(Stroke. 1999;30:1290.)
© 1999 American Heart Association, Inc.
Letters to the Editor |
Early Ischemic Recurrence and Microembolic Signals Detected by Transcranial Doppler
Division of Neurology, Department of Medicine, Queen Mary Hospital, Hong Kong
To the Editor:
I read with great interest the article, "Microembolic Signals and Risk of Early Recurrence in Patients With Stroke or Transient Ischemic Attack," by Valton and colleagues.1 In this study, an early, random, 20-minute sampling of microembolic signals (MES) by use of transcranial Doppler (TCD) over the middle cerebral artery on the side of recent stroke or transient ischemic attack (TIA) in a selected cohort of 73 patients was found to be an independent risk factor for early ischemic recurrence (EIR). In the article, EIR was arbitrarily defined as (1) recurrent TIA or stroke in patients presenting with TIA or (2) clinical deterioration not due to hemorrhagic transformation and cerebral edema in those presenting with stroke.1 Their findings and interpretations raised 2 interesting points. First, their data indicated that the prediction of EIR by detecting MES on TCD has a sensitivity of 62.5% (5/8), a specificity of 84.6% (55/65), a positive predictive value of 33.3% (5/15), and a negative predictive value of 94.8% (55/58). Thus, absence of MES on TCD is reassuring in an individual patient, but presence of MES does not have a high predictive value for EIR. Second, the protective effect of antiplatelet therapy on the risk of EIR in the Cox model suggests that early use of antiplatelet therapy may lessen the risk of EIR and perhaps reduce the need of performing TCD to detect MES. In this regard, I fully agree with Valton and colleagues that stroke patients are often restless and uncooperative in the acute stage, making prolonged TCD monitoring technically difficult.
Before we consider routine screening of MES using TCD in stroke patients who are similar to the cohort of the present study, several methodological issues are worth remembering. First, the timing (2±2 days, mean±SD) and duration (20 minutes) of TCD monitoring raised a serious concern of sampling error.1 Of the 7 patients who had MES on the initial monitoring, only 4 were found to have MES on repeat TCD monitoring. The practical issue is, therefore, how soon after cerebral ischemia we should arrange TCD monitoring. In addition, the 20-minute period of sampling is short, because presence or absence of MES was used in the study to separate the cohort into 2 groups and the number of MES reflects the intensity of arterial thromboembolism.2 I would like to know the mean, median, and range of the number of MES in the 15 patients, the reason that bilateral TCD was performed in a subgroup of 33 patients, the reason for selecting a subgroup of 26 patients for repeat TCD monitoring, and the time interval between the TCD detection of MES and the occurrence of EIR. Specifically, I wonder whether a large number of MES is predictive of an imminent risk of EIR.
Second, EIR consisted of 2 strokes and 6 TIAs. Apparently, the TIAs occurred in 6 patients presenting with TIA, and the strokes affected 2 patients presenting with stroke.1 Initial deterioration following stroke is not uncommon and can result from many causes other than EIR.3 Diagnosing EIR in stroke patients simply by excluding hemorrhagic transformation and cerebral edema on CT scan is insufficient. I wonder whether the findings and conclusions would be the same when only the 23 patients presenting with TIA are considered and whether detecting MES on TCD is relevant only for recurrent TIA. Furthermore, I am interested in knowing how many of the 50 patients presenting with stroke had initial deterioration due to hemorrhagic transformation and cerebral edema on CT scan.
Third, the importance of carotid plague ulceration is equal to or greater than that of the degree of stenosis in the association with MES.4 5 I wonder whether carotid plaque ulceration is better than MES in predicting EIR in the cohort.
Fourth, the interobserver agreement for diagnosing MES in the study was excellent, but the recordings of only 32 patients were analyzed by a second independent observer.1 My concern is how the selection was made and why the recordings of all patients were not independently analyzed.
References
-
Valton L, Larrue V, le Traon AP, Massabuau
P, Géraud G. Microembolic signals and risk of
early recurrence in patients with stroke or transient
ischemic attack. Stroke.. 1998;29:2125–2128.
[Abstract/Free Full Text] -
Georgiadis D, Lindner A, Manz M, Sonntag M, Zunker P,
Zerkowski HR, Borggrefe M. Intracranial microembolic
signals in 500 patients with potential cardiac or carotid embolic
source and in normal controls. Stroke.. 1977;28:1203–1207.
[Abstract/Free Full Text] - Oppenheimer SM, Hachinski V. Complications of acute stroke. Lancet.. 1992;339:721–724.[Medline] [Order article via Infotrieve]
-
Valton L, Larrue V, Arrué P, Géraud G,
Bès A. Asymptomatic cerebral embolic signals in
patients with carotid stenosis: correlation with appearance of
plaque ulceration on angiography. Stroke.. 1995;26:813–815.
[Abstract/Free Full Text] -
Sitzer M, Müller W, Siebler M, Hort W, Kniemeyer
H, Jäncke 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.
[Abstract/Free Full Text]
Response
Department of Neurology, Rangueil Hospital, University of Toulouse, Toulouse, France
Key Words: cerebral embolism • ultrasonography • Doppler
We thank Dr Cheung for his interest in our study. We admit that earlier and longer-duration monitorings might have detected more patients with MES. The mean, median, and range values of MES were 2.4, 1, and 1 to 7, respectively. Several factors limited the number of bilateral and repeat transcranial Doppler scannings: poor patient cooperation, lack of a bone window, and short duration of stay in our department. The mean±SD delay between detection of MES and EIR was 4±5 days. Four of the 5 patients with MES who had EIR had only 1 MES. Therefore, our data do not permit exploration of the relationship between the number of MES and the delay from monitoring to recurrence.
We are confident in our diagnosis of EIR. Six patients who were included after TIA (3 patients) or reversible ischemic neurological deficit (3 patients) were asymptomatic when EIR occurred. A patient had a stroke in the anterior cerebral artery territory after a stroke in the middle cerebral artery territory. Another patient had sudden and transient worsening of his hemiparesis. A control CT scan excluded hemorrhagic transformation and cerebral edema in all cases, and general causes of deterioration were also carefully excluded. Subgroup analysis of patients who presented with TIA is not feasible because only 3 such patients had EIR. Of the 50 patients who presented with stroke, 5 deteriorated because of cerebral edema. There were no symptomatic hemorrhagic transformations.
We have previously reported1 that microembolism is associated with angiographic evidence of carotid plaque ulceration. Therefore, it is rational to speculate that plaque ulceration could also be associated with increased risk of early recurrence. However, our data do not permit us to address this issue because only 32 patients had selective carotid angiography.
References
- Valton L, Larrue V, Arrué P, Géraud G, Bès A. Asymptomatic cerebral embolic signals in patients with carotid stenosis: correlation with appearance of plaque ulceration on angiography. Stroke.. 1995;26:813–815.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||