Descending thoracic aorta-to-iliofemoral artery bypass as an alternative to aortoiliac reconstruction E n r i q u e Criado, M D , G e o r g e J o h n s o n Jr., M D , Steven J. B u r n h a m , M D , Jeffrey Buehrer, M D , and Blair A. Keagy, M D , Chapel Hill, N.C. During the last 3 decades subcutaneous extraanatomic bypass, despite its limited durability, has been the favored alternative to infrarenal aortofemoral bypass. Meanwhile, the descending thoracic aorta has been scarcely used as an inflow source for aortoiliac reconstruction. Over the past 8 years we performed 16 bypasses from the descending thoracic aorta to the iliofemoral vessels for occlusive disease. Our experience combined with that found in the English-language literature totaled 141 patients. In 79 patients (56%) the indication for surgery was failure or infection of an abdominal aortofemoral graft. Previous abdominal operations, sepsis, radiation therapy, the presence of abdominal stomas, or an unsuitable infrarenal aorta were the indications in the remaining cases. The combined operative mortality rate was 6.4%. The life-table primary graft patency was 98% at i year, 88% at 2 years, and 70.4% at 5 years. Bypass from the descending thoracic aorta to the iliofemoral artery uses an inflow source superior to other extraanatomic reconstructions, does not require aortic cross-clamping, avoids the abdominal cavity, and places the graft remote to the skin and intestine. The operative mortality and patency rates compare favorably to those of other extraanatomic or remedial aortic reconstructions. Descending thoracic aorta to iliofemoral artery bypass is a superb alternative to abdominal aortofemoral bypass, carries a low mortality rate, has an excellent short-term patency, and features unique characteristics for long-term durability. (J VAsc SURG 1992;15:550-7.)

On June 7, 1956, Lester R. Sauvage performed a bypass graft between the descending thoracic aorta and the femoral arteries in a patient with a history o f disabling intermittent daudication and a thrombosed aortoiliac graft. During the operation a transabdominal approach with a median sternotomy extending into the left fourth intercostal space was used, and the graft was routed through the peritoneal cavity. Two preserved aortic homografts were anastomosed and interposed between the descending thoracic aorta and the femoral arteries. The patient returned to unrestricted ambulation, and the graft remained patent for 20 months. This operation, however, was not reported until September o f 1961.1 One month later, Blaisdell et al. 2 published a case report o f a patient requiring removal o f an infected aortoiliac polytetrafluoroethylene prosthesis in whom a 14 m m Dacron graft was anatomosed to the descending From the Department of Surgery, Schoolof Medicine, University of North Carolina at Chapel Hill. Reprint requests: Enrique Criado, MD, University of North Carolina Department of Surgery, 210 Burnett-WomackBldg., CB No. 7210, Chapel Hill, NC 27599-7210. 24/1/32727 550

thoracic aorta through a low anterolateral thoracotomy by use o f an extraperitoneal tunnel anastomosed to the left femoral artery with a suprapubic branch graft to the right femoral artery. Unfortunately, the patient died 4 weeks later o f intraabdominal sepsis. These initial reports were attempts to use the descending thoracic aorta as an alternative inflow source in situations where the use o f infrarenal aorta was ill advised. Shortly after these communications, the development and popularization o f the axillofemoral bypass for extraanatomic aortoiliac reconstruction dissipated interest in the use o f the descending thoracic aorta as an inflow source. Therefore during the next 20 years only two small series and a few case reports describing variations o f the original techniques were published, a-9 Furthermore, until January o f 1991 only five series were reported o f 10 or more patients with descending thoracic aorta to iliofemoral artery bypass grafts (thoracofemoral bypass)) °-14 During the last 2 decades axillofemoral bypass has been the most widely used alternative for aortoiliac reconstruction in patients in whom the abdominal approach is contraindicated or unduly risky, or in whom survival is judged to be limited. Meanwhile,

Volume 15 Number 3 March 1992

thoracofemoral bypass has remained ignored. The accumulated experience with axillofemoral bypass, however, has shown that this procedure is associated with significant morbidity and mortality rates, has a low patency rate, and requires reoperation frequently25-18 Despite this, it has remained the favored alternative for aortoiliac reconstruction. The thoracofemoral bypass offers an optimal inflow source and is a fairly benign operation. In many situations it has notable advantages over conventional aortofemoral bypass and may offer as good or better patency rates. Therefore thoracofemoral bypass constitutes a superior alternative to axillofemoral bypass in many cases. The purpose of this study was to review our experience with this type of surgery at the University of North Carolina at -,_;hapel Hill and that recorded in English-language literature. PATIENTS A N D M E T H O D S

