Carotid Artery Stenting Outcomes in Comprehensive Stroke Hospitals (CASSH): A Prospective Multicenter Study

Abstract

BACKGROUND:

The CASONI study (Carotid Artery Stenting Outcomes by Neurointerventional Surgeons) showed that proceduralist experience significantly reduces complications in carotid artery stenting. The CASSH study (Carotid Artery Stenting Outcomes in Comprehensive Stroke Hospitals) prospectively evaluates real-world carotid artery stenting outcomes by fellowship-trained neurointerventionalists at comprehensive stroke centers across the United States to validate and expand on CASONI’s findings.

METHODS:

CASSH is a multicenter, prospective observational study conducted across 15 US comprehensive stroke centers from January 2023 to December 2024. Adults with symptomatic ≥50% or asymptomatic ≥70% carotid stenosis undergoing carotid artery stenting by fellowship-trained neurointerventionalists were included. The primary outcome was a 30-day composite of procedure-related death, stroke, or myocardial infarction. Secondary outcomes included nonprocedural mortality, access site complications, stent thrombosis, and other adverse events. Logistic regression identified predictors of adverse outcomes.

RESULTS:

Among 889 patients (mean age 70.3±9.9 years; 61.4% male), 87.1% had hypertension and 63.1% were symptomatic. The 30-day composite primary outcome occurred in 1.2% (mortality 0.8%, ischemic stroke 0.3%, hemorrhagic stroke 0.2%, myocardial infarction 0.2%). Composite secondary outcome occurred in 5.4%, most commonly access site complications (1.7%) and nonprocedural mortality (1.5%). Higher preprocedural modified Rankin Scale (odds ratio [OR], 1.42), National Institutes of Health Stroke Scale score (OR, 1.09), and longer fluoroscopy times (OR, 1.02) were associated with increased complication risk. Mortality was independently predicted by elevated modified Rankin Scale (OR, 1.72), higher National Institutes of Health Stroke Scale score (OR, 1.15), older age (OR, 1.05 per year), and lower ejection fraction (OR, 0.96). Postprocedural antiplatelet therapy was protective, reducing both complications (OR, 0.03) and mortality (OR, 0.07).

CONCLUSIONS:

Carotid artery stenting performed by fellowship-trained neurointerventionalists at comprehensive stroke centers is associated with low rates of periprocedural stroke, myocardial infarction, and death. These outcomes align with the landmark CREST-2 trial (Carotid Revascularization and Medical Management for Asymptomatic Carotid Stenosis Trial), particularly in asymptomatic patients, and are strongly influenced by preprocedural disability, stroke severity, age, and cardiac function, underscoring the importance of patient selection and optimized perioperative care.

CLINICAL PERSPECTIVE.

What Is New?

• Carotid artery stenting performed by fellowship-trained neurointerventionalists at comprehensive stroke centers is associated with low rates of periprocedural stroke, myocardial infarction, and death.

• Higher preprocedural disability (modified Rankin Scale), stroke severity (National Institutes of Health Stroke Scale), older age, and reduced cardiac function are independently associated with adverse outcomes after carotid artery stenting.

What Are the Clinical Implications?

• Careful patient selection based on preprocedural disability, stroke severity, age, and cardiac function is critical to optimize outcomes in carotid artery stenting.

• Carotid artery stenting performed by a fellowship-trained interventionist at high-volume centers may improve procedural safety and reduce periprocedural complications.

• Adherence to appropriate postprocedural antiplatelet therapy should be prioritized to minimize adverse events and enhance patient safety.

Ischemic stroke remains a leading cause of morbidity and mortality, with carotid artery atherosclerosis accounting for 20% to 25% of stroke incidence.¹ There are several modalities of carotid artery revascularization, including carotid endarterectomy (CEA), transcarotid artery revascularization, and carotid artery stenting (CAS). Advances in endovascular technology, improved training, and refined techniques have increased the feasibility and safety of CAS.²

The role of proceduralist experience in reducing complications of CEA has been established in the literature, with a learning curve required to decrease periprocedural complications.^3,4 Multidisciplinary additional training for neurointerventionalists after residency is focused solely on cerebrovascular pathology. This allows these physicians to be well-versed in carotid anatomy, crossing high‐risk lesions, and deploying embolic protection devices, in addition to providing experience with emergent thrombectomy, which offers a unique perspective. The increased familiarity with the technical demands and clinical nuances of cerebrovascular interventions makes neurointerventionalists particularly well-suited to perform CAS.⁵ A recent multicenter retrospective study, CASONI (Carotid Artery Stenting Outcomes by Neurointerventional Surgeons), demonstrated the safety of CAS performed exclusively by fellowship-trained neurointerventionalists, with a 30-day complication rate of 1.8%.⁶

The growing body of CAS literature often aggregates outcomes from a wide range of specialties. Neurointerventionalists remain underrepresented in large-scale trials and national registries,^7,8 which may impact the results of those studies. The unique contributions and potential advantages offered by specialists with focused expertise in both stroke care and endovascular techniques may also be obscured.

