Zhao XQ, Yuan C, Hatsukami TS, Frechette EH, Kang XJ, Maravilla KR, Brown BG. Effects of prolonged intensive lipid-lowering therapy on the characteristics of carotid atherosclerotic plaques in vivo by MRI: a case-control study. Arterioscler Thromb Vasc Biol. 2001; 21: 1623–1629.
Zhang S, Cai J, Luo Y, Han C, Polissar NL, Hatsukami TS, Yuan C. Measurement of carotid wall volume and maximum area with contrast-enhanced 3D MR imaging: initial observations. Radiology. 2003; 228: 200–205.
Thackray BD, Burns DH, Ferguson MS, Gordon D, Beach KW, Hatsukami T, Detmer PR, Primozich JF, Strandness DE Jr. A new method for studying plaque morphology. Am J Card Imaging. 1995; 9: 149–156.
Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantitation of high-density-lipoprotein cholesterol. Clin Chem. 1982; 28: 1379–1388.
Warnick GR. Enzymatic methods for quantification of lipoprotein lipids. Methods Enzymol. 1986; 129: 101–123.
Hatsukami TS, Ross R, Polissar NL, Yuan C. Visualization of fibrous cap thickness and rupture in human atherosclerotic carotid plaque in vivo with high-resolution magnetic resonance imaging. Circulation. 2000; 102: 959–964.
Sadahira Y, Kanzaki A, Wada H, Yawata Y. Immunohistochemical identification of erythroid precursors in paraffin embedded bone marrow sections: spectrin is a superior marker to glycophorin. J Clin Pathol. 1999; 52: 919–921.
Arbustini E, Morbini P, D’Armini AM, Repetto A, Minzioni G, Piovella F, Vigano M, Tavazzi L. Plaque composition in plexogenic and thromboembolic pulmonary hypertension: the critical role of thrombotic material in pultaceous core formation. Heart. 2002; 88: 177–182.
Kockx MM, Cromheeke KM, Knaapen MW, Bosmans JM, De Meyer GR, Herman AG, Bult H. Phagocytosis and macrophage activation associated with hemorrhagic microvessels in human atherosclerosis. Arterioscler Thromb Vasc Biol. 2003; 23: 440–446.
Yuan XM, Anders WL, Olsson AG, Brunk UT. Iron in human atheroma and LDL oxidation by macrophages following erythrophagocytosis. Atherosclerosis. 1996; 124: 61–73.
Corti R, Osende JI, Fallon JT, Fuster V, Mizsei G, Jneid H, Wright SD, Chaplin WF, Badimon JJ. The selective peroxisomal proliferator-activated receptor-gamma agonist has an additive effect on plaque regression in combination with simvastatin in experimental atherosclerosis: in vivo study by high-resolution magnetic resonance imaging. J Am Coll Cardiol. 2004; 43: 464–473.
Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis: a randomized controlled trial. JAMA. 2004; 291: 1071–1080.
Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology. 1993; 189: 15–26.
Brajtburg J, Elberg S, Schwartz DR, Vertut-Croquin A, Schlessinger D, Kobayashi GS, Medoff G. Involvement of oxidative damage in erythrocyte lysis induced by amphotericin B. Antimicrob Agents Chemother. 1985; 27: 172–176.
Yang E, Huestis WH. Oxidative interactions between the erythrocyte membrane and phosphatidylcholine vesicles. J Biol Chem. 1994; 269: 14518–14524.
Imparato AM, Riles TS, Mintzer R, Baumann FG. The importance of hemorrhage in the relationship between gross morphologic characteristics and cerebral symptoms in 376 carotid artery plaques. Ann Surg. 1983; 197: 195–203.
Kerwin W, Hooker A, Spilker M, Vicini P, Ferguson M, Hatsukami T, Yuan C. Quantitative magnetic resonance imaging analysis of neovasculature volume in carotid atherosclerotic plaque. Circulation. 2003; 107: 851–856.(Presence of Intraplaque H)
Department of Cardiology, Juntendo University School of Medicine, Tokyo, Japan (N.T.)
Mountain-Whisper-Light Statistical Consulting, Seattle, Wash (N.L.P.)
Department of Surgery, University of Washington, and Veterans Affairs Puget Sound Health Care System, Seattle, Wash (C.I., J.M., C.N., T.S.H.).
