Retrograde Target Vessel Catheterization as a Salvage Procedure in Fenestrated/Branched Endografting

Study Design and Patient Cohort

A prospectively maintained database was queried to identify all patients treated with F/B-EVAR (available since 2001) who required retrograde target vessel access after failure of antegrade cannulation. Of the total 671 patients who were treated with F/B-EVAR (214 TAAA and 457 JAA/SAA) from November 2003 to November 2014, antegrade catheterization of one or more target vessels was not feasible in 27 patients. In 12, no further action was undertaken, and the target vessel lost. In 4 patients, the fenestration was not stented but was well aligned with the target vessel and patent without signs of endoleak. The remaining 11 (1.6%) patients (mean age 64±7.8 years; 8 men) had a retrograde approach to cannulate the target vessels as a bailout measure after failure of antegrade cannulation; these patients form the study cohort.

Six patients were classified as American Society of Anesthesiologists (ASA) 3 and 5 were ASA 2. Eight patients were treated for TAAA (mean diameter 73.1±14.0 mm): Crawford type II (n=2), type III (n=2), type IV (n=3), and type V (n=1). The other 3 patients had JAAs, all with an aneurysm diameter of 60 mm. Nine patients were treated on an elective basis with custom-made stent-grafts, one patient was treated for a symptomatic TAAA and the last patient for a contained rupture in a TAAA. In both acute cases, an off-the-shelf Zenith t-Branch device (William A Cook Australia, Ltd., Brisbane, Australia) was used. The customized stent-grafts were based on the Zenith platform, fitting fenestrations and/or branches for the visceral vessels according to preoperative computed tomographic angiography (CTA) measurements. A variety of endograft configurations was used to accommodate individual patient anatomy, including stent-grafts with fenestrations only, branches only, and branches and fenestrations. Fenestrations were preferred for right-angle takeoff visceral arteries (more common in the renal arteries) and when the stent-graft body was against the aortic wall. Branches were preferentially planned in larger aortic diameters, when the graft was not against the aortic wall, and when the target vessels had a downward path.

F/B-EVAR Procedure

Procedures were performed either in the operating theater using a mobile C-arm [OEM 9800 (General Electric Medical Systems, Salt Lake City, UT, USA); Arcadis Avantic (Siemens AG, Forchheim, Germany)] or (later) in a hybrid operating room with a fixed C-arm system (Artis Zeego; Siemens AG, Forchheim, Germany). The operation was always done under general anesthesia.

The operative technique for fenestrated endografting has been described in detail previously.^7,8 In brief, surgical access was obtained via bilateral femoral cutdown in cases involving stent-grafts with fenestrations only and with an additional left axillary cutdown for stent-grafts with fenestrations and branches or branches only. Target vessel catheterization was performed under fluoroscopic control in perpendicular orientation to the orifice of each target vessel. For fenestrations, a catheterization attempt was initially carried out with a 0.035-inch Terumo Standard Glidewire (Terumo Medical, Somerset, NJ, USA) after positioning a Cobra 1 catheter at the ostium of the respective vessel through the fenestration. In case of difficult catheterization, other types of catheters were used, including VS2, Omni-flush, and SOS-shaped catheters, as well as a 0.018-inch Terumo Standard Glidewire (Terumo Medical) to facilitate catheterization of stenotic vessels. Additional support during catheterization was applied by advancement of a 7-F ANL Guiding Sheath (William A Cook Australia). For vessels targeted with branches, an indwelling wire was planned for the top branch. A 12-F, 45-cm-long sheath and a coaxial 8-F, 70-cm-long or 7-F, 55-cm-long sheath were both advanced into the stent-graft via the upper access. A Soft-Vu Berenstein Angiographic Catheter (Angiodynamics, Latham, NY, USA) was advanced to the tip of the respective branch. Subsequently, a 0.035-inch Terumo Standard Glidewire was used to cannulate the target vessel followed by advancement of the catheter into the target vessel and replacement of the Terumo Glidewire with an Amplatz super-stiff wire (Boston Scientific, Marlborough, MA, USA). When catheterization proved problematic, attempts were made to cannulate the vessels in an antegrade manner with the use of other catheter shapes as mentioned above.

