Background and Purpose Transcranial Doppler sonography (TCD) has been used to detect microembolic signals in a variety of clinical situations. We studied the prevalence of TCD-detected microemboli in 38 acute stroke patients.
Methods Consecutive patients with acute anterior circulation stroke were stratified into high-risk (group 1), medium-risk (group 2), and low-risk (group 3) groups based on their risk factors for cerebral embolism.
Results Microemboli were detected in 11% of patients. They were present in 17% of group 1, 10% of group 2, and 0% of group 3 patients. Emboli were present in patients with mechanical prosthetic valves, carotid stenosis (>70%), and mitral valve strands with a patent foramen ovale. Patients with microemboli more frequently had a history of cerebral ischemia compared with patients without microemboli (P<.05). They also more frequently had recent (<3 months) symptoms compared with patients without microemboli (P<.05). In patients with a cardiac source of embolization, the number of microemboli detected was directly proportional to the acuity of previous symptoms.
Conclusions These data suggest that TCD-detected microemboli are associated with an increased prevalence of prior cerebrovascular ischemia. The presence of TCD-detected microemboli could be a risk factor for cerebrovascular ischemia.
TCD can detect cerebral microemboli. These microemboli were originally identified during monitoring of carotid1 and cardiac2 surgery. Subsequently, they have been detected in a variety of conditions including prosthetic valve replacement,3 4 5 high-grade carotid stenosis,6 7 8 AF,9 acute myocardial infarction,10 and ischemic stroke.11 12 Similar emboli have been described in experimental models of embolization.13 14 The significance of these microemboli is unclear. However, growing evidence suggests that they are associated with an increased risk of embolic stroke.8 11 15 Because cerebral embolism from arterial and cardiac sources may account for as many as 30% to 60%16 17 18 of all strokes, the detection of microemboli could be of great importance for the identification of patients who are at high risk for cerebral embolism.
The purpose of this study was to determine the prevalence of microemboli in acute stroke patients admitted to the Stanford Stroke Service. We hypothesized that microemboli would be more frequently detected in patients with known risk factors for cerebral embolism. Another goal of the study was to investigate whether there is a relationship between microemboli and a history of stroke or TIA. We suspected that patients with microemboli would be more likely to have a previous TIA or stroke compared with patients without microemboli and that the number of microemboli detected during the monitoring session would be greater in patients with recent neurological symptoms.
Subjects and Methods
This protocol was approved by the institutional Human Subjects Committee. Between July 1993 and January 1994, patients admitted to the Stanford Stroke Service with an acute ischemic stroke in the anterior circulation were prospectively screened as possible candidates for the study. Informed consent was obtained from all patients before enrollment. Data regarding the patients’ medical and neurological history were recorded. Results of neurodiagnostic tests performed during the hospitalization including CT, MRI/MR angiography, echocardiography (transesophageal or transthoracic), carotid ultrasonography, contrast angiography, and laboratory tests were recorded.
Patients were divided into three categories based on the presumed mechanism of stroke. Group 1 patients had an established high-risk carotid or cardiac source of cerebral embolism, group 2 patients had an embolic source of uncertain clinical significance, and group 3 patients had no known embolic risk factor. Patients with prosthetic heart valves, high-grade ipsilateral (>70%) carotid stenosis, acute myocardial infarction, atrial thrombus, significant congestive heart failure, and AF were included in group 1. Patients with a PFO, atrial septal aneurysm, mitral valve strands, chronic (>6 months old) left ventricular thrombus, and carotid occlusion were included in group 2. Patients with stroke and no known embolic risk were included in group 3. Patients with vasculitis were included in group 3 because the mechanism of infarction in these cases is presumed to be local thrombosis rather than embolism.
