Technical Solutions for Common Problems in TEVAR:
Thoracic endovascular aortic repair (TEVAR) is rapidly emerging as an important treatment option for several indications, and it would not be unreasonable to predict that endograft treatment may well become the predominant form of therapy for many, if not the majority, of patients. However, several unresolved issues remain, and the need for further improvements and technological refinements will not cease any time soon. Ranking high among these issues are the challenges related to endovascular access and aortic branch management, which constitute the main focus of this review.
Achieving safe and successful endovascular access for introduction and deployment of the stent-graft device is a crucially important and often challenging step during TEVAR, but arterial injury has been, and continues to be, an all-too-common occurrence to this day. A clear understanding of the relevant issues and available technical solutions can go a long way toward preventing such catastrophes.
A preponderance of thoracic aortic pathologies tend to develop adjacent to or within the branched segments. It is therefore not surprising that branch management issues have risen to the top of the entire TEVAR field. Debranching and vessel relocation techniques have added a whole new dimension to the therapy because they can expand or create suitable landing zones proximally and distally, thereby broadening the applicability of endograft technologies to a much larger number of patients.
2008 was the “coming-out” year for thoracic endovascular aortic repair (TEVAR), as 2 additional stent-graft devices (Cook TX2 and Medtronic Talent) received marketing approval in the US. They joined the Gore TAG stent-graft, which was the first to be granted regulatory approval in 2005.1–3 The therapy overall is just beginning to reach a level of maturity, as the 3 major thoracic endograft manufacturers now aggressively support this technology. Many more physicians are likely to be trained in the required knowledge and skills necessary to perform thoracic endograft procedures. No doubt, many more patients will be offered such interventions in the future, and a true “revolution” (of sorts) in the field of thoracic aortic surgery looms on the horizon. TEVAR is rapidly emerging as an important treatment option for several indications,4 and it would not be unreasonable to predict that endograft repair may well become the predominant form of therapy for many, if not the majority, of patients (Table). Indeed, it is poised to eclipse the current open surgery standard of care, which has been dogged by continued high morbidity and significant perioperative mortality. Focal lesions in the mid descending thoracic aorta and perhaps blunt trauma injuries at the isthmus represent some of the most compelling indications for TEVAR; in such cases, the endovascular approach seems headed for “total triumph” and likely will replace surgical treatment in the near future. However, the reader is forewarned of the fact that these predictions may be somewhat premature, as we still lack the kind of scientific evidence that is generally required before a new therapy becomes truly a “game-changer” and replaces a time-honored standard of care.4
The above statements notwithstanding, several unresolved issues remain, and the need for further improvements and technological refinements will not cease any time soon. Ranking high on this list are the challenges related to endovascular access and aortic branch management, which constitute the main focus for this review.
Achieving safe and successful endovascular access for introduction and deployment of the stent-graft device is a crucially important and often challenging step during TEVAR. The reasons are twofold. First, currently available thoracic endografts have a large profile: for instance, for a 40-mm-diameter device, the TAG is 27.6 F (introducer sheath required), the TX2 is a 25.5-F device (sheath not required), and the Talent is 24 F (sheath not required). Second, female patients, with their notoriously small iliac and femoral arteries, make up a very significant segment of the TEVAR patient population (>30%5). Consequently, arterial access injury has been, and continues to be, an all-too-common occurrence, with external iliac artery (EIA) rupture ranking high as a cause of procedure-related mortality to this day. A clear understanding of the relevant issues and available technical solutions can go a long way toward preventing such catastrophes.
♦ (A) A focal aneurysm in the mid portion of the descending thoracic aorta is ideally suited for endovascular treatment. (B) Assessment of arterial access anatomy by CTA.
