Comparison of Transcranial Color-Coded Sonography and Magnetic Resonance Angiography in Acute Stroke
Background and Purpose We sought to compare the ability of transcranial color-coded Doppler sonography (TCCS) and magnetic resonance angiography (MRA) to identify circulatory changes that occur after acute stroke.
Methods Forty-four patients with a clinical diagnosis of acute stroke were studied with both TCCS and MRA within 24 hours of stroke onset. The appearances of all vessels identified on MRA were divided into three categories: normal, attenuated, and absent vessels. The basal cerebral arteries were identified with the use of TCCS, and their velocities were measured with pulsed-wave Doppler. The side-to-side asymmetry was calculated and expressed as an asymmetry index.
Results Five patients could not be studied with TCCS because of lack of a suitable acoustic window. An additional 4 patients were too restless to tolerate MRA scanning. Three patients had intracerebral hemorrhages, 2 patients had intracerebral gliomas, and the remaining 30 patients had cerebral infarctions. In the group of patients with acute cerebral infarction, TCCS identified 10 patients with normal asymmetry indices, 1 patient with an increased asymmetry index, 9 patients with decreased asymmetry indices, and 10 patients with occlusion of the symptomatic middle cerebral artery (MCA). MRA identified 8 normal angiograms, 6 patients had attenuated MCA branch vessels, 4 patients had MCA branch occlusions, 2 angiograms demonstrated MCA main stem attenuation, and 10 angiograms showed MCA occlusion.
Conclusions TCCS and MRA are both noninvasive techniques that can be used to study the acute stroke patient. They both yield information regarding the pathophysiology of acute stroke and may be useful techniques in deciding on therapeutic interventions. The findings agree closely with each other, and these techniques may be useful in the long-term follow-up of stroke patients.
Conventional TCD is a noninvasive method of studying the basal cerebral arteries.1 Serial study of the MCA enables analysis of the circulatory changes occurring after acute cerebral infarction.2 3 4 5 More recently, TCCS has been introduced as a new method of imaging the basal cerebral arteries.6 As a result of its B-mode facility and color-coded Doppler facilities, the vessels can be more readily and confidently identified than with TCD. Additionally, the angle of insonation can be measured and corrected for, yielding velocity measurements that are closer to the true value compared with TCD.7 TCD has been shown to be a useful tool in identifying angiographically proven MCA stenoses and occlusions in patients with cerebrovascular disease.8 9
MRA is another noninvasive technique that has been used to visualize the circle of Willis in acute stroke.10 It has been used in imaging the carotid vessels in the assessment of carotid stenosis11 and also to visualize the intracranial vasculature in cases of arteriovenous malformations,12 aneurysms,13 and venous thrombosis.14 MRA has been shown to have a high sensitivity and specificity for detecting intracranial abnormalities compared with intra-arterial digital subtraction angiography.15
Many new therapeutic interventions, such as thrombolysis, must be administered within a short time after the onset of symptoms. While novel imaging techniques may provide important information regarding the cerebral circulation, these must be readily available and rapidly acquired. Additionally, monitoring of response to such intervention is not always possible with MRI scanning, even in the setting of a large city hospital, because of scanner availability.
We studied 44 patients who presented with an acute stroke syndrome with both TCCS and MRA within 24 hours of stroke onset. The aims of the study were to assess the practicality of using these techniques in the setting of acute stroke and to assess the agreement between them. In addition, it was hypothesized that if TCCS and MRA yielded similar results, then TCCS may be used to monitor the cerebral circulation in stroke patients who have undergone an initial comprehensive MR examination, since it is relatively inexpensive and may be used at the bedside.
Subjects and Methods
Forty-four patients were entered into the study (21 men, 23 women; age range, 40 to 88 years; median, 74.8 years). All patients entering the hospital with a clinical diagnosis of carotid territory stroke who could be investigated within 24 hours of stroke onset were eligible for the study. Exclusion criteria included vertebrobasilar stroke, inability to undergo MR scanning because of either a contraindication or an inability to cooperate, and failure to obtain informed consent from either the patient or the next of kin. Patients with vertebrobasilar strokes were excluded because of technical reasons, ie, it was not possible to maintain the patient in the required position (ie, lying on one side with the neck flexed) for long enough to obtain adequate images of the vertebrobasilar system. T2-weighted axial MR images of the brain were taken at the same time as the MRA to differentiate ischemic from hemorrhagic stroke and to exclude other causes of acute neurological deficit.