A TCD study was performed within 48 hours of admission. A TCD Cerebrovascular Diagnostic System low-profile monitoring system (Medasonics) was used for all investigations. Patients were placed in a supine or sitting position, and the middle cerebral artery/anterior cerebral artery bifurcation was insonated. The symptomatic hemisphere was insonated first. However, if an interpretable ultrasound signal was not obtainable, then the contralateral hemisphere was insonated (2 patients). The total duration of insonation was 30 minutes. The TCD probe was held in place with an elastic headband to reduce the possibility of movement artifact. If an ultrasound signal was unobtainable with headband monitoring, then hand-held monitoring was permitted (1 patient).
Microemboli were identified with use of the criteria described by Spencer19 as a guideline. These criteria include brief duration (<0.1 second), intensity greater than 3 dB above the background, and variable location in the TCD waveform. Emboli also have a distinctive “chirp” or “whistle,” which can be audibly discerned during monitoring. Signals were excluded if the patient moved or if there was any possibility of probe movement artifact. Signals were also excluded if there was not the combination of both the distinctive sound and visual signal simultaneously detected at the time of insonation. No automated detection programs were used. Emboli were recorded on computer disk for off-line analysis, and a hard copy was printed. All monitoring was done in real time to rule out signals caused by movement artifact. If there was any doubt about the validity of the signal detected, it was excluded from the emboli count. The observer was blinded to the results of clinical evaluation at the time of TCD monitoring. All monitoring was done by the same investigator (D.C.T.).
Fisher’s exact test was used to compare associations between categorical variables, and the Wilcoxon rank sum test was used to compare continuous variables between groups. The level of statistical significance was P<.05.
Thirty-eight consecutive patients were studied within 48 hours of acute stroke. Of these subjects, 22 were male and 16 were female. The prevalence of selected stroke risk factors is summarized in Table 1⇓. There was no significant relationship between these risk factors and the detection of microemboli. Eighteen patients were assigned to group 1, 10 patients to group 2, and 10 patients to group 3. The embolic risk factors that led to specific group assignments are summarized in Table 2⇓. All patients underwent MRI or CT scanning. Other diagnostic tests were performed at the discretion of the treating physicians. Twenty-nine patients had carotid ultrasound, and 15 underwent echocardiography.
Microemboli (Figure⇓) were detected in 4 of 38 patients (11%). Of these patients, 3 were in group 1 (3 of 18 or 17%), 1 in group 2 (1 of 10 or 10%), and none in group 3. Two of the emboli-positive patients had mechanical prosthetic valves. One patient had a high-grade carotid stenosis, and 1 patient had a small PFO associated with mitral valve strands. No other source of embolization was detected in any of these patients after appropriate diagnostic tests were performed.
None of the 13 patients with AF had microemboli, nor did the 2 patients with a chronic left ventricular thrombus. Three patients had prosthetic heart valves. Two patients had St Jude mechanical aortic valves; both had microemboli detected. One patient had a porcine aortic valve; no microemboli were detected in this patient.
No microemboli were detected in the 2 patients with carotid occlusions. Neither of these patients had a history of prior carotid occlusion. One of the 2 patients with significant (>70%) carotid stenosis had microemboli.
Overall, 47% of patients had a previous TIA (n=9) or stroke (n=9). The results are summarized in Table 3⇓. Among patients with emboli, 100% had a history of stroke (n=2) or TIA (n=2). In patients without emboli, 41% had a history of stroke or TIA. This difference was statistically significant (P<.05, Table 4⇓). Only 3% of the patients without microemboli had a history of TIA or stroke in the 3 months before admission, compared with 50% of the patients with microemboli. This was also statistically significant (P<.05, Table 4⇓).
The number of emboli detected per 30-minute insonation period varied from 1 to 20. The mean number of emboli was 8.5/30 min. There was a positive correlation between the proximity of previous cerebral ischemic events (stroke or TIA) and the embolism rate in the patients with cardiac abnormalities. Emboli-positive patients with a cardiac source of embolization and more recent cerebral ischemic symptoms had more microemboli. The results are summarized in Table 5⇓. The carotid stenosis patient with emboli had no history of prior symptoms.
