Carotid Artery Disease

Meet the patient

A 70-year-old man walks into the vascular clinic two days after a 20-minute episode of right-hand weakness and slurred speech that resolved completely on the way to hospital. He also recalls a "curtain" descending over his left eye for a few seconds last week. Blood pressure 168 over 96, no atrial fibrillation on ECG, carotid duplex awaited. He smokes 20 a day and has not seen a doctor in years.[1]

The two questions that decide this stem are the two that decide every carotid stem: is the carotid significantly stenosed? — the duplex and CTA answer within hours — and do the benefits of CEA outweigh the peri-operative stroke risk? — the NASCET band, the time from symptom, and the unit's audited outcomes answer together. Hold those two and the rest of the page slots into place.[2]

Symptomatic stenosis, not asymptomatic, drives urgency

Carotid artery disease is atherosclerotic narrowing of the extracranial carotid arteries, almost always at the bifurcation, and the single largest surgically preventable cause of ischaemic stroke. It sits alongside small-vessel lipohyalinosis and cardioembolism as one of the three great causes of ischaemic stroke, and accounts for roughly 15 to 20 percent of all such events — one in six.[1]

The disease runs on a spectrum from asymptomatic — an incidental bruit, or stenosis found on screening or after a contralateral event — to symptomatic, defined as a focal neurological deficit of ischaemic origin in the territory of the affected carotid within the preceding six months. Symptomatic disease declares itself as transient ischaemic attack (TIA), amaurosis fugax, or a completed stroke.[1]

The clinical question is never whether to image — always image. The question is whether to intervene, by CEA, CAS, or best medical therapy alone — a decision driven by the degree of stenosis, the symptom status, and the patient's perioperative risk and life expectancy. One fact dominates the rest: the danger of a carotid plaque is embolic first, haemodynamic second. Most carotid strokes begin when an unstable plaque ruptures or ulcerates, releasing platelet-fibrin or cholesterol debris into the cerebral circulation; pure flow limitation matters only in critical "string-sign" stenosis. This embolic truth is why the magnitude of benefit from surgery depends so sharply on how recently the patient became symptomatic.[2]

Etymology, for viva gold: carotid comes from the Greek karos, "heavy sleep" — the ancient Greeks knew that compression of these arteries caused sudden syncope, which is exactly why wrestlers and executioners targeted them. The name has carried the warning for two and a half thousand years.[1]

Why the bifurcation, and why the MCA territory

The plaque sits at the bifurcation because that is where the blood stirs, and it strikes the middle cerebral artery territory because that is where the embolus lands. Both facts fall out of the geometry of flow.[1]

The carotid bifurcation is a region of turbulent, oscillatory shear stress: blood dividing into the external and internal branches creates low and oscillatory wall shear stress at the outer wall of the carotid bulb. Low shear promotes endothelial dysfunction, permits LDL infiltration and retention in the intima, and slows the clearance of adhesion molecules — the seedbed for a plaque. This is why the plaque sits at the bulb and proximal ICA, and why the external carotid and the distal ICA are usually spared.[1]

A plaque matures from a fatty streak into a fibrous-cap atheroma with a lipid-rich necrotic core. The danger is not the plaque's size alone but its stability: a thin fibrous cap over a large lipid core, with intra-plaque haemorrhage and macrophage activity, is a vulnerable plaque prone to rupture.[1]

Two mechanisms carry the danger, and their order is the whole story of why surgery is urgent:[1]

1. Artery-to-artery embolism — the commonest mechanism. The fibrous cap ruptures and exposes the thrombogenic lipid core to blood. Platelets adhere and aggregate; fragments break off and embolise distally. A platelet-fibrin embolus lodging transiently in a cortical or deep artery produces a TIA; a larger or more persistent embolus causes a completed infarct. The destination is usually the middle cerebral artery (MCA) territory — hence the motor and language signature.
2. Haemodynamic insufficiency — less common. In critical "string-sign" stenosis or acute occlusion, flow distal to the stenosis falls below the brain's autoregulatory reserve, especially if the circle of Willis is incomplete. A drop in perfusion pressure then produces a watershed infarct between arterial territories.[1]

