Computed tomography versus color duplex ultrasound for surveillance of abdominal aortic stent-grafts.

PURPOSE: To compare the ability of computed tomography (CT) and color duplex ultrasound (CDUS) to detect endoleak and accurately measure aortic aneurysm diameters after endovascular repair.
METHODS: Between February 2000 and October 2004, 178 consecutive patients (156 men; mean age 74 years, range 49-89) were treated with aortic stent-grafts (86 Ancure, 55 AneuRx, and 37 Excluder). The follow-up protocol included serial CT and CDUS at 1 month and every 6 months thereafter. Sensitivity, specificity, positive predictive value, negative predictive value, and Kappa statistics (kappa) were calculated using CT as the gold standard; Bland-Altman analysis was used to determine the 95% limits of agreement. Paired and unpaired t tests and correlation coefficients were used to compare the methods.
RESULTS: Follow-up ranged from 1 to 53 months (mean 16), during which 367 paired CT and CDUS studies were acquired. The mean diameter of the AAA sac after repair was 5.15 cm by CT versus 4.99 cm by CDUS (p=0.07); 93% of paired studies were somewhat similar (<or=5 mm). Mean pre to postoperative AAA size changes throughout follow-up were -0.60 mm for CT versus -0.58 mm for CDUS (p=0.78). Thirty-four (19%) endoleaks were detected (26 early and 8 late). Versus CT, the sensitivity, specificity, positive predictive value, and negative predictive value of CDUS for detecting endoleaks were 68%, 99%, 85%, and 97%, respectively (kappa=0.73). CDUS was more accurate in detecting type I endoleak than type II (88% versus 50%, p=0.046).
CONCLUSIONS: Although CDUS has good correlation to CT in measuring the size of AAAs, it has a lower sensitivity in detecting endoleak, particularly type II. Therefore, CT scans should remain the primary imaging modality for the diagnosis of endoleak.