Abstract
Background— Previous studies suggest that erythrocyte membranes from intraplaque hemorrhage into the necrotic core are a source of free cholesterol and may become a driving force in the progression of atherosclerosis. We have shown that MRI can accurately identify carotid intraplaque hemorrhage and precisely measure plaque volume. We tested the hypothesis that hemorrhage into carotid atheroma stimulates plaque progression.
Methods and Results— Twenty-nine subjects (14 cases with intraplaque hemorrhage and 15 controls with comparably sized plaques without intraplaque hemorrhage at baseline) underwent serial carotid MRI examination with a multicontrast weighted protocol (T1, T2, proton density, and 3D time of flight) over a period of 18 months. The volumes of wall, lumen, lipid-rich necrotic core, calcification, and intraplaque hemorrhage were measured with a custom-designed image analysis tool. The percent change in wall volume (6.8% versus –0.15%; P=0.009) and lipid-rich necrotic core volume (28.4% versus –5.2%; P=0.001) was significantly higher in the hemorrhage group than in controls over the course of the study. Furthermore, those with intraplaque hemorrhage at baseline were much more likely to have new plaque hemorrhages at 18 months compared with controls (43% versus 0%; P=0.006).
Conclusions— Hemorrhage into the carotid atherosclerotic plaque accelerated plaque progression in an 18-month period. Repeated bleeding into the plaque may produce a stimulus for the progression of atherosclerosis by increasing lipid core and plaque volume and creating new destabilizing factors.
Key Words: magnetic resonance imaging ; carotid arteries ; hemorrhage ; atherosclerosis ; plaque
Introduction
The progression of atherosclerosis is a complex and often longstanding phenomenon. Many studies have shown that inflammation and the subsequent immune response contribute to atherogenesis. The recruitment of inflammatory cells in atherosclerotic lesions is a constitutive phenomenon throughout the process of plaque growth.1,2 Intraplaque hemorrhage is commonly observed in atherosclerotic plaques and is considered by some to be caused by rupture of plaque neovasculature. These microvessels are fragile because of lack of support by smooth muscle cells and focal discontinuity of the endothelial lining.3
In a recent histopathological study of coronary artery specimens, Kolodgie et al4 suggested that intraplaque hemorrhage may represent a potent atherogenic stimulus by contributing to the deposition of free cholesterol, macrophage infiltration, and enlargement of the necrotic core. Immunohistochemical staining with the antibody to glycophorin A, a protein specific to erythrocyte membranes, was strongly associated with the size of the necrotic core and degree of macrophage infiltration. They also noted that rabbit lesions with induced intramural hemorrhage had significantly greater lipid content than control lesions without hemorrhage. Other investigators have noted that erythrocyte membranes contain more free cholesterol than any other cell in the body.5
To test the hypothesis that intraplaque hemorrhage is a predisposing factor in progression of human atherosclerosis, an imaging tool is needed that can serially examine the morphological and compositional characteristics of the lesion over time. High-resolution MRI is ideally suited for this purpose because it is noninvasive and has proven capability for distinguishing tissue characteristics.6–9 Our laboratory recently demonstrated that multicontrast MRI can accurately detect the presence and age of carotid intraplaque hemorrhage.10 Furthermore, MRI is capable of precisely quantifying overall lesion size and plaque compositional features such as the lipid-rich necrotic core.11–13 Thus, this imaging tool provides a unique opportunity to examine the temporal relationship between baseline plaque characteristics and subsequent progression in lesion size and composition.
The aim of this study was to test the hypothesis that intraplaque hemorrhage, as detected by high-resolution MRI, is associated with greater progression in both lipid-rich necrotic core and plaque volume over a period of 18 months.
Methods
Study Population
Cases and controls were selected from an ongoing prospective serial carotid MRI study referred to as PRIMARI. Subjects for PRIMARI were recruited from the diagnostic vascular ultrasound laboratories at the University of Washington Medical Center and Veterans Affairs Puget Sound Health Care System after informed consent was obtained. The study procedures and consent forms were reviewed and approved by each site’s institutional review board. Inclusion criteria for the PRIMARI study included (1) 50% to 79% carotid stenosis by duplex ultrasound examination and (2) asymptomatic status with regard to carotid artery disease within the 6 months before enrollment in PRIMARI. Exclusion criteria included (1) prior carotid endarterectomy (CEA) on the side of the index carotid artery, (2) prior neck irradiation, and (3) contraindication for MRI. Subjects undergo bilateral carotid artery MRI every 18 months. At the time of analysis, 142 subjects had at least 2 MRI scans (baseline and 18 months). Among the 142 patients (284 arteries) available, 66 arteries were excluded because of marginal image quality, insufficient plaque size, or absence of the common carotid bifurcation in the field of view. In this serial MRI study, the common carotid artery on the index side was used as an internal fiducial marker to match cross-sectional locations between MRI examinations. In some cases, the bifurcation on the contralateral, nonindex side occurred at a different axial level, outside the field of view. Four arteries with total occlusion were excluded, as were 11 arteries with MRI evidence of fibrous cap rupture or ulceration with mural thrombus formation. The latter were excluded to focus the analysis of this study on the association between intraplaque hemorrhage (presumably from rupture of plaque neovasculature) and plaque progression.
