Infrainguinal reconstruction for patients with chronic renal insufficiency

Infrainguinal reconstruction for patients with chronic renal insufficiency Anthony D. Whittemore, MD, Magruder C. Donaldson, MD, and John A. Mannick, MD, Boston, Mass. Purpose: The efficacy of autogenous infrainguinal reconstruction for patients with varying degrees of chronic renal insufficiency was reviewed because it has not been well defined. Methods: Pertinent data were retrieved from hospital records and the vascular registry for all patients with chronic renal insufficiency (serum creatine > 2 mg/dl) who required surgical intervention for ischemic lower limbs during the past 15 years at Brigham and Women's Hospital. Results: During the past 15 years, 56 patients underwent 70 autogenous vein bypass procedures, and 31 underwent 42 primary major amputations. All patients had serum creatinine levels greater than 2 mg/dl. Limb salvage was the indication for intervention in 84% of reconstructed limbs, and 48% required infrapopliteal bypass. No significant difference was found in the operative mortality rate associated with primary amputation (17%) and reconstruction (11%), but overall 5-year survival for the group undergoing reconstruction (40%) was significantly superior to that observed in the group undergoing amputation (9%). The 5-year cumulative limb salvage rate for the group having reconstruction was 80%, with a primary graft patency of 74% and a secondary patency of 77%. Patients with diabetes had lower rates of both limb salvage (63%) and survival (21%). No patient on dialysis survived beyond the third-year interval, but the limb salvage rate after 2 years was 76%. Conclusion: Patients with chronic renal insufficiency achieved graft patency and limb salvage results comparable to those obtained in patients with normal lrddneyfunction, but they sustained higher perioperative morbidity and reduced survival. (J VAsc SURG 1993;17:32-41.)

Infrainguinal arterial reconstruction for peripheral vascular occlusive disease provides anticipated 5-year graft patency rates approaching 80% and slightly higher limb salvage rates when autogenous vein is used as the conduit. 1-4 Results have not been frequently analyzed with respect to the specific subgroup of patients with renal insufficiency. The limited number of reports in the current literature that focus on ischemic limbs in the presence of end-stage renal disease (ESRD) present conflicting primary graft patency rates, ranging from 68% to 89% and limb salvage rates from 76% to 89% at 2 years. 5,6 One overriding observation, however, is the discouraging patient survival rate that is limited to 18% to 39% at 3 years. Optimal management of ischemic limbs in patients with varying degrees of From the Division of Vascular Surgery, Brigham and Women's Hospital and Harvard MedicalSchool, Boston. Presented at the Forty-sixthAnnual Meeting of the Societyfor Vascular Surgery, Chicago, Ill., June 8-9, 1992. Reprint requests: Anthony D. Whittemore, MD, Brigham & Women's Hospital, 75 Francis St., Boston,MA 02215. 24/6/42889 32

renal insufficiency has not been clarified, and to this end we have retrospectively reviewed our experience with this specific cohort of patients. Because patients with overt diabetic nephroarteriosclerosis may have a more advanced stage of vascular occlusive disease, probably compounded by the consequences of medial calcification as a result of kidney impairment, the impact of diabetes on infrainguinal reconstruction in the presence of renal insufficiency is of particular interest. PATIENT P O P U L A T I O N A N D METHODS Pertinent data were retrieved from hospital records and the vascular registry for all patients with chronic renal insufficiency (serum creatine > 2 mg/dl) who required surgical intervention for ischemic lower limbs during the past 15 years at Brigham and Women's Hospital. The upper limit of normal for serum creatinine in our laboratory facility is 1.6 mg/dl, and a sustained elevation in excess of 2.0 mg/dl is therefore considered significant. Although patients with ESRD on long-term maintenance hemodialysis represent the extreme of this patient 0741-5214/93/$1.00 + .10.

Volume 17 Number 1 January 1993

Infrainguinal bypass in patients with renal insufficiency 33

Table I. Demographic characteristics and associated risk factors in patients with chronic kidney failure requiring intervention for ischemic lower limbs Primary amputation

Infrainguinal reconstruction

No. of patients No. of procedures

31 42

56 70

Mean Range Sex Male Female Risk factors Hypertension CAD Tobacco Diabetes CVA Chronic kidney failure Nondialysis patients Serum creatinine (mean) 2-4 mg/dl >4 mg/dl Dialysis patients

65 37-88

69 38-95

25 (60%) 17 (40%)

47 (67%) 23 (33%)

24 26 9 20 5

(57%) (62%) (21%) (48%) (12%)

57 (81%) 49 (70%) 37 (53%) 37 (53%) 9 (13%)

32 (76%) 3.1

54 (77%) 2.8

30 (71%) 2 (5) 10 (24%)

46 (67%) 8 (11%) 16 (23%)

Age

CAD, Coronary artery disease; CVA, cerebrovascular accident.

