A comparative evaluation of externally supported polytetrafluoroethylene axillobifemoral and axillounifemoral bypass grafts

A comparative evaluation of externally supported polytetrafluoroethylene axillobifemoral and axillounifemoral bypass grafts C h i t t u r R. M o h a n , M D , William J. Sharp, M D , Jamal J. Hoballah, M D , T i m o t h y F. Kresowik, M D , Michael T. Schueppert, M D , and J o h n D. Corson, M B , ChB, Iowa City, Iowa

Purpose: We analyzed a current 78-month e~perience with externally supported (ringed) polytetrafluoroethylene (PTFE) axillobifemoral (AxBF) and axillounifemoral (AxUF) bypass grafts to address the controversy about whether the addition o f a femorofemoral limb to an axillofemoral bypass graft improves the patency results. &I~thods: Between lanuary 1988 and June 1994, 36 AxBF and 22 AxUF externally supported PTFE ringed bypass grafts were performed at our institution. The age of the patients in the AxBF group was 67 +- 11 years and 69 -+ 11 years in the AxUF group. The male/female ratio was 22:13 (AxBF) and 8 - 9 (A.xUF). In 71% of cases (29/36 AxBF, 12122 AxUF), the operations were performed for aortoiliac atherosclerotic occlusive disease in patients with significant medical risk factors or a "hostile" abdomen. The remaining 29% were patients requiring revascularization during treatment of an infected aortic graft. Bypass patency was assessed in the follow-up period by clinical evaluation, color-flow duplex imaging, or segmental limb pressure measurements. Results: There was no significant difference in the 30-day operative mortality rate for all AxBF bypasses (11%) and all AxUF bypasses (6%) (p = 0.89 by chi-squared testing). The primary and secondary patency rates for the whole group of bypasses were 80% and 89% at 3 years, respectively (SE < 0.1). Between the AxBF and AxUF groups, there were no significant differences in either primary patency (80% for each group) or secondary patency (91% in AxBFs vs 85 0~ in AxUFs) (SE < 0.1) at 2 years (Wilcoxon rank sum test). Conclusions: These data show no differences in the patency of externally supported PTFE AxBF and AxUF bypass grafts up to 2 years after implantation. (J VASC SURG 1995; 21:801-9.)

In 1961, Lewis! reported the use of the subclavian artery as an inflo w source for lower extremity revascularization, T h i s unique:COnCept was further modified by Originating ~ e i ~ o w frOm the axillary artery and isubcut~eously routing ~ e bypass to the femoral vesselS: The ~ l l o f e m o r a l bypass was first performed in 1962 b y Biaisdell From the Section of Vascu!ar Surgery, University of Iowa Hospitals & Clinics, Iowa City. Presented at the Eighteenth Annual Meeting Of the Midwestern Vascular SurgicalSociety,Cincinnati, Ohio, Sept. 23-24, 1994. Reprint requestsi lohn D. Corson,MB, ChB, Sectionof Vascular Surgery, Universityof Iowa Itospitals & Clinics, 200 Hawkins Dr,, Iowa City, IA 52242. Copyright 9 1995 by The Society for Vascular Surgery and International Societyfor CardiovascularSurgery,North American Chapter. 0741-52~4/95/$3.00 + 0 24/6/62882

and HalIZ in the United

States and in 1963 bY Louw 3 in SoUth Africa~ The further addition o f a femorOfemoral crossover fimb as initially Suggested by Sauvage and W o o d 4 offered a simple o~tion ifor reVasc~arization o f both lower ex~emitiesi ~ l l 6 ~ i f e m o r a l (AxUF) or axiUobifemoral ( ~ F ) bypass procedures have been Used mainly for fimb Salvage in patients at high risk w i t h aortoiliac occlusive disease or similar patients With a::~h0stiie ', a b d o m e n or for revascularization o f patients requiring removal o f a functioning infected aortiC: graft.4! ~ Controversy remains regarNng whether the addition o f a femorofemoral limb will i m p r o v e late patency results. 7,913 T o address this issue, we evaluated a current 78-month experience with externally supported polytetrafluoroethylene (PTFE) AxBF and AxUF bypass grafts. 801

802

JOURNAL OF VASCULAR SURGERY May 1995

fl/lohan et al.

