From the Departments of Cardiology and Radiology, University of Bonn, Germany
In most cases, the severity of valvular aortic stenosis may be accurately assessed noninvasively by echocardiography. Cardiac catheterization, however, is still often performed to determine the aortic valvular area and pressure gradient invasively.1-4 This procedure carries a potential risk of apparent cerebral embolism.5 The risk of clinically inapparent cerebral embolism is not yet known. Diffusion-weighted magnetic resonance imaging (MRI) has been shown to be highly sensitive and specific for the detection of acute ischemic cerebral lesions.6-8 We conducted a prospective, randomized study using diffusion-weighted MRI to determine the incidence of clinically apparent and silent cerebral embolism after retrograde catheterization of the aortic valve in patients with severe valvular aortic stenosis. Patients and methods To determine the incidence of embolic events, 152 consecutive patients with valvular aortic stenosis were randomly selected in a 2:1 relationship to receive either cardiac catheterization with (n = 101) or without (n = 51) passage of the aortic valve. The patients underwent cranial MRI imaging and neurological assessment within 48 hours before and after the procedure.9-11 Exclusion criteria included contraindication to MRI or transesophageal echocardiography, or the inability to give written informed consent. Patients with unclear echocardiographic findings, or for whom we were unable to obtain reliable measurements of the aortic valve area, were also excluded from the study. The control group consisted of 32 patients without valvular aortic stenosis who underwent coronary angiography and levocardiography.
All patients were examined clinically, and any history of previous embolism was assessed. Patients underwent transthoracic and transesophageal echocardiography, a 12-lead surface electrocardiogram, an ultrasonography examination of the carotid arteries, as well as coronary angiography.1,3,12-14
Patient groups with and without passage of the aortic valve did not differ regarding age, sex, history of embolism, occurrence of coronary artery disease, other cardiovascular risk factors, and atrial fibrillation. Compared with these two groups, control subjects more frequently had coronary artery disease and a history of embolism (table 1).
Table 2 shows the patients’ hemodynamic parameters on echocardiography. Aortic valvular area, mean pressure gradient across the valve, left ventricular and atrial dimensions, and left ventricular ejection fraction did not differ between patients with and without passage of the valve. No differences in left ventricular and atrial dimensions between the control and the study groups were found. Patients in the study group had a significantly higher left ventricular ejection fraction (P < .01).
Table 3 shows the frequency of other possible cardiac sources of cerebral embolism. Patient groups did not differ with respect to aortic atheroma, patent foramen ovale, thrombi, dense left atrial spontaneous echocardiographic contrast, and peak emptying velocities of
the left atrial appendage. Control
subjects, however, had more severe left atrial spontaneous echocar-
diographic contrast (P < .02) than study patients.
MRI scans showed that no patients had focal diffusion abnormalities before cardiac catheterization. Of 101 patients with passage of the stenotic aortic valve, 22 (22%) showed acute cerebral diffusion abnormalities in a pattern consistent with embolic lesions (99% confidence interval [CI], 12%—34%) after the procedure. Four of these patients had more than one embolic lesion.
Of 101 patients with passage of the stenotic aortic valve, three (3%) had an acute, clinically apparent complication after cardiac catheterization (99% CI, 0%—13%). Focal diffusion abnormalities were shown on MRI for all patients with symptoms. The locations of these embolic
lesions were concordant with the neurological findings of all three patients. For one patient, the neurologic examination after the procedure showed a supranuclear disorder of gaze and scattered hypoesthesia of the left arm. These deficits completely resolved within 3 months. One patient developed diplopia, which was still present at the 3-month follow-up examination. Another patient developed a right hemiparesis, which was also present at 3-month follow-up.
After 3 months, MRI scans demonstrated an acute diffusion abnormality after cardiac catheterization in 17 of 22 patients. These patients developed a focal signal hyper-
attenuation in the region corre-sponding to the original index lesion, indicating infarcted brain tissue. Patients without passage of the stenotic aortic valve and control subjects had no acute diffusion abnormalities after the procedure. No deaths, perforation of the left ven-
tricle, cardiac tamponade, acute myocardial infarction, or pulmonary embolism during cardiac catheterization were recorded in the study patients or control group.
The mean (± SD) fluoroscopy time was longer for patients with passage of the stenotic aortic valve (6.1 ± 1.1 minutes) than for patients without passage (2.9 ± 0.9 minutes; P < .01) or control subjects (2.8 ± 0.6 minutes; P < .01). No difference in the fluoroscopy time between patients without passage of the valve and control subjects was noted. There was no correlation between fluoroscopy time and frequency of acute cerebral diffusion abnormalities after the procedure for patients with passage of the aortic valve (5.8 ± 1.4 minutes with embolism versus 6.2 ± 1.1 minutes without embolism).
We assessed the frequency of clinically apparent and silent cerebral embolism in patients with degenerative aortic valvular stenosis undergoing cardiac catheterization with transvalvular passage using modern MRI techniques. The chief findings of our study are that retrograde catheterization of valvular aortic stenosis is associated with a high rate of clinically silent cerebral embolism (22%) and can also lead to clinical symptoms of neurological deficits in 3% of patients.
A retrospective analysis showed similar findings for the rate of clinically apparent cerebral embolism, with a 1.7% frequency of neurological complications.5 These data confirm the increased risk of cerebral embolism with retrograde passage of valvular degenerative aortic stenosis. Our observation that the patients with aortic valvular stenosis who underwent coronary angiography without retrograde catheterization of the valve and the control group did not have cerebral embolism strengthens this conclusion. These findings show that the passage of the valve itself is responsible for emboli.
Another important finding of our study is the high frequency of clinically silent cerebral embolism. A potential explanation for this finding might be that most lesions were located in noneloquent brain areas. Emboli could have produced se-
vere neurological complications
in eloquent brain areas, such as the Wernicke temporal speech area,
the motor cortex, or Broca’s motor speech area. Although clinically silent embolic lesions have no clinical effect and clinical stroke is a different entity, strong evidence links these lesions with structural ischemic injury.6,8,9 These data are similar to our observation that all patients with acute cerebral diffusion abnormalities had structural tissue changes at follow-up.
There were no differences between study patients with and without passage of the valve regarding other cardiovascular sources of
embolism, that is, low left ventricular ejection fraction, aortic atheroma, patent foramen ovale, cardiac thrombi, left atrial spontaneous echocardiographic contrast, and carotid stenosis. We were careful to avoid the formation of thrombi on the catheter or guidewire, which would have presented another possible cause of embolism, by flushing catheters, cleaning the wire regularly, and giving unfractionated heparin to patients undergoing passage through the aortic valve. There was no correlation, however, between fluoroscopy time and cerebral embolism for patients who underwent cardiac catheterization with passage through the stenotic valve.
This study shows that patients undergoing retrograde catheterization of a stenotic aortic valve are at substantial risk for clinical neurological complications. The frequency of clinically silent cerebral embolic lesions we noted was also far greater than clinical embolism in these patients. In the past, therefore, the considerable risk of structural ischemic brain damage related to the passage of a stenotic aortic valve has been underestimated. Patients should be informed about these procedure-related risks, and retrograde catheterization of the stenotic aortic valve should be limited to those patients with unclear echocardiographic findings or to instances when additional information is essential for clinical management.