The records, noninvasive arterial studies, and arteriograms of all patients who underwent aortoiliac reconstruction with use of the descending thoracic aorta as the inflow source from July 1982 to lanuary 1991 were reviewed. During this same 81/2-year period 78 infrarenal aortofemoral bypass grafts were performed at our institution. The collected data included age and sex of the patient at the time of surgery, significant medical illness, previous aortoiliac or distal arterial reconstructions and presence of infrainguinal occlusive disease, preoperative and postoperative anlde-brachial indexes (ABIs), and material and type of graft used. The indication for surgery was categorized into limb salvage or severe claudication and also according to the reason for the use of the thoracic aorta for inflow. The type of operation was classified according to the distal anastomotic site and whether it was the first aortoiliac reconstruction or a secondary procedure after previous graft failure. The postoperative complications, 30-day operative mortality rate, late mortality rate, and graft patency were recorded. Graft failure was considered when at least one limb of the graft occluded. Follow-up was obtained from routine clinic visits and by telephone contact in some cases. A search of the English-language literature was conducted, and those cases of aortoiliac reconstruction that used the descending thoracic aorta as inflow source were collected. The indications, type of graft used, operative mortality rates, complications, primary graft patency, and patient survival were recorded. The cumulative patency rate was obtained by the life-table method. Statistical comparison between

Descending thoracic aorta to femoral artery bypass 551

groups of paired data was done by means of the Student paired t test. Surgical technique The patient is positioned with the left hemithorax elevated 30 to 45 degrees and the pelvis as flat as possible to allow access to both groins. A bean vacuum bag is optimal for patient positioning. Endotracheal intubation is performed with a double lumen tube to facilitate aortic exposure by collapsing the left lung. An epidural catheter is placed in the operating room for intraoperative and postoperative pain control. To reduce operative time, a double-team approach is ideal for this procedure. Dissection of the distal vessels is done first to minimize the time of exposure of the pleural cavity. Infraingninal longitudinal incisions are used if the femoral arteries are selected for the distal anastomosis. The left inguinal incision is carried 5 to 7.5 cm above the inguinal ligament, and the external oblique muscle aponeurosis is opened in the direction of the fibers, the internal oblique and transverse muscles are divided, the preperitoneal space is entered, and the external iliac vessels are exposed by reflecting the peritoneum and its contents medially. By use of blunt dissection, a tunnel is created lateral to the iliac vessels and posterior to the left kidney up to the insertion of the diaphragm. A second runnel is created between the left suprainguinal preperitoneal space, already exposed, and the right groin. The latter runnel courses cephalad and anterior to the bladder and posterior to the rectus muscles. The medial aspect of the right inguinal ligament is partially divided to prevent graft limb compression. Alternatively, when the iliac arteries are chosen for distal anastomosis, the incisions are placed above and parallel to the inguinal ligaments, and the iliac vessels are exposed extraperitoneally. An additional left suprainguinal extraperitoneal incision is useful to facilitate tunneling; however, with experience in the procedure, this may be avoided. Once the distal arteries are exposed and the tunnels created, attention is turned to the thoracic incision. An anterolateral thoracotomy through the eighth intercostal space is used. Rib resection is not required. Although the sixth through the ninth intercostal spaces are also adequate for exposure of the descending aorta, the lower incisions allow placement of the proximal anastomosis at a more distal site shortening the total length of the graft. The left lung is deflated, and the inferior pulmonary ligament is taken down. The pleura is dissected off

552

Journal of VASCULAR SURGERY

Criado et al.

Table I. Patients with descending thoracic aorta to femoroiliac bypass grafts Best ABI increase Patient No./age/race/sex

Symptoms

Indication

1/58/W/F 2/69/W/M 3/68/W/M 4/38/B/M 5/52/W/M 6/59/W/M 7/66/W/F 8/56/W/M 9/81/B/M 10/68/W/M

Rest pain Dis claud Rest pain Rest pain Rest pain Dis claud Foot ulcer Dis claud Foot ulcer Rest pain

lI/57/W/F

Dis claud Dis claud Rest pain Rest pain Dis claud Dis claud

Graft occlusion Hx diverticulitis Graft occlusion Graft occlusion Graft occlusion Graft occlusion Inadequate IA Horseshoe kidney Inadequate IA Previous abdominal surgery Inadequate IA Inadequate IA Graft occlusion Inadequate IA Inadequate IA Inadequate IA

12/63/W/F 13/51/B/M 14/51/B/F 15/53/W/M 16/56/W/F

Graft type

Distal anastomosis

(R)

(L)

