Histologic characteristics of carotid artery plaque T. Martin Feeley, MCh, FRCSI, Edward J. Leen, MB, Mary-Paula Colgan, MD, Dermot J. Moore, MD, FRCSI, Dermot O'Brien Hourihane, MD, F R C P I , FRCPath, and Gregor D. Shanik, MD, MCh, FRCSI, FACS, Dublin, Ireland Carotid plaque characteristics associated with the production of symptoms were identified with quantification of carotid plaque constituents in high-grade stenotic asymptomatic (n = 8) and symptomatic (n = 44) plaques. Asymptomatic plaques contained significantly more fibrous/collagen material (88%) than symptomatic plaques (66%) (p < 0.05). Hemorrhage constituted 2% and 1% of asymptomatic and symptomatic plaques, respectively. The predominant nonfibrous material was a pink amorphous material mixed with cholesterol, which composed 7% of asymptomatic and 27% of symptomatic plaques (p < 0.05). No relationship was found between plaque composition and the number of ipsilateral ischemic neurologic events, nor was there evidence of a healing process. B-mode ultrasound scanning had a sensitivity of 94% in identifying plaque with > 80% fibrous content. We believe that plaque composition may be a useful discriminating factor in selecting asymptomatic patients for carotid endarterectomy. (J VAsc Smtc 1991;13: 719-24.)
Most strokes are associated with carotid bifurcation disease, and most are thromboembolic in origin. Since many of the strokes caused by thromboembolization are not heralded by a transient ischemic attack it behooves us to identify which carotid bifurcation plaques are the harbingers of future disasters. In 1982 Lusby et al.1 reported a close association between intraplaque hemorrhage and ipsilateral cerebral symptoms, a finding since confirmed by a number of studies. 2"3We have recently demonstrated a strong correlation between focal cerebral events or cerebral infarction as detected by CT scanning and heterogenous or echolucent plaques as classified by B-mode ultrasonography, 4 thus confirming that ischemic cerebral events are associated with hemorrhagic or echolucent plaques. Given this association the following questions must be answered: (1) Is the echolucent material hemorrhage? (2) What plaque characteristics or contents predispose to embolization? and (3) What is the natural history of carotid plaque? In this study plaque constituents were identified From the Departments of Vascular Surgery and Pathology, St. James' Hospital, Dublin, Ireland. Reprint requests: Prof. Gregor Shanik, FRCSI, Department of Vascular Surgery,St. James'Hospital, P.O. Box 580, James' St., Dublin 8, Ireland. 24/1/27206
histologically, and plaque composition was quantitatively assessed. With these data we set out to (1) Identify differences between symptomatic and asymptomatic plaques; (2) Identify a relationship, if ar,T, between plaque composition and number of symptoms caused by the plaque; (3) Establish if a healing process occurs in plaques after the occurrence of symptoms; and (4) Assess the accuracy of duplex scanning in identifying potentially dangerous plaques. METHODS Forty-four symptomatic and eight asymptomatic plaques from 51 patients were studied. Patients' symptoms, the number of ischemic neurologic events, and the interval between the most recent event and surgery were recorded. All carotid arteries were scanned before operation (Hoffrel 516SD; Hoffrel Instruments, Inc., South Norwalk, Conn.), and plaques were classified according to criteria similar to those recently described by Gray-Weale et al.~: I = Large echolucent component II = Mixed echolucent and echogenic areas III = Echogenic with minimal echolucent component IV = Uniformly echogenic After operation plaques were fixed in 10% formaldehyde, decalcified and multiple sections 719
720
;ourna~ of VASCULAtl SURGERY
Feeley et al.
8O
60
It
2
.ID
,T
o I week
2-6 weeks
7-12 weeks
> 12 weeks
Symptom - Surgery Inlerval
Fig. 1. Comparison of percent fibrous plaque content and time interval since the most recent ischemic event.
