Carotid endarterectomy and carotid stenting both reopen a narrowed carotid artery, but they are not interchangeable. Surgery removes the plaque through an incision in the neck; stenting leaves the plaque in place and holds it against the artery wall with a metal mesh delivered by catheter. For a typical patient with symptoms and severe stenosis, surgery remains the better-supported option. For younger patients, and for many without symptoms, the tradeoffs are closer.

The mechanism: two different ways to protect the same brain

In endarterectomy, the surgeon opens the artery, lifts out the plaque, and sews a patch into the wall. The plaque is gone. The wall is reshaped. Blood flows through a clean lumen, and the artery keeps its natural elasticity and diameter. The neck incision is the cost of that completeness.

Stenting works by a different principle. A balloon presses the plaque outward, then a metal mesh holds it against the wall. The plaque stays, the wall stretches, and a foreign surface now lines the lumen. That surface is a clot magnet. That is why every stent patient needs antiplatelet therapy, and why the timing of that therapy is part of the procedure itself.

The two mechanisms explain the two risk patterns. Stenting concentrates risk in the minutes around the procedure, when plaque fragments can break off and travel to the brain. Endarterectomy concentrates risk in the days after it, when the neck wound can bleed or a clot can form at the repair site.

Quantified context: what the trials and registries show

The large randomized trials define the landscape. The North American Symptomatic Carotid Endarterectomy Trial (NASCET) and the European Carotid Surgery Trial (ECST) established surgery's benefit for symptomatic patients with severe stenosis. In NASCET, surgery roughly halved the risk of ipsilateral stroke in patients with 70 to 99 percent stenosis over two years. Those trials also showed harm in patients with near-occlusion and in patients below the benefit threshold.

The Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) compared the two head to head. The composite of stroke, heart attack, and death did not differ significantly between groups. But the components split. Stenting had more strokes, while surgery had more heart attacks. In CREST, periprocedural stroke risk was higher with stenting, and the difference matters more in older patients.

Age is the sharpest divider in CREST. Patients 70 and older did better with surgery. Patients younger than 70 did better with stenting. This was a subgroup finding, but it has been consistent enough that guideline writers reference it.

A practical frame for the numbers: both procedures carry a periprocedural stroke risk measured in a few percentage points, not fractions of a point. Whether that risk is acceptable depends on what it prevents. For an asymptomatic patient with 75 percent stenosis, the yearly stroke risk without intervention is on the order of a few percent, so a procedural risk of similar size must be justified by years of protection.

One unit system throughout this article is US customary and percent stenosis as used in NASCET-style measurement.

Edge cases: when the standard comparison breaks down

The classic comparison assumes a patient who can have either procedure. That assumption fails in several recognizable situations.

Restenosis after prior surgery changes the calculus. A second neck operation on a scarred field is technically harder, and its complication rates are higher. Stenting avoids the scarred neck. Whether it is the safer choice depends on the anatomy, not the theory.

Radiation-induced carotid disease is another exception. These arteries are fibrotic and poorly suited to surgery. Stenting is often preferred because the field is hostile to dissection.

A "hostile neck" is shorthand for anatomy that makes open surgery risky. Prior neck dissection, a tracheostomy, or a short thick neck all qualify. In these patients, stenting may be the only reasonable option.

Tandem lesions are a third case. When stenosis exists at the carotid origin and also in the intracranial circulation, the treatment strategy changes, and some patients are treated with stenting plus intracranial clot removal rather than surgery.

Near-occlusion is a trap. An artery that is nearly closed but still patent behaves differently from a critical stenosis. In the trials, surgery showed harm in some of these patients, so "70 percent or worse" is not a simple trigger.

The counterintuitive professional caveat: a higher stenosis percentage does not always mean more benefit. The relationship between stenosis severity, benefit, and harm is a threshold relationship, not a straight line.