TCCS was performed with the use of a 2.25-MHz curved phase array probe (Diasonics). The transtemporal acoustic window was used to visualize the basal cerebral arteries. The vessels were then interrogated with the use of pulsed-wave Doppler and the velocities measured from the spectral display. The side-to-side difference in blood flow velocity (AI) was calculated according to the formula proposed by Zanette et al9 : where MV1 and MV2 represent the mean velocities in the symptomatic MCA and contralateral MCA, respectively. A negative AI indicates a reduction in the mean flow velocity in the symptomatic MCA; conversely, a positive AI indicates an increase in the mean flow velocity in the symptomatic MCA. An AI of −200% results from occlusion of the main stem of the symptomatic MCA (ie, a flow velocity of 0 cm/s). The threshold value for asymmetry was defined as ±21%. This was based on both our previous studies16 and those by Zanette et al.2 9 Our data were derived from 29 healthy subjects older than 60 years, which yielded data on 53 MCAs (in five hemispheres the MCA could not be insonated because of lack of a suitable acoustic window). The absolute MCA velocity in the study patients was also compared with our reference values,17 which were calculated from the mean MCA velocity ±2 SD and yielded a value of 58±18 cm/s.
An occluded vessel was inferred by lack of signal on the color display despite visualization of at least two other vessels and lack of any signal when interrogated with pulsed-wave Doppler. The mean time from symptom onset to TCCS was 15.4 hours (range, 4 to 24 hours). TCCS was performed within 4 hours of MRA (range, 15 minutes to 4 hours; median, 2 hours).
MRA was performed on a 1-T Magnetom system (Siemens) with the use of a circularly polarized transmit-receive head coil with 26-cm inner diameter. Three-dimensional time of flight images were obtained with the use of a fast imaging with steady state free precession three-dimensional sequence, in the majority of cases with the use of magnetization transfer and tilted optimized nonrandom excitation to improve contrast between flowing blood and background tissue (repetition time, 36 ms; echo time, 10 ms; flip angle, 20°; acquisition, 1; field of view, 230×230; matrix 256×256; slab thickness, 50 mm; time of acquisition, 9 minutes, 52 seconds. With the use of magnetization transfer contrast and tilted optimized nonrandom excitation, a 192×256 matrix in conjunction with a repetition time of 42 ms gave a total acquisition time of 8 minutes, 30 seconds). During postprocessing, MIP was used to produce angiographic-like images. These were viewed in 12 projections rotated about a side-to-side axis through the imaging volume. The source data and targeted MIPs were used if there were equivocal results from the full MIP data set. The vessels were classified into three groups according to their MRA appearances: normal, attenuated, and absent. MRA was performed with the investigators blinded to the results of the clinical examination and TCCS. TCCS was performed with the investigators blinded to the results of MRA and MRI but not to the clinical state of the patient (the same investigator [A.R.K.] performed both the TCCS and the clinical examination).
Fourteen patients were excluded from the study: 5 patients lacked a suitable acoustic window for TCCS, 4 patients were too restless to obtain adequate MRA scans, 3 patients had primary intracerebral hemorrhages, and 2 patients had intracerebral gliomas.
Twenty-six patients were diagnosed as having a cerebral infarction on MRI, and in 4 patients who had a normal MRI, the diagnosis was made on clinical grounds after other causes had been excluded.
A summary of the results is presented in Table 1⇓.
Of the patients presenting with acute cerebral infarction, 10 were found to have normal AIs in the MCA (ie, ±20%). Average mean MCA blood flow velocities were 53.4 cm/s (range, 34.0 to 88.8 cm/s) in the symptomatic vessel and 55.5 cm/s (range, 35.6 to 88.6 cm/s) in the asymptomatic vessel. There was no significant difference between these velocities. Eight of these patients had normal MRA scans. T2-weighted MRI scans of these 8 patients were normal in 2 patients, revealed scattered high-signal areas (indicative of small-vessel disease) in 2 patients, revealed a small high-signal area in the internal capsule in 2 patients, and revealed a small high-signal area in the parietal lobe in 2 patients. MRA in the remaining 2 patients with a normal AI revealed MCA branch vessel attenuation on the symptomatic side; T2-weighted MRI was normal in 1 patient and revealed small-vessel disease in the other patient.
