Endarterectomy achieves lower stroke and death rates compared with stenting in patients with asymptomatic carotid stenosis

REVIEW ARTICLES Ronald M. Fairman, MD, SECTION EDITOR

Endarterectomy achieves lower stroke and death rates compared with stenting in patients with asymptomatic carotid stenosis Stavros K. Kakkos, MD, PhD, RVT,a,b Ioannis Kakisis, MD, PhD,c Ioannis A. Tsolakis, MD, PhD,a and George Geroulakos, MD, PhD,b,c Patras and Athens, Greece; and London, United Kingdom

ABSTRACT Background: It is currently unclear if carotid artery stenting (CAS) is as safe as carotid endarterectomy (CEA) for patients with significant asymptomatic stenosis. The aim of our study was to perform a systematic review and meta-analysis of trials comparing CAS with CEA. Methods: On March 17, 2017, a search for randomized controlled trials was performed in MEDLINE and Scopus databases with no time limits. We performed meta-analyses with Peto odds ratios (ORs) and 95% confidence intervals (CIs). Quality of evidence was assessed with the Grading of Recommendations Assessment, Development, and Evaluation method. The primary safety and efficacy outcome measures were stroke or death rate at 30 days and ipsilateral stroke at 1 year (including ipsilateral stroke and death rate at 30 days), respectively. Perioperative stroke, ipsilateral stroke, myocardial infarction (MI), and cranial nerve injury (CNI) were all secondary outcome measures. Results: The systematic review of the literature identified nine randomized controlled trials reporting on 3709 patients allocated into CEA (n ¼ 1479) or CAS (n ¼ 2230). Stroke or death rate at 30 days was significantly higher for CAS (64/2176 [2.94%]) compared with CEA (27/1431 [1.89%]; OR, 1.57; 95% CI, 1.01-2.44; P ¼ .044), with low level of heterogeneity beyond chance (I2 ¼ 0%). Also, stroke rate at 30 days was significantly higher for CAS (63/2176 [2.90%]) than for CEA (26/1431 [1.82%]; OR, 1.63; 95% CI, 1.04-2.54; P ¼ .032; I2 ¼ 0%). MI at 30 days was nonsignificantly lower for CAS (12/1815 [0.66%]) compared with CEA (16/1070 [1.50%]; OR, 0.53; 95% CI, 0.24-1.14; P ¼ .105; I2 ¼ 0%); however, CNI at 30 days was significantly lower for CAS (2/1794 [0.11%]) than for CEA (33/1061 [3.21%]; OR, 0.13; 95% CI, 0.07-0.26; P < .00001; I2 ¼ 0%). Regarding the long-term outcome of stroke or death rate at 30 days plus ipsilateral stroke during follow-up, this was significantly higher for CAS (79/2173 [3.64%]) than for CEA (35/1430 [2.45%]; OR, 1.51; 95% CI, 1.02-2.24; P ¼ .04; I2 ¼ 0%). Quality of evidence for all stroke outcomes was graded moderate. Conclusions: Among patients with asymptomatic stenosis undergoing carotid intervention, there is moderate-quality evidence to suggest that CEA had significantly lower 30-day stroke and also stroke or death rates compared with CAS at the cost of higher CNI and nonsignificantly higher MI rates. The long-term efficacy of CEA in ipsilateral stroke prevention, taking into account perioperative stroke and death, was preserved during follow-up. There is an urgent need for highquality research before a firm recommendation is made that CAS is inferior or not to CEA. (J Vasc Surg 2017;66:607-17.)

At present, there is no trial evidence available to support the hypothesis that carotid artery stenting (CAS) for asymptomatic carotid stenosis is either superior or inferior to carotid endarterectomy (CEA). The observations that CAS for symptomatic patients is indeed inferior in the short term to CEA, as demonstrated by previous randomized controlled trials (RCTs),1,2 have From the Department of Vascular Surgery, University of Patras Medical School, Patrasa; the Department of Surgery and Cancer, Imperial College London, Londonb; and the National and Kapodistrian University of Athens, Athens.c Author conflict of interest: none. Correspondence: Stavros K. Kakkos, MD, PhD, RVT, Department of Vascular Surgery, University Hospital of Patras, Hippocrates Avenue, Patras 26504, Greece (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2017.04.053

not been replicated so far for asymptomatic stenosis because the original trials might have been underpowered or had a noninferiority design.3-5 This might have also been a result of the fact that asymptomatic patients have a very low risk of events likely to be similar in the CAS and CEA group owing to the mostly stable nature of the carotid atheroma, which is less likely to produce distal embolization as a result of wire and catheter manipulations in CAS compared with symptomatic patients. Therefore, a clinical equipoise of CAS and CEA in asymptomatic patients may well be hypothesized. The aim of this investigation was to perform a systematic review of the literature and meta-analysis of RCTs comparing CAS with CEA, restricted to asymptomatic carotid stenosis; this was an update of a similar metaanalysis performed by the second author 5 years ago,6 thought to be necessary in view of the recently published RCTs in asymptomatic patients.5,7 607

