Time-Resolved, Dynamic Computed Tomography Angiography for Characterization of Aortic Endoleaks and Treatment Guidance via 2D-3D Fusion-Imaging

Endoleak still remains the major challenge after endovascular aortic aneurysm repair, and it can be misdiagnosed in terms of the location and the types of vessels which are feeding the endoleak. Unlike conventional CT imaging, dynamic time-resolved CT angiography has the advantage of visualizing the aneurysm sac during multiple time points after iodinated contrast injection. So this technique helps identify all the vessels contributing to the endoleak and the source of the endoleak after EVAR, and can be used for guiding endoleak therapy once the vessels have been imaged.

During dynamic-CTA imaging, multiple scans are required after an intravenous bolus of iodinated contrast agent. To mitigate radiation exposure, the imaging is tailored to focus only in the region of previously implanted stent graft. Studies have shown that this modality has better diagnostic accuracy than conventional triphasic CTA imaging for endoleaks.

Utilizing diagnostic CTR MR images for intra-procedural interventional guidance is not yet routine and standard of care. Diagnosing and treating aortic endoleaks often involves multiple 2D angiographic images acquired at different C-Arm angulations. In this dynamic-CT imaging, the 3D information from the imaging can be combined with intraoperative 2D fluoroscopic images using image fusion techniques such as 2D-3D image fusion to facilitate interventional treatment, and this often leads to limited radiation exposure and contrast usage during the interventional procedure.

Before starting the actual scan, review the prior imaging studies for the presence of an endoleak in stent graft type. For image acquisition, position the patient supine on the CT scanner table, then perform the topogram and non-contrast CT image acquisition using a stannum 100 tin filter to reduce the radiation exposure and select the region of interest in the dynamic-CTA scan. Next, perform a timing bolus to check the contrast arrival time by placing a region of interest above the stent graft in the abdominal aorta.

Then inject 10 to 20 milliliters of the contrast through the peripheral venous access, followed by 50 milliliters of saline injection at a 3.5 to 4 milliliter per minute flow rate and acquire the timing bolus scan. Next, plan the distribution and number of scans based on the contrast arrival time from the timing bolus and prior imaging findings. Optimize the imaging parameters including kilovolts, scan range and so on to reduce radiation exposure.

Then inject 70 to 80 milliliters of the contrast material for dynamic-CTA acquisition, followed by 100 milliliters of saline injections through the peripheral access. Start the dynamic-CTA image acquisition using the delay time based on the timing bolus described earlier. Breath hold is not necessary and the duration of image acquisition ranges from 30 to 40 seconds.

For image analysis, minimize the respiratory motion artifacts between the dynamic-CTA images by selecting the dedicated software's align body motion correction menu item. Next, perform a qualitative analysis. Check axial slices of the CT images when the maximum opacification of aorta occurred to interpret any obvious endoleak.

Then analyze the scans in multiplanar reconstruction mode. If an endoleak is suspected, focus on the endoleak and use the time scale to watch time-resolved images and infer the source of the endoleak. For quantitative analysis, click on the time attenuation curve function, select a region above the stent graft and draw a circle using the TAC function.

Then select the endoleak region and draw a circle there as well. Analyze the acquired time attenuation curve to determine the endoleak characteristics. Subtract the time to peak value of the endoleak in the aortic ROI curves to get the delta time to peak value.

Load the selected dynamic-CTA scan with the best visibility of the endoleak in the hybrid OR workstation. Then manually annotate the critical landmarks such as the renal arteries ostia, internal iliac arteries ostia, endoleak cavity, lumbar arteries, or inferior mesenteric artery. Select 2D to 3D image fusion in the workstation and inquire an anteroposterior and an oblique fluoroscopic image of the patient.

To do this, move the C-arm to the required angle with the joystick on the operating table and step on the scene acquisition pedal. Electronically align the stent graft with markers from a 3D dynamic-CTA scan with fluoroscopic images using automated image registration, followed by manual refinement, if necessary, in the 3D post-processing workstation. Check and accept the 2D to 3D image fusion and overlay the markers from the dynamic-CTA on the real-time 2D fluoroscopic image.

Qualitative analysis following dynamic-CTA imaging showed a persisting type 1A endoleak in an 82-year-old male patient with chronic obstructive pulmonary disease and hypertension. Further, quantitative time attenuation curve analysis showed a 12.2 second time to peak value for ROI aorta and a 15.4 second time to peak value for ROI endoleak, creating a 3.2 second time to peak value. In a 62-year-old male patient with a medical history of obesity, stroke, renal insufficiency, hypertension, hyperlipidemia, and coronary artery disease, dynamic-CTA revealed sac enlargement with a type two endoleak from bilateral L3 lumbar arteries as inflow vessels.

Time attenuation curve analysis showed a 7.3 second time to peak value for ROI aorta and 24.6 seconds for ROI endoleak at the level of L3 vertebra creating a 17.3 second time to peak value. Appropriate kilovolt selection is crucial for ensuring adequate image quality. Too low kilovolts result in suboptimal images.

Timing of scans is also important because acquisitions launched later result in a timing error and may influence qualitative analysis. To address the underlying clinical question, imaging protocols, whether it's CT or MRI, can be customized really to the individual patient. Often it is not done based on prior imaging findings.

In this particular case of aortic endoleaks, if there is a suspicion for type one endoleaks, it is recommended to have more scans acquired during the earlier phase of the time attenuation curve. For example, if the patient is suspected to have type two endoleak that appears often later in the aneurysm sac, it is recommended to have more spaced scans during the later phase of the time attenuation curve. If no prior imaging studies are available, for example, if it's an index follow-up scan, the scans can be distributed equally along the time attenuation curve from the timing bolus.

Using dynamic-CT angiography, multiple inflow and outflow vessels can be identified. This helps in better understanding of the endoleak and in targeting our treatment. This technique enables a quantitative approach to diagnose aortic endoleaks.

The time to peak can be used to differentiate type one versus type two endoleaks. Using dynamic-CT fluoroscopy image fusion, guidance for endoleak embolization, radiation exposure, and contrast volume consumption can be reduced. Dynamic time-resolved CTA can adequately characterize endoleak types and inflow vessels.

This is especially useful in complex endoleak cases when using the combination of qualitative dye arrival times and quantitative analysis, we can distinguish between endoleak types. In our experience, dynamic-CT imaging has also been shown to provide additional image fusion guidance during endoleak treatment. Such dynamic time-resolved CT imaging can also be helpful in the future of imaging other dynamic disease processes such as aortic dissection, peripheral arterial disease, arteriovenous malformations or intramural hematoma.