INTRODUCTION

Aortic reconstruction is an index operation, one whose outcome is used to compare surgeons and centers. With validated tools, the surgeon can be made the statistical variable.¹ In an era of ranking and league tables, there are obvious implications. For patients, the implications are more critical. But aortic surgery is difficult. Even in the best hands, complications occur. Despite the advent of endovascular procedures, there are still those who are unsuitable for, or unwilling to undergo, stent grafting, and even this procedure is not without complications.

Infrarenal aortic surgery can be divided into three areas: (1) aortobifemoral bypass for occlusive disease, (2) elective abdominal aortic aneurysm (AAA) repair, and (3) repair of ruptured AAAs. There are two approaches to the aorta: transabdominal and retroperitoneal. In this chapter, we describe how we do aortic surgery at Albany Medical Center, by the retroperitoneal approach. We share the strategies we use to minimize complications. More pertinently, we describe what we do when complications arise. Young surgeons will read of these and doubt their relevance. Older surgeons will read them and empathize. In the surgical literature, papers detailing success outnumber those documenting failure. Therefore, much of this chapter is based on our own complications or dealing with the aftermath of others’.

We use the retroperitoneal approach for all aortic repairs including ruptured aneurysms.² We accept this is a minority practice. Most aortic surgery is performed via a transperitoneal approach. Despite our bias, we still employ the transabdominal approach when other pathology needs to be addressed at the same operation. However, before describing techniques in detail, it is worthwhile addressing some recurring arguments.

CONTINUING CONTROVERSIES

End-to-Side or End-to-End Anastomosis in Aortobifemoral Bypass? (Fig. 59-1)

In Europe, the end-to-side anastomosis is favored. In North America, end-to-end is preferred. It is argued that an end-to-end anastomosis is easier to perform. It is more anatomic and avoids competitive flow between the graft and the native arteries. Therefore, it ought to have better patency and carry low risk for duodenal fistulas. The argument in favor of the end-to-side approach is that it preserves anterograde flow to the pelvic viscera, and in the event of graft occlusion, the patient’s status reverts to that prior to surgery rather than being profoundly worse, as might be the case if an end-to-end anastomosis occluded.

Figure 59-1 End-to-side aortobifemoral bypass with Dacron.

Both opinions steer clear of the facts. There have been randomized, controlled studies comparing both approaches. An initial study of 79 patients in 1982 from Chicago showed small but significant advantages for end-to-end anastomoses over end-to-side.³ However, a larger study from Becquemin’s group⁴ in Paris in 1990 of 158 patients refuted this, as did a study from Toronto of 120 patients.⁵ The most impressive statistic, however, was just how durable aortic bypasses were, regardless of the means used to sew them together. Actuarial primary 5-year patency rates were 90%. Secondary patency rates were 98%. So there is really little to choose between either approach. The best advice came from the French who suggested that “as we could not find any difference between the results in the two groups, we suggest choosing the simplest procedure which maintains adequate pelvic and colonic blood supply, according to angiographic findings.”

Polytetrafluoroethylene or Dacron?

Most centers use Dacron for aortic surgery. These grafts are either knitted collagen-coated or knitted gelatin-coated. Woven Dacron grafts are not widely used now. The alternative is expanded stretch polytetrafluoroethylene (PTFE). Since 1991, we have used stretch Gore-Tex grafts for all our aortic anastomoses. Concerns about needle hole bleeding are unfounded. We find that stretch PTFE has handling characteristics that are at least as good as those of Dacron. Theoretically, PTFE is also less likely to dilate over time.

Again, there are good randomized, controlled data comparing the various materials. A well-performed prospective study from Cornell in 1995 compared PTFE and Dacron and showed no difference in outcomes.⁶ A large multicenter, prospective, randomized, controlled trial of gelatin-impregnated Dacron, collagen-impregnated Dacron, and PTFE involving 315 patients from Vienna in 2001 also failed to show a difference.⁷

Clamp the Aneurysm Neck or the Common Iliacs First?

The sequence of applying clamps to AAAs has exercised some of the better minds in vascular surgery. Whether to clamp the proximal neck prior to the iliacs or vice versa? At the outset, we should declare an interest—we always clamp the iliacs first in elective cases.