Days on ventilator

8 mrn KD 16 x 8 KD 8 mm Yd) 14 x 7 KD 10 mm KD 10 mm WD 16 x 8 WD 16 x 8 KD 16 x 8 KD 18 x 8 KD

Left CFA Bil CFA Left deep femoral Bil CFA Right SFA Left deep femoral Bil CFA Right CFA, left EI Right EI, left CFA Bil SFA

0.66 0.54 1.11 1.02 0.67 0.22 0.84

0.76 0.49 0.68 0.07

1 12 1 1

-

1

0.76 0.71 0.97 0.23 0.72

1 6 0 1 1

14 14 18 14 16 14

Bil CFA Bil CFA Bil CFA Right CFA, left EI Right EI, left CFA Bil CFA

1.11 0.58 0.74 0.54 0.74 0.30

1.06 0.38 0.74 0.26 1.06 0.53

2 1 1 1 0 0

x x x x x x

7 7 9 7 8 7

KD WD KD KD KD KD

Dis claud, Disabling claudication;/A, infrarental aorta; KD, knitted Dacron; WD, woven Dacron; CFA, common femoral artery; SFA, superficial femoral artery; E/, external iliac artery; Bil, bilateral.

the distal aorta in a length of approximately 7.5 cm. Circumferential control of the aorta proximally is optional. Extreme caution is taken to prevent intercostal artery injury. The posteromedial costophrenic sulcus is entered bluntly, and the left pleuroiliac retroperitoneal tunnel is completed. At this point systemic heparinization is given, and proximal control of the aorta is obtained with a partial occlusion clamp. The selected bifurcated graft is beveled and, using its entire length, anastomosed to the descending aorta in an end-to-side fashion. The graft is then tunneled to the left preperitoneal space and from there to both groins where the distal anastomoses are completed. If the right limb of the graft is short, as happens in some large patients, it can be lengthened by using the redundant segment from the left limb. The pleural space is drained. RESULTS

From July 1982 to January 1991 at The University of North Carolina Hospitals 16 patients underwent aortoiliac reconstruction for occlusive disease with use of the descending thoracic aorta as inflow source. The age of the patients ranged from 38 to 81 years, with a mean of 59 years. Ten patients were men and 12 were white. Eleven patients were smokers at the time of surgery, nine were on medication for hypertension, six had documented coronary artery disease, three of w h o m had had myocardial infarctions, three had cerebrovascular disease, one of w h o m had had multiple previous strokes, and three had a mild elevation of the serum creatinine. N o

patients in this group were diabetic. In nine patients the operation constituted the first arterial reconstruction for aortoiliac occlusive disease, whereas in six the operation was a secondary procedure after failure of one or more previous aortofemoral grafts, and in one patient it followed failure of an axillofemoral bypass. Eight patients had preoperative evidence of infrainguinal occlusive disease, four of w h o m were in the group undergoing primary reconstruction. In addition, 29 previous revascularizations, including axillofemoral, femorofemoral, and infrainguinal bypass grafts, had been performed in these patients. Three patients had a previous lower extremity amputation, two above and one below the knee. The indications for surgery were rest pain in seven patients, severe bilateral claudication in seven, and nonhealing foot ulceration in two (Table I). The reason to elect the descending thoracic aorta for inflow source was a previous failed transabdominal aortic reconstruction in six patients, previous intraabdominal sepsis or surgery in two, horseshoe kidney in one, and surgeon's preference because of the presence of heavy calcification and plaque in the abdominal aorta in seven cases. The arteriographic studies revealed that the abdominal aorta was completely occluded in 12 patients, occluded at a juxtarenal level in eight, and at a lower level in four. In four patients the infrarenal aorta was severely diseased but not totally occluded. The origin of the inferior mesenteric artery was occluded in 11 patients. The superior mesenteric artery was stenotic in two patients with

Volume 15 Number 3 March 1992

Days in intensive care unit

Descending thoracic aorta to femoral artery bypass

Days in hospital

Complications

3 13 1 5 2 3 18 1 3 3

9 I3 10 10 10 5 18 6 11 7

None Death None None None None Death None Groin seroma None

3 3 2 6 1 2

6 10 22 16 5 7

None Left lung collapse Groin infection Stroke None Groin infection

Status at follow-up Dead at 34 m o Well at 78 m o Inf. at 28 m o Inf. at 20 m o Well at 46 mo Well at 30 mo Well at 19 mo Well at 20 mo Well Well Well Well Well Well

