| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Departments of Neurology (W.K.B., B.M.D., F.S., T.H., U.B.), Cardiology (S.R.H.), and Psychiatry (H.J.K.), University of Regensburg, Universitätsstrasse 84, 93053 Regensburg, Germany. Received September 24, 2001; revision requested November 21; final revision received May 7, 2002; accepted May 16. Address correspondence to W.K.B. (e-mail: wendelin.blersch@klinik.uni-regensburg.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Forty consecutive patients with stroke or transient ischemic attack were admitted to the hospital and were examined by using c-TCCS and c-TEE. High-intensity transient signals (HITS) were counted for 25 seconds after the end of the Valsalva maneuver, and the numbers of HITS were classified in one of four categories (zero HITS, one to 10 HITS, >10 HITS without curtain, and curtain). A statistically significant difference was calculated with the Fisher exact test.
RESULTS: HITS were counted in 21 (52%) patients by using c-TCCS and c-TEE. With both tests, no HITS were counted in 15 (38%) patients. In two (5%) patients, no HITS were counted with c-TEE but three HITS in one patient and five HITS in the other were counted with c-TCCS. In two (5%) patients, no HITS were counted with c-TCCS, but a small patent foramen ovale (PFO) was seen at c-TEE. With c-TCCS, the sensitivity was 91% (21 of 23) and the specificity was 88% (15 of 17). In 23 patients examined with c-TCCS, 14 (61%) patients had category 1 PFO; seven (30%) patients, category 2 PFO; and two (9%) patients, category 3 PFO. Mean HITS count in patients with category 1 PFO was 4.4 and that for those with category 2 PFO was 27.6.
CONCLUSION: c-TCCS is a sensitive noninvasive method for detecting cardiac right-to-left shunt and is as sensitive as c-TEE.
© RSNA, 2002
Index terms: Brain, infarction, 10.78 • Brain, US, 10.1298, 10.12988, 10.12989 • Heart, diseases, 514.1411 • Heart, US, 50.1298, 50.12988, 50.12989 • Ultrasound (US), contrast media, 10.12988
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Different methods are used in clinical practice to detect cardiac right-to-left shunt. Findings in several studies indicated that contrast material–enhanced transesophageal echocardiography (c-TEE) is superior to contrast-enhanced transthoracic echocardiography and that it has a high sensitivity and specificity in the detection of PFO (13–15). However, disadvantages of the method are that it is semiinvasive and that it depends on the patient’s cooperation and ability to swallow.
Since the early 1990s, contrast-enhanced transcranial Doppler ultrasonography (US), introduced by Teague and Sharma (16), has become an optional method for detecting a cardiac right-to-left shunt. By using this method, high-intensity transient signals (HITS) passing through the middle cerebral artery can be detected noninvasively. Results of studies in which contrast-enhanced transcranial Doppler US was compared with c-TEE as the reference standard were convincing in regard to sensitivity and specificity for detection (17–20). Standardization in contrast-enhanced transcranial Doppler US procedures for cardiac right-to-left shunt detection was proposed in the Venice Consensus Paper in 1999, as mentioned in the study of Jauss and Zanette (21). The most limiting factor of contrast-enhanced transcranial Doppler US is the lack of a sufficient temporal bone window in a proportion of patients (22,23).
Recently, contrast material–enhanced transcranial color-coded sonography (c-TCCS) has evolved as an additional tool for diagnosis of cerebrovascular diseases (22,24). The technique allows generation of a two-dimensional color image of the cerebral vasculature in combination with the angle-corrected Doppler mode. The purpose of the present study was to determine the sensitivity of c-TCCS for the detection of cardiac right-to-left shunt and to compare it with that of the reference standard, c-TEE.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Cerebrovascular risk factors in patients were as follows: established arterial hypertension, 14 patients; smoking, six patients; diabetes mellitus, six patients; and hyperlipidemia, five patients. One patient had three concurrent cardiovascular risk factors (ie, hypertension, hyperlipidemia, and smoking), six patients had two risk factors (ie, hypertension and diabetes mellitus, hypertension and smoking, or hyperlipidemia and diabetes mellitus), and 16 patients had only one cardiovascular risk factor.
Hematologic tests revealed evidence of thrombophilia (ie, S protein deficiency or elevated phosphatidylserine antibody level) in three patients.
In each patient, the pathologic mechanism of stroke or TIA was classified in five modified categories according to the Trial of Org 10172 in Acute Stroke Treatment, or TOAST, criteria (25). The five categories were large- or small-vessel disease, cardioembolism (embolus of cardiac origin), dissection, paradoxic embolism, and stroke or TIA of unknown cause. Paradoxic embolism was suspected when all other causes of stroke (particularly pathologic findings in the ascending aorta and myocardium), other than a PFO seen at c-TEE, were excluded. In addition, radiologic and clinical examinations were required to show ischemic lesions in two distinctly separate vessel territories (ie, carotid and/or vertebrobasilar arteries). A systematic search for thrombosis was not performed in this study.
