Occurrence of Transcranial Doppler High-Intensity Transient Signals in Patients With Potential Cardiac Sources of Embolism
A Prospective Study
Background and Purpose Cerebral emboli can be recognized by typical “high-intensity transient signals” (HITS) in the transcranial Doppler (TCD) spectral curves. Patients with potential cardiac sources of embolism are at higher risk for stroke.
Methods We examined the frequency of HITS in the left middle cerebral artery (MCA) with TCD over periods of 30 minutes in 100 patients having potential cardiac sources of embolism, as indicated by transthoracic or transesophageal echocardiography.
Results Thirty-six (36%) of the patients presented with HITS. Sex, age, sufficient anticoagulation level, antiplatelet therapy, neurological symptoms, and a history of thrombosis had no influence on the prevalence and number of HITS. The patients with a single echocardiographic diagnosis were separated into eight echocardiographically defined groups: patients with (1) atrial fibrillation, (2) coronary artery disease plus ejection fraction of more than 30% including at least three wall segments of hypokinesia/akinesia, (3) coronary artery disease with less than 30% ejection fraction, (4) dilated cardiomyopathy, (5) infectious endocarditis, (6) aortic stenosis, (7) mitral stenosis, and (8) patent foramen ovale. A significant difference in HITS occurrence could not be found in any of the defined groups. Only patients with infectious endocarditis showed a tendency for a higher HITS prevalence.
Conclusions HITS are common phenomena in patients with potential cardiac sources of embolism. The clinical relevance of these HITS remains unclear.
Following the first description of “high-intensity transient signals” (HITS) by Spencer and colleagues1 in the late 1960s, several authors have reported the occurrence of HITS in different patient groups. A high incidence of these stereotyped transcranial Doppler (TCD) signals has been documented during cardiopulmonary bypass,2 during carotid endarterectomy,3 4 in symptomatic and asymptomatic carotid artery stenosis,5 6 7 in symptomatic vertebral artery stenosis,8 in patients with atrial fibrillation,9 and in patients with heart valve prostheses.10 11 12 13
Cardioembolic ischemic strokes are one of the major complications in patients with potential sources of embolism. Of ischemic strokes, 15% to 30% are of cardiac origin.14 Thus, a predictor for the potential risk for cerebral embolism would be of great value in stroke prevention, especially in patients with potential cardiac sources of embolism.
This study was designed to evaluate the prevalence of HITS in patients defined as having different cardiac sources of embolism estimated either by transesophageal or transthoracic echocardiography. Another objective was to identify subgroups of heart disease that may have an increased risk of HITS occurrence. Last, the data were correlated with hematologic parameters and neurological events. Taken together, the data could help to define the usefulness of TCD monitoring and HITS detection in the management of patients at high risk for cardioembolic stroke.
Subjects and Methods
During a period of 6 months, 109 consecutive patients with atrial fibrillation or other potential cardiac sources of embolism estimated either by transthoracic or transesophageal echocardiography were studied. Two-dimensional echocardiography was performed in all patients with either a 5.0-MHz multiplane probe for transesophageal echocardiography or a 3.5-MHz probe for transthoracic echocardiography. All examinations were performed using a Hewlett-Packard Sonos 1500.
Nine of the 109 patients had an insufficient transtemporal bone window and were excluded from the study. The remaining 100 patients were tested for neurological deficits defined as stroke, sufficient levels of anticoagulation, a history of thrombosis (deep vein thrombosis or a history of proven embolic event to the lung), and HITS occurrence (Tables 1⇓ and 2⇓). To identify subpopulations having a higher prevalence of HITS, separate analysis was performed only in those patients with a single clearly defined echocardiographic diagnosis. Sixty-six patients had a single diagnosis and were easily placed into one of the following defined groups, whereas 34 of the 100 patients had multiple diagnoses and did not fit into one of the eight groups: (1) lone or nonvalvular atrial fibrillation, 18 patients; (2) infectious endocarditis, 5 patients; (3) dilated cardiomyopathy with an ejection fraction <30%, 5 patients; (4) coronary artery disease plus ejection fraction >30% with at least three segments of hypokinesia/akinesia in a 16-segment model of the left ventricle, 15 patients; (5) coronary artery disease plus ejection fraction <30%, 6 patients; (6) aortic stenosis, 8 patients; (7) mitral stenosis, 4 patients; and (8) patent foramen ovale, 5 patients.