To address gaps in existing literature, the CASSH study (Carotid Artery Stenting Outcomes in Comprehensive Stroke Hospitals) was developed as a multicenter prospective registry of 15 comprehensive stroke centers (CSCs) in the United States. A prospective multicenter study will help further validate and expand on earlier findings.

Methods

The data that support the findings of this study are available from the corresponding author on reasonable request. This study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting observational research.

Study Design and Setting

CASSH is a prospective, multicenter observational study evaluating outcomes of CAS performed by neurointerventionalists at CSCs across the United States. Baseline patient data, including demographics, procedural details, and outcomes, were extracted from electronic health records across 15 CSCs from January 2023 to December 2024. A list of the contributing centers is provided in Supplemental Material S1. Procedures and testing followed the standard of care at the participating hospital.

Patient Selection

Eligible patients were adults aged >18 years undergoing CAS for carotid artery stenosis secondary to atherosclerosis. Symptomatic stenosis was defined by the presence of anterior circulation ischemic events (eg, transient ischemic attack, amaurosis fugax, retinal artery occlusion, or ischemic stroke) occurring within 180 days of evaluation and confirmed to be associated with ≥50% stenosis per NASCET criteria (North American Symptomatic Carotid Endarterectomy Trial).⁹ Asymptomatic stenosis was defined as ≥70% stenosis in the absence of clinical events within 6 months before their CAS. Key exclusion criteria included: life expectancy <5 years, pregnancy, contraindication to antiplatelet or anticoagulation therapy, intracranial hemorrhage on initial imaging, and CAS performed for nonatherosclerotic causes, such as trauma, dissection, carotid web, or tandem lesions undergoing acute intracranial procedures with concomitant emergent CAS.

Data Collection and Definitions

Data were prospectively collected on patient demographics (site, sex, age, race), comorbid conditions (eg, hypertension, diabetes, atrial fibrillation, hyperlipidemia, heart failure, prior stroke or myocardial infarction [MI], peripheral artery disease, chronic kidney disease, smoking and alcohol use), and baseline clinical metrics including body mass index, ejection fraction, preprocedural modified Rankin Scale (mRS) score, National Institutes of Health Stroke Scale (NIHSS) score, and P2Y12 reaction unit. Procedural variables included symptom duration, degree of internal carotid artery stenosis, aortic arch type (including bovine variants), anesthesia type, vascular access, antiplatelet and anticoagulant administration, activated clotting time, internal carotid artery tortuosity, plaque calcification length, stent characteristics, embolic protection devices use, angioplasty (prestenting and poststenting), fluoroscopy time, contrast volume, and technical success. Postprocedural data included blood pressure monitoring, NIHSS score, and mRS score assessments at 24 hours and discharge, as well as hospital length of stay and follow-up neurological scores at 30 and 90 days. The 2021 American Heart Association/American Stroke Association stroke prevention guidelines were followed for all enrolled patients’ stroke risk management.¹⁰ Physician background training data was also collected. Supplemental Material S2 contains definitions of variables and angiographic classifications.

Outcomes

The primary outcome is procedure-related composite adverse event, defined as the occurrence of any one of the following events: (1) 30-day procedure-related mortality, (2) symptomatic ischemic stroke, (3) symptomatic hemorrhagic stroke, and (4) MI. Each patient was counted only once in the composite outcome, even if they experienced more than 1 of these events.

The secondary outcome is also a composite outcome includes the following events: (1) 30-day nonprocedure-related mortality, (2) asymptomatic intracranial hemorrhage, (3) transient ischemic attack including Amaurosis Fugax and minor nondisabling ischemic stroke, (4) distal embolization successfully addressed with no residual neurological deficits, (5) access site complications, (6) symptomatic bradycardia including pulseless electrical activity, (7) early in-stent thrombosis (within 30 days), (8) late in-stent thrombosis (>1 up to 12 months), (9) dissection secondary to CAS, and (10) other complications including retinal artery occlusion, hemodynamic instability, and very late stent thrombosis. All-cause 30-day mortality was defined as the combined occurrence of outcomes: 30-day procedure-related mortality and 30-day nonprocedure-related mortality. Details of the complications are presented in Table S1.