Investigations with Echocardiography
All 40 patients underwent c-TEE performed by an experienced echocardiographer (S.R.H.) from the Department of Cardiology who was blinded to the results of the c-TCCS study. c-TEE was performed with an imaging system (Sonos 5500; Hewlett-Packard, Palo Alto, Calif) and a 4–7-MHz transesophageal multiplanar probe. For this examination, the patient fasted, and local pharyngeal anesthesia was induced with 0.02 mg of orally administered lidocaine (Xylocaine Pumpspray; AstraZeneca, Wedel, Germany). If necessary, patients were sedated with 5 mg of intravenously administered midazolam hydrochloride (Dormicum; Hoffmann-LaRoche, Grenzach-Wyhlen, Germany).
A cardiac right-to-left shunt was diagnosed when the US galactose-based contrast agent (Echovist; Schering, Berlin, Germany), 10 mL administered intravenously, passed from the right to the left atrium. A PFO was considered when at least one microbubble in the left atrium was detected during the Valsalva maneuver or within three cycles from the appearance in the right atrium. The number of microbubbles and the presence of intrapulmonary shunt were not systematically investigated in this study. Other pathologic findings at c-TEE in addition to PFO were documented.
Investigations with Extra- and Transcranial Duplex Sonography
The extra- and transcranial color duplex sonographic studies were performed with a standard US system (Elegra; Siemens Medical Systems, Erlangen, Germany). A linear-array transducer (Sonoline 7.5L40; Siemens Medical Systems) was used for extracranial studies and a phased-array transducer (Sonoline 2.5PL20; Siemens Medical Systems) was used for transcranial studies. Patients were placed in the supine position, and an 18-gauge needle was placed into the cubital vein, predominantly in the right arm, for studies with contrast material. Unilateral insonation of the middle cerebral artery was performed. For each patient, the side of the cranium with the superior temporal bone window was chosen.
At c-TCCS, the mesencephalic scanning plane was identified with unequivocal imaging of the brainstem and the contralateral skull and of the ipsilateral middle cerebral artery. Flow toward the transducer was colored red, and flow away from the transducer was colored blue. After identification of the ipsilateral middle cerebral artery, the Doppler sample volume was placed in the M1 segment. With a sample volume 10 mm long, the Doppler spectrum was measured, and by using angle correction, flow velocities were determined. All data were documented by using a commercially available archiving system (UltraPACS 4.4; ALI Technologies, Corona, Calif).
At c-TCCS, the galactose-based US contrast agent was administered by using a mechanical injection system (Pulsar; Medrad Medizinische Systeme, Volkach, Germany). The patient received 10 mL of the galactose-based US contrast agent in two fractions of 5 mL each. Injection velocity was 2.5 mL/sec.
The investigation was started with a first session of Valsalva maneuvers over 10 seconds without injection of the contrast agent. By using this procedure, the effectiveness of the Valsalva maneuver was tested; a decreased peak maximum flow velocity in the Doppler spectrum was demonstrated during the Valsalva maneuver, and a "rebound" was demonstrated after the Valsalva maneuver. In a second session, 5 mL of the contrast agent was administered, and the Valsalva maneuver was started 5 seconds after the end of contrast agent injection and lasted 10 seconds. During a period of 25 seconds after the maneuver was completed, the Doppler spectrum was continuously recorded. In the last session, the Valsalva maneuver was performed immediately after the end of administration of 5 mL of the contrast agent. Again, during a period of 25 seconds, the Doppler spectrum was recorded (Table 1).
