Thromboembolic Events during Endovascular Coil Occlusion of Cerebral Aneurysms

In this issue of the AJNR, Rordorf et al (page 5) report their experience with the use of diffusion-weighted MR imaging in detecting clinically silent thromboembolic events that occurred during endovascular coil occlusion of unruptured aneurysms. In a recent issue of the AJNR, Biondi et al (1) reported their experience using diffusion-weighted imaging after endovascular coil occlusion of both ruptured and unruptured aneurysms. Rordorf's group imaged 14 patients within 48 hours after elective Gugliemli detachable coil (GDC) occlusion of their unruptured aneurysms. Small diffusion-weighted imaging abnormalities, suspected to represent embolic infarctions, were noted in eight (57%) patients. Six patients had multiple diffusion-weighted imaging abnormalities. All patients were clinically asymptomatic, except for one in whom there was coil stretching and herniation into the parent vessel, resulting in a symptomatic infarct. Biondi's group imaged 20 patients before, 2 to 4 hours after, and 48 hours after GDC occlusion of their aneurysms, 11 of which were ruptured. Diffusion-weighted imaging abnormalities were seen after treatment in only two patients (10%), both of whom were asymptomatic. Clinical management was not altered by the imaging findings in either series.

As is the situation with clinically evident thromboembolic events, which are reported in the literature to range from 2.5% to 28% of cases (2, 3), we are once again left with a discrepancy in the frequency of clinically silent thromboembolic events associated with the GDC treatment of aneurysms. Many interacting variables may account for the disparities in the reported frequency of both clinically evident and silent thromboembolic events. These include, but are not limited to, the anticoagulation regimen before, during, and after treatment; aneurysm location, size, and neck morphology; number of guiding catheters and microcatheters; clinical status of the patient; and skill and experience of the operator. On the basis of these variables, two differences between the Rordorf and Biondi series deserve mention. In Rodorf's series, intravenous heparin was administered to prolong the activated clotting time to greater than 2.5 times baseline. In Biondi's series, the target activated clotting time was five times baseline. In addition, Biondi's series included intravenous aspirin administration during the procedure in selected patients. There was also a notable difference in the degree of final aneurysm occlusion between the two series. Rodorf et al reported only three (21%) patients with complete occlusion and five (36%) patients with residual aneurysm filling, whereas Biondi reported aneurysm occlusion in 18 (90%) patients and small neck remnants in two (10%) patients. The numbers are small, but it is reasonable to postulate that differences in the anticoagulation regimen and degree of aneurysm occlusion may at least partially account for the difference between the two series in the percentage of new diffusion-weighted imaging abnormalities seen after treatment.

What is there to be learned from these two series? First, it is unlikely that routinely obtaining diffusion-weighted images after endovascular treatment of cerebral aneurysms is going to be cost-effective in significantly impacting subsequent patient management. Combining the two series, the diffusion-weighted imaging abnormalities in nine of the ten patients were “clinically silent,” and did not prompt a change in the posttreatment care in any of the patients. The remaining patient had a clinically evident infarct, which was probably known before posttreatment imaging was performed. Nonetheless, it is apparent that diffusion-weighted imaging will be a very important tool to help us understand and prevent future complications. Diffusion-weighted imaging provides us with a very sensitive and objective measure of thromboembolic events that occur during treatment. Much uncertainty currently exists regarding the optimal anticoagulation regimen that should be employed for the endovascular treatment of both ruptured and unruptured aneurysms. Investigators should be encouraged to seek funding that will allow routine diffusion-weighted imaging after endovascular treatment of aneurysms, which I hope will result in objective data regarding optimal anticoagulation regimens.

Finally, I agree with Rordorf et al that it is unlikely that the new diffusion-weighted imaging abnormalities revealed in their patients would cause abnormalities detectable on neuropsychological testing. It must be remembered, however, that the unexpectedly high rate of measurable cognitive impairment after surgical clipping of unruptured cerebral aneurysms was unsuspected prior to the publication of the results of The International Study of Unruptured Intracranial Aneurysms (4), primarily because cognition in postoperative patients had not been previously evaluated in a scientific and systematic fashion. In the continuing process of evaluating endovascular treatment versus surgical treatment for cerebral aneurysms, we need to document in a fashion that is undeniable to the skeptics that cognition is not significantly impaired after endovascular treatment.