Cerebral Microembolism and Early Recurrent Cerebral or Retinal Ischemic Events
Background and Purpose We investigated whether cerebral microembolism as detected by transcranial Doppler ultrasonography (TCD) identifies patients at an increased risk for early, recurrent cerebral or retinal ischemic events.
Methods Records of consecutive patients examined during a 40-month period in the Neurovascular Laboratory were reviewed for the presence of cerebral microembolism. Of the original 302 patients, 229 with 310 arteries met inclusionary criteria. Follow-up information was obtained from the laboratory’s database as well as the hospital records. Microembolus detection studies were performed on TC-2000 or TC-2020 instruments equipped with special software, and criteria established a priori were used for microembolus selection. TCD testing was performed a median interval of 9 days after the initial symptoms of cerebral ischemia. Severity of arterial stenosis was determined by cerebral angiography or noninvasive methods.
Results Microembolic signals were detected more frequently in symptomatic (40/140; 28.6%) than asymptomatic (21/170; 12.4%) arteries (P<.001). Ten recurrent ischemic events occurred during a median follow-up of 8 days after TCD examination, all in the territories of symptomatic arteries. Nine events occurred in the territories of microembolic signal–positive arteries (9/61; 14.8%) and one in the territory of a microembolic signal–negative artery (1/249; 0.4%) (P<.001). No association was detected in the subgroup with known cardiac lesions. Microembolic signals were more frequent in arteries with lesions causing 70% or more stenosis or occlusion (26/99; 26.3%) than in those with a degree of stenosis less than 70% (17/126; 13.5%) (P=.016).
Conclusions In this retrospective study, microembolic signals were more common in the territories of symptomatic arteries and particularly those with severely stenotic lesions. During a short follow-up, recurrent ischemic events were more common along the territories of arteries with TCD-detected microembolism and previous symptoms of cerebral or retinal ischemia.
The increased incidence of recurrent cerebral ischemic events during the days and weeks after a TIA or a cerebral infarct has long been recognized.1 “Stroke in evolution,” “progressive stroke,” and early “recurrent stroke” are terms frequently used to describe the development of new neurological symptoms and imply the extension of ischemia to previously spared cerebral regions.2 3 Clinically identified conditions such as atrial fibrillation and severe, extracranial internal carotid artery stenosis increase the risk of early recurrent events.4 5 It is, however, frequently impossible for the treating physician to identify or monitor intravascular flow changes and embolism that may be the immediate cause of recurrent ischemia in an individual patient. Available techniques, such as cerebral angiography, are associated with a risk of complications. Brain CT or MR imaging and angiography are almost always in use. Still, these methods do not show direct, real-time evidence of microembolism. The ability to diagnose the latter is of clinical interest since its identification may potentially be helpful in selecting specific treatment modalities.
High-intensity transient signals are frequently detected during TCD testing in patients with symptoms of cerebral ischemia.6 7 8 These signals, also called microembolic signals, have also been observed during operative procedures.9 10 Laboratory and animal models suggest that they correspond to microbubbles or formed element particles composed of platelet-rich aggregates, atheromatous material, or fat.11 Examination of carotid endarterectomy specimens shows that they may correspond to platelet and fibrin particles originating at stenotic lesions of the cervical carotid artery.12 The ability to detect these signals is a relatively new development of TCD technology and provides a means to monitor cerebral microembolism in vivo.
In this study we investigated whether in patients with cerebrovascular disease, the presence of microembolic signals along the internal carotid or basilar artery territories was associated with new or recurrent cerebral ischemic events.
Subjects and Methods
The records of consecutive patients examined between March 29, 1993, and July 31, 1996, at the Neurovascular Laboratory of our hospital were reviewed. Three hundred two patients with 404 arteries had adequate ultrasonic windows and were examined during the study period. Most of these patients were either hospitalized (Stroke or Vascular Surgery Services) or were followed at the outpatient Stroke Clinic. Ninety-four arteries were excluded for the following reasons: incomplete TCD data sheets or unavailable medical records (n=17), clinical and laboratory data complete but insufficient to establish diagnosis of ischemic cerebrovascular disease with reasonable certainty (n=35), follow-up duration of less than 24 hours (n=14), arterial territory of TIA not determined with reasonable certainty (n=10), and recent carotid endarterectomy or angioplasty ipsilateral to the insonated side (n=9). In addition, 9 arteries were excluded because of the diagnosis of concomitant neurological or hematologic diseases including subdural hematoma (n=5), giant cell arteritis (n=1), hypercoagulable state (n=1), migraine (n=1), and intracranial asymptomatic aneurysm (n=1). The remaining 310 arteries in 229 patients constituted the study sample. The study population had a mean age of 67.7 years (range, 30 to 87 years) and consisted of 223 men and 6 women. Patients were diagnosed with TIAs (n=28 arteries), transient monocular blindness or other retinal ischemic syndromes (n=36 arteries), or cerebral infarction (n=76 arteries). One hundred seventy arteries were asymptomatic.