Statistical Analysis

Continuous variables were reported as means with SD. Categorical variables were summarized as frequencies and percentages. To evaluate predictors of outcomes, logistic regression was performed for 3 end points: primary outcome, secondary outcome, and 30-day mortality. Covariates were selected a priori based on clinical relevance and prior literature. Results were expressed as odds ratios (OR) with 95% CIs, and statistical significance was set at P<0.05.

Patients with missing data for the primary or secondary outcomes were excluded entirely from the study. For other covariates, analyses were conducted using complete-case data without imputation, as the proportion of missing data was low and missingness was not outcome dependent.

Receiver Operating Characteristic curve analysis was conducted to assess the discriminative performance of preprocedural mRS score, NIHSS score, and age. Area under the curve (AUC), sensitivity, specificity, and optimal cutoff points (based on Youden index) were reported for each predictor across the 3 outcomes. All statistical analyses were performed using R version 4.5.5.

Institutional Review Board Oversight and Consent

This research activity was determined to be exempt or excluded from institutional review board oversight in accordance with current regulations and institutional policy (reference no. 2023-149). The study used fully deidentified data; therefore, informed consent was not required. All research materials are subject to copyright protection by HCA Healthcare and its subsidiaries. One author has full access to all study data and takes responsibility for the integrity of the data and accuracy of the analysis.

RESULTS

Baseline Characteristics

Between January 2023 and December 2024, the participating hospitals reported 1007 consecutive CAS cases performed by 32 neurointerventionalists. The distribution of annual CAS volume per neurointerventionalists was as follows: 17 proceduralists (53.1%) performed 1 to 24 cases per year, 10 (31.3%) performed 25 to 50 cases, and 5 (15.6%) performed 51 to 75 cases. Regarding years of experience postfellowship training, 5 (15.6%) of neurointerventionalists had 1 to 5 years of experience, 10 (31.3%) had 6 to 10 years, and the majority, 17(53.1%), had >10 years of experience. Out of the total 1007 patients with CAS included, 43 were excluded due to the following: 4 CAS cases aborted, 6 dissections, 4 carotid web, 9 acute stenting during mechanical thrombectomy for tandem lesions, 3 transcarotid artery revascularization, 2 duplicates, and 15 were performed outside the time window. Eleven patients were missing the primary outcome, and 64 were lost to follow-up, leaving 889 carotid artery stenosis patients undergoing CAS included (Figure).

Figure.

STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) patient flowchart. CASSH study (Carotid Artery Stenting Outcomes in Comprehensive Stroke Hospitals) patients, follow-up status, and reasons for exclusion. CAS indicates carotid artery stenting; CSC, comprehensive stroke center; and TCAR, transcarotid artery revascularization.

The mean age was 70.33 years (SD, 9.89), and the cohort was predominantly male (61.4%). The racial distribution was primarily White (73.5%), followed by Hispanic (15.1%), Black (8.5%), and Other (2.9%). Hypertension was observed in 87.1% of patients, whereas diabetes was present in 43.3%. Atrial fibrillation and congestive heart failure were reported in 12.7% and 14.4% of patients, respectively. Hyperlipidemia was common, affecting 77.3% of patients, and a history of prior stroke was noted in 42.1%. The mean preprocedural mRS and NIHSS scores were 1.30 (SD, 1.54) and 2.41 (SD, 4.42), respectively. The mean duration of symptoms before stenting was 31.34 days (SD, 176.93). A symptomatic presentation before the procedure was reported in 63.1% of patients. Detailed Baseline characteristics of the patients are presented in Table 1.

Table 1.

Baseline Characteristics of Included Patients

Preprocedural Details

Among the 889 patients included, preprocedural P2Y12 reaction unit testing was performed in 34.8% of cases. Of these, 86.4% achieved therapeutic P2Y12 reaction unit levels, whereas 13.6% were nontherapeutic. The degree of internal carotid artery stenosis was high across the cohort, with a mean of 79.2% (SD, 14.46). Aortic arch configuration was most commonly classified as type II (34.7%) and type I (34.3%), followed by type III (21.5%). Variants of bovine arch anatomy were observed in a smaller subset, including type 1 bovine (5.5%), type 2 bovine (2.2%), and type 3 bovine (1.7%). Preprocedural hemodynamic measurements revealed a mean systolic blood pressure of 147.42 mm Hg (SD, 26.94) and a mean diastolic blood pressure of 76.74 mm Hg (SD, 36.08; Table S2).

