Does dextran 40 improve the early patency of autogenous infrainguinal bypass grafts?☆☆☆★

Article Outline

• Abstract
• METHODS
• RESULTS
• DISCUSSION
• Acknowledgements
• Discussion
• References
• Copyright

Abstract 

Purpose: We determined whether the administration of dextran 40 would increase the early (30-day) patency of autogenous infrainguinal bypass grafts. Methods: During a 4-year period, 244 patients undergoing 273 autogenous infrainguinal bypass grafts were prospectively enrolled into and completed this study. Patients were randomized into two groups; one of the groups received a 72-hour infusion of dextran 40 after surgery, and the other did not. Comparisons were made between those patients who did and did not receive dextran 40 with respect to risks factors, demographics, and early graft patency. Results: One hundred twenty-six procedures were accompanied by the use of dextran; 147 were not. There was no significant difference between the two groups with respect to patient age, gender, perioperative risk factors, indication for surgery, or location of bypass graft (popliteal vs tibial). Among those patients receiving dextran, there were eight early occlusions (6.4%) and four deaths (3.2%); 89.7% of the patients were alive with patent grafts 30 days after surgery. In the group not receiving dextran, there were 10 early occlusions (6.8%) and 3 deaths (2%); 90.5% of the patients were alive with patent grafts 30 days after surgery. There was no significant difference between the two groups with respect to rate of early occlusion (p = 1.00), death (p = 0.71), or 30-day patency (p = 0.84). Conclusions: The administration of dextran 40 does not increase the early patency of autogenous infrainguinal bypass grafts. Its routine use during these procedures cannot be recommended. (J Vasc Surg 1998;28:23-7.)

Early closure remains a significant problem after infrainguinal bypass grafting; it occurs in 3% to 15% of cases.¹, ², ³, ⁴, ⁵, ⁶ In some instances technical error or poor runoff is found to be responsible.¹, ² However, in a number of cases neither of these factors appears to account for graft occlusion. Increased platelet reactivity to collagen and an increase in factor VIII activity after femorodistal bypass grafting have been described.⁷ It has been hypothesized that this increase in thrombogenicity has been responsible for the early failure of a number of bypass grafts.

Dextran 40 is composed of long-chain carbohydrate polymers with a mean molecular weight of 40,000. In vivo it has been shown to be an effective plasma expander that lowers blood viscosity and increases peripheral blood flow.⁸, ⁹ In addition, it has been shown to inhibit platelet adhesion and to specifically interfere with the interaction of factor VIII, platelets, and endothelial cells.¹⁰, ¹¹ Several studies have suggested that the administration of dextran 40 increased the early patency of infrainguinal bypass grafts.¹², ¹³ As a result, many surgeons began to use dextran as a routine adjunct to lower extremity arterial reconstruction. With the hope of better delineating its role in lower extremity revascularization, we began a prospective and randomized study to determine whether dextran 40 improved the early patency of autogenous infrainguinal bypass grafts.

Back to Article Outline

METHODS 

Between July 1992 and June 1996 we performed 292 autogenous infrainguinal bypass grafts. After informed consent had been obtained, 244 patients were entered into and completed the study protocol. These patients underwent a total of 273 operations. After the drawing of presealed opaque envelopes each patient was randomized to either the dextran or the non-dextran group. Nineteen additional patients were excluded from the study; these included two patients who refused randomization, four patients who had their dextran discontinued because of complications, six patients with recent bouts of congestive heart failure, and seven patients with creatinine levels above 2 mg/dL not yet requiring dialysis.

Patient demographics, preoperative risk factors, and indication for surgery were noted in all cases. Graft patency at 30 days or early closure was confirmed by duplex scan. Initial graft occlusion was used as the endpoint of the study, and secondary graft patency was not evaluated.

