Carotid Endarterectomy

31.1Decision frame before operative benefit

Carotid endarterectomy is not a procedure chosen from the angiogram alone. The first decision is whether the carotid lesion is truly linked to a recent retinal or hemispheric ischemic event, whether stenosis severity is in a range where operative benefit has been shown, whether the patient can be treated within the period in which benefit is greatest, whether the expected 30-day stroke or death rate at the treating center is acceptable, and whether the neck and lesion are suitable for open repair. Only after those elements are separated should a recommendation threshold be applied. Editor's Choice European 2023 · CREST 2010

For symptomatic disease, the strongest operative lane is recent ipsilateral TIA, amaurosis fugax, or non-disabling stroke with severe carotid stenosis. The NASCET high-grade symptomatic cohort established the classic benefit signal for 70–99% stenosis, while ECST showed a major stroke-or-death reduction in patients with symptomatic stenosis of at least 80% by the ECST measurement method; the two measurement systems are not interchangeable and should not be treated as numerically identical. Randomised trial endarterectomy 1998 · NASCET · 1991

Timing is central. Pooled NASCET and ECST data showed that the number needed to treat rises sharply when carotid endarterectomy is delayed beyond 2 weeks after the index TIA or minor stroke. The practical implication is that symptomatic patients who are neurologically stable, medically fit, and anatomically suitable should move through imaging, risk review, and operative scheduling as an urgent pathway rather than as an elective referral. Rothwell pooled CEA analysis 2004 · Randomised trial endarterectomy 1998 · NASCET · 1991

Current European and North American framing is consistent in treating symptomatic 70–99% stenosis as the clearest indication for carotid endarterectomy in an average-risk patient, with more selective benefit in the 50–69% band. The European Stroke Organization recommends CEA for symptomatic 70–99% stenosis and suggests it for 50–69% stenosis, with treatment best delivered within 2 weeks of the qualifying retinal or cerebral event when feasible. The SVS 2022 guideline frames CEA as the default modality for average-surgical-risk symptomatic 70–99% stenosis. SVS 2022 · European Stroke Organisation 2021

The moderate symptomatic stenosis lane requires more discipline. In the NASCET 50–69% subgroup, benefit was smaller over 5 years and concentrated in selected patients, particularly men and patients treated within a few weeks of the qualifying event. This is not a “never operate” category, but it is a category in which the surgeon must be explicit about symptom credibility, stenosis measurement, timing, competing stroke mechanisms, and local perioperative results. NASCET · 1991

Sex modifies expected benefit but should not be converted into a simplistic exclusion rule. The pooled NASCET/ECST analysis reported an approximate 5-year number needed to treat of 9 for men and 36 for women with symptomatic carotid stenosis, reflecting a smaller absolute benefit in women in those trial data. In contemporary practice, women with appropriate symptomatic stenosis can still be candidates for CEA, but the operative-risk gate and timing discipline are especially important. Rothwell pooled CEA analysis 2004 · Randomised trial endarterectomy 1998 · NASCET · 1991

Asymptomatic disease is a different benefit lane. The SVS 2022 guideline frames asymptomatic CEA as appropriate for approximately 70% or greater stenosis by validated imaging in patients with at least 3 to 5 years of life expectancy and acceptable surgical risk. This framing should be applied only after confirming that the patient is not being misclassified as asymptomatic, that imaging is reliable, and that the treating surgeon and institution can meet the low complication rates required for net benefit. SVS 2022

The CEA-versus-stenting boundary should be kept narrow in this chapter. Stenting is not the default alternative whenever open surgery is inconvenient; SVS 2022 reserves carotid artery stenting for patients with documented high surgical risk for CEA or hostile cervical anatomy. CREST showed an age-treatment interaction, with patients aged 70 years or older doing better with endarterectomy than stenting on the periprocedural stroke axis, and ESVS 2023 similarly frames CEA as preferred over stenting in average-risk symptomatic patients aged 70 years or older. Editor's Choice European 2023 · SVS 2022 · CREST 2010