Because the ICA supplies the anterior circulation — the eye via the ophthalmic artery, and the MCA and anterior cerebral artery via the circle of Willis — emboli from the bifurcation produce anterior-circulation symptoms: contralateral hemiparesis and hemisensory loss (face and arm more than leg, the MCA pattern), aphasia in the dominant hemisphere, homonymous hemianopia, and neglect in the non-dominant hemisphere.[1]

Amaurosis fugax is the retinal signature of the same embolic process. A cholesterol crystal embolus — the shiny, refractile Hollenhorst plaque — from the carotid bifurcation lodges transiently in a branch of the central retinal artery. The patient describes a "curtain descending" or a "shade" over one eye, lasting seconds to minutes, with full recovery. It is, for practical purposes, pathognomonic of ipsilateral carotid disease and must be treated as a carotid-territory TIA.[1]

70 to 99 percent symptomatic = CEA wins (NASCET)

The whole surgical decision turns on two axes — the degree of stenosis and the symptom status — and both are read off the NASCET method. Get the measurement right and the band falls out.[1]

The NASCET method

Stenosis is measured on angiography (conventional, CTA, or MRA) by the NASCET (North American Symptomatic Carotid Endarterectomy Trial) formula:[4]

Percent stenosis = (D-normal − D-minimal) / D-normal × 100 — where D-normal is the diameter of the distal, disease-free internal carotid artery (where the walls are parallel, beyond the bulb) and D-minimal is the narrowest residual lumen at the stenosis.[4]

The older European Carotid Surgery Trial (ECST) method used the estimated original diameter at the bulb as the denominator, which inflates the percentage for the same anatomy; the two methods agree closely only at severe stenosis. After Rothwell's pooled re-analysis reconciling NASCET and ECST data in NASCET terms, the NASCET method became the international standard — and every threshold below is a NASCET value.[3][5]

Severity bands and the decision

The severity band sets the surgical decision for both symptomatic and asymptomatic disease.[1]

[4] [8]

The classic trap: treating an asymptomatic stenosis as though it carried the same urgency as a symptomatic one. It does not. Symptomatic 70 to 99 percent disease is a surgical emergency within two weeks; asymptomatic disease is an outpatient risk-benefit conversation. Confuse the two and you either miss the window on a patient who needs theatre or operate on a patient whom modern medical therapy would have served better.[1]

The 2-week rule from TIA to endarterectomy

Timing is the single most important modifiable factor in carotid surgery, and the target is two weeks from the index symptom to the operating table. The benefit of CEA after a TIA or minor stroke is greatest in the first two weeks and declines steeply thereafter — the unstable plaque stabilises with time, and so the strokes surgery was meant to prevent have already happened. "Time is brain."[2]

The number needed to treat to prevent one ipsilateral stroke is approximately 5 if CEA is performed within 2 weeks, rising to 125 if delayed beyond 12 weeks. That single pair of numbers is the highest-yield fact in the topic.[2]

[4] [8]

Consultant confession: the two-week target is the hardest number to hit in vascular surgery, because it depends on three things none of us fully control — primary-care referral speed, duplex-slot availability, and a theatre list with a surgeon and anaesthetist whose audited outcomes sit under the ceiling. The centres that hit it are the centres that do enough volume to keep the audit honest. If your unit cannot publish a perioperative stroke-or-death rate under 6 percent for symptomatic and under 3 percent for asymptomatic disease, you should not be offering CEA — the operation will cause more strokes than it prevents.[2]

Anterior is face and arm and aphasia; posterior is diplopia and vertigo

Map the deficit to the circulation before you reach for the carotid duplex — the anterior circulation is carotid territory, the posterior is vertebrobasilar, and only the anterior earns a carotid workup.[1]

Symptomatic disease is defined as a focal neurological deficit of ischaemic origin in the territory of the affected carotid within the preceding six months, and presents in three ways:[1]

• Transient ischaemic attack (TIA) — sudden-onset focal neurological deficit lasting under 24 hours (typically minutes), with complete resolution, in the anterior (carotid) circulation: contralateral hemiparesis and hemisensory loss (face and arm more than leg), aphasia (dominant hemisphere) or neglect (non-dominant), and monocular visual loss (amaurosis fugax). Use the ABCD squared score to stratify early stroke risk after a TIA.
• Amaurosis fugax — transient monocular blindness, already described; the retinal embolic equivalent of a hemispheric TIA.
• Completed stroke — a persistent focal deficit from established cerebral infarction in the carotid or MCA territory.[1]

[1]