608

Journal of Vascular Surgery

Kakkos et al

August 2017

METHODS On May 30, 2016, we conducted a literature search on MEDLINE, Cochrane Central Register of Controlled Trials, and Scopus electronic databases using the keywords “carotid endarterectomy and carotid stenting,” with no date limit, to identify RCTs comparing CEA with CAS, followed by a manual search of the reference list of the full-text articles and similar meta-analyses to identify additional trials. Our search was fully updated on March 17, 2017. We opted to include only RCTs in our metaanalysis, which is the best way to adjust potential bias introduced by case-controlled trials in which cases not favorable for CAS or CEA are allocated to the alternative group, respectively. Studies published in languages other than English were not considered, but unpublished data (eg, from the authors of the included studies) were requested. RCTs were identified by two of the authors (S.K.K. and I.K.), and any disagreement was resolved by discussion. The study selection process (Preferred Reporting Items for Systematic Reviews and MetaAnalyses flow diagram) is shown in Fig 1. For each RCT included in the study, raw data (number of patients who developed an end point and total number of patients in the intervention and control groups) were independently extracted by two of the authors (S.K.K. and I.K.) and entered into data sheets. All entries were compared; any disagreement was resolved by discussion, and once confirmed, they were entered into the metaanalysis software. Outcome measures. Stroke or death rate at 30 days was the primary safety outcome, and ipsilateral stroke at 1 year (including ipsilateral stroke and death rate at 30 days) was the primary efficacy outcome of our study. In case long-term results were not available, the corresponding short-term results were provided. To minimize excessive imbalance in follow-up, because of the variable length of observations for the included studies, only midterm results (ie, up to 5 years) were considered. Secondary outcomes included perioperative stroke, ipsilateral stroke, myocardial infarction (MI), and cranial nerve injury (CNI). We planned subgroup analyses for studies performed in standard and high surgical risk patients. Methodologic assessment. The methodologic quality of the RCTs was assessed with the risk of bias assessment tool of the Review Manager software (RevMan version 5.3.5; The Nordic Cochrane Centre, Copenhagen, Denmark). An independent assessment was performed by two of the authors (S.K.K. and I.K.); any disagreement was resolved by discussion. Data synthesis. Meta-analysis was performed with RevMan, which calculated the odds ratio (OR, Peto method) and its 95% confidence interval (CI), produced forest plots, and provided inconsistency (I2) statistics to evaluate the heterogeneity of the included studies. The

95% CIs of I2 statistics were calculated with StatsDirect statistical software (version 2.8.0 or later; StatsDirect Ltd, Altrincham, Cheshire, UK), where the number of studies and events made this calculation possible. Significance levels (P values) for ORs were calculated with both software types. A nonsignificant P value for the Cochrane Q statistic indicates that the included studies are homogeneous. An I2 value of 0% indicates no heterogeneity, whereas larger values are consistent with increasing heterogeneity. An I2 >50% is indicative of substantial heterogeneity. Sensitivity analysis was planned in case heterogeneity was present, but this was not required. The quality of evidence was graded high, moderate, low, and very low on the basis of risk of bias, directness of evidence, heterogeneity, precision of effects estimates, and risk of publication bias; this was performed using the method developed by the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) Working Group. A “summary of findings” table was created using GRADEprofiler (version 3.6.1), which included the two primary and all secondary outcome measures. The assumed control intervention risks were calculated from the mean number of events in the control groups of the included studies for each outcome.

RESULTS We identified 3871 records (after removing the duplicates) on literature search (Fig 1). Some 3824 of them were excluded on the basis of their title and abstract, and an additional 12 records were excluded because they were RCTs in symptomatic patients (n ¼ 4), had a nonrandomized (n ¼ 2) or retrospective (n ¼ 1) design, were published in a non-English language (n ¼ 2), had an irrelevant study protocol (n ¼ 1), or were a comment to a study (n ¼ 1) and because full text could not be acquired in a study with mixed symptomatic and asymptomatic patients (n ¼ 1), which left 35 records in the search. An additional record identified through hand searching was added so that the total number of relevant records was 36, which reported on nine RCTs included in the study.3-5,7-38 A summary of these nine RCTs is provided in Table I. Five additional RCTs were identified for this update of the meta-analysis.6 These nine RCTs included a total of 3709 patients allocated to CEA (n ¼ 1479) or CAS (n ¼ 2230). One study was excluded from quantitative meta-analysis because only death rates at 3 months were provided,29 which left eight RCTs for quantitative analysis. Additional information for the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS) was obtained from a metaanalyses the authors have published.39 Risk of bias and summary graphs are shown in Figs 2 and 3. These demonstrated a low risk of bias in most attributes; as expected, risk for performance bias (blinding of participants and personnel) was high for all these interventional trials. Detection bias (blinding of outcome

Journal of Vascular Surgery

Kakkos et al

Volume 66, Number 2

Fig 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram showing the selection process of suitable randomized controlled trials (RCTs) for meta-analysis.