Images from the baseline MRI examination were first reviewed to identify cases and controls blinded to the follow-up MR studies. Among the 203 arteries remaining, 14 arteries (99 cross-sectional locations) demonstrated clear MRI evidence of early or recent intraplaque hemorrhage. Fifteen arteries (103 cross-sectional locations) with comparable baseline plaque volume that clearly had no evidence of early, recent, or chronic intraplaque hemorrhage were selected as controls. The baseline and the 18-month follow-up MRI examinations were used to compare cases and controls for progression in overall plaque volume and plaque constituents. Two patients underwent CEA at 2 months and 1 year after the second MRI examination, and detailed histological analysis was performed for comparison with the MRI findings with the use of a previously published protocol.14
Baseline Clinical and Laboratory Data
Before the baseline MRI examination, study subjects were asked to complete a detailed health questionnaire and physical examination to assess risk factors for atherosclerosis (hypertension, diabetes, smoking, family history), history of coronary artery and peripheral vascular disease, current medications, and body mass index. Lipid measurements were performed on fasting whole plasma. Standard enzymatic methods were used to determine levels of total cholesterol, triglycerides, and HDL cholesterol on an Abbott Spectrum analyzer.15,16 LDL cholesterol was calculated by the Friedewald equation.
MRI Protocol
Patients were imaged with a custom-designed phased-array surface coil in a 1.5-T GE Signa Scanner (Horizon EchoSpeed, version 5.8, GE Medical Systems). A standardized protocol was used to obtain 4 different contrast-weighted images (time of flight [TOF], T1, proton density [PD], and T2) of the carotid arteries centered at the common carotid bifurcation on the index side.8 MRI parameters were as follows: T1-weighted (T1W): double inversion recovery, black blood, 2D fast spin-echo, repetition time (TR)/effective echo time (TE)=800/10 ms, echo train length=8; PD-weighted (PDW) and T2-weighted (T2W): fast spin echo, cardiac gated, TR=3 or 4 cardiac R-R intervals, effective TE=20 ms for PDW and 40 ms for T2W, echo train length=6; and 3D TOF: TR/TE 23/3.8 ms, flip angle 25°. Fat suppression was used for T1W, PDW, and T2W images to reduce signals from subcutaneous fat. Images were obtained with a field of view of 13 to 16 cm, matrix size of 256, slice thickness of 2 mm, and 2 signal averages. Interslice spacing was 0 mm for T1W, PDW, and T2W and –1 mm for 3D TOF (1 mm overlapping between adjacent slices). The longitudinal coverage of each carotid artery was 24 mm (12 slices) for T1W, 30 mm (15 slices) for PDW and T2W, and 40 mm (40 slices) for 3D TOF. A zero-filled Fourier transform was used to reduce pixel size (0.25x0.25 to 0.31x0.31 mm2) and minimize partial-volume artifacts.
MRI Review and Criteria
Image quality was rated per artery for each contrast weighting on a 5-point scale (1=poor, 5=excellent) dependent on the overall signal-to-noise ratio and clarity of the vessel wall boundaries.17 Images with an image quality 2 were excluded from the study. After identifying cases and controls, 1 reviewer (N.T.), who was blinded to time points, examined all of the MR images.
Area measurements of the lumen, outer wall, and tissue components (lipid-rich necrotic core, calcium, and hemorrhage) were obtained with a custom-designed imaging analysis tool (QVAS).13 Wall area was calculated as the difference between outer wall and lumen area. The carotid bifurcation was used as the landmark for matching the 4 different contrast weightings of the 2 image sets at 2 time points. Only matched locations covered by both time points were used for analysis.