Asymptomatic carotid stenosis has no referable symptoms. It is detected as a carotid bruit on auscultation or as an incidental finding on duplex done for another reason — pre-coronary bypass, contralateral symptomatic disease, or screening. A bruit is a systolic, high-pitched murmur best heard at the angle of the jaw. The crucial point: the presence and loudness of a bruit correlates poorly with stenosis severity. A very tight stenosis may produce no bruit at all because flow is so reduced ("silent occlusion"), and a moderate stenosis may be loud. A bruit is a reason to image, not a measure of severity.[1]

Everyone forgets: crescendo TIA (two or more TIAs in 24 hours) and stroke-in-evolution denote a highly active, embologenic plaque and are urgent — they shorten the window for CEA, not lengthen it. Elderly and diabetic patients may report vague symptoms (limb heaviness, word-finding difficulty) that are misattributed to age; a careful history and imaging are essential.[2]

The bedside round — listen at the angle of the jaw

A focused assessment answers three questions: is this a stroke or TIA, is it carotid-territory, and is there a surgical lesion? Examination rarely makes the diagnosis — imaging does — but the bedside round sets the trajectory.[1]

Neurological examination begins with the Glasgow Coma Scale (GCS) to grade consciousness and the NIH Stroke Scale (NIHSS) — the standardised stroke severity score (0 normal, 42 deepest coma) that also stratifies thrombolysis and must be documented on every suspected stroke. A focused cortical and motor exam then maps the deficit to anterior versus posterior circulation; carotid-territory stroke favours face-and-arm weakness, aphasia, and visual field loss. Fundoscopy looks for a Hollenhorst plaque — a refractile cholesterol crystal at a retinal artery bifurcation, the retinal stigmata of carotid embolism.[1]

Cardiovascular examination centres on auscultating for a carotid bruit at the angle of the jaw along the carotid course — systolic, high-pitched, present in roughly half of significant stenoses, and whose absence does not exclude stenosis. Then pulse, blood pressure in both arms, and heart sounds: AF (irregularly irregular), murmurs (valvular source), pulse deficit (dissection), unequal BP (subclavian or aortic disease). Because carotid disease is systemic, also examine the peripheral arteries — palpate for PAD and abdominal aortic aneurysm, and listen for renal bruits. Document smoking pack-years, HbA1c, lipid profile, blood pressure, and exercise tolerance: these drive the medical therapy you will prescribe regardless of surgery.[1]

Two non-invasive images before the knife

Confirm the stenosis with two concordant non-invasive modalities before surgery; reach for the catheter only when they disagree. The vascular ladder runs duplex, then CTA or MRA, then — rarely — digital subtraction angiography.[8]

• Carotid duplex ultrasound — first-line. Non-invasive, cheap, bedside, widely available. It measures peak systolic velocity (PSV) and end-diastolic velocity (EDV) at the stenosis and in the distal ICA; velocity criteria correlate with severity (PSV over 230 cm/s and ICA/CCA PSV ratio over 4 suggest over 70 percent stenosis). It also reads plaque morphology — echolucency, ulceration, intra-plaque haemorrhage are markers of an embologenic, vulnerable plaque. It is operator-dependent and less reliable distally and at very high (near-occlusion) grades.
• CT angiography (CTA) — the usual confirmatory test after an abnormal duplex. Fast, widely available, gives anatomical detail and allows formal NASCET measurement on a 3D dataset; also visualises the aortic arch and intracranial circulation. Requires iodinated contrast (renal caution).
• MR angiography (MRA) — contrast-enhanced MRA is highly accurate and avoids radiation; useful when CTA is contraindicated and for assessing the intracranial circulation and plaque composition. Time-of-flight MRA can overestimate stenosis due to turbulence.
• Conventional (catheter) digital subtraction angiography (DSA) — the gold standard for stenosis measurement, but invasive and carries an approximately 1 percent risk of procedure-related stroke. Reserved for discordant non-invasive imaging, suspicion of unusual lesions (dissection, fibromuscular dysplasia, vasculitis), or as part of CAS.