Table Il. Location of anastomoses in 70 autogenous vein grafts Proximal anastomosis

Distal anastomosis

Location

No.

Common femoral Superficial femoral Proximal Distal Deep femoral Popliteal Above knee Below knee Tibial/peroneal Inframalleolar

56 (80%)

16 20 31 3

(23%) (29%) (44%) (4%)

Table Ill. Thirty-day operative morbidity associated with primary amputation (N = 42) or infrainguinal arterial reconstruction (N = 70) Complication Systemic Myocardial infarction Congestive failure Arrhythmia Pulmonary failure

No. of No. of amputations revascularizations 4 (10%) 1 (2%) 5 (12%) 7 (17%)

CVA Short-term dialysis Coagulopathy

population, a significant and sustained elevation of serum creatinine indicates some degree of functional kidney impairment, not necessarily resulting in significant physiologic derangement, but potentially reflective of underlying vascular disease. The hypothesis under consideration queries whether these individuals with mild kidney impairment harbor more extensive systemic atherosclerosis that will adversely effect perioperative morbidity, graft patency, limb salvage, and overall survival. In a prior report we investigated a similar hypothesis with regard to patients undergoing abdominal aortic aneurysm repair. 7 Patients not undergoing dialysis who have creatinine levels in excess of 4 mg/dl demonstrated increased operative morbidity and a marked reduction in 1-year survival (33%) when compared with either the group undergoing dialysis or patients with lower but abnormal creatinine levels. A similar finding in patients undergoing infrainguinal reconstruction might alter the relative priorities of various treatment options. Although patients with ESRD are undoubtedly important, milder degrees of kidney failure may also affect operative mortality and overall survival rates adversely. Three patients with functioning kidney transplant were excluded because they represent a unique population, and their limited

10 (14%) 2 (3%) 2 (3%)

7 (10%) 2 (3%) 10 (14%) 4 (6%)

-

1 0%)

3 (7%) 2 (5%)

2 (3%) -

Local

5 (12%)

8 (11%)

Wound infection/seroma/hematoma Graft occlusion Major amputation Death

5 (12%)

6 (9%)

7 (17%)

4 (6%) 4 (6%) 8 (11%)

CVA, Cerebrovascular accident.

number in our experience precluded valid analysis as a distinct group. Thiw[-one (36%) of 87 patients so defined required 42 primary amputations, whereas 56 (64%) underwent 70 infrainguinal arterial reconstructions. Primary amputation was carried out when advanced tissue necrosis precluded limb salvage and for those who refused revascularization. Although this number of amputations seems high when compared with our recent series of all patients requiring infrainguinal reconstruction, the present study spans 15 years, and many patients with advanced distal tibial peroneal disease or extensive tissue necrosis were not universally considered appropriate candidates for reconstruction during the earlier years. After our collective favorable experience with long distal autogenous reconstructions, the number of amputations has diminished considerably. Furthermore, our population of patients undergoing inpatient dialysis consists

34

Journal o f VASCULAR SURGERY

Whittemore,Donaldson, and Mannick

100" 90" 80"

"E t.r}

27

70"

20

60"

•

13

7

T

50 40-

E L)

30.

t

2010. 0

T

!(40%)

'

1

5

• Primary Reconstruction(N=56) O PrimaryAmputation (N=31)

'~ ~

1~

~(9%1

24

i6

d8

~o

Months Fig. 1. Cumulative survival curves after primary amputation or infrainguinal reconstruction in patients with chronic renal insufficiency. After 5 years the difference proved statistically significant (p < 0.004). 100-

T46 37 31 90- 1

I I

~

22

I

is I

11

7

4

2~

3'6

4~

~o

7060-

e~

._E ..J

so-

>~ 40E

30-

~3

20" 10-

0~

~

;2

Months Fig. 2. Cumulative limb salvage curve associated with 70 autogenous infrainguinal reconstructions in patients with chronic renal insufficiency. of an exceptionally ill group of patients who are not reflective of a stable maintenance dialysis population. The number of beds available in our inpatient dialysis facility to limited to six and necessarily restricts their use to those individuals requiring hospitalization either for initiation of dialysis during an acute illness or for significant acute comorbid disease. Fifty-six patients (64%) underwent 70 infrainguinal arterial reconstructions. Indications for reconstruction consisted of disabling claudication in 11 (16%) patients with minimal kidney impairment and

critical ischemia in the remainder as evidenced by severe rest pain in 21 (30%) or tissue necrosis in 38 (54%). Demographic characteristics and associated risk factors listed in Table I demonstrate no significant differences between patients requiring amputation and those undergoing reconstruction. Conduits used for reconstruction consisted of saphenous vein in 65 (93%), composite vein grafts in four (6%), and arm vein in a single case. O f the saphenous veins, 32 (46%) were reversed, 3 (4%) were nonreversed and translocated, and 30 (43%)

