The chimney technique (Ch-EVAR) allows off-the-shelf endovascular repair of complex aortic pathologies involving the visceral arteries, such as aneurysms involving the juxtarenal ( Fig. 16.1 ), pararenal, and thoracoabdominal aorta and the aortic arch ( Fig. 16.2 ). First described by Greenberg et al. , Ch-EVAR was reported as a bailout procedure to maintain renal artery patency in case of unintended coverage during standard endovascular aneurysm repair. Bare or covered stents are deployed into visceral branches parallel to the main aortic endograft in order to achieve a cranial extension of the proximal sealing zone with preservation of renal or splanchnic arteries. A periscope graft is a distally oriented Ch-EVAR that allows retrograde flow into an aortic side branch. The sandwich technique first deploys a tubular endograft to create an artificial proximal neck, which is intended to accommodate the proximal end of the Ch-EVAR in its lumen. A second aortic endograft is then overlapped to the first in order to repair the underlying diseased aorta. With this technique, Ch-EVAR remains eventually sandwiched between the two aortic endografts. It was initially described to preserve the hypogastric artery and later for visceral arteries in thoracoabdominal aortic aneurysm.
Urgent aortic cases unfit for open repair and fenestrated-branched endografting because of manufacturing delays can be managed by these techniques, if an adequate sealing zone is absent. In some centers, Ch-EVAR is considered a possible first-line treatment, even in elective cases. The available literature on chimney/periscope/sandwich techniques is mainly about case reports and single-center series with a limited number of patients and little follow-up data. Furthermore, patients have been treated with a variety of off-the-shelf devices and different follow-up protocols, leading to a difficult overall analysis of the procedure. The largest worldwide Ch-EVAR experience was published in 2015 by Donas et al. with data on 898 snorkel stents placed in 517 patients, resulting in a technical success rate of 97.1%, elective 30-day mortality of 3.6%, and primary branch vessel patency of 94% at a mean follow-up of 17.1 months.
Choice of Device
A variety of combinations of different materials have been used to achieve satisfactory results with this technique. Mestres et al. have shown that results in terms of apposition, stent compression, and in-folding may vary between different combinations. Endografts with a low radial force behave better in terms of apposition and stent compression but are at higher risk of in-folding. As a matter of fact, the risk of endograft in-folding is maximum if a 40% oversize is applied to a low-radial force endograft, such as the Excluder (Gore, W. L. Gore, Flagstaff, Arizona) endograft. On the other hand, this kind of endograft allows lower stent compression, and this is particularly evident when parallel stent grafts with low radial force (Viabahn, W. L. Gore, Flagstaff, Arizona) are used. By using a greater radial force main endograft, such as the Endurant (Medtronic, Santa Rosa, California) together with the Viabhan device, a significant stent compression is more likely to occur. The best combination in terms of endograft and stent apposition is achieved with an oversize of 30% of the Excluder endograft, with the Advanta V12 (Atrium/Maquet, Hudson, New Hampshire) parallel stent graft, which are characterized by a greater radial force.
Recently, the CE mark for Ch-EVAR procedures has been assigned to the Endurant device, following publication of the PROTAGORAS study. In that study, 128 patients were treated with a combination of Endurant device and Advanta or Viabahn parallel stent grafts, with a technical success of 100% and only 1.6% Type I endoleak needing a second procedure. At a mean follow-up time of 24.6±17.4 months, a 93.1% freedom from chimney graft-related reinterventions was achieved , indicating that this technique may be effective in the midterm, even if the long-term outcome is yet to be assessed.
Technical Failure: Target Visceral Vessels Loss/Lesions/Early Occlusions
Inability to cannulate and stenting target visceral vessels (TVVs) cause an intraoperative TVV loss. As a result of manipulations with guidewires, catheters, balloons, and stent/stent graft, TVVs lesions such as dissection, rupture, or thrombosis are possible. The use of stiff guidewires with a short soft tip (i.e., Rosen or Amplatz guidewire: Cook, Bloomington, Indiana) is advantageous once access has been obtained in order to gain sufficient stability for insertion of the stent graft in the visceral arteries, without causing excessive endothelial trauma to their distal segment. Ruptures or dissections of distal renal artery branches with subsequent hemorrhage or parenchymal infarct may occur, as well as intestinal bleeding or ischemia. The consequences of these complications are variable but clearly can be fatal. Undesired coverage of aortic side branches can also occur in case of imprecise planning or intraoperative technical problems. Accidental coverage of target visceral vessels by stent graft can occur as a result of planning or intraoperative errors ( Fig. 16.3 ).
A recent literature review reports a risk of 1% TVV-loss in Ch-EVAR for abdominal/thoracoabdominal aneurysms and arch pathologies, leading to an overall early TVV-patency of Ch-EVAR of 97%–99%. Acute TVV occlusion can be either symptomatic or asymptomatic; however, when occurring in the superior mesenteric artery, it can easily cause fatal massive bowel ischemia. Alternatively, renal artery occlusion may cause worsening of the renal function and necessitate permanent dialysis when occurring bilaterally or in patients with single kidneys, or it may remain completely asymptomatic. Extraanatomic bypass has been extensively reported in the treatment of target visceral failures.
Gutters/Proximal Type I Endoleak
A critical issue for Ch-EVAR is the presence of gutters, defined as channels between the visceral and the main aortic endografts, causing Type I–III endoleak and failure of excluding the aneurysm sac. Gutters are described as a lack of apposition between the main body of the endograft, the chimney stent graft, and the aortic wall, and can be classified according to their location. Type A gutters originate at the proximal fabric of the endograft, Type B are determined by apposition failure between the chimney stent graft and the branch vessel and Type C are those starting below the fabric of the endograft.
Independently of their origin, postprocedural Type I–III endoleaks are frequently reported and can be considered the Achilles’ heel of all parallel graft strategies. Proximal Type I endoleaks occur in approximately 13% of cases (range 0%–33%) of juxtarenal or pararenal and TAAA and 11% (range 0%–44%) of thoracic arch treated with Ch-EVAR. They can be stratified into high-flow endoleaks, which appear immediately after contrast injection, and low-flow endoleaks, which appear after some delay ( Fig. 16.4 ). The latter may be compatible with hemorrhage control even in patients with ruptured aneurysms and usually disappear during early follow-up when normal coagulative parameters are restored. Visceral Type I endoleaks have been observed in as many as 70% of cases, with subsequent spontaneous sealing. High flow Type I–III endoleaks are potentially fatal in cases of aortic rupture, as demonstrated by cases of AAA-related mortality after Ch-EVAR repair. Up to 30% of early postoperative imaging studies show the presence of gutter-related Type Ia endoleaks , with spontaneous resolution in 44.3%, 65.2%, and 88.4% of patients after 6, 12, and 18 months respectively, with secondary intervention required in only 3% of cases. At a mean radiologic follow-up of 20.9 months, there were no differences in mean aneurysm sac size change between cases, with or without early Type Ia gutter endoleak. The low number of gutter-related reinterventions required and the low rate of endoleaks associated with aneurysm sac growth lead to speculation that the natural history of Type I endoleak in Ch-EVAR may be more benign than initially expected.