609

610

Journal of Vascular Surgery

Kakkos et al

August 2017

Table I. Baseline characteristics of the nine trials comparing carotid angioplasty with stenting No. of participants

Study, year

Endarterectomy

Stenting

Asymptomatic stenosis severity

Surgical risk

Distant symptoms (>180 days) allowed

Additional inclusion of patients with recent symptoms (<180 days) Yes

CAVATAS, 2001

23

30

$50% on angiography

Standard

Yes

Kentucky, 2004

42

44

>80% on angiography

Standard

No

No

SAPPHIRE, 2004

120

117

$80% on ultrasound

High

Yes

Yes

CREST, 2010

587

594

$60% on angiography, $70% or more on ultrasound or $80% on CTA or MRA

Standard

Yes

Yes

Li, 2014 Kuliha, 2015

65

65

$50% on angiography

Standard

Not known

No

7

26

$70% on ultrasound and confirmed by CTA

Standard

Not known

Yes

ACT 1, 2016

364a

1089a

$70% on ultrasound or angiography

Standard

Yes

No

SPACE 2, 2016

203

197

$70% on ultrasound

Standard

Yes

No

68

68

$70% on ultrasound corroborated with CTA or MRA

Standard

Yes

No

Mannheim, 2017

ACT, Asymptomatic Carotid Trial; CAVATAS, Carotid and Vertebral Artery Transluminal Angioplasty Study; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial; CTA, computed tomography angiography; MRA, magnetic resonance angiography; SAPPHIRE, Stenting and Angioplasty with Protection in Patients with High Risk for Endarterectomy; SPACE, Stent-Protected Angioplasty vs Carotid Endarterectomy. a A 1:3 randomization ratio.

Fig 2. Risk of bias graph demonstrating a low risk of bias in most attributes. Risk was high for performance bias.

assessment) was considered low for 67% of all trials, with assessment being performed by an independent assessor or the clinical event committee. One study did not provide 30-day outcome results for 62 patients. Other bias in the form of early stop was noted in two trials. Main findings. Stroke or death rate at 30 days was significantly higher for CAS (64/2176 [2.94%]) than for CEA (27/1431 [1.89%]; OR, 1.57; 95% CI, 1.01-2.44; P ¼ .044; Fig 4). I2 was 0% (95% CI, 0%-61%), indicating low evidence of heterogeneity beyond chance. No difference was detected between the standard surgical risk and high surgical risk subgroups (P ¼.80). Also, stroke rate at 30 days was significantly higher for CAS (63/2176 [2.90%]) than for CEA (26/1431 [1.82%]; OR, 1.63; 95% CI, 1.04-2.54; P ¼ .032; Fig 5; I2 ¼ 0% [95% CI, 0%-61%]); this

included fatal and nonfatal strokes. No difference was detected between the standard and high surgical risk groups (P ¼ .75). There was no significant difference in ipsilateral stroke rate at 30 days with CAS (27/1238 [2.2%]) or CEA (6/499 [1.2%]; OR, 1.56; 95% CI, 0.71-3.43; P ¼ .27; Fig 6; I2 ¼ 0%). There was a nonsignificant trend for a lower MI rate at 30 days for CAS (12/1815 [0.66%]) compared with CEA (16/1070 [1.50%]; OR, 0.53; 95% CI, 0.24-1.14; P ¼ .105; Fig 7; I2 ¼ 0%). Similarly, CNI at 30 days was significantly lower for CAS (2/1794 [0.11%]) than for CEA (33/1061 [3.21%]; OR, 0.13; 95% CI, 0.07-0.26; P < .00001; Fig 8; I2 ¼ 0% [95% CI, 0%-72.9%]). Regarding the long-term outcome of stroke or death rate at 30 days plus ipsilateral stroke during follow-up, this was significantly higher for CAS (79/2173 [3.64%]) than for CEA (35/1430 [2.45%]; OR, 1.51; 95% CI, 1.02-2.24;

Journal of Vascular Surgery

Kakkos et al

Volume 66, Number 2

Fig 3. Risk of summary graph demonstrating the risk assessment of the bias attributes for the individual studies. One study did not provide outcome results for 62 patients, and two more studies were discontinued, classified as “other bias.” ACT, Asymptomatic Carotid Trial; CAVATAS, Carotid and Vertebral Artery Transluminal Angioplasty Study; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial; SAPPHIRE, Stenting and Angioplasty with Protection in Patients with High Risk for Endarterectomy; SPACE, Stent-Protected Angioplasty vs Carotid Endarterectomy.