Volume 17 Number 1 January 1993

Infrainguinal bypass in patients with renal insufficiency

35

Table IV. Five-year cumulative primary patency of 70 autogenous infralnguinal reconstructions in 56 patients with chronic kidney failure Interval (mo)

No. of grafts at risk

No. of grafts failed

No. of limbs lost

No. of grafts withdrawn

Interval failure (%)

Interval patency (%)

Cumulative patency (%)

SE (%)

0-1 1-3 3-6 6-12 12-24 24-36 36-48 48-60

70 54 44 37 26 18 12 10

4 2 2 1 1 1 0 0

3 0 0 0 0 0 0 0

9 8 5 10 7 5 2 3

6.1 4.0 4.8 3.1 4.4 6.4 0 0

93.9 96.0 95.2 96.9 95.6 93.6 100.0 100.0

93.9 90.1 85.8 83.2 79.5 74.4 74.4 74.4

3.0 3.9 4.8 5.3 6.2 7.6 7.6 7.6

were placed in situ with use of methods previously published) Most grafts originated in the common femoral artery and terminated in the popliteal, but a significant number (44%) were carried to the infrapopliteal level (Table II). Eighty-seven percent of the reconstructions were primary procedures, whereas 13% were secondary. Patients were monitored at regular intervals after the initial postoperative visits, usually every 4 months for 2 years and every 6 months thereafter. Graft patency was assessed by history and physical examination supplemented by segmental Doppler pressures, pulse volume recordings, and most recently by duplex ultrasonography. Recurrent ischemic symptoms, a decrease in ankle/brachial index of greater than 0.15, or a graft flow velocity diminished below 0.4 m/sec constituted grounds for suspecting primary graft failure and mandated repeat angiography to assess the graft before revision as indicated. 8 Graft patency, fimb salvage, and survival rates were determined by standard actuarial life-table analysis, expressed as mean + SE as determined by the Greenwood method, and reported in accordance with the Ad Hoc Committee on Reporting Standards. 9-11 In addition to the entire cohort, subpopulations were analyzed for the specific variables of diabetes and the requirement for maintenance hemodialysis. Comparisons of life-table curves used the Mantel-Cox log rank test to determine significance. Primary graft patency is defined by those grafts requiring no subsequent revision after leaving the operating room after the initial procedure. Primary revised patency includes primarily patent grafts as well as those requiring some form of intervention while retaining patency. Secondary patency consists of all grafts including those requiring no intervention (primary patency), those requiring revision while patent (primary revised patency), and those revised after graft occlusion.

RESULTS

Analysis of morbidity with the number of procedures as the denominator demonstrated 40 significant operative complications after 31 (44%) of the 70 reconstructions (Table III). Eight of these complications, seven of which were myocardial infarctions, proved fatal, for a 30-day operative mortality rate of 11%. Twenty-seven complications were observed after 22 (52%) of the 42 primary amputations. These included seven operative deaths (17%) as a result of three myocardial infarctions, one disseminated malignancy, and three patients with progressive kidney failure who were not suitable candidates for, or who refused, long-term dialysis. Postoperative kidney failure was considered a major complication if short-term temporary dialysis was required or hospital stay was prolonged. Neither morbidity nor mortality rates proved significantly different when these two groups were compared. As illustrated in Fig. 1, patients undergoing revascularization demonstrated a 5-year cumulative survival rate of 40%, significantly better than the 9% rate for those undergoing primary amputation (p < 0.004). Infrainguinal reconstruction yielded a 5-year cumulative limbs salvage rate of 80% (Fig. 2), and a primary graft patency rate of 74% (Table IV, Fig. 3). As anticipated, the 100% primary 5-year patency rate in the small group of patients with claudication was significantly better (p < 0.001) than that observed for the group with critical ischemia (67% primary, 72% secondary). Of the 59 limbs reconstructed from critical ischemia, salvage was achieved in 80% for 5 years. Reversed saphenous veins were used exclusively until 1984 when our preference shifted to the in situ method. The 5-year patency rate associated with the 32 reversed grafts (74%) was not significantly different (p --- 0.171) from that associated with the 30 in situ grafts (87%). It should be noted, however,

36

Journal of VASCULAR SURGERY

Whittemore, Donaldson, and Mannick

1 _

-

-

io

e

e

-"

(74%)

i

= ,41 i |

0..........