[1] [8]

Brain imaging confirms ischaemia and excludes haemorrhage and mimics before any antithrombotic or surgical decision. Non-contrast CT brain is the first test — its job is to exclude haemorrhage (a dense area); early ischaemia may be subtle (loss of the insular ribbon, obscuration of the lentiform nucleus, the hyperdense MCA sign). MRI brain with diffusion-weighted imaging is the most sensitive test for early ischaemia and identifies small and lacunar infarcts missed by CT.[1]

Because cardioembolism is the chief alternative source, every patient with a carotid-territory event is screened for a cardiac cause: ECG for atrial fibrillation (with prolonged Holter monitoring if paroxysmal), echocardiogram for mural thrombus, vegetations, patent foramen ovale or low ejection fraction, and bloods — FBC, U and E, lipid profile, HbA1c, clotting, ESR and CRP (giant cell arteritis in the older amaurosis patient), and glucose (the hypoglycaemia mimic).[1]

Differential — when it isn't the carotid

The differential of a focal neurological deficit is wide, and the task is to separate carotid disease from its mimics and from the other causes of stroke. Two questions do most of the work: is there haemorrhage on CT, and is there a cardioembolic source?[1]

[1]

The classic trap: missing atrial fibrillation as the real cause of the stroke. AF is the commonest alternative embolic source, and anticoagulation — not endarterectomy — is the treatment. Always check the ECG, image the heart, and consider prolonged monitoring for paroxysmal AF before assuming the stenotic carotid is the culprit.[1]

Resuscitate the stroke first — carotid surgery is elective

The acute presentation of carotid disease is the stroke or TIA itself, and resuscitation is generic stroke care. Carotid surgery, if any, is elective and deferred even in symptomatic patients — it is never the immediate resuscitative intervention.[1]

• ABCDE approach; secure the airway if GCS is depressed or bulbar failure is present.[1]
• Oxygen only if hypoxic (SpO₂ under 94 percent) — routine high-flow oxygen is not beneficial and may worsen outcome.
• Blood glucose — treat hypoglycaemia (a stroke mimic) promptly; avoid hyperglycaemia (worsens outcome).
• Blood pressure — do not lower aggressively in the acute phase; permissive hypertension preserves penumbral perfusion. Lower only if over 220/120 mmHg, or if thrombolysis is planned (target under 185/110 mmHg).
• Thrombolysis — if an acute ischaemic stroke meets criteria (within 4.5 hours of onset, no haemorrhage on CT, no exclusions), intravenous alteplase 0.9 mg/kg (max 90 mg, 10 percent bolus then infusion over 60 min) is given before any carotid workup. Thrombectomy is considered for large-vessel anterior-circulation occlusion.
• Admit to a stroke unit; keep NPO until swallow assessed; give VTE prophylaxis; arrange continuous cardiac monitoring for AF.

[1]

Once the patient is stabilised and the diagnosis of a carotid-territory event confirmed, the definitive carotid question is addressed: how severe is the stenosis, and is the patient a candidate for CEA or CAS?[1]

Best medical therapy is now everyone's baseline

Best medical therapy (BMT) is the foundation for every patient, whether or not they proceed to surgery, and it has markedly lowered the stroke risk of asymptomatic carotid stenosis in the modern era. A patient on optimal BMT today faces a lower medical-arm risk than the patients randomised in ACAS thirty years ago — which is exactly why the asymptomatic-surgery debate has shifted.[1]

[1]

Everyone forgets: BMT does not stop at surgery. Even after a perfect CEA, the patient keeps the antiplatelet, the high-intensity statin, the blood-pressure target, and the smoking-cessation support for life. The plaque you did not endarterectome — the contralateral carotid, the coronaries, the aorta — still needs the medical therapy.[1]

CEA: open the artery, peel the plaque, patch the closure

Carotid endarterectomy is the open surgical removal of the atheromatous plaque from the carotid bifurcation and proximal ICA, and it remains the evidence-based standard for revascularisation. Its benefit depends on two variables: the degree of stenosis and the time from the last symptom.[4][5]