at at at at at at

20 mo 18 mo 18 m o 6 mo 3 mo 3 rno

juxtarenal aortic occlusion who had no symptoms suggestive of mesenteric ischemia. The renal arteries were patent in all patients, with mild stenosis present in five. Three of the five had juxtarenal aortic occlusion, and only one had a slight elevation of the creatinine level. No evidence existed of renovascular hypertension in any patient. One patient was found to have a horseshoe kidney with multiple renal arteries arising from the anterior aspect of the aorta. The descending thoracic aorta was evaluated arteriographically in only three instances and found to be normal in all cases. During surgery, a bypass to both lower extremities was constructed in 12 patients, and a unilateral bypass was constructed in four. The eighth intercostal space was the most common approach to the thorax. An additional left flank incision for guidance of the graft through the retroperitoneal runnel was used in 12 patients, and was deemed unnecessary in four. Dacron grafts were used in all instances. Average blood loss was 946 ml (range, 350 to 2000 ml), and average operative time was 4 hours and 10 minutes. Partial aortic occlusion clamp time for construction of the proximal anastomosis was recorded in 11 patients and ranged from 15 to 65 minutes. Perioperative homologous blood transfusion averaged 2.8 units and ranged from none to eight units. Other details of surgery are given in Table I. After operation 14 patients were extubated in the operating room or within 24 hours of surgery. The average intensive care unit stay was 4 days; however, 12 patients stayed for 3 days or less. Ten patients

553

resumed oral intake within 3 days of surgery. The average length of hospital stay was 10 days. Two patients died in the hospital (i2.5% operative mortality rate), one of a severe intraoperative myocardial infarction with hypotension that resulted in multiorgan failure, and the other of respiratory failure followed by massive intestinal infarction. Five other complications occurred: one patient had a left lung collapse requiting bronchoscopy but not reintubation, one patient with multiple previous right hemispheric strokes had a recurrent stroke from which she recovered well, two patients had an infected seroma in the groin, one of them required surgical drainage but both resolved without graft compromise, and one developed a wound seroma (Table I). The hemodynamic results of the bypass were excellent, as reflected by a significant increase (p < 0.0001) in the mean ABI from 0.18 (SD = 0.19) before operation to 0.84 (SD = 0.19) after operation. Twenty-one limbs had an improvement in the ABI of at least 0.5 and seven of less than 0.5. Only one limb had a not significant increase in the ABI (< 0.15). Mean follow-up of these patients was 24 months, ranging from 3 to 78 months. During the follow-up period there was no documentation of any hemodynamic graft failure; two grafts, however, had to be removed because of infection, one patient died, and two were lost to follow-up. The 18-month cumulative primary patency rate was 100%, and at 2 years it was 83.3% (Table II). Our experience combined with that accumulated in the English-language literature totaled 141 cases in which the descending thoracic aorta was used as an inflow source for aortoiliac reconstruction (Table III). Branchereau et al.~4 recently overestimated the accumulated experience by considering 20 patients from two series that were included in a larger one (personal communication, R. D. Schultz, MD, February 1991). The overall operative mortality rate was 6.4%. The most common indication for the operation was previous abdominal aortic graft failure (40.4%), followed by aortic graft infection (15.6%). Other indications for surgery were multiple previous abdominal operations, previous abdominal radiation therapy, presence of abdominal stomas or previous sepsis, and unsuitability of the abdominal aorta as inflow source. The characteristics of these series are described in Table III. One hundred twenty-four patients were monitored from 1 month to 9 years. With use of the available follow-up data, the cumulative patency was obtained by life-table analysis. The overall cumulative

554

lournalof VASCULAR SURGERY

Criado et al.

Table II. Life-table analysis of cumulative patency data of 124 descending thoracic aorta to iliofemoral artery bypasses Grafts lost to observation Death

Withdrawn *

Grafts at risk

Grafts failing

Interval (mo)

UNC

All

UNC

All

UNC

All

UNC

All

UNC

All

UNC

All

0-1 1-3 3-6 6-12 12-18 18-24 24-30 30-36 36-42 42-48 48-54 54-60 60-66 66-72 72-78 78-84 84-90 90-96 96-102 102-108

14 14 12 11 11 6 5 3 2 1 1 I I 1 1 -

124 123 113 105 75 64 47 38 31 24 21 i6 i3 12 10 9 7 6 6 6

0 0 0 0 0 1 1 0 0 0 0 0 0 0 0

0 1 I 0 3 4 2 2 i 0 1 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 1 0 0 0 0 0 0 0

