Stapled aortic anastomoses: a minimally invasive, feasible alternative to videoscopic aortic suturing?

Widespread applications of totally laparoscopic aortic reconstructions have been limited by the long cross-clamp time required to suture the aortic anastomosis despite improvement in instrumentation. The authors' hypothesis was that a "one-step anastomosis concept" using an intraluminal stapler would allow shorter cross-clamp time but similar patency and imperviousness as videoscopic suturing techniques. An intraluminal stapler (Endopath-ILS, Ethicon) with a modified anvil was used to perform videoscopic-assisted thoracic aorta-to-iliac artery bypass with a 21 mm by 8 mm polytetrafluoroethylene (PTFE) graft in 22 sheep through a minimally invasive approach using a 5 cm thoracotomy. The graft-to-iliac artery anastomoses were hand sutured through a flank incision. Twelve sheep were used to establish the technique and 10 subsequent animals constituted the study group. Aortic cross-clamp time, imperviousness, and need for additional sutures were recorded and compared to previously reported data using videoscopic suturing in pigs. Patency was assessed by comparing lower limb arterial pressures. Macroscopic and microscopic examinations of the anastomoses were performed at different time-points within the first 3 months. Videoscopic-assisted stapled anastomoses were also performed on atherosclerotic aortas of 3 human cadavers. Stapled anastomoses between the thoracic aorta and PTFE graft were completed in 8 of 10 animals. Two animals were euthanized after stapler failure and anastomotic bleeding. Sutures to strengthen the anastomosis had to be used in 4 cases. Mean aortic cross-clamp time in 8 successful cases was 4.3 +/-2.9 minutes (range 2-11 minutes) and was significantly shorter than clamp time of videoscopic suturing technique (48.7 +/-9.4 minutes, p < 0.0001). Imperviousness was good or excellent in 4 animals and fair in 4 animals. All anastomoses were patent at the end of the procedure. Examination of the anastomosis of the 2 failed interventions showed medial aortic tear surrounding the anastomosis in 1 case and misfired staples in the other. No graft occlusion was noted during follow-up ranging from 0 to 12 weeks. At the time of harvest, no bleeding was noted after epinephrine and volume infusion to increase mean arterial pressure to 200 mm Hg for 15 minutes. Macroscopic examination of the anastomoses revealed adequate healing with circumferential stapling of the prosthesis to the aortic wall and no stenosis or thrombus except in 1 false aneurysm (1/7, 14%). Surface electron microscopy showed cells coverage of the anastomosis surface. When applied on human cadaver thoracic and abdominal aorta with atherosclerotic changes, clamping times of less than 5 minutes were achieved. However, imperviousness tested with saline was poor. An automatic stapling device allows performance of a graft-to-aorta anastomosis through a minimally invasive approach with shorter clamping time than a videoscopic suturing technique. However, the current technique of aortic stapling is unreliable and further improvements are needed.