were stained with hematoxylin and eosin, Martius scarlet blue, Perl's Prussian blue, Verhoefl%, and Van Gieson's stains. Plaque contents were identified as fibrous/collagen, hemorrhage, fibrin, thrombus, calcium, lipid/cholesterol. The percentage of plaque occupied by each constituent was calculated by planimetry with a CALCOMP 2000 (California Computer Products, Anaheim, Calif.) linked to an Apple 2C (Apple Computer, Inc., Cupertino, Calif.) desktop computer. Multiple cross sections from each plaque were projected on a tablet where each plaque constituent was outlined as was the outline of the plaque cross section and lumen. The percentage cross-sectional area occupied by each constituent was computed. For the purposes of analysis, fibrous material was considered "hard" as this is identified as echogenic on B-mode scanning. Nonfibrous material, hemorrhage, fibrin, cholesterol, and cholesterol compound were considered "soft" as they are identified as echolucent on B-mode uhrasonography. Observers responsible for plaque analysis were unaware of clinical presentation or results of duplex scanning. RESULTS Eighteen of the 44 symptomatic plaques were associated with 1 cerebral event, 7 with 2 events,
8 with 3 to 5 events, and 11 with more than 5 events. The interval from last ischemic event to operation varied from less than i week (n = 9) to i year, with a mean of 16 weeks (Fig. 1). All plaques, asymptomatic and symptomatic, caused greater than 50% luminal stenosis, confirmed by angiography, and there was no difference in the degree of carotid stenosis in the two groups (94% + 2% vs 88% -+ 6%), respectively. Fig. 2 shows the plaque composition in symptomatic and asymptomatic plaques. No difference was found in hemorrhage or fibrin content between plaque types. There was, however, significantly greater fibrous content with correspondingly less pink amorphous/cholesterol material in asymptomatic plaques (Figs. 3, A and B). Recent hemorrhage was seen in three of the eight asymptomatic plaques and 25 of the 44 symptomatic plaques, but occupied greater than 1% of the plaque content in only 13 (29%). Fibrin was present in 4 asymptomatic and 25 symptomatic plaques. The dominant nonfibrous plaque constituent was an amorphous material that stained pink with hematoxylin and eosin. The material was always mixed with variable amounts of cholesterol (Figs. 4 A and B). This was present in 3 asymptomatlc and 36 symptomatic plaques. In both asymptomatic and symptomatic plaques
Volume 13 Number 5 May 1991
Carotid plaque morphology 721
100
80
60" p 5) did contain slightly more soft material than those associated with one event, but this difference was not significant. This study confirms the accuracy of B-mode ultrasonography in identifying soft plaque. It is surprisingly accurate when one considers that more than 60% of the symptomatic plaque (so-called soft plaque) is fibrous material. Taking plaque soft content of 20% or greater as "soft," duplex scanning
724
Journal of VASCULAR SURGERY
Feeley et al.
had a sensitivity of 94% in characterizing plaque type. We have previously shown that heterogenous plaque, as detected ultrasonographically, is a greater risk factor for development of cerebral symptoms and cerebral infarction than the degree of luminal stenosis. 4 This study confirms that heterogenous plaques are associated with cerebral symptoms, and that duplex scanning can accurately identify these potentially dangerous plaques. This has important implications for the m.anagement of asymptomatic carotid stenosis. More detailed information on the natural history of atheromatous carotid plaques is now required. In summary, we have shown that the histologic difference between symptomatic and asymptomatic plaque is the greater content of amorphous/cholesterol material in symptomatic plaques. In spite of the small numbers of asymptomatic plaques the percentage fibrous content was significantly different. We were unable to establish a relationship between plaque content and the number of ischemic events, nor were we able to identify a healing process. Further work is required to fully establish the natural history of carotid artery plaque. We acknowledge the help of Dr. Scan O'Briain in carrying out the planimetry. REFERENCES
1. Lusby RJ, Ferrell LD, Ehrenfield WK, Stoney RJ, Wylie EJ. Carotid plaque hemorrhage: its role in the production of cerebral ischemia. Arch Surg 1982; 117:1479-88.
2. Ammar AD, Wilson RL, Travers H, Lin ll, Farha SJ, Chang FC. Intraplaque hemorrhage: its significance in cerebrovascular disease. Am J Surg 1984;148:840-3. 3. 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;197:195203.
4. Leahy AL, McCollum PT~ Feeley TM, et al. Duplex ultrasonography and selection of patients ['or carotid endarterectomy: plaque morphology or luminal narrowing? J VASe SURG 1988;8:558-62. 5. Gray-Weale AC, Graham JC, Burnett JR, Byrne K, Lusby RJ. Carotid artery,atheroma: comparison of preoperative B-mode ultrasound appearance with carotid endarterectomy specimen pathology. J Cardiovasc Surg 1988;29:676-81. 6. Bassiouny HS, Davis H, Massawa N, Gewertz BL, Gtagov S, Zarins CK. Critical carotid stenoses: morphologic and chemical similarity between symptomatic and asymptomatic plaques. J VAse SURG 1989;9:202-12. 7. Reilly LM, Lusby RJ, Hughes L, Ferrdl LD, Stoney RJ, Ehrenfeld WK. Carotid plaquc histology using real-time ultrasonography. Clinical and therapeutic implications. Am J Surg 1983; 146:188-93. 8. AldooriMI, Baird RN, AI-Sam SZ, Cole SEA, Mera S, Davies JD. Duplex scanning and plaque histology, in cerebral ischemia. Eur J Vase Surg 1987; 1: 159-64. 9. Persson AV, Robichaux WT, Silverman M. The natural history of carotid plaque development. Arch Surg 1983; 118: 1048-52. 10. Leen EJ, Feeley TM, Colgan MP, et al. "Haemorrhagic" carotid plaque does not contain haemorrhage. Eur J Vase Surg 1990;4:123-8. Submitted June 4, 1990; accepted Nov. 26, 1990.