Table I. Patient demographics and indications for the axillofemoral grafts AxBF bypass

AxUF bypass

36 35 67 + 11 22 : 13 36% 64% 29 (80%) 12 (33%) 1 (3%) 3 (8%) 13 (36%) 7 (20%) 6 (17%) 1 (3%)

22 17 69 -+ 11 8 :9 41% 59% 12 (55%) 9 (41%) 0 (0%) 0 (0%) 1 (5%) 10 (45%) 10 (45%) 0 (0%)

No. of procedures No. of patients Age (yr) Male/female Emergency procedures Elective procedures Atherosclerotic disease Acute ischemia Gangrene Nonhealing ulcer Rest pain Infection Aortic graft infection Aortoduodenal fistula

p Values

p p p p p

= = = = =

0.54* 0.43t 0.931t 0.931t 0.07t

p = 0.07t

*By t testing. tBy chi-squared testing.

Table II. Associated medical risk factors Condition

AxBF

AxUF

Hypertension Diabetes meUitus Coronary artery disease COPD Chronic kidney failure Previous aortic surgery Age >75 years

51% 37% 57% 34% 9% 17% 23%

65% 22% 59% 24% 12% 29% 35%

p Value* p p p p p p p

= = = = = = =

0.544 0.505 0.854 0.640 0.893 0.513 0.528

COPD, Chronic obstructive pulmonary disease. *By chi-squared testing.

PATIENTS A N D METHODS We evaluated the results of all the externally supported (ringed) PTFE AxBF and AxUF bypasses done between January 1988 and June 1994 at the University of Iowa Hospitals & Clinics. The medical records of all of these patients were reviewed. Information gathered from the medical records included patient demographics, associated medical illnesses, indication for surgery, and prior and subsequent vascular procedures. Relevant preoperative diagnostic study results such as segmental limb pressures, duplex study results, and arteriograms were reviewed. The patients' records were also reviewed for the type and diameter of graft, bypass configuration, outflow artery, runoff status, and perioperative and late complications. Information regarding the postoperative use of aspirin or warfarin (Coumadin) was also gathered from the medical records. Patients were evaluated in the vascular surgery clinic at i month after the procedure, then at 6-month intervals for the first year, and yearly thereafter. A similar protocol was followed if a patient underwent a bypass revision or new bypass

procedure. In addition to a clinical evaluation, segmental limb pressure measurements or color-flow duplex imaging of the bypass graft was done during each clinic visit. Brachial systolic blood pressures in both arms were measured before operation. If both pressures were normal, the inflow site chosen was usually on the side of the most ischemic lower limb. If there was evidence of unilateral inflow axillary artery disease the normal contralateral axillary artery was used as the inflow artery. Arteriographic studies were not routinely used to evaluate the inflow, but duplex scanning was occasionally used for this purpose. Each axillary anastomosis was to the anterior aspect of the first part of the axillary artery. The vertical limb of the bypass was tunneled along the posterior axillary line and a counterincision was avoided, if possible. The femorofemoral component of the AxBF graft originated from the lower end of the vertical limb of the bypass and was tunneled subcutaneously to the native artery in the contralateral groin by use of an inverted C configuration in 28 bypasses and in a lazy g configuration in eight bypasses.