<) . . . . . . . . . .

o ..........

o

(55%)

50"

E

40"

(.9

30"

fi

20-

• ..........

• (22N

• Tot~l ( N = 7 0 )

10-

o o ~ t e s (N=37) • Dialysis ( N = I O )

0

i

i

i

3g

4h

6b

Months

Fig. 3. Primary cumulative graft patency curves for entire group of 70 autogenous reconstructions in patients with chronic renal insufficiencyand for those subgroups of patients with diabetes and those on dialysis. Interrupted lines indicate SEM greater than 10%. Table V. Thirty-day cardiac morbidity associated with infrainguinal arterial reconstructions in patients with chronic kidney failure No, of

No. of

procedures

MIs (fatal)

70

7 (10%)

Overall Diabetes + -

37

3 (9%)

33

4 (12%)

p < 0.436 Dialysis + -

16 54

2 (13%) 5 (9%) p < 0.198

MY, Myocardial infarction.

that the distal anastomosis was located at the infrapopliteal level in only 25% of reversed grafts, whereas 70% of the in situ veins were carried to the distal tibial or peroneal vessels. After initial reconstruction, 11 of the 70 grafts sustained primary failure. Four of these failures occurred within 30 days and resulted in two immediate major amputations without operative deaths. Three of the 11 failures were revised while still patent, yielding a primary revised patency rate of 78% + 7%. The remaining eight grafts were occluded and resulted in immediate amputation in five. Only a single thrombosed graft was revised, and the secondary patency rate remained 78% + 8%.

No significant differences were noted in operative mortality and morbidity rates when patients with diabetes were compared with patients without diabetes. With respect to cardiac morbidity in particular, patients with diabetes had a lower incidence of fatal myocardial infarctions (9%) than patients without diabetes (12%), but again this proved statistically insignificant (Table V). In the absence of diabetes, patients with renal insufficiency demonstrated a 5-year survival rate of 57%, significantly better (p = 0.018) than the 21% rate observed in the diabetic group (Table VI). Limb salvage proved similar at the 2-year interval, but patients without diabetes achieved a greater limb salvage rate (91%) after 5 years than patients with diabetes (63%). Because of the small number of patients at the 5-year interval, however, this difference was not significant. Primary graft patency rates in patients with diabetes (Table VII, Fig. 3) were not significantly worse than in patients without diabetes at the 2-year interval (73% and 84% respectively, Table VIII). After 5 years, however, patients without diabetes sustained this patency, whereas that associated with patients with diabetes diminished to 55%. This difference approached but did not achieve significance (p = 0.061). This lower primary patency rate in patients with diabetes was improved somewhat with revision to yield a secondary 5-year patency rate of 65%. Although there was no significant difference in operative morbidity and mortality rates when dialysis

Volume 17 Number 1 January 1993

Infrainguinal bypass in patients with renal insufficiency

37

Table VI. Patient survival and limb salvage rates (mean + SEM) after infrainguinal arterial reconstruction in patients with chronic kidney failure 2-year

Total Diabetes +

No. of patients

No. of limbs

Limb salvage

Survival

Limb salvage

Survival

56

59

87 + 5

60 -+ 8

80 + 8

40 + 9

35 24

84+ 7 91-+-6

4 7 + 11 73+9

63 + 19 9 1 + 6

21_+ 11 57-+ 12

16 43

76 -+ 12 90-+ 5

32 _+ 19 66-+ 8

83-+ 8

46--- 10

-

Dialysis +

5-year

12

-

-

Table VII. Cumulative primary patency associated with autogenous infrainguinal arterial reconstruction in patients with renal insufficiency and diabetes (n = 37) and with ESRD on hemodialysis (n = 16) Interval (mo) 0-1 1-3 3-6 6-12 12-24 24-36 36-48 48-60

Diabetes Dialysis Diabetes Dialysis Diabetes Dialysas Diabetes Dialysis Diabetes Dialysis Diabetes Dialysis Diabetes Dialysis Diabetes Dialysis

No. of grafts at risk

No. of grafts failed

No. of limbs lost

No. of grafts withdrawn

Cumulative patency (%)

SE (%)

37 16 26 10 21 6 18 4 11 2 5

3 :1 i 11 1 1 0 0 1 1 1

2 1 0 0 0 0 0 0 0 0 0

6 4 4 3 2 1 7 2 5 1 2

91 93 87 82 83 67 83 67 73 22 55

5.0 7.1 6.1 12,0 7.2 16.6 7.2 16.6 11.1 26.2 17.2

2

0

0

1

55

17.2

1

0

0

1

55

17.2

Table VIII. Graft patency rates for infraingulnal reconstruction in patients with chronic kidney failure 2-year cumulative patency rate No. of grafts Total Diabetes + -

