Bombay Hospital Journal Issue SpecialContentsHomeArchivesSearchBooksFeedback


Consultant, Vascular Surgeon, Gangaram Hospital, New Delhi.


Carotid surgery as we know it today was developed in the 1950s. A number of different procedures were described for treating atheromatous stenoses of carotid arteries. These included resection of the stenosis and replacement with a homograft or saphenous vein graft. These have been largely abandoned and carotid endarterectomy has become the operation of choice. Since then, indications for carotid endarterectomy, as well as technical aspects of the operation, have been subjected to constant evaluation.

However, the causes of stroke are multifactorial, and the link between asymptomatic carotid lesions or minor degrees of carotid stenosis and stroke had not been investigated or elucidated. Towards the end of the 1980s considerable uncertainty arose concerning the value of carotid endarterectomy. Improvements in medical management were set against considerable geographical variation in the number of carotid endarterectomies performed, and striking differences in the rates of postoperative disabling stroke and death.

There was clearly a need for large scale, randomised trials to examine the relationship between degree of carotid stenosis and medical or surgical treatment for both symptomatic and asymptomatic patients.

Symptomatic Carotid Artery Disease

The European carotid surgery trial (ECST) and the North American symptomatic carotid endarterectomy trial (NASCET) published their interim results in 1991. Patients who had suffered a non-disabling stroke, hemispheric (carotid territory) transient ischaemic attack (TIA), or an episode of fleeting monocular blindness (amaurosis fugax) were included in the trials. Randomisation was either to carotid endarterectomy or best medical management. The degree of internal carotid artery stenosis in both trials was measured by carotid angiography. ECST stratified patients into three groups : mild stenosis (0-29%), moderate stenosis (30-69%) or severe stenosis (70-99%). NASCET randomised only those with moderate or severe stenosis, using the same percentage categories as ECST.

Conclusions from ECST and NASCET

For the first time these two large, randomised trials clarified the indications for carotid endarterectomy in symptomatic patients, and quantified the expected benefit or lack of benefit for surgery depending on the degree of carotid stenosis. Carotid endarterectomy prevents strokes in patients with severe carotid stenosis. Interim results suggested that surgery was beneficial for stenoses greater than 70%, but the final results of ECSTsuggest that this threshold should be 80% for men and 90% for women. This subgroup analysis requires confirmation by pooling NASCET and ECST data to make the conclusion more robust.

Fig 1
Fig 1

Fig 2

Fig 3

Fig 4
Fig 4

Fig 5
Fig 5

Fig 6
Fig 6


There is strong evidence that carotid endarterectomy prevents strokes in patients with symptomatic internal carotid artery stenosis greater than 80%. This threshold has been revised from the original recommendation of 70% in the light of longer follow-up from ECST. Appropriate symptoms are amaurosis fugax, retinal infarction, carotid territory TIAs and non-disabling stroke. Final data from NASCET are awaited to confirm the revised thresholds for carotid endarterectomy, based on age, sex and degree stenosis, that have arisen from the final ECST results. Symptomatic patients with mild (0-29%) or moderate (30-69%) stenoses should not undergo carotid endarterectomy. Asymptomatic patients with stenoses from 50-99% benefit from carotid endarterectomy, but the benefit is too marginal for carotid surgery to berecommended for all asymptomatic patients in this group.

Further data are required before firm conclusions can be drawn in a number of areas. No data from randomised trials exist to evaluate the role of carotid endareterectomy for other cerebral symptoms such as vertebrobasilar ischaemia, non-hemispheric symptoms and eye signs other than amaurosis fugax. Existing data from trials of carotid endarterectomy for asymptomatic disease are too few to allow identification of high-risk subgroups in whom carotid surgery could be indicated. There are no data from randomised trials to support combined carotid endareterectomy and coronary artery bypass grafting for symptomatic cardiac and carotid disease. Meta-analysis suggests that a combined operation can be performed without excess risk.

Surgical technique - Important aspects

Although specific factors that may influence the outcome of CEA are discussed below, certain aspects of general technique should be considered. Whilst not subjected to critical assessment, the author believes them to be important. They include:

Endarterectomy technique

Three techniques have been described for CEA of which standard endarterectomy via a longitudinal arteriotomy (sCEA) is most widely used. Whilst eversion endarterectomy (eCEA) has gained in popularity, interposition garfting should be reserved for revisional carotid surgery. Data from both prospective randomised trials and retrospective studies have failed to show a difference in early ICA thrombosis and 30 day combined stroked and mortality rates, or in the frequency of cranial nerve injury and myocardial infarction, between sCEA and eCEA.

Inappropriate eCEA, particularly if the atheromatous disease extends > 3 cm into the ICA, may lead to difficulties with assessment and tacking of the distal endarterectomy site. Similarly, shunt insertion and retention is not easy during eCEA. Despite this, the only prospective randomised trial of eCEA versus sCEA, in which 96% of patients underwent completion angiography, angioscopy or duplex ultrasound, showed no difference in the incidence of technical defects (8% v 9%) or the need for surgical revision (4% v 3%) 8. However, the duration of carotid clamping was shorter and recurrent ICA stenoses less common (non-significant trend) after eCEA. Patch closure is not required after eCEA.


Patch angioplasty following sCEA may reduce the rate of early ICA thrombosis (reduced thrombogenicity of the endarterectomy site, improved ICA diameter) and late restenosis (Neointimal hyperplasia) and thus improve both early and long-term stroke rates. However, patching increases the carotid clamp time and the risk of patch related complications (rupture, false aneurysm, sepsis), and so the role of patching remains uncertain.

