Staged Fenestrated/Branched Repair of Postdissecting Thoracoabdominal Aneurysm With Candy-Plug False Lumen Occlusion for Spinal Cord Preconditioning

Introduction

False lumen (FL) thrombosis and aortic remodeling after stent-graft coverage of the proximal entry tear in chronic type B aortic dissection patients is achieved in only 40% of cases, ¹ and persistent perfusion from distal thoracic and abdominal tears is associated with FL growth and poor long-term patient survival.^2,3 This high failure rate has pushed different investigators to evaluate new endovascular therapies aimed at occluding the distal tears of the FL. Postdissecting aneurysms limited to the descending thoracic aorta have been managed with thoracic FL occlusion techniques such as the Candy-Plug (CP) and the knickerbocker techniques.^4–6 Conversely, postdissecting aneurysms extending through the entire thoracoabdominal aorta (TAAA) have been successfully treated with fenestrated or branched stent-grafts.^7–10 This technical note proposes the use of a combined approach with prior intentional occlusion of the thoracic FL with a CP occluder, followed by fenestrated or branched endovascular exclusion of the postdissecting TAAA as a staged procedure in order to precondition the spinal cord.

Technique

The technique is described in 3 consecutive patients (ages 47, 54, and 61 years; 2 men) with postdissecting type II TAAA following type A (n=2) or type B aortic dissection who had been treated by means of a multistage fenestrated/branched endovascular approach with CP interposition. All patients had histories of smoking and hypertension; 2 had chronic obstructive pulmonary disease and one was diabetic. All had previous aortic repairs. Three and 12 years prior, respectively, the 2 patients with residual type A aortic dissection had undergone a frozen elephant technique (FET) zone 0 anastomosis with an arch branch procedure using a modified Jotec Evita Open Plus hybrid stent-graft system (Jotec GmbH, Hechingen, Germany) ¹¹ preceded by a left carotid–subclavian bypass and endovascular occlusion of the left subclavian artery. ¹² The other patient had a standard thoracic stent-graft implanted for complicated dissection 6 years before.

All the steps and procedures were performed in a hybrid room (Artis Pheno; Siemens Healthineers, Erlangen, Germany), preferably under local anesthesia (steps 1, 2, and 4) via percutaneous femoral accesses with the preclose technique (Perclose Proglide; Abbott Vascular, Santa Clara, CA, USA). A percutaneous transaxillary access was used when required. ¹³ Routine perioperative cerebrospinal fluid drainage was not done during staged approaches, but it was planned if a neurological deficit were detected postoperatively. ¹⁴

First Step

All 3 patients received a Zenith Alpha stent-graft (William Cook Europe, Bjaeverskov, Denmark) in the true lumen landed distally 1 to 3 cm above the celiac trunk; the stent-graft was oversized with regard to previous thoracic or FET stent-grafts and deployed with adequate overlapping. In 1 case, the celiac trunk originated from the FL, so a neofenestration in the dissecting lamella was created 3 cm above the celiac trunk origin ¹⁵ to facilitate easy catheterization during the third step (Figure 1).

OPEN IN VIEWER

Figure 1.

(A) Multiplanar reconstruction (MPR) of postdissecting thoracoabdominal aortic aneurysm following surgical replacement of the aortic arch (frozen elephant trunk) with persistent thoracic false lumen (FL) perfusion and aneurysmal dilatation. (B) Volume rendering, (C) axial view, and (D) MPR showing a small tear localized 3 cm above the celiac trunk originating from the FL. (E) Intraoperative angiography of thoracic true lumen (TL) stent-grafting and fenestration of the supraceliac tear in the dissecting lamella using a 14-mm noncompliant balloon. (F) Postoperative volume rendering and (G) axial view after thoracic true lumen stent-grafting with an enlarged neofenestration in the dissecting lamella to ease access during the following steps.

Second Step

Two weeks after the first step, the thoracic FL was occluded with a custom-made CP (William Cook Europe, Bjaeverskov, Denmark) using version II in 1 case and version III in the other two.^5,6,16 The CP was placed into the thoracic FL at the same level as the distal end of the previous true lumen stent-graft. The access to the FL was obtained by catheterizing the distal reentry tear at the level of the common (1 case) or external iliac artery (2 cases). In these 2 latter cases, a standard iliac branch device (Zenith Branch Endovascular Graft-Iliac Bifurcation; William Cook Europe) was implanted in the true lumen to occlude the distal entry tear used to access the FL located in the external iliac artery (Figure 2).

OPEN IN VIEWER

Figure 2.

(A) Multiplanar reconstruction (MPR) of the distal reentry tear at the level of the external iliac artery used to access the false lumen (FL). (B) MPR and (C) volume rendering images of the postdissecting thoracoabdominal aortic aneurysm showing perfusion of the thoracic and abdominal FLs, true lumen (TL) stent-graft from previous frozen elephant trunk (FET) procedure, and extension of the dissecting lamella into the right external iliac artery. (D) Postoperative multi-intensity reconstruction showing implantation of an iliac branch device (IBD) to close the distal reentry tear in the external iliac artery. (E) MPR showing postoperative implantation of the Candy-Plug device, placed at the same level as the TL stent-graft, excluding the thoracic FL. (F) Postoperative volume rendering showing exclusions of the thoracic FL by the Candy-Plug device and the iliac reentry tear by the IBD; only the abdominal FL is perfused.

