Refining the indications for carotid endarterectomy in asymptomatic carotid stenosis: a systematic review and metaanalysis

TMMastracci (1), G Arena (2), CM Clase (3), CS Cinà (1)
(1) Division of Vascular Surgery and Department of Clinical Epidemiology and Biostatistics, McMaster University
(2) Department of Surgery, Division of General Surgery, McMaster University
(3) Departments of Medicine and Clinical Epidemiology and Biostatistics, McMaster University

Objective:  We undertook a systematic review and metaanalysis to refine the indications for carotid endarterectomy (CE) in asymptomatic carotid stenosis.

Methods:  Database searching, study eligibility, quality assessment, and data extraction were performed independently in duplicate.  Electronic databases were searched using a modified Cochrane search strategy.  Reference lists of retrieved and review articles, relevant textbook chapters, and authors’ personal files were also searched.  The outcomes chosen were ‘all stroke (long term) and perioperative death’ and ‘major or disabling stroke (long term) and perioperative death’.  Secondary analyses and subgroups for analysis were determined a priori.  Agreement was calculated with Cohen’s Kappa.  Random effects models were used to estimate relative risk (RR).

Results:  The search yielded 594 citations; 44 articles were retrieved, and 7 randomized controlled trials (5961 patients) met eligibility criteria.  Agreement for judgment of relevance was excellent (K = 0.92, 95% CI 0.85 – 0.96).  The degree of stenosis required for eligibility in the original studies was variable (>50% stenosis to >60% stenosis).  One study was excluded post hoc from the meta-analysis because of methodologic differences.  The pooled results for all outcomes are presented in Table 1.  The number needed to treat presented in this table is calculated based on pooled absolute risk reduction across all studies.  Event rates in the control groups for all stroke (long term) and perioperative death vary from 5.5% to 12.9%.  Applying the pooled relative risk reduction to these control event rates, the number needed to treat varies from 56 to 24, respectively.  Outcomes stratified by age, degree of stenosis, and sex were available for 2 of the 6 randomized controlled trials did not reveal statistically significant differences.

Conclusion:  Carotid endarterectomy modestly reduces the risk of stroke or perioperative death in patients with asymptomatic carotid artery stenosis. 

Table 1: Outcomes for metaanalysis of randomized controlled trials comparing carotid endarterectomy with best medical management in asymptomatic carotid stenosis.