Areas of hemorrhage, lipid-rich necrotic core, and calcification were identified as described previously,10 with modification in terminology, in which the terms type I and type II rather than fresh and recent were used to describe the age of intraplaque hemorrhage. Briefly, fresh intraplaque hemorrhage (type I) appears as a hyperintense signal on T1W and TOF images and as an isointense signal on PD/T2W images compared with the signal of the adjacent sternocleidomastoid muscle. Recent hemorrhage (type II) is identified by a hyperintense signal on all 4 contrast weightings. The lipid-rich necrotic core without hemorrhage is identified by a hyperintense to isointense signal on T1W and PDW images, a varied signal on T2W images, and an isointense signal on TOF. Calcium is characterized by well-defined hypointense signal areas in all 4 contrast weightings. Figure 1 shows representative images of a lipid-rich necrotic core without hemorrhage, a necrotic core containing type I hemorrhage, and a necrotic core containing type II hemorrhage.
Histology Processing
After the second MRI scan, 2 patients underwent CEA. The specimens were fixed in formalin, decalcified, and embedded in paraffin. Samples were sectioned (10 μm thick) every 0.5 to 1.0 mm throughout the length of the specimen and stained (hematoxylin-eosin, Mallory’s trichrome). The slides were independently evaluated by a reviewer (M.S.F) who was unaware of the imaging results. Immunohistochemical staining of glycophorin A (1:500 dilution; Daco) and Mallory’s iron stain were performed.4,18 The age of hemorrhage was classified as previously described.10
Statistical Analysis
Data were expressed as mean±SD. Percent changes in measurements over time were calculated for individual subjects. The Student paired t test was used for comparison of means within the same group, and the Student unpaired t test was used for comparison between 2 groups. Linear regression was used to compare percent change between groups while controlling for potentially confounding variables. Categorical variables were presented as counts and percentages per category and compared by means of Fisher exact test. All calculations were made with SPSS 12.0 for Windows. Statistical significance was defined as a value of P<0.05. Volumes were measured including all slices available with American Heart Association grade 3. Volumes were then normalized for varying numbers of slices by adjustment to yield a volume for 10 slices: adjusted volume=observed volumex(10 slices/observed number of slices).
Results
Patient Characteristics
The mean observational periods of both groups were 17.8 months (range, 16.6 to 19.6 months) and 18.0 months (range, 16.4 to 19.5 months) for the hemorrhage and control groups, respectively. There was no significant difference in the observation period between the 2 groups (P=0.5).
Two patients in the hemorrhage group underwent CEA during the course of the study. One of the CEA patients developed transient hemispheric ischemia in the distribution of the index carotid artery 3 weeks before surgery, and the other patient had asymptomatic progression of carotid stenosis to >80%. In the control group, 1 patient described an episode of amaurosis fugax on the side of the index carotid artery during the observational period. However, CEA was not performed because of prohibitive cardiac risk to surgery.
Percent Volume Change
The percent change in wall volume (6.8±7.9% versus –0.15±5.1%; P=0.009) and lipid-rich necrotic core volume (28.4±29.7% versus –5.2±17.3%; P=0.001) was significantly higher in the hemorrhage group than in the control group (Table 3). Lumen volume was reduced in the hemorrhage group compared with the control group (–8.5±12.2% versus 1.5±7.9%; P=0.001). Maximum and mean wall thickness increased significantly in the hemorrhage group versus the control group (8.3±11.3% versus –3.2±10.7%, P=0.009; 8.5±8.8% versus –1.6±4.7%, P=0.001). Figure 2 is an example of intraplaque hemorrhage, lumen narrowing, and wall increase over the 18-month follow-up period.
In the hemorrhage group, wall volume and lipid-rich necrotic core volume increased despite statin therapy. However, the percent change in volume was smaller in the subgroup treated with statins compared with those not receiving statin therapy (5.9±6.3% versus 8.4±10.9% for wall volume and 25.4±14.7% versus 33.9±48.7% for lipid-rich necrotic core volume, respectively), but these differences were not significant (P=0.6).
Hemorrhage Signal Intensity Change
Baseline examination of the 29 patients produced 80 MR slices (83 locations) of hemorrhage. Type I hemorrhage was present in 6 slices (7.5%), and type II hemorrhage was observed in 71 slices (88.8%). Three slices (3.8%) contained both type I and type II hemorrhages in the same slice level. Interestingly, at the 18-month examination, 78 locations (94.0%) demonstrated signal intensity, indicating that the hemorrhage stage remained unchanged. Eight locations (88.9%) of type I hemorrhage and 70 locations (94.5%) of type II hemorrhage remained constant. Figure 3 shows a representative case of type II hemorrhage that retained the same hyperintense signal in all 4 weightings over a period of 18 months. CEA was performed in this case 1 year after the second MRI examination, and the specimen was processed and serially sectioned. Histology examination demonstrated recent hemorrhage (type II) in the same regions as in the matched MR images (Figure 3C). Those areas also had abundant cholesterol crystals and showed extensive staining for glycophorin A within the necrotic core (Figure 3D, 3E).
