Clinical Relevance of Detecting Asymptomatic Embolic Signals in Acute Stroke
To the Editor:
Kaposzta and colleagues successfully performed bilateral transcranial Doppler (TCD) examination and other measurements in 100 of 119 patients with acute anterior circulation infarction within 72 hours of stroke onset.1 Asymptomatic embolic signals (AES) were detected over the side of stroke in 16 patients; presence of AES was associated with significant carotid stenosis/occlusion, whereas absence of AES was associated with lacunar infarction in patients without significant carotid disease and potential cardiac embolic sources.1 I wish to make the following comments.
First, the authors reported that 19 of the 119 consecutive patients did not have an acoustic window for the bilateral TCD examination.1 Exclusion of other patients was not mentioned in the article. Did any patient/relative refuse to participate in the study? Was there any patient who could not tolerate the 1-hour TCD examination because of unstable conditions and/or restlessness?
Second, early use of aspirin has been shown to be beneficial in acute ischemic stroke,2 3 but antiplatelet agents and anticoagulants were withheld in the 30 patients who underwent the serial TCD examinations.1 Although there was no difference in the frequency of prior use of aspirin between patients with AES and those without AES, I am interested to know whether the dosage of aspirin is related to the presence of AES.
Third, it is important to note how the authors classified the stroke subtypes according to the presumed etiology. Ten of the 16 patients positive for AES were classified under the etiological category of significant carotid artery disease, and 2 of them had coexistent potential cardiac embolic sources but were not classified under the category of “coexistent causes.” In addition, patients were put under the category of “lacunar stroke” only when significant carotid artery disease and potential cardiac embolic sources were excluded. Thus, the classification of patients into the different etiological subtypes had been different from what was stated in the Methods section. More than 22 patients of the cohort had potential cardiac embolic sources, and more patients would be in the category of “coexistent causes.” It is interesting to note that patients with significant carotid artery disease can present with lacunar strokes and that carotid endarterectomy reduces the risk of recurrent stroke of both lacunar and nonlacunar types.4 5 On the other hand, it may be impossible to confirm the real cause of stroke in patients who have both significant carotid artery disease and potential cardiac embolic sources.6
Fourth, I found the results subsection entitled “Localization of Embolic Source” rather confusing. In my opinion, the results of the entire 90-minute period of AES detection should be presented together because localization of the embolic source would be interpreted on an individual basis according to the available information. In addition to a cardiac or aortic arch source of emboli, simultaneous AES over both the middle cerebral artery and the common carotid artery can also arise from the common carotid artery proximal to the site of ultrasonic monitoring.
Fifth, the stroke severity was omitted in the article. The information is of great relevance since the presence of AES was found to be related to a higher risk of death. It is possible that the presence of AES was associated with carotid occlusion and that carotid occlusion resulted in severe strokes with a high case fatality rate.
Finally, the authors’ results indicate that presence of AES is associated with carotid occlusion rather than potential cardiac embolic sources, but the clinical relevance of early TCD detection of AES is unclear. Only 1 of 16 patients with AES (6.3%) had a recurrent stroke while 12 patients without AES (14.3%) had recurrent strokes (10 patients) or transient ischemic attacks (2 patients), suggesting that AES are really asymptomatic events with little clinical consequence. There is no documented role of urgent carotid endarterectomy, and the benefit of carotid endarterectomy is unknown in patients with established carotid occlusion.7 Although long-term anticoagulation is indicated in stroke patients with potential cardiac embolic sources such as atrial fibrillation, the controversy lies in when to start anticoagulation.6 Thus, early extracranial carotid sonography for significant carotid artery disease and appropriate screening for potential cardiac embolic source are clinically relevant, and detection of AES by TCD may not alter the clinical management.
- Copyright © 2000 by American Heart Association
Kaposzta Z, Young E, Bath PM, Markus HS. Clinical application of asymptomatic embolic signal detection in acute stroke. Stroke.. 1999;30:1814–1818.