Retrograde Technique

If antegrade catheterization failed, retrograde access to the target vessel was carried out via laparotomy or a retroperitoneal incision from the lateral edge of the left rectus abdominis muscle to the tip of the 11th rib. Surgical access and retrograde puncture technique area illustrated in Figure 1. The respective vessel was punctured, and a Terumo Glidewire was introduced and advanced retrogradely into the aneurysm sac over a 6-F sheath. This wire was retrieved over the fenestration or branch using an 8-F Indy OTW snaring device (Cook Inc, Bloomington, IN, USA) to obtain a through-and-through wire. Subsequently, the corresponding target vessel was stented with a balloon-expandable or self-expanding covered bridging stent. For fenestrations, balloon-expandable Atrium iCast stent-grafts (Maquet Getinge Group, Hudson, NH, USA) of 22- or 38-mm length were routinely used. For branches, Atrium iCast stent-grafts of 59-mm length were used whenever possible. If a longer covered bridging stent was needed, an 80-mm-long Fluency stent-graft (Bard, Murray Hill, NJ, USA) was used. Any kinked stent-graft was relined with a self-expanding Smart Control stent (Cordis Corporation, Bridgewater, NJ, USA) at the discretion of the operator.

OPEN IN VIEWER

Figure 1.

Intraoperative images of retrograde access to the left renal artery via (A) laparotomy and (B) retroperitoneal incision.

Postoperative Management

Postoperatively, patients were monitored with clinical and laboratory examination, including thoracic and abdominal radiography in standardized anteroposterior and oblique views as reference prior to discharge. CTA controls were performed at 1 month, 1 year, and annually thereafter, depending on each patient’s characteristics. On suspicion of endoleak or branch vessel malperfusion, additional digital subtraction angiography was carried out for further evaluation and possible reintervention.

Data Analysis

Variables are presented as mean ± standard deviation in case of normal distribution and median (range) in case of skewed distribution. Analyzed outcomes included operative mortality, major morbidity, and late events with regard to target vessel stent patency, reintervention, endoleak, and death. Major morbidity was defined as the occurrence of clinically relevant pulmonary or cardiac complications, postoperative glomerular filtration rate deterioration >30%, paraplegia/paraparesis, and local complications requiring surgical treatment. The threshold of statistical significance was p<0.05. SPSS for Windows (version 20.0; IBM Corporation, Somers, NY, USA) was used for statistical analysis.

Technical Issues and Outcome

Target vessels requiring retrograde access were the left renal artery (LRA) in 6 patients, the celiac artery (CA) in 3 patients, the right renal artery (RRA) in 1 patient, and both renal arteries in the last patient. Aneurysm and target vessel characteristics, the technical issues that prompted retrograde catheterization, and the surgical access applied in each case are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Details of the 11 F/B-EVAR Patients Who Underwent Retrograde Target Vessel Catheterization.

Patient	Procedure Year	Aneurysm Type	Stent-Graft Configuration	Target Vessel	Failure	Technical Issue	Access	Outcome
1	2008	JAA	3 Fen	LRA	Small fen	Intraoperative occlusion	RP incision	Success
2	2009	TAAA	4 Br	LRA	↓ Branch	Adverse anatomy	RP incision	Success
3	2009	TAAA	2 Fen, 2 br	CA	↓ Branch	Low branch design	Laparotomy	Success
4	2010	TAAA	4 Fen	RRA, LRA	Small fen ×2	Adverse anatomy	Laparotomy	Success
5	2012	TAAA	2 Br	CA	↓ Branch	Low branch deployment	Laparotomy	Success
6	2012	JAA	1 Fen	LRA	Small fen	Intraoperative occlusion	RP incision	Success
7	2012	TAAA	3 Fen, 1 br	LRA	Small fen	Adverse anatomy	RP incision	Success
8	2012	TAAA	t-Branch	LRA	↓ Branch	Distance from branch	RP incision	Success
9	2013	JAA	3 Fen	LRA	Small fen	Adverse anatomy	RP incision	Success
10	2014	TAAA	t-Branch	CA	↓ Branch	False SMA catheterization	Laparotomy	Success
11	2014	TAAA	3 Fen	RRA	Small fen	False orientation, impossible repositioning	RP incision	Iliac-renal bypass

Abbreviations: CA, celiac artery; F/B-EVAR, fenestrated/branched endovascular aneurysm repair; fen, fenestration; JAA, juxtarenal aortic aneurysm; LRA, left renal artery; RP, retroperitoneal; RRA, right renal artery; SMA, superior mesenteric artery; TAAA, thoracoabdominal aortic aneurysm; ↓, downward facing.