Intraprocedural Details

Monitored anesthesia care was the most commonly used anesthetic strategy (70.5%), followed by general anesthesia (25.0%) and local or no anesthesia (4.5%). The femoral artery was the predominant access site (81.2%), with radial access used in 18.3% and other sites in 0.4%. Intravenous antiplatelet agents were used in only 3.6% of procedures, while procedural anticoagulation was administered in 98.7% of patients. Among those assessed, the activated clotting time fell within the therapeutic range of 250 to 300 seconds in 95.1%. Regarding carotid anatomy, extracranial internal carotid artery tortuosity categorized using the Koge et al¹¹ grading system (straight; tortuous: >15° angulation or S/C-shaped course; coiled: exaggerated S-shaped or circular configuration; kinked: acute <90° angulation) was reported as tortuous in 46.35% of cases, straight in 40.6%, coiled in 7.8%, and kinked in 5.0%. The calcified lesion length was >15 mm in 41.7% of patients, 10 to 15 mm in 29.6%, and <10 mm in 28.7% (Table S3).

Postprocedural Details

After CAS, the mean degree of residual stenosis was 9.35 (SD, 12.40). The mean NIHSS score at 24 hours postprocedure was 1.76 (SD, 3.92). A total of 55 patients (6.2%) experienced complications, of whom 23 patients (46.0%) required active treatment. Hemorrhagic complications were rare, occurring in 3 patients (0.3%): hemorrhagic infarction type 1 accounted for 50.0% of the reported hemorrhages, whereas parenchymal hematoma type 1 and combined parenchymal hematoma type 2 with subarachnoid hemorrhage each accounted for 25.0%. Postprocedural oral antiplatelet therapy was administered in 99.6% of patients, and 10.9% received oral anticoagulation. Immediately after the procedure, mean systolic and diastolic blood pressures were 122.80 mm Hg (SD, 26.95) and 63.59 mm Hg (SD, 15.48), respectively. At 24 hours, blood pressures remained stable at 123.26 mm Hg (SD, 17.13) systolic and 64.14 mm Hg (SD, 12.35) diastolic. The mean hospital length of stay was 4.97 days (SD, 5.44). At discharge, the median mRS score was 1.39 (SD, 1.62) and the NIHSS score was 1.33 (SD, 3.02; Table S4).

Study Outcomes

A primary outcome occurred in 11 patients (1.2%), whereas 48 patients (5.4%) experienced a secondary outcome. When stratified by symptomatic status, the primary outcome occurred in 9 of 561 symptomatic patients (1.6%) compared with 2 of 328 asymptomatic patients (0.6%). Secondary outcomes occurred in 33 symptomatic patients (5.9%) and 15 asymptomatic patients (4.6%). For individual primary complications, procedure-related deaths occurred in 7 patients (0.8%), symptomatic ischemic stroke in 3 patients (0.3%), symptomatic hemorrhagic stroke in 2 patients (0.2%), and MI in 2 patients (0.2%). Among secondary outcomes, non–procedure-related deaths occurred in 13 patients (1.5%), asymptomatic intracranial hemorrhage in 3 patients (0.3%), transient ischemic attack or minor ischemic stroke in 5 patients (0.6%), access site complications in 15 patients (1.7%), symptomatic bradycardia in 4 patients (0.4%), and distal embolization in 1 patient (0.1%). Early and late stent thrombosis were rare, in 1 patient (0.1%) and 2 patients (0.2%), respectively. Dissection attributable to CAS occurred in 2 patients (0.2%), whereas other complications were reported in 5 patients (0.6%; Table 2). All-cause 30-day mortality occurred in 22 patients (2.2%). At follow-up, patients exhibited favorable neurological and functional outcomes, with a mean NIHSS score of 0.90 (SD, 2.69) and a mean 30-day mRS score of 1.12 (SD, 1.61).

Table 2.