There are two questions: (1) Does clamping the iliacs first protect against emboli traveling down the leg? (2) Does clamping the aortic neck first protect against embolization to the renal and visceral arteries from the aortic sac? A study from Leicester in the United Kingdom in 2004 examined the first question by comparing the rate of embolization down the superficial femoral arteries of patients undergoing aortic surgery using a transcranial Doppler.⁸ They showed no difference between the “aorta-first” group and the “iliacs-first” group. The second question was most recently addressed by the Montefiore Medical Center study in 1999.⁹ Although this was an animal study in nonatherosclerotic aortas, they suggested that clamping the aorta first could protect against embolization to the renal arteries.

Retroperitoneal or Transperitoneal? (Fig. 59-2)

Why do we favor the retroperitoneal approach? Aside from the theoretical considerations of quicker return of bowel and respiratory function, we feel it is more versatile. It provides easy access to the left common iliac and internal iliac arteries. Once mastered, it also provides easier access to the left renal artery and the aortic neck. It avoids the left renal vein, which is reflected anteriorly out of the operative field, and also avoids the gonadal veins, which can be a problem when the aorta is approached from the front. Many of the aneurysm repairs we perform are those rejected for endovascular repair and are really juxtarenal or suprarenal aneurysms. Suprarenal clamping, therefore, becomes an important issue. Once the lumbar branch of the left renal vein is ligated and the peritoneal contents and kidney are retracted cephalad and medial, access to the infradiaphragmatic aorta can easily be obtained by incising the left crus. This allows the proximal aortic clamps to be placed above, below, or between the renals as well as on the supraceliac aorta. However, the procedure has a definite learning curve. Using a left flank incision, access to the right common and internal iliac arteries is difficult and, in many cases, impossible. To access these vessels, we perform a separate right counterincision. Reimplantation or bypass of the right renal artery, when required, is also technically difficult but can be performed with experience.¹⁰ The retroperitoneal approach is more time-consuming when performing a straightforward infrarenal tube graft. The areolar tissue around the aorta is also vascular and can result in blood loss that is usually not encountered in the conventional approach.

Figure 59-2 Transperitoneal exposure of an abdominal aortic aneurysm (AAA).

There are randomized, controlled data. Initial reports were equivocal. In 1990, the Massachusetts General Hospital group showed little difference in outcomes between the two techniques.¹¹ However, in 1995, Sicard and coworkers¹² reported the results of a randomized, controlled trial of 145 patients. Whereas there was no difference in mortality rates, the retroperitoneal approach was associated with fewer postoperative complications, shorter hospital and intensive care stays in the hospital, and lower cost. In 1999, Kirby and colleagues¹³ from Atlanta reported on 92 high-risk American Society of Anesthesiologists Class IV (ASA IV) patients randomized to either transabdominal or retroperitoneal aortic repair. Complications were significantly lower in the retroperitoneal group. So, it would seem that proponents of the retroperitoneal approach, including ourselves, are vindicated by the literature. Unfortunately, in the interests of balance, we must also include Lawrence-Brown and associates’ trial from Perth in 1997 ¹⁴ involving 100 patients that showed no difference at all between the two techniques. In the end, it seems that either approach is justified as long as outcomes are acceptable and pitfalls avoided.

Aortobifemoral Bypass by the Retroperitoneal Approach

OPERATIVE STEPS

The left flank approach was first described by Williams and coworkers at Johns Hopkins in 1980.¹⁵ They described an incision through the 11th intercostal space with division of the left crus of the diaphragm. Our technique is based on this original description. We do not use the so-called anterior retroperitoneal approach of Schumacker, whereby the retroperitoneal space is developed using a vertical midline incision.¹⁶

Step 1 Patient positioning

Step 2 Exposure of both femoral arteries

Step 3 Left flank skin incision

Step 4 Reflection of peritoneum and creation of retroperitoneal space

Step 5 Dissection of aorta

Step 6 Fashioning of proximal anastomosis

Step 7 Tunneling of graft limbs and performing femoral anastomoses

Step 8 Inspection of peritoneal cavity and closure of flank incision