1. Anaesthesia — general or local or regional. Regional lets the patient perform tasks during clamping, detecting ischaemia in real time.
2. Position and incision — supine, head turned away; a longitudinal incision along the anterior border of sternocleidomastoid from the angle of the jaw downward.
3. Exposure — divide the facial vein; expose and control the common, external, and internal carotid arteries, slinging each. Identify and protect the hypoglossal (XII), vagus (X), and marginal mandibular branch of the facial (VII) nerves.
4. Heparinisation — intravenous heparin before clamping.
5. Clamping and cerebral protection — clamp the common, external, then internal carotid. If intra-operative monitoring shows ischaemia (or stump pressure is low, under 40 to 50 mmHg, or under regional anaesthesia the patient loses function), deploy a temporary intraluminal shunt to maintain cerebral perfusion while the artery is opened.
6. Endarterectomy — longitudinal arteriotomy from CCA across the bulb into the ICA; dissect the plaque in the subintimal plane, removing it cleanly with a feathered endpoint in the distal ICA to avoid a flap.
7. Closure — close the arteriotomy, ideally with a patch angioplasty (vein or prosthetic) to widen the lumen and reduce restenosis. Flush thoroughly to remove debris before final closure.
8. Haemostasis and closure in layers over a drain; observe for haematoma.

[1] [8]

CAS: a stent through a filter, when the neck is hostile

Carotid artery stenting (CAS) is the endovascular alternative — femoral access, an embolic protection device deployed distal to the stenosis, pre-dilatation, then a self-expanding stent across the lesion. It trades a higher peri-procedural stroke rate for a lower MI and cranial-nerve-injury rate, and it earns its place when open surgery is anatomically or clinically hostile.[7]

CAS is preferred when open surgery is high-risk: previous neck surgery or radiotherapy, recurrent stenosis after CEA, a "hostile neck", high lesions (above C2, surgically inaccessible), or contralateral laryngeal nerve palsy. The trade-off is a higher peri-procedural stroke risk than CEA, partly offset by lower MI and cranial-nerve-injury rates.[1]

[7] [8]

Asymptomatic over 60 percent: only if your hand is safer than your medicine

For asymptomatic stenosis, surgery is far more selective — the benefit is modest, the perioperative ceiling is tighter, and modern medical therapy has narrowed the gap.[6]

The original ACAS (1995) and ACST (2004) showed that CEA reduced the 5-year stroke risk from about 11 to 12 percent (medical) to roughly 5 to 6 percent (CEA) — an absolute risk reduction of about 5 to 6 percent (NNT 17) — but only if the perioperative stroke-or-death rate was under 3 percent. With modern intensive medical therapy narrowing the medical-arm risk further, the net benefit of operating on asymptomatic patients has shrunk and is individualised.[6]

Consider CEA for asymptomatic stenosis over 60 to 70 percent only in a fit patient with a life expectancy over 5 years and a surgeon and unit with audited perioperative risk under 3 percent. If those three conditions are not all met, the operation causes more strokes than it prevents.[6]

[6]

The three landmark trials — NASCET, ECST, and ACAS

The modern management of carotid disease rests on three landmark randomised trials and the pooled analysis that reconciled them. Name them, the bands they established, and their absolute risk reductions, and you have answered most of the viva.[5]

NASCET — symptomatic severe stenosis (the surgical winner)

NASCET (North American Symptomatic Carotid Endarterectomy Trial, 1991 and 1998) randomised symptomatic patients to CEA plus medical therapy versus medical therapy alone. For 70 to 99 percent stenosis, CEA reduced 2-year ipsilateral stroke from 26 percent (medical) to 9 percent (CEA) — an absolute risk reduction of 17 percent (NNT 6), durable over time. For 50 to 69 percent a more modest benefit (5-year ARR about 6 percent); under 50 percent no benefit. NASCET also fixed the measurement method — the NASCET formula became the international standard.[4]

ECST — the European confirmatory trial (cite id 3)

ECST (European Carotid Surgery Trial, 1991) ran in parallel as the European confirmatory RCT of CEA versus medical therapy in symptomatic carotid stenosis. Its interim results confirmed CEA benefit for severe (70 to 99 percent) stenosis and no benefit for mild (under 30 percent) stenosis. ECST used a different stenosis denominator — the estimated original diameter at the bulb — which inflated percentages for the same anatomy. Together with NASCET, ECST defined the surgical thresholds; the two methods were later reconciled in Rothwell's pooled analysis (2003), which re-expressed all stenosis in NASCET terms and confirmed the greatest benefit at 70 to 99 percent (NNT as low as 3 to prevent one stroke at 2 years in the first days after symptoms), a modest benefit at 50 to 69 percent, and no benefit under 50 percent.[3][5]

