Contrast Enhanced Ultrasound Endoleak Identification and Classification

Overview

Surgery for abdominal aortic pathology usually occurs in the setting of advanced cardiovascular disease. The repair can be relatively simple or complex with multiple steps including open repair, placing a stent with a catheter, and placing a complicated stent that allows for crossing vessels without occluding them. All these repairs require imaging follow up. The most commonly accepted tool for follow up is Computed Tomography (CT) scan which involves ionizing radiation and potentially nephrotoxic iodinated contrast. Recommendations for the time interval for follow up, as well the radiology imaging technique vary. Routine ultrasound with Doppler, CT, and MRI has all been employed utilizing various imaging protocols. Clinicians use non-contrast CT, arterial phase and delayed phase CT, ultrasound, and various combinations based on personal experience and patient pathology. Concerns over cost, potential nephrotoxicity of contrast agents and repeated radiation exposure has led to investigation of alternate imaging modalities such as contrast-enhanced ultrasound (CEUS). CEUS represents an improvement of ultrasound imaging but comparisons against CT report widely varying results, likely due to technical factors of CEUS and limitations of single-phase CTA. Contrast ultrasound has been used effectively to diagnose leaks in the aorta post repair and is without the radiation and potential nephrotoxicity of iodinated contrast. Of yet, no large prospective studies have compared CT and contrast US and no studies have looked at the more complicated staged or fenestrated repairs. This study proposes to perform a contrast ultrasound at the same time as a contrast CT using a standardized protocol. This protocol would include a non-contrast CT, angiographic CT, and a CT in a delayed phase in all patients as standard of care. We will compare the results of a contrast US with the various data derived from a three phase CT.

Introduction: Surgery for abdominal aortic pathology usually occurs in the setting of advanced cardiovascular disease. The repair can be relatively simple or complex with multiple steps including open repair, placing a stent with a catheter, and placing a complicated stent that allows for crossing vessels without occluding them. All these repairs require imaging follow up. Rationale: The most commonly accepted tool for follow up is Computed Tomography (CT) scan which involves ionizing radiation and potentially nephrotoxic iodinated contrast. Routine ultrasound with Doppler, CT, and MRI has all been employed utilizing various imaging protocols. Concerns over cost, potential nephrotoxicity of contrast agents and repeated radiation exposure has led to investigation of alternate imaging modalities such as contrast enhanced ultrasound (CEUS). CEUS also allows continuous (dynamic) or real-time monitoring of the aneurysm and endoleak throughout the study. Objectives: Aim 1: To determine the consistency/discrepancy of endoleak detection (positive/negative) for contrast enhanced ultrasound (CEUS) versus various phases of CTA versus delayed phase CTA. Aim 2 (Exploratory): To explore the consistency/discrepancy of endoleak type (type 1-5) between CEUS and delayed phase CTA. Aim 3 (Exploratory): To explore the consistency/discrepancy of endoleak source detected (positive/negative for each source) for type II endoleaks between CEUS and CTA. Aim 4 (Exploratory): Re-assess Aims 1-3 for the follow-up scans. The goal for this aim is to examine whether the consistency/discrepancy patterns discovered in Aims 1-3 can be repeated at the follow-up scans. Population: 40 patients who have undergone an EVAR or FEVAR for abdominal aortic aneurysms and are expected to have CTA as part of their standard of care will be enrolled in the study. Methodology: Patients will be evaluated with research contrast ultrasound at one month post-op, six months post-op and at one year. Data Analysis Plan: Binomial 95% confidence interval will be calculated for sensitivity and specificity. One- sided binomial test will be conducted to test whether the target sensitivity and specificity are superior to the null value, e.g. a minimal acceptable level of sensitivity and specificity. In detecting the rate of endoleak positive on CEUS or other phases of the CTA but negative on CTA in the delayed phase, we will compute Clopper Pearson Exact 95% confidence interval since such a rate could be small. To assess consistency of endoleak type (type 1-5) between CEUS and delayed phase CTA, Kappa coefficient will be used. In the situation with high agreement biased towards one endoleak type, or extremely unbalanced distribution of endoleak types, prevalence-adjusted bias-adjusted kappa coefficient (PABAK) (Byrt, Bishop, and Carlin 1993) will be used. The discrepancy of endoleak type between CEUS and delayed phase CTA will be reported as rate, for example percent of case rated as type 1 in CTA but type 2 in delayed phase, with Clopper Pearson Exact 95% confidence interval. SAS9.4 will be used for all statistical analyses.