IOURNAL OF VASCULAR SURGERY Volume 21, Number 5

The CUlnulative primary and secondary parency rates for the bypass grafts were calculated by life-table analysis as per the recommendations of the Society for Vascular Surgery/International Society for Cardiovascular Surgery Ad Hoc Committee o n Suggested Standards in Reporting of Lower Extremity Bypasses. 14 Because a patient with bilateral AxUF bypasses has two separate bypass grafts, the patency results were calculated and analyzed per graft and not per patient to compare the outcome of all AxUF and AxBF bypasses. In the case of AxBF bypasses, a graft was considered patent only when both the axillofemoral and the femorofemor{tl components were patent. Patient demographics between the AxBF and AxUF bypass groups were compared for any statistical differences b y r testing. Both the primary and secondary patency rates between AxBF and AxUF bypasses were analyzed for statistical differences by the Wilcoxon rank stun test. Other variables such as indications for surgery and theass0ciated medical risk factors between AXBF and AxUF bypass groups were compared by use of chi-square testing. A p value less than 0.05 was considered significant. RESULTS Thirty-six AxBF and 22 AxUF externally supported (ringed) PTFE bypass grafts were done at the University of Iowa Hospitals & Clinics from January 1988 to June i994. Patient demographics and the indications for the bypasses are depicted in Table I. Five patients (three inthe AxBF group and two in the AxUF group) t!ad a prior AxBF that had failed. The secondary axillofemoral bypass grafts originated from the opposite axiUary artery. The medical risk factors for both the groups are compared in Table II. Twenty-two patients required emergency revascularization. In the AxBF group, both lower extremity superficial femoral arteries were patent in 19 cases (53%), only one superficial femoral artery was patent in 10 cases (28%), and both were occluded in four cases (11%). The status o f the superficial femoral arteries was not knownl before three emergency procedures (8%) because arteriography was not performed. In the AxUF bypass group, the ipsilateral superficial femoral artery was patent in 14 cases (64%), occluded in four cases (18%), and of unknown status in four emergency cases (18%). In both the AxBF and AxUF bypass groups, all limbs with a patent femoropopliteal/tibial bypass were listed as having a patent superficial femoral artery. All the bypasses in this series were performed with use of PTFE grafts with externally supported rings.

3/lohan et al.

803

The graft diameter in the vertical limb of 33 AxBF bypasses was 8 mm and in the other three was 6 mm. An 8 mm-diameter graft was used in 16 AxUF bypasses and a 6 mm diameter graft in the remaining six.

There were four deaths in the AxBF group and one death in the AxUF group during the 30-day postoperative period, resulting in an overall (both elective and emergency procedures) operative mortality rate of 11% for the/LxBF group and 6% for the AxUF group (p = 0 . 8 9 b y chi-squared testing). In the AxBF group three deaths occurred after emergency revascularization procedures, and one occurred after an elective procedure. The causes of death in the three patients who died after emergency AxBF bypass grafting Were septic shock (n - 1) and multisystem organ failure (n = 2). The death after an elective AxBF bypass was due to a postoperative myocardial infarction. One patient in the AxUF group died 1 day after emergency revascularization as a result of a myocardial infarction. Thus, for elective revascularization procedures, t h e 30-day operative mortality rate for AxBF and AxUF bypass grafts was 4% and 0%, respectively. During the postoperanve period one patient in the AxBF group required a femoropopliteal bypass on day l0 for distal ischemia of the lower extremity despite the patent AxBF bypass. Two patients with AxBF bypasses underwent evacuation of a hematoma in the axillary region on postoperative day 2 and 10, respectively. The latter patmnt was receiving Coumadin. One patient with an AxUF bypass underwent exploration and evacuation of an axillary hematoma on postoperative day 2. This patient was not receiving anncoagulation. One patient with rest pain of the left leg underwent an AxBF bypass and 13 months later, despite a patent bypass, underwent a right below-knee amputation for a chronic infected venous stasis ulcer with osteomyelitis. Another patient who had a thrombosed AxUF bypass underwent an above-knee amputation on postoperative day 1 because of progressive advanced ischemia. There was no significant difference in the amputation rates between the AxBF and AxUF bypass groups (p = 0.69 by chi-squared testing). Disruption at an axillary and femoral anastomosis (one each) was seen in two patients with an AxBF graft. The patient with a proximal anastomotic disruption had development of this 2 weeks after operation, possibly as a result of forced exertion of the upper limb. The patient with a femoral anastomotic disruption had development of this 3 weeks

804

JOURNAL OF VASCULARSURGERY May 1995

A/lohan et al.