ACAS and ACST — the asymptomatic evidence, and why it has narrowed

ACAS (1995) and ACST (2004) randomised asymptomatic patients with over 60 percent stenosis to CEA plus medical versus medical alone. CEA halved the 5-year stroke risk (from about 11 to 12 percent down to 5 to 6 percent; ARR 5 to 6 percent, NNT 17) — but only if the perioperative stroke-or-death rate was under 3 percent. The benefit is real but modest, and because contemporary intensive medical therapy has lowered the medical-arm risk toward 1 to 2 percent per year, the net advantage of prophylactic CEA has narrowed. Modern practice is selective: reserve asymptomatic CEA for fit patients with over 60 to 70 percent stenosis, a life expectancy over 5 years, and an audited perioperative risk under 3 percent.[6]

Subtypes and scenarios that change the decision

Each subtype changes the surgical decision and must be recognised on imaging and history.[1]

• Bilateral carotid stenosis — CEA on the symptomatic side first, within 2 weeks if symptomatic. The contralateral CEA is staged 4 to 6 weeks later: simultaneous bilateral CEA risks bilateral cranial nerve injury, bilateral hyperperfusion, and haemodynamic instability.
• Carotid near-occlusion or "string sign" — a thread-like residual lumen with collapse of the distal ICA ("slim" distal ICA). The peri-procedural stroke risk is higher and the benefit of CEA less certain than for discrete 70 to 99 percent stenosis; managed in a specialist unit, often with careful DSA definition.
• Complete carotid occlusion — no lumen to repair, so CEA is not useful. Medical therapy; chronic occlusion carries an annual stroke risk of 2 to 5 percent (higher if symptomatic). Extracranial-to-intracranial bypass is investigational.
• Carotid dissection — younger patient, neck pain, Horner's syndrome (ptosis, miosis, anhidrosis), trauma or connective-tissue disease. Managed medically (antiplatelet or anticoagulant); stenting reserved for high-risk or evolving deficit.
• Post-CEA restenosis — myointimal hyperplasia within 2 years (smooth, non-atherosclerotic). CAS is often preferred over re-do CEA (hostile scarred neck).
• Carotid body tumour (paraganglioma) — a differential of a neck mass at the bifurcation; a separate entity, not atherosclerotic, but anatomically relevant.
• Crescendo TIA or stroke-in-evolution — denotes a highly active embologenic plaque; treat as urgent, with CEA within days if stenosis is severe and the deficit is not devastating.

[1] [2]

Named traps — the recurring trainee errors

These are the errors that lose benefit or cause harm. Name them and you will not make them.[1]

• Treating asymptomatic stenosis like symptomatic — asymptomatic disease is an outpatient risk-benefit conversation, not a surgical emergency. Confuse the two and you either miss the window or operate unnecessarily.
• Missing atrial fibrillation as the real cause — AF is the commonest alternative embolic source; always check the ECG and image the heart before assuming the stenotic carotid is the culprit. Anticoagulation, not endarterectomy, is the treatment.
• Ignoring peri-operative stroke risk — if the unit's audited rate exceeds 6 percent symptomatic or 3 percent asymptomatic, the surgery causes more strokes than it prevents. The audit is non-negotiable.
• Delaying CEA beyond 2 weeks — benefit declines sharply with time; the strokes surgery was meant to prevent have already happened. NNT 5 within 2 weeks versus 125 beyond 12 weeks.
• Treating a bruit as a measure of severity — a bruit mandates imaging, not a percentage; a tight stenosis can be silent and a moderate one loud.
• Lowering blood pressure aggressively in acute stroke — permissive hypertension protects the penumbra; lower only if over 220/120 mmHg, or for thrombolysis.
• Forgetting BMT in surgical patients — even after CEA, antiplatelet, statin, BP control, and smoking cessation are lifelong.

[1] [2]

Complications — of the disease and of the operation

Carotid disease harms the patient directly through stroke, and the operation harms them too — chiefly through peri-operative stroke, myocardial infarction, and cranial nerve injury. The unit's audited peri-operative stroke-or-death rate is the central metric: above the ceiling, the surgery causes more strokes than it prevents.[1]

The recognised complications of CEA, in order of clinical importance:[1]