New Hemorrhage
Six of the 14 patients (43%) with intraplaque hemorrhage at baseline developed intraplaque hemorrhage in new locations within the plaque during the 18-month study period, whereas no new incidents of hemorrhage were detected in the control group (P=0.006, Fisher exact test) (Table 4). In location-based analysis, new hemorrhage was observed in 12 sites (12.6%). Further evidence of repeated intraplaque hemorrhage is provided by the increase in volume of intraplaque hemorrhage over the 18-month period, from 146.6±118.9 to 181.5±161.8 mm3, but the increase was not statistically significant (P=0.06).
Arteries that exhibited evidence of repeated intraplaque hemorrhage were also associated with significant increase in percent change of mean wall thickness (14.8% versus 3.7%; P=0.013) as well as a significant decrease in percent change of lumen volume (–16.4% versus –2.5%; P=0.013) compared with plaques without evidence of repeated intraplaque hemorrhage. The change in wall and lipid-rich necrotic core volume absent the volume of new hemorrhage was still significantly greater than change in the control group (5.7% versus –0.2%, P=0.021; 24.0% versus –5.2%, P=0.002). In addition, a comparison of intraplaque hemorrhage patients without new hemorrhage during follow-up (n=8) with controls (n=15) also showed that the percent changes of wall and lipid-rich necrotic core volumes were greater than the changes of the control group (5.3% versus –0.2%, P=0.049; 19.0% versus –5.2%, P=0.03).
Histological findings from the excised carotid specimens provided further evidence of repeated intraplaque hemorrhage. Figure 4 shows extensive microvasculature with surrounding intraplaque hemorrhage of an earlier phase than that deep within the core. Mallory’s stain for iron was used on sections from each level of the 2 excised specimens to provide confirmation of red cell breakdown and erythrophagocytosis. Iron, in the form of inclusions, was predominantly found in macrophages in tissues peripheral to the necrotic core.
Discussion
To our knowledge, this study is one of the first to demonstrate an association between baseline plaque compositional characteristics, specifically intraplaque hemorrhage, and progression of human atherosclerosis in a prospective, serial fashion with the use of high-resolution MRI. In this study carotid plaques with evidence of early or recent intraplaque hemorrhage at baseline were significantly more likely to increase in overall wall volume and lipid-rich necrotic core volume and more likely to decrease in lumen size over an 18-month period. Furthermore, patients with intraplaque hemorrhage at baseline had a greater probability of repeated carotid intraplaque hemorrhage.
Findings from this study are consistent with the suggestion by Kolodgie et al4 that intraplaque hemorrhage may stimulate atherogenic activity by being a source of free cholesterol and macrophage activation. Other studies based on histopathological examination of excised plaques also suggest that intraplaque hemorrhage may be a powerful instigator of progression.19–21 Kockx et al20 reported that intraplaque microhemorrhages initiated phagocytosis of platelets and erythrocytes, leading to activation of inducible NO synthase expressing foam cells. They also found that clusters of macrophages located near microvessels appeared to have ingested red cell fragments. Yuan et al21 noted that erythrophagocytosis enhances the capacity of macrophages to oxidize LDL, leading to conversion of LDL to a form that promotes lipid accumulation in macrophages. This longitudinal, in vivo study corroborates these in vitro findings by showing not only an increase in the total volume of the atherosclerotic vessel wall (6.8±7.9%) but also an increase in the lipid-rich necrotic core (28.4±29.7%) among plaques with intraplaque hemorrhage at baseline. In addition, iron staining of the endarterectomy specimens in this study demonstrated that the iron deposits within macrophages were frequently observed in tissues peripheral to the necrotic core, suggesting that erythrophagocytosis might contribute to the formation of foam cells, as previously described.4,20
In contrast, the control group did not exhibit an increase in plaque volume (–0.15±5.1%) and showed a decrease in the lipid-rich necrotic core volume (–5.2+17.3%). These results suggest that plaques without hemorrhage are less active or perhaps that statin treatment (73.3% of control subjects) truly did abolish progression of arteriosclerosis, as previously described in studies by Corti et al22 and Nissen et al.23 However, because plaques in the hemorrhage group progressed rapidly despite treatment with statins (64.3% of subjects), this may indicate that the physical increase in plaque volume and cholesterol supply by repeated intraplaque hemorrhage may exceed the effect of statin therapy. On the other hand, statin-treated plaques in the hemorrhage group had a smaller percent change of wall (5.9±6.3% versus 8.4±10.9%; P=0.6) and lipid-rich necrotic core (25.4±14.7% versus 33.9±48.7%; P=0.6) volume, which suggests that statin therapy may suppress the progression of plaques even in the presence of hemorrhage. However, neither change in these volumes was statistically significant. Further study with a larger group of patients with intraplaque hemorrhage is needed for this aspect of the study.
