Increased recognition of type II endoleaks using a modified intraoperative angiographic protocol: implications for intermittent endoleak and aneurysm expansion.

Retrograde arterial perfusion of the aneurysm sac (type II endoleak) may complicate endovascular abdominal aortic aneurysm (AAA) repair and may lead to AAA expansion and rupture. Aneurysm expansion may also occur in the absence of a demonstrable endoleak. Current intraoperative assessment techniques may underrepresent the incidence of type II endoleaks. This study evaluated the incidence and impact of previously unrecognized type II endoleaks using a modified intraoperative angiographic protocol. A total of 391 patients undergoing endovascular AAA repair were evaluated. In 264 patients standard completion angiograms were performed. In 127 patients a modified angiographic protocol was used to visualize collateral lumbar and inferior mesenteric arteries as well as the aneurysm sac. The modified protocol uses digital subtraction fluoroscopy continuously for 60 sec after injections of 20 mL iodinated contrast both in the pararenal aorta and within the endovascular graft. Postoperative CT scans were performed at 1, 6, and 12 months and annually thereafter. The average age was 73.3 years; 324 patients were men and 67 were women. Mean follow-up was 11.4 months (range, 1-60 months). Type II endoleaks were documented intraoperatively in a significantly increased proportion of patients in whom the modified angiographic protocol was used: modified, 53/127 = 41% vs. standard, 17/264 = 6%; p < 0.001. No significant difference in the incidence of type II endoleaks was present on CT scan at 6 or 12 months after surgery (6 months: modified, 6/72 = 8% vs. standard, 10/159 = 6%, p = NS; 12 months: modified, 2/36 = 5% vs. standard, 6/138 = 4%, p = NS). Forty-six type II endoleaks resolved spontaneously (10 in the standard cohort, 36 in the modified cohort). One patient had a 10-mm increase in AAA diameter after spontaneous thrombosis of a type II endoleak 18 months postoperatively. One patient had a type II endoleak intraoperatively and at 12 months after surgery but the endoleak was absent at 1 and 6 months. Thirteen patients from the standard protocol cohort and 1 from the modified protocol cohort developed newly visualized type II endoleaks during follow-up. These findings may imply intermittent patency of the artery supplying the type II endoleak. The overall morbidity rate was 14% and the perioperative mortality rate was 1.8%. Retrograde (type II) endoleaks originating from AAA side branches occur intraoperatively more frequently than is currently recognized. Intermittent patency and thrombosis of these vessels may also occur and may contribute to AAA expansion. The full significance of these previously unrecognized endoleaks with respect to risk of aneurysm rupture remains to be definitively determined.