after the surgical procedure. In this case the suture line was intact, but the PTFE graft at the distal anastomosis was torn. Both these disruptions were revised with an interposition PTFE graft to bridge the defect. The former patient underwent a successful thrombectomy for graft thrombosis 6 months later that subsequently occluded, requiring a new AxUF bypass with inflow from the contralateral axillary artery. The latter patient with the femoral anastomotic disruption has a patent graft at follow-up 12 months later. Five of the AxBF bypass grafts thrombosed during the follow-up period (Table III). They were all 8 mm-diameter grafts. Three thromboses occurred within 24 hours (perioperative thrombosis), and the other two were late thromboses at 18 and 31 months after graft implantation. All the thromboses in the AxBF group involved the vertical limb of the bypass, as well as the femorofemoral limb except in one patient who had thrombosis of the femorofemoral component alone detected by duplex examination 1 month after the surgery. Because the patient did not have any symptoms, he did not undergo further intervention for the thrombosed crossover limb. One patient with an AxUF bypass graft had development of a perioperative graft thrombosis (< 24 hours) and another patient had development of a graft thrombosis at 3 weeks. Two other patients had development of late graft thromboses (4 and 9 months later). All four graft thromboses in the AxUF bypass group occurred in patients with 8 mm diameter grafts, and none in 6 mm diameter grafts. There was no statistically significant difference between the thrombosis rate in either the AxBF and AxUF groups with regard to whether the superficial femoral arteries were patent or occluded (/9 = 0.38 and 0.80 by chi-squared testing). Similarly, with both groups combined, there was no significant difference between the rate of thrombosis whether the superficial femoral artery was patent or occluded (p = 0.31 by chi-squared testing). Postoperative anticoagulation therapy alone with Coumadin was instituted, in 11% and 36% of the cases in the AxBF and AxUF groups, respectively. Similarly, 19% of the cases in the AxBF group and 5% of the cases in the AxUF group were on aspirin therapy alone after the bypass procedure. In the AxBF group 36% of the cases were on both Coumadin and aspirin therapy compared with 18% of the cases in the AxUF group. Thirty-three percent of AxBF and 41% of AxUF cases did not receive either Coumadin or aspirin (p = 0.76 by chi-squared testing). The effect of C0umadin or aspirin therapy

on the patency of the AxBF and AxUF bypasses is depicted in Table IV. Because there were only small numbers of grafts in each of these groups, a comparison of the patency of the AxBF and AxUF bypass in each category was not done. The patency of the grafts in both the AxBF and AxUF group with regard to the status of the superficial femoral artery and the Coumadin therapy is depicted in Table V. For all the bypasses combined together the cumulative primary and secondary patency rate by life-table analysis was 80% and 89% at 3 years (SE < 0.1) (Fig. 1). The primary and secondary patency rates of AxBF and AxUF bypasses are depicted in Figs. 2 and 3. When the two groups were compared, it was found that there was no significant difference in the primary patency rate at 2 years (80% for each group, SE < 0.1) nor any significant difference between secondary patency rates in the two groups (91% for AxBF and 85% for AxUF) at 2 years

(SE < 0.1). DISCUSSION

Controversy still exists regarding whether the addition of a femorofemoral limb improves the patency of an axillofemoral bypass graft. In contradistinction to the positive reports ofLoGerfo et al.,15 Rutherford et al., 1~ Kalman et al., 16 Burrell et al., 7 and Hepp et al.17 that favor the use of an AxBF over an AxUF graft, Ascer et al.13 demonstrated no difference in patency rates between such grafts. In the report by Ascer et al. 13nonringed AxBF PTFE bypass grafts had a 5-year cumulative primary rate and secondary patency rate of 50% and 77%, which were not significantly different from the patency rates of AxUF bypasses (44% and 71%). The 2-year primary patency rate of AxBF bypasses in our series was 80% and is comparable to the 2-year data for similar bypasses of 70% to 90% reported by LoGerfo et al.,i5 Rutherford et al.]4 Kalman et al.,16 Burrell et al.,7 and Hepp et al.17 However, the 2-year primary patency rate of AxUF bypasses in our study was markedly superior to the 2 year results of earlier reports (80% vs 30% to 60%). LoGerfo et al. 15 suggested that the improved patency of their AxBF bypass grafts was due to increased blood flow through the vertical limb of the axillofemoral bypass proximal to the take off of the crossover femoral limb. Several reports from other centers had confirmed the findings of LoGerfo et al. that AxBF grafts had patency rates of 71% to 78% at 3 years compared with 37% to 58% for AxUF bypasses. 7,i~ Ray et al.,t2 however, reported contradictory findings in their series of USGI Sauvage filamentous velour

JOURNAL OF VASCULAR SURGERY Volume 21, Number 5

214ohan et al.

