Dinah B. Quiachon

Preliminary experience with the Endovascular Technologies* bifurcated endovascular aortic prosthesis in a calf model

PURPOSE The purpose of this study was to develop a bifurcated endoluminal prosthesis for transfemoral placement in the aortoiliac position with a large-animal model that would simulate human implantation. METHODS Fifteen calves (160 to 200 kg) underwent bilateral femoral artery exposure and transfemoral placement in the aortoiliac position of a bifurcated Dacron prosthesis, inserted through a 26F sheath with an over-the-bifurcation guide wire to retrieve the contralateral limb and secured proximally and distally with self-expanding attachment systems. The preferred location was determined before implantation and compared with final implant location by caliper measurements and angiography. Events during implantation, maneuvers used to accomplish accurate deployment, and final results, as judged by angiography and clinical evaluation, were recorded. Four animals survived and were used for chronic evaluation and healing by gross and microscopic studies. RESULTS All grafts (n = 15) were patent at the end of the procedure. All (n = 7) noncrimped grafts had minor kinks in areas of curvature, whereas eight of eight crimped grafts (device modification) had none. Torque control of the body and individual limbs was necessary to correct twists before deployment in 10 of 15 grafts, with two greater than 30-degree twists remaining, which did not appear to affect flow. One implant was entirely deployed in the aorta because of parallax error, subsequently avoided by use of a marker board placed dorsally. Three premature deployments occurred, corrected by attachment system lock modification. The mean final implant position was within 2.9 +/- 1.4 mm (aortic), 1.6 +/- 1.1 mm, and 1.5 +/- 0.8 mm (contralateral and ipsilateral iliac limbs, respectively) of the intended position. Three of four animals intended for long-term evaluation were killed prematurely because of clinically evident spinal cord ischemia. Histologic sections at 2 weeks showed early wall repair without inflammatory cells and pannus ingrowth across the anastomosis. CONCLUSION We conclude that implantation of a bifurcated endovascular prosthesis through the bilateral femoral approach is possible, provided the intended aortic implantation site (neck) is at least 12 mm in length (mean +/- 2 SD each direction). Torque control of each portion of the device will be needed in the majority of instances, with attention to parallax effect necessary for optimal placement. This animal model is not suitable for chronic graft evaluation because of its sensitivity for spinal cord ischemia. Healing data suggest graft incorporation similar to that of a surgically placed prosthesis.