|
Statistical Analysis
For statistical analysis, c-TEE served as the reference standard. Statistically significant differences for distribution in cross tables were calculated by using the Fisher exact test. Sensitivity was determined as the percentage of true-positive findings (ie, findings positive for PFO at c-TCCS and c-TEE) compared with true-positive plus false-negative findings (ie, findings negative for PFO at c-TCCS plus findings positive for PFO at c-TEE). Specificity was calculated as the percentage of true-negative findings (ie, findings negative for PFO at c-TCCS and c-TEE) compared with true-negative plus false-positive findings (ie, findings positive for PFO at c-TCCS plus findings negative for PFO at c-TEE). The positive predictive value was determined as the percentage of true-positive findings compared with true-positive plus false-positive findings. The negative predictive value was determined as the percentage of true-negative findings compared with true-negative plus false-negative findings. Diagnostic accuracy was calculated as the percentage of true-positive plus true-negative findings compared with the total number of examined patients. For all results, 95% confidence limits (CLs) were indicated. A P value of less than .05 was considered to represent a statistically significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
With c-TEE as the reference standard, we determined a sensitivity of 91% (21 of 23; 95% CLs: 79%, 102%) and a specificity of 88% (15 of 17; 95% CLs: 72%, 103%) for c-TCCS. The positive predictive value of detecting cardiac right-to-left shunt with c-TCCS was 91% (21 of 23; 95% CLs: 79%, 102%), and the negative predictive value was 88% (15 of 17; 95% CLs: 72%, 103%). Compared with c-TEE, the overall diagnostic accuracy was 90% (36 of 40; 95% CLs: 80%, 99%). A P value less than .001 was considered to indicate a statistically significant difference (Table 3).
|
|
|
|
|
|
|
|
|
Of the two patients who had positive c-TEE results without abnormalities at c-TCCS, one (patient 26) had atrial fibrillation as a probable cause of embolism. This patient was 70 years old and had an infarction of the left posterior inferior cerebellar artery at cranial CT. At examination in the other patient (patient 27), who was 62 years old, an infarction in the left middle cerebral arterial territory was revealed. In this patient, large-vessel disease with an embolic origin of stroke was diagnosed.
Patient 35, one of the two patients with PFO at c-TCCS and negative findings for PFO at c-TEE, was 51 years old and had a TIA in the left middle cerebral arterial territory. The TIA in this case was of unknown origin. Patient 40, the other patient, was 38 years old and had infarctions in the right middle cerebral arterial territory. In this patient, a dissection of the right internal carotid artery was diagnosed (a HITS count at c-TCCS was performed in the left middle cerebral artery). In both patients, at c-TCCS, a category 1 PFO was diagnosed.
Patient 18, the patient with positive results for PFO at a second c-TEE examination, had a category 2 PFO at c-TCCS. MR imaging in this patient revealed multiple ischemic lesions in the vertebrobasilar territory, without findings of additional embolic causes other than PFO. His condition was classified as a stroke of unknown origin.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
With increasing availability of c-TCCS, this technique allows access to the cerebral circulation for initial routine assessment. Nearly all patients with stroke or TIA undergo extracranial and transcranial duplex sonographic examination in most hospitals, and not only in hospitals that specialize in stroke treatment (24,26).
The purpose of this study was to compare c-TCCS and the standard technique of c-TEE in regard to detection of cardiac right-to-left shunt. In this study, we determined a sensitivity of 91% and a specificity of 88% with c-TCCS. These results are comparable with previously published (15,16,27,28) data in regard to use of contrast-enhanced transcranial Doppler US; with the latter modality, sensitivity was 68%–100%, and specificity was 67%–100%.
Our c-TCCS examinations were based on experience concerning paradigm settings used in previous contrast-enhanced transcranial Doppler US studies (17–22,27,29) to evaluate cardiac right-to-left shunt. Contrast agents were administered only in combination with the Valsalva maneuver. Previous data (27) in regard to the use of contrast-enhanced transcranial Doppler US showed low detection rates of HITS when it was performed without the Valsalva maneuver. In this study, the contrast agent was administered via the cubital vein. The sensitivity in detecting a cardiac right-to-left shunt by using c-TCCS may be increased when injections are performed through the femoral vein, as proposed by Hamann and colleagues (30). However, this may not be suitable for routine diagnostic examinations and is potentially hazardous because of an increased risk of damage to the femoral artery and a potentially higher infection rate. In our study, insonation of the unilateral middle cerebral artery was performed. Findings in previous studies (29,31) of contrast-enhanced transcranial Doppler US for evaluation of HITS distribution showed no side preference in cerebral arteries for HITS.
In two patients, findings at c-TEE were negative and did not suggest a diagnosis of PFO, although in both patients, findings at c-TCCS disclosed a few HITS and therefore indicated a diagnosis of PFO. As discussed in preceding studies (12,15) of contrast-enhanced transcranial Doppler US, this discrepancy may have occurred for different reasons. First, the intracardiac shunt might have been small so that detection by using c-TEE failed, even after contrast agent administration and use of the Valsalva maneuver. Also, in patients with a tube in the esophagus who were sedated, the Valsalva maneuver could not be performed as well as it could be performed during c-TCCS. Finally, the presence of a small pulmonary shunt should be considered, and minor shunts thus may not be detected at c-TEE.