Dialysis + -

5-year cumulative patency rate

Primary

Secondary

Primary

Secondary

70

80 + 6

83 + 6

74 + 8

77 - 8

37 33

73 + 1 84-+ 8

87+6 84-+ 8

55-+ 17 81+ 8

65-+ 19

16 54

22+26 86-+5

82-+ 12 84+6

81-+ 8

requirement was analyzed as a specific variable, two patients (13%) undergoing dialysis sustained fatal myocardial infarctions (Table V). Although the four wound infections all occurred in patients not undergoing dialysis, three of whom had diabetes, this did not prove a significant finding. The degree of kidney failure adversely affected overall survival such that

8 1 +

9

-

81-+

8

patients not undergoing dialysis demonstrated a 3-year survival rate of only 13%, and no patient requiring maintenance hemodialysis survived the third year interval (Table VI). Nevertheless, the limb salvage rate for this small group of patients proved to be 76% at the completion of 2 years. The 2-year primary patency rate for the 16 grafts was limited,

38 Whittemore, Donaldson, and Mannick

however, to 22% (Table vii, Fig. 3), but successful revision of two of four failed grafts improved the secondary patency rate to 82% (Table VIII). The standard errors even at the early 2-year interval for mean salvage and patency rates were greater than 10%. DISCUSSION Recent results with autogenous infrainguinal arterial reconstruction suggest that the overall primary graft patency rate of 65% to 75% is associated with salvage of 80% to 90% of threatened limbs for 5 years at the expense of a 2% operative mortality rate. 1~ Although most of these series specifically single out the diabetic population, which consistently shows no significant differences in these parameters, the population with chronic renal insufficiency has been less frequently examined,s,6 Our experience with this subpopulation as reported in this study suggests that although overall limb salvage rates may remain comparable, operative mortality rates are considerably higher. In addition, patient survival after successful reconstruction approached 70% at 5 years in our recent overall experience, significantly higher than the 40% 5-year survival rate observed in our patients with chronic renal insufficiency. The mortality rate associated with patients on hemodialysis is even higher, in that none of our 11 patients survived 3 years, in keeping with a similar 18% rate reported by Edwards et al.6 but in contrast to the 40% rate reported by Harrington et alfi ,6 Thus the decision to revascularize a threatened limb in a patient with chronic renal insufficiency must take into account a more limited life expectancy and higher operative mortality rate, even though ultimate limb salvage rates for s~rvivors seem comparable to the general population. These considerations seem particularly important for diabetic patients with chronic kidney failure. Factored into this decision must be the consequences, both positive and negative, associated with primary amputation. Unfortunately, since a prospective randomized study cannot be ethically justified, decisions must be based on retrospective analysis of past experience, which is further confounded by the inherent incomparability of the two population groups. Those patients offered primary amputation more often have more severe concurrent disease and are more likely to have tissue necrosis of greater magnitude. The incidence of specific risk factors is identical in both groups, yet it is difficult to quantitate the severity of any individual factor. Although the operative morbidity and mortality rates

Journal of VASCULAR SURGERY

associated with either primary amputation or revascularization appear to be comparable, graft patency and limb salvage rates are consistently high enough to preclude accurate analysis of the added morbidity should amputation ultimately be required after failure of primary revascularization. In this study for instance, the 11 graft failures resulted in five amputations, with no operative deaths. Cost-benefit analyses of reconstruction versus amputation have usually favored primary revascularization, but again the two groups are not comparable. Finally, a variety of potential predictive factors have been sought, but as yet to no avail. 12'13 ESRD, however, has been suggested as one such predictor, particularly for patients with diabetes and with advanced tissue necrosis. 6 It has seemed reasonable, however, in all but the most extenuating circumstances to consider primary infrainguinal reconstruction when there is a realistic chance for salvage of limb viability. Patients with chronic renal insufficiency and diabetes sustained similar operative morbidity as patients without diabetes but achieved lower limb salvage rates (63%) than either patients without diabetes (91%) or patients with normal renal function. Furthermore, given their limited 5-year survival rate (21%), a more temporizing endovascular method such as balloon angioplasty for patients with short segment lesions might be an appropriate alternative warranting further investigation. Primary amputation is also an alternative, especially for patients with diabetes on dialysis with advanced tissue necrosis, as suggested by Edwards et al.,16 who reported all such patients in their series required major amputation in spite of patent grafts. Patients with advanced renal insufficiency requiring maintenance hemodialysis fared very poorly. Although operative morbidity was not dissimilar to that observed in patients not requiring dialysis, overall survival was poor in that no patient survived 3 years. This is in part attributed to the highly selected tertiary nature of our population undergoing dialysis and the limited number of dialysis beds available in our institution, which preclude long-term hemodialysis. Hemodialysis is reserved for patients requiring either initiation of dialysis treatment or hospitalization for acute concurrent disease. In all probability then, our population cannot compare with that recently reported by Harrington et alfi whose results in patients undergoing dialysis were more encouraging. Although our limb salvage rate of 76% after 2 years was comparable to the 84% reported after 3 years in their series, 40% of their patients remained alive after the third year interval. Two of the 39