611

612

Kakkos et al

Journal of Vascular Surgery August 2017

Fig 4. Forest plot demonstrating that carotid artery stenting (CAS) was associated with a higher stroke (2.94%) or death rate at 30 days than carotid endarterectomy (CEA; 1.89%). No difference was detected between the standard and high surgical risk groups (P ¼ .80). ACT, Asymptomatic Carotid Trial; CAVATAS, Carotid and Vertebral Artery Transluminal Angioplasty Study; CI, confidence interval; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial; SAPPHIRE, Stenting and Angioplasty with Protection in Patients with High Risk for Endarterectomy; SPACE, Stent-Protected Angioplasty vs Carotid Endarterectomy.

Fig 5. Forest plot demonstrating that carotid artery stenting (CAS) was associated with a higher stroke rate (2.90%) at 30 days than carotid endarterectomy (CEA; 1.82%). This included fatal and nonfatal strokes. No difference was detected between the standard and high surgical risk groups (P ¼ .75). ACT, Asymptomatic Carotid Trial; CAVATAS, Carotid and Vertebral Artery Transluminal Angioplasty Study; CI, confidence interval; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial; SAPPHIRE, Stenting and Angioplasty with Protection in Patients with High Risk for Endarterectomy; SPACE, Stent-Protected Angioplasty vs Carotid Endarterectomy.

Journal of Vascular Surgery

Kakkos et al

613

Volume 66, Number 2

Fig 6. Forest plot demonstrating that carotid artery stenting (CAS) was associated with similar ipsilateral stroke rate at 30 days (2.2%) with carotid endarterectomy (CEA; 1.2%). ACT, Asymptomatic Carotid Trial; CI, confidence interval.

Fig 7. Forest plot demonstrating that carotid artery stenting (CAS) was associated with smaller short-term risks for myocardial infarction (MI; 0.66%) than carotid endarterectomy (CEA; 1.50%). ACT, Asymptomatic Carotid Trial; CI, confidence interval; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial.

Fig 8. Forest plot demonstrating that carotid artery stenting (CAS) was associated with smaller short-term risks for cranial nerve injury (CNI; 0.11%) than carotid endarterectomy (CEA; 3.21%). ACT, Asymptomatic Carotid Trial; CI, confidence interval; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial.

P ¼ .04; Fig 9; I2 ¼ 0% [95% CI, 0%-61%]). No difference was detected between the standard and high surgical risk groups (P ¼ .87). Quality of the evidence according to the GRADE system. A detailed assessment of all study outcomes appears in Table II. The early outcome of stroke or death rate at 30 days and also the late outcome of stroke or death rate at 30 days plus ipsilateral stroke during followup were both downgraded because of risk of bias and imprecision and upgraded because effects were immediately apparent and consistent with RCTs on symptomatic patients. There was moderate-quality evidence for these main outcome measures of our meta-analysis. The outcome of MI was downgraded because of risk of bias and impression, leading to low-quality evidence.

High-quality evidence was present for CNI, which was downgraded because of risk of bias and imprecision and upgraded by 2 points because of the large magnitude of the effect.

DISCUSSION Our meta-analysis has demonstrated that among patients undergoing carotid intervention for asymptomatic stenosis, there is evidence to suggest that CEA had lower 30-day stroke and also stroke or death rates compared with CAS. This safety issue dominated the relative efficacy of the two procedure types during follow-up. Because the quality of the research is moderate, there is a need for further high-quality research before a firm recommendation is made that CAS is indeed inferior or not to CEA.

614

Kakkos et al

Journal of Vascular Surgery August 2017

Fig 9. Forest plot demonstrating that carotid artery stenting (CAS) was associated with a higher stroke or death rate at 30 days plus ipsilateral stroke during follow-up (3.64%) than carotid endarterectomy (CEA; 2.45%). No difference was detected between the standard and high surgical risk groups (P ¼ .87). ACT, Asymptomatic Carotid Trial; CAVATAS, Carotid and Vertebral Artery Transluminal Angioplasty Study; CI, confidence interval; CREST, Carotid Revascularization Endarterectomy vs Stenting Trial; SAPPHIRE, Stenting and Angioplasty with Protection in Patients with High Risk for Endarterectomy; SPACE, Stent-Protected Angioplasty vs Carotid Endarterectomy.

Our meta-analysis has demonstrated that the OR of 30day stroke and also stroke or death with CAS is about 1.6 compared with CEA; the magnitude of this effect size clearly indicates the presence of a clinically significant difference. Indeed, stroke or death rate at 30 days was significantly higher for CAS (2.94%) than for CEA (1.89%), and this 1% difference may be viewed as being excessively high for the baseline risk of CEA. Nonetheless, this should be taken into account when CAS is suggested for a patient who is at “high risk” for CEA. Contemporary medical treatment, preoperative imaging and selection of patients, surgical and anesthesia techniques, and postoperative care have all significantly improved patient outcomes, as recently outlined by a metaregression analysis.39 The low evidence of heterogeneity beyond chance in our meta-analysis indicates that the direction of the effect size is similar across studies that span two decades. Of note, a similar increase in risk (relative risk of about 1.5) with CAS has been reported by meta-analyses of symptomatic patients.6,40-42 The fact that the magnitude of the risk estimate is about the same in RCTs of asymptomatic and symptomatic patients indicates that the increased risk with CAS is a result of the increased risk in the control group of CEA, which is higher in symptomatic patients. Nevertheless, the complexity of the mechanisms responsible for perioperative stroke after CEA or CAS,43,44 including the