Summary of the Study Outcomes

Regression

Primary Outcome

Higher preprocedural mRS scores were independently associated with increased odds of the primary outcome (OR, 1.42 [95% CI, 1.01–2.00]; P<0.05), as were higher preprocedural NIHSS scores (OR, 1.09 [95% CI, 1.01–1.18]; P<0.05). Longer fluoroscopy time during the procedure was also significantly associated with the primary outcome (OR, 1.02 [95% CI, 1.00–1.03]; P<0.05). The odd of 24-hour NIHSS score worsening (OR, 1.20 [95% CI, 1.12–1.28]; P<0.05), higher discharge (OR, 4.92 [95% CI, 2.59–9.36]; P<0.05), higher discharge NIHSS score (OR, 1.19 [95% CI, 1.05–1.35]; P<0.05) and increased length of stay (OR, 1.09 [95% CI, 1.04–1.14]; P<0.05) were associated with primary outcomes. Importantly, postprocedural use of oral antiplatelet therapy was associated with significantly reduced odds of the primary outcome (OR, 0.03 [95% CI, 0.00–0.36]; P<0.05; Table 3).

Table 3.

Logistic Regression for Variables Predicting the Primary Outcome

Secondary Outcomes

Higher preprocedural mRS score (OR, 1.37 [95% CI, 1.16–1.62]; P<0.05) and higher preprocedural NIHSS score (OR, 1.11 [95% CI, 1.07–1.16]; P<0.05) were both associated with increased odds of the secondary outcome. Procedurally, prestenting angioplasty was linked to significantly higher odds of the secondary outcome (OR, 2.05 [95% CI, 1.11–3.80]; P<0.05), whereas right-sided carotid stenting was associated with reduced odds compared with left-sided procedures (OR, 0.51 [95% CI, 0.28–0.93]; P<0.05). Postprocedurally: elevated NIHSS score at 24 hours postprocedure (OR, 1.16 [95% CI, 1.11–1.21]; P<0.05), discharge mRS score (OR, 1.74 [95% CI, 1.47–2.06]; P<0.05), and discharge NIHSS score (OR, 1.15 [95% CI, 1.08–1.22]; P<0.05) were all significantly associated with increased odds of secondary outcomes. Longer hospital stay was also modestly predictive (OR, 1.08 [95% CI, 1.04–1.12]; P<0.05; Table S5).

Mortality

Older age was significantly associated with increased odds of mortality (OR, 1.05 [95% CI, 1.00–1.11]; P<0.05). Lower ejection fraction was also predictive (OR, 0.96 [95% CI, 0.92–1.00]; P<0.05). Among baseline functional scores, both higher preprocedural mRS score (OR, 1.72 [95% CI, 1.32–2.24]; P<0.05) and higher NIHSS score (OR, 1.15 [95% CI, 1.09–1.22]; P<0.05) were independently associated with mortality. Intraprocedural predictors showed that right-sided carotid stenting was associated with significantly lower odds compared with left-sided procedures (OR, 0.21 [95% CI, 0.07–0.65]; P<0.05), whereas longer fluoroscopy time conferred increased risk (OR, 1.02 [95% CI, 1.01–1.03]; P<0.05). Postprocedural predictors included NIHSS score at 24 hours (OR, 1.19 [95% CI, 1.12–1.25]; P<0.05), length of hospital stay (OR, 1.09 [95% CI, 1.05–1.14]; P<0.05), mRS score at discharge (OR, 3.40 [95% CI, 2.28–5.06]; P<0.05), and NIHSS score at discharge (OR, 1.15 [95% CI, 1.04–1.27]; P<0.05). Notably, postprocedural oral antiplatelet use was associated with significantly reduced odds of mortality (OR, 0.07 [95% CI, 0.01–0.66]; P<0.05; Table 4).

Table 4.

Logistic Regression for Variables Predicting Mortality

Discriminative Performance of Predictors

Receiver operating characteristic analysis demonstrated that preprocedural NIHSS score had the highest predictive value for mortality, with an AUC of 77.0%, sensitivity of 84.2%, and specificity of 64.9% at a cutoff point of 2. Preprocedural mRS score also performed well for mortality prediction (AUC, 73.1%; sensitivity, 65.0%; specificity, 76.7%) with a cutoff of 3. Age showed a lower predictive value (AUC, 64.2%) for mortality, with an optimal threshold of 74 years. For the primary outcome, preprocedural NIHSS score remained the strongest predictor (AUC, 70.0%), followed by mRS score (63.4%) and age (52.9%). Predictive performance for the secondary outcome was more modest across all variables, with NIHSS score (AUC, 63.9%) and mRS score (63.4%) showing similar discrimination at a cutoff of 3, whereas age had limited predictive value (AUC, 54.3%) at a threshold of 80 years (Table S6).