Third Step

The next step consisted of the deployment of custom-made preloaded fenestrated/branched stent-grafts (William Cook Europe) with bridging of all target vessels. The procedure was performed whenever the graft was shipped to the institution (mean 44 days after the first 2 steps). During this step all patients also received a ZFEN or unibody abdominal aortic bifurcated component (William Cook Europe) as well as a Zenith Spiral-Z AAA iliac leg (William Cook Europe) in the ipsilateral iliac artery, leaving the contralateral limb open and unbridged according to our staging STEAR protocol (ClinicalTrials.gov identifier NCT03342755) for elective thoracoabdominal repair. ¹⁴

The custom-made stent-grafts had 4 fenestrations in 2 cases and 3 fenestrations and a proximal branch in the case in which a neofenestration in the dissecting lamella was performed to access the vessel from the true lumen to the FL. In all cases the preloaded wires were used to target the renal arteries; the superior mesenteric artery (SMA) was catheterized from the contralateral groin and the celiac trunk from the axillary artery access. Bridging devices for the renovisceral vessels were BeGraft (Bentley Innomed, Hechingen, Germany), Advanta V12 (Getinge, Gothenburg, Sweden), or Covera Plus (BD Bard, Covington, GA, USA) stent-grafts.

All the cases were conducted under local anesthesia to have continuous neurological monitoring during the entire procedure (Figure 3). From the contralateral groin via a 20-F introducer, the SMA was catheterized, and the graft was molded with a balloon before bridging the target vessels to disrupt the intimal lamella. First, the SMA was bridged and the contralateral groin closed percutaneously. Second, the abdominal bifurcated component and the ipsilateral iliac limb were deployed after the renal arteries were bridged. The femoral access was downsized to promptly restore blood flow to both the pelvic and limb arteries. Third, the celiac trunk was bridged from an upper extremity access with a percutaneous transaxillary access (2 cases from the right side). In 1 patient the deployment of a bifurcated component occluded the distal entry tear located in the common iliac artery.

OPEN IN VIEWER

Figure 3.

(A) Technical drawing and picture of the preloaded fenestrated device used for case 1 with highlighted catheter pathway (inset). (B) Intraoperative angiography showing catheterization of all target vessels (the renal arteries from the preloaded device, the superior mesenteric artery from the contralateral groin, and the celiac trunk from above). (C) Completion digital subtraction angiography in anteroposterior (left) and lateral-lateral projections showing exclusion of the false lumen (FL) with patency of all target vessels. (D) Postoperative volume rendering and axial scans demonstrating complete FL exclusion after the completion of the 4 steps that consisted of a thoracic stent-graft deployment (TEVAR) followed by the thoracic FL exclusion with a Candy-Plug (CP) and exclusion of the iliac reentry with an iliac branch device (IBD). The procedure was completed with exclusion of the abdominal FL with a fenestrated stent-graft (FEVAR).

Discussion

Despite the complexity of the procedure, fenestrated/branched stent-graft repair of postdissecting TAAA is associated with high technical success and acceptable mortality. SCI remains an important concern during endovascular treatment of extensive TAAAs, reaching up to 15.5% in postdissecting TAAAs in experienced centers despite standard staging protocols.^7–10 The higher SCI incidence might not only be related to the extensive aortic coverage but also to unpredictable FL thrombosis and consequent inconsistent spinal cord preconditioning during staging. ¹⁰ Persistent retrograde FL perfusion can prevent thoracic FL thrombosis, jeopardizing spinal cord preconditioning after proximal entry tear coverage and possibly explaining the higher SCI incidence in fenestrated/branched repair in postdissecting TAAAs despite staging.

The intentional occlusion of the thoracic FL with a CP and subsequent occlusion of distal reentry tears with fenestrated/branched repairs was first reported by Rohlffs et al ¹⁶ in 2 cases. Different CP versions have been used. ¹⁶ Thanks to a self-occluding mechanism (Figure 4), versions II and III do not require occlusion of the central channel with a plug to reduce the risk of inadequate sealing, thus simplifying the procedure. ⁶ The 2 self-occluding CPs versions were designed to cope with regulatory aspects: the CP II is made on the Cook Alpha platform, which is not registered for treating dissections. Furthermore, a custom-made product cannot be replicated many times.

OPEN IN VIEWER

Figure 4.

Technical drawings and pictures of the different custom-made Candy-Plug (CP) versions. CP version I requires subsequent occlusion of the central channel. CP versions II and III have a self-closing central channel mechanism that collapses, occluding the CP lumen. Different designs have the central fabric inside the graft in version II or outside in version III.

Staging the true lumen stent-graft procedure (first step) and the thoracic FL occlusion with a CP (second step) when an iliac branch deployment is also needed is controversial. Even though staging allows separate occlusion of the TL and FL intercostals and reduces the complexity of the procedures, FL thrombosis could occur after step 1. This would be a drawback because thrombosis may hamper full expansion of the CP and interfere with the sealing of the thoracic FL by preventing the CP from adhering properly to the aortic wall. Further investigation of this staging sequence is required.

Aside from spinal cord preconditioning, other potential benefits might be considered. First, the risk of aneurysm rupture is intrinsic to any staging protocol, and oftentimes in postdissecting TAAA, the FL lumen is mostly dilated in the thoracic region so the CP FL occlusion becomes a temporary treatment of the thoracic FL while waiting for the custom-made device. ¹⁸ Second, fenestrated or branched endovascular treatment for postdissecting TAAA is burdened by a significant rate of endoleak (up to 32.4%) due to the presence of unsealed distal reentry tears or patent lumbar/intercostal arteries, leading to a high rate of reinterventions. ¹⁰ By separating the thoracic and abdominal FLs, the propagation of these endoleaks and the pressurization is confined to a limited area of the FL and not to its whole extent. In the future, the potential benefits of CP deployment during fenestrated/branched repair of postdissecting TAAA as a staged procedure should be balanced against of the risk of interval rupture as well as greater material expenses and longer total length of hospital stay.