Consensus statement: Carotid endarterectomy

1.   Carotid endarterectomy (CEA) has been shown in two major trials (ECST and NASCET) to prevent stroke in patients with carotid territory TIA or non-disabling stroke during the last 6 months and a severe stenosis of the ipsilateral carotid artery. However, the trial results were not consistent for moderate stenosis, although this may have been due to differences between the trials in the method of measurement of the degree of stenosis and in the definition of outcomes. The individual patient data from these trials and a third trial (VA 309) have now been pooled, normalising for these differences. The NASCET method of measuring stenosis was used in the pooled analysis and is used below.

2.   The benefit of CEA increases with the degree of stenosis, and is substantial for 70-99% stenosis of the symptomatic carotid artery (Grade A evidence): six patients were needed to treat to prevent one ipsilateral stroke (NNT=6; 95% CI 5-9) or any stroke or surgical death (NNT=6; 95% CI 5-10).

      The benefit from surgery for patients with narrowing of the ICA distal to a severe stenosis (as shown by angiography) is much less.

3.   The benefit from CEA is lower in patients with 50-69% stenosis than for severe stenosis (grade A evidence); twenty-four patients were needed to treat to prevent one ipislateral ischaemic stroke (NNT=24; 95% CI 13-50), or fourteen to prevent one stroke of any origin or death (NNT=14; 95% CI 9-35).

      There is no benefit for patients with 30-49% stenosis and surgery is harmful for patients with <30% stenosis (grade A evidence).

 4.  Patients with carotid TIA or non disabling stroke require urgent investigation. This should include history, neurological and cardiovascular examination and imaging of the carotid arteries. CT and or MR brain imaging should be done at least in cases with cerebral symptoms but may be omitted for patients with only ocular symptoms. Poorly controlled hypertension should be treated preoperatively since it is a risk factor for postoperative stroke.

 5. The benefit from CEA is highly dependent on time since the presenting symtom. The absolute risk reduction from CEA is reduced by half if surgery is delayed beyond 2 weeks and further reduced by half if it is delayed beyond 4 weeks

 6. Other variables in addition to the degree of stenosis and time may be useful in the selection of patients for surgery. This is particularly important for patients with 50-69% stenosis. In patients with 70-99% stenosis without an ICA narrowing, benefit is present for all subgroups. In patients with 50-69% stenosis the following variables may influence the benefit of surgery: sex, age, stroke vs. cerebral or ocular TIA and angiographic plaque surface morphology.

      Additional studies are needed to identify other possible variables that may influence the outcome e.g. ultrasonographic appearance of the plaque and recording microemboli with TCD. Preliminary analysis demonstrates that these risk factors can be used in a prognostic model.

7.   Surgery should be restricted to centres at which the practice and complication rates are regularly monitored, and at which complication rates of surgery are comparable to those found in the clinical trials demonstrating efficacy. Monitoring should be prospective and independent and should take account of case mix.

8.   Although trial evidence is based on angiography, most centers now use ultrasonography in the diagnosis of carotid artery stenosis. Ultrasonographic duplex technique measures the degree of stenosis with high accuracy, which is improved by a low doppler angle (<50^o). The degree of stenosis has to be determined by angle specific peak systolic velocity cut off points.

      A repeat ultrasound examination is strongly recomended prior to surgery as part of a quality control.

9.   Carotid angioplasty with or without stenting has not been properly compared with CEA in terms of safety, effectiveness and durability. The results of ongoing studies are awaited.

10. Surgery is also effective for severe asymptomatic stenosis (Grade A evidence), but the benefit is much less than for severe symptomatic stenosis (NNT = 20 for any stroke or surgical death; 95% CI=14-50) according to ACAS, but metaanalysis indicates that the number is higher.

      A large trial is currently ongoing in order to determine which patient groups benefit most. Ideally surgery for asymptomatic carotid stenosis should be performed within such a trial.

Abbreviations:

ACAS, Asymptomatic carotid atherosclerosis study

CEA, carotid endarterectomy

TIA, transitory ischaemic attack

ECST, European Carotid Surgery Trial

NASCET, North American Surgery Carotid Endarterectomy Trial

NNT, number necessary to treat

CI, confidence interval

ICA, internal carotid artery

TCD, transcranial doppler

CT,computer tomography

MR, magnetic resonance

Comparison between degree of stenosis for the two large trials