Fourteen of 18 (78%) group 1 patients were receiving antithrombotic therapy at the time of TCD monitoring. One patient with a mechanical valve and microemboli was receiving both heparin and warfarin treatment (PTT, 48.7; INR, 4.4). The other patient with a mechanical valve and microemboli was receiving heparin (PTT, 105) at the time of study. The patient with a bioprosthetic valve and no microemboli was receiving heparin (PTT, 104) plus ASA (80 mg/d) therapy. Of the AF patients, 8 of 13 were receiving treatment (4 ASA, 2 warfarin, 1 heparin, and 1 heparin and ASA [80 mg/d]). Both heparin patients were therapeutically anticoagulated. One patient receiving warfarin therapy was subtherapeutic (INR, 1.4) and did not have microemboli. All patients taking ASA indicated compliance with their medication. Two of the paroxysmal AF patients were not in AF at the time of TCD. Of the 3 remaining untreated AF patients, none had emboli detected. All of these patients had a history of prior neurological symptoms. The patient with bilateral 90% carotid stenoses and microemboli was not on antithrombotic therapy at the time of TCD. The patient with an ulcerated 80% carotid stenosis was on heparin (PTT, 46) and did not have emboli.
In group 2, 3 of 10 patients (30%) were receiving antithrombotic therapy at the time of TCD examination. The patient with mitral valve strands, a PFO, and microemboli was taking ASA (325 mg/d). One of the patients with a chronic left ventricular thrombus was taking ASA (325 mg/d); the other was not receiving antithrombotic treatment. None of the patients with a PFO or old myocardial infarction were receiving antithrombotic treatment (n=4). One patient with carotid occlusion was receiving adequate warfarin therapy (INR, 2.0 to 3.0); the other was not on treatment.
In group 3, 6 of 10 (60%) patients were receiving treatment (4 ASA, 1 heparin, 1 warfarin). All were compliant and/or in the therapeutic range (PTT, 45 to 65; INR, 2.0 to 3.0). The dosage of ASA was 325 mg/d. No relationship between antithrombotic treatment and the presence of microemboli was found.
Microemboli were found in 11% (4/38) of patients admitted with acute anterior circulation stroke in this series. Emboli were seen in 2 of 5 patients with prosthetic valves, 1 of 2 patients with a high-grade carotid stenosis, and 1 patient with mitral valve strands associated with a small PFO. Patients with emboli had a significantly higher prevalence of prior cerebrovascular symptoms. There was also a relationship between the number of microemboli detected and the acuity of these prior symptoms. Emboli were seen despite therapeutic anticoagulant or antiplatelet therapy.
The prevalence of microemboli detected in this study is lower than in some previous reports. Grosset et al12 studied 41 patients with acute stroke using a similar protocol. Microemboli were detected in 70% of patients studied. All of the microemboli-positive patients had known carotid or cardiac disease. Tegeler et al11 reported microemboli in 16% of patients with recent stroke. Emboli were present in 30% (13/44) of AF patients and 25% (4/16) of patients with other cardioembolic sources. Using Doppler monitoring of the carotid artery, Tegeler et al9 also reported that 89% (8/9) of patients with AF and acute stroke had microemboli.
There are several possible explanations for the difference between our results and those previous studies. First, the study populations seem to be different. TCD-detected microemboli are particularly common in patients with mechanical prosthetic heart valves3 4 5 15 and carotid stenosis.6 7 8 The present study had relatively few patients with these findings. In the studies of Tegeler et al9 11 and Grosset et al,12 the prevalence of these findings was high. A large number of these high-risk lesions could explain the larger percentage of positive examinations in those investigations.
The effect of treatment with anticoagulant and antiplatelet agents must also be considered. In the previous studies cited, the use of antithrombotic therapy usually was not clearly specified. Although the effects of antithrombotic treatment on microemboli detection are unclear, such treatment could influence the number of positive studies. In group 1 of this study, 78% (14/18) of patients were receiving antithrombotic treatment at the time of TCD study. Two additional patients had paroxysmal AF and were in sinus rhythm at the time of TCD.