The patients were placed into the above-mentioned eight different groups according to echocardiographic criteria that were clearly defined. The subgroup of 66 patients with clearly defined single diagnoses was analyzed separately (Table 3⇓). The other 34 individuals with more than one type of cardiac disease were excluded from this separate analysis. A further classification of these patients was not done.
All patients were examined for neurological deficits. The left middle cerebral artery (MCA) was monitored over 30 minutes by TCD in 15 control subjects without cerebrovascular or cardiac disease.
TCD monitoring was performed with a DWL Multidop X with a 2-MHz probe. The left MCA was insonated at a depth between 50 and 55 mm according to standard criteria15 for a period of 30 minutes. The probe was fixed on the head with a specially designed spectacle frame. This guaranteed a minimum of artifact disturbances and a constant angle of insonation. For demonstration of HITS, all of the following criteria were required: (1) random occurrence, (2) brief duration (<0.1 second), (3) high intensity (minimum 3 dB above background intensity), (4) primarily unidirectional quality within the Doppler spectrum, (5) causing a spike in the power/intensity trace, and (6) accompanied by an audible “chirp” or “pop.”16 17 All HITS were recorded on hard disk. The patients were examined in the supine position by the same investigator. The investigator did not know the cardiological status of the patient, the diagnosis, or whether the subject belonged to the control population.
Extracranial carotid artery disease was ruled out by continuous-wave Doppler sonography. A nonexisting bone window for the TCD examination was a criterion for exclusion.
Parametric data were expressed as mean±SD. Pearson correlation coefficients were calculated between quantitative data. The significance of frequency distributions within fourfold tables was calculated by Fisher’s exact test.
In the 15 control subjects (median age, 52 years; range, 23 to 75 years) without cerebrovascular or cardiac disease, no HITS were detected during the 30 minutes of TCD monitoring of the left MCA. In other words, in a total TCD investigation time of 7.5 hours, no HITS occurred.
Sixty-eight of the patients were men. The average age for both sexes was 56.0±16.5 years. The numbers of patients presenting with a sufficient level of anticoagulation, with neurological symptoms defined as stroke, or with a history of thrombosis defined as deep vein thrombosis or a history of proven embolic event to the lung are summarized in Table 2⇑.
HITS were detected in 36 of the 100 patients. The average HITS frequency of these 36 patients was 2.69±2.7 per 30 minutes (range, 1 to 12). Twenty-five of the 66 patients with a single diagnosis had HITS (38%), whereas 11 of the 34 patients with multiple diagnoses had HITS (32%). These figures are calculated from the data in Tables 1⇑ and 3⇑.
There were no statistically significant correlations between the presence of HITS and age, sex, anticoagulation, neurological symptoms, or thrombotic events (Table 1⇑).
In patients positive for HITS, there was no significant correlation of the number of HITS with age, sex, coagulation parameters, or history of thrombotic events.
Table 3⇑ shows the statistical analysis for the subgroup of 66 patients with single echocardiographic diagnoses. The patients of each group were compared with those in groups 1 through 8.
A significant correlation between HITS occurrence and specific echocardiographic diagnosis could not be found for any of the eight subgroups mentioned in Table 3⇑. However, patients with infectious endocarditis showed a nonsignificant (P=.064) trend for a higher HITS prevalence.
The probability that HITS would occur in the patients investigated by our group is suggested by the data of others, who reported HITS in patients with atrial fibrillation.9 The present study is the first echocardiographic controlled investigation to search for HITS in patients with different potential cardiac sources of embolism.