In 10 patients no color-coded or spectral Doppler signal could be detected from the MCA despite obtaining adequate signals from both the ACA and PCA (Fig 1a⇓). Main stem MCA occlusion was therefore diagnosed. The AI of all these patients was −200%. The average mean MCA velocity in the asymptomatic vessel was 41.1 cm/s (range, 34.3 to 48.4 cm/s). All 10 patients had absent MCAs on the symptomatic side on MRA (Fig 1b⇓). In addition, 1 patient had an occlusion of the ipsilateral ICA (tandem lesion), confirmed by color-coded duplex ultrasonography. In 7 of the patients with MCA occlusion, velocities in the symptomatic ACA were significantly greater than the contralateral hemisphere (AI >+27%).2 MRA failed to show any differences between the ACAs in these patients. T2-weighted MRI scans revealed large MCA territory infarctions in 8 patients (Fig 1c⇓), a basal ganglia infarction in 1 patient, and scattered high signal areas in 1 patient.
Nine patients were found on TCCS to have a reduced AI (range, −20.1% to −78.9%; mean, −44.8%) (Fig 2a⇓). The average MCA velocity was 30.5 cm/s (range, 21.1 to 44.2 cm/s) on the symptomatic side and 48.5 cm/s (range, 30.5 to 69.4 cm/s) on the asymptomatic side; the difference in the average velocity in the two groups was statistically significant (P<.006, Student’s t test). MRA revealed 2 patients to have marked attenuation of the proximal segment of the MCA, and MRI scans showed a parietal lobe infarction in 1 patient and a haemorrhagic parietal lobe infarction in the other. Four patients had evidence of M2 branch vessel occlusion on the symptomatic side (Fig 2b⇓). One patient also had an occluded ICA diagnosed by color-coded duplex ultrasonography, and TCCS showed retrograde filling of the ipsilateral ACA combined with a high flow velocity in the anterior communicating artery, indicating interhemispheric collateral flow. MRI scans showed high signal areas consistent with infarction in MCA branch vessel territory in all patients (Fig 2⇓ c). The remaining 3 patients with a reduced AI had M2 branch vessel attenuation on MRA; MRI was normal in 1 patient and showed evidence of a small parietal lobe and basal ganglia infarction in 2 patients.
One patient presented with a high AI (+62.4%); mean MCA velocity was 55.4 cm/s. MRA indicated MCA branch attenuation on the symptomatic hemisphere, and MRI revealed a parietal lobe infarction.
The average mean symptomatic MCA velocity in all patients with an abnormal MRA was significantly lower (P<.02, Student’s t test) than that of patients with a normal MRA (Table 2⇓).
One of the main limitations of TCD has been the number of patients that lack a suitable acoustic window, which has been estimated at approximately 5%.18 This is slightly less than for TCCS19 which, as a result of B-mode imaging, lacks the penetration of conventional TCD. The failure rate increases with age and is also higher in women because of the higher prevalence of temporal hyperostosis. In our study population, the majority of the patients were older than 65 years (34 of 43), and the failure rate was 16.3%.
Conventional TCD and MRA have both been shown to have a high specificity and sensitivity compared with conventional angiography and digital subtraction angiography in patients with cerebrovascular disorders.2 20 21 22 In a previous report23 it was shown that the combination of TCD and MRA was more reliable in identifying MCA stenoses than digital subtraction angiography alone. At this time there have been no reported studies comparing TCCS with MRA or conventional angiography in acute stroke in adults. We did not use standard angiography in this study because it was thought not to be ethically justified because of the associated mortality and morbidity. Thus, there is no gold standard as such, and therefore it is not possible to state which imaging technique gave the “correct” result when the two methods produced different results.
It is not always possible to visualize the M2 branches of the MCA using TCCS. This is due to the slower flow in these branches compared with the M1 portion and possibly due to the increased angle that the branches subtend to the ultrasound probe. Therefore, since it is not usually known exactly how many M2 branches there are, it is not possible to make a certain diagnosis of M2 branch occlusion with TCCS. A previous study comparing TCD and intra-arterial angiography showed that multiple branch occlusion was associated with a low AI.9 This was confirmed in our study, in which all 4 patients with branch occlusion had low AI’s. In addition, these patients had low absolute MCA velocities in the symptomatic hemisphere.
There was exact agreement between the two techniques for diagnosing MCA main stem occlusion. However, this finding can only be reported with confidence with TCCS if the other vessels can be clearly identified. Otherwise, failure to identify the MCA may be a result of poor ultrasound penetration. This highlights another area in which TCCS is superior to conventional TCD: with the B-mode facility and color-coding of TCCS, the major vessels can usually be identified rapidly and with more certainty than with TCD, and therefore the lack of an MCA signal is apparent immediately. This reduces the lengthy “search” required with TCD to image the ACA and PCA to make a definite diagnosis of MCA occlusion. An associated finding in acute MCA occlusion is elevated blood flow velocity in the ipsilateral ACA. This is due to a shunting of blood from the occluded MCA to the patent ACA and has been reported in previous studies.4 9 24 This finding was present in 7 of our patients with MCA main stem occlusion.