• Perioperative stroke or death — the principal risk, from intraoperative embolisation (plaque manipulation), hypoperfusion during clamping (mitigated by shunting), or postoperative thrombosis or embolism. The audited rate must be under 6 percent symptomatic and under 3 percent asymptomatic.
• Myocardial infarction — the commonest cause of perioperative death; these patients have coronary disease too. Continue statin and beta-blocker, optimise cardiac status.
• Cranial nerve injury — hypoglossal (XII) (tongue deviation to the operated side, dysarthria), vagus or recurrent laryngeal (X) (hoarseness, vocal cord palsy), and the marginal mandibular branch of VII (mouth droop). Usually transient; bilateral injury (for example staged surgery) can compromise the airway.
• Wound haematoma — can expand rapidly and compromise the airway. This is a true surgical emergency: open the wound at the bedside to decompress before returning to theatre.
• Cerebral hyperperfusion syndrome — days after revascularisation: severe headache, seizures, focal deficit, hypertension, and risk of intracerebral haemorrhage from restored flow exceeding autoregulation. Manage with tight BP control and seizure treatment.
• Restenosis — myointimal hyperplasia (early, under 2 years) or recurrent atherosclerosis (late). Reduced by patch closure. Symptomatic restenosis is often treated by CAS.
• Bleeding, infection, hypertensive crisis (carotid sinus manipulation) — perioperative.

[1] [8]

CAS carries its own distinct risk profile centred on embolic and access complications:[7]

• Peri-procedural stroke — higher than CEA, from embolic debris during wire and stent manipulation despite embolic protection devices.
• Bradycardia or hypotension — carotid sinus stimulation during balloon inflation; have atropine ready.
• Contrast nephropathy, access-site complications (groin haematoma), and in-stent restenosis or thrombosis.

[7]

The post-CEA wound haematoma — recognise the airway emergency

A patient who is recovering well after CEA develops neck swelling, discomfort, hoarseness, or stridor in the first hours. This is a wound haematoma expanding under tension until it compresses the airway. Do not wait for a CT and do not attempt difficult intubation in a distorted airway. Open the wound at the bedside to release the haematoma and decompress the airway, then return to theatre for haemostasis and re-closure. The classic error is to delay while arranging imaging.[1]

Special populations

The surgical choice and perioperative risk shift with the patient, so revascularisation is individualised rather than protocolised.[1]

• Elderly (over 70) — higher peri-procedural stroke risk with CAS, so CEA is preferred (the CREST age effect). Overall operative risk is higher from comorbidities, so selection is careful; age alone is not a contraindication.
• Women — smaller carotid arteries and possibly higher perioperative risk; the benefit of CEA in asymptomatic women is debated and generally more selective.
• Diabetics — higher stroke risk and worse outcomes; aggressive risk-factor control. Check renal function before CTA contrast.
• Renal impairment — iodinated contrast for CTA is a risk; prefer duplex first, and consider MRA or pre-hydration and contrast-sparing protocols.
• Anticoagulated patients (AF, valves) — manage warfarin or DOAC peri-CEA: bridge or withhold as appropriate. AF remains the alternative embolic source to exclude.
• Previous neck surgery or radiotherapy — "hostile neck": CAS preferred over re-do open surgery.
• Young patients (under 50) — suspect non-atherosclerotic causes: carotid dissection, fibromuscular dysplasia, vasculitis, hypercoagulable state; investigate accordingly. CAS is more acceptable in younger patients needing revascularisation.

[1] [7]

Regional guidelines and the live controversies

The two-week target is a global consensus, but the asymptomatic threshold and the CEA-versus-CAS trade-off still vary by region and by patient.[8]

Two live debates shape contemporary practice. Asymptomatic stenosis and modern BMT: with contemporary intensive medical therapy lowering the medical-arm stroke risk to 1 to 2 percent per year, the net benefit of prophylactic CEA has narrowed; some centres no longer operate routinely on asymptomatic stenosis, reserving surgery for high-risk plaque features (echolucent, progressing, silent infarction on imaging). Ongoing trials (ACST-2, ECST-2) inform this debate. CEA versus CAS: the peri-procedural trade-off (stroke with CAS versus MI and nerve injury with CEA) is individualised by age and anatomy; there is no single best procedure for all.[1]

The mantra, and the mnemonic

[3] [4] [3] [5] [6]

The mantra: symptomatic severe stenosis, CEA within two weeks, on best medical therapy.[2][4]

Ward-round test — three stems

Stem 1 — symptomatic 80 percent stenosis, two days after a TIA (answer)