Antithrombotic management should support, not delay, the urgent symptomatic pathway. The AHA/ASA secondary stroke prevention guideline frames short-course dual antiplatelet therapy for minor stroke or high-risk TIA, followed by single antiplatelet therapy before or after CEA according to bleeding and ischemic-risk balance. For the surgeon, the key decision is to coordinate the operative date, neurological stability, and hemorrhagic risk rather than applying a fixed antiplatelet rule detached from the patient’s presentation. AHA 2021 · AHA/ASA secondary stroke prevention 2021

The historical lineage of CEA begins with open reconstruction for recurrent hemispheric ischemic events, described by Eastcott, Pickering, and Rob in 1954. Modern decision-making, however, is governed by randomized trials, contemporary guidelines, and local outcome audit rather than historical precedent. The case is useful because it reminds the surgeon that CEA is a procedure for preventing future stroke in a selected patient, not simply for correcting a stenosis. RECONSTRUCTION INTERNAL CAROTID 1954

31.2Symptomatic and asymptomatic benefit lanes

The evidence base for carotid endarterectomy is best understood as two separate bodies of evidence. Symptomatic CEA trials such as NASCET and ECST address patients with recent ipsilateral ischemic symptoms, in whom untreated early recurrent stroke risk can be high and timing is decisive. Asymptomatic trials such as ACAS and ACST address screening-era stenosis, where the absolute stroke-prevention benefit is smaller and depends heavily on low perioperative complication rates and sufficient life expectancy. ACST-1 (10-year) 2010 · Endarterectomy asymptomatic carotid 1995 · NASCET · 1991

In symptomatic 70–99% stenosis, the benefit of CEA is large enough to drive urgent pathways. NASCET reported an approximate 17 percentage-point absolute reduction in 2-year ipsilateral stroke for CEA compared with medical therapy in this high-grade symptomatic band. This is the reference point result behind the familiar recommendation that an average-risk patient with recent symptoms and severe ipsilateral stenosis should generally be offered CEA promptly when the operative-risk gate is met. SVS 2022 · NASCET · 1991

ECST reinforces the same principle through a different stenosis-measurement method. In symptomatic stenosis of at least 80% by the ECST method, CEA produced an approximate 11.6% absolute reduction in 3-year major stroke or death. The practical caveat is that ECST percentages and NASCET percentages should not be casually substituted for one another; the surgeon must know how the stenosis was measured before applying a threshold. Randomised trial endarterectomy 1998

The timing signal is one of the most important bedside lessons in carotid surgery. Pooled NASCET and ECST data showed that benefit declines when CEA is delayed beyond 2 weeks after the qualifying TIA or minor stroke, with the number needed to treat rising sharply as time passes. The patient who is stable enough for surgery but waits in a routine outpatient queue may lose the very benefit that justified intervention. Rothwell pooled CEA analysis 2004 · Randomised trial endarterectomy 1998 · NASCET · 1991

The operational rationale for urgent assessment is supported by the EXPRESS experience, in which an urgent TIA and minor-stroke workup pathway reduced early recurrent stroke compared with a prior delayed-care era. For carotid surgeons, this means that “urgent CEA” begins before the operating room: rapid symptom recognition, vascular imaging, neurological triage, medical optimization, and theatre access are all part of the treatment effect. CEA 2007

Asymptomatic CEA has a narrower margin. ACAS demonstrated an approximate 5.9% absolute reduction in 5-year aggregate stroke for asymptomatic 60–99% stenosis assigned to CEA compared with medical therapy, provided perioperative complications were kept near 3%. ACST-1 later confirmed long-term benefit, but again only where perioperative stroke or death was held below approximately 3%. ACST-1 (10-year) 2010 · Endarterectomy asymptomatic carotid 1995

The ACST-1 figures illustrate why asymptomatic CEA is so sensitive to operative quality. The immediate-CEA arm had an approximately 3.0% 30-day perioperative stroke or death rate, and the 5-year non-perioperative stroke rate was approximately 4.1% after immediate CEA versus 10.0% with deferred CEA medical management. These numbers are clinically useful only when the contemporary center can reproduce low perioperative risk; otherwise, the small long-term advantage is eroded at the time of operation. ACST-1 (10-year) 2010