Volume 17 Number 1 January 1993

patients incorporated in their series had functioning kidney transplants. O n l y 30% o f their patients u n d e r g o i n g dialysis required infrapopliteal reconstruction as o p p o s e d to all o f ours, and 50% o f their patients received a prosthetic graft. I t should be n o t e d that in their series as well as in ours, SEs exceeded 10% after 18 m o n t h s with respect to patency rates and at 30 m o n t h s with regard to limb salvage and patient survival. Patients with renal insufficiency are appropriate candidates for infrainguinal reconstruction as evidenced by acceptable 5-year graft patency (70%) and limb salvage (80%) rates. O u r approach to these patients, therefore, does n o t deviate substantially f r o m that applied to patients with n o r m a l renal function. Operative morbidity, however, is higher in patients with impaired renal function, and patient survival is reduced, especially for patients with diabetes and those o n maintenance h e m o dialysis. The authors acknowledge the assistance of Ms. Julie Lombara for data retrieval. REFERENCES

1. Donaldson MC, Mannick JA, Whittemore AD. Femoraldistal bypass with in situ greater saphenous vein. Ann Surg 1991;213:457-65. 2. Taylor LM, Edwards JM, Porter JM. Present status of reversed vein bypass grafting: five-year results of a modern series. J VAsc SURG 1990;11:193-206. 3. Bandyk DF, Schmitt DD, Seabrook GR, et al. A comparison of in situ and reversed saphenous vein bypasses: the impact of

Infrainguinal bypass in patients with renal insufficiency 39

a surveillance protocol and elective revision. J VASe SURG 1989;9:286-96. 4. Leather RP, Shah DJ, Chang BB, et al. Resurrection of the in situ saphenous vein bypass: 1000 cases later. Ann Snrg 1988;208:435-42. 5. Harrington EB, Harrington ME, Schanzer H, Haimov M. End-stage renal disease-is infrainguinal limb revascularization justified? J VAsc SURG 1990;12:691-6. 6. Edwards JM, Taylor LM, Porter JM. Limb salvage in end-stage renal disease (ESRD). Arch Surg 1988;123: 1164-8. 7. Cohen JR, Mannick JA, Couch NP, Whittemore AD. Abdominal aortic aneurysm repair in patients with preoperative renal failure. J VAse SURG 1986;3:867-70. 8. Bandyk DF, Cato RF, Towne JB. A low flow velocity predicts failure of femoropopliteal and femorotibial bypass grafts. Surgery 1985;98:799-809. 9. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Reports 1966;50:163-70. 10. Greenwood M. The natural duration of cancer. Reports on Public Health and Medical Subjects (London: His Majesty's Stationery Office, 1926), No. 33. 11. Ad HocCommittee on Reporting Standards, Society for Vascular Surgery/North American Chapter, International Society for Cardiovascular Surgery: standards for reports dealing with lower extremity ischemia. J VASC SURG 1986; 4:80-94. 12. Ascer E, White SA, Gupta SK, Veith FJ. Amputation vs revascularization. In: Rutherford RB, ed. Vascular surgery. 3rd ed. Philadelphia: WB Saunders, 1989:1760. 13. Hobson RW II, Lynch TG, Jamil Z, et al. Results of revascularization and amputation in severe lower extremity ischemia: a five-year clinical experience. J VASCSUING1985; 2:174-85. Submitted June 12, 1992; accepted Sept. 25, 1992.

