Image fusion performed with noncontrast computed tomography scans during endovascular aneurysm repair.

We report two endovascular aneurysm repair procedures achieved under image fusion guidance accomplished with noncontrast injected preoperative computed tomography scans. Such use of this advanced imaging application reduces contrast media injection volume (respectively, 27 and 24 mL throughout the patients' hospital course). No changes in creatinine clearance occurred after the procedures. Contrast-enhanced ultrasound imaging confirmed technical success in both cases.

Renal insufficiency after endovascular aneurysm repair (EVAR) is associated with increased morbidity. Iodinated-contrast media injection, required during the implantation procedure and preoperative and postoperative computed tomography (CT), is a major source of renal impairment after EVAR. Renal impairment is observed in up to 24% of patients after EVAR, especially in those with pre-existing renal insufficiency.2, 3

We report two EVAR cases in patients with chronic renal insufficiency performed under fluoroscopic guidance with three-dimensional (3D) road mapping fusion. In both cases, the 3D overlay was constructed from a noncontrast preoperative CT scan, performed for diagnosis and sizing purposes, to minimize the total volume of contrast media injected to the patients throughout their hospital course. Consent to publish was obtained from both patients.

Case report

Patient 1 was a 77-year-old man with a 55-mm-wide abdominal aortic aneurysm (AAA). Medical history reported a right lobectomy for a T2 N0 M0 lung adenocarcinoma 6 months earlier and a severe renal insufficiency, noted as stage 3 chronic kidney disease (CKD), with an estimated glomerular filtration rate of 35 mL/min/1.73 m2 calculated with the Modification of Diet in Renal Disease method. Patient 2 was a 79-year-old man with a 54-mm AAA and a 35-mm right common iliac artery aneurysm. His medical history included two major abdominal surgeries—a subtotal gastrectomy and a radical cystoprostatectomy with ureterostomy—and coronary occlusive disease. In addition, he had a stage 4 CKD (estimated glomerular filtration rate of 21 mL/min/1.73 m2).

Rapid growth (>5 mm at 6-month interval) was observed in both patients, at the aortic level in patient 1 and at the iliac level in patient 2. We generally perform contrast-enhanced CT for patients with AAA undergoing preoperative workup, but preoperative CTs were performed in both patients without contrast media injection. The CTs were loaded in our Advantage Workstation 4.6, volume share 5 (GE Healthcare, Waukesha, Wisc) and reconstructed in multiplanar reformatted (MPR) views. Adjusting contrast and brightness in various MPR views provides an accurate assessment of the landing zones.

A 3D volume-rendering model of the bone structures was automatically generated. Then, a 3D volume-rendering model of the aorta and its main branches was manually reconstructed (Fig 1, A). Schematically, the arterial wall was identified on several consecutive MPR slices (each at least 5 mm) in axial, coronal, and sagittal views, from above the renal arteries to the iliac bifurcations and delimited with contouring software (Fig 1, B and C). Data inside this segmented area were extracted from the complete CT acquisition and reconstructed separately in 3D. Voxels intensity was adjusted to improve visibility (Fig 1, D). This stage required a trained operator and lasted ∼5 minutes.

Fig 1

A, Three-dimensional (3D) volume rendering of the bone structures automatically generated by the workstation. B and C, The arterial wall is identified on several consecutive slices and delimited with contouring software. D, Data inside the segmented area are extracted from the complete computed tomography (CT) image and reconstructed separately in 3D. Voxels intensity is adjusted to improve visibility.

At the beginning of the procedures, two fluoroscopic orthogonal shots (anteroposterior and lateral) were used to perform a 3D-over-2D registration of the bone subvolume from the noncontrast injected CT scan on the X-ray bone structures. We then switched from the bone 3D model to the vascular model. This second step was performed by a scrubbed physician from tableside and required ∼2 more minutes. After rigid guidewires insertion, fine adjustment was achieved with an injection of 7 mL isosmolar iodixanol contrast media (Visipaque; GE Healthcare, Princeton, NJ) to accurately locate the level of the renal ostia.