805

Table III. AxBF and AxUF bypass graft thrombosis: cause and outcome No.

Occurrence

AxBF graft thrombosis 1 < 24 hours

Cause

Management

2

< 24 hours

Platelet plus at distal anastomosis Thrombosis of profunda femoris

3

< 24 hours

Hypotension + CVA

Thrombectomy, aspirin, Coumadin Thrombectomy, jump graft to distal profunda femoris, Coumadin No intervention

Subtherapeutic anticoagulation CVA + seizures

Thrombectomy, Coumadin No intervention

Recurrent thrombosis of femoral artery

Above-knee amputation for continued ischemia New AxUF (unsuccessful thrombectomy) Successful thrombectomy at 4, 6, and 17

4

18 months

5

31 months

AxUF graft thrombosis 1 < 24 hours 2

3 weeks

3

4 months

4

9 months

Recurrent thrombosis of femoral artery Unknown

Current status Patent at 11 months Patent at 21 months

Died because of CVA 2 months later Patent at 36 months Died because of CVA 4 days later Alive Patent Patent

mos~

Thrombosis of fem-pop bypass

Ilioprofunda bypass

Patent at 21 months

CVA, Cerebrovascular accident; fern-pop, femoropopliteal.

Table IV. Effect of aspirin or Coumadin on patency of grafts

Aspirin therapy Coumadin therapy Aspirin + Coumadin therapy None

Table V. Effect of SFA status and Coumadin therapy on patency of grafts

AxBF (patent grafts/total number)

AxUF (patent grafts/total number)

5/7 3/4 12/13

1/1 6/8 4/4

11/12

7/9

Dacron grafts (United States Catheter and Instrument, C. R. Bard, Inc., Billerica, Mass.). Their secondary patency rate at 9 years for AxBF bypass grafts was 77%, not significantly different from 67% for AxUF bypass grafts. 12 Several other authors more recently have reported no significant difference in patency rates between AxBF and AxUF bypass grafts, ls1921 In the small sample size of our study, there was no statistically significant difference between the patency rates between AxBF and AxUF grafts. One may speculate, however, whether there would be a significant difference if there was a larger sample size to compare the patency of AxBF and AxUF bypasses. Although it was not statistically significant, a greater percentage of the AxUF grafts in our series were performed for aortic graft infection. However, the original indication for most previous

AxBF (patent gratis/total number)

AxUF (patent grafts/total number)

6/7

7/8

10/11

2/3

8/10

5/5

8~9

4 '6

Patent SFA on Coumadin Patent SFA off Coumadin Occluded SFA on Coumadin Occluded SFA off Coumadin

SFA, Superficial femoral artery.

aortic grafts was atherosclerotic occlusive disease. Additionally the runoffstatus of the AxBF and AxUF groups in our series were similar. In our study, the indication for the bypass had no bearing on the patency results. In 1979 Campbell et al. 22 first reported the use of PTFE for extraanatomic reconstructive procedures. A major advantage of PTFE is that a thrombectomy can be performed more easily than with other grafts, especially crimped Dacron. Most early reports of axillofemoral bypass grafting had a 35% to 50% incidence of thrombosis.6,7 Hence, the ease of thrombectomy of PTFE grafts in this location was perceived as an advantage. 7'232s In 1978, Dacron grafts

JOURNAL OF VASCULAR SURGERY May 1995

806 M o h a n et al.

58 100

89%

'n148

46

80

44 0 E3 42

42

33 OI

i~ 39

22

12

0 I

6

~

D

I

D

--

----~

1 I

__ __

31

1 0

21

11 80%

W _

_

6

- - - -

1"7

- - - -

1

60 o n

O c

Secondary Primary

40

20

0

I

I

I

1

I

I

I

I

I

1

3

6

12

18

24

36

48

60

Months

Fig. 1. Cumulative patency rate of axillofemoral grafts.