Findings from this study also suggest the importance of repeated hemorrhage into the plaque as an additional mechanism of progression. New hemorrhages occurred in at least 43% of patients with previous intraplaque hemorrhage, whereas no new hemorrhages appeared in the control group. Plaques with MRI evidence of repeated intraplaque hemorrhage were associated with a significant increase in mean wall thickness (14.8% versus 3.7%; P=0.013) and decrease in lumen volume (–16.4% versus –2.5%; P=0.013) compared with plaques without evidence of repeated intraplaque hemorrhage. This difference suggests that repeated intraplaque hemorrhage may be a marker for plaque instability and may be a contributing factor toward a more rapid progression of atherosclerosis. The disruption of thin-walled microvessels with discontinuous endothelium is believed to be a cause of intraplaque hemorrhage. These fragile microvessels are implicated in repeated intraplaque capillary rupture.3,20 Indeed, close histological examination of the necrotic core of the excised CEA specimens showed extensive microvasculature with surrounding intraplaque hemorrhage of an earlier phase than that deep within the core (Figure 4). On the other hand, the change in wall and lipid-rich necrotic core volume absent the volume of new hemorrhage was still significantly greater than the change of the control group. Furthermore, intraplaque hemorrhage patients without new hemorrhage during follow-up also had greater change in wall and lipid-rich necrotic core volumes than the controls. These data suggest that increase in wall and lipid-rich necrotic core volume may result not only from the influx of new hemorrhage but also from other mechanisms, possibly from increase in plaque cholesterol and activation of macrophages, as suggested by Kolodgie et al.4
Our previously published work established the criteria for the identification of the stages of intraplaque hemorrhage with the use of multicontrast MRI.10,12 The concept of the change in signal intensity of intraplaque hemorrhage was based on the detection and staging of cerebral hemorrhage with the use of multicontrast MR images. The stages of hemorrhage are based on red blood cell content and the oxygenation of hemoglobin. The criteria used to classify early subacute cerebral hemorrhage were applied to the carotid intraplaque hemorrhage and given the term fresh. Morphologically, this includes the presence of intact red blood cells, macrophage and inflammatory cell infiltrate, and evidence of early thrombus formation, which by MRI appears hyperintense on TOF and T1W images. The second category, termed recent, used the late subacute criteria from brain hemorrhage, which includes hemorrhagic debris, intact and degenerating red blood cells, and the increased interstitial fluid associated with inflammation and early organization. These latter components combine to create hyperintense signal patterns in all contrast weightings. Our use of the terms type I and type II rather than fresh and recent, respectively, is based on an evolving understanding of the distinct time course of intraplaque hemorrhage in the carotid artery. This study demonstrates that in the majority (94%) of locations with intraplaque hemorrhage at baseline, the signal intensity associated with hemorrhage remained unchanged at the 18-month time point. MRI studies of intracranial hemorrhage demonstrate a rapid evolution of signal intensity changes within days after the hemorrhagic event that are related to the oxidative state of hemoglobin and the eventual lysis of red blood cells.24 Erythrocyte lysis and hemoglobin oxygenation are heavily dependent on local factors such as oxygen tension and pH.25,26 Our results suggest that atherosclerotic intraplaque hemorrhage may have a longer course for change due to inherent differences in the tissue environment within the necrotic core. Intraplaque hemorrhage may attract macrophages that eventually become trapped within the core and are unable to survive.4 Our histological assessment showed that macrophages with iron inclusions were found in fibrous areas around the core. The infrequency of macrophage iron-laden cells in the core indicates that the degradation of hemorrhage may be slow because of reduced activation of macrophages. Alternatively, repeated hemorrhage into the same plaque region may account for the absence of expected signal intensity changes over the 18-month interval between scans. Because our study included a relatively small number of type I hemorrhage patients at baseline, further analysis using more patients will be needed to solidify our understanding of the signal intensity change of type I hemorrhage. Additionally, and to establish the time course of the signal intensity changes of intraplaque hemorrhage, a more frequent schedule of MRI examinations after the intraplaque hemorrhage event is needed.