operator’s experience and temporal proximity with the ischemic event in symptomatic patients,45,46 is responsible for the wide variation of stroke and death in contemporary data sets.45 The loss of significance for ipsilateral stroke alone at 30 days in the comparison of CEA vs CAS can be explained by the reduced number of events compared with the all strokes variable that includes contralateral strokes as well, which are more common in CAS because of brain embolism as a result of aortic arch manipulation. We have not performed subgroup analyses for stroke severity (major vs minor) because the criteria for minor strokes do not indicate full recovery, and therefore there still may be significant residual disability.5 The outcomes of MI and CNI were less common at 30 days after CAS than after CEA. However, the difference in MI rates between the two groups was not significant, perhaps as a result of type II error; further limitations include the ambiguity and diversity of criteria for MI diagnosis used in the trials. We have opted not to present composite outcomes, as disability after an MI is clearly not equivalent with what is experienced after a postprocedural stroke. In our meta-analysis, CNI at 30 days was significantly lower for CAS compared with CEA (0.11% vs 3.21%, respectively), which has been also observed in a meta-analysis of mainly symptomatic patients.47 Nevertheless, CNI after CEA has been reported to be temporary

Journal of Vascular Surgery

Kakkos et al

615

Volume 66, Number 2

Table II. Summary of findings table using Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology for randomized controlled trials (RCTs) comparing carotid artery stenting (CAS) with carotid endarterectomy (CEA) for asymptomatic stenosis Illustrative comparative risksa (95% CI)

Outcomes

Assumed risk

Corresponding risk

CEA

CAS

Relative effect (95% CI)

No. of participants Quality of the evidence (studies) (GRADE)

Stroke or death at 30 days 19 per 1000

29 per 1000 (19-45)

OR, 1.57 (1.01-2.44)

3607 (8 studies)

4442 Moderateb,c,d

Stroke at 30 days

29 per 1000 (19-45)

OR, 1.63 (1.04-2.54)

3607 (8 studies)

4442 Moderateb,c,d

Ipsilateral stroke at 30 days 12 per 1000

19 per 1000 (9-40)

OR, 1.56 (0.71-3.43)

1737 (4 studies)

4442 Moderateb,c,d

MI

15 per 1000

8 per 1000 (4-17)

OR, 0.53 (0.24-1.14)

2885 (5 studies)

4422 Lowb,c

CNI

31 per 1000

4 per 1000 (2-8)

OR, 0.13 (0.07-0.26)

2855 (4 studies)

4444 Highb,c,e

37 per 1000 (25-53)

OR, 1.51 (1.02-2.24)

3603 (8 studies)

4442 Moderateb,c,d

18 per 1000

Stroke or death at 30 days 24 per 1000 plus late ipsilateral stroke

GRADE Working Group grades of evidence: High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. CI, Confidence interval; CNI, cranial nerve injury; MI, myocardial infarction; OR, odds ratio. a The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). b Downgraded because of risk of bias as a result of the nonblinded nature of the RCTs. c Downgraded because of small number of events. d Upgraded because effects were immediately apparent and consistent with RCTs of symptomatic patients. e Upgraded because of the very large effect size.

in the majority of cases,32,48 which should not defer patients from having CEA in view of the increased stroke rate with CAS. Our meta-analysis has demonstrated that the longterm outcome of stroke or death rate at 30 days plus ipsilateral stroke during follow-up was significantly higher for CAS than for CEA (3.64% vs 2.45%, respectively). A similar finding has been reported in a meta-analysis of trials with predominantly symptomatic patients.49 Although this increased long-term risk may be attributed to ischemic brain lesions after CAS, which increase future cerebrovascular risk,50 the results of our meta-analysis indicated that most of the risk is the result of perioperative events in both CAS and CEA. Meta-analyses of CAS compared with CEA are abundant.6,39-42,47,49,51 Most of them included only symptomatic patients. A few of them included asymptomatic patients analyzed separately, but because of the small study size, they were underpowered to detect significant differences. Our meta-analysis has some limitations; quality assessment using the GRADE method of the present meta-analysis that included only RCTs revealed low and moderate evidence for the outcomes of MI and stroke, respectively. This downgrade of the quality of evidence was the result of serious imprecision due to the small number of events detected, indicating that

further research may be required with continuing enrollment into ongoing RCTs on asymptomatic carotid stenosis (eg, Asymptomatic Carotid Surgery Trial [ACST] 2). Also, the number of MIs was too low to permit statistical significance, and more trials are thus required. Nevertheless, the result of our meta-analysis implies that CEA should be considered in preference to CAS, after discussing its indications, benefits, and disadvantages with the patients and referring physicians, taking into account the already known negative effect of age on CAS outcome.52 Future work could formally explore the effects of the patient’s age and publication year on the effect size of the outcome measures; however, the lack of heterogeneity in all forest plots indicates the lack of any changes in the relative safety and effectiveness of CAS over CEA during the period in which these studies were conducted. In addition, the relatively small number of trials included in the present meta-analysis did not allow us to proceed with an adequate, formal estimation of publication bias.