Modern asymptomatic decision-making must also acknowledge that historical medical therapy is not contemporary medical therapy. ACAS-era and ACST-era medical management differed from current intensive vascular prevention, so the absolute benefit of prophylactic CEA should not be assumed to be identical in every current patient. The surgeon should therefore treat asymptomatic intervention as a selective prevention procedure requiring validated stenosis measurement, adequate life expectancy, low local complication rates, and a patient-specific discussion of competing risks. SVS 2022 · ACST-1 (10-year) 2010 · Endarterectomy asymptomatic carotid 1995

Comparative trials help define the modality discussion but should not blur the CEA indication itself. CREST randomized 2502 patients and found similar 4-year primary composite outcomes overall for CEA and carotid stenting, with more periprocedural stroke after stenting and more periprocedural myocardial infarction after endarterectomy. ACST-2 randomized 3625 asymptomatic patients and reported similar 5-year non-disabling stroke rates between CEA and stenting in centers meeting trial-entry standards. Second asymptomatic carotid 2021 · CREST 2010

Symptomatic stenting trials provide a cautionary contrast. EVA-3S reported a 30-day stroke or death rate of approximately 9.6% after carotid artery stenting versus 3.9% after CEA in symptomatic severe stenosis, an approximately 2.5-fold relative risk increase with stenting, and the trial was stopped early for safety. Although device era and operator experience matter, this trial remains an important reminder that avoiding a neck incision does not necessarily reduce early neurological risk. Endarterectomy versus Stenting 2006

Center-level outcomes determine whether trial evidence can be translated safely. The ESO carotid guideline advises that in-hospital stroke or death after CEA or carotid stenting should ideally remain below 2% to maintain a favorable risk-benefit profile, with site-level audit. ESVS 2023 frames approximate 30-day perioperative stroke or death caps of 6% for symptomatic CEA and 3% for asymptomatic CEA; exceeding these thresholds should trigger review of case selection, timing, technique, and operator performance. European Stroke Organisation 2021

31.3Technique and comparator boundaries

A technically good carotid endarterectomy is not a single maneuver but a sequence of choices that must be made deliberately: conventional endarterectomy with patch closure or another closure strategy, eversion versus longitudinal arteriotomy technique, routine or selective shunting, method of monitoring, and recognition of cranial-nerve risk. These choices should be documented as separate decision lanes because each has different failure modes and different strength of evidence.

The closure decision is one of the most visible technical choices. Patch angioplasty and primary closure have been studied in randomized trials and systematic review, and the 2010 Cochrane review by Rerkasem and Rothwell is the admitted evidence reference point for the patch-versus-primary-closure lane. In practice, a unit should define when patch closure is standard, when primary closure is acceptable, and how restenosis or occlusion after either approach is captured in audit. CEA 2009

Patch material is a second, distinct question. The Cochrane patch-material review included 14 randomized trials and 2278 CEA patch closures, but most comparisons had low or very low certainty because event counts were small and methods limited confidence. Thus, material selection should be protocolized and audited rather than justified by overconfident claims that one material is universally superior. Cochrane review carotid 2021

The available patch-material evidence is mixed and clinically imperfect. The review found little or no clear difference in perioperative or long-term ipsilateral stroke between vein and synthetic patch materials, while Dacron showed worse signals than other synthetic materials for combined perioperative stroke or transient ischemic attack and restenosis or occlusion. Conversely, another extracted comparison reported PTFE associated with higher long-term stroke/death than Dacron, with low-certainty evidence and wide confidence intervals. The safest conclusion is that the surgeon should know the institution’s material-specific outcomes rather than relying on a universal hierarchy. Cochrane review carotid 2021

Vein patch remains a special case because the late failure mode differs. In the Cochrane review, vein-patch closure was associated with higher pseudoaneurysm formation than synthetic patch, while perioperative ipsilateral stroke did not show a clear difference and the certainty was very low. When a vein patch is chosen, especially in settings where infection risk influences material choice, the operative note and follow-up plan should acknowledge pseudoaneurysm surveillance. Cochrane review carotid 2021