Navigation and stent graft deployment were accomplished with a minimum quantity of contrast (27 mL and 24 mL, respectively). In addition to the 7 mL contrast media to check the accuracy of the 3D-over-2D registration, 10 mL contrast media was injected twice to locate the internal iliac arteries origin in patient 1, and 10 mL to locate the left internal iliac artery origin and 7 mL to perform a right internal iliac artery embolization in patient 2 (Fig 2). The procedure time was 75 minutes in patient 1 and 60 minutes in patient 2, with 7 minutes and 16 minutes of fluoroscopy time, respectively. The radiation dose measured by the dose area product was 1648 cGy.cm2 and 1010 cGy.cm2, respectively.

Fig 2

Right internal iliac artery embolization performed with image fusion guidance.

Postprocedural control was performed with a contrast-free cone-beam CT scan (CBCT; Fig 3) and a contrast-enhanced 3D ultrasound assessment before discharge. No type I or type III endoleaks were observed, and all renal arteries were patent. Occlusion of the origin of the right internal iliac artery after embolization was confirmed. No changes in renal function were noted throughout the hospital course.

Fig 3

At the end of the procedure, a contrast-free cone-beam computed tomography (CBCT) scan is performed.

Discussion

Renal impairment is a relative contraindication for EVAR because it is associated with increased morbidity. Renal dysfunction is frequently observed after EVAR and has a multifactorial origin, including embolization from guidewire and catheter manipulation and injection of contrast media. In case of pre-existing renal impairment, contrast media injection during EVAR must be restricted to a minimum to minimize the risk of renal function deterioration.

Modern hybrid operative rooms are equipped with advanced imaging applications such as image fusion. This technology allows superimposition of a 3D volume generated from the preoperative CT angiography on the live operative fluoroscopy.7, 8, 9 This vascular overlay is a helpful tool to assist navigation, catheterization, and graft deployment, particularly in complex cases. In 2011, Kobeiter et al reported the first successful endovascular repair of a thoracic aneurysm with zero-contrast injection during the procedure under fusion guidance using CT angiography.

Since then, image fusion has been widely accepted as having the potential to decrease iodinated-contrast media use11, 12 and, thus, nephrotoxicity. Image fusion is traditionally performed from a preoperative contrast-enhanced CT scan or CBCT. We report the first use of fusion based on a preoperative CT scan without contrast. This technique is identical to the previously described fusion technique apart from the fact that we constructed a 3D aortic volume by manual segmentation.

To generate a 3D aortic volume, we had to develop a specific workflow on the 3D workstation. In patients with CKD, the use of image fusion generated from a CT scan without contrast enables EVAR repair with minimum contrast media injection. Postoperative control was also achieved without iodinated contrast media use because it was performed with a noncontrast CBCT associated with a contrast-enhanced 3D ultrasound examination. In patients with a major risk of renal impairment after EVAR, iodinated contrast media injection was restricted to the strict minimum to prevent worsening of the renal disease, new onset of dialysis, and to lower postoperative morbidity. To minimize the risk of contrast-induced nephropathy in patients with CKD, preoperative magnetic resonance angiography-gadolinium registered to an intraoperative rotational system is also an option to perform image fusion. Only one case report has described this technique performed under an experimental protocol. The identification of small branch vessels was not validated.

The use of carbon dioxide as a contrast agent in patients with CKD or severe allergy to iodinated contrast media has also been suggested. However, the high pressure required to inject carbon dioxide can dislodge mural thrombus and create embolization in the visceral arteries and the lower limbs. Excessive volume injection can also place strain on the right side of the heart.

Finally, precatheterization of renal arteries before stent graft deployment is also an option. This later technique can be challenging and associated with complications such as dissections and embolization.

Conclusions

Image fusion improves visualization during image-guided procedures and allows a low volume of contrast injection. In patients with pre-existing renal impairment, image fusion can be achieved from a noncontrast injected preoperative CT scan, minimizing the risk of nephrotoxicity without compromising the accuracy of the EVAR.