The evolution of aortic endovascular therapies has been dramatic in the past decade. Understanding this evolution requires examination of three critical components; these components depend on each other to obtain favorable results for endovascular treatment of the entire aorta and its major branches. The three components are (1) the ability to obtain and process detailed preprocedural imaging to clearly understand a particular anatomic situation, (2) improvements in intraprocedural imaging to clearly perform what is intended, and (3) development of the devices in terms of accuracy of position and achieving a durable seal. These components, which have progressed while successfully treating increasingly complex cases, are interwoven with the skill of the operator performing these cases.

The purpose of this article is to address the first component above, which considers the preprocedural imaging involved in complex cases. Further, this article offers a perspective as to how three-dimensional (3D) computed tomography angiography (CTA) contributes to a better understanding of what is being accomplished. The ultimate goal is accurate and durable control of life-threatening aortic disease. This is challenged by the fact that treatment is being performed in a minimally invasive fashion. The basis of this imaging has been the increase in the quality of fine-slice resolution (the raw CT axial slices) and the ability to postprocess virtual 3D models of a particular anatomic situation in terms of overall morphology and functional measurements applied to that model of a real case.When these models are appropriately studied, the actual execution of the case becomes a simpler, more predictable, faster, and safer exercise. This occurs because of the second and third components: better intraprocedural imaging and better products.

The time spent carefully planning a case is time spent well. Attempts to completely automate case planning have been problematic for a number of reasons. In a study by Wyss et al, the average time spent by a physician operator reviewing and planning a case was 17 minutes.1 They found that the intraobserver consistency was very good at planning with little variability. The interobserver error was a bit higher but still demonstrated excellent consistency when the interventionist was involved in the planning of a particular case.

The routine use of automation and set protocols is an excellent starting point for automated 3D software. The basis of most of these is the center lumen line (CLL). The CLL is an easily reproducible reference that extends along the entire length of the aorta, into the major branch vessels, and all the way to access points. However, the purpose of operator review is to take into account the variations necessary to carefully plan a particular case. These can include alterations of where the endograft will lie varying from the CLL. The situations in which this can be a major factor are if the aortic pathology is (1) particularly saccular, (2) very large aneurysms—especially if no thrombus is contained, (3) tortuous aortas, (4) complex thrombus in the aortic sac, and (5) abrupt angulation of the iliacs entering the aortic flow channel. Additional processing of the CLL needs consideration of how a guidewire and endograft delivery system will align in one of the above situations and vary from the CLL.

Standard 3D software is an excellent way of seeing the full axis of the aorta and its branch relationships. The importance of this will be discussed in this article with specific considerations for the abdominal and thoracic aorta.

SPECIFIC ABDOMINAL CONSIDERATIONS
Endovascular treatment of abdominal aortic disease has several fundamental considerations; these mainly involve a keen understanding of (1) the adequacy of access anatomy, (2) the proximal neck quality and renal relationships, and (3) the distal landing zones.

The access anatomy must have adequate quality, caliber, and lack of tortuosity to safely introduce selected devices. Three-dimensional imaging allows a more thorough investigation of these variables to help determine which delivery systems could be safely used. If two out of the three previously mentioned variables are considered challenging, alternative methods should be considered. These possibilities would include percutaneous transluminal angioplasty/stenting, endoconduits, and formal surgical conduits.

The proximal neck quality is a very complex consideration of the actual quality of the aortic tissue at that level and the degree of thrombus/calcification, the shape of the neck (3D), the angle of the neck (3D), and the relationship of the renal arteries to each other and their position on the neck.2 The quality of the neck is first determined in terms of whether the neck is concentric or irregular at the seal zone. Commonly, there is some degree of irregularity at the posterolateral quadrants due to the aorta lying on the spine, which, along with the degree of calcification/thrombus, must be carefully looked at when both sizing and positioning a particular case. Recall that the renal arteries have identifiable accessory vessels in 15% of kidneys. There is frequently linear offset, and one of the two renal arteries (left) is usually lower than the other. A clear understanding of the 3D takeoff of the renal arteries assists in optimal alignment of the II angle to be truly orthogonal to the takeoff the lowest renal. Frequently, the right renal artery will originate in the right anterior quadrant, and the left renal artery will originate in the left posterior quadrant of the transverse view of the aorta. Optimal visualization for endograft deployment occurs most often with some degree of a left anterior oblique (LAO) projection, usually in the range of 10º to 45°.

The infrarenal aneurysm dilates, and as it does, the neck is lifted and angles anteriorly to a variable degree, mandating some cranial angulation of the II to get orthogonal to the neck itself.3

The complexity of this arrangement is often grossly underestimated when performing a particular case (Figure 1). For example, when the endograft is delivered to a 25-mm-diameter neck that is only 20 mm long and has an anterior angulation of 45°, presumed optimal visualization will only require the II to have minimal or no cranial angulation to appear orthogonal to the upper end of the constrained endograft. However, when the graft is deployed, it will foreshorten in the angulation of the neck and ultimately land several millimeters lower than intended. This concept is important in thoughtful 3D assessment of a particular case. This also argues for tip capture/control of grafts in more challenging anatomies to be able to fine-tune the exact position of an opened but not yet anchored endograft.

Further, the relationship of the renal arteries relative to how far ventral or dorsal they come off the aorta needs to be considered to assess the functional length of the aortic neck (Figures 2 and 3). In the same model as previously described, consider the following variation. If the renal artery appears to come off the more ventral portion of the aorta, there will be less neck available for a landing/seal zone than if it comes off more dorsal. Both of these renals will come off on the same axial CT slice and yet affect the case and its ultimate outcome very differently. These relationships are better understood with a virtual 3D model.