Misconceptions: what the marketing and the short explainers get wrong

The first misconception is that the newer procedure must be the gentler one. Stenting is less invasive at the neck, but it is not less risky to the brain. The procedural stroke risk with stenting is real and, in older patients, higher than with surgery.

The second is that "less invasive" means "no recovery." Stent patients still need access-site care, antiplatelet adherence, and follow-up imaging. The recovery profile differs, but it is not zero.

The third is that surgery is obsolete. For symptomatic severe stenosis in a fit patient, surgery has the strongest evidence base of any option. Newer technology does not automatically displace older evidence.

The fourth is that both procedures are interchangeable for everyone. Age, anatomy, symptom status, and prior treatment all shift the balance. A single "best" procedure is a simplification that fails at the bedside.

The fifth is that the choice is purely medical. Operator volume and institutional experience matter. A center with excellent surgical results and modest stent experience will produce different outcomes than the reverse.

Practical actions: what actually happens, and the thresholds that drive it

Start with the symptom question. Symptomatic patients with severe stenosis are the group with the clearest surgical benefit. Asymptomatic patients need a more individualized discussion, because the absolute stroke risk is lower.

Know the measurement. Stenosis percentage depends on the method used to calculate it. NASCET-style measurement compares the narrowest point to the normal distal internal carotid artery. A number without a method is not a threshold.

Check the operator and the center. Ask for periprocedural stroke rates and volumes, not just procedure availability. The same operation in different hands has different outcomes.

Red flags that should change the conversation:

A stenosis percentage given without the measurement method used. A stenting recommendation for a patient over 70 without discussion of the stroke-versus-heart-attack tradeoff. Antiplatelet therapy planned for a stent patient without a clear start and stop plan. A recommendation for either procedure in an asymptomatic patient with no discussion of the absolute risk. A near-occluded artery described as if it were an ordinary severe stenosis.

If the patient is a candidate for both, the practical sequence is: confirm symptoms, confirm the stenosis measurement, review age and anatomy, then compare local outcomes. That sequence produces a defensible choice.

The procedures on a timeline

Phase Endarterectomy Stenting
Access Neck incision Groin or radial catheter
Plaque Removed Left in place, held by mesh
Periprocedural brain risk Concentrated at surgery Concentrated at device deployment
Main early complication Neck bleeding, nerve injury Stroke, access-site problems
Antiplatelet need Short-term, individualized Mandatory dual therapy for a defined period
Recovery emphasis Wound care, activity Access-site care, adherence
Follow-up focus Wound, restenosis over years In-stent narrowing, adherence
Time point What is evaluated
Before the procedure Symptom status, stenosis measurement, anatomy, operator outcomes
Day of procedure Periprocedural stroke and cardiac events
First 30 days Early complications, antiplatelet adherence for stents
Months 1 to 6 Wound healing, access-site healing, symptom recurrence
Year 1 and beyond Restenosis or in-stent narrowing, stroke-free survival

Limitations: what the evidence cannot settle

The trials enrolled specific populations. Patients with severe heart failure, recent heart attack, or complex anatomy were underrepresented. The results apply best to the patients the trials studied.

Subgroup analyses are weaker than headline results. The age split in CREST is a subgroup finding. Guidelines cite it, but subgroup findings can weaken when tested again.

Technology has moved. Embolic protection devices and newer stent designs were not uniform across the trials. Some of the stent risk in older trials may not carry over to current practice.

Operator skill is a confounder no trial can eliminate. A well-run center and a poorly run center produce different outcomes from the same procedure.

Long-term comparisons beyond a few years are limited. Most of the evidence covers the periprocedural period and the first few years, not decades.

The honest conclusion, and the one this article delivers, is structural rather than rhetorical: for symptomatic severe stenosis in a fit patient, surgery has the stronger evidence base. For younger patients and selected asymptomatic patients, stenting is a defensible alternative. Every recommendation should be tied to a measured stenosis, a named operator outcome, and an explicit discussion of absolute risk. When any of those three is missing, the comparison is not being made, it is being assumed.