100

36 27

80

29

24

22

26

24

18 o

91% 12 o

6

2 ]

13

7

I[_

80%

1 o --

L r7

60

2

- -

1

0 r"

n

0 1-1

40

Secondary Primary

20

I

I

I

I

I

I

I

I

I

1

3

6

12

18

24

36

48

60

Months

Fig. 2. Cumulative patency rate o f AxBF grafts.

with externally supported rings became available, and two series reported a markedly improved patency rate with these conduits for AxBF and AxUF bypasses. 19,26In 1980, Alexander et at. 2areported a 79% primary patency rate at 21 months with nonringed PTFE grafts for 29 axillofemoral bypasses. A similar patency rate at 2 years has been reported in a nonringed PTFE series consisting of 34 AxBF and 23

AxUF bypasses.27 The use of externally supported PTFE grafts, which became available in 1981, have also been reported to be responsible for the improved patency rates of both AxUF and AxBF bypasses compared with historic controls. 2s'29 The improved results reported with the use of externally supported grafts could be due to the fact that these grafts are resistant to external graft compression. On the

JOURNAL OF VASCULAR SURGERY Volume 21, Number 5

214ohan et al.

807

22 100

85%

EI]19

18

17

19

18

J rq 18

80

o

13

9

I

5

4 e-

0

I

r--i

14

9

5

4

80% 60

0

Secondary

IC]

Primary

0

~

40

20

0

I

I

I

I

I

1

I

I

1

6

12

18

24

36

48

60

Months

Fig. 3. Cumulative patency rate of AxUF grafts.

contrary, Jarowenko et al. 3~failed to demonstrate any changes in the pulse volume recording and segmental Doppler pressure measurements in the calf and ankle by attempted graft compression of nonringed axillofemoral bypasses, by having the patient lie on their graft. The low incidence of graft thrombosis noted in our series and those of others using externally supported ringed grafts suggests that an externally supported conduit is an important factor that contributes to improved patency rates. LoGerfo et al.iS and Rutherford et al. to suggested that the patency of an AxUF or AxBF bypasses is influenced by the patency of the superficial femoral arteries. However, reports from other authors have failed to confirm this finding. 13,19,29 In our current study, the patency of the superficial femoral arteries did not appear to influence patency of the extraanatomic bypasses in either group. A severely diseased yet patent superficial femoral artery could be present in some patients, thus explaining why patency alone is not adequate to assess runoff. We conclude that a patient at high risk with unilateral limb ischemia is well served with an AxUF bypass graft when a simple, local endovascular, iliofemoral or femorofemoral bypass option is not available. The addition of a femorofemoral component to an AxUF bypass graft merely to lower the outflow resistance appears unnecessary unless the opposite limb is also ischemic and symptomatic. An AxBF bypass remains a suitable option when revas-

cularization of both the lower extremities is required in a patient at high risk with atherosclerotic occlusive disease or after removal of an infected aortic graft. The use of externally supported ringed conduits may be a factor that is important in maintaining long-term patency of AxUF and ExBF bypass grafts. REFERENCES

1. Lewis CD. A subclavian artery as the means of blood supply to the lower half of the body. Br J Surg 1961;48:574-5. 2. Blaisdell FW, Hall AD. AxiUary-femoral artery bypass for lower extremity ischemia. Surgery 1963;54:563-8. 3. Louw JH. Splenic to femoral and axillary to femoral bypass grafts in diffuse atherosclerotic occlusive disease. Lancet 1963;1:1401-2. 4. Sauvage LR, Wood SJ. Unilateral axillary bilateral femoral bifurcation graft: A procedure for the poor risk patient with aortoiliac disease. Surgery 1966;60:573-7. 5. Gorman JF, Douglas FM. Axillary femoral artery bypass. Arch Surg 1965;91:509-12. 6. Moore WS, Hall AD, Blaisdell FW. Late results of axiUaryfemoral bypass grafting. Am J Surg 1971;!22:143-54. 7. Burell MJ, Wheeler JR, Gregory RT, Snyder ST, Gayle RG, Mason MS. Axillofemoral bypass: a ten-year review. Ann Surg 1982; 195:796-9. 8. Sharp WJ, Hoballah JJ, Mohan CR, et al. The management of the infected aortic prosthesis: a current decade of experience. J VASC SURG 1994;19:844-50. 9. Donaldson MC, Louras JC, Bucknam CA. Axillofemoral bypass: a tool with a limited role. J VAsc SUING1986;3:75763. 10. Rutherford RB, Patt A, Pierce WH. Extra-anatomic bypass: a closer view. J VASC SURG 1987;6:437-46. 11. Johnson WC, LoGerfo FW, Vollman RW, et al. Is axillo-