1 1 1 1 3 2 3 2 0 1 1 i 0 0 0 0 0 0 0 1

0 2 1 0 5 0 I 0 1 0 0 0 0 0 1

0 8 6 29 5 11 4 3 6 2 3 2 1 2 1 2 1 0 0 0

100 100 100 100 100 83.3 80 100 100 100 100 100 100 100 100 -

100 99.2 99.1 100 96 93.7 95.7 94.7 96.8 100 95.2 100 I00 I00 100 100 100 100 100 100

100 100 100 100 100 83.3 63.3 63.3 63.3 63.3 63.3 63.3 63.3 63.3 63.3 -

100 99.2 98.3 98.3 94.3 88 83.7 78.4 75.2 75.2 70.4 70.4 70.4 70.4 70.4 70.4 70.4 70.4 70.4 70.4

-

-

-

Interval patency rate (%)

Cumulative patency rate (%)

U N C , G r a f t s f r o m U n i v e r s i t y o f N o r t h C a r o l i n a experience. *Withdrawn because of duration of follow-up.

patency rate was 98.3% at 1 year, 88% at 2, 78.4% at 3, and 70.4% at 5 years (Table II). DISCUSSION

Infrarenal aortobifemoral bypass is currently considered the "gold standard" procedure for aortoiliac reconstruction in occlusive disease. The cumulative patency rates in large series of aortofemoral and aortoiliac bypass range from 83% to 92% at 5 years; 71% to 83% at 10 years; and 61% to 74% at i5 years. 2628 Late aortofemoral bypass occlusion, however, develops in 6% to 13% of patients after primary abdominal aortic reconstruction. 26'29'3° Approximately 45% of patients with aortoiliac occlusive disease have associated distal occlusive disease at the femoral, popliteal, or tibial vessels. Progressive atherosclerotic involvement of the outflow and/or inflow tract, therefore, is the most common cause of late graft occlusion. 31 Graft infection is also a cause of late graft failure in a small number of aortic reconstructions for occlusive or aneurysmal disease. Reoperative abdominal aortic surgery, therefore, is not infrequently required for relief of recurrent ischemic symptoms or limb salvage after abdominal aortic graft failure. The surgical options for aortoiliac reconstruction after infrarenal aortic graft failure include repeat abdominal aortic grafting, graft

thrombectomy and profundaplasty, subcutaneous extraanatomic bypass, bypass from the ascending aorta to the iliofemoral vessels, and thoracofemoral bypass originating in the descending thoracic aorta. Remedial aortoiliac reconstruction for graft occlusion or infection, however, is a difficult task; it carries a significant morbidity, 31 and the long-term results are generally worse than those of the original procedure. Bernhard et al.32 reported a 75% 3-year cumulative patency rate after aortofemoral graft revision with use of thrombectomy and profundaplasty, aortic graft replacement, or extraanatomic bypass. Fulenwider et al., 33 however, reported a 14% amputation rate at i year after repeat aortobifemoral bypass in patients with an occluded aortofemoral or aortoiliac bypass. Crawford et al.30 in a series of 64 aortofemoral and aortoiliac bypasses performed for previous graft failure, had a 14% operative mortality rate, and a 6% early amputation rate. The ascending aorta for remedial aortoiliac reconstruction has been sporadically used in those situations where the entire aorta distal to the left subclavian artery and the axillary arteries are unsuitable as inflow source. 3436 This operation, however, requires a laparotomy, a median sternotomy, and groin incisions when appropriate; therefore it is a procedure of much greater magnitude than the thoracofemoral bypass.

Volume 15 Number 3 March 1992

Descending thoracic aorta to femoral artery bypass

555

Table I I L Collective experience with descending thoracic aorta to iliofemoral artery bypass Indication First author Stevenson 1 B l a i s d e l l2 Robicsek 3 Reichle4 Nurm ~ Froysaker 6 Jarret 7 Cevese 8 Buxton 9 L a k n e r 19 R e i l l y 2° E n o n 21 F e l d h a u s 1° H a a s 22 D e L a u r e n t i s 12 " M c C a r t h y 13 H u s s a i n 23 S c h e l l a c k 24 B r a d h a m zs Bowes H B r a n c h e r e a u ~4 P r e s e n t series Total

Year

No. of patients

1961 1961 1967 1970 1972 1973 1975 1975

1 1 1 1 3 6 2 6 1 2 5 3 18 3 10 13 8 3 2 26 10 16 141

1976 1983 1984 1985 1985 1985 1986 1986 1988 1988 1989 1990 1991 1991

Aortic graft failure 1

Aortic graft infection

Other

-

-

-

1 1 1 1

1 -

1 6

2 1

12 -8 5 2 3 9 9 6 57 (40.4%)