Iliac Access ConduitA graft conduit anastomosed to the CIA constitutes the most frequently used non-femoral access option during TEVAR. The following basic technical description is based on the author’s >10-year experience with the use of access conduits for aortic stent-graft intervention.23 Retroperitoneal exposure of the CIA is achieved via an oblique incision in the lower quadrant of the abdomen. CTA assessment of CIA anatomy provides the necessary information to choose the better side for conduit-graft attachment: the 3 most important features are vessel size, calcification, and length . The least diseased, larger, and longer vessel should be targeted. The distal abdominal aorta may occasionally be the only reasonable or available target for graft anastomosis. We tend to use a similar incision in such cases, but with a wider-field retroperitoneal exposure. A left-sided incision is preferred for aortic exposure, for the same reasons that most surgeons choose a left-side flank approach when performing retroperitoneal abdominal aortic aneurysm (AAA) repair. A 10-mm-diameter Dacron graft is the conduit of choice because it is easy to use and provides a large enough lumen for introduction of all delivery systems (even the largest size), including passage of additional parallel diagnostic catheters if necessary. The anastomosis is end to side between the graft and the CIA. The conduit is exited through the abdominal wall via a small stab incision placed just above the inguinal ligament to provide a smoother angle of entry for device introduction and delivery). The operator now has (in-hand) a true and direct extension of the CIA. The subsequent steps are not different from those generally followed when accessing a native vessel
♦ (A) Choice of the right or left iliac artery should be based on anatomy and disease, targeting the longest, least-diseased, and largest CIA for conduit graft anastomosis. (B) A 10-mm Dacron graft anastomosed end to side to the CIA (left) exteriorized through a lower-placed stab incision (right). (C) TEVAR sheath and device placed via puncture of the side of the conduit. Note that the 5-F sheath and diagnostic catheter have also been inserted into the same conduit graft. (D) Oversewn conduit stump may appear like a pseudoaneurysm on CTA. (E) The iliac conduit has been retained as a permanent implant, brought down to the groin as an iliofemoral bypass.
- Direct needle puncture of the Dacron graft and introduction of a soft-tipped, steerable, 0.035-inch guidewire (i.e., Bentson, Wholey, or similar) that is advanced under fluoroscopic guidance to the ascending aorta, followed by catheter exchange for the stiff TEVAR wire of choice (i.e., Lunderquist).
- Next, a second (slightly higher) needle puncture is made to insert a 5-F short sheath (over a guidewire) directly into the conduit. The wire is then advanced retrograde (under fluoroscopy) to the aortic arch or ascending aorta and used to support introduction of a long pigtail catheter. The conduit is thus used for both delivery of the endograft system and introduction of the necessary diagnostic catheter, obviating the need for additional vessel punctures. The introducer (or access) sheath is next passed into and through the conduit over the stiff guidewire. A #11 blade is used to enlarge the graft puncture hole only slightly, allowing the introducer sheath itself to enlarge the opening as the device is carefully advanced in retrograde fashion. Such a technique, as opposed to a wide opening into the graft or introduction through the end of the conduit, facilitates the procedure significantly and minimizes blood loss
The iliac conduit is excised at the end, leaving behind only a short stub that is carefully oversewn with a running polypropylene suture. The stub should be of a length that comfortably allows placement of a clamp on the graft itself (not the native vessel or suture line), but not so long as to create a pseudoaneurysm-like image that may lead to confusion and unnecessary concern when visualized on CTA postoperatively
The iliac conduit technique is not without potential pitfalls and complications:
- Retroperitoneal exposure of the CIA can be a very difficult operation on obese individuals and those who have had prior operations in the same region of the abdomen.
- Densely calcified iliac arteries must be avoided.
- Suture-line disruption can and does happen on occasion while passing large rigid devices through relatively narrow anastomotic openings, especially when the CIA is small and/or diseased and thin-walled. We find it helpful to support and actually hold the anastomosis with one’s hand during such maneuvers.
If necessary, the conduit graft can be retained as a permanent iliofemoral bypass when flow through the EIA appears compromised at the end of the procedure, or when the need for a redo or repeat future aortic endograft procedure is anticipated or felt to be likely because it would be easily accessible through a short femoral incision.
Other Transfemoral Access Techniques for Extensively Diseased and/or Small EIAsWhile techniques to gain endovascular access through the iliac arteries (or more proximally) have been used extensively and with acceptable success rates, it is important to recognize they all involve surgical retroperitoneal exposure of the iliac arteries and/or the distal aorta. Related morbidity is unquestionably higher than that of the groin-only procedure.25It makes sense, therefore, to focus on expanding the applicability of the transfemoral access, even to patients with extensive atherosclerotic disease and/or small femoral artery or EIAs. Essentially, two techniques have been described that may be worth considering:
- Queral and Criado26 first described the technique of “retrograde iliofemoral endarterectomy facilitated by balloon angioplasty” in 1995. The underlying concept is that an angioplasty balloon, when inflated, can be used effectively as an instrument to disrupt and “dissect” heavy atheromatous plaques that may be lining the full length of the EIA. The procedure is completed with a combination of (endarterectomy) rings and clamp-mediated “blind” extraction through the open femoral arteriotomy. While the original indication was revascularization for treatment of ischemic limbs, Yano et al.27 used the same principles and tools to create a larger lumen that would allow access for delivery of an endograft system. They also introduced the concept of relining the lumen with a stented graft.