The functional positioning of the endograft relative to the abdominal aortic aneurysm sac and neck must be carefully considered.4 The gait of the graft (in most systems) will land 8 cm below the top of the graft. The quality of the abdominal aortic aneurysm lumen can have a significant impact on the ability to cannulate the gait and the ultimate function of the gait to contralateral limb docking.

The orientation of the aortic bifurcation and the ostia of the common iliacs will also affect the alignment of the graft and the ease of performing the procedure. Frequently, the left iliac is more dorsal than the right. I will often use the left side for the ipsilateral access. The gait and the right iliac ostia can then be aligned for easier cannulation. However, more importantly, there seems to be better alignment of the endograft components and less twisting and abrupt angulation of devices.

The final consideration is the amount of dilatation of the common iliacs relative to the external iliacs. The more dilated the common iliacs, the more elongation and tortuosity present. When this occurs, the angle between the common and external iliac arteries will be more acute and can actually be < 90º. This can be a significant concern if a large common iliac limb (eg, 18 or 20 mm) encroaches the ostia of a standard size external iliac (eg, 7–9 mm). The ventral part of the limb will lie closer to the external iliac and potentially compromise a smooth flow channel. A clear understanding of this distal relationship before the case with the 3D software is important. During the procedure, appropriate obliquity for visualization needs to be understood ahead of time to maximize this fluoroscopic view. Careful preprocedural planning improves the efficacy of radiation and contrast doses and the accuracy of the procedure (Figure 4).

SPECIFIC THORACIC CONSIDERATIONS
A complete discussion of the complexities of the thoracic aorta and the possibilities for dissections and deceleration injuries is beyond the scope of this article. However, there are several key relationships that first must be understood when dealing with descending thoracic aneurysms. The aortic arch itself has a predictable progression with age; the arch will dilate, elongate, and rotate. Normal relationships include the fact that the innominate artery is typically more dorsal, the common carotid straight off the mid-apex, and the subclavian more ventral. With age, the arch dilates and then rotates from a typical LAO relationship to the body of 20° to 30° up to a full 75° or greater LAO angle over time. This angle for deployment of the graft is fundamental to the performance of thoracic endografting. The preoperative 3D images will assist in the positioning of the patient so that an optimal functional imaging angle can be obtained. The positioning on the table is very important with regard to not only how far the body is rotated but also for side-to-side positioning so that there is a clear fluoroscopic view free of side rails, arm boards, arms, and monitoring devices. An older patient frequently has a larger, more gradual radius of curvature for delivery of the graft, which leads to less infolding and kinking of the endograft. This concept is a primary problem with endografting younger patients with a tighter radius curvature in smaller-diameter aortas.

The distal aorta in particularly large aneurysms can fold laterally in the left chest because of relative fixation at the level of the diaphragm. This raises concern for delivery and/or kinking of the lower end of the endograft. This will also significantly affect the ability to land at the intended target relative to the celiac artery. The 3D images will clearly show the CLL, and the full consideration of length and angulation can be more clearly appreciated.

The distal landing at the celiac/superior mesenteric artery level is better understood with accurate imaging. The anterior wall of the visceral segment of the aorta is commonly of better quality because of the embryologic interlacing of fibers due to the takeoff of the ostia of the four visceral vessels, whereas the posterior wall is frequently more diseased (eccentric). An appreciation of this will assist particularly in the preoperative planning of size and landing site of devices.

CONCLUSION
Aortic endograft therapy is a rapidly developing field in vascular medicine. A clear understanding of the vascular anatomic-to-graft relationships is fundamental to the success of the procedures. The description of relationships involved and the use of 3D imaging to understand them before the procedures should be a comfortable component in the abilities of an endovascular specialist who wishes to treat the aorta in this manner. Unfortunately, the ability to convey the advantages of a 3D platform is limited by the very fact that I am communicating this in a 2D format. I have tried to illuminate some of the key concepts that I have learned in doing these cases and thinking about them in a 3D perspective. I would encourage all operators to consider carefully working to understand the advantages of the 3D format. The best recommendation is to work with vendors who are supporting the spread of this technology and spend time with local representatives who want to share this knowledge with you to gain a stronger understanding for your individual practice.

Douglas Massop, MD, FACS, is a vascular surgeon at The Iowa Clinic, PC, in Des Moines, Iowa. He has disclosed that he is a paid consultant to Medtronic, Inc., and Cordis Corporation. Dr. Massop may be reached at (515) 875- 9090; d.massop@iowaclinic.com.

  1. Wyss TR, Dick F, England A, et al. Three-dimensional imaging core laboratory of the endovascular aneurysm repair trials: validation of methodology. Eur J Vasc Endovasc Surg. 2009;38:724-731.
  2. Filis KA, Arko FR, Rubin GD, Zarins CK. Three-dimensional CT evaluation for endovascular abdominal aortic aneurysm repair. Quantitative assessment of the infrarenal aortic neck. Acta Chir Belg. 2003;103:81-86.
  3. Diehm N, Katzen BT, Samuels S, et al. Sixty-four-detector CT angiography of infrarenal aortic neck length and angulation: prospective analysis of interobserver variability. J Vasc Interv Radiol. 2008;19:1283-1288.
  4. Sprouse LR 2nd, Meier GH 3rd, Parent FN, et al. Is three-dimensional computed tomography reconstruction justified before endovascular aortic aneurysm repair? J Vasc Surg. 2004;40:443-447.