Shunting policy is another area where strong local practice often exceeds the strength of trial evidence. The Cochrane shunting review included six trials with 1270 participants and found no randomized trials directly comparing selective shunting with no shunting during CEA under general anesthesia. The review found no statistically significant difference between routine shunting and no shunting for 30-day stroke or death, with small event numbers and poor methodological reporting. Routine selective carotid 2014

The practical shunting lesson is not that shunts do not matter; it is that evidence does not establish a universally superior routine or selective policy. Small trials of routine shunting versus no shunting suggested lower early stroke-related outcomes in best-case analyses, but the Cochrane authors judged the evidence too limited to support or reject routine or selective shunting, and no monitoring method proved superior. A surgeon should therefore use a clear institutional protocol, ensure the team can insert and manage a shunt safely, and audit neurological events by shunt policy. Routine selective carotid 2014

Intra-operative monitoring should be treated as an adjunct to a policy, not a substitute for one. Because no monitoring method proved superior in the shunting review, the monitoring strategy must be chosen according to anesthetic approach, local expertise, and the ability to act on the result. A monitoring signal that does not change clamp management, shunt insertion, or technical inspection is not a meaningful safety system. Routine selective carotid 2014

Cranial-nerve injury belongs in the operative-risk discussion even though the dominant trial endpoints are stroke, death, and myocardial infarction. ESVS 2023 explicitly separates cranial-nerve injury as a technique-related decision lane, and it should be part of consent, operative exposure planning, and postoperative assessment. In the CEA-versus-stenting conversation, cranial-nerve risk is one of the open-surgery trade-offs that must be weighed against the periprocedural stroke signal seen with stenting. CREST 2010

Comparator boundaries should be stated precisely. CREST showed similar 4-year primary composite outcomes overall for CEA and stenting, but the periprocedural trade-off differed: more strokes after stenting and more myocardial infarctions after endarterectomy. The Bonati Cochrane review likewise concluded that CEA reduces periprocedural stroke compared with stenting but increases periprocedural myocardial infarction, with similar long-term ipsilateral stroke rates between modalities. Carotid artery stenting 2020 · CREST 2010

Durability is part of the comparator boundary. The Bonati Cochrane review reported lower long-term carotid restenosis rates after CEA than after carotid artery stenting across pooled randomized carotid intervention trials. This does not eliminate the need for surveillance after CEA, but it supports describing open endarterectomy as the more durable modality on the restenosis axis in the randomized evidence base. Carotid artery stenting 2020

Technique choices should be converted into a reproducible operative culture. The operative report should specify symptom status, stenosis measurement basis, closure method, patch material if used, shunt policy and whether a shunt was inserted, monitoring method if used, and any cranial-nerve or neurological concern. These details allow postoperative events, restenosis, and reintervention to be traced back to modifiable technical choices rather than treated as isolated complications. Cochrane review carotid 2021 · Routine selective carotid 2014

31.4Clinical integration, follow-up, and evidence boundaries

The post-referral workflow for recently symptomatic carotid stenosis should be built around neurological stability and access to imaging, not around a rigid day count alone. ESVS 2023 frames CEA as ideally performed within 14 days for symptomatic 50–99% stenosis, while also recognizing that hyperacute operation may carry higher perioperative risk and requires individualized review. The surgeon’s task is to identify the patient who should be accelerated, the patient who should be stabilized before operation, and the patient whose symptoms or imaging suggest another mechanism.

A practical symptomatic pathway begins with confirmation that the event was ipsilateral and carotid-territory, followed by validated duplex, CTA, or MRA assessment of stenosis severity. The European Stroke Organization guideline recommends CEA for symptomatic 70–99% stenosis and suggests CEA for 50–69% stenosis, with intervention ideally within 2 weeks when feasible. The ESVS and ESO positions should be used as complementary European framing for the same urgent-treatment concept. European Stroke Organisation 2021

The urgent pathway is justified because early recurrence can be altered by system design. EXPRESS showed that urgent assessment and treatment after TIA or minor stroke reduced early recurrent stroke compared with a prior delayed pathway. For CEA services, this supports direct communication between stroke physicians, vascular imaging, anesthesia, and the operating team, so that eligible patients are not lost between diagnosis and treatment. CEA 2007

Perioperative-risk thresholds are not abstract quality metrics; they are treatment-selection rules. For symptomatic CEA, ESVS 2023 frames an operator-level 30-day stroke or death cap of approximately 6%. For asymptomatic CEA, the cap is approximately 3%, reflecting the smaller underlying benefit in patients without recent symptoms. If a center cannot meet these thresholds, the evidence supporting CEA in otherwise eligible patients no longer applies in the same way.

