Prospective nonrandomized study to evaluate cone beam computed tomography for technical assessment of standard and complex endovascular aortic repair

Objective: The objective of this study was to analyze the utility of cone beam computed tomography (CBCT) for technical assessment of standard and complex endovascular aneurysm repair (EVAR). Methods: Data of consecutive patients who underwent standard or complex EVAR in 2016 and 2017 at our institution were entered into a prospective database and analyzed retrospectively. There were 154 patients (126 male; mean age, 74 +/- 8 years) enrolled in a prospective study between 2016 and 2017. A total of 170 aortic procedures were investigated, including 85 fenestrated-branched EVARs (F-BEVARs), 42 abdominal and thoracic EVARs, 32 EVARs with iliac branch devices, and 11 aorta-related interventions. Technical assessment was done using CBCT with and without contrast enhancement, digital subtraction angiography (DSA), and computed tomography angiography (CTA). Patients with stage 3B or stage 4 chronic kidney disease had CBCT without contrast enhancement. Radiation exposure (mean dose-area product), effective dose (ED), and amount of iodine contrast agent were analyzed. End points were presence of any endoleak, positive findings warranting possible intervention (stent kink or compression, type I or type III endoleak, dissection, thrombus), and need for secondary intervention. Results: Radiation exposure and amount of iodine contrast agent were significantly higher (P < .05) for F-BEVAR compared with other aortic procedures (174 +/- 101 Gy.cm(2) vs 1135 +/- 113 Gy.cm(2) and 144 +/- 60 mL vs 122 +/- 49 mL). ED averaged 74 +/- 36 mSv for the aortic procedure, 18 +/- 18 mSv for fluoroscopy, 7 +/- 7 mSv for DSA acquisition, 15 +/- 7 mSv for CBCT, and 34 +/- 17 mSv for CTA imaging (P < .001). Endoleak detection was significantly higher (P < .001) with CBCT (53%) compared with DSA (14%) and CTA (46%). CBCT identified 52 positive findings in 43 patients (28%), higher for F-BEVAR compared with other aortic procedures (35% vs 16%; P = .01). Positive findings included stent compression or kink in 29 patients (17%), type I or type III endoleak in 16 patients (10%), and arterial dissection or thrombus in 7 patients (5%). Of these, 28 patients (18%) had positive findings that prompted an intraoperative (17%) or delayed intervention (1%). Another 15 patients (10%) with minor positive findings were observed with no clinical consequence. DSA alone would not have detected positive findings in 34 of 43 patients (79%), including 21 patients (49%) who needed secondary interventions. CTA diagnosed two (1%) additional endoleaks requiring intervention (one type IC, one type IIIC) that were not diagnosed by CBCT. Replacing DSA and CTA by CBCT would have resulted in 53% +/- 13% reduction in amount of iodine contrast agent and 55% +/- 12% reduction in ED (P < .05). Conclusions: CBCT reliably detected positive findings prompting immediate revisions in nearly one of five patients, with the highest rates among F-BEVAR patients. Detection of any endoleak was higher with CBCT compared with DSA or CTA, but most endoleaks were observed. DSA alone failed to detect positive findings warranting revisions.