Purpose: To evaluate the optimal timing of delayed phase imaging for detecting low-flow endoleaks.
Materials and methods: Fifty-eight patients with unruptured abdominal aortic aneurysm treated with endovascular repair underwent 1- and 6-month follow-up multidetector row computed tomography (CT) performed during unenhanced, arterial, and delayed phase. At 6-month follow-up, delayed phase imaging, focused on stent graft, was performed with a delay of 60 (early delayed enhanced phase) and 300 seconds (late delayed enhanced phase) after intravenous injection of 120 mL of iodinated nonionic contrast medium (iomeprol 300 mgI/mL, Iomeron), at a flow rate of 3 mL/s via an antecubital vein, with a detector-row configuration of 4 x 1-mm, a 1.25-mm slice width, and a pitch of 6. Six-month follow-up CT images were independently evaluated by 2 readers during 2 different reading sessions: sets A (unenhanced, arterial, and early delayed phase images) and B (unenhanced, arterial, and late delayed phase images). Sensitivity and diagnostic accuracy of both reading sessions were compared. The standard of reference was represented by the combined evaluation of 1- and 6-month CT scans.
Results: At standard of reference, 24 of 58 patients had an endoleak classified as type 1 in 2 cases, type 2 in 21 cases, and type 3 in the last 1 case. Seven of 21 type 2 endoleaks were classified as low-flow endoleaks. Set A reading session, including early delayed enhanced phase, allowed the detection of 19 of 24 endoleaks (5 false-negative cases represented by low-flow endoleaks), whereas all endoleaks were detected during set B reading session, including late delayed enhanced phase. Differences between sets A and B in terms of sensitivity and diagnostic accuracy were statistically significant (P < 0.05).
Conclusions: For optimal multidetector CT detection of low-flow endoleaks in patients who underwent endovascular repair, delayed phase should be acquired 300 seconds after injection of contrast medium.