Arteries were considered “symptomatic” when symptoms of cerebral or retinal ischemia had occurred along their territories during the 6-month period preceding TCD testing. They were otherwise labeled “asymptomatic.” Arteries associated with silent infarcts detected by brain CT or MRI were considered asymptomatic.
Follow-up information was obtained from the Neurovascular Laboratory’s patient records. The latter are routinely completed for each patient within 24 hours of the study and, for hospitalized patients, are updated as needed throughout the hospital stay. They contain summaries of each patient’s neurological diagnoses, other medical conditions, follow-up information, and neuroradiological and TCD test results. In addition, the medical record or, for hospitalized cases, the discharge summary of each patient was reviewed to complete information regarding follow-up events. These documents were reviewed by two of the investigators (V.L.B., C.A.C.W.) using a structured protocol.
A diagnosis of recurrent event was made when a focal deficit of abrupt onset was detected during the period after TCD testing; when applicable, it was confirmed by neuroimaging studies to rule out cerebral edema or bleeding. For patients who had more than one follow-up cerebrovascular event, the dates of only the first event and that of the last follow-up visit were extracted for analysis. Diagnoses were established by the treating physician, almost always either a stroke neurologist or an experienced, board-certified neurology consultant. Each case was reviewed at the time of data collection to confirm the initial clinical impression.
The diagnosis of cardiac disease was based on clinical histories and electrocardiograms in all patients and either transthoracic or transesophageal echocardiograms in 120 patients. Sonos 1000 or Sonos 1500 instruments (Hewlett Packard) were used for echocardiographic testing. With the exception of one patient with atrial fibrillation and severe extracranial carotid stenosis who had recurrent, stereotypical TIAs ipsilateral to the stenosed carotid and whose TIAs resolved after carotid endarterectomy, arteries in patients with atrial fibrillation (n=34), prosthetic heart valves (n=20), ventricular aneurysms (n=9), left ventricular thrombi (n=4), and severe akinetic segments of the left ventricle (n=13) were considered “cardiac” for the purpose of analysis in this study. There were no patients with infective endocarditis, recent myocardial infarction, atrial septal defect, or left atrial myxoma in the study group.
At the time of TCD testing, most patients were receiving antiplatelet or anticoagulant therapy, in addition to their regular medications. Analyses probing the possible effect of treatment on clinical outcome or the prevalence of microembolism were not performed.
TCD studies were performed on either a TC-2000 or a TC-2020 instrument (Nicolet/EME) equipped with special software for microemboli detection. With the help of a specially designed headband, a 2-MHz probe was immobilized against the temporal bone of each patient, and the distal internal carotid artery or proximal middle cerebral artery was insonated. For studies of the basilar artery, the probe was handheld. The technique of TCD testing has been described before.13 Each microemboli detection study lasted 30 minutes. An experienced technologist monitored the patient as well as the instrument throughout each study and kept a log of potential sources of artifact. Potential microembolic signals were identified as they occurred and were saved on floppy disk for subsequent analysis.
Criteria for microembolic signals were established before data collection. Accepted signals were more than 25 milliseconds in duration, most lasting less than 100 milliseconds; they had an intensity of at least 9 dB above that of the background blood flow, were unidirectional within the Doppler velocity spectrum, and were accompanied by a “chirp” on the audio output. In individuals with more than one TCD examination, only the findings of the first study were included for data analysis.
Cerebrovascular Imaging Studies
The carotid arteries and intracranial vessels were imaged by contrast cerebral angiography in 93 arteries, Duplex ultrasonography in 154, and MR angiography in 43. Imaging studies were not obtained in 20 arteries. With the exception of 35 cases in which the original films could not be located, all imaging studies were reviewed. The original radiology report was used when films could not be located.
The severity of arterial stenosis, when determined by cerebral angiography or MR angiography, was based on a formula modified from Wiebers et al14 and presented in another report.8 It was divided into four categories: mild (0% to 29%), moderate (30% to 69%), and severe (70% to 99%) stenosis and complete occlusion.