DISCUSSION

Dr. Calvin B. Ernst (Detroit, Mich.). In an attempt to refine indications for infrainguinal reconstruction, Dr. Whittemore and his colleagues have provided valuable data that suggest that with the exception of patients on dialysis, we should remain optimistic and anticipate favorable results after bypass in patients with azotemia, including diabetic patients. From our 15-year experience ending in 1991 at the Henry Ford Hospital, with 97 infrainguinal bypass procedures in 78 patients with azotemia, defined as individuals with a serum creatinine of greater than 1.5 mg/dl, our overall operative mortality rate was 1.3%, 0% for nondiabetic patients, and 3.4% for diabetic patients. During this interval, 240 amputations were performed on 176 patients with azotemia for nonreconstructible disease, with an operative mortality rate of 11.4%. Our 5-year primary graft patency rate in patients with

diabetes was 55%, identical to the authors, and 84% in patients without diabetes. Five-year survival was 45% in patients with diabetes and 72% in patients without diabetes. Our overall 5-year survival rate of 64% was somewhat better than that reported, but our patients may have been less severely azotemic, because only nine of our patients underwent long-term dialysis, and seven had functioning kidney transplants. Of the nine patients undergoing dialysis, however, only three survived beyond 3 years. Therefore we agree with Dr. Whittemore that patients with azotemia not requiring dialysis fare well after infraingulnal arterial reconstruction and should be treated like patients without azotemia. However, as we have just heard, survival and primary graft patency rates among patients undergoing dialysis are dismal, and reoperation, either to restore graft patency or

40

Whittemore,Donaldson, and Mannick

to amputate, sometimes even with a patent graft, is frequently required. So this clearly begs the question, is the most appropriate treatment for a patient undergoing dialysis primary amputation rather than revascularization? We have limited performance of infrainguinal bypasses on patients undergoing dialysis to those requiring limb salvage only. What was the indication for operation in your patients on dialysis? • Also, I suspect that when an autogenous vein is not available, you would recommend primary amputation in the group undergoing limb salvage; am I correct in this assumption? Although primary amputation versus revascularization for patients undergoing dialysis has been widely debated, no definitive answer has yet emerged, and until endovascular techniques are refined and prove successful for the mukisegment advanced disease that all of these patients have, Dr. Whittemore, what do you recommend in these patients? Dr. Anthony Whittemore. First, no patient undergoing dialysis had reconstruction for clandication; the indication for intervention was limb salvage. Second, there are differences, as you pointed out, in patient populations. The overall patient mortality rate for those maintained on long-term dialysis is reported to be in the range of 10% per year. The mortality rate in our series was 50% per year, reflecting extensive concurrent disease. With that wide range of survival in patients undergoing dialysis, I think one has to individualize assessment of treatment plans appropriately. As Drs. Taylor and Porter pointed out in their group of patients who had reconstruction and were undergoing dialysis, six of their patients with diabetes required amputation despite patent grafts. These were patients with advanced tissue necrosis. Thus if a patient maintained on dialysis, particularly a patient with diabetes, has advanced tissue necrosis, perhaps the more humane management is primary amputation in that very select group. And short of that, balloon angioplasty has achieved a reasonable shortterm patency rate and remains an option when possible. We would still favor very aggressive revascularization with autogenous tissue in most of these patients, with the exception of the diabetic patient on dialysis with extensive tissue necrosis. In the absence of autogenous vein, perhaps one of the endovascular methods ought to be considered. I do not think that synthetic prostheses, particularly below the knee to the infrapopliteal level, have much to offer. Dr. Lloyd Taylor (Portland, Ore.). I agree with almost everything the authors had to say. Our own experience in patients with chronic kidney failure is almost exactly similar in size, with 70 procedures performed in 50 patients over the past 10 years. The results, too, are almost exactly similar, with respect to graft patency, limb salvage, and the considerable difference between patients on long-term dialysis and those who are not.

Journal of VASCULAR SURGERY

I am concerned by the operative mortality rates reported by the authors, with an overall mortality rate in thiss patient group of 11%. We associate your institution with multiple reports describing meticulous preoperative risk assessment and intraoperative management with intensive care and hemodynamic monitoring to minimize operative risk. So my first question is, do you devote the same attention to these patients and to their operative risk as you do to your very well-reported series of patients undergoing aortic surgery? Do they all go to intensive care? Did they all have hemodynamic monitoring? Did this make a difference? And finally, I am astounded by your recommendations to consider endovascular procedures in the treatment of this difficaalt patient group. You have accurately documented results inferior to those obtained in the overall group of patients undergoing infrainguinal bypass. I fail to understand the logic leading to your suggestion that we use an even more inferior procedure in hopes of improving the results in this patient group. Dr. Kilt Nichols (Columbia, Mo.). I remind this audience that the population with ESRD is increasing, and approximately 32,000 patients enter long-term dialysis each year. Many of these patients have ischemic vascular disease or aneurysmal disease or other diseases develop. Diabetes is often a comorbid condition or the origin of the ESRD in approximately 36%. The authors have carefully identified the significant mortality and morbidity rates in dealing with this segment of the population with lower extremity disease. We at the University of Missouri are increasingly called on to operate on patients with ESRD for ischemic lower extremities. Our results are similar to those reported today, with high mortality and morbidity rates in the population on dialysis. Many of our patients have had multiple angioaccess procedures. Several of them have, or have had, an angioaccess in the ischemic limb. Have you also had that experience? Dr. Whittemore, would you comment on the contribution of a fistula or shunt to the ischemia? H o w would such a patient be handled in Boston? Do they ligate or band the fistula, move it to another location, or proceed with revascularization with the fistula in place? We have found that no simple answer exists. Also, a number of our patients who undergo long-term dialysis have heparin-induced thrombocytopenia, with early postoperative graft failure, from being previously sensitized to heparin during their dialysis. Would you comment as to whether you have found that the heparininduced thrombocytopenia syndrome contributes to the graft failures that occur in patients on dialysis? Dr. Elizabeth Harrington (New York, N.Y.). Our series differs from yours in that it dealt mainly with patients undergoing dialysis, and we found satisfactory revascularization results in terms ofpatency and salvage. In our series as well as yours, a significant number of patients had a fatal