CONCLUSIONS Among patients with asymptomatic carotid stenosis who are candidates for carotid intervention, there is evidence to suggest that CAS had higher 30-day stroke or death rates compared with CEA. The long-term efficacy of CEA in ipsilateral stroke prevention, taking into account perioperative stroke and death, was preserved

616

Journal of Vascular Surgery

Kakkos et al

August 2017

during follow-up. Because the research is judged to be of moderate quality, due to serious risk of bias and imprecision, there is an urgent need for further high-quality research.  We would like to thank David Skoloudík for providing information on asymptomatic patients included in his trial.

11.

12.

AUTHOR CONTRIBUTIONS Conception and design: SK, IK, IT, GG Analysis and interpretation: SK, IK, IT, GG Data collection: SK, IK Writing the article: SK, IK, IT, GG Critical revision of the article: IK, IT, GG Final approval of the article: SK, IK, IT, GG Statistical analysis: SK Obtained funding: Not applicable Overall responsibility: SK

13.

14.

REFERENCES 1. Eckstein HH, Ringleb P, Allenberg JR, Berger J, Fraedrich G, Hacke W, et al. Results of the Stent-Protected Angioplasty versus Carotid Endarterectomy (SPACE) study to treat symptomatic stenoses at 2 years: a multinational, prospective, randomised trial. Lancet Neurol 2008;7:893-902. 2. Ederle J, Dobson J, Featherstone RL, Bonati LH, van der Worp HB, de Borst GJ, et al. Carotid artery stenting compared with endarterectomy in patients with symptomatic carotid stenosis (International Carotid Stenting Study): an interim analysis of a randomised controlled trial. Lancet 2010;375:985-97. 3. Brooks WH, McClure RR, Jones MR, Coleman TL, Breathitt L. Carotid angioplasty and stenting versus carotid endarterectomy for treatment of asymptomatic carotid stenosis: a randomized trial in a community hospital. Neurosurgery 2004;54:318-24; discussion: 324-5. 4. Brott TG, Hobson RW 2nd, Howard G, Roubin GS, Clark WM, Brooks W, et al. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med 2010;363:11-23. 5. Rosenfield K, Matsumura JS, Chaturvedi S, Riles T, Ansel GM, Metzger DC, et al. Randomized trial of stent versus surgery for asymptomatic carotid stenosis. N Engl J Med 2016;374: 1011-20. 6. Kakisis JD, Avgerinos ED, Antonopoulos CN, Giannakopoulos TG, Moulakakis K, Liapis CD. The European Society for Vascular Surgery guidelines for carotid intervention: an updated independent assessment and literature review. Eur J Vasc Endovasc Surg 2012;44:238-43. 7. Eckstein HH, Reiff T, Ringleb P, Jansen O, Mansmann U, Hacke W. SPACE-2: a missed opportunity to compare carotid endarterectomy, carotid stenting, and best medical treatment in patients with asymptomatic carotid stenoses. Eur J Vasc Endovasc Surg 2016;51:761-5. 8. CAVATAS investigators. Endovascular versus surgical treatment in patients with carotid stenosis in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS): a randomised trial. Lancet 2001;357:1729-37. 9. Yadav JS, Wholey MH, Kuntz RE, Fayad P, Katzen BT, Mishkel GJ, et al. Protected carotid-artery stenting versus endarterectomy in high-risk patients. N Engl J Med 2004;351: 1493-501. 10. McCabe DJ, Pereira AC, Clifton A, Bland JM, Brown MM. Restenosis after carotid angioplasty, stenting, or

15.

16. 17. 18.

19.

20.

21.

22.

23.

24.