With continuous-wave Doppler sonography, we excluded patients with middle or high-grade stenosis of the carotid artery. Some authors report HITS in patients with high-grade stenosis of the internal carotid artery.5 6 7 In patients with nonstenotic internal carotid artery lesions, HITS could not be detected by others, which is in agreement with our results.24 In addition, in patients with these types of minimal lesions of the carotid artery, which are not detectable by continuous-wave Doppler sonography, the risk of artery-to-artery stroke due to carotid artery disease is very low. The data from the above-mentioned studies, the clinical observations, and the fact that we ruled out every stenosis of the internal carotid artery ≥60% are strong arguments for the assumption that the HITS detected by unilateral TCD monitoring in our study are from cardiac and not carotid origin.
However, since equipment for bilateral monitoring of the MCA is available, searching for HITS simultaneously in both MCA will be the method of choice in the future. It is obvious that the total number of HITS detected during bilateral monitoring will be approximately twice that observed during unilateral monitoring.
We demonstrated that HITS detected by TCD are not uncommon in a large population of patients with different potential cardiac sources of embolism, as indicated by echocardiography. Although the majority of our patients were receiving antiplatelet therapy or therapeutic anticoagulation, HITS were still detected in 50% of the patients with therapeutic levels of anticoagulation (Table 1⇑). This lack of correlation between anticoagulation therapy and the occurrence of HITS is in accordance with the results of others25 and suggests that the underlying embolic material in the patient group undergoing anticoagulation therapy is not thrombotic. These data suggest that HITS in this patient group represent particles that do not respond to anticoagulation treatment with vitamin K antagonists. As yet, the exact type and size of the embolic material causing HITS cannot be clearly defined.
One of the major questions was whether the occurrence of HITS is a direct marker for stroke risk. Although it is clear that our population had a higher risk for cardioembolic strokes, we were unable to find a correlation between the occurrence of HITS and their neurological parameters. Although 35% of the patients with a focal neurological deficit in their history presented with HITS, HITS were detected also in 36% of patients without a previous neurological event. This lack of correlation between the occurrence of HITS and neurological deficits has been described previously.11 Nevertheless, some studies have described positive and significant correlations between HITS and the occurrence of neurological symptoms.13 26 The different results regarding the clinical significance of HITS in the published studies may be due to the problem of limited monitoring time and the small numbers of patients studied. The cumulative effect of HITS may be clinically important. Over time, such microscopic embolization could conceivably cause diffuse small-vessel ischemia or could compromise the ability of microcirculatory collateral blood supply to withstand vascular insults. The occurrence of minor brain damage has to be investigated in further studies.
Another question was whether there are subgroups of heart disease that have an increased HITS occurrence. To answer this question, we excluded all patients with multiple echocardiographic diagnoses and performed separate analyses of those with only one clearly defined echocardiographic diagnosis. The benefit of generating clearly defined subgroups caused a problem of small patient numbers per group. To deal with this problem from the statistical point of view, we used Fisher’s exact test to compare each subgroup with the remaining 66 patients. Significant differences in HITS occurrence were not found in any of our echocardiographically defined groups. A tendency for higher HITS prevalence was only evident in those patients having documented infectious endocarditis. These patients tended to have more HITS than the other 66 patients. This result is in agreement with the multiple thromboembolic complications of patients suffering from infectious endocarditis. It is possible that the inclusion of larger numbers of patients within each subgroup would lead to statistically significant differences.
In conclusion, we believe that TCD is a helpful and noninvasive method for investigating clinically silent microemboli in patients with potential cardioembolic disease. We detected HITS in a substantial number of patients, even though the majority were receiving specific treatment to avoid embolic complications. Our data demonstrate that patients with known cardiac sources of emboli present with a number of subclinical emboli identified as HITS but that only a minority of HITS are associated with proven neurological events. Whether this is also true for neuropsychological deficits, or “minor brain damage,” has to be evaluated in further studies. We were unable to demonstrate that the frequency of HITS in our population differs significantly among different cardiac diseases or among neurologically symptomatic and asymptomatic patients. Therefore, we do not think that HITS are an independent risk factor for recurrent stroke. Additional analyses of the HITS signals are needed. The aim of such studies should be to further investigate the underlying material that produces HITS and determine whether HITS of different intensity and duration represent different risks. Such investigations will help to validate TCD monitoring in patients with stroke and to assess stroke risk.