Discussion

Baseline

This multicenter, prospective observational study demonstrates that the 30-day composite primary outcome of CAS, when performed by fellowship-trained neurointerventionalists at CSCs, is lower than previously reported. Higher preprocedure disability score, stroke score, and low ejection fracture are associated with 30-day mortality. The expertise of these specialists is evidenced by several key outcomes that surpass those reported in earlier literature.

Postprocedure

The degree of residual stenosis has been recognized as an important factor in CAS outcomes. In our cohort, the mean residual stenosis was 9.35, which represents an excellent procedural outcome. This likely reflects the expertise of the neuroendovascular specialists performing these interventions in our study. There has been mixed evidence regarding the association between clinical outcomes and the technical success of reducing residual stenosis. In a previous study, higher residual stenosis was significantly associated with periprocedural complications, including hemodynamic instability, asymptomatic dissection, and asymptomatic embolization, as well as with 30-day clinical outcomes such as stroke, MI, and all-cause mortality.¹² Many prior studies have used a threshold of 30% to 40% residual stenosis when assessing clinical outcomes.^12–14 In contrast, one study reported that residual stenosis was not associated with either short- or long-term postprocedural adverse events, with comparable 2-year rates of stroke or death between patients with ≥40% residual stenosis (6.2%) and those with <40% (6.7%).¹³

This likely reflects the expertise of the neuroendovascular specialists performing these interventions in our study. Further studies investigating the degree of different residual stenosis and the clinical outcome are warranted.

Complication Predictor

Preprocedural mRS score was a critical predictor in our cohort, independently predicting mortality, primary, and secondary outcomes. Our findings align with previous studies demonstrating that higher preprocedural mRS score is associated with increased in-hospital stroke and death after carotid revascularization.¹⁵ Supporting evidence shows that 46% of patients with an mRS score ≤2 experienced improved follow-up functional outcomes compared with only 23% with an mRS score >2, highlighting that functional gains are more likely when baseline disability is minimal.¹⁶

These findings require careful interpretation in light of the October 2023 Centers for Medicare & Medicaid Services coverage update for CAS, which significantly broadened eligibility criteria to include symptomatic stenosis ≥50% and asymptomatic stenosis ≥70%.¹⁷ This policy expansion may lead to future studies that include a larger proportion of asymptomatic and functionally independent patients, potentially biasing results toward better outcomes and overestimating the protective effect of low baseline mRS score. In contrast, our study reflects deliberate patient selection at CSCs with experienced proceduralists. The substantial effect sizes we observed (OR, 1.43) for primary outcomes and (OR, 1.37) for secondary outcomes related to preprocedural mRS score should therefore be emphasized as clinically meaningful. Future publications reporting different magnitudes of association warrant scrutiny for case-mix differences, particularly regarding symptomatic status, baseline disability distribution, and center expertise.

Oral antiplatelet therapy was identified as an independent factor associated with lower rates of primary complications after CAS. This observation is consistent with prior literature suggesting that antiplatelet therapy may reduce the risk of periprocedural neurological events, although causality cannot be established in this observational study. A landmark randomized controlled trial from 2005 demonstrated that dual antiplatelet therapy reduced adverse neurological outcomes after CAS without increasing major bleeding complications.¹⁸ In our study, the association between antiplatelet use and reduced complications persisted even when hemorrhagic stroke was included as an adverse outcome; however, these findings should be interpreted cautiously, given the potential for confounding by indication and other unmeasured factors. A comparison with the preceding CASONI study provides important context for our findings. CASONI reported an overall complication rate of 4.1% and all-cause 30-day mortality of 1.2%, with primary events accounting for 1.7% and secondary events accounting for 2.8%.⁶ In contrast, CASSH observed a slightly higher overall complication rate of 6.6% and all-cause 30-day mortality of 2.2%. Although procedure-related complications were comparable between studies, the modestly higher rates in CASSH likely reflect its smaller sample size (889 versus 1445), shorter duration (2 versus 10 years), and more rigorous prospective design, which captured real-world events with greater granularity.

CASSH demonstrated an overall primary complication rate of 1.2%. When stratified by symptom status, primary complications occurred in 1.6% of symptomatic patients and 0.6% of asymptomatic patients. In our asymptomatic cohort (n=328), early and short-term event rates were low: overall mortality occurred in 4 patients (1.2%), ischemic stroke in 1 patient (0.3%), and no hemorrhagic strokes were identified. These findings are notable alongside the CREST-2 trial (Carotid Revascularization and Medical Management for Asymptomatic Carotid Stenosis Trial),¹⁹ in which 7 strokes and 1 death (8/616) occurred in the stenting arm during the periprocedural period (1.3% of patients [95% CI, 0.6–2.5]). Importantly, CREST-2 was conducted under highly controlled conditions, with protocol-driven intensive medical therapy, centralized adjudication, and standardized postprocedural surveillance. In contrast, our study reflects real-world practice without mandated postoperative care pathways. Despite these differences, ischemic and hemorrhagic stroke rates in our asymptomatic population remained low. These data underscore the procedural safety of contemporary carotid stenting in selected asymptomatic patients under routine clinical conditions.