Infrainguinal bypass grafts were performed with patients under regional anesthesia unless the harvesting of arm veins was required; for these cases, general anesthesia was used. All grafts were performed in reversed fashion. Dextran 40 infusions were begun after induction of anesthesia. The first bottle of dextran (500 mL) was administered over 5 hours as a loading dose. A constant infusion of dextran 40 was then begun immediately, and the patient received 500 mL during each 24-hour period for a total of 72 hours. All patients were anticoagulated with 5000 units of sodium heparin before arterial clamping. Protamine sulfate was not used to reverse the effect of heparin. Patients on chronic oral anticoagulants or antiplatelet agents had their medications discontinued 1 week before surgery and restarted 1 week after surgery.

Randomization was performed by the lottery method. Five hundred opaque envelopes were prepared before the study was undertaken. One half of the envelopes contained dextran cards; the other half contained non-dextran cards. The envelopes were then placed in a bin and shuffled. For each patient eligible for inclusion in the randomization process, a single card was drawn before surgery. Preparation and selection of the envelopes was performed by office personnel not involved in the research protocol.

Randomization was performed before every operation, and for the purposes of statistical analysis each procedure was considered separately. Comparisons of demographic data, risk factors, indication for surgery, and rates of early graft occlusion (30 days) were made between the two groups. χ² and Fisher exact tests were used to analyze the categorical factors. The Student t test was performed to test for differences in the mean level of specific factors. Statistical significance was assumed at a p level less than 0.05. Sample sizes were selected to enable a difference of 30% in the rate of early graft occlusion between the two groups to be detected at a power of 80%. Differences less than this were considered to be clinically insignificant.

Back to Article Outline

RESULTS 

One hundred twenty-six procedures in the trial were accompanied by the use of dextran; 147 procedures were not. Of the patients included in the study, 63% were male and 37% were female; their ages ranged from 43 to 94 years, with a mean of 72.2 years. Forty-five percent of the patients had diabetes, 42% had symptomatic heart disease, and 41% were under treatment for hypertension. Forty-five percent of the patients smoked 20 or more cigarettes per day. There was no significant difference in demographic data or risk factors between patients who did and patients who did not receive dextran (Table I).

Table I. Demographics and risk factors

Indications for surgery included claudication in 26% of patients, nonhealing ulcer in 26%, ischemic rest pain in 25%, and gangrene in 23%. Fifty-three percent of the grafts were to the popliteal artery (38% above knee and 62% below knee), and 47% of the grafts were to the infrapopliteal vessels. Ninety-one percent of the grafts were performed with greater saphenous vein, 8% with arm vein, and 1% with lesser saphenous vein. There was no significant difference in indication for surgery, distal target vessel, or type of conduit between patients who did and patients who did not receive dextran (Table II).

Table II. Type of operation and conduit

Five patients had complications that were felt to be secondary to dextran administration 48 to 96 hours after surgery. Four patients (three with heart failure and one with renal failure) had their dextran discontinued before infusion was complete, and they were excluded from the study. None of these patients had early graft occlusion, but one died within 30 days of surgery. Another patient experienced an elevation of her creatinine level to 5 mg/dL with oliguria soon after finishing her course of dextran therapy. Her renal function returned to normal in 12 days.

Overall, 18 patients (6.6%) had closure of their grafts within 30 days of operation. Eight patients to whom dextran was administered (6.4%) and 10 patients to whom it was not (6.8%) experienced early graft occlusion. Seven patients (2.6%) died within 30 days of surgery; four of these patients had received dextran and three had not. There was no significant difference between the two treatment groups with respect to rate of early graft occlusion (p = 1.00) or rate of perioperative death (p = 0.71). One patient who received dextran and two patients who did not underwent revision of hemodynamically failing but patent grafts within 30 days of operation. One patient in each group required graft removal for infection. Ninety percent of patients were alive with functioning grafts 30 days after surgery (Tables III and IV).

Table III. Indication for surgery

Table IV. Rates of early graft failure

	No. patients (%)
	Dextran	Non-dextran
Variable	(n = 126)	(n = 147)	p value
Early	8 (6.4)	10 (6.8)	1.00
occlusion*
Early occlusion	4 (7.3)	6 (8.1)	1.00
of tibial grafts
Grafts patent	113 (89.7)	133 (90.5)	0.84
at 30 days
Death	4 (3.2)	3 (2.0)	0.71

*Occlusion by gender: female, 8 (7.9%) ; male, 10 (5.8%); p = 1.00.