A 1.5-T Signa unit (General Electric Medical Systems) was used for MRI. An Ultramark 9-HDI instrument (Advanced Technology Laboratories) was used for duplex imaging. The severity of stenosis determined by ultrasound studies was based on the criteria published in 1996 by Hood et al,15 and studies performed before that date were reread.
All data were stored on a personal computer with the use of Microsoft Excel software (version 5.0). Group comparisons were made with the χ2 test, Fisher’s exact test (two-tailed), or the Wilcoxon rank sum test. Statistical analyses were performed at the Data Coordinating Center of the Boston University School of Public Health.
Microembolic signals were detected in 61 of 310 arteries (19.7%). They were more frequent in symptomatic (40/140; 28.6%) than asymptomatic (21/170; 12.4%) vessels. The difference between the two groups was significant (P<.001). Further analyses showed that this association did not reach statistical significance in patients with identified cardiac lesions (Table 1⇓).
Two cerebral infarcts and 8 transient cerebral or retinal ischemic events occurred along the territories of 10 arteries within 30 days after the TCD testing. All of these recurrent events occurred in the territories of symptomatic arteries. Nine events occurred in the territories of 61 microembolic signal–positive arteries and 1 in the territory of 1 of the 249 microembolic signal–negative arteries. The latter, a TIA, occurred along the distribution of an internal carotid artery with a severe extracranial stenosis. The difference between microembolic signal–positive and –negative arteries with regard to recurrent events after TCD testing was significant (P<.001; RR, 36.7; 95% CI, 4.7 to 284.5). Additional analyses of subgroups of the study sample are summarized in Table 2⇓. No significant association was found between the presence of microembolic signals and recurrent ischemic events in patients with identified cardiac lesions.
The median follow-up after TCD testing for microemboli was 8 days for the group of 310 arteries; it was 3 days for the subgroup of arteries with recurrent events and 8 days for those without recurrence. TCD testing was performed a median interval of 9 days after the initial symptoms of cerebral ischemia.
The relationships between the severity of arterial stenosis, the presence of microembolism, and recurrent clinical events were studied in patients who did not have identified cardiac lesions. Two hundred thirty arteries were designated as “noncardiac”; in 5, imaging studies of the cerebral vasculature were not obtained. The remaining 225 were divided into the following categories: mild stenosis (n=60 arteries), moderate stenosis (n=66 arteries), severe stenosis (n=76), and occlusion (n=23). The frequency of microembolic signals was significantly higher in the groups with occlusion or stenosis equal to or exceeding 70% (26/99; 26.3%) than in the groups with stenosis less than 70% (17/126; 13.5%) (P=.016; RR, 1.95; 95% CI, 1.1 to 3.4). Further analyses regarding the effect of increasing degrees of stenosis on microembolic signals are presented in Table 3⇓.
No significant association was found between the degree of arterial stenosis and recurrent cerebral or retinal ischemic events.
The findings of this study indicate that in patients without cardiac embolic sources, microembolic signals in the territory of a symptomatic cerebral artery can be precursors of recurrent ischemic cerebrovascular events in the distribution of that artery. In addition, the findings show that microembolic signals are more prevalent along the territories of cerebral arteries with severely stenotic lesions.
Several limitations must be taken into consideration when the results of this study are interpreted. First, the retrospective nature of the investigation may have affected the data. The need for a prospective investigation is recognized. Second, the selection of a microembolic signal intensity of 9 dB probably introduced a bias in the study sample toward large particles. Third, the majority of patients were receiving some form of antiplatelet or anticoagulant therapy at the time of TCD testing and during the follow-up period. These medications may have affected the rate of both recurrent ischemic events and microembolic signals and may have introduced an unmeasurable error in the study’s findings. Fourth, multiple imaging modalities were used to measure the severity of arterial stenosis, introducing a degree of inconsistency to our methods. However, even when a relatively crude measure of stenosis based on the criterion of 70% was used rather than stratified values based on small, incremental increases, a significant association was found. The need to complete a similar study using one imaging technology is also recognized.
