Objectives: To evaluate the reduction of embolic signals after the initiation of an antithrombotic secondary prevention in patients with recent arterioembolic stroke and to determine the predictive value of decreased microembolism on the risk of early stroke recurrence.
Methods: Eighty six consecutive patients (55 men, 31 women; mean age 60.6 years) with a non-disabling arterioembolic ischaemic event in the anterior circulation within the last 30 days and a medium grade or high grade stenosis (≥50%) of the ipsilateral carotid or middle cerebral artery underwent 1 hour transcranial Doppler monitoring as part of the admission examinations. Antithrombotic secondary prevention was started after completion of admission. Patients in whom embolic signals were detected underwent a second monitoring within 4 days (mean time 1.8 days). All patients were followed up prospectively to evaluate the relation between presence and persistence of embolic signals and the risk of recurrent transient ischaemic attack (TIA) and stroke within the next 6 weeks.
Results: In 44 patients, embolic signals were detected at admission, a mean 5.4 days (range 0 to 21 days) after the initial event. Twenty five were positive for embolic signals also at the second monitoring, in 19 signals had ceased. Forty two patients without embolic signals at admission served as controls. During follow up, six ischaemic events (two stroke, three TIA, one amaurosis fugax) occurred in 25 patients with persisting embolic signals but none in 19 patients in whom signals had ceased by the second monitoring. One patient in the control group had a TIA. The incidence of a recurrent event was 0.45 per 30 patient-days if embolic signals persisted compared with 0.015 if signals could not be detected or had ceased. Persistence of embolic signals was an independent predictor of a recurrent TIA or stroke (adjusted odds ratio 37.0; 95% confidence interval (95% CI) 3.5 to 333; p<0.003). Cessation and decrease of embolic signals was associated with the administration of antiplatelet agents but not with anticoagulation with intravenous heparin (p<0.001).
Conclusions: Rapid cessation of embolic signals detected in patients with recently symptomatic arterial stenosis decreases increased risk of an early ischaemic recurrence. Effect of antithrombotic agents on embolic signals might serve as a marker for their efficacy on preventing stroke recurrence.
- carotid artery stenosis
- cerebral embolism
- transcranial Doppler sonography
- TCD, transcranial Doppler sonography
- MCA, middle cerebral artery
- ASA, acetylsalicylic acid
- ES, embolic signals
- TIA, transient ischaemic attack
Statistics from Altmetric.com
- TCD, transcranial Doppler sonography
- MCA, middle cerebral artery
- ASA, acetylsalicylic acid
- ES, embolic signals
- TIA, transient ischaemic attack
In patients with symptomatic arterial stenosis, stroke is most likely caused by emboli arising from the stenotic lesion. Subsequent antithrombotic therapy aims to prevent recurrent thromboembolism. However, even in the presence of an antithrombotic prevention, recurrence of an ischaemic event within the first month has been described in 10% to 15% of patients.1,2 Presence of microemboli as detected by transcranial Doppler sonography (TCD) after an initial event has been reported as an independent predictor of an early ischaemic recurrence with a ninefold increase of risk,2 and was associated with the absence of an antiplatelet medication.3 These findings suggest a common pathophysiology of clinically symptomatic embolism and clinically silent particulate microemboli, both probably reflecting a still active emboligenic process at the site of the stenotic lesion. This assumption is supported by a few case studies which followed up patients by repeated clinical and TCD examinations, in which the frequency of embolic signals reflected the course of recurrent ischaemic symptoms.4–7 We therefore hypothesised that embolic signals may be used as a marker of the efficacy of secondary prevention—that is, that a reduction of signals by an antithrombotic therapy is accompanied by a reduction of risk for an early stroke recurrence in patients with recently symptomatic arterial stenosis.
In the study presented here, we monitored the effect of an antithrombotic prevention on embolic signals in consecutive patients with a recent arterioembolic ischaemic event. We prospectively followed up these patients to determine the predictive value of a decrease of signals on the risk of stroke recurrence.