When remedial abdominal aortic surgery is undesirable because of severe disease at the inflow site, graft infection, multiple previous abdominal operations, or abdominal sepsis, axillofemoral bypass is currently the most common alternative procedure. The results of axillofemoral bypass, however, are far from ideal; it carries an operative mortality rate ranging from 2% to 13%, and the 3-year cumulative primary patency rate ranges from 40% to 54%, a~~s and 19% to 47% at 5 years. 16'37 The thoracofemoral bypass has major advantages over the axillofemoral because it provides better inflow, requires a shorter graft length, offers better protection of the graft from infection and mechanical trauma, and carries a superior patency rate without increased morbidity or mortality rates. The overall results of our collective experience with thoracofemoral bypass (Tables II and III) in a group of patients with a large proportion of multiple previous aortic graft failure (56%), compare favorably to the patency rates and morbidity and mortality rates of reoperative aortoiliac reconstruction by use of repeat aortofemoral or aortoiliac and axillofemoral bypass presented above. Furthermore, in our experience thoracofemoral bypass produces a uniform increase in the ABIs, which as shown by Garrett et al.3a is a good predictor of improvement of ischemic symptoms. Complete aortic occlusion occurs in approxi-

-

6 2

5

-

-

3 3 3 2 1 6

3 7 -

1

22 (15.6%)

2 16 1 10 62 (43.9%)

Operative deaths 0 1 0 0 0 0 2 0 0 0 0 0 1 1 0 0 0 0 0 1 1 2 9 (6.4%)

Mean length offoliow-up (mo) 12 6 54 21 28 24 15 55 40 16 44 22 36 21 53 14 24

mately 8% of patients admitted with chronic aortoiliac disease, 39 and juxtarenal aortic occlusion represents an advanced stage of aortoiliac occlusion2° Its surgical management requires adequate exposure of the suprarenal aorta; classically a transabdominal approach with thromboendarterectomy and bypass graft from the infrarenal aorta to the femoral or iliac arteries has been used for this purpose. 41 More recently, Saner and Stoney42 have recommended transabdominal exposure of the pararenal aorta using medial visceral rotation. Nonetheless, the direct aortic approach for juxtarenal occlusion constitutes a surgical challenge. Starrett and Stoney43 reported the onset of renal failure in six of a group of 13 patients with juxtarenal aortic occlusion who were monitored without operation, which occurred from 1 5 clays to 17 years after diagnosis. Therefore, the risk of subsequent renal artery thrombosis is a concern with the use of thoracofemoral bypass in patients with juxtarenal aortic occlusion. In our experience and that found in our literature review, however, only one case of renal failure as a result of proximal progression of the aortic thrombus was documented, n Furthermore, Cevese and Gallucci 8 using radioisotope renography prospectively confirmed normal renal perfusion up to 5 years after thoracofemoral bypass in a group of six patients with complete juxtarenal aortic occlusion. Nevertheless, the long-term visceral ischemic complications in patients with juxtarenal aortic

556

Criado et al.

Table IV. Advantages of thoracic aorta-to-iliofemoral bypass Optimal inflow Favorable hemodynamics High patency rate Avoids diseased aorta Prevents aortoenteric fistula Avoids aortic cross-clamping Reduces risk of operative embolization Minimizes splanchnic ischemia Avoids transabdominal route Negligible risk of spinal cord ischemia Excellent graft protection Avoids sexual dysfunction

Table V. Indications for thoracic aorta to iliofemoral bypass Previous abdominal aortic procedures Unsuitable infrarenal aorta Failed axillofemoral bypass Unsuitable axillary arteries Abdominal aortic graft infection Abdominal septic source Abdominal stomas Abdominal radiation therapy Multiple abdominal operations

occlusion undergoing thoracofemoral bypass remain uncertain. Thoracofemoral bypass features a unique combination of characteristics that make it most suitable for reoperative aortoiliac reconstruction or for those situations in which the abdominal approach is not advisable (Table IV). It allows the use of an invariably excellent inflow source, and it avoids dissection in scarred or potentially infected areas, minimizes the risk of embolization by using an aortic segment that is uniformly free of severe atherosclerosis, eliminates the threat of aortoenteric fistula by placing the anastomosis and body of the graft remote to the abdominal contents, and decreases the morbidity of aortic cross-clamping by using partial aortic occlusion. Furthermore, end-to-side anastomosis placed at the descending thoracic aorta level forms an antegrade acute angle that is hemodynamically advantageous. It is advisable, however, to perform a preoperative CT scan or aortography to ensure the adequacy of the thoracic aorta for proximal anastomosis. Thoracofemoral bypass can be kept as a fairly benign operation by observing several technical factors important in minimizing systemic and organspecific repercussions of surgery. First, a limited anterolateral thoracotomy suffices for excellent ap-