Most strokes are associated with carotid bifurcation disease, and most are thromboembolic in origin. Since many of the strokes caused by thromboembolization are not heralded by a transient ischemic attack it behooves us to identify which carotid bifurcation plaques are the harbingers of future disasters. In 1982 Lusby et al.1 reported a close association between intraplaque hemorrhage and ipsilateral cerebral symptoms, a finding since confirmed by a number of studies. 2"3We have recently demonstrated a strong correlation between focal cerebral events or cerebral infarction as detected by CT scanning and heterogenous or echolucent plaques as classified by B-mode ultrasonography, 4 thus confirming that ischemic cerebral events are associated with hemorrhagic or echolucent plaques. Given this association the following questions must be answered: (1) Is the echolucent material hemorrhage? (2) What plaque characteristics or contents predispose to embolization? and (3) What is the natural history of carotid plaque? In this study plaque constituents were identified From the Departments of Vascular Surgery and Pathology, St. James' Hospital, Dublin, Ireland. Reprint requests: Prof. Gregor Shanik, FRCSI, Department of Vascular Surgery,St. James'Hospital, P.O. Box 580, James' St., Dublin 8, Ireland. 24/1/27206
histologically, and plaque composition was quantitatively assessed. With these data we set out to (1) Identify differences between symptomatic and asymptomatic plaques; (2) Identify a relationship, if ar,T, between plaque composition and number of symptoms caused by the plaque; (3) Establish if a healing process occurs in plaques after the occurrence of symptoms; and (4) Assess the accuracy of duplex scanning in identifying potentially dangerous plaques. METHODS Forty-four symptomatic and eight asymptomatic plaques from 51 patients were studied. Patients' symptoms, the number of ischemic neurologic events, and the interval between the most recent event and surgery were recorded. All carotid arteries were scanned before operation (Hoffrel 516SD; Hoffrel Instruments, Inc., South Norwalk, Conn.), and plaques were classified according to criteria similar to those recently described by Gray-Weale et al.~: I = Large echolucent component II = Mixed echolucent and echogenic areas III = Echogenic with minimal echolucent component IV = Uniformly echogenic After operation plaques were fixed in 10% formaldehyde, decalcified and multiple sections 719
720
;ourna~ of VASCULAtl SURGERY
Feeley et al.
8O
60
It
2
.ID
,T
o I week
2-6 weeks
7-12 weeks
> 12 weeks
Symptom - Surgery Inlerval
Fig. 1. Comparison of percent fibrous plaque content and time interval since the most recent ischemic event.
were stained with hematoxylin and eosin, Martius scarlet blue, Perl's Prussian blue, Verhoefl%, and Van Gieson's stains. Plaque contents were identified as fibrous/collagen, hemorrhage, fibrin, thrombus, calcium, lipid/cholesterol. The percentage of plaque occupied by each constituent was calculated by planimetry with a CALCOMP 2000 (California Computer Products, Anaheim, Calif.) linked to an Apple 2C (Apple Computer, Inc., Cupertino, Calif.) desktop computer. Multiple cross sections from each plaque were projected on a tablet where each plaque constituent was outlined as was the outline of the plaque cross section and lumen. The percentage cross-sectional area occupied by each constituent was computed. For the purposes of analysis, fibrous material was considered "hard" as this is identified as echogenic on B-mode scanning. Nonfibrous material, hemorrhage, fibrin, cholesterol, and cholesterol compound were considered "soft" as they are identified as echolucent on B-mode uhrasonography. Observers responsible for plaque analysis were unaware of clinical presentation or results of duplex scanning. RESULTS Eighteen of the 44 symptomatic plaques were associated with 1 cerebral event, 7 with 2 events,
