Chapter 40 Endovascular Therapy for Abdominal Aortic Aneurysms
Surgical treatment of abdominal aortic aneurysms (AAAs) dates back centuries. Some of the initial approaches involved techniques similar in some fashion to modern endovascular techniques. In 1684, Moore reported on the use of large quantities of wire placed intraluminally into the aneurysm sac to induce thrombosis of an AAA.1 Later, electric currents were passed through the wire to further promote thrombosis. A self-expanding endoluminally placed umbrella device was reported by Colt in the early 1900s to treat AAA.2 In the mid-1900s, the use of endoluminally placed wire with the passage of electricity through it was revived and remained the procedure of choice until conventional operative therapy of AAA was introduced.3 Operative repair of AAA evolved during the second half of the 20th century. Early techniques ranged from simple aortic ligation to aortic wrapping with cellophane.4,5 Neither was successful. In 1951, the first replacement of an aortic aneurysm with an aortic homograft was described by Dubost et al.6 Homografts, however, became aneurysmal over time, and the procedure evolved to the use of prosthetic material to reconstruct the aorta.7,8 This technique was later modified by Creech, who reported on endoaneurysmorrhaphy with intraluminal graft placement, leaving the aneurysm sac in situ9; this has become the mainstay of treatment.
Although excellent results have been obtained with conventional aneurysm repair, it remains a complex, challenging operation that initiates great physiological stress for patients. The pursuit of a less invasive approach to AAA repair has subsequently evolved. Parodi et al.10 reported the first use of endovascular AAA repair (EVAR). This approach allowed for intraluminal exclusion of an aneurysm with placement, through the femoral arteries, of an endograft. The hope was that this would decrease the morbidity and mortality of aneurysm repair and allow repairs to be performed in patients with significant comorbidities. The original endograft was constructed of a Dacron tube sutured to a Palmaz stent. Several generations of endografts have since been developed, tested, and put into general clinical use. With the evolution of aortic endografting, our knowledge about the pathophysiology of AAA has changed. Our understanding of the complexities of this mode of treatment is only just being realized and examined. This chapter reviews what is currently understood about endograft repairs of abdominal aortic aneurysms.
Indications
The indications for endovascular repair of an AAA remain the same as conventional repair with regard to the size of the aneurysm and its rate of growth. The classic teaching is that rupture rates for aneurysms depend on the size of the aneurysm. Rupture rates of 5% to 7% per year are estimated for aneurysms between 5 and 7 cm in diameter, and a greater than 20% rupture rate per year is estimated for larger aneurysms.11 Compared with observation, surgical treatment for patients with these larger aneurysms significantly improves mortality.12 Although it is known that small aneurysms do have the potential to rupture, the U.K. Small Aneurysm Trial, which randomized patients with AAAs between 4.5 and 5.5 cm to either surgery or observation, suggested that early repair did not improve survival.13 The operative arm of this study, however, had a mortality rate of 5.8%, which is high compared with other large series of elective open AAA repair.14 Perhaps with lower mortality rates in the operative arm, the conclusions of the study would have been reversed.15
Endovascular AAA repair is a less invasive technique than open surgery and offers several potential benefits over conventional AAA repair. It requires small femoral incisions instead of a large abdominal incision, which may decrease the incidence of postoperative pulmonary complications. Avoidance of extensive retroperitoneal dissection decreases the risk for perioperative bleeding. The period of aortic occlusion is minimal and accounts for the lower incidence of intraoperative hemodynamic and metabolic stress compared with patients undergoing open surgery.16 Given these differences, endovascular aneurysm repair may be reasonable in patients who are “unfit” for conventional AAA surgery.17 Proving its durability as a replacement for conventional surgery in relatively healthy patients is the aim of many clinical trials, the results of which will be discussed in more detail later.