Our outcomes are also markedly superior to those reported in earlier symptomatic trials, including EVA-3S (Endarterectomy Versus Stenting in Patients With Symptomatic Severe Carotid Stenosis; 9.6% stroke/death with CAS versus 3.9% with CEA), SPACE (Stent‑Protected Angioplasty Versus Carotid Endarterectomy; 6.8% versus 6.3%), and ICSS (International Carotid Stenting Study; 7.0% versus 3.3%).^20–25 In asymptomatic populations, our results also compare favorably: ACT I (Asymptomatic Carotid Trial I) reported 30-day stroke or death rates of 2.9% with CAS and 1.7% with CEA, whereas ACST-2 (Second Asymptomatic Carotid Surgery Trial) observed rates of 3.7% versus 2.7%.^7,26 The landmark CREST trial, which included both symptomatic and asymptomatic patients, reported periprocedural stroke, death, or MI in 6.7% of symptomatic patients with CAS versus 5.4% with CEA, and 3.5% versus 3.6% in asymptomatic patients.^27–29 More contemporary data from the CREST-2 Registry showed 30-day stroke or death rates of 2.8% in symptomatic patients and 1.4% in asymptomatic patients.³⁰ Even in high-risk populations such as the SAPPHIRE trial (Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy), 30-day CAS outcomes included death in 1.2%, stroke in 3.6%, and MI in 2.4%.^31,32 Collectively, these comparisons demonstrate that CASSH complication rates are substantially lower than earlier symptomatic randomized controlled trials, more favorable than recent asymptomatic trial outcomes, and consistent with contemporary registry results. This evidence strongly supports the safety and efficacy of CAS when performed by fellowship-trained neurointerventionalists at CSCs.

Mortality Predictor

Several factors were strongly predictive of mortality after CAS. Most notably, older age was independently associated with higher mortality. Although some prior studies reported that patients older than 75 years did not experience significantly more major adverse events within 30 days than younger patients undergoing CAS,³³ these results may be confounded by selection bias toward healthier elderly patients. In contrast, our study was more representative of real-world patients undergoing CAS.

Reduced left ventricular ejection fraction was another important mortality predictor, consistent with research showing that cardiovascular comorbidities significantly affect long-term survival after CAS.³³ In addition, baseline functional and neurological status were critical; higher preprocedural mRS and NIHSS scores were both independently associated with increased mortality, aligning with studies demonstrating that preprocedural disability and stroke severity strongly determine survival and recovery.^34–36

These findings emphasize the need for careful patient selection under the new Centers for Medicare & Medicaid Services guidelines for CAS.¹⁷ The updated guidance has broadened access by removing high-surgical-risk requirements, potentially increasing procedures among older patients and those with significant cardiovascular and neurological comorbidities. In such cases, procedural mortality risk may outweigh potential benefits, particularly when life expectancy and functional independence are limited, and these factors are not adequately considered in the decision-making process.

Limitations

This prospective observational study has several important limitations. First, the nonrandomized design introduces potential selection bias and unmeasured confounding in treatment allocation. Second, outcomes were site-reported, and procedural variations across participating centers may have introduced reporting variability and influenced results. Third, the 30- to 90-day follow-up period precludes assessment of late restenosis or long-term outcomes. Fourth, comparisons with prior studies are limited by inherent design differences between randomized trials, registries, and observational cohorts that cannot be fully accounted for statistically.

Despite these limitations, this study has several notable strengths. The large sample size (N=889) provides sufficient statistical power to detect clinically meaningful differences and supports robust outcome analyses. The multicenter design across geographically diverse CSCs enhances external validity and generalizability of findings to real-world clinical practice. The uniform requirement for fellowship-trained neurointerventionalists ensures a consistent level of expertise while still capturing natural variations in technique and patient selection.

Building on this prospective observational study, the next essential step is a multicenter randomized controlled trial with independent core laboratory adjudication, directly comparing CAS by fellowship-trained neurointerventionalists against alternative treatment modalities to definitively establish relative safety and efficacy in the contemporary era.

