Risk Assessment for a Learning Curve in Endovascular Abdominal Aortic Aneurysm Repair with the Zenith Stent-Graft: The First Year in Japan

(Department of Cardiovascular Surgery, Tokyo Womenfs Medical University*, Tokyo, Japan and Department of Vascular Surgery, Tokyo Medical University* Tokyo, Japan)

Takashi Azuma*,** Satoshi Kawaguchi** Taro Shimazaki**
Kenji Koide** Masataka Matsumoto** Hiroshi Shigematsu**
Akihiko Kawai* and Hiromi Kurosawa*
In Japan, doctors inexperienced stent-graft new devices are required to secure agreement on criteria and choice of the device size in endovascular aneurysm repair (EVAR) from experienced doctors. It was hoped that strict patient selection might reduce the learning curve for initial successes in given procedures. In a leading center in Japan, a number of cases which were scheduled for operation at other institutes were evaluated anatomically. We surveyed the initial success of Zenith AAA system implantation in the remaining cases by inexperienced doctors and evaluated the results. This study aimed to verify the validity of strict patient selection in improving the success rate of inexperienced doctors. We enrolled 112 consecutive patients from 19 institutes, who were scheduled for repair between January and October in 2007. All patients were evaluated on the basis of a less-than-3mm reconstructed CT image. Mean patient age was 76}5.7 years. All cases satisfied the Zenithfs anatomic prerequisites. Fifteen cases were excluded for various reasons, the major reason being insufficiency of the proximal landing zone (LZ) length, angle and contour. The second reason was difficulty to approach via the iliac artery. Ninety seven cases were included, of which 17 cases were low-risk candidates for EVAR. Medium-risk seventy two cases requiring some advice to avoid problems with device size, technique of implantation and choice of main-body side. Eight cases were high-risk, requiring the presence of an experienced surgeon. Excluded cases had significantly shorter proximal LZ, larger aortic diameters 15mm below the renal artery and tortuous access routes on preliminary measurement by inexperienced doctor. Perioperative mortality was 0%, while the major complications were injury to the iliac artery in one high-risk case and thromboembolism of the superficial femoral artery in another. Perioperative proximal type I endoleak occurred in 5 cases. In 3 of these cases, the endoleak was eliminated by implantation of a Palmatz stent. In the other 2 cases, it disappeared within a month without additional procedures. These cases had a significantly greater angle between the proximal LZ and the suprarenal aorta and significant amount of mural thromboses in the proximal LZ. Perioperative type III endoleak occurred in 3 cases. In all cases the endoleak was eliminated by additional procedure. Perioperative type II endoleak occurred 8 cases. In 3 of these cases, the endoleak disappeared within a month. In the 5 other cases, the endoleak did not disappear. Mid-term results showed iliac leg thromboembolism in one case and new type II endoleaks in 3 caces. Type II endoleak occurred in cases which had significantly greater angles between the proximal LZ and the aneurysm. The results which were evaluated in our center had excellent perioperative and mid-term outcomes. We think this evaluation system is effective for risk assessment and reduces the learning curve in EVAR. In anatomically marginal cases, it is possible for proximal type I endoleak and injury of the iliac artery to occur. It is impossible to exclude these marginal cases if treatment need for EVAR is a priority. In these cases, lessexperienced operators should be trained in troubleshooting techniques in advance.
@Jpn. J. Cardiovasc. Surg. 37: 311-316 (2008)