Anatomical Requirements
Imaging
Historically, contrast aortography was used as a routine adjunct to axial imaging because it was felt to allow for a more accurate determination of vessel length and angulation before computed tomography (CT) reformatting was widely available. Preoperative angiography is now rarely employed and reserved for cases where an adjunctive therapeutic intervention (i.e., coil embolization) is necessary.18 Recently the advent of flat panel technology has allowed for development of rotational angiography techniques that facilitate construction of three-dimensional (3D) images found to be comparable to standard multidetector computed tomographic angiography (CTA). This technique has been termed C-arm cone-beam CT or fluoro-CT. Fluoro-CT uses a modified C-arm with specialized software and allows for precise measurements to be performed without using standard CT imaging.19 Its routine use in preoperative planning prior to EVAR is currently under investigation.20
Spiral CT of the abdomen and pelvis is the mainstay of aortoiliac imaging. The imaging protocol is different from the standard protocol for most abdominal CT scans. Acquisition should use a 1:1.5 helical pitch and 3- to 5-mm collimation.21 Two- to 3-mm slices are ideal for providing adequate information for stent graft planning. The two-dimensional (2D) images, however, can often be misinterpreted. The axial images may “cut” vessels at an angle, particularly iliac arteries that have some degree of tortuosity, thus creating an ellipse as opposed to visualization of the true lumen diameter. Due to this problem, some physicians recommend 3D image processing as a better method to evaluate aortoiliac anatomy for endograft therapy.22 Although it was common initially to perform an angiogram in addition to a CT scan, recent evaluations suggest that high-quality 3D CT scans alone may provide sufficient data for endovascular graft planning.18,23 Currently, proprietary products for CT postprocessing provide the ability to evaluate the 3D reconstruction of the aortoiliac system rapidly, rotate the images on the screen to obtain better vessel diameter measurements, and provide “virtual endograft” simulation.
Aortic Neck
The aortic neck is defined as the area of the aorta cephalad to the aneurysm in which the aortic endograft is placed (Fig. 40-1). This zone of the aorta is important for two reasons during aortic endografting. First, it is the site of proximal fixation that will prevent the device from migrating distally. Second, a circumferential seal must be obtained between the graft and the aorta in this area to prevent leakage of blood into the aneurysm sac. The exact length of aortic neck required is somewhat device dependent, but most commercially available devices require a 10- to 15-mm length of aortic neck below the level of the most caudal renal artery. Some investigational devices may allow for shorter necks. Several devices employ the use of a suprarenal uncovered (or bare) stent to provide additional protection against graft migration. Suprarenal stent fixation may be useful, particularly in patients who have a shorter aortic neck, because it transfers protection against migration to a more normal segment of aorta. The suprarenal stent, however, does not provide any function with regard to creating a circumferential seal.
In addition to the length of the neck, other anatomical characteristics are important when determining whether patients are suitable candidates for endovascular aneurysm repair. These include aortic neck angulation, the shape of the neck, and the quality of the neck. Neck angulation refers to an alteration in the direction the aorta takes with regard to the centerline pathway. Acute angulation of the aortic neck can greatly affect the endograft’s ability to obtain a proximal seal. Aortic neck angulation of greater than 60 degrees compared to the centerline is often considered prohibitive for endovascular aneurysm repair. The shape of the aortic neck also affects the ability of the graft to obtain a seal as well as fixation. A conical-shaped neck (Fig. 40-2) is generally felt to be unstable and predisposes to distal migration.24 An increase in diameter from the top of the neck to the bottom of greater than 10% is often believed to be a contraindication to routine aortic endografting. Presence of circumferential thrombus or aortic calcification can also negatively affect an endograft’s ability to obtain a proximal seal.
