[Atheroma: which carotids should be surgically treated?].

Atheroma of the carotid bifurcation can be held responsible for 15% of the 150 000 cases of cerebrovascular attack recorded annually in France. Most of these are caused by an embolus detached from the thrombus formed around the atheromatous lesions. The natural history of asymptomatic carotid stenosis is imperfectly known because the patients are very heterogeneous. However, published series suggest that the risk of ischaemic complication in cases with more than 50% stenosis is about 0.7% per annum. No medical treatment has been tested against this disease. On the other hand, several uncontrolled studies have highlighted the therapeutic value of endarterectomy. The tendency towards surgery may become more widespread as non-invasive detection by ultrasounds is increasingly used. A randomized therapeutic trial is required to determine the value of this operation and the maximum tolerable morbidity rate. In the case of transient ischaemic attacks, a joint study on extracranial arterial occlusion has demonstrated the value of successful endarterectomy and shown that the maximum tolerable morbidity from surgery is 3%. Slightly higher figures are probably acceptable for cases with tight stenosis. Such an approach cannot determine whether endarterectomy would be of value in completed stroke, because the pathological conditions involved are extremely variable. Reasoning by analogy with transient ischaemic attacks, it would seem that tight or ulcerated stenoses could benefit from the operation some time after a regressive stroke. Ultrasonic techniques and intravenous angiography have made examination of the cervical arteries considerably easier. The risk of wounding the left carotid artery should be carefully weighed. Preoperative cardiac assessment is necessary in view of the high incidence of heart disease in the follow-up of these patients. Thus, natural history data set high standards for surgical performance and less risk for pre-operative investigations.