Finally, findings from this study also provide insight into the remodeling of the carotid atherosclerosis. Lumen stenosis progressed more rapidly in the hemorrhage group than in the control group. Percent change of lumen volume was –8.5±12.2% for the hemorrhage group and 1.5±7.9% for control subjects (P=0.014). However, in comparison, the outer wall volume increased by only a small amount among cases compared with controls (1.5±6.2% versus 0.2±46%, respectively; P=NS), suggesting that positive remodeling played little role in the increase in wall volume associated with intraplaque hemorrhage. Both of these trends are consistent with previous histopathological studies indicating that intraplaque hemorrhage has a significant correlation primarily with increased luminal narrowing.27
As demonstrated in this study, the clinical implications of serial MR examinations are significant. MRI can give information not only about the change in lumen size and plaque volume but also about plaque components,11 inflammation,28 and fibrous cap status17 noninvasively. Findings from these studies may provide new insight into the pathophysiology of atherosclerosis and permit direct assessment of the effect of pharmacological therapy.
Limitations of Study
At baseline, the lipid-rich necrotic core volume was larger in the hemorrhage group than in the controls, which was anticipated, given that intraplaque hemorrhage is typically admixed within the necrotic core of the plaque. Of note, however, when we controlled for baseline volume of lipid-rich necrotic core and baseline wall volume, the association of intraplaque hemorrhage with increased plaque volume was still statistically significant (P=0.009). It is important to remember that however intriguing these results may be, this study focused only on the carotid artery, and therefore extrapolation to other vascular beds is not possible at this time.
Also of note, 56 (19.7%) of the 284 arteries available for analysis were excluded because of marginal image quality. New technological developments in hardware and image acquisition techniques should improve overall image quality and reduce the number of excluded cases.
Conclusions
This in vivo MRI progression study demonstrates that hemorrhage into the carotid atherosclerotic plaque is significantly associated with progression in wall and lipid-rich necrotic core size and a decrease in lumen volume. Furthermore, lesions that had intraplaque hemorrhage at baseline had a greater probability of repeated intraplaque hemorrhage at 18 months. We propose that repeated hemorrhage and the resulting consequences of cholesterol released from trapped erythrocytes and activated macrophages may be important factors in the rapid progression of carotid atherosclerosis.
Acknowledgments
This study was supported by National Institutes of Health grants R01 HL61851, R01 HL073401, and HL67406.
References
Libby P. Inflammation in atherosclerosis. Nature. 2002; 420: 868–874.
de Boer OJ, van der Wal AC, Teeling P, Becker AE. Leucocyte recruitment in rupture prone regions of lipid-rich plaques: a prominent role for neovascularization; Cardiovasc Res. 1999; 41: 443–449.
Virmani R, Narula J, Farb A. When neoangiogenesis ricochets. Am Heart J. 1998; 136: 937–939.
Kolodgie FD, Gold HK, Burke AP, Fowler DR, Kruth HS, Weber DK, Farb A, Guerrero LJ, Hayase M, Kutys R, Narula J, Finn AV, Virmani R. Intraplaque hemorrhage and progression of coronary atheroma. N Engl J Med. 2003; 349: 2316–2325.
Yeagle PL. Cholesterol and the cell membrane. Biochim Biophys Acta. 1985; 822: 267–287.
Toussaint JF, LaMuraglia GM, Southern JF, Fuster V, Kantor HL. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo. Circulation. 1996; 94: 932–938.
Yuan C, Mitsumori LM, Ferguson MS, Polissar NL, Echelard D, Ortiz G, Small R, Davies JW, Kerwin WS, Hatsukami TS. In vivo accuracy of multispectral magnetic resonance imaging for identifying lipid-rich necrotic cores and intraplaque hemorrhage in advanced human carotid plaques. Circulation. 2001; 104: 2051–2056.
Cai JM, Hatsukami TS, Ferguson MS, Small R, Polissar NL, Yuan C. Classification of human carotid atherosclerotic lesions with in vivo multicontrast magnetic resonance imaging. Circulation. 2002; 106: 1368–1373.
Fayad ZA. MR imaging for the noninvasive assessment of atherothrombotic plaques. Magn Reson Imaging Clin N Am. 2003; 11: 101–113.