A 70-year-old man has a 20-minute episode of right-hand weakness and slurred speech resolving fully, with a curtain over the left eye the week before. Duplex and CTA show 80 percent left ICA stenosis by NASCET; ECG is in sinus rhythm; CT brain shows no infarct. What is the management, and what is the target? Model: This is a symptomatic carotid-territory TIA with severe (70 to 99 percent) stenosis. Start best medical therapy immediately — aspirin 75 mg (or aspirin plus clopidogrel for the first 21 days after high-risk TIA), atorvastatin 80 mg, BP control, smoking cessation — and refer urgently for carotid endarterectomy within 2 weeks of the index symptom. The benefit is large (ARR 17 percent, NNT 6; NNT 5 if done within 2 weeks versus 125 if delayed beyond 12). Confirm the unit's audited perioperative stroke-or-death rate is under 6 percent for symptomatic disease. Amaurosis fugax is a carotid-territory TIA, not a separate entity — it carries the same urgency.[4][2]

Stem 2 — asymptomatic 60 percent stenosis on a screening duplex (answer)

A 72-year-old woman has a carotid duplex before coronary bypass that shows 60 percent right ICA stenosis by NASCET. She has never had a focal neurological deficit. What is the management? Model: This is asymptomatic carotid stenosis — do not treat it like symptomatic disease. Start best medical therapy (antiplatelet, high-intensity statin, BP control, smoking cessation) and assess the three conditions that would justify CEA: stenosis over 60 to 70 percent (she is at the threshold), a life expectancy over 5 years, and a surgeon and unit with audited perioperative stroke-or-death rate under 3 percent. If all three are met, selective CEA halves the 5-year stroke risk (from about 11 to 12 percent down to 5 to 6 percent; ARR 5 to 6 percent, NNT 17) — but modern BMT has narrowed this benefit, and many centres now reserve asymptomatic CEA for high-risk plaque features. The decision is individualised, not protocolised.[6]

Stem 3 — new deficit two days after CEA (answer)

A 68-year-old man is two days post right-sided carotid endarterectomy for symptomatic 85 percent stenosis. The nurse calls you because he has developed new left-arm weakness and the wound looks tense and swollen. What are the two emergencies, and what do you do in the next five minutes? Model: Two emergencies are competing. First, the tense swollen wound is a haematoma threatening the airway — open the wound at the bedside immediately to decompress before it obstructs, then return to theatre for haemostasis; do not wait for imaging. Second, the new left-arm weakness is a peri-operative stroke until proven otherwise — urgent CT brain to exclude haemorrhage (especially if hyperperfusion syndrome is brewing) and vascular review for thrombosis or embolism at the endarterectomy site. Address the airway first, then the brain. Both are time-critical; neither tolerates delay for a scan.[1][8]

References

1. [1]Bonati LH, Jansen O, de Borst GJ, Brown MM. Management of atherosclerotic extracranial carotid artery stenosis Lancet Neurol, 2022.PMID 35182512
2. [2]Naylor AR. Time is brain! Surgeon, 2007.PMID 17313125
3. [3]European Carotid Surgery Trialists' Collaborative Group. MRC European Carotid Surgery Trial: interim results for symptomatic patients with severe (70-99%) or with mild (0-29%) carotid stenosis. European Carotid Surgery Trialists' Collaborative Group Lancet, 1991.PMID 1674060
4. [4]Barnett HJ, Taylor DW, Eliasziw M, et al. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators N Engl J Med, 1998.PMID 9811916
5. [5]Rothwell PM, Eliasziw M, Gutnikov SA, et al. Analysis of pooled data from the randomised controlled trials of endarterectomy for symptomatic carotid stenosis Lancet, 2003.PMID 12531577
6. [6]Halliday A, Mansfield A, Marro J, et al. Prevention of disabling and fatal strokes by successful carotid endarterectomy in patients without recent neurological symptoms: randomised controlled trial Lancet, 2004.PMID 15135594
7. [7]Brott TG, Hobson RW 2nd, Howard G, et al. Stenting versus endarterectomy for treatment of carotid-artery stenosis N Engl J Med, 2010.PMID 20505173
8. [8]Naylor R, Rantner B, Ancetti S, et al. Editor's Choice - European Society for Vascular Surgery (ESVS) 2023 Clinical Practice Guidelines on the Management of Atherosclerotic Carotid and Vertebral Artery Disease Eur J Vasc Endovasc Surg, 2023.PMID 35598721