|Home | About | Journals | Submit | Contact Us | Français|
The aim of carotid endarterectomy is to reduce the risk of stroke and death in patients with carotid artery disease. The benefit to society of surgeons undertaking this operation lies in the balance between the risks of the natural history of the carotid disease and the risks of the surgical treatment. Carotid endarterectomy is probably the vascular surgical procedure with the most scientific evidence in its favour.
In this article the reliability of evidence is graded A-C according to the classification in Box 1.
The natural history of carotid disease is quite different for patients with symptomatic and asymptomatic carotid arterial stenoses. Symptomatic stenoses include those causing transient ischaemic attack and amaurosis fugax. Also included are patients who have had a previous completed stroke. Asymptomatic patients may have a significant carotid stenosis found incidentally, they may have a moderate stenosis that progresses on duplex follow-up, or finally they may have a stenosis identified as part of a routine preoperative work-up before major surgery such as coronary artery bypass grafting.
Two huge international trials have been done on symptomatic carotid artery disease1,2, and the resultant data have been extensively reviewed and reanalysed3. It is of some comfort that the results of these trials are similar. In the European Carotid Surgery Trial, 3024 patients were randomized2; and in the group of patients with a greater than 70% carotid stenosis the 3-year death and stroke rate was 26.5% with best medical therapy and 14.9% after carotid endarterectomy. The average postoperative death and stroke rate after carotid endarterectomy was 7%. There is grade A evidence therefore that patients with symptomatic carotid disease and an ipsilateral stenosis greater than 70% have their risk of stroke reduced by surgery. It is estimated that 7 operations are required to prevent 1 stroke.
There are several small trials on asymptomatic carotid stenoses and only one large one, the ACAS study4. A trial is currently underway in Europe and results should emerge in the next year or two. In the ACAS trial 1659 patients with a greater than 60% carotid stenosis were randomized to surgery or aspirin. The risk of stroke and death after 5 years was 11% with best medical therapy and 5.1% following surgery, a statistically significant advantage that equates to just over 17 operations to prevent 1 stroke. In the surgical group, the risk of perioperative stroke or death following carotid endarterectomy was 2.3%. For patients with asymptomatic carotid stenosis greater than 60%, surgery does indeed reduce the risk of stroke and death, providing that surgeons undertaking this work have a low post-operative morbidity rate. There is grade A evidence of this advantage. About 50 operations are required to prevent 1 disabling stroke, and it is for society to say whether the price is worth paying5.
Box 1 Classification of publications by reliability
Many cardiac surgeons believe that patients who undergo coronary artery bypass grafting have a rising stroke risk with increasing carotid disease. From a critical review of the published work Naylor et al.6 conclude that patients with no significant carotid disease have a less than 2% chance of perioperative stroke during and after coronary artery bypass, whereas patients with a unilateral occlusion have an approximately 10% chance. However, only 40% of these strokes occur within 24 hours of surgery and are thus likely to be preventable. There are many open trials with excellent results from combined carotid endarterectomy and coronary artery bypass grafting, but randomized trials are lacking and no clear evidence of benefit was identified in a recent review7. This is surely fertile ground for further research.
Whilst information about the natural history of carotid disease is becoming more widely known, the advantage to society of this operation only comes with low perioperative morbidity and mortality. There are several things surgeons can do to optimize their performance. First, all those who undertake carotid endarterectomy must know their results, for otherwise it is impossible to say whether their technique is satisfactory. One difficulty arises if a surgeon performs comparatively few operations. The average surgeon in the Vascular Surgical Society of Great Britain and Ireland (VSSGBI) performs 18 carotid endarterectomy operations per year (Vascular Surgical Society of Great Britain and Ireland Registry 1999 [www.vssgbi.org ]). The minimum number of operations per year to ensure adequate experience is not known but there is evidence, in particular from North America, that surgeons should perform at least 10 per year for optimal results. The VSSGBI is collecting national outcome audit statistics for carotid endarterectomy using POSSUM risk scoring and Bayesian analysis to compare performance between individual surgeons8. One way that more experienced surgeons can affect outcome is by selecting patients who are likely to benefit most from operation. Various risk classification schemes exist to enable selection of patients, to optimize individual advantages9,10.
As far as the technical aspects of the procedure are concerned, the need for preoperative angiography is uncertain: it carries a 1% risk of stroke and in several prospective studies did not improve outcome. A pilot study suggests that surgery can be based on duplex diagnosis alone11 (grade B evidence), though some argue for angiography in patients shown by ultrasound to have potentially operable disease12,13. Most surgeons do open carotid endarterectomy but some prefer the eversion endarterectomy technique. A couple of randomized trials, including the EVEREST study, suggested that the eversion endarterectomy method carries equivalent or lower death and stroke rates14,15. However, the technique is not suitable for all patients and a carotid shunt is not easily inserted during the procedure. Several small series demonstrate that the procedure has a significant learning curve that has to be overcome. No randomized trials exist validating the use of carotid shunts16, but routine shunt insertion with carotid patching can give low perioperative death and stroke rates17. There is no strong consensus on this matter. Use of a carotid shunt can be minimized by performing the operation under local anaesthesia, where awake neurological testing is the ideal means to determine shunt requirement. Local anaesthesia can reduce postoperative stay, but whether it improves the overall results is not clear. A careful review suggested that stroke and death rates may be reduced by local anaesthesia18, and this is the subject of a continuing randomized trial. There is, however, evidence that routine carotid patch closure reduces the risk of perioperative stroke and death. In addition, it reduces the late restenosis rate19 (grade A evidence). A randomized trial from America has suggested that there is no significant difference between vein, Dacron or polytetrafluoroethylene as the patch material20. The risks of vein patches are in patch disruption, and the risk of prosthetic materials are in patch infection. If transcranial doppler monitoring is continued postoperatively, 5% of patients are seen to be having cerebral emboli; a substantial number of these progress to a completed stroke—a risk that can be reduced by Dextran 40 infusion17 (grade A evidence).
Finally, it is no longer appropriate to write about carotid surgery without mentioning carotid angioplasty, which may well take over from carotid endarterectomy in the next ten years21. It is cheaper, reduces hospital stay, avoids surgical complications and is more comfortable and tolerable for patients. If periprocedural complication rates were similar, most patients would clearly choose the endovascular method. The CAVATAS trial has reported similar stroke and death rates for surgery and carotid angioplasty22; however, the stroke and death rate of 9.9% in the surgical arm of this trial was disturbingly high. It remains to be seen whether, with use of cerebral protection devices, angioplasty will match the low risk reported in single centre surgical series. On a cautionary note it seems that carotid angioplasty likewise has its learning curve, and in Leicester the early results were tragic23.
Carotid endarterectomy still has a secure role in the repertoire of a vascular surgeon. From grade A evidence, patients with a good chance of benefit can be identified. Experienced surgeons who audit their own results and have acceptable outcomes should continue to perform this operation. Patch closure should be routine. There is a good case for continuing transcranial doppler evaluation postoperatively, to detect continued emboli and to treat with Dextran if they occur. Endovascular therapists have yet to prove that angioplasty techniques will take over from this established surgical procedure.