Iliac Arteries
The iliofemoral arterial system is important in endograft placement for two reasons. First, most endografts are placed through the common femoral artery (CFA) and must traverse the iliofemoral system to reach the aorta. Iliac artery diameter and tortuosity can adversely affect the ease with which the endograft traverses this course. This topic is covered in more detail below. Certainly the presence of significant atherosclerotic disease can cause arterial narrowing that inhibits placement of the device. In addition, tortuosity of the iliac arteries can hinder placement of the grafts (Fig. 40-3). Second, the iliofemoral system is important because it is the site of the distal seal between the endograft and the iliac artery, preventing retrograde flow of blood into the aneurysm sac. Many of the features necessary for an adequate aortic neck are also necessary for the distal landing zone. Presence of thrombus, calcification, and tortuosity can significantly hinder the iliac limb seal. Ectatic or aneurysmal iliac arteries obviously affect the ability of the graft to seal against the iliac limb. Most available endograft systems require at least a 15-mm segment of iliac artery to be of adequate caliber and free of significant disease to obtain a distal seal. If this is not present, adjunct interventions can be performed to assist in placing the device (i.e., iliac artery conduit placement, coil embolization of the internal iliac artery [IIA]). Management of these complicated situations is discussed in more detail later.
Endograft Design
Delivery System
Standard endograft insertion involves placement of the device through an arteriotomy in the common femoral artery, from where the graft traverses the external iliac and common iliac arteries (CIAs). The ability to deliver the endograft safely and effectively in this fashion is a prerequisite for effective repair. Three factors are important determinants of device delivery.25
Flexibility
Tortuosity, another anatomical variant, affects the ability to adequately deliver the endograft system. Tortuous iliac vessels can be “straightened” with the use of stiff guidewires, but this is not always possible or desirable. The ideal delivery system easily traverses these arteries on the basis of an intrinsic degree of flexibility. Again, different delivery systems have different abilities to track through tortuous iliac arteries, and some may be more successfully placed than others in this anatomical variant. Delivery systems composed of long, flexible, tapered tips pass more easily than those with short, stiff, blunt tips. In addition, other aspects of device construction, such as metallic struts that provide columnar strength, increase device rigidity and limit use in tortuous vessels.25
Deliverability
A number of features have been noted to affect the deliverability of endograft devices. As stated previously, long, flexible, tapered tips pass more easily than short, blunt, stiff ones. This allows for easier maneuverability through tortuous vessels, as well as past sites of narrowing. Larger-caliber devices are also more difficult to deliver, particularly in patients with smaller-diameter arteries.25 Some delivery systems allow for placement of the endograft system through alternate sheaths, whereas other systems necessitate use of the manufacturer’s own delivery system. This can greatly affect placement of specific endografts in specific anatomical variants. A thorough understanding of the patient’s arterial anatomy and the limitations of different endograft systems are important. The complexity of the delivery system also affects the ease with which it is placed. Some devices generally provide a simple maneuver to deploy the graft, whereas others have several complicated steps.
Specific Grafts
Various endografts are currently commercially available or in clinical trials in the United States. A brief description of the currently commercially available endograft systems (in the United States) is outlined in Table 40-1 and depicted in Figure 40-4.
Graft Placement and Postoperative Management
Once the patient is deemed an endograft candidate, the best graft has been chosen, and the device properly sized, the patient can undergo implantation. The majority of endografts are placed through the femoral arteries that have been operatively exposed. The majority of surgeons prefer the use of the transverse incision as it associated with a lower rate of wound complications (12.7% in transverse incision vs. 47.5% in vertical incisions).26 Percutaneous access for EVAR is growing in popularity, and its use will become even more widespread with the further development of low- profile devices. Suture-mediated closure devices facilitate this process, and using a “preclose” technique has been described to allow closure of sheaths as large as 24 F.27 Use of this procedure has been associated with 70% to 100% technical success, and immediate failures mandate surgical exploration of the femoral artery. Prospective analysis has demonstrated that use of a percutaneous approach may shorten operating times and reduce the rate of wound-related complications, without a significant increase in overall procedural cost.27–29 The aorta is then cannulated with a guidewire and catheter. Small boluses of contrast agent are delivered to further define the anatomy and localize the renal arteries. With an angulated aorta, it is important to remember that the best view of the renal arteries and visualization of the fixation zone may not be in a direct anterior-posterior plane but at a more cranial-caudal angle. The device is then generally advanced over a stiff guidewire and correctly positioned to allow the most extensive coverage within the aortic neck without intruding on the orifice of the renal arteries. Each device has its own unique instruction for actual deployment. Once the main body and ipsilateral limb have been placed, the contralateral limb has to be placed. The sequence of events for this varies depending on graft design—whether unibody or modular.