Two patients had a PFO that was clearly seen at c-TEE but could not be verified at c-TCCS, not even after a second examination: In both patients, c-TEE showed a small cardiac right-to-left shunt with few bubbles during the Valsalva maneuver. A possible explanation for the negative results at c-TCCS may be the low number of bubbles, which were scattered into other arterial territories rather than into the unilateral middle cerebral artery monitored with our examination. Another explanation may be fast bubble destruction, as seen in some patients.
Of all patients, 23 (58%) showed evidence of cardiac right-to-left shunt at c-TCCS. Of these, nine (39%) patients had a PFO classified as category 2 or 3. Considering this group, four (44%) patients had a stroke or TIA of unknown cause, three (33%) patients fulfilled the criteria for a diagnosis of paradoxic embolism, and two (22%) patients had large-vessel disease. The four patients who had a stroke of unknown cause and a PFO classified as category 2 or 3 did not show ischemic lesions in two distinctly separate vascular territories. However, a single paradoxic embolism might have occurred in these patients. These data, as well as findings of preceding contrast-enhanced transcranial Doppler US studies (10–12,15), suggest a correlation between PFO size and embolic implications. Therefore, further prospective investigations are required for studying selected patients with larger PFOs (category 2 and 3) to clarify the realistic embolic implications of such pathologic findings.
One advantage of the c-TCCS technique is that it can help to correctly identify the middle cerebral artery and to distinguish that artery from other large intracranial vessels, particularly the ipsilateral posterior cerebral artery. Meves and colleagues (32) showed that only 21.4%–23.7% of the HITS detected in the middle cerebral artery could be found in the basilar artery. A clear disadvantage of classic contrast-enhanced transcranial Doppler US is that the ipsilateral posterior cerebral artery may be mistaken for the middle cerebral artery, even by an experienced examiner (33). Potentially, this may lead to reduced HITS counts or lack of detection of any HITS.
In conclusion, findings of this study demonstrated that c-TCCS is a sensitive, fast, reliable method to evaluate cardiac right-to-left shunt. Compared with contrast-enhanced transcranial Doppler US, c-TCCS allows more accurate location of the middle cerebral artery. The technique may be integrated into the routine diagnostic protocol for cerebrovascular assessment following stroke. If the postulation that a higher incidence of cardiac right-to-left shunt is correlated with a higher embolic risk for stroke is true, a correct HITS count may have important implications for either conservative management or intervention. The results of this study demonstrate that c-TCCS is a potential alternative to contrast-enhanced transcranial Doppler US in the detection of cardiac right-to-left shunt and that it has comparable clinical sensitivity and specificity. However, c-TCCS should be considered as a complementary diagnostic tool to c-TEE, since c-TEE may be used to detect cardiac abnormalities, especially a septal aneurysm, other than PFO in this context.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, W.K.B.; study concepts and design, B.M.D., W.K.B.; literature research, W.K.B.; clinical studies, W.K.B.; data acquisition, B.M.D., W.K.B., S.R.H.; data analysis/interpretation, W.K.B.; statistical analysis, H.J.K.; manuscript preparation, W.K.B.; manuscript definition of intellectual content, W.K.B., F.S., T.H., U.B.; manuscript editing, T.H., W.K.B.; manuscript revision/review, W.K.B., F.S., T.H., U.B.; manuscript final version approval, T.H., W.K.B., U.B.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Ghosh, A. K. Ghosh, and S. K. Ghosh Patent foramen ovale and atrial septal aneurysm in cryptogenic stroke Postgrad. Med. J., March 1, 2007; 83(977): 173 - 177. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Souteyrand, P. Motreff, J.-R. Lusson, R. Rodriguez, E. Geoffroy, C. Dauphin, J.-Y. Boire, D. Lamaison, and J. Cassagnes Comparison of transthoracic echocardiography using second harmonic imaging, transcranial Doppler and transesophageal echocardiography for the detection of patent foramen ovale in stroke patients Eur J Echocardiogr, March 1, 2006; 7(2): 147 - 154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Hara, R. Virmani, E. Ladich, S. Mackey-Bojack, J. Titus, M. Reisman, W. Gray, M. Nakamura, M. Mooney, A. Poulose, et al. Patent Foramen Ovale: Current Pathology, Pathophysiology, and Clinical Status J. Am. Coll. Cardiol., November 1, 2005; 46(9): 1768 - 1776. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F J Pinto When and how to diagnose patent foramen ovale Heart, April 1, 2005; 91(4): 438 - 440. [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Wu, J. F. Malouf, J. A. Dearani, D. J. Hagler, G. S. Reeder, G. W. Petty, and B. K. Khandheria Patent Foramen Ovale in Cryptogenic Stroke: Current Understanding and Management Options Arch Intern Med, May 10, 2004; 164(9): 950 - 956. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