We wish to thank Professor John Krasney (University of Buffalo, NY) and Dr Gary Brook (RWTH Aachen, Department of Neurology) for helpful suggestions.
Reprint requests to Dr Ulrich Sliwka, MD, RWTH Aachen, Department of Neurology, Pauwelsstr 30, D-52057 Aachen, Germany.
- Received April 17, 1995.
- Revision received August 17, 1995.
- Accepted August 18, 1995.
- Copyright © 1995 by American Heart Association
Stump DA, Tegeler CH, Rogers AT, Coker LH, Newman SP, Wallenhaupt SL, Hammon JW. Neuropsychological deficits are associated with the number of emboli detected during cardiac surgery. Stroke. 1993;24:509. Abstract.
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.
Siebler M, Sitzer M, Steinmetz H. Detection of intracranial emboli in patients with symptomatic extracranial carotid artery disease. 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.
Siebler M, Kleinschmidt A, Sitzer M, Steinmetz H, Freund H-J. Cerebral microembolism in symptomatic and asymptomatic high-grade internal carotid artery stenosis. Neurology. 1994;44:615-618.
Diehl RR, Sliwka U, Rautenberg W, Schwartz A. Evidence for embolization from a posterior cerebral artery thrombus by transcranial Doppler monitoring. Stroke. 1993;24:606-608.
Anzola GP, Magoni M, Costa A, Cobelli M, Guindani M, Coletti G, Alfieri D. Transcranial Doppler monitoring in atrial fibrillation. Stroke. 1995;26:174. Abstract.
Georgiadis D, Grosset DG, Kelman A, Faichney A, Lees KR. Prevalence and characteristics of intracranial microemboli signals in patients with different types of prosthetic cardiac valves. Stroke. 1994;25:587-592.
Sliwka U, Diehl RR, Meyer B, Schöndube F, Noth J. Transcranial Doppler ‘High Intensity Transient Signals’ (HITS) in the acute phase and long term follow-up of mechanical heart valve implantation. J Stroke Cerebrovasc Dis. 1995;5:139-146.
Streifler JY, Furlan AJ, Barnett HJM. Cardiogenic brain embolism incidence, varieties, treatment. In: Barnett HJM, Mohr JP, Stein BM, Yatsu FM, eds. Stroke: Pathophysiology, Diagnosis, and Management. 2nd ed. New York, NY: Churchill Livingstone Inc; 1992:967-994.
Tegeler CH. High-intensity transient signals detected by Doppler ultrasonography: searching for answers. Cerebrovasc Dis. 1994;4:379-382.
Pugsley W. The use of Doppler ultrasound in the assessment of microemboli during cardiac surgery. Perfusion. 1989;4:115-122.
Russel D, Madden KP, Clark WM, Sabdset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke. 1991;22:253-258.
Markus HS, Brown MM. Differentiation between different pathological cerebral embolic materials using transcranial Doppler in an in vitro model. Stroke. 1993;24:1-5.
Markus H, Loh A, Brown MM. Computerized detection of cerebral emboli and discrimination from artifact using Doppler ultrasound. Stroke. 1993;24:1667-1672.
Markus H. Transcranial Doppler detection of circulating cerebral emboli. Stroke. 1993;24:1246-1250.
Brown MM, Markus HS. Transcranial Doppler detection of asymptomatic cerebral microemboli. J Heart Valve Dis. 1993;3:126-127.
Eicke BM, von Lorentz J, Paulus W. Emboli detection in different degrees of carotid disease. Stroke. 1995;26:731. Abstract.
Georgiadis D, Mallinson A, Grosset DG, Lees KR. Coagulation activity and emboli counts in patients with prosthetic cardiac valves. Stroke. 1994;25:1211-1214.
Grosset DG, Georgiadis D, Abdullah I, Bone I, Lees KR. Doppler emboli signals vary according to stroke subtype. Stroke. 1994;25:382-384.