808

12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

2Vlohan et al.

bilateral femoral graft an effective substitute for aorticbilateral iliac femoral grafts? Ann Surg 1977;186:123-9. Ray LI, O'Connor JB, David CC, et al. Axillofemoral bypass: a critical reappraisal of its role in the management of aortoiliac occlusive disease. Am J Surg 1979;138:117-28. Ascer E, Veith FJ, Gupta SK, et al. Comparison of axillounifemoral and axillobifemoral bypass operations. Surgery 1985;97:169-74. Rutherford RB, Flanigan DP, Gupta SK, et al. Suggested standards for reports dealing with lower extremity ischemia. J VASCSURG 1986;4:80-94. LoGerfo FW, Johnson WC, Corson JD, et al. A comparison of the late patency rates of axillobilateral femoral and axillounilateral femoral grafts. Surgery 1977;81:33-40. Kalman PG, Hosang M, Cina C, et al. Current indications for axillounifemoral and axiHobifemoral bypass grafts. J VASC SUV,G 1987;5:828-32. Hepp W, de Jonge K, Pallua N. Late results following extra-anatomic bypass procedures for chronic aortoiliac occlusive disease. J Cardiovas Surg 1988;29:181-5. Eugene J, Goldstone J, Moore WS. Fifteen year experience with subcutaneous bypass grafts for lower extremity ischemia. Ann Surg 1977;186:177-83. EI-MassryS, Saad E, Sauvage LR, et al. Axillofemoral bypass with externally supported, knitted Dacron grafts: a follow-up through twelve years. J VASCSURG1993;17:I07-15. Harrington ME, Hartington EB, Haimov M, Schanzer H, Jacobson JJ. Axillofemoral bypass: Compromised bypass for compromised patients. J VASCSUV,G 1994;20:195-201. Schneider JR, McDaniel MD, Walsh DB, Zwolak RM, Cronenwett JL. Axillofemoral bypass: outcome and hemodynamic results in high-risk patients. J VAsc SUV,G 1992;15: 952-63.

JOURNALOF VASCULARSURGERY May 1995

22. Campbell CD, Brooks DH, Siewers RD, et al. Extra-anatomic bypass with expanded polytetrafluoroethylene. Surg Gynecol Obstet 1979;148:525-30. 23. Alexander RH, Selbyby JH. Axillofemoralbypass grafts using polytetrafluoroethylene. South Med J 1980;73:1325-30. 24. Broome A, Christenson JT, EklofB, Norgren L. Axillofemoral bypass reconstructions in sixty-one patients with leg ischemia. Surgery 1980;88:673-6. 25. Chtistenson JT, Broome A, Norgren L, Eklof B. The late results after axillofemoral bypass grafts in patients with leg ischemia. J Cardiovasc Surg 1986;27:131-5. 26. Schultz GA, Sauvage LR, Mathisen SR, et al. A five- to seven-year experience with externally supported Dacron prostheses in axillofemoral and femoropopliteal bypass. Ann Vasc Surg 1986;1:214-23. 27. Pradhan DJ, Juanteguy JM, Michelson E, Abosch O. Extraanatomic bypass: review of 60 cases using PTFE grafts. Am Surg 1982;48:258-60. 28. Harris EJ, Taylor LM, McConnell, Moneta GL, Yeager RA, Porter JM. Clinical results of axillobifemoral bypass using externally supported polytetrafluoroethylene. J VASC Sl:lt~ 1990; 12:416-21. 29. Taylor LM, Moneta GL, McConnel D, Yeager RA, Edwards JM, Porter JM. Axillofemoral grafting with externally supported polytetrafluoroethylene. Arch Surg 1994; 129:588 -95. 30. Jarowenko MV, Buchbinder D, Shah DM. Effect of external pressure on axillofemoral bypass grafts. Ann Surg 1981;193: 274-6. Submitted Sept. 27, 1994; accepted Dec. 16, 1994.