Volume 17 Number 1 January 1993

Infrainguinal bypass in patients with renal insufficiency 41

postoperative myocardial infarction. What are your recommendations for cardiac evaluation in patients with chronic kidney failure requiring infrainguinal bypass? And have these changed over the last decade? Most of our patients who underwent dialysis and who were admitted with limb-threatening ischemia did not appear to be amenable to balloon angioplasty. Did you find that many of your patients with this degree of ischemia were amenable to such therapy? Finally, we noted that in your article you had a 22% primary and an 82% secondary patency rate for the dialysis group. Does that mean that all or most of your patients with successful revision were in this group? Was there a reason for this? Dr. H a r r y Schanzer (New York, N.Y.). As is evident from this presentation, the critical question is whether it is worthwhile to revascularize patients with ESRD on hemodialysis? In the past few years, four articles coming from Portland; Mt. Sinai, New York; Montefiore, New York; and now Boston, have tried to clear this question. If the answer could be obtained by scoring, we would have a tie. Two are for revascularization, and two are against it. I propose that we have to change the approach to clarify this question. We should really try to identify the characteristics that make a patient on dialysis an acceptable risk for revascularization. It is diabetes, coronary artery disease, obesity, degree oftisue injury, runoff, all of the above, some of them, combinations? By recognizing the indMdual risk factor on each patient, the physician could make a more intelligent judgment in terms of deciding on revascularization or amputation. Do you have any feeling for which patients in your population undergoing dialysis were the ones that were really at severe risk and that you would clearly eliminate as candidates for revascularization? Dr. Whittemore. With respect to Dr. Taylor's query regarding our monitoring techniques, they have not differed from those used over the years concurrently with our patients undergoing routine vascular procedures. Although we have reported consistently on these techniques using aneurysm repair as a study model, we carry out the same kind of risk evaluation and intensive monitoring in all patients with vascular disease. Our higher mortality rates in this group with kidney failure reflects the advanced

stage of comorbidity in our unique population undergoing dialysis. Perhaps the reason for your being so astounded with regard to the selective use ofendovascular techniques lies in the difference in our operative mortality rates. Yours are commendable. Ours are much higher, fivefold higher than observed in our population undergoing standard vascular procedures without kidney failure, again a reflection o f their advanced coronary and other concomitant disease. Although endovascular results are clearly inferior to those achieved with conventional autogenous venous reconstruction, the limited life expectancy of our particular population undergoing dialysis may not require optimal durability in peripheral vascular reconstruction. Because we have tried to avoid angioaccess in the lower extremity at all costs, we have very few such patients, and I am therefore unable to answer your first question. Second, we have not routinely looked for or identified coagulopathy in most of our patients on dialysis, As you know, 10% to 15% of our patients undergoing routine vascular procedures have some form of concurrent coagulopathy, the most common of which is heparin-induced platelet aggregation. I am unable to provide comparable data in the group with kidney failure or dialysis, since this present study spanned 15 years and we did not routinely look for hypercoagulability until recently. Dr. Harrington, I hope I have addressed the issue of cardiac evaluation, which should proceed just as for any other patient undergoing vascular surgery. Only a few patients have peripheral lesions amenable to balloon angioplasty. Their disease is typically far advanced and most often requires distal surgical revascularization, Thus for those patients with markedly advanced tissue necrosis primary amputation should receive serious consideration. The increase to 82% s e c o n d ~ patency rate from the 22% primary rate reflects the successful secondary revision of only two grafts in our series. We cannot therefore speak to the efficacy of revision with any statistical va~dity but merely report that it seemed worthwhile in our small group. And finally, Dr. Schanzer, we were unable to come up with any particular predictor that would help guide us in reaching a decision as to whether to amputate or to perform primary reconstruction.