Currently most surgeons favour selective patching based on data that suggest that complications following direct repair are greater when the diameter of the ICA is 5 mm or less, particularly in women, in whom the mean diameter of the ICA (4.9 æ 0.6 mm) is 8-15% less than in men (5.3 æ 0.7 mm). With this protocol, approximately 50% of patients will need a patch. There is no randomised trial to support this view although a non-randomised study reported a lower incidence of residual/recurrent ICA stenosis when selective patching was applied in this way.

In summary, available evidence favours patch angioplasty after sCEA, although debate continues as to which type of patch to use. It remains uncertain whether a patch should be used routinely, or only in patients with an ICA diameter of 5 mm or less. If these questions are subjected to a randomised trial one treatment arm should include eCEA if all uncertainties are to be answered. Such a study would require 3000-6000 patients.

The future

The current status of carotid endarterectomy has taken 44 years to achieve. There is little to indicate that this operation will change greatly in the future. Performing carotid endarterectomy under local anaesthesia may remove some of the risks of surgery, but as yet data do not exist to substantiate this claim.

By contrast, the endovascular management of carotid disease began in 1983 and has developed only in earnest over the last 4-5 years. The future is exciting. Pharmaceutical cover surrounding the procedure can be optimised. A great deal of time and money is currently being invested in cerebral protection systems designed to limit embolisation. One such device, based on Therons’ original concept of balloon occlusion, already has a CE mark and is undergoing extensive investigation around the world. Other systems designed to limit embolisation by filtration are already developed and entering clinical trials. Dedicated stents will have low profile delivery systems, small interstices, and a coating to limit embolisation and restenosis.

What of the future?

It is inevitable that CA will have a role in the management of selected patients with carotid artery disease but appropriate selection criteria will only ever evolve from properly conducted randomised trials where inclusion and exclusion criteria are clearly stipulated. Until this is done, the results will never be generalisable. The problem will remain the ability reliably to predict the high risk carotid plaque. Amid the haste to pioneer aids and technological advancements to make both CEA and CA safer, it should not be forgotten that only high risk patients require intervention. Surgeons and radiologists cannot justify intervening in low risk patients in order to ensure optimal procedural outcomes.

In the future, it seems inevitable that both surgeons and interventionists will become subject to increased independent audit and accountability. Those who quote the results of other trials or centres to justify individual practice face the daunting prospect of medicolegal action. It also seems likely that the volume of symptomatic patients being referred for CEA or CA will be greatly reduced in the future, thereby paving the way for centralisation of carotid surgery and/or angioplasty into larger centres. Such a trend would serve to optimise outcomes, training, multi-disciplinary team input, cost and audit/research. The question as to whether CA will be able to meet these increasing demands in the future remains to be seen. At present there is no evidence that, even with future technological advances, CA will be any more generalisable or safer than CEA. Over the last eight years (1992-2000) we evaluated 340 patients with symptoms of MCA territory ischaemic events along with high grade Carotid stenosis.

Colour Doppler was used as a screening modality and the degree of stenosis was confirmed by angiography. 270 patients had excellent correlation of MR Angiography (MRA) with Colour Doppler whereas 70 had to be subjected to conventional Angiography (DSA) for two reasons:

1.Cases where MRI failed to demonstrate distal patency in very tight stenosis.

2.Patients with concomitant severe coronary artery disease which required Coronary Angiography (CAG).


Only 62 patients needed general Anaesthesia whereas 278 patients were operated under cervical block. 8 patients had to be converted to GA from cervical block due to non-cooperative nature of the patient. Of the 62 under GA, 12 underwent combined Coronary Artery Bypass (CABG), at the same sitting.

All patients were operated under indwelling shunt cover using either a Javid or Pruitt-in-Hara or recently, Argylle tube shunts.

Five patients were subjected to Eversion Endarterectomy due to significant curve and kinks in the ICA.

290 patients underwent PTFE patch angioplasty while the remaining underwent primary closure following CEA. One patient had CCA stenosis, which was patched.

Of the operated patients, 272 are on follow up with us or neurologists (maximum 6 yrs) with post operative Doppler examination and then yearly Doppler evaluations.

Five patients presented after 2-3 years with recurrent TIA’s which on re-evaluation were found to have re-stenosis which required a redo CEA. All the re-stenosis occurred in patients who had undergone primary closure.

One patient was found to have ICA aneurysmal dilatations on follow up after 2 years and he was managed by a PTFE interposition graft.

Three patients died in the immediate postoperative period. 2 had massive MI and one had advanced liver disease. Fourteen patients had postoperative TIA’s of whom 6 had irreversible deficit. 3 patients had major disabling deficits due to infarcts, possibly embolic and 1 had a major intra-cerebral haemorrhage.

Four patients needed re-exploration due to haematoma formation in the immediate post operative period ranging from 1-5 days, all were on preoperative Ticlopidine.

In 10 patients there was excessive bleeding in the immediate post operative period as they were on Ticlopidine and this was controlled by transfusing platelet rich plasma.

Most of our neurological complications occurred in the first 50 cases.


To conclude CEA is a safe procedure with a beneficial role in stroke prevention but strict selection criteria and precise and safe operative details must be adhered to, to prevent the procedure from going into disrepute. It has a learning curve, which is evident from our early experience.

To Section TOC
Sponsor-Dr.Reddy's Lab