The criteria used for microemboli detection probably also influenced emboli detection results. In this study patients were required to have unequivocal microemboli to be included in the microemboli-positive group. Any unusual or atypical signals were excluded to reduce the possibility of false-positive results. Because the rate of microembolization can be as low as 1 embolus per 30 minutes or less, false-negative studies are a significant possibility. In previous studies, monitoring time has typically ranged from 15 to 60 minutes,3 4 5 8 10 12 15 although some patients with carotid stenosis have been monitored for several hours.6 7 Whether microembolism occurs at a constant rate or is intermittent remains to be determined. This could lead to considerable sampling error with 30-minute monitoring periods. Less stringent emboli detection criteria could also result in more positive studies. Moreover, in at least some prior studies, carotid monitoring rather than middle cerebral artery monitoring was used.8 10 This could result in a higher emboli detection rate because of the more proximal location of the probe to a carotid or cardioembolic source.
The detection of microemboli in a patient with mitral valve strands and a PFO has not previously been reported. The clinical significance of mitral strands is uncertain. They are believed to consist of fibrin, small thrombi, and/or connective tissue.20 Some investigators have found an increased incidence of stroke and TIA in these patients.21 22 PFOs are also known to be associated with stroke, particularly in young patients.23 24 25 26 However, a recent study suggests that the size and degree of shunting of the PFO may be most important for predicting embolic risk.27 Therefore, the small PFO in this patient was probably not significantly related to the patient’s neurological symptoms.
The significance of TCD-detected microemboli is unclear. Georgiadis et al3 compared the presence of microemboli and prevalence of stroke in 179 prosthetic valve patients and found no relationship between the presence or rate of microembolization and prevalence of stroke. However, the emboli-positive and emboli-negative groups were not directly compared with each other. Moreover, there were large differences in the rate of embolization between valve types, which could possibly have influenced the results.28
In contrast, Siebler et al7 noted that carotid stenosis patients with a higher embolization rate were more likely to have a history of recent stroke or TIA. A significant decrease in microembolism after carotid endarterectomy was also reported. Babikian et al8 also found a relationship between symptomatic carotid lesions and recent symptoms. Similarly, Tegeler et al11 found that patients with microemboli had a higher incidence of recurrent stroke, TIA, or death. In our previous study of patients referred for echocardiography, we also found a higher prevalence of prior TIA and stroke in patients with microemboli.15
This study supports a predictive role of microemboli monitoring in patients with cerebrovascular disease. Microemboli-positive patients more commonly had a history of prior stroke or TIA (P<.05). The patients with the most microemboli also had a history of more recent neurological symptoms (P<.05). We believe that in aggregate this evidence supports a role of TCD in identifying patients at higher risk for cerebrovascular ischemia. However, whether microemboli are an independent risk factor or whether the risk is solely related to the patients’ underlying medical condition cannot be definitively determined from these data. Further prospective studies with larger numbers of patients are needed to evaluate this question. Such studies will help to better define the role of TCD in the evaluation of patients at risk for cerebrovascular disease, potentially identifying those patients who would benefit the most from aggressive treatment.
Selected Abbreviations and Acronyms
|INR||=||international normalized ratio|
|PFO||=||patent foramen ovale|
|PTT||=||partial thromboplastin time|
|TCD||=||transcranial Doppler sonography|
|TIA||=||transient ischemic attack|
- Received April 3, 1995.
- Revision received May 22, 1995.
- Accepted May 22, 1995.
- Copyright © 1995 by American Heart Association
Spencer MP, Thomas GI, Nicholls SC, Sauvage LR. Detection of middle cerebral artery emboli during carotid endarterectomy using transcranial Doppler ultrasonography. Stroke. 1990;21:415-423.
Pugsley W, Klinger L, Paschalis C, Aspey B, Newman S, Harrison M, Treasure T. Microemboli and cerebral impairment during cardiac surgery. Vasc Surg. 1990;24:34-43.