Recovery following EVAR is generally rapid and uncomplicated, and most patients are discharged home on the first or second postoperative day. Return to activities of daily living has been shown to be quicker following endovascular repair than open surgery. In addition, most patients report less postoperative pain. Aortic remodeling following EVAR, however, is a slow process that continues for several years. Anatomical changes in the native vessel, particularly at the proximal neck, can cause conformational changes in the implanted device that mandate close follow-up. In addition, late failures have been identified that have required reintervention.30 Given these facts, routine surveillance following EVAR is universally recommended, although there are no standard regimens, and the requirements of a standard intensive regimen are debated. Nordon et al. performed a meta-analysis evaluating secondary intervention rates based on contemporary graft implants.31 Their findings demonstrated that surveillance imaging alone initiated the secondary intervention in 1.4% to 9% of cases. Over 90% of EVAR cases, however, received no benefit from surveillance scans. Based on these findings, the group recommended that surveillance should be directed toward those patients identified as having a high risk for postoperative complications. Identification of this group, however, is not obvious but may be necessary in patients with complicated aortic neck anatomy or in patients in whom the device was used outside of the indications for use (IFU).
Contrast-enhanced CT is the most widely used modality for follow-up after EVAR. It is widely available, has rapid data acquisition, reproducibility, and is uniform across institutions. The major concerns associated with this modality are use of a contrast agent and the potential associated nephrotoxicity, radiation exposure, and cost. It is considered the gold standard for assessing aortic diameter, with nearly 100% sensitivity and specificity. Sensitivity and specificity rates for endoleak detection with CT are better than those with conventional angiography: 92% and 90% for CT vs. 63% and 77% for angiography, respectively.32–34 Triphasic CT (noncontrasted phase, arterial phase, and delayed phase) results in the greatest amount of information but at the cost of increased radiation exposure. Unenhanced CT imaging is useful for differentiation of endoleaks from calcifications from the metallic portion of a stent graft, and can help detect small perigraft leaks better than arterial-phase images. Use of arterial phase alone has a lower diagnostic value than combined arterial and delayed-phase scanning.35
Repeated CT scanning subjects the patient to potential carcinogenic risks associated with ionizing radiation exposure. The estimated lifetime attributable risk of death from cancer following an abdominal CT scan in a patient older than 50 years of age is 0.02%.31,36 Although this effect in itself is small, the cumulative effects over time with repeat imaging can be significant. Repetitive use of iodinated contrast can have a cumulative deleterious effect on renal function, especially in the elderly and those patients with preexisting renal impairment.37 Given this, as well as the expense, the use of alternate modes of surveillance has been evaluated.
Magnetic resonance imaging (MRI) and MRA provide much of the same imaging information that can be acquired by CT scanning. Multiple format MR images (T1, T2, gadolinium-enhanced) can be viewed in 2D and be reformatted into 3D volumes, allowing dimensional measurements, assessment of luminal patency, device positioning, and detecting the presence of an endoleak. Limitations of MRA/MRI include potential magnet-induced in vivo metallic heating or motion, which may prevent imaging in the immediate postimplant time period. In addition, postimplant artifacts, in particular with ferromagnetic metallic stents, will limit morphological assessments. Furthermore, there is a risk of nephrogenic systemic fibrosis associated with gadolinium contrast use in patients with renal insufficiency.38 Benefits of MRI are related to the lack of exposure to ionizing radiation and low nephrotoxicity of MR contrast medium. Disadvantages of MRI are its lack of wide availability, longer procedure time than CT, patient claustrophobia, and contraindications for patients with cardiac pacemakers.
