In Vitro Geometry Analysis of Fenestrations in Endovascular Aneurysm Repair

PURPOSE: Changes in the flared end of balloon-expandable covered stent (BECS) may precede BECS-associated complications but are not regularly assessed with computed tomographic angiography (CTA) after fenestrated endovascular aneurysm repair (FEVAR). Validation of the flare geometric analysis (FGA) and assessment of intraobserver and interobserver variability are investigated in this study. METHODS: Two series of 3 BeGraft BECSs (Bentley InnoMed GmbH, Hechingen, Germany) and 1 series of 3 Advanta V12 BECSs (Getinge AB, Göteborg, Sweden) were deployed in 3 side branches (45°, 60°, and 90° aortic branch angles) of an aorta phantom model. A standard post-FEVAR CTA scan was acquired. Computed tomographic angiography–derived measurements consisted of centerline reconstructions and placement of 3-dimensional coordinate markers by 2 observers in a vascular workstation. Flare geometric analysis calculates 3 BECS parameters: the circumferential flare-to-fenestration distance (FFD), which is the distance from the proximal end of the flare to fenestration, and diameters at the proximal end of the flare (Dflare) and at the fenestration (Dfenestration). Computed tomographic angiography–derived measurements were validated against microscopy measurements. Bland-Altman plots were used to determine the intraobserver and interobserver variability of the BECS parameters and intraclass correlation coefficient (ICC). RESULTS: For each BECS, the FFD at 4 equidistant quadrants of the circumference, Dflare, and Dfenestration were calculated. The mean difference and repeatability coefficient (RC) of the validation were 0.8 (2.1) mm for FFD, 0.4 (1.0) mm for Dflare, and −0.2 (1.2) mm for Dfenestration. The mean intraobserver and interobserver difference (RC) was 0.5 (1.6) mm and 0.7 (2.6) mm for FFD, 0.1 (0.6) mm and 0.1 (0.7) mm for Dflare, and −0.1 (0.8) mm and −0.8 (1.0) mm for Dfenestration. The mean ICC of intraobserver variability was 0.86 for FFD, 0.94 for Dflare, and 0.78 for Dfenestration. The mean ICC of interobserver variability was 0.77 for FFD, 0.92 for Dflare, and 0.48 for Dfenestration. CONCLUSION: This study showed that FGA of the flared ends of BECS can be performed with high accuracy in a phantom model, with good intraobserver and interobserver variability. Flare geometric analysis can be used to determine flare geometry of the BECS on standard post-FEVAR CTA scans.