Ammad Mahmood, MBChB
Howard VJ, Algra A, Howard G, Bonati LH, de Borst GJ, Bulbulia R, Calvet D, Eckstein H-H, Fraedrich G, Greving JP, et al. Absence of Consistent Sex Differences in Outcomes From Symptomatic Carotid Endarterectomy and Stenting Randomized Trials. Stroke. 2021;52:416–423.
Management of carotid artery stenosis has been approached differently in women and men since the 1990s following the NASCET criteria.1 More recently, trials comparing carotid artery stenting with endarterectomy have shown the modification of treatment effect by sex to be variable. The Carotid Revascularisation Endarterectomy vs Stenting Trial (CREST)2 suggested women receiving stenting were at higher periprocedural risk of stroke, myocardial infarction or death compared to men, though there was no difference in the primary endpoint (periprocedural risks above or ipsilateral stroke within 4 years). Conversely, the International Carotid Stenting Study (ICSS)3 suggested a borderline higher periprocedural risk for men undergoing stenting.
This meta-analysis pooled data from four trials of carotid artery stenting vs endarterectomy — CREST, ICSS, Endarterectomy versus Angioplasty in Patients with Symptomatic Severe Carotid Stenosis (EVA-3S)4 and Stent-Protected Angioplasty versus Carotid Endarterectomy in Symptomatic Patients (SPACE).5 A total of 4754 patients, 1437 women (30%), were included. The majority of events were periprocedural (335/433). Combining periprocedural risk and risk of ipsilateral stroke thereafter, the risk for women of stenting versus endarterectomy was found to be significantly lower in EVA-3S, non-significantly lower in ICSS and non-significantly higher in both SPACE and CREST. Overall, they found heterogeneity and inconsistency between trials, making the pooling of data inappropriate to draw conclusions from. Pooled data showed a trend to higher risk of stenting which did not reach significance in women but did in men, with no significant difference between women and men. Similar results were seen when considering periprocedural risk alone.