- Yano et al.27 described construction of a homemade “endoluminal conduit” from a Palmaz stent sutured into a 6-mm polytetrafluoroethylene (PTFE) graft and backloaded into a 6-F delivery system. It was used on 5 patients with severe iliac artery disease to facilitate access for stent-graft intervention. More aggressive iliac artery dilatation could then be pursued, within the PTFE conduit, with little fear of uncontained arterial rupture. In truth, this description represented another application of the “stented graft” concept for treatment of aortoiliac occlusive disease championed by Marin et al.28 in 1994. A recent publication by Peterson and Matsumura29 described further refinements to the creation of such “internal endoconduits.” They used an iliac limb component of an AAA stent-graft device to reline the extensively disease EIA. The device was 14 cm in length and tapered from 16 mm proximally to 12 mm distally. Proximal fixation was within the relatively normal distal CIA (covering the hypogastric artery origin); distally, the endograft was brought down into the femoral region. Once in place, and having achieved complete exclusion of the full length of the EIA, the vessel could be aggressively dilated to a diameter of 12 mm or more . “Controlled rupture of the re-lined EIA” may be a more apt name to characterize the procedure. It seems that a 12-mm lumen is necessary for placement of a 24-F outer diameter introducer sheath. A smaller sheath would require less aggressive dilation. The technique has its merits as it can truly expand the capabilities of the transfemoral access. The potential downside of covering the hypogastric origin may be a moot point since this vessel is often severely stenosed or occluded in the majority of patients with extensive iliac artery disease. Furthermore, coverage of the hypogastric artery, as opposed to direct coil embolization of the vessel, tends to be very well tolerated.
♦ Iliac endoconduit technique: (A) The guidewire is advanced retrograde through the lumen of a severely diseased EIA, (B) with subsequent placement of a suitable covered stent or stent-graft (such as the iliac limb of an AAA endograft). (C) Aggressive balloon dilation of the endograft-relined EIA, up to a 12-mm diameter if necessary, results in a large transluminal endoconduit (D).
Antegrade Transcarotid AccessThere are occasional patients who present with situations that may be truly impossible for retrograde endovascular access to the thoracic aorta. Such unusual conditions tend to be the result of various combinations of unsuitable anatomy, significant calcified atherosclerotic disease, and previous endovascular and/or surgical procedures. Use of an antegrade transcarotid access may constitute a viable alternative for some of these.19,31–33It is based on the observation that the proximal common carotid arteries (CCA) tend to be large and essentially free of disease, even in the setting of severe arterial atherosclerosis elsewhere. Both anatomically and geometrically, the right CIA tends to be the better choice, but the final decision (right versus left) should be based on careful assessment of the anatomy of the aortic arch and its branches.The technique begins with surgical exposure of the proximal right CCA through a short longitudinal incision that parallels the anterior border of the sternocleidomastoid muscle at the base of the neck. Retrograde needle puncture of the exposed artery allows passage of an access-type 0.035-inch guidewire. Because the wire will often go in the direction of the ascending aorta, a multipurpose or similar curved catheter can be used to steer it in the direction of aortic flow, down into the descending thoracic aorta.
Transcatheter exchange for a very stiff guidewire will (hopefully) result in a relatively gentle and navigable pathway for easy introduction of the large sheath down to the proximal portion of the descending thoracic aorta . TEVAR can then proceed in standard fashion, but attention must be paid to the design and configuration of the chosen endograft. While the TAG stent-graft can be placed (indistinctly) retrograde or antegrade, other devices may require preliminary re-loading into the sheath “upside-down,” representing clearly an off-label use for such a device. Following completion of endograft placement and balloon molding, the large sheath can be removed, and the CCA is repaired using standard vascular suturing technique. Cross-clamping of the proximal CCA for a few minutes is, of course, required, a maneuver that is generally felt to be well-tolerated by most if not all patients. Potential problems and pitfalls of this technique are:
- The procedure is (or seems) “awkward” because it involves use of long wires and devices that will extend up from the patient’s neck region. Careful attention must be paid to operating room configuration and table setup.
- Displaying and understanding arch branch anatomy with clarity and precision is obviously paramount. We have found both CTA and conventional angiography to be valuable in this setting. Certain configurations are particularly challenging or possibly risky, for instance, when the left CCA originates directly from the innominate artery (bovine anatomy). In such a patient, retrograde introduction through the right CCA could potentially impede flow in both carotid arteries.