25.

endarterectomy in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS). Stroke 2005;36:281-6. Gurm HS, Yadav JS, Fayad P, Katzen BT, Mishkel GJ, Bajwa TK, et al. Long-term results of carotid stenting versus endarterectomy in high-risk patients. N Engl J Med 2008;358:1572-9. Bonati LH, Ederle J, McCabe DJ, Dobson J, Featherstone RL, Gaines PA, et al. Long-term risk of carotid restenosis in patients randomly assigned to endovascular treatment or endarterectomy in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS): long-term follow-up of a randomised trial. Lancet Neurol 2009;8:908-17. Ederle J, Bonati LH, Dobson J, Featherstone RL, Gaines PA, Beard JD, et al. Endovascular treatment with angioplasty or stenting versus endarterectomy in patients with carotid artery stenosis in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS): long-term follow-up of a randomised trial. Lancet Neurol 2009;8:898-907. Ederle J, Featherstone RL, Brown MM. Long-term outcome of endovascular treatment versus medical care for carotid artery stenosis in patients not suitable for surgery and randomised in the Carotid and Vertebral Artery Transluminal Angioplasty study (CAVATAS). Cerebrovasc Dis 2009;28:1-7. Reiff T, Stingele R, Eckstein HH, Fraedrich G, Jansen O, Mudra H, et al. Stent-protected angioplasty in asymptomatic carotid artery stenosis vs. endarterectomy: SPACE2da three-arm randomised-controlled clinical trial. Int J Stroke 2009;4:294-9. Correction. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med 2010;363:498. Correction. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med 2010;363:198. Cohen DJ, Stolker JM, Wang K, Magnuson EA, Clark WM, Demaerschalk BM, et al. Health-related quality of life after carotid stenting versus carotid endarterectomy: results from CREST (Carotid Revascularization Endarterectomy Versus Stenting Trial). J Am Coll Cardiol 2011;58:1557-65. Mahoney EM, Greenberg D, Lavelle TA, Natarajan A, Berezin R, Ishak KJ, et al. Costs and cost-effectiveness of carotid stenting versus endarterectomy for patients at increased surgical risk: results from the SAPPHIRE trial. Catheter Cardiovasc Interv 2011;77:463-72. Silver FL, Mackey A, Clark WM, Brooks W, Timaran CH, Chiu D, et al. Safety of stenting and endarterectomy by symptomatic status in the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST). Stroke 2011;42:675-80. Voeks JH, Howard G, Roubin GS, Malas MB, Cohen DJ, Sternbergh WC 3rd, et al. Age and outcomes after carotid stenting and endarterectomy: the carotid revascularization endarterectomy versus stenting trial. Stroke 2011;42:3484-90. Hill MD, Brooks W, Mackey A, Clark WM, Meschia JF, Morrish WF, et al. Stroke after carotid stenting and endarterectomy in the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST). Circulation 2012;126:3054-61.  Kuliha M, Skoloudík D, Roubec M, Herzig R, Procházka V, Jonszta T, et al. Risk reduction of brain infarction during carotid endarterectomy or stenting using sonolysisd prospective randomized study pilot data. AIP Conference Proceedings 2012;1503:239. Lal BK, Beach KW, Roubin GS, Lutsep HL, Moore WS, Malas MB, et al. Restenosis after carotid artery stenting and endarterectomy: a secondary analysis of CREST, a randomised controlled trial. Lancet Neurol 2012;11:755-63. Vilain KR, Magnuson EA, Li H, Clark WM, Begg RJ, Sam AD 2nd, et al. Costs and cost-effectiveness of carotid stenting versus endarterectomy for patients at standard

Journal of Vascular Surgery

Kakkos et al

617

Volume 66, Number 2

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

surgical risk: results from the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST). Stroke 2012;43:2408-16. Timaran CH, Mantese VA, Malas M, Brown OW, Lal BK, Moore WS, et al. Differential outcomes of carotid stenting and endarterectomy performed exclusively by vascular surgeons in the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST). J Vasc Surg 2013;57:303-8. Bonati LH, Ederle J, Dobson J, Engelter S, Featherstone RL, Gaines PA, et al. Length of carotid stenosis predicts periprocedural stroke or death and restenosis in patients randomized to endovascular treatment or endarterectomy. Int J Stroke 2014;9:297-305. Brooks WH, Jones MR, Gisler P, McClure RR, Coleman TC, Breathitt L, et al. Carotid angioplasty with stenting versus endarterectomy: 10-year randomized trial in a community hospital. JACC Cardiovasc Interv 2014;7:163-8. Li FM, Zhong JX, Jiang X, Wang QZ, Guo Y. Therapeutic effect of carotid artery stenting versus endarterectomy for patients with high-risk carotid stenosis. Int J Clin Exp Med 2014;7: 2895-900. Malas M, Glebova NO, Hughes SE, Voeks JH, Qazi U, Moore WS, et al. Effect of patching on reducing restenosis in the Carotid Revascularization Endarterectomy versus Stenting Trial. Stroke 2015;46:757-61. Sheffet AJ, Voeks JH, Mackey A, Brooks W, Clark WM, Hill MD, et al. Characteristics of participants consenting versus declining follow-up for up to 10 years in a randomized clinical trial. Clin Trials 2015;12:657-63. Hye RJ, Mackey A, Hill MD, Voeks JH, Cohen DJ, Wang K, et al. Incidence, outcomes, and effect on quality of life of cranial nerve injury in the Carotid Revascularization Endarterectomy versus Stenting Trial. J Vasc Surg 2015;61:1208-14. Kuliha M, Roubec M, Prochazka V, Jonszta T, Hrbac T, Havelka J, et al. Randomized clinical trial comparing neurological outcomes after carotid endarterectomy or stenting. Br J Surg 2015;102:194-201. Kuliha M, Roubec M, Prochazka V. Randomized clinical trial comparing neurological outcomes after carotid endaterectomy or stenting. J Vasc Surg 2015;62:519. Voeks JH, Howard G, Roubin G, Farb R, Heck D, Logan W, et al. Mediators of the age effect in the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST). Stroke 2015;46:2868-73. Brott TG, Howard G, Roubin GS, Meschia JF, Mackey A, Brooks W, et al. Long-term results of stenting versus endarterectomy for carotid-artery stenosis. N Engl J Med 2016;374: 1021-31. Moore WS, Popma JJ, Roubin GS, Voeks JH, Cutlip DE, Jones M, et al. Carotid angiographic characteristics in the CREST trial were major contributors to periprocedural stroke and death differences between carotid artery stenting and carotid endarterectomy. J Vasc Surg 2016;63:851-8.e1. Mannheim D, Karmeli R. A prospective randomized trial comparing endarterectomy to stenting in severe asymptomatic carotid stenosis. J Cardiovasc Surg (Torino) 2017; [Epub ahead of print]. Munster AB, Franchini AJ, Qureshi MI, Thapar A, Davies AH. Temporal trends in safety of carotid endarterectomy in