Chu B, Kampschulte A, Ferguson MS, Kerwin WS, Yarnykh VL, O’Brien KD, Polissar NL, Hatsukami TS, Yuan C. Hemorrhage in the atherosclerotic carotid plaque: a high-resolution MRI study. Stroke. 2004; 35: 1079–1084.
Saam T, Ferguson MS, Yarnykh VL, Takaya N, Xu D, Polissar NL, Hatsukami TS, Yuan C. Quantitative evaluation of carotid plaque composition by in vivo MRI. Arterioscler Thromb Vasc Biol. 2005; 25: 234–239.
Zhao XQ, Yuan C, Hatsukami TS, Frechette EH, Kang XJ, Maravilla KR, Brown BG. Effects of prolonged intensive lipid-lowering therapy on the characteristics of carotid atherosclerotic plaques in vivo by MRI: a case-control study. Arterioscler Thromb Vasc Biol. 2001; 21: 1623–1629.
Zhang S, Cai J, Luo Y, Han C, Polissar NL, Hatsukami TS, Yuan C. Measurement of carotid wall volume and maximum area with contrast-enhanced 3D MR imaging: initial observations. Radiology. 2003; 228: 200–205.
Thackray BD, Burns DH, Ferguson MS, Gordon D, Beach KW, Hatsukami T, Detmer PR, Primozich JF, Strandness DE Jr. A new method for studying plaque morphology. Am J Card Imaging. 1995; 9: 149–156.
Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantitation of high-density-lipoprotein cholesterol. Clin Chem. 1982; 28: 1379–1388.
Warnick GR. Enzymatic methods for quantification of lipoprotein lipids. Methods Enzymol. 1986; 129: 101–123.
Hatsukami TS, Ross R, Polissar NL, Yuan C. Visualization of fibrous cap thickness and rupture in human atherosclerotic carotid plaque in vivo with high-resolution magnetic resonance imaging. Circulation. 2000; 102: 959–964.
Sadahira Y, Kanzaki A, Wada H, Yawata Y. Immunohistochemical identification of erythroid precursors in paraffin embedded bone marrow sections: spectrin is a superior marker to glycophorin. J Clin Pathol. 1999; 52: 919–921.
Arbustini E, Morbini P, D’Armini AM, Repetto A, Minzioni G, Piovella F, Vigano M, Tavazzi L. Plaque composition in plexogenic and thromboembolic pulmonary hypertension: the critical role of thrombotic material in pultaceous core formation. Heart. 2002; 88: 177–182.
Kockx MM, Cromheeke KM, Knaapen MW, Bosmans JM, De Meyer GR, Herman AG, Bult H. Phagocytosis and macrophage activation associated with hemorrhagic microvessels in human atherosclerosis. Arterioscler Thromb Vasc Biol. 2003; 23: 440–446.
Yuan XM, Anders WL, Olsson AG, Brunk UT. Iron in human atheroma and LDL oxidation by macrophages following erythrophagocytosis. Atherosclerosis. 1996; 124: 61–73.
Corti R, Osende JI, Fallon JT, Fuster V, Mizsei G, Jneid H, Wright SD, Chaplin WF, Badimon JJ. The selective peroxisomal proliferator-activated receptor-gamma agonist has an additive effect on plaque regression in combination with simvastatin in experimental atherosclerosis: in vivo study by high-resolution magnetic resonance imaging. J Am Coll Cardiol. 2004; 43: 464–473.
Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis: a randomized controlled trial. JAMA. 2004; 291: 1071–1080.
Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology. 1993; 189: 15–26.
Brajtburg J, Elberg S, Schwartz DR, Vertut-Croquin A, Schlessinger D, Kobayashi GS, Medoff G. Involvement of oxidative damage in erythrocyte lysis induced by amphotericin B. Antimicrob Agents Chemother. 1985; 27: 172–176.
Yang E, Huestis WH. Oxidative interactions between the erythrocyte membrane and phosphatidylcholine vesicles. J Biol Chem. 1994; 269: 14518–14524.
Imparato AM, Riles TS, Mintzer R, Baumann FG. The importance of hemorrhage in the relationship between gross morphologic characteristics and cerebral symptoms in 376 carotid artery plaques. Ann Surg. 1983; 197: 195–203.
Kerwin W, Hooker A, Spilker M, Vicini P, Ferguson M, Hatsukami T, Yuan C. Quantitative magnetic resonance imaging analysis of neovasculature volume in carotid atherosclerotic plaque. Circulation. 2003; 107: 851–856.(Presence of Intraplaque H)