Surveillance computed tomographic arteriogram does not change management before 3 years in patients who have a normal post-EVAR study.

BACKGROUND: Standard surveillance after endovascular abdominal aortic aneurysm repair (EVAR) consists of periodic computed tomographic arteriographies (CTAs) usually performed at postoperative months 1, 6, and 12, and then annually. This imaging regimen is expensive and exposes patients to the hazards of radiation and intravenous contrast. We hypothesized that a normal 1-month CTA after EVAR with no endoleak or other significant abnormality predicts a low rate of future complications, which would justify a reduction in frequency of subsequent CTAs.

METHODS: We identified 106 consecutive patients who underwent EVAR at a single hospital from 2003 to 2010 and reviewed all their CTAs. Fifteen patients for whom we could not review a postoperative CTA were excluded. Of the remaining 91 patients, 70 (76.9%) had no abnormality on their CTA at 1 month after EVAR. The medical records of these 70 patients were analyzed for subsequent complications and interventions related to EVAR.

RESULTS: The mean patient follow-up was 3.4 ± 2.1 years. Five of the 70 (7.1%) patients with a normal post-EVAR CTA developed late complications consisting of 1 type I endoleak, 3 type II endoleaks, and 1 case of endotension. Only the type I endoleak and one of the type II endoleaks met criteria for intervention, and in both cases, the endoleaks were discovered >3 years after EVAR. Log-rank test showed a statistically significant increased freedom from aneurysm sac expansion in patients with a normal compared with an abnormal 1-month CTA (P < 0.001).

CONCLUSIONS: For patients who have a normal CTA with no endoleak 1 month after EVAR, it is reasonable to consider less-frequent CTA surveillance because no significant complications requiring intervention occurred before 3 years. This would decrease unnecessary CTAs and health care expenditures as well as minimize patient exposure to radiation and intravenous contrast.