Stroke Risks with Carotid Stenosis
abstracts & commentary
Sources: Henderson RD, et al. Angiographically defined collateral circulation and risk of stroke in patients with severe carotid stenosis. Stroke 2000;31:128-132; Demchuk AM, et al. Specific transcranial Doppler flow findings related to the presence and site of arterial occlusion. Stroke 2000;31:140-146.
The nascet1 and ecst2 trials both demonstrated that the risk of stroke in symptomatic patients increased with enlarging degrees of internal carotid artery (ICA) stenosis. In the present report, the NASCET group assessed the effect of collaterals on the risk of stroke and TIA in the presence of severe ICA stenosis. Henderson and colleagues determined angiographic collateral filling through anterior communicating (ACoA) and posterior communicating (PCoA) arteries as well as retrograde filling through ophthalmic arteries (OA) in 339 medically treated and 342 surgically treated patients entered into the NASCET study.
The percentage of patients with collaterals increased with the degree of ICA stenosis from 0.5% at less than 50% stenosis to 64% at the highest degree of stenosis (P < 0.001). In medically treated patients with severe ICA stenosis, the two-year risk of hemispheric stroke was reduced in the presence of collaterals: 28% to 11% (P = 0.005). A similar reduction was observed for hemispheric TIA (36% vs 19%; P = 0.008). In surgically treated patients, the reductions of preoperative risk and of the two-year stroke risks were not significant.
The results of this study suggest that collateral circulation is important in reducing the risk of hemisphere stroke and TIA in patients with symptomatic severe ICA stenosis. The reduction in stroke risk was demonstrated when the presence of collaterals was defined as the angiographic visualization of one or more collateral pathways via AcoA, PcoA, or OA. Leptomeningeal collaterals over the hemispheres were not considered for technical reasons. In this study, the AcoA was the most important collateral pathway.
Demchuk and associates evaluated the frequency of specific transcranial Doppler (TCD) flow findings in patients with angiographically proved arterial occlusion. A standard TCD insonation was performed and TCD was interpreted independently of angiographic results. Angiographic occlusion was demonstrated in 48 patients. There were 17 proximal ICA occlusions, 13 distal ICA occlusions, 17 middle cerebral artery (PCA) occlusions, nine distal vertebral artery (VA) occlusions, and five basilar artery (BA) occlusions. The TDC findings in the 17 patients with proximal ICA occlusion are representative. MCA wave forms were abnormal in 67%, flow was reversed in OA in 71%, there was cross-filling via the ACoA in 79%, and PCoA in 71%. There was a contralateral compensatory velocity increase in 85%. Only one of 17 patients with proximal ICA occlusion had no abnormal TCD findings.
Assessing collateral pathways may be useful in identifying patients with severe ICA stenosis who are at a lower risk of stroke and TIA and therefore may be spared carotid endarterectomy. Nevertheless, clinicians are unlikely to recommend invasive assessment with transfemoral angiography to patients already possessing severe ICA stenosis already identified by duplex Doppler and magnetic resonance angiography. In this setting, TCD appears to be a method that can be clinically important to identify compensatory flow increases in the ACoA, PcoA, and OA.
In the present study, Demchuk et al found that proximal ICA occlusion commonly produced abnormal wave forms at the ICA siphon but its effect on the ipsilateral MCA wave form depended on the amount of collateral blood flow. TCA reliably detected reversed flow in the OA and both communicating arteries. TCD offers a noninvasive alternative to transfemoral angiography in assessing underlying stroke risks, as well as developing treatment options for prognosis and secondary stroke prevention. —jjc
In medically treated patients with severe ICA stenosis, the annual risk of hemispheric stroke in the presence of collaterals is approximately: