The treatment of symptomatic carotid artery stenosis (CAS) has been evolving over many decades. From being a condition with dismal outcomes, CAS management turned a new leaf with the successful conduct of the first carotid surgery by the doyen of vascular surgery himself (DeBakey, 1953 [1]).

Symptomatic CAS, a hitherto medical disease, was now being managed by surgical carotid endarterectomy (CEA) which marked a sea change, making it the first “stroke prevention procedure”. Gradually, CEA started to gain widespread acceptance. The number of CEA procedures being performed in the USA increased from 350/year in 1961 to 34,000/year by 1976 [2], a whopping 100-fold increase! However, this had its drawbacks too; a retrospective analysis from 1977 pegged the perioperative stroke and mortality rate at 21.1% [3], and clearly checks and balances had to be put in place at this stage and alternative options were simultaneously being explored.

Endovascular treatment of CAS was the next logical step and Mathias had been successful in performing the first carotid balloon angioplasty in 1981 [4]. This was soon followed by adoption of stents drawing parallel from coronary intervention. In the late 1990s, carotid artery stenting came up in a big way and a large number of interventionists started performing stenting in all symptomatic/asymptomatic patients of  > 60% stenosis. Naturally, stenting was compared to its predecessor-CEA in the form of various randomized trials (RCTs).

The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS) compared CEA to stenting and found equivalence in terms of stroke outcomes [5]. The Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) study actually favoured stenting in high-risk patients [6].The International Carotid Stenting Study (ICSS) too reported similar stroke prevention outcomes between the surgery and endovascular groups [7]. Clinical equipoise, thus established between both groups, was the beginning of a turf war with each side having their proponents who promulgated the advent of newer techniques in open as well as endovascular surgery.

Evolution of open surgery (CEA)

The technique of CEA has undergone many modifications after its first report whereby a longitudinal arteriotomy was utilized for endarterectomy followed by primary closure using silk sutures [1]. Neuroprotection was the next big challenge, shunts were devised to maintain perfusion during surgery, and soon shunting gained universal acceptance. As its use increased, proponents started to realize that its use was not without complications. Selective shunting became the logical alternative [8]. Studies suggested that an internal carotid artery (ICA) stump pressure of  > 50 mmHg was adequate to maintain cerebral perfusion during clamping and shunting could be avoided in this scenario [9]. Other investigators proposed intentional elevation of mean arterial pressure (MAP) to  > 20% above baseline to optimize collateral cerebral blood flow and reduce risk of ischemic stroke [10]. In this direction, a recent publication from a multicenter Russian registry also suggested that pharmacologically increased blood pressures (systolic blood pressure (BP) 180–200 mmHg) was cerebroprotective especially in the setting of low ICA stump pressures [11]. Additional parameters to monitor adequacy of perfusion were simultaneously being explored. Somatosensory evoked potentials (SSEP)/transcranial Doppler (TCD) and near-infrared spectroscopy (NIRS) were being studied in their utility to reflect adequacy of cerebral perfusion. Recent publication from a multicenter Russian registry also suggested that pharmacologically increased blood pressure (systolic BP 180–200 mmHg) was cerebroprotective especially in the setting of low ICA stump pressures.

In this bouquet of choices, it started to become evident that if CEA could be done under local anesthesia, clinical neurologic monitoring of a conscious patient could be taken as an indicator of cerebral perfusion [12]. While the results from the General Anesthesia vs. Local Anesthesia for Carotid Surgery (GALA) trial [13] suggested equivalence in terms of adverse events when CEA under general anesthesia (GA) was compared to local anesthesia (LA), the latter option afforded the opportunity to talk to the patient during clamping in order to assess neurological integrity, which could not be done under GA. Increasingly, CEA under LA began to be espoused at all major centers the world over including ours to the extent that it is the default choice at our institution.

One important focus was arterial stenosis post primary closure. Autologous vein patch closure was proposed as a measure to counter this issue; systematic reviews documented that patch angioplasty was associated with a reduced risk of stroke as compared to non-patched closure [14]. Standard guidelines, including the most recent ones, have suggested that wherever conventional CEA is being performed, patch closure is superior to primary repair (European Society for Vascular Surgery (ESVS) 2023 [15]). A variety of substitute materials were also being promulgated in cases where autologous veins were suboptimal with reports suggesting equivalent results. It was thus evident that patch was sine qua non to conventional CEA and its absence was detrimental to long-term patency.

Although patching did improve patency, it added additional foreign material to the repair site with necessary prolongation of clamp time which had its own complications. To solve this issue, another school of thought had proposed a different method of CEA wherein patching was not necessary—this was the eversion endarterectomy (EEA). As the name suggests, the ICA was transected at the bifurcation and everted over in order to “deliver” the cylindrical plaque after which it was resewn onto the bulb. Once this technique gained traction, there was a significant switch from patch CEA to EEA propagated by Dhiraj Shah and his group [16]. This found favour in most of the European countries except the UK which continued with the standard repair using a patch.