Georgiadis D, Grosset D, Kelman A, Fairchney A, Lees K. Prevalence and characteristics of intracranial microemboli signals in patients with different types of prosthetic cardiac valves. Stroke. 1994;25:587-592.
Deklunder GM, Dauzat MM, Aldis A, Burte FM, Stewart SF, Houdas YJ. Microbubbles cerebral emboli detected by transcranial Doppler sonography in patients with prosthetic heart valves. Circulation. 1993;88(pt 2):I-223. Abstract.
Siebler M, Sitzer M, Steinmetz H. Detection of intracranial emboli in patients with symptomatic extracranial carotid stenosis. Stroke. 1992;23:1652-1654.
Siebler M, Sitzer M, Rose G, Bendfeldt D, Steinmetz H. Silent cerebral embolism caused by neurologically symptomatic high-grade carotid stenosis. Brain. 1993;116:1005-1015.
Babikian V, Hyde C, Pochay V, Winter M. Clinical correlates of high-intensity transient signals detected on transcranial Doppler sonography in patients with cerebrovascular disease. Stroke. 1994;25:1570-1573.
Tegeler CH, Hitchings LP, Eicke M, Leighton J, Fredericks RK, Downes TR, Stump DA, Burke GL. Microemboli detection in stroke associated with atrial fibrillation. J Cardiovasc Tech. 1990;9:283-284. Abstract.
Kitzman D, Tegeler C, Barber C, Dalley G, Downes T, Stewart K, Nomeir AM, Burke G. Detection of carotid microemboli following acute myocardial infarction. Circulation. 1993;88(pt 2):I-223. Abstract.
Tegeler CH, Hitchings LP, Eicke M, Leighton J, Fredericks RK, Downes TR, Stump DA, Burke GL. Carotid emboli predict poor outcome in stroke. Stroke. 1993;24:186. Abstract.
Grosset DG, Georgiadis D, Abdullah I, Bone I, Lees KR. Doppler emboli signals vary according to stroke subtype. Stroke. 1994;25:382-384.
Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke. 1991;22:253-258.
Markus H, Brown M. Differentiation between different pathological cerebral embolic materials using transcranial Doppler in an in vitro model. Stroke. 1993;24:1-5.
Tong D, Bolger A, Albers G. Incidence of transcranial Doppler–detected cerebral microemboli in patients referred for echocardiography. Stroke. 1994;25:2138-2141.
Wolf P, Cobb J, D’Agostino R. Epidemiology of stroke. In: Barnett H, Mohr J, Stein B, Yatsu F, eds. Stroke: Pathophysiology, Diagnosis and Management. 2nd ed. New York, NY: Churchill Livingstone Inc; 1992:4-6.
Mohr JP, Caplan LR, Melski JW, Goldstein RJ, Duncan GW, Kistler JP, Pessin MS, Bleich HL. The Harvard Cooperative Stroke Registry: a prospective registry. Neurology. 1978;28:754-762.
Spencer MP. Detection of cerebral arterial emboli. In: Newell DA, Aaslid R, eds. Transcranial Doppler. New York, NY: Raven Press Publishers; 1993:215-230.
Heath D, Best P, Davis B. Papilliferous tumors of the heart valves. Br Heart J. 1961:23;20-24.
Orsinelli DA, Pearson AC. Detection of prosthetic valve strands by transesophageal echocardiography: clinical significance in patients with suspected cardiac source of embolus. J Am Coll Cardiol. 1993;21:218A. Abstract.
DiTullio M, Sacco R, Gopal A, Mohr J, Homma S. Patent foramen ovale as a risk factor for cryptogenic stroke. Ann Intern Med. 1992;117:461-465.
Hausmann D, Mugge A, Nikutta P, Daniel WG. Patent foramen ovale: characteristics of the atrial septum associated with an increased risk for paradoxical embolism. J Am Coll Cardiol. 1993:21:263A. Abstract.
Tong D, Albers G. Microembolic signals in patients referred for echocardiography. Stroke. 1995;26:525-526. Letter.