In our study group many of the patients were restless after the event, and the quality of the MRA was consequently reduced. This can make interpretation of the finer branches of the MCA difficult and may lead to an overestimation of the number of attenuated MCA branch vessels present. In addition, some patients with acute stroke have very low absolute MCA velocities, and this may give rise to reduced flow-related enhancement with this time-of-flight technique and thus an overestimation of the number of MCA branch occlusions.
Although MRA in this study was able to provide detailed anatomic information, it did not provide any quantitative data regarding blood flow velocity. The use of phase contrast techniques25 may, however, provide flow information that not only relates to the direction of flow but may also provide true quantitative flow velocity data.
There was a significant difference between the mean MCA velocities in the patients with normal MR angiograms and the 12 patients with attenuated or occluded MCA branch vessels on MRA (Table 2⇑). Eight of these 12 patients had a mean MCA velocity outside the normal reference range (ie, <40 cm/s). The lower velocities in patients with an abnormal MRA probably reflected the presence of distal flow disturbances in the MCA. The mean MCA velocity in patients with a normal MRA was 56.4 cm/s, and none of these patients had a mean MCA velocity under 40 cm/s. The value of 56.4 cm/s agrees closely with reference data for healthy subjects older than 60 years (mean MCA velocity, 58 cm/s; 95% confidence interval, 55 to 61 cm/s).17 All patients with a normal MRA had either a hemiparesis or a facial palsy, and none had evidence of higher cortical dysfunction. The normal MCA velocities recorded in this group are in keeping with the presumed small-vessel pathology underlying such lacunar infarctions26 and agree with previous studies.27
By using the threshold value of 40 cm/s for the lower end of the normal range for mean MCA velocity, it was possible to identify 8 of the 12 patients with an abnormal MRA (excluding proximal MCA occlusion). This compares well with the use of the AI, which identified 9 of the 12 abnormal MRAs.
Not surprisingly, patients with MCA branch occlusions tended to present with milder neurological deficits and a correspondingly smaller area of infarction on MRI than those patients presenting with a proximal occlusion of the MCA. Nine of the 10 patients presenting with proximal MCA occlusion had severe neurological deficits, including hemiparesis, homonymous hemianopia, and, if the dominant hemisphere was affected, dysphasia. All these patients had large MCA territory infarctions on MRI. However, 1 patient (patient 2) presented with a hemiparesis and facial palsy only and was found to have an infarction in the basal ganglia. A perfusion MRI scan of this patient showed only a deficit in the basal ganglia. This implies that the MCA occlusion may have been of long duration.
TCCS is a useful tool in the assessment of the acute stroke patient. The equipment is portable and can therefore be used as a bedside investigation. This may prove advantageous in certain patient groups, such as those too ill to undergo conventional neuroimaging. TCCS can be easily repeated and may prove useful in monitoring the patient’s response to therapeutic intervention. The scans can be performed rapidly: the cerebral vessels are imaged within minutes if there is an adequate acoustic window. This may have implications for the assessment of patients undergoing treatment that must be performed within a strict time limit. TCCS can be performed early in the clinical course of the stroke and, like MRA, may show flow abnormalities, the effects of which are not yet apparent on either CT or standard T1- and T2-weighted MRI. However, there is quite a substantial failure rate as a result of lack of a suitable acoustic window, especially in elderly women. This is a particular problem in cerebrovascular disease since most of the patients are elderly.
MRA is more able to demonstrate the anatomy of the circle of Willis, especially the finer branches of the basal cerebral arteries. An advantage of MRA is that it is not limited by the availability of an acoustic window, which precluded TCCS in 5 patients. However, in our study 4 patients had to be excluded because they were too restless for adequate imaging once in the MR scanner.