8 with 3 to 5 events, and 11 with more than 5 events. The interval from last ischemic event to operation varied from less than i week (n = 9) to i year, with a mean of 16 weeks (Fig. 1). All plaques, asymptomatic and symptomatic, caused greater than 50% luminal stenosis, confirmed by angiography, and there was no difference in the degree of carotid stenosis in the two groups (94% + 2% vs 88% -+ 6%), respectively. Fig. 2 shows the plaque composition in symptomatic and asymptomatic plaques. No difference was found in hemorrhage or fibrin content between plaque types. There was, however, significantly greater fibrous content with correspondingly less pink amorphous/cholesterol material in asymptomatic plaques (Figs. 3, A and B). Recent hemorrhage was seen in three of the eight asymptomatic plaques and 25 of the 44 symptomatic plaques, but occupied greater than 1% of the plaque content in only 13 (29%). Fibrin was present in 4 asymptomatic and 25 symptomatic plaques. The dominant nonfibrous plaque constituent was an amorphous material that stained pink with hematoxylin and eosin. The material was always mixed with variable amounts of cholesterol (Figs. 4 A and B). This was present in 3 asymptomatlc and 36 symptomatic plaques. In both asymptomatic and symptomatic plaques
Volume 13 Number 5 May 1991
Carotid plaque morphology 721
100
80
60" p 5) did contain slightly more soft material than those associated with one event, but this difference was not significant. This study confirms the accuracy of B-mode ultrasonography in identifying soft plaque. It is surprisingly accurate when one considers that more than 60% of the symptomatic plaque (so-called soft plaque) is fibrous material. Taking plaque soft content of 20% or greater as "soft," duplex scanning
724
Journal of VASCULAR SURGERY
Feeley et al.
had a sensitivity of 94% in characterizing plaque type. We have previously shown that heterogenous plaque, as detected ultrasonographically, is a greater risk factor for development of cerebral symptoms and cerebral infarction than the degree of luminal stenosis. 4 This study confirms that heterogenous plaques are associated with cerebral symptoms, and that duplex scanning can accurately identify these potentially dangerous plaques. This has important implications for the m.anagement of asymptomatic carotid stenosis. More detailed information on the natural history of atheromatous carotid plaques is now required. In summary, we have shown that the histologic difference between symptomatic and asymptomatic plaque is the greater content of amorphous/cholesterol material in symptomatic plaques. In spite of the small numbers of asymptomatic plaques the percentage fibrous content was significantly different. We were unable to establish a relationship between plaque content and the number of ischemic events, nor were we able to identify a healing process. Further work is required to fully establish the natural history of carotid artery plaque. We acknowledge the help of Dr. Scan O'Briain in carrying out the planimetry. REFERENCES
1. Lusby RJ, Ferrell LD, Ehrenfield WK, Stoney RJ, Wylie EJ. Carotid plaque hemorrhage: its role in the production of cerebral ischemia. Arch Surg 1982; 117:1479-88.
2. Ammar AD, Wilson RL, Travers H, Lin ll, Farha SJ, Chang FC. Intraplaque hemorrhage: its significance in cerebrovascular disease. Am J Surg 1984;148:840-3. 3. 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;197:195203.
4. Leahy AL, McCollum PT~ Feeley TM, et al. Duplex ultrasonography and selection of patients ['or carotid endarterectomy: plaque morphology or luminal narrowing? J VASe SURG 1988;8:558-62. 5. Gray-Weale AC, Graham JC, Burnett JR, Byrne K, Lusby RJ. Carotid artery,atheroma: comparison of preoperative B-mode ultrasound appearance with carotid endarterectomy specimen pathology. J Cardiovasc Surg 1988;29:676-81. 6. Bassiouny HS, Davis H, Massawa N, Gewertz BL, Gtagov S, Zarins CK. Critical carotid stenoses: morphologic and chemical similarity between symptomatic and asymptomatic plaques. J VAse SURG 1989;9:202-12. 7. Reilly LM, Lusby RJ, Hughes L, Ferrdl LD, Stoney RJ, Ehrenfeld WK. Carotid plaquc histology using real-time ultrasonography. Clinical and therapeutic implications. Am J Surg 1983; 146:188-93. 8. AldooriMI, Baird RN, AI-Sam SZ, Cole SEA, Mera S, Davies JD. Duplex scanning and plaque histology, in cerebral ischemia. Eur J Vase Surg 1987; 1: 159-64. 9. Persson AV, Robichaux WT, Silverman M. The natural history of carotid plaque development. Arch Surg 1983; 118: 1048-52. 10. Leen EJ, Feeley TM, Colgan MP, et al. "Haemorrhagic" carotid plaque does not contain haemorrhage. Eur J Vase Surg 1990;4:123-8. Submitted June 4, 1990; accepted Nov. 26, 1990.