In asymptomatic disease, follow-up implications begin before the operation. The surgeon should confirm that the patient’s life expectancy is sufficient to realize delayed stroke-prevention benefit and that the institutional perioperative stroke or death rate is low enough to preserve net benefit. SVS 2022 frames asymptomatic CEA for approximately 70% or greater stenosis in patients with at least 3 to 5 years of life expectancy and acceptable surgical risk, while ACST-1 shows why the 3% perioperative-risk gate matters. SVS 2022 · ACST-1 (10-year) 2010

Postoperative surveillance should be linked to the technical choices made at operation. Patch closure, primary closure, patch material, and shunt policy all create audit variables for later restenosis, occlusion, pseudoaneurysm, and neurological events. The Cochrane patch and shunt reviews do not provide certainty strong enough to mandate a single universal technique, but they do support careful documentation and center-level review of outcomes by technique. Cochrane review carotid 2021 · Routine selective carotid 2014 · CEA 2009

Restenosis findings after CEA should change future planning in two ways. First, they should prompt review of the original operation, including closure method and patch material, because the randomized patch evidence signals differences in restenosis or occlusion trends without a definitive material winner. Second, they should influence the modality discussion for reintervention, since pooled randomized evidence shows lower long-term restenosis after CEA than after carotid stenting overall, while stenting carries a different periprocedural risk profile. Cochrane review carotid 2021 · Carotid artery stenting 2020

The CEA-versus-CAS trade-off should be revisited whenever follow-up reveals recurrent stenosis, new symptoms, or new comorbidity. CREST and the Bonati Cochrane review both support the broad contrast: CEA has lower periprocedural stroke risk than stenting, while CEA carries more periprocedural myocardial infarction risk, and long-term ipsilateral stroke rates are similar. In an older symptomatic patient, particularly at age 70 years or above, ESVS 2023 and the CREST age interaction support CEA as the preferred default when surgical risk is acceptable. Editor's Choice European 2023 · Carotid artery stenting 2020 · CREST 2010

European guideline framing has evolved and should be cited with care. The 2017 ESC/ESVS peripheral arterial disease guideline includes an extracranial carotid section that provides a broader PAD-era European lane, while the 2023 ESVS carotid and vertebral guideline is the more specialized contemporary carotid source. Historical SVS 2011 guidance should be treated as prior-version context rather than current operative direction, with SVS 2022 used for present North American framing. SVS 2022 · ESC 2017 · SVS 2011

Evidence boundaries should be made explicit in operative teaching. The symptomatic high-grade benefit is robust and time-sensitive; the moderate symptomatic benefit is smaller and subgroup-dependent; the asymptomatic benefit is real only when perioperative stroke or death is very low and life expectancy is adequate; and technical choices such as shunting and patch material remain areas where local protocol and audit matter because randomized evidence is limited or low certainty. Cochrane review carotid 2021 · Routine selective carotid 2014 · Rothwell pooled CEA analysis 2004 · ACST-1 (10-year) 2010 · Randomised trial endarterectomy 1998 · Endarterectomy asymptomatic carotid 1995 · NASCET · 1991

For longitudinal practice, every carotid program should know its own denominator. The minimum dataset should include symptom status, stenosis band and measurement method, timing from qualifying event, operative technique, closure method, patch material, shunt policy, postoperative stroke, death, myocardial infarction, cranial-nerve injury, restenosis, and reintervention. This is how the trial and guideline thresholds become a living quality standard rather than a literature citation in the clinic letter. European Stroke Organisation 2021