Color duplex ultrasonography (US) is a convenient, noninvasive, inexpensive portable means of postimplant surveillance. Its reliability as a useful surveillance tool, however, is still debated. Grayscale US is accurate for measurement of aortic aneurysm diameter. Endoleak detection by US requires color duplex. The reported specificity rates of color duplex US for endoleak detection are high (89%-97%), but the sensitivity and diagnostic power of color duplex US for endoleak detection compared with CT is still debated.39 The use of duplex US to detect an endoleak has a sensitivity of approximately 77%, with a specificity approaching 94%.40 The addition of US contrast agents increases sensitivity to 98%, with no significant change in specificity.40 Contrast agents are useful in identifying slow leaks that are not readily discernible on CT.41,42 Detection of flow direction of the endoleak is also an advantage of US that is not easily discernable with CT.43 Presence of a “to-and-fro” flow pattern is associated with spontaneous closure of the endoleak, whereas a monophasic or biphasic waveform is consistent with endoleak persistence.43 Limitations of color duplex US include its operator dependence, variation based on patient physical size, and the need for optimal patient preparation. The substitution of duplex US imaging for CT, however, may result in long-term cost savings.44
Problems with Endografting and Management
Iliac Artery Disease
When iliac artery disease is present, whether it be aneurysmal disease, atherosclerotic disease, or severe tortuosity, the use of an iliac conduit can provide a safe route to deliver the endograft.45 In cases of iliac artery lumen narrowing resulting from atherosclerotic disease or increased vessel tortuosity, advancement of the device, despite the presence of resistance, can result in rupture of the iliac artery. Iliac artery rupture has been reported in 1% to 2% of cases.46,47 To circumvent prohibitive iliac artery anatomy, an iliac conduit can be used. An iliac conduit involves suturing a prosthetic graft (generally 8-10 mm in diameter) to the mid–CIA even if it is aneurysmal. This can be done in an end-to-end or end-to-side fashion, although the latter often provides a greater lumen for passage of the device. The device is placed through the prosthetic graft, and the iliac limb of the endovascular graft traverses the CIA and anastomosis and seals within the conduit. The distal end of the graft is tunneled along the natural course of the iliac artery and anastomosed to the femoral artery. The distal end of the CIA is oversewn to allow retrograde flow through the EIA to supply the ipsilateral hypogastric artery. Alternatively, the hypogastric artery can be anastomosed directly to the conduit.
Iliac artery ectasia or aneurysms can present a problem in obtaining a distal seal. Enlarged CIAs are present in up to 30% of patients presenting for endovascular aneurysm repair.48–51 Many available endografts do not allow for ectatic or aneurysmal common iliac arteries, so the distal seal may have to be obtained within the external iliac artery, which is often of normal caliber. If the distal seal occurs in the external iliac artery, the hypogastric artery is generally sacrificed using coil embolization. Presence of a hypogastric artery aneurysm would necessitate the same approach. Rarely, bilateral hypogastric artery embolization is required. Hypogastric artery embolization can occur before aneurysm repair or concurrently. If bilateral embolization is planned, it is generally performed in a staged fashion, although its occlusion is not always planned. Hypogastric artery embolization is not without risk, and side effects can occur in up to 50% of patients.49 Buttock claudication is the predominant complaint after hypogastric artery occlusion. This occurs in 12% to 50% of patients, but in most it generally resolves after several months.48–53 Some 5% to 25% of men complain of new-onset erectile dysfunction (ED).51,52 Buttock ischemia and bowel ischemia requiring resection are of theoretical concern, but they have not been described in any of the larger series. Patients requiring embolization in the more distal branches of the hypogastric artery (as might be done with the presence of an IIA aneurysm) and those in whom coil placement was not adequately controlled are at higher risk of developing pelvic symptoms.53 Bilateral hypogastric artery embolization has not been associated with increased symptoms when compared with unilateral occlusion.49,50,52 Coil embolization of the IIA is not necessary if it is not aneurysmal. In the face of CIA aneurysms, Wyers et al.54 have shown that if there is a 5-mm neck of iliac artery proximal to the hypogastric artery in addition to a 15-mm neck in the external iliac artery, coil embolization of the hypogastric artery is not necessary to obtain a distal seal. This may be possible in up to two thirds of patients requiring coverage of the hypogastric artery.