SUBJECTS AND METHODS
Eighty six patients (55 men, 31 women; mean age 60.6 (SD (13)) years) with a recent (≤30 days) arterioembolic stroke in the anterior circulation who underwent a 1 hour TCD monitoring at admission were eligible for the study. Fifteen had had amaurosis fugax, 34 a transient ischaemic attack, and 37 a minor stroke a mean 7.3 days (range 0 to 30 days) earlier. Eligible patients were selected from a series of 238 consecutive cases presented at the Department of Neurology, Magdeburg University, with a non-disabling ischaemic event in the anterior circulation within the past 30 days, a medium grade or high grade stenosis of the ipsilateral carotid or middle cerebral artery (MCA), and without findings associated with cardioembolic stroke—for example, atrial fibrillation, cardiomyopathy, acute myocardial infarction, and prosthetic valve. Patients were not included if 1 hour TCD monitoring could not be performed at admission—that is, insufficient temporal bone windows not allowing transcranial insonation (n=50), patients rejecting the examination (n=7), when the device was not available (n=11), or when outpatient management did not allow time consuming monitoring (n=42). Six patients were excluded because exhausted Doppler carbon dioxide reactivity suggested haemodynamic rather than embolic stroke pathology.8,9 Of the remaining 122 patients with probable arterioembolic stroke and 1 hour TCD monitoring at admission, 36 were excluded from the investigation because antithrombotic prevention had been initiated or modified (by a pre-existing medication) before the initial monitoring. This might have influenced the spontaneous rate of embolic signals so that they no longer reflected the emboligenic status of the stenosis at the time of the ischaemic event.3 Overall, included patients were younger (mean age 61 v 67 years), less often diabetic (24 v 40%), under antiplatelet prevention (47 v 65%), and more often had transient (57 v 42%) and recurrent (42 v 17%) ischaemic events than excluded patients. In addition, patients excluded because of transtemporal bone window failure more often were women (56 v 36%) and had medium grade stenosis (70 v 48%), and patients in whom monitoring could not be arranged (in the majority presented from regional physicians and hospitals for ambulant reference sonography) were examined later after the ischaemic event (median, 7.7 v 3.8 days) than selected ones.
Selected patients underwent a standardised admission and follow up procedure. At admission a detailed medical history including cardiovascular risk factors was taken from all patients. Neurological deficit was quantified according to the NIHSS score and was repeated after 24 hours and 4 days. Extracranial and intracranial vascular pathology was assessed by colour coded duplex sonography (SSH 380, Toshiba). Degree of extracranial carotid stenosis was classified as medium grade (≥50% local diameter reduction) at angle corrected peak systolic velocities >120 cm/s, and as high grade (≥80% local diameter reduction) at peak systolic velocities >300 cm/s or end diastolic velocities >135 cm/s.10 Cut off values of angle corrected peak systolic velocities for the diagnosis of MCA stenosis were 155 cm/s (<50%) and 220 cm/s (≥50%), respectively.11 Cerebrovascular reactivity was determined at the middle cerebral artery distal to the symptomatic stenosis by inhalation of a mixture of 5% carbon dioxide and 95% oxygen and subsequent hyperventilation to induce hypercapnia and hypocapnia. Brain CT or MRI was performed in all patients and was repeated after 24 to 48 hours if hemispheric signs persisted for more than 24 hours and initial scan failed to demonstrate a corresponding infarction. Holter monitoring and transoesophageal echocardiography (49 patients), both performed subsequently after admission according to clinical decision criteria, as well as a 12 lead ECG at admission showed no findings suggestive of cardioembolic stroke.
Also as part of the admission examinations, 1 hour bilateral simultaneous MCA monitoring with dual gated 2 MHz pulsed wave probes (Multi-Dop X, DWL) was performed after informed consent. Periods of interest for embolic signals were assessed automatically, registered on line by the software (detection threshold of the relative intensity increase ≥12 dB, calculated propagation distance 0.5 to 10 mm at a gate distance of 5 mm), and stored on hard disk. Subsequent off line review by two independent observers blinded for patients and monitoring data included only events with unidirectional signal within the Doppler velocity spectrum (interobserver agreement >99%). Estimated sensitivity for artifact identification evaluated on ∼300 produced artifacts in control subjects was >98%,12 estimated sensitivity for the detection of an embolic signal evaluated by reference DAT tape within the scope of a between centre comparison was ∼95%. Patients in whom embolic signals were detected distal to the symptomatic stenosis at the first monitoring underwent a second examination 1.8 days (mean) after secondary prevention had been initiated or modified. Device setting of the initial monitoring was maintained at follow up recording.
Antithrombotic prevention was decided on by the treating clinician without knowledge of the monitoring results and was started after the admission investigations. Sixty four of the 86 patients received an antiplatelet agent. In 31 without antithrombotic premedication, secondary prevention was started with acetylsalicylic acid (ASA) (500 mg, continued at 300 mg/day; n=24), or, in cases of known gastric ulcer or allergy to ASA, ticlopidine (500 mg/day; n=3) or clopidogrel (75 mg/day; n=4). The last was used since it became available instead of ticlopidine and was started without a loading dose. In 33 patients with a pre-existing ASA medication (100 mg/day) already before ischaemia, ASA dosage was increased (500 mg, continued by 300 mg/day; n=23) or replaced/ supplemented by ticlopidine (n=3/ n=1) or clopidogrel (n=1/ n=5). Nineteen of the 86 patients initially received heparin, which increased mean activated partial thromboplastin time to 2.4 ((SD) 0.6) times the normal value. Twelve of these patients had no antithrombotic premedication. In seven patients who at the time of their ischaemic event already had received ASA (100 mg/day), this medication was stopped before anticoagulation in four but was continued in three
Recurrence of an ischaemic event during follow up was defined as the occurrence of neurological symptoms or clinical signs corresponding to the distribution area of the initially symptomatic stenosis that appeared after the initial deficit had completely resolved or a sudden deterioration of the initial or occurrence of new focal symptoms. All patients with a recurrent event subsequently underwent an additional CT or MRI, extracranial and intracranial duplex sonography, ECG, and routine laboratory evaluation, which were all required to be compatible with the diagnosis.
