J Vasc Surg 2020 Nov 2;72(5):1593-1601. Epub 2020 Apr 2.
Division of Vascular and Endovascular Surgery, Beth Israel Deaconess Medical Center, Boston, Mass. Electronic address:
Objective: The stroke rate after endovascular aneurysm repair (EVAR), particularly complex EVAR such as fenestrated EVAR (FEVAR) and chimney EVAR (chEVAR), is not well defined. Whereas stroke is a well-established risk of thoracic endovascular aortic repair (TEVAR), the impact of procedural characteristics on stroke remains unclear. Therefore, we characterized the risk of stroke after endovascular aortic interventions in the Vascular Quality Initiative database and identified procedural characteristics associated with stroke.
Methods: We performed a retrospective cohort study of patients undergoing infrarenal EVAR, complex EVAR, and TEVAR within the Vascular Quality Initiative between 2011 and 2019. Complex EVAR included FEVAR (with either a Food and Drug Administration-approved custom-manufactured device or physician-modified endovascular graft) and chEVAR. We excluded emergent procedures. The primary outcome was in-hospital stroke. We used multivariable logistic regression to identify procedural characteristics associated with stroke.
Results: We identified 41,540 EVARs, 1371 complex EVARs, and 4600 TEVARs. The in-hospital stroke rate was 0.1% after EVAR, 0.9% after complex EVAR, and 2.9% after TEVAR. In patients undergoing EVAR, aneurysm diameter >6.5 mm (odds ratio [OR], 1.7; 95% confidence interval [CI], 1.1-2.7; P = .03) and use of a proximal extension cuff (OR, 3.3; 95% CI, 1.4-7.9; P < .01) were independently associated with stroke. Among complex EVARs, stroke rate was 0.7% after FEVAR with a custom-manufactured device, 0.4% after FEVAR with a physician-modified endovascular graft, and 2.1% after chEVAR (P = .08). In multivariable analysis, arm access was associated with 8.4-fold higher odds of stroke (95% CI, 1.7-41; P < .01). Whereas chEVAR was associated with higher odds of stroke in crude analysis, this association did not persist after adjustment for arm access (OR, 1.0; 95% CI, 0.2-4.4; P = .99). In patients undergoing TEVAR, more proximal landing zones were associated with higher risk of stroke compared with zone 4/5 (zone 3: OR, 2.0 [95% CI, 0.9-4.2]; zone 2: OR, 3.8 [95% CI, 1.8-8.2]; zone 0/1: OR, 6.3 [95% CI, 2.8-14]). In terms of procedural characteristics, any involvement of the left subclavian artery was associated with stroke (bypass: OR, 2.5 [95% CI, 1.5-4.0]; stent: OR, 2.7 [95% CI, 0.9-8.5]; covered or occluded: OR, 2.5 [95% CI, 1.5-4.1]).
Conclusions: Stroke, although rare after elective EVAR, is substantially more common after complex EVAR and TEVAR. Increasing procedural complexity in complex EVAR and TEVAR is associated with a higher stroke rate, a risk that should be factored into clinical decision-making. The strong association between stroke and upper extremity access during complex EVAR is alarming and warrants further study.