- Antegrade transcarotid access is only (potentially) useful for treatment of aortic lesions that are located distal to the aortic arch, ideally, in the mid descending thoracic aorta or lower.
- Lastly, it is only fair to state that this technique is far from “accepted” or “conventional.” Some may argue its safety is unproven. However, preliminary and anecdotal reports, as well as our own limited recent experience with 3 cases, would seem to indicate that it does represent a viable and reasonable option for TEVAR access. However, it should probably be considered only in situations that preclude use of the more standard retrograde access options.
Endograft repair will likely become the preferred treatment option for patients presenting with aneurysms and other pathologies involving the non-branched segment of the thoracic aorta, if it has not already done so. However, challenges and unresolved issues abound because many aortic lesions tend to develop adjacent to or within the branched segments. It is, therefore, not surprising that branch management issues have risen to the top of the entire TEVAR field. Debranching and vessel relocation techniques have added a whole new dimension to the therapy as they can expand or create suitable landing zones (necks) proximally and distally, thereby expanding applicability of endograft technologies to a much larger number of patients. However, the aortic arch and visceral segments could not be more different when the issues of branch vessel accessibility and the invasiveness of the required extra-anatomical bypasses are taken into account.
Arch BranchesThe aortic arch has been acknowledged repeatedly as “the Achilles’ heel” of TEVAR, and the presence of branches is the main reason for such a designation. Mapping the various arch zones (Fig. 5) has proven useful for reporting and documenting the all-important proximal endograft fixation site. It also serves as a platform for discussion of arch branch-related issues. For instance, in our 12-year TEVAR experience with > 400 endograft thoracic implants, approximately three quarters of the cases involved a proximal fixation site within arch zones 1, 2, or 3. While “conquering Zone 0” is not generally considered to be within the realm of current endovascular capabilities,stent-graft placement in the mid and even proximal arch is being reported increasingly at present.36The left subclavian artery (LSA) is the vessel that must be dealt with most frequently. The potential consequences of endograft coverage of the LSA (without revascularization) relate to ischemia of the hindbrain, spinal cord, and ipsilateral arm. Currently, opinions and practice trends seem to be coming back full circle, with a majority of experts expressing the view that simple “overstenting” (exclusion) is probably borderline unacceptable in many patients and that revascularization should be performed more often than not to minimize or eliminate potentially serious complications.37Arm claudication is the most frequent consequence but tends to be self-limited, with a majority of patients experiencing significant improvement or symptom resolution within a few months. The need to perform a reintervention for treatment of persistent symptoms of disabling arm claudication is rare: in our experience of >65 instances of LSA exclusion by endograft, we have encountered only 2 such examples.The indications for LSA revascularization (prior to or at the time of TEVAR) are now quite clear and almost universally agreed upon: prior left internal mammary artery-to-coronary artery bypass operation, occluded or absent right vertebral artery, dominant left vertebral artery, or extensive endograft coverage of the thoracic aorta. Less commonly, anatomical anomalies (such as presence of an aberrant right subclavian artery) may also require preliminary revascularization. While the carotid-subclavian bypass is considered widely as the standard technique for LSA revascularization, we have for many years preferred the carotid-axillary bypass instead because of technical ease and the avoidance of potential lymphatic and nerve complications associated with subclavian artery exposure.More proximal device landing into the mid or proximal arch often implies the need for left carotid artery debranching. We have found the crossover right-to-left carotid-carotid retropharyngeal bypass to be the most satisfactory technique for such purpose.
