The value of the initial post-EVAR computed tomography angiography scan in predicting future secondary procedures using the Powerlink stent graft

Objective

Current long-term surveillance after endovascular abdominal aortic aneurysm repair (EVAR) is based on high-resolution contrast-enhanced computed tomography (CT) scans at scheduled, lifelong intervals. The cancer and nephrotoxicity risks of interval CT scanning and prolonged radiation exposure are concerning. We sought to determine if surveillance CT angiography (CTA) can be safely reduced.

Methods

From July 2000 to November 2007, 345 patients were enrolled in U.S. Food and Drug Administration trials of the Powerlink System (Endologix, Irvine, Calif). An independent core laboratory analyzed 1519 post-EVAR CT scans (N = 1519) to 5 years to evaluate aneurysm size, migration, presence of endoleak, and evidence of graft obstruction. Analyses were conducted to determine the value of the initial CTA scan in predicting future secondary procedures in enrolled patients.

Results

At any time during follow-up, CTA identified endoleak in 123 patients (36%), with 95% of endoleaks being type II. In addition, 49 patients underwent 72 secondary procedures at a mean of 22 ± 21 months (range, 2-2007 days) after initial EVAR. These were based on clinical identification of limb ischemia in 13 interventions (18%) or core laboratory identification of abnormal CT finding in 58 interventions (81%). Of the 58 core laboratory identified findings, the inciting abnormality was present on the initial postoperative scan in 49 (84%). Of the remaining nine CT-driven procedures, three (5.2%) were due to late sac expansion attributed to type II endoleak (n = 2) or endotension (n = 1); two (3.4%) were for prophylactic reasons in the absence of endoleak; and four (6.8%) were in patients with type II endoleak not observed by the core laboratory and without sac expansion. The negative predictive value of the initial postoperative CTA for the need for a secondary procedure is therefore 96.4%, which can be improved to 97.6% with duplex ultrasound surveillance to detect sac expansion. Thus, a negative initial postoperative CTA is highly predictive of long-term freedom from secondary intervention.

Conclusions

Among enrolled patients with suitable anatomy for EVAR, most abnormalities that result in a secondary procedure are detected on the initial postoperative CTA or present with clinical symptoms. Long-term surveillance CTA may therefore be replaced by duplex ultrasound imaging if the initial postoperative CTA shows no abnormalities.