Journal of VASCULAR SURGERY

proach to the descending thoracic aorta; the use of a thoracoabdominal incision as advocated by others is unnecessary and increases the magnitude and potential morbidity of the operation. 5,13,24 Similarly, a formal posterolateral thoracotomy is not required. The proximal anastomosis should be done under partial aortic occlusion; aortic cross-clamping is not needed, and its use introduces the risk of spinal cord ischemia as reported by Branchereau et al. 14 In summary, the thoracofemoral bypass is a superb alternative operation for aortoiliac reconstruction. Technically it is a rather simple operation that provides a durable graft with reasonable morbidity and mortality rates. Its main indications (Table V) are those situations in which the use of the abdominal aorta is contraindicated, or predictably difficult or~ hazardous. In patients with severely diseased infrarenal aorta at the predicted anastomotic site consideration should be given to thoracofemoral bypass as an excellent choice for primary aortic reconstruction. REFERENCES 1. Stevenson JK, Sauvage LR, Harkins HN. A bypass homograft from thoracic aorta to femoral arteries for occlusive vascular disease. Am Surg 1961;27:632-7. 2. Blaisdell FW, DeMattei GA, Gauder PJ. Extraperitoneal thoracic aorta to femoral bypass graft as replacement for an infected aortic bifurcation prosthesis. Am J Surg 1961;102: 583-5. 3. Robicsek F, McCall M~VI, Sanger PW, Daugherty HK. Recurrent aneurysm of the abdominal aorta: insertion of a vascular prosthesis from the distal aortic arch to the femoral arteries. Ann Thorac Surg 1967;3:549-52. 4. Reichle FA, Tyson RR, Soloff LA, Lautsch EV, Rosemond GP. Salmonellosis and aneurysm of the distal abdominal aorta: case report with a review. Ann Surg 1970;171:219-28. 5. Nunn DB, Kamal MA. Bypass grafting from the thoracic aorta to femoral arteries for high aortoillac occlusive disease. Surgery 1972;72:749-55. 6. Froysaker T, Skagseth E, Dundas P, Hall KV. Bypass procedures in the treatment of obstructions of the abdominal aorta. J Cardiovasc Surg 1973;14:317-21. 7. Jarretr F, Darling RC, Mundth ED, Austen WG. Experience with infected aneurysms of the abdominal aorta. Arch Surg 1975;110:I281-6. 8. Cevese PG, Gallucci V. Thoracic aorta-to-femoral artery bypass. J Cardiovasc Surg 1975;16:432-8. 9. Buxton B, Simpson L, Johnson N, Myers K. Descending thoracic aortofemoral bypass for distal aortic reconstruction after removal of an infected Dacron prosthesis. Med J Aust 1976;2:133-6. 10. Feldhaus RJ, Sterpetti AV, Schultz R.D, Peetz DJ Jr. Thoracic aorta-femoral artery bypass: indications, technique, and late results. Ann Thorac Surg 1985;40:588-92. 11. Bowes DE, Youkey JR, Pharr WP, Goldstein AM, Benoit CH. Long-term follow-up of descending thoracic aorto-iliac/ femoral bypass. J Cardiovasc Surg 1990;31:430-7. 12. DeLaurentis DA. The descending thoracic aorta in reoperatire aortic surgery. In: Bergan JJ, Yao JST, eds. Reoperative

Volume 15 Number 3 March 1992

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

26.

27. 28.