Conclusions

CAS performed by fellowship-trained neurointerventionalists at CSCs was associated with lower periprocedural complication rates compared with those reported in prior randomized trials. These findings are consistent with recently published data from CREST-2, underscoring the safety of contemporary CAS in real-world practice. Increased preprocedural disability, greater stroke severity, advanced age, and impaired cardiac function were independently associated with early mortality, highlighting the critical importance of careful patient selection. Collectively, these results emphasize the value of procedural expertise at high-volume centers and support the need for future multicenter randomized studies to further refine best practices and optimize patient outcomes.

ARTICLE INFORMATION

Sources of Funding

None.

Disclosures

This research was supported (in whole or in part) by HCA Healthcare and an HCA Healthcare–affiliated entity. The views expressed in this publication represent those of the author(s) and do not necessarily represent the official views of HCA Healthcare or any of its affiliated entities. Dr Hassan reports the following: consultant/speaker for Medtronic, Microvention, Stryker, Penumbra, Cerenovus, Genentech, GE Healthcare, Scientia, Balt, viz.ai, Insera Therapeutics, Proximie, NeuroVasc, NovaSignal, Vesalio, Rapid Medical, Imperative Care, Galaxy Therapeutics, Route 92 and Perfuze. Dr Hassan is principal investigator for COMPLETE study—Penumbra (Endovascular Treatment of Acute Ischemic Stroke With the Penumbra System in Routine Practice: COMPLETE Registry Results), LVO SYNCHRONISE—viz.ai (SYNCHRONISE: LVO Triage Timing and Outcome Study), and Millipede Stroke Trial—Perfuze (Millipede Aspiration for Revascularization in Stroke [MARRS] Study), and RESCUE ICAD—Medtronic (Resolute Stensing in the treatement of Intracranial Atherosclerotic Disease). Dr Tanweer reports the following: consulting agreements from viz.AI Inc, Penumbra Inc, Balt Inc, Stryker Inc, Imperative Inc, and Q’Apel Inc, and Proctor for Microvention Inc and Medtronic Inc. Dr Alaraj reports being a consultant for Cerenovus. Dr Martucci reports being a consultant for Cerenovus (Scientific Advisory Board, Clinical Events Committee), Stryker (Data Safety Monitoring Board), Rapid Medical (Clinical Events Committee), and Core Lab Principal Investigator for Medtronic and Microvention. Steering Committee/Publication committee member for SELECT (Optimizing Patientʼs Selection for Endovascular Treatment in Acute Ischemic Stroke), DAWN (Clinical Mismatch in the Triage of Wake Up and Late Presenting Strokes Undergoing Neurointervention With Trevo), SELECT 2 (A Randomized Controlled Trial to Optimize Patientʼs Selection for Endovascular Treatment in Acute Ischemic Stroke), EMBOLISE (A Study of the Embolization of the Middle Meningeal Artery With ONYX Liquid Embolic System in the Treatment of Subacute and Chronic Subdural Hematoma), CLEAR study (The Vesalio NeVa Stent Retriever Study for Treatment of Large Vessel Occlusion Strokes), ENVI RCT (Envi-SR Randomized Controlled Trial for Endovascular Treatment of Ischemic Stroke), DISTALS (Distal Ischemic Stroke Treatment With Adjustable Low-Profile Stentriever), and COMMAND (COMMAND Early Feasibility Study: Implantable BCI to Control a Digital Device for People With Paralysis). Dr Ezzeldin reports the following: consultant for Stryker, viz AI, Medtronic and Imperative Care. Dr Ezzeldin has a small investment at Galaxy Therapeutics. Dr Zaidat serves on the editorial board of Stroke: Vascular and Interventional Neurology. Editorial board members are not involved in the handling or final disposition of submissions. Dr Zaidat also reports the following: consultant and speaker for Cerenovus, Stryker, Penumbra, Medtronic, and research grants from Stryker, Medtronic, Cerenovus, Penumbra, and Genentech. Dr Siddiq reports being a consultant for Microvention and grants from the National Institutes of Health, Child Neurology Society (CNS), and The Society of NeuroInterventional Surgery (SNIS), not related to this project. The other authors report no conflicts.

Supplemental Material

Supplemental Material 1: List of Contributing Centers

Supplemental Material 2: Definitions of Variables and Angiographic Classifications

Tables S1–S6

References 9,11,29,35,37–45

STROBE Checklist

Supplementary Material

svi2-6-e002201-s001.docx ^{(29.9KB, docx)}