Passing a Mechanical Aortic Valve With a Short Tip Dilator to Facilitate Aortic Arch Endovascular Branched Repair

Introduction

Open surgical repair of aortic arch pathologies is the benchmark surgical method to which all new techniques should be compared. Contemporary sophisticated surgical techniques, improvements in perioperative management, and increasing surgeon experience contribute to good postoperative results.^1,2 However, open surgery becomes more challenging to perform in high-risk patients and especially in those that have already undergone previous thoracic aortic surgery. This is an especially common occurrence among certain patient groups, such as connective tissue disorder patients. ³

Endovascular repair of the aortic arch using fenestrated ⁴ or branched ⁵ stent-grafts is an attractive alternative option for these patient groups showing good mortality and morbidity outcomes. This technique is particularly useful in patients who have previously undergone open thoracic aortic surgery and re-do sternotomy is intended to be avoided. Previously performed aortic valve replacements using a mechanical prosthesis may turn a subsequent endovascular repair into a challenge.^6,7

In this report, we present the endovascular repair of an ascending aortic anastomotic aneurysm with a branched endograft in a patient with a mechanical aortic valve and describe a technique to pass the valve without destroying it.

Case Presentation and Technique Description

A 55-year-old female patient with genetically verified Marfan syndrome presented with a 5-cm anastomotic aneurysm of the proximal aortic arch after previous open aortic replacement (Figure 1). The patient underwent mechanical aortic valve replacement in 1991, supracommissural ascending aorta and proximal arch replacement due to type-A aortic dissection in 2007, and open surgical thoraco-abdominal aortic replacement in 2013.

OPEN IN VIEWER

Figure 1.

Preoperative computed tomography angiography with (a) axial, (b) coronal, and (c) sagittal reconstructions.

After an interdisciplinary consensus, the patient was deemed at high risk for another open surgical treatment. Thus, a 3-stage hybrid repair with bilateral cervical de-branching and endovascular repair with a double-branched stent-graft was planned. The main challenge in this case would be to pass the mechanical prosthesis with the wire and the tip of the sheath dilator without damaging the leaflet and without causing severe regurgitation. A typical mechanical aortic valve is depicted in Figure 2. Our goal was to pass one leaflet through the lateral side (Figure 3a), which would allow the second leaflet to continue working, and as such only mild valve regurgitation during the deployment of the graft should be noticed. Valve functionality should then be reverted to normal after retrieval of the device. It was imperative to avoid cannulation at the center of the prosthesis (Figure 3b) between the 2 leaflets as this may lead to destruction of the leaflets. The behavior of the leaflets during such a procedure was tested on a benchtop model and is shown in Video 1 (Supplemental Material, available online). The patient was informed about the risks of the procedures, and especially about the potential risk of damaging the valve and gave written consent for all procedures.

OPEN IN VIEWER

Figure 2.

Depiction of a typical mechanical aortic valve. Blue arrows indicate the desired cannulation spaces on the side of the leaflets.

OPEN IN VIEWER

Figure 3.

Cannulation of the mechanical aortic valve with an 18-F dilator through the side of the leaflet (a) and through the center between the 2 leaflets (b).

In the first step, a carotid-subclavian bypass on the left side was performed using an 8-mm Dacron graft. Due to extensive tortuosity and dissection of the innominate artery that extended into the right common carotid artery (RCCA), we planned the distal landing of the first branch in the RCCA with coverage of the right subclavian artery. In the second stage 3 weeks later, a carotid-subclavian bypass on the right side was performed with concomitant interposition of the RCCA to achieve a stable landing zone for the branch extension.

In the third procedure, a double-branched endovascular stent-graft with branches for the RCCA and the left common carotid artery (LCCA) was implanted. The stent-graft was custom-made (COOK Medical, Bloomington, IN, USA) for the patient with a short 35-mm introducer tip to facilitate proximal landing as close as possible to the coronary arteries (Figure 4). Access for this procedure was established first through a cutdown of both common carotid arteries and the right common femoral artery, as well as puncture of the left brachial and left common femoral artery. Through the right femoral access, a 0.035″ hydrophilic wire (Terumo Medical, Tokyo, Japan) was used to pass the mechanical valve. As the valve could not be adequately visualized in the fluoroscopy images, the wire was advanced with the support of a pigtail catheter (Cordis, Santa Clara, CA, USA) along the outer curvature of the aortic arch and through the lateral side of the leaflet, avoiding the central portion of the valve. Correct positioning of the wire was also confirmed by transesophageal echocardiography, as well as fluoroscopy in multiple projections. The guiding wire was then exchanged for a stiff Lunderquist wire (COOK Medical), which was placed in the left ventricle (Figure 5a). The mechanical valve was then passed with the tip of the dilator on the lateral side of the prosthesis (Figure 5b) with only mild aortic regurgitation, as one leaflet continued to work. This allowed the deployment of the stent-graft directly distally of the coronary arteries (2.5 cm distally from the aortic valve).

