Voeks, Jenifer H.

Introduction: Increasing stenosis has been questioned as a factor increasing risk of carotid endarterectomy (CEA) or carotid stenting (CAS) in patients with severe asymptomatic carotid stenosis. Hypothesis: Greater severity of carotid stenosis is associated with higher rates of periprocedural stroke and death following revascularization for asymptomatic patients. Methods: Asymptomatic patients with carotid stenosis ≥ 70% by ultrasound or ≥ 60% by angiogram were eligible for the Carotid Revascularization Endarterectomy vs Stenting Trial (CREST). Patients who had a catheter-angiogram were divided into tertiles based on the degree of stenosis. Outcomes were the occurrence of any stroke or death at 30 days. Proportional hazards models adjusting for age and treatment were used to assess risk of 30 day stroke or death by tertile of stenosis. Results: Among 1181 asymptomatic patients, qualifying angiograms were done for 662 patients who had assigned procedure performed within 30 days of randomization. Median % stenosis was 62.6, 73.4, and 83.0 for the tertiles that otherwise differed only for female sex (40% female in tertile 1, 36% in tertile 2, 29% in tertile 3, p=0.01). The 30-day stroke and death rates did not differ significantly by severity of stenosis (Table), but the number of stroke and deaths was only 14 across the tertiles (Table). Similarly meaningful comparison of CEA vs CAS was not possible. Conclusion: This is the largest contemporary study of carotid angiograms performed in patients with severe asymptomatic carotid stenosis. No relationship was detected between severity of stenosis and 30-day stroke and death. The safety of CEA and CAS in asymptomatic patients limits detection of other factors that may increase risk because so few events complicate these procedures.

Objective: To compare the perioperative risk of any stroke, death and MI and the long-term risk of re-stenosis and stroke following CEA with patching (a prosthetic or vein patch sutured to the arterial opening to widen the closure) versus CEA with primary closure in CREST. Methods: We included patients who were randomized and underwent CEA in CREST. We compared peri-procedural and 4-year event rates and 2-year restenosis rates, in addition to rates of reoperation between the two groups. Re-stenosis was defined as elevated peak systolic velocity ≥3.0 m/s on carotid duplex US. Treatment differences were assessed using proportional hazard models adjusting for age, sex, and symptomatic status. Results: There were 1,148 patients who underwent CEA (755 {66%} patch; 328 {29%} primary closure). We excluded 43 patients who underwent eversion CEA and 24 patients missing CEA data (5%). There was no significant difference between the two groups with respect to age, sex, race and comorbidities (Table 1). There were significant reductions in the risk of any periprocedural stroke and death (HR 0.35, 95% CI 0.15-0.82, P<0.02), in re-operation (OR 0.33, 95% CI 0.14-0.80, P<0.01), in the 4-year risk of stroke (HR 0.52, 95% CI 0.28-0.97, P<0.04) and in the 2-year risk of re-stenosis (HR 0.26, 95% CI 0.14-0.45, P<0.0001) in the patch compared to the primary closure group (Table 2). Conclusion: Patch closure should be considered for all CEAs because of the association of patch closure with reductions in risk of periprocedural stroke and death, re-operation, and long-term clinical and anatomic durability.

Background: Hormone replacement therapy (HRT) was among factors considered in a post-hoc assessment of the risk of revascularization by carotid artery stenting (CAS) or carotid endarterectomy (CEA) in women enrolled in CREST. Methods: Of the 872 women in CREST, 775 (89%) were post-menopausal and current HRT status was known in 739 (95%). The study composite outcome was defined as any stroke, myocardial infarction, or death during the periprocedural period, plus ipsilateral stroke over a 4-year follow-up period. Treatment groups were combined and differences in the composite outcome were assessed by current HRT at baseline. Results: There were 77/739 (10%) current HRT users. HRT users were more likely to be symptomatic at baseline than non-users (65% vs. 52%; p = 0.027), and were less likely to be diabetic (19% vs. 32%; p = 0.024). However, they were similar in most other factors including age, race, hypertension, dyslipidemia, smoking status, and length of follow-up (p > 0.10). There were no (0) periprocedural events among the 77 HRT users while there were 43 (6.5%) events among the 662 non-users (p = 0.017). Among those event-free at the end of the periprocedural period, there was little evidence of an impact of baseline HRT on the risk of ipsilateral stroke with 2 (2.6%) post-procedural events among 77 HRT users and 12 (1.9%) among 619 non-users. Conclusion: We found that women on current HRT have lower periprocedural risk associated with CAS or CEA than non-users, but little difference in risk after the procedural period. The mechanism for this unexpected finding is elusive. We urge caution in the interpretation of this post-hoc observation and encourage replication or refutation by other revascularization studies with similar data.

Background An increased risk of periprocedural stroke and death has been reported for older patients treated with carotid stenting (CAS). Selection of patients for CAS could be improved if patient or artery characteristics responsible for this higher risk could be identified. Methods The 1123 CREST patients randomized to CAS who received CAS within 30-days were analyzed. From the CREST sites, risk factors for atherosclerosis (hypertension, diabetes, and dyslipidemia) and arterial characteristics by angiography (lesion length, eccentric lesions, ulcerated lesions, and percent stenosis) were assessed. From the CREST core labs, potential mediators of the age effect included peak systolic velocity (ultrasound core, available on 853 patients) and arterial characteristics (angiographic core, available on 1042 patients, narrow/no ulcer or wide mouth, lesion location, and distal tortuosity). Mediation analysis assessed impact of adjustment for these factors on the magnitude of the estimated increased risk for CAS at older ages. Results After adjustment for sex and symptomatic status, there was a 1.72-times (95% CI: 1.26 - 2.37) increased risk of periprocedural stroke and death associated with each decade of patient age. Adjustment for lesion length as assessed in the clinical center significantly attenuated this increased risk at older ages to 1.66 (95% CI: 1.20 - 2.29, p = 0.039), an 8.3% attenuation. No other factor showed evidence as a contributor to the periprocedural increased risk of stroke and death in older patients (p > 0.10). Conclusion These data suggest that a longer carotid artery lesion length is a significant contributor to the increased risk of CAS in the elderly. However, attenuation of the age effect was modest, less than a 10% change in the hazard ratios. Other potential mediators such as tortuosity and atherosclerosis of the aortic arch and great vessels were not assessed. Non-anatomic factors that are more common with advancing age such as cerebral white matter disease and subclinical dementia were also not assessed and may play a role in how elderly patients respond to ischemic insults.