For the investigation presented here, follow up was terminated 6 weeks after admission. Outpatients without early endarterectomy or recurrent stroke are routinely re-evaluated at this time including recent medical history, neurological examination, and duplex sonography. The end point of follow up for stroke free survival analysis in the 86 patients was a recurrent ischemia in seven, carotid endarterectomy in 27, a change of antithrombotic medication in 10, and censored follow up in 39. Three patients were lost to follow up.
All findings were documented in the local stroke data base, which covers all inpatients and outpatients with cerebrovascular diseases presented at the Department of Neurology. Results of TCD monitorings were stored separately and not referred to inpatients' medical records chart or disclosed to clinicians involved in the treatment of the patient.
Statistical analysis was performed using SPSS for Windows, Version 10.0. Frequency of clinical characteristics, embolic signals, and time periods were compared by χ2 test, Fisher's exact test, and non-parametric tests for independent and related samples as appropriate. The relation between and adjusted relative risk of persistence of embolic signals and early ischaemic recurrence was analysed by Kaplan-Meier curves and estimated by Cox regression analysis. Significance was set at a p value of less than 0.05.
At admission, embolic signals distal to the symptomatic stenosis were detected in 44 of the 86 patients. After initiation of an antithrombotic treatment, embolic signals persisted in 25 patients and had ceased in 19 by follow up monitoring. Patient groups did not differ by their baseline clinical characteristics, initial number of embolic signals, time between ischaemic event, antithrombotic prevention, and TCD monitoring (table 1). For calculation of recurrent stroke risk in relation to cessation or persistence of embolic signals, the 42 patients without signals at the initial monitoring served as controls. Mean time of follow up was 15.5 days in patients with persisting embolic signals, 35.2 in those in whom signals had ceased, and 32.2 in the control group. Six recurrent ischaemic events (two stroke, three TIA, one amaurosis fugax) occurred in the 25 patients with persisting embolic signals, all within 3 to 7 days after initiation of antithrombotic prevention (3 to 10 days after the initial ischaemic event). No event was seen in the 19 patients in whom signals had ceased. One patient of the control group had a recurrent TIA 39.1 days after the start of ASA (39.5 days after the initial event), 1 day before endarterectomy was scheduled. The persistence of embolic signals during secondary prevention was associated with an early recurrence of a TIA or stroke whereas their cessation within 4 days (mean 1.8 days) after the start of an antithrombotic therapy decreased this risk to a level as low as in patients without detectable signals at admission (p=0.0002, Kaplan-Meier survival analysis; fig 1). The incidence of a recurrent event was 0.45 per 30 patient-days if embolic signals persisted compared with 0.015 if signals could not be detected or had ceased. Persisting embolic signals remained a significant predictor of a recurrent event also after controlling for potential covariates (adjusted odds ratio 37.0; 95% confidence interval (95% CI) 3.5 to 333; p<0.003) (table 2) with increasing risk with increasing signal frequency (adjusted odds ratio 1.2; 95% CI 1.1 to 1.4; p<0.006, fig 2).
Ten of 25 patients with persisting embolic signals and all 19 patients without signals at the second monitoring had received antiplatelet agents for secondary prevention, whereas heparin was administered to 15 with persisting and none without persisting signals (p<0.001). In the six patients with recurrent events and persisting embolic signals, anticoagulation was performed in four (added on ASA in one) and was sufficient in three (activated partial thromboplastin time 1.8 to 2.0 of normal baseline values) but insufficient in one (1.1) at the time of the event. One patient had a TIA 4.1 days after the start of an ASA prevention, one an amaurosis fugax 4.9 days after clopidogrel.
The frequency and number of embolic signals before and after the start of a secondary prevention in relation to administered antithrombotic medication—that is, antiplatelet agents or intravenous heparin—are listed in table 3. In all 44 patients as well as in the subgroup of 33 without pre-existing antithrombotic medication, embolic signals had ceased in most patients under antiplatelet agents but persisted in all of those who were anticoagulated with heparin (p<0.001). In patients with persisting embolic signals, their number had significantly decreased after antiplatelet agents but not after heparin. There was a trend to a higher rate of patients with cessation of embolic signals after the administration of ticlopidine/clopidogrel added on pre-existing ASA compared with ASA or ticlopidine/clopidogrel alone (p=0.068).