In 4 patients, failure of antegrade catheterization was a result of adverse target vessel anatomy. These vessels were severely stenotic and tortuous, with an anterior takeoff hindering an antegrade approach (Figure 2). In 4 patients, the need for retrograde catheterization was a result of planning or intraoperative technical mistakes. In one of these patients, treated with a device featuring 2 fenestrations and 2 branches, the 2 fenestrations of the device were correctly aligned with the renal arteries but the branch intended for the CA was designed lower than the artery orifice, resulting in inability to catheterize. In the second patient, treated with a device featuring 2 branches, the stent-graft was deployed too low, which resulted in the CA branch being lower than the artery orifice. In the third patient, who had a very tortuous and stenotic RRA, the fenestrated graft was not deployed in the correct orientation; repositioning of the graft was impossible since the fenestrations were completely against the aortic wall. In the fourth case, involving a patient treated with a t-Branch device, the superior mesenteric artery (SMA) was secured through the branch originally intended for the CA. The other branch, intended for the SMA, was subsequently too low for the CA (Figure 3).

OPEN IN VIEWER

Figure 2.

Computed tomographic angiography demonstrating small and angulated renal arteries. Antegrade catheterization of both renal arteries was not possible, prompting laparotomy for retrograde catheterization.

OPEN IN VIEWER

Figure 3.

(A) Intraoperative angiography of a patient whose superior mesenteric artery was secured through the branch originally intended for the celiac artery. (B) As a result the remaining branch is lower (lower arrow) than the orifice of the celiac artery (upper arrow).

Of the remaining 3 patients, the LRA orifice was situated significantly lower than the respective branch in the patient treated emergently with a t-Branch device for a symptomatic TAAA. The distance and angulation of the LRA made antegrade cannulation impossible (Figure 4). In the final 2 cases, retrograde access was applied to treat an intraoperative target vessel occlusion; in both, the LRA was targeted with a fenestration. The vessel was initially successfully catheterized and secured with an iCast stent-graft. In the first case, a stent-graft with a single fenestration was applied. An occlusion that could not be explained was visualized on completion angiography. In the second case, the LRA stent-graft was crushed during balloon flaring of the stent-graft applied for the more cranially situated RRA, leading to target vessel occlusion. In both cases, the occlusion was immediately treated by retroperitoneal incision for retrograde access to preserve the kidney. Estimated occlusion time until successful reopening was 30 to 45 minutes.

OPEN IN VIEWER

Figure 4.

(A–C) Computed tomographic angiography and (D) intraoperative angiography of a patient treated with a t-Branch device for a thoracoabdominal aneurysm with contained rupture. The angulation and distance between the renal arteries (A–C) led to wire instability and inability to advance the angiographic catheter into the left renal artery. (D) After introduction of a through-and-through wire, the catheter could be advanced into the target vessel.

Ten of the 11 target vessels could successfully be cannulated and secured with stent-grafts. In one patient, with a tortuous and stenotic RRA, snaring of the retrograde wire over the fenestration was impossible because the fenestration was completely against the aortic wall in the wrong position. This patient was ultimately treated with an iliac-renal bypass for the RRA. The respective fenestration was left open, and a secondary occlusion of the fenestration planned. Mean total operative time was 396±117 minutes. Median estimated blood loss was 1050 mL (range 300–2000). Median fluoroscopy time was 103 minutes (range 84–130) and median iodinated contrast volume used 298 mL (range 280–420).

Perioperative Mortality and Morbidity

One patient died within 30 days. He underwent elective treatment for a type II TAAA with a custom-made device featuring 4 branches. Retrograde catheterization was required for the LRA due to a very stenotic and tortuous orifice. While the target vessels were patent postoperatively, the patient died of multiple organ failure 8 days after the procedure.

Six patients had major perioperative complications. Of the 3 with renal complications, one with acute intraoperative occlusion of the LRA had a known occlusion of the RRA and developed a retroperitoneal hematoma postoperatively. Renal function deteriorated, requiring temporary dialysis. The second patient with LRA occlusion also suffered temporary renal function deterioration, but he did not need dialysis. In both cases, renal function returned to baseline prior to discharge. The third case involved the patient in whom retrograde cannulation was unsuccessful. This patient suffered an occlusion of the iliac renal bypass 3 days postoperatively and ultimately became dialysis dependent. The three other complications included 2 abdominal wound dehiscences with evisceration, which both required surgical revision and fascial closure, and one case of pneumonia in the patient treated for contained TAAA rupture. Median hospital stay was 20 days (range 7–60) and median intensive care unit stay was 2.5 days (range 0–9).

Follow-up

Mean follow-up was 26 months (range 1–60). The patient with occlusion of an iliac renal bypass refused to continue dialysis and died due to uremia 2 months after the procedure. No other deaths and no target vessel occlusions were noted during follow-up; no endoleaks were detected.

There was one reintervention during follow-up in a patient with progression of an aneurysm of the right iliac artery. This patient was treated at 14 months postoperatively with stent-graft extension using a Zenith Iliac Branch Device (William A Cook Australia).