Barnett HJ, Taylor DW, Eliasziw M, Fox AJ, Ferguson GG, Haynes RB, Rankin RN, Clagett GP, Hachinski VC, Sackett DL, Thorpe KE, Meldrum HE, The North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. N Engl J Med.. 1998;339:1415–1425.
Cheung RT, Hachinski V. Cardiology. In: Samuels MA, ed. Hospitalist Neurology. Boston, Mass: Butterworth-Heinemann; 1999:305–330.
Moore WS, Barnett HJ, Beebe HG, Bernstein EF, Brener BJ, Brott T, Caplan LR, Day A, Goldstone J, Hobson RW, Kempczinski RF, Matchar DB, Mayberg MR, Nicolaides AN, Norris JW, Ricotta JJ, Robertson JT, Rutherford RB, Thomas D, Toole JF, Trout HH, Wiebers DO. Guidelines for carotid endarterectomy: a multidisciplinary consensus statement from the Ad Hoc Committee, American Heart Association. Stroke.. 1995;26:188–201.
We thank Dr Cheung for his interest in our article. In reply to his specific questions, first, no patients were unable to tolerate the 1-hour TCD examination. We found that a commercially available transducer fixation device (Spencer Technologies) was great assistance in probe fixation. Second, all patients who were taking aspirin had been treated with doses ranging from 75 to 300 mg per day. The number of patients who were embolic signal–positive means that the study is not sufficiently powered to examine relationships between aspirin dose prior to admission and rates of embolization. In addition, we didn’t feel that this was an issue of major interest. Third, Cheung is very critical of our stroke subtyping, but we feel he has not read our manuscript with sufficient care and has failed to distinguish between carotid stenosis and occlusion. Stroke subtyping was performed as described in our paper, with a diagnosis of carotid artery stroke being made if there was ≥50% stenosis in the symptomatic internal carotid artery territory. If there were other potential stroke causes in addition to the carotid stenosis, a diagnosis of coexistent stroke was made. There were 10 patients who had carotid artery stroke, but an additional 2 patients had carotid stenosis with other coexistent causes of stroke. These were tabulated in our Table 1 as coexistent. In addition to describing the group of patients with carotid artery disease, who could have either stenosis or occlusion, we specifically looked at patients with carotid occlusion to determine whether embolization distal to the stump could occur. This group of patients included the 2 with other potential stroke causes.
Fourth, we attempted to localize the embolic source in 2 ways. (1) In all patients we recorded simultaneously from both middle cerebral arteries. Bilateral embolic signals, or embolic signals on the asymptomatic side, are suggestive of a central or cardiac cause of embolism. (2) In patients in whom embolic signals were detected at the time of recording, we insonated the symptomatic MCA and the ipsilateral CCA simultaneously for an additional 30 minutes. We are unclear as to where his figure of 90 minutes comes from, but if Dr Cheung means that we should add the additional 30 minutes of middle cerebral artery recording to the 120 minutes already performed, we feel this would be inappropriate and make it more difficult to analyze the data. We can compare the detection of embolic signals among the group of 100 patients only if recordings of similar duration have been made throughout the group.
Finally, as stated in our article, “the number of strokes and deaths in our study was very small, making firm conclusions difficult to draw” on the relationship between stroke and death and asymptomatic embolic signals. Our data have suggested that this is an area worth looking at in more detail, but further larger studies would first need to be performed. Nevertheless, our results suggest this is an area worth investigating further. The analysis performed by Cheung on our data is misleading. While, indeed, only 1 of 16 patients with embolic signals did have a recurrent stroke, the 3 causes of death may have been related to further stroke. Postmortem examinations were not performed. Additionally, in his reanalysis he assumes that all emboli-negative patients who had recurrent stroke and/or death had recurrent stroke rather than death. This was not the case.
We would entirely agree that the use of asymptomatic emboli signal detection in this context is a research technique, and further studies are required to evaluate its clinical usefulness. These will be aided by technical advances in the technique, particularly analysis systems that will allow automated analysis of the large quantities of Doppler data produced.