DISCUSSION

Dr. Terry A. King (Cleveland, Ohio). The authors addressed a question that many of us face on a regular basis, namely whether the cross-femoral portion of the bypass is necessary for long-term patency when considering axillofemoral bypass for lower extremity revascularization. They provide a large, recent series of patients and bypasses with current surgical techniques, graft material, and perioperative care. They also had the fortitude to perform a significant number of AxUF bypasses when the general consensus is that the cross-femoral limb is important to long-term patency. There are several considerations in the use of axillofemoral bypass, including patient selection, distal runoff, and the use of outflow procedures, as well as graft selection and the use of anticoagulation. With respect to the patient selection, I presume the large number of patients with infected aortic graft placement reflects the tertiary referral nature of your practice. Regarding distal runoff and the importance of outflow, several authors that you quote show that the superficial femoral artery patency is important to long-term patency of

the axillofemoral bypass. Taylor and his colleagues 29 from Portland show the generous use of concomitant outflow procedures to improve patency rates of their axillofemoral grafts. Certainly the procedure is technically easier in the face of a patent superficial femoral artery. Regarding graft selection and type, it seems to be inferential, but the use of externally supported PTFE appears to have contributed to long-term patency of this axillofemoral procedure. You show no clear significance that outflow is important regarding superficial femoral artery patency, or for that matter, the importance of the cross-femoral llmb. What component does outflow resistance have with respect to patency? My second question has to do with your indications and the use of anticoagulation. It is not extremely clear which patient receives Coumadin versus aspirin. D o your data support the preferential use of externally supported PTFE grafts without any further randomization or studies comparing it with other materials? Dr. Chittur R. Mohan. Outflow resistance probably

JOURNAL OF VASCULAR SURGERY Volume 21, Number 5

plays an important roie in the patency; however, it is not necessary to add a cross-over femoral limb to decrease the outflow resistance. Patency is dependent on the outflow resistance on the side of the ischemic limb. Other people have correlated superficial femoral artery (SFA) patency with results. I n thi s study we did not find any significant difference~ A patent, SFA ma.y still be diseased; plus outflow resistance may be:higher on the same side because of disease other than the SFA below the knee level: So certainly outflow probably plays an important :role, but pateney of SFA itself did not correlate in Our Study, With regard to your question regarding anticoagulation, most of o u r patients are on aspirin or coumadin. therapy and probably this contributed: to the good results that we had in our study.. We certainly believe in starting everybody on aspirin therapy unless there i s a contraindication. As far as~.anticoagulation w i t h Coumadin is concerned, we believe that patients with poor outflow should be on Coumadin therapy. If the patient has any history suggestive of a hypercoagulable state in the past, we

Alohan et al.

809

definitely put them on Coumadin therapy. If the patient has a single episode of thrombosis, even if we can't find a cause for that, we will put them on Coumadin therapy m the follow-up period. With regard to the use of externally supported PTFE grafts, there has been no difference between externally supported PTFE versus externally supported Dacron. We didn't compare them in our study because we don't have much data on externally supported Dacron. We preferentially use externally supported PZFE because the patency results with externally supported Dacron and PTFE are compatible, plus PTFE also has the added advantage of ease of ~rombectomy in case:an occlusion should occur. So, because of these two factors, we believe m using an externally supported PTFE graft; The data on axillofemoral bypass have certainly improved with the introduction of externally supported grafts. As .far as the patency is concerned, there is no difference between Dacron or PTFE as long as they are externally supported: however, thrombectomy is easier with PTFEI