Our study provides evidence that a cessation of embolic signals found within 4 days after the start of an antithrombotic prevention, is associated with a risk reduction of a recurrent arterioembolic TIA or stroke within the next 6 weeks to a level as low as in patients without embolic signals after an initial recent ischaemic event. Patients in whom embolic signals persisted remained at high risk, which according to our investigation is estimated at almost 40 times the risk of patients in whom embolic signals could not be detected or had ceased. Embolic signals have been found to be a predictor of an early ischaemic recurrence in recent prospective studies on patients with symptomatic arterial stenosis with an adjusted relative risk about 10 times that of signal negative patients.1,2 By contrast with our study, these included patients with a wider time window since the index event,1 and with arterial lesions also in localisations other than the carotid or middle cerebral artery,2 and did not differentiate whether an antithrombotic medication had been altered after the TCD monitoring during follow up, probably not allowing direct comparison of risk estimations. However, as both studies only performed a single monitoring they did not yield data on the effect of a decrease of embolic signals on TIA and stroke risk.
Decrease and cessation of embolic signals in our patients were most likely caused by the administered antithrombotic agents. In consideration of the unknown spontaneous course of embolic signals in an individual patient, it has to be stressed that time intervals between both monitorings were short and did not differ in patients with and without embolic signals at the second monitoring. Further variables known or assumed to influence the prevalence of embolic signals—for example, degree and localisation of the stenosis, type of the initial ischaemic event, time since this event, and multiple recurring initial events3,13–15—also did not differ between the groups. Therefore, even considering that our patients were not randomised to the type of the antithrombotic prevention, our data strongly suggest that embolic signals are inhibited by antiplatelet agents rather than intravenous heparin. A failure of anticoagulation as well as the potential of antiplatelet agents to decrease embolic signals of arterial origin already has been suggested by previous case reports. In two patients with ongoing cerebral microembolism (and ischaemic events) during anticoagulation, replacement of warfarin/heparin by ASA/ticlopidine stopped both embolic signals and stroke recurrence.6,16 Embolic signals detected in three patients with recently symptomatic atherosclerotic carotid stenosis persisted (as recurrent ischaemic events did in two) despite sufficient anticoagulation with intravenous heparin.5,17 On the other hand, embolic signals significantly dropped in seven of nine patients (with four patients becoming signal negative) within 2 days after the start of a secondary prevention with ASA.12
With respect to selection criteria, sample size, and timing of the examinations, some reservations need to be made about the validity of our results. Differences between selected patients and those excluded relating to the frequency of multiple initial events and the presence of pre-existing antiplatelet medication might have biased selection to patients positive for embolic signals.3 However, as the primary intention of the study was to monitor the effect of an antithrombotic prevention on embolic signals and their relation to the risk of stroke recurrence, exclusion of patients with insufficient transtemporal bone window, other obstacles to TCD monitoring, and not determinable effects of recently started antithrombotic prevention on embolic signals was mandatory. As baseline characteristics did not differ between patients in whom embolic signals persisted and those signals had ceased (table 1) and Cox regression was used to correct prediction of ischaemic recurrence for baseline characteristics (table 2) this is not expected to alter the presented results. Although our study is one of the largest to have followed up patients with embolic signals after an ischaemic cerebrovascular event, the confidence interval of risk estimation is wide for the risk calculated for TIA and stroke. Confirmation of our results by a randomised to treatment study with higher patient numbers is required. The time of the second TCD monitoring was not prefixed in relation to the initial one or the start of antithrombotic prevention. This might be of minor relevance in patients who had received intravenous heparin and in whom monitoring was repeated when therapeutic anticoagulation has been reached as documented by an aPTT within the target range. In patients on antiplatelet agents, time of repeat TCD was chosen more arbitrarily. In previous follow up studies, reischaemia occurred within 2 to 25 days after the detection of embolic signals,1,2 and the maximal effect of ASA on embolic signals in responding patients has been shown to occur after 1–2 days.12 Therefore, the 1−2 days was considered as an adequate target time period. At least in patients who had received ticlopidine or clopidogrel the effect on embolic signals still might be incomplete at this time.18 However, with respect to the early recurrence of symptoms in these patients an early effect of antithrombotic agents seems clinically relevant.
In summary, in patients in whom cerebral embolic signals distal to a recently symptomatic stenosis indicate an increased risk of recurrent TIA and stroke, cessation of these signals after the start of an antithrombotic prevention is associated with a substantial risk reduction. Early decrease of embolic signals in these patients with atherosclerotic arterial disease is seen after antiplatelet agents but not with intravenous heparin.