Another adaptation was the modified eversion endarterectomy (MEE)—the arteriotomy was limited to the bulb and the ICA plaque was fished out much like EEA—thus it was eversion without transection. It offered the best of both worlds—adequate plaque clearance without need to restitch the ICA, also the limited incision was amenable to primary closure—therefore shorter clamp times. These advantages translated into better real-world outcomes as demonstrated by own experience [17] and that of others [18].

Evolution of carotid stenting

The technique of carotid stenting has also evolved over the last two decades in terms of using embolic protection devices (EPDs) and the type of stents being used for the same. After its introduction as the able competitor to CEA in high-risk patients unfit for surgery, CAS didn’t perform as expected in terms of complications. Practically, poor patient selection with rampant stenting in asymptomatic patients on finding incidental stenoses (the so-called oculostenting reflex) caused the procedure to be mired in controversy. With coming of age, streamlining of indications, increased operator experience, and standardization of technique (e.g., abandonment of predilation and balloon-expandable stents), CAS once again found its foothold. The most important contributor in its revival was the use of the EPDs as iatrogenic embolization causing procedural stroke was the Achilles heel for CAS.

Distal filter-type EPDs, like the SpiderFx and the Emboshield Nav 6, were the most commonly used as the learning curve was relatively shorter and deployment/retrieval were easier. The filter-type EPDs, when retrieved, inevitably showed trapped debris which could lead to stroke, if embolization occurred, and therefore their use was logically justified. EPDs eventually gained widespread acceptance to the extent that protected CAS has become the norm rather than the exception with results from multiple registries providing sufficient evidence for their use [19].

The stent design has undergone many modifications in terms of configuration and architecture. Generally, carotid stents are divided into either open cell design (large metal free pores, fewer strut interconnections) or closed cell configurations (smaller metal-free area and dense interconnections) although with the advent of hybrid and dual layer mesh stents, the architectural lines have started to blur. With regard to choice of stents, guidelines advice operator discretion [15], although certain factors favour one over the other—e.g., in high-risk plaques, the closed cell design provides protection from “cheesewiring” and distal plaque embolization while open cell stents are preferred for their flexibility in tortuous anatomy.

Despite these advances, CAS, even with filter protection, was not free from complications. The Carotid Revascularization Endarterectomy versus Stent Trial (CREST), one of the largest RCTs comparing protected CAS performed by experienced operators to CEA, still found slightly higher periprocedural stroke rates in CAS compared to CEA [20]. Incapacitating or even lethal stroke, was theorized to be caused possibly due to hardware navigation through an unfavourable aortic arch despite the use of EPDs. To offset these problems, a new hybrid technique was introduced—transcarotid revascularization (TCAR). This involved exposure of the supraclavicular common carotid artery (CCA) via a small neck incision with the introduction of a flow reversal filter system enabled by establishing an extracorporeal arteriovenous circuit shunting blood away from the CCA to the femoral vein via a filter in order to convincingly reduce procedural embolic stroke. TCAR has shown promising results [21], but the significant cost and expertise/facilities required to perform a carotid hybrid procedure may preclude its use in developing countries such as ours. Nonetheless, with the speed with which innovation is occurring in the field of carotid intervention, it is clear that the next decade is going to witness unprecedented progress on this therapeutic front.

What is the way forward?

The most recent ESVS 2023 carotid guidelines provide a detailed overview of this disease and give recommendations on evidence-based management of different scenarios pertaining to this condition.

At present, the procedural risk of CCA/EEA is less than 3% in terms of stroke or neurological damage whereas the same with CAS is still higher [20]. Surgery still continues to be a gold standard as the results are consistently better as compared to CAS over many decades. Carotid stenting requires a much higher learning curve to get optimum results and is still limited by periprocedural stroke despite use of filters [22]. Due to these reasons and on the basis of evidence, the ESVS recommends CEA in symptomatic patients with  > 70% stenosis with a role for CAS if a patient is unfit for surgery (Fig. 1).

Fig. 1
figure 1

Management of patients with carotid artery stenosis-ESVS 2023 guidelines

Full size image

Patients with other comorbidities like unstable coronary artery disease (CAD) and left ventricular (LV) dysfunction and patients with local factors like previous neck radiation are best managed by CAS. Restenosis after CEA is also a candidate for CAS if one wants to avoid re-exploration of a surgically scarred neck. TCAR is another lucrative option although RCT level comparison with either CAS or CEA has not yet been published.

Asymptomatic patients usually are the ones where screening has identified lesions causing concern (e.g., pre-coronary artery bypass (CABG) duplex testing). As detailed earlier, asymptomatic patients should only undergo intervention if the benefits outweigh the risks and that threshold has been shown by evidence to be at a stenosis  > 70% in a patient with a reasonable (> 5 years) life expectancy. Practically, asymptomatic patients are not benefitted from intervention at lower degrees of stenosis and therefore discretion is especially important in this subgroup.