♦ The “arch map” has proven of value for documenting and reporting proximal fixation sites of TEVAR devices. Complexity of the procedure and its outcome tend to correlate mostly with the proximal implantation zone
Ligation of the left CCA just below the graft anastomosis is an important component of this procedure as well. The operation has proven sound and safe, with excellent long-term durability.While it is true that all these cervical bypasses (or transpositions) work well and serve their purpose, they also imply the need for additional operative procedures, some morbidity, and the potential for delaying the definitive repair of the thoracic aorta, sometimes by several weeks. These ideas, together with personal experience in 2 cases (in early 2002) where unintentional endograft coverage of the left CCA occurred (Fig. 7), became the foundation for a different perspective and practice vis-à-vis arch branch management during TEVAR. It has evolved into a strategy that focuses on branch vessel preservation instead of debranching. The technique, conceived initially as a troubleshooting maneuver, consists of stenting the vessel origin to re-establish or preserve normal antegrade flow by creating an antegrade parallel channel outside of the aortic endograft.39The expanded stent “breaks” the endograft seal to the aortic wall in that focal area adjacent to the vessel ostium, thereby re-opening an antegrade channel for normal branch flow. Access for the procedure is via percutaneous retrograde catheterization of the left CCA or LSA; a similar approach (right-sided) can be used for stenting of the innominate artery.Typically, a micropuncture technique is used for placement of a fine guidewire that is advanced into the ascending aorta. The wire can be placed “pre-emptively” in cases where encroachment of a given arch branch is possible. If stenting becomes necessary, or when it is part of the operative plan, the microaccess wire is then exchanged for a more standard 0.035-inch guidewire that supports placement of a short 6-F sheath. A 6-mm-diameter angioplasty balloon is advanced retrograde (over the wire), inflated across the vessel origin, and used as a sizing tool to enable selection of an appropriate balloon-expandable stent. In most cases, an 8-mm-diameter by 29- or 30-mm-long device has been used. A self-expanding nitinol stent may need to be considered when dealing with a very large vessel (>10-mm diameter). In either case, the proximal end of the stent (residing inside the lumen of the aortic arch) must be positioned flush with the proximal border of the endograft fabric (or more proximally) to ensure formation of a long enough antegrade channel for unimpeded normal flow into the branch vessel . Our experience with 20 such “chimney grafts” in the arch (8 left CCA and 12 LSA associated with Talent and TAG stent-grafts) has been quite encouraging and satisfactory; over a mean 24-month follow-up (range 1–60), all the stents have remained patent, with only 2 showing in-stent stenosis. While creating a proximal type I endoleak is the most frequently voiced concern related to this technique, we have not encountered any in our experience. In truth, there have been essentially no complications or problems whatsoever, and this is in line with the experience recently reported by Ohrlander et al. However, it is important to note that several unanswered questions surrounding chimney grafts remain, mainly, the potential for integrity issues or device damage as a result of interaction between the aortic endograft and the adjacent branch stent over the long haul. This is the main reason why such a strategy cannot yet be recommended for wide adoption. Proximal endoleaks can probably be avoided as long as there is still (in the aorta) a circumferential neck area for fixation and seal distal to the stent. Expansion of these concepts for the creation of longer conduits through the lumen of the aortic arch is beginning to be explored . Only a few such procedures have been recently performed using a self-expanding covered stent (Viabahn) that has been re-lined with bare metal nitinol stents to enhance crush resistance.The above-described techniques for partial arch debranching (or vessel preservation) are generally felt to be reasonable if they lead to the creation of a minimum 2-cm-long proximal neck distal to the innominate artery origin (or beyond) to optimize or enable endograft intervention. However, for those patients with lesions that involve most or all of the aortic arch, only total arch debranching will do if endovascular repair is to be performed. This is perhaps the only area in the entire debranching scenario where only little if any disagreement exists; most surgeons concur that an ascending aorta-based bypass to the innominate and left CCA (or all 3 branches) performed through a median sternotomy approach should be the prescribed strategy in most cases. The operation involves only side clamping of the aorta and tends to be well tolerated.Finally, for the occasional patient who cannot have or withstand a median sternotomy/ascending aorta bypass and, at the same time, has good aortoiliac arterial inflow, there is yet one additional extra-anatomical and completely extra-thoracic option for total debranching. It uses the femoral artery or EIA to construct a retrograde femoroaxillary bypass, with simultaneous axillary (or subclavian) artery to carotid bypass, a crossover graft to the opposite-side carotid artery, plus or minus bypass to the contralateral subclavian artery. Proximal ligation of all arch branches (below the graft anastomoses) is obviously necessary to prevent backflow endoleak
♦ (A) A patient with large ductus arch aneurysms that involved the subclavian artery; the left CCA origin is located a short distance proximally. (B) Percutaneous puncture of the left CCA, with placement of a 0.018-inch guidewire (arrow) into the ascending aorta. Note also the diagnostic catheter that was advanced from a left brachial artery sheath. (C) Once deployed, the thoracic endograft covered the carotid artery almost completely. (D) After placement of the 6-F sheath into the left CCA, a balloon-expandable stent was deployed across the origin of the vessel, with the proximal end of the stent flush with the proximal border of the aortic endograft fabric to create an antegrade channel for normal arch branch flow
♦ Potential technical options for placement of longer chimney grafts using covered-stent devices.