arterial surgery. Orlando: Grime and Strarton, 1986:195203. McCarthy WJ, Rubin [IR, Flinn WR, Williams LR, Bergan JJ, Yao JST. Descending thoracic aorta-to-femoral artery bypass. Arch Surg 1986;121:681-8. Branchereau A, Espinoza H, Rudondy P, Magnan P-E, Reboul J. Descending thoracic aorta as an inflow source for late occlusive failures following aortoiliac reconstruction. Ann Vasc Surg 1991;5:8-15. Donaldson MC, Louras JC, Bucknam CA. Axillofemoral bypass: a tool with a limited role. J VASCSUING1986;3:75763. Eugene J, Goldstone J, Moore WS. Fifteen-year experience with subcutaneous bypass grafts for lower extremity ischemia. Ann Surg 1977;186:177-83. Ascer E, Veith FJ, Gupta SK, et al. Comparison of axillounifemoral and axillobifemoral bypass operations. Surgery 1985;97:169-75. Ward RE, Holcroft JW, Conti S, Blaisdell FW. New concepts in the use of axillofemoral bypass grafts. Arch Surg 1983; 118:573-6. Lakner G, Lukacs L. High aortoiliac occlusion: treatment with thoracic aorta to femoral arterial bypass. J Cardiovasc Surg 1983;24:532-4. Reilly LM, Ehrenfeld WK, Stoney RJ. Delayed aortic prosthetic reconstruction after removal of an infected graft. Am J Surg 1984;148:234-9. Enon B, Chevalier JM, Moreau P, Lescalie F, Pillet J. Revascularisation des membres inf6rieurs ~ partir de l'aorte thoracique descendante. [i Chir (Paris) 1985;122:539-43. Haas KL, Moulder PV, Kerstein MD. Use of thoracic aortobifemoral artery bypass grafting as an alternative procedure for occlusive aortoiliac disease. Am Surg 1985;51:573-6. Hussain SA. Descending thoracic aorta to bifemoral by-pass graft without laparotomy. Int Surg 1988;73:263-6. Schellack If, Fulenwider lIT, Smith RB III. Descending thoracic aortofemoral-femoral bypass: a remedial alternative for the failed aortobifemoral bypass. J Cardiovasc Surg 1988;29:201-4. Bradham RR, Locldair PR Jr, Grimball A. Descending thoracic aorta to femoral artery bypass. J S C Med Assoc 1989;85:283-6. Szilagyi DE, Elliott JP, Smith RF, Reddy D[I, McPharlin M. A thirty-year survey of the reconstructive surgical treatment of aortoiliac occlusive disease. [i VAsc SUV,G 1986;3:42136. Brewster DC, Darling RC. Optimal methods of aortoiliac reconstruction. Surgery 1978;84:739-48. Crawford ES, Bomberger RA, Glaeser DH, Saleh SA, Russell WL. Aortoiliac occlusive disease: factors influencing survival

Descending thoracic aorta to femoral artery bypass 5 5 7

29.

30.

31.

32. 33.

34.

35.

36. 37. 38.

39.

40. 41.

42.

43.

and function following reconstructive operation over a twenty-five-year period. Surgery 1981;90:1055-67. Thompson WM, Johnsrud IS, Jackson DC, Older RA, Wechsler AS. Late complications of abdominal aortic reconstructive surgery: roentgen evaluation. Ann Surg 1977;185: 326-34. Crawford S, Manning LG, Kelly TF. "Redo" surgery after operations for aneurysm and occlusion of the abdominal aorta. Surgery 1977;81:41-52. Robbs JV, Wylie EJ. Factors contributing to recurrent lower limb ischemia following bypass surgery for aortoiliac occlusive disease, and their management. Arch Surg 1981;193: 346-52. Bernhard VM, Ray LI, Towne JB. The reoperation of choice for aortofemoral graft occlusion. Surgery 1977;82:867-74. Fulenwider JT, Smith RB III, Johnson RW, Johnson RC, Salam AA, Perdue GD. Reoperative abdominal arterial surgery-a ten-year experience. Surgery 1983;93:20-7. Frantz SL, Kaplitt MJ, Beil AR [ir, Stein HL. Ascending aorta-bilateral femoral artery bypass for the totally occluded infrarenal abdominal aorta. Surgery 1974;75:471-5. Wukasch DC, Cooley DA, Sandiford FM, Nappi G, Reul GJ Jr. Ascending aorta-abdominal aorta bypass: indications, technique, and report of 12 patients. Ann Thorac Surg 1977;23:442-8. Gelfand ET, Callaghan JC, Sterns LP. Extended aortic bypass. J Thorac Cardiovasc Surg 1980;79:381-7. Rutherford RB, Patt A, Pearce WH. Extraanatomic bypass: a closer view. J VASC SUV,G 1987;6:437-46. Garrett WV, Slaymaker EE, Heintz SE, Barnes RW. Intraoperative prediction of symptomatic resuk of aortofemoral bypass from changes in ankle pressure index. Surgery 1977; 82:504-9. Corson JD, Brewster DC, Darling RC. The surgical management of infrarenal aortic occlusion. Surg Gynecol Obstet 1982;155:369-72. Bergan [i[i, Trippel OH. Management of juxtarenal aortic occlusions. Arch Surg I963;87:230-8. Liddicoat JE, Bekassy SM, Dang MH, De Bakey ME. Complete occlusion of the infrarenal abdominal aorta: management and results in 64 patients. Surgery 1975;77:467-72. Saner L, Stoney RJ. Management ofjuxtarenal aortic occlusive disease by transabdominal exposure of the pararenal and suprarenal aorta by medial visceral rotation. In: Ernst CB, Stanley [IC, eds. Current therapy in vascular surgery. Philadelphia: BC Decker, 1991:406-9. Starrert RW, Stoney RJ. Juxtarenal aortic occlusion. Surgery 1974;76:890-7.

Submitted May 31, 1991; accepted July 25, 1991.