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

asymptomatic patients: systematic review. Neurology 2016;86:312-3. Liu Z, Shi Z, Wang Y, Chen B, Zhu T, Si Y, et al. Carotid artery stenting versus carotid endarterectomy: systematic review and meta-analysis. World J Surg 2009;33:586-96. Murad MH, Shahrour A, Shah ND, Montori VM, Ricotta JJ. A systematic review and meta-analysis of randomized trials of carotid endarterectomy vs stenting. J Vasc Surg 2011;53: 792-7. Zhang L, Zhao Z, Ouyang Y, Bao J, Lu Q, Feng R, et al. Systematic review and meta-analysis of carotid artery stenting versus endarterectomy for carotid stenosis: a chronological and worldwide study. Medicine (Baltimore) 2015;94:e1060. Gaunt ME. Transcranial Doppler: preventing stroke during carotid endarterectomy. Ann R Coll Surg Engl 1998;80: 377-87. Huibers A, Calvet D, Kennedy F, Czuriga-Kovacs KR, Featherstone RL, Moll FL, et al. Mechanism of procedural stroke following carotid endarterectomy or carotid artery stenting within the International Carotid Stenting Study (ICSS) randomised trial. Eur J Vasc Endovasc Surg 2015;50: 281-8. Paraskevas KI, Kalmykov EL, Naylor AR. Stroke/death rates following carotid artery stenting and carotid endarterectomy in contemporary administrative dataset registries: a systematic review. Eur J Vasc Endovasc Surg 2016;51:3-12. Rantner B, Goebel G, Bonati LH, Ringleb PA, Mas JL, Fraedrich G. The risk of carotid artery stenting compared with carotid endarterectomy is greatest in patients treated within 7 days of symptoms. J Vasc Surg 2013;57:619-26.e2; discussion: 625-6. Economopoulos KP, Sergentanis TN, Tsivgoulis G, Mariolis AD, Stefanadis C. Carotid artery stenting versus carotid endarterectomy: a comprehensive meta-analysis of short-term and long-term outcomes. Stroke 2011;42:687-92. Fokkema M, de Borst GJ, Nolan BW, Indes J, Buck DB, Lo RC, et al. Clinical relevance of cranial nerve injury following carotid endarterectomy. Eur J Vasc Endovasc Surg 2014;47:2-7. Vincent S, Eberg M, Eisenberg MJ, Filion KB. Meta-analysis of randomized controlled trials comparing the long-term outcomes of carotid artery stenting versus endarterectomy. Circ Cardiovasc Qual Outcomes 2015;8(Suppl 3): S99-108. Gensicke H, van der Worp HB, Nederkoorn PJ, Macdonald S, Gaines PA, van der Lugt A, et al. Ischemic brain lesions after carotid artery stenting increase future cerebrovascular risk. J Am Coll Cardiol 2015;65:521-9. Bonati LH, Lyrer P, Ederle J, Featherstone R, Brown MM. Percutaneous transluminal balloon angioplasty and stenting for carotid artery stenosis. Cochrane Database Syst Rev 2012;9:CD000515. Antoniou GA, Georgiadis GS, Georgakarakos EI, Antoniou SA, Bessias N, Smyth JV, et al. Meta-analysis and metaregression analysis of outcomes of carotid endarterectomy and stenting in the elderly. JAMA Surg 2013;148:1140-52.

Submitted Nov 16, 2016; accepted Apr 10, 2017.