The Assessment of Carbon Dioxide Automated Angiography in the Type II Endoleaks Detection: Comparison with Contrast-Enhanced Ultrasound

Abstract

Introduction - The presence of type II endoleak (ELII) after endovascular aortic repair (EVAR) can be investigated with different methods. Since intraoperative completion angiography (ICA) may underestimate their presence, particularly if they are at low flow, contrast-enhanced ultrasound (CEUS) has been proposed as the gold standard during EVAR follow-up. Intra-procedural carbon dioxide (CO2) angiography has been shown to be useful in this setting, however no comparative studies including these three techniques are currently available. Aim of our study is to investigate the accuracy of a new automated CO2 angiographic (CO2A) system in the detection of ELII, by comparing it with ICA and CEUS. Methods - A series of consecutive patients undergoing EVAR for Abdominal Aortic Aneurysm (AAA) were enrolled and submitted to ICA and CO2A during the procedure. The iodinate contrast media were delivered through an automatic injector connected to a pigtail catheter in the suprarenal aorta, with constant flow velocity and volume injection. The CO2 was delivered through a recently available automatic injector connected to a 10F, 11mm long sheet positioned in the external iliac artery, with constant CO2 injection in terms of flow velocity and volume rate. All patients were blindly evaluated by CEUS on postoperative day1. The presence of ELII was defined as contrast enhancement into the residual sac appearing with a ≥ 5 seconds delay from contrast injection. The ICA and CO2A ability to detect ELII was compared with that of CEUS through the Cohen's concordance Index (K).s Results - Twenty-one patients with a mean AAA size of 56±8mm, were entered in the study. There were no intra or peri-operative complications related to ICA or CO2A. No type I-III endoleak were detected by ICA or CO2A, as confirmed also by CEUS. One (5%), 7(33%) and 4(19%) ELII, were detected by ICA, CO2A and CEUS, respectively. The only ELII detected by ICA was also detected by CO2A and CEUS. Three cases of ELII detected by CO2A were not detected by CEUS. No cases of ELII undetected by CO2A were visualized by CEUS. A perfect agreement between CEUS and both ICA and CO2A was observed for type I/III endoleak (Cohen's K:1). CEUS and ICA showed a poor agreement for ELII detection (Cohen's K:0.35). A substantial agreement was observed between CEUS and CO2A for ELII (Cohen's K:0.65). Conclusion - CO2A is safe and effective method for ELII detection in EVAR, with a significantly higher agreement with CEUS if compared with ICA.