OPEN IN VIEWER

Figure 4.

Intraoperative photo of the stent-graft tip.

OPEN IN VIEWER

Figure 5.

Intraoperative angiogram of the aortic arch, with the arrows showing the Lunderquist wire (a) and the dilator of the stent-graft (b) passing though the side of the leaflet into the left ventricle.

The 2 branches were cannulated through the carotid arteries on each side. The RCCA branch was bridged with 2 “thoracic extension” stent-grafts (13×124 and 13×73 mm; COOK Medical) that were relined with a 10×38 mm Advanta V12 (Getinge AB, Gothenburg, Sweden) and a Viabahn 13×50 mm (W. L. Gore & Associates, Inc, Flagstaff, AZ, USA). The bridging stent-graft used for the LCCA branch was a COOK “Thoracic extension,” 13×90 mm. The left subclavian artery was occluded with a vascular plug (Amplatzer Vascular Plug II, St. Jude Medical, St. Paul, MN, USA) to avoid retrograde endoleaks.

The final angiogram and postoperative computed tomography (CT) angiography showed no endoleaks and patent coronary and supra-aortic vessels (Figure 6). Echocardiography showed normal function of the aortic valve after the procedure. No complications arose in the postoperative period and the patient was discharged to a rehabilitation facility on the sixth postoperative day.

OPEN IN VIEWER

Figure 6.

Final intraoperative angiogram (a) and postoperative computed tomography angiography (b).

In 6 months the patient returned for follow-up and the CT scan showed exclusion of the aneurysm with regression of the sac. The aortic valve function was not compromised.

Discussion

Endovascular repair of aortic arch pathologies is an alternative solution for high-risk patients, such as the presented case, showing promising results regarding mortality and morbidity. ⁵ However, if concomitant valve pathology was replaced by a mechanical prosthesis formerly, endovascular arch repair is formally contraindicated according to the recommendations of the stent-graft manufacturers. Spear et al published a report in 2014 on the endovascular aortic arch repair in a patient with a mechanical aortic valve; however, a sufficient proximal landing was achieved without passing the prosthesis with the graft dilator. They introduced a bullet-nose dilator as an alternative for reaching the mechanical valve. ⁸ A minimum distance of at least 3 cm from the valve to the proximal landing zone is still mandatory in such cases. This design of the delivery system is not available by the company anymore. Moreover, since we did not intend to advance the tip of the sheath dilator in its whole through the valve, we opted for a spear-shaped tip that was more suitable in our case.

In the presented case, the mechanical prosthesis had to be passed with the graft dilator during endovascular repair to achieve a sufficient proximal seal. We decided to proceed with an endovascular arch repair and present herein the feasibility of our intended approach, which, to the best of our knowledge, is the first description thereof in the literature.

The stent-graft we used was custom-made based on the preoperative CT angiography of the patient and had a short 35-mm tip, so that the advancement of the full sheath through the valve would not be necessary. As the device was custom-made for the patient, no official IFU exist for it. It is difficult to precisely calculate the maximum width of the dilator that passed the valve, but we estimate that approximately two thirds of the dilator tip were inserted, so 18 to 20 Fr was accommodated through the side of the leaflet. The most critical step of the endovascular procedure was the passage through the mechanical valve at the correct site, namely, the lateral side of one of the leaflets. Since most of the mechanical valves are not clearly visible in fluoroscopy, passage of the valve through visualization of the target structure is challenging.

The key to cannulating the valve at the intended side of the leaflets is understanding the 3-dimensional (3D) spatial relations of the aortic root in the specific patient, knowledge of the preoperative CT scan including 3D reconstructions, usage of multiple fluoroscopy projections, and intraoperative echocardiography.

Using a flexible hydrophilic wire for the cannulation, which is fairly atraumatic, provides the advantage of having more than one attempt at passing the valve in the correct place. Once the wire is across the valve, visualization with plain fluoroscopy and at a certain projection can, with a high degree of certainty, confirm the position of the wire based on the knowledge of the orientation of the valve in the preoperative CT scan and 3D reconstruction. Echocardiography is then used to ensure correct wire placement. At the latest after introduction of the tip of the graft, it is possible to identify on echocardiography how the valve leaflets behave and where the wire has passed the valve.

This technique should be therefore only attempted by experienced teams of endovascular and cardiac surgeons since errors during valve passage may cause prosthetic leaflet destruction followed by severe regurgitation. The detrimental results of advancing a sheath, if the passage occurred in the mid-portion of the mechanical valve, should be made clear and communicated with the patients beforehand.

Conclusion

Passing a mechanical aortic valve prosthesis at the proper position is feasible and allows adequate endovascular treatment in complex arch anatomy. However, caution should be taken during positioning of the endovascular graft as the tip may potentially damage the valve prosthesis.