We have shown that both MRA and TCCS can identify proximal MCA occlusion. MRA can identify attenuation and occlusion of distal MCA branches that cannot be imaged satisfactorily with TCCS. The latter findings can be inferred if mean MCA velocity is below the threshold value of 40 cm/s or if the AI is greater than −20%. With a trend toward an increasing number of stroke patients undergoing MR imaging, the addition of MRA is a valuable tool for determining the patience of the major cerebral vessels. MRA may also be used not only in conjunction with standard T1- and T2-weighted imaging but also with the newer techniques of perfusion MRI and diffusion-weighted imaging. This combination of investigations would provide the clinician with a powerful tool for the comprehensive assessment of the stroke patient. TCCS may also be used in the acute situation, and although it does not provide the anatomic detail of MRA, it can reliably diagnose proximal MCA occlusion. In the routine clinical setting, MRA may be used as the initial imaging technique in combination with standard T2-weighted imaging to diagnose the cause of the stroke and to determine vessel patency. If MRI is contraindicated, TCCS may be used instead. If further follow-up is required, TCCS may be the technique of choice since it can be performed as a bedside test, is relatively inexpensive, and may be more accessible than MRA in major city hospitals.
Selected Abbreviations and Acronyms
|ACA||=||anterior cerebral artery|
|ICA||=||internal carotid artery|
|MCA||=||middle cerebral artery|
|MRA||=||magnetic resonance angiography|
|PCA||=||posterior cerebral artery|
|TCCS||=||transcranial color-coded Doppler sonography|
|TCD||=||transcranial Doppler sonography|
This study was supported by the Stroke Association. and by a Leicester Royal Infirmary NHS Trust Research Fellowship Award (Dr Kenton). The authors gratefully acknowledge the help of the staff in the MRI suite.
- Received December 28, 1996.
- Revision received March 18, 1997.
- Accepted April 11, 1997.
- Copyright © 1997 by American Heart Association
Zanette EM, Roberti C, Mancini G, Pozzilli C, Bragoni M, Toni D. Spontaneous middle cerebral artery reperfusion in ischemic stroke: a follow-up study with transcranial Doppler. Stroke. 1995;26:430-433.
Ringelstein EB, Biniek R, Weiller C, Ammeling B, Nolte PN, Thron A. Type and extent of hemispheric brain infarctions and clinical outcome in early and delayed middle cerebral artery recanalization. Neurology. 1992;42:289-298.
Kaps M, Damian MS, Teschendorf U, Dorndorf W. Transcranial Doppler ultrasound findings in middle cerebral artery occlusion. Stroke. 1990;21:532-537.
Bogdahn U, Becker G, Winkler J, Greiner K, Perez J, Meurers B. Transcranial color-coded real-time sonography in adults. Stroke. 1990;21:1680-1688.
Rorick MB, Nichols FT, Adams RJ. Transcranial Doppler correlation with angiography in detection of intracranial stenosis. Stroke. 1994;25:1931-1934.
Zanette EM, Fieschi C, Bozzao L, Roberti C, Toni D, Argentino C, Lenzi GL. Comparison of cerebral angiography and transcranial Doppler sonography in acute stroke. Stroke. 1989;20:899-903.
Turski PA, Korosec FR, Graves VB, Strother CM. Vascular malformations. In: Potchen JE, Haacke EM, Siebert JE, Gottschalk A, eds. Magnetic Resonance Angiography: Concepts and Applications. St Louis, Mo: Mosby; 1993:380-404.
Riles TS, Eidelman EM, Litt AW, Pinto RS, Oldford F, Schwartzenberg GW. Comparison of magnetic resonance angiography, conventional angiography, and duplex scanning. Stroke. 1992;23:341-346.
Martin PJ. Evaluation and Application of Transcranial Colour-Coded Sonography. Leicester, UK: Leicester University; 1994. Thesis.
Martin PJ, Evans DH, Naylor AR. Transcranial color-coded sonography of the basal cerebral circulation: reference data from 115 volunteers. Stroke. 1994;25:390-396.
Wada T, Kodaira K, Fujishiro K, Maie KI, Tsukiyama E, Mikawa H, Shimuzu H, Fukumoto T, Okamura T. Quantitative measurement of middle cerebral artery flow velocity using transcranial Doppler sonography. Arters et Veines. 1990;9:682-685.
von Reutern GM, von Budingen HJ. Doppler examination of the basal cerebral arteries (transcranial or intracranial sonography). In: Ultrasound Diagnosis of Cerebrovascular Disease. Stuttgart, Germany: Georg Thieme Verlag; 1993:176-248.
Mattle HP, Grolimund P, Huber P, Sturzenegger M, Zurbrugg HR. Transcranial Doppler sonographic findings in middle cerebral artery disease. Ann Neurol. 1988;45:289-295.
Horowitz DR, Tuhrim S, Weinberger JM, Rudolph SH. Mechanisms in lacunar infarction. Stroke. 1992;23:325-327.