The finding of a significant association between microembolic signals and recurrent cerebral ischemia in symptomatic patients without clinically detected cardiac sources for embolism considerably extends the observations of previous studies. Although other investigators have found an association between microembolic signals and symptoms of cerebral ischemia, in most studies TCD testing was performed after these symptoms were clinically evident,7 8 16 an observation confirmed by this investigation. Two published studies present data about patients who were prospectively monitored after TCD testing for microembolism. Siebler et al17 followed 48 asymptomatic patients with severe carotid stenosis after TCD testing and found that a middle cerebral artery microembolic rate of 2 or more per hour is associated with an increased risk of developing cerebral ischemia in the territory of the ipsilateral internal carotid artery. Tegeler et al18 reported a trend toward an increased risk for recurrent cerebral ischemic events among those of the 66 patients with cerebral infarction who had positive microemboli studies at baseline. An association between microembolic signals during carotid endarterectomy and perioperative cerebral infarction has also been reported,9 and microembolic signal counts have been linked to the neuropsychological deficit after cardiopulmonary bypass surgery.10 When our observation is reviewed in conjunction with those of the preceding studies, it indicates that the presence of microembolic signals may have diagnostic value by identifying patients at an increased risk for early, recurrent cerebral or retinal ischemia.
Worsening of the neurological condition during the days after the initial symptoms of cerebral ischemia can be seen in 4% to 50% of patients.1 2 19 20 21 22 23 Recurrent cerebral embolism,21 22 impaired collateral flow,24 and other factors20 can contribute to further progression of the original neurological deficit. The exact impact of each of these factors in an individual patient is often not known. In this study, 9 of 10 recurrent ischemic events occurred along the distribution of arteries with ultrasonic evidence of microembolism, suggesting that the latter has a high prevalence in this setting. However, given the study’s technical limitations and the small number of recurrent events during the follow-up period, a cause and effect relationship between microembolism and recurrent cerebral ischemia could not be proven. In addition, although the association’s RR values were high, wide 95% confidence intervals did not establish reliable estimates of these values.
The 4% incidence of new cerebrovascular events during follow-up in this study is low compared with the results of investigations referenced above. It is not unexpected, because only retinal or cerebral ischemic events occurring after TCD testing were included in this analysis. TIAs and cerebral or retinal infarcts that occurred during the 9-day interval between the onset of symptoms and TCD testing were excluded, thus limiting the total number of recurrences. In addition, only recurrent ischemic events that occurred in the territories of insonated arteries were taken into account, and patients with clinical diagnoses of neurological worsening secondary to other causes were excluded. Because this is a retrospective study and not all patients were monitored for prolonged periods of time, it is possible that some recurrent events were not detected. Given the exhaustive nature of our review, we doubt that this limitation had a severe impact on the main findings.
Extensive clinical experience provides support to the notion of a cause and effect relationship between severe arterial stenosis and cerebral infarction.4 25 26 27 28 Severe arterial stenosis was associated with microembolic signals in this study, suggesting that microembolism can be considered a link in the cause and effect chain. Pathological examination of specimens obtained at carotid endarterectomy consistently shows that atheromatous debris and intraluminal thrombi are regularly present in these lesions.26 29 Plaque ulceration and lumen thrombus are also the main sources of microembolic signals in high-grade internal carotid artery stenosis.30 The lack of association between severity of stenosis and recurrent events in this study is to be noted. Taken in conjunction with the significant association between microembolic signals and recurrent events, it supports the previously established notion of embolism as a risk factor for ischemic events irrespective of the severity of stenosis. Its exact causes remain undetermined.
The clinical relevance of microembolic signals in patients at an increased risk for cardioembolic stroke remains unknown. Although microembolic signals have predictive value in subjects with left ventricular assist devices,31 in patients with prosthetic cardiac valves they are not associated with stroke.6 No significant association was detected in cardiac patients in this investigation either; however, the number of studied arteries was too small to derive any conclusions. The reasons for this apparent disparity between patients with or without cardiac sources for embolism are undetermined. Cardiac lesions form a heterogeneous group, with subgroups associated with microemboli of variable composition.6 7 32 The potential for cerebral ischemic damage from microembolism may vary among these subgroups.
In summary, microembolic signals are more common in the territories of symptomatic cerebral arteries, particularly those with severely stenotic lesions, and are associated with an increased risk of recurrent ischemic events. The exact extent of this risk and other specific features of microembolism, such as emboli count or particle size, remain unknown.
Selected Abbreviations and Acronyms
|TCD||=||transcranial Doppler ultrasonography|
|TIA||=||transient ischemic attack|
- Received March 11, 1997.
- Revision received April 9, 1997.
- Accepted April 24, 1997.
- Copyright © 1997 by American Heart Association
Marshall J. The natural history of transient ischaemic cerebrovascular attacks. Quart J Med. 1964;33:309-324.
Haley EC, Kassell NF, Torner JC. Failure of heparin to prevent progression in progressing ischemic infarction. Stroke. 1988;19:10-14.
Bassetti C, Barth A, Bogousslavsky J. Progressive stroke: a prospective study of patients hospitalized in a stroke unit within the first 24 hours. Cerebrovasc Dis. 1996;6(suppl 2):49-50. Abstract.
Wolf PA, Kannel WB, McGee DL, Meeks SL, Bharucha NE, McNamara PM. Duration of atrial fibrillation and imminence of stroke: the Framingham Study. Stroke. 1983;14:664-667.
Babikian VL, Hyde C, Pochay V, Winter MR. Clinical correlates of high intensity transient signals detected on transcranial Doppler sonography in patients with cerebrovascular disease. Stroke. 1994;25:1570-1573.
Jansen C, Ramos LMP, van Heesewijk JPM, Moll FL, van Gijn J, Ackerstaff RGA. Impact of microembolism and hemodynamic changes in the brain during carotid endarterectomy. Stroke. 1994;25:992-997.
Pugsley W, Klinger L, Paschalis C, Treasure T, Harrison M, Newman S. The impact of microemboli during cardiopulmonary bypass on neuropsychological functioning. Stroke. 1994;25:1393-1399.
Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke. 1991;22:253-258.
Babikian VL, Rosales R, Pochay V. Composition of particles associated with embolic signals on transcranial Doppler ultrasonography. J Stroke Cerebrovasc Dis. 1994;4:86-90.
Saver JL, Feldmann E. Basic transcranial Doppler examination: technique and anatomy. In: Babikian VL, Wechsler LR, eds. Transcranial Doppler Ultrasonography. St Louis, Mo: Mosby–Year Book; 1993:11-28.
Braekken SK, Russell D, Brucher R, Svennevig J. Incidence and frequency of cerebral embolic signals in patients with a similar bileaflet mechanical heart valve. Stroke. 1995;26:1225-1230.
Siebler M, Nachtmann A, Sitzer M, Rose G, Kleinschmidt A, Rademacher J, Steinmetz H. Cerebral microembolism and the risk of ischemia in asymptomatic high grade internal carotid artery stenosis. Stroke. 1995;26:2184-2186.
Tegeler CH, Burke GL, Dalley GM, Stump DA. Carotid emboli predict poor outcome in stroke. Stroke. 1993;24:186. Abstract.
Petty GW, Brown RD, Sicks J, O’Fallon M, Whisnant JP. Survival and recurrence after first cerebral infarction in Rochester, Minnesota, 1975-1989. Cerebrovasc Dis. 1996;6(suppl 2):6-7. Abstract.
Koller RL. Recurrent embolic cerebral infarction and anticoagulation. Neurology. 1982;32:283-285.
Yasaka M, Yamaguchi T, Oita J, Sawada T, Shichiri M, Omae T. Clinical features of recurrent embolization in acute cardioembolic stroke. Stroke. 1993;24:1681-1685.
Fisher CM, Gore I, Okabe N, White PD. Atherosclerosis of the carotid and vertebral arteries: extracranial and intracranial. J Neuropathol Exp Neurol. 1965;24:455-476.
Castaigne P, Lhermitte F, Gautier JC. Role des lesions arterielles dans les accidents ischemiques cerebraux de l’atherosclerose. Rev Neurol. 1965;113:1-32.
Imparato AM, Riles TS, Gorstein F. The carotid bifurcation plaque: pathologic findings associated with cerebral ischemia. Stroke. 1979;10:238-245.
Sitzer M, Muller W, Siebler M, Horst W, Kniemeyer HW, Jancke L, Steinmetz H. Plaque ulceration and lumen thrombus are the main sources of cerebral microemboli in high-grade internal carotid artery stenosis. Stroke. 1995;26:1231-1233.
Nabavi DG, Georgiadis D, Mumme T, Schmid C, Mackay TG, Scheld HH, Ringelstein EB. Clinical relevance of intracranial microembolic signals in patients with left ventricular assist devices. Stroke. 1996;27:891-896.
Kaps M, Hansen J, Weiher M, Tiffert K, Kayser I, Droste DW. Clinically silent microemboli in patients with artificial prosthetic aortic valves are predominantly gaseous and not solid. Stroke. 1997;28:322-325.