Back to Article Outline

DISCUSSION 

The immediate perioperative period remains a precarious time for infrainguinal bypass grafts. Elevated levels of factor VIII, increased activation of platelets, and decreased levels of antithrombin III create a hypercoagulable state during the first days after graft implantation, and it is during this time that a number of unexplained graft failures occur.⁷ It has been postulated that some of these grafts would achieve long term patency if early occlusion could be avoided. Surgeons have been reluctant to use full anticoagulation with heparin or warfarin in the immediate perioperative period for fear of bleeding complications, and traditional antiplatelet therapy has not been convincingly shown to decrease early graft thrombosis. This has led to the search for a drug that would selectively neutralize the hypercoagulable state found in the early perioperative period without creating a prohibitive risk of hemorrhage.

Dextran appeared to be an attractive alternative to heparin anticoagulation and antiplatelet agents. Its ability to inhibit factor VIII activity, increase clot lysability, and decrease the aggregation of platelets in response to collagen seemed ideal in combating the postoperative hypercoagulable state.¹⁰, ¹¹, ¹³ Dextran had proved its efficacy in the prevention of deep vein thrombosis and had been successful in preventing thrombosis of small vessels in experimental models.⁸, ⁹ Several clinical studies suggested that dextran was effective in preventing early thrombosis of infrainguinal bypass grafts, particularly those to the infrapopliteal vessels and those in which autologous vein had not been used.¹², ¹³ This evidence led many surgeons to use dextran as an adjunct to almost all infrainguinal arterial reconstructions, even though its efficacy in improving the patency of autogenous bypass grafts had not been conclusively proved.

Unfortunately, dextran is not without its untoward effects. Acute renal failure, congestive heart failure, thrombocytopenia, and anaphylaxis after its use have all been reported, and as with all pharmacologic agents the risks and benefits of its administration must be carefully weighed.¹⁴, ¹⁵, ¹⁶

The dosage and method of administration of dextran have not been well standardized. Various investigators have given patients 500 to 1000 mL during each 24-hour period for times ranging from 1 to 5 days. Dextran has been given either by intermittent rapid bolus or by continuous infusion. Early in our experience, we attempted to follow the intermittent rapid infusion protocol described by Rutherford¹³; this led to an unacceptably high incidence of congestive heart failure in our patient population. We then adopted our current method of administering an initial bolus of dextran during the induction of anesthesia followed by the continuous administration of 500 mL every 24 hours. This method has the advantage of rapidly achieving a high plasma drug concentration and maintaining this level throughout the treatment course. Even with this approach several episodes of heart failure appeared to be precipitated by the dextran. The choice of a 72-hour period of therapy reflects the fact that most of our patients were discharged on the third postoperative day; we could not justify prolonging the hospital stay solely for the sake of our research protocol.

In contrast to other studies, our trial focused exclusively on patients undergoing autogenous bypass grafting. The rate of graft closure and the makeup of the patient population in this series are comparable to those reported by others.¹, ², ³, ⁴, ⁵, ⁶ In addition, the two treatment groups were closely matched for risk factors, patient demographics, and type of operation performed, and the sample sizes were large enough to have made possible the detection of a relatively small difference in early occlusion rate (30%) between the two groups. Still we were unable to observe any difference in early graft failure rate between patients who did and patients who did not receive dextran. It is probable that any observed benefit of dextran therapy is conferred only on patients who receive prosthetic bypass grafts.

In summary, we conclude that dextran does not improve the early patency of autogenous infrainguinal bypass grafts. Although they are rare, serious and potentially fatal side effects of its administration do occur. There may be subsets of patients who might benefit from dextran administration (those with veins of poor quality, those with poor outflow, and those undergoing prosthetic bypass grafting), but its routine use as an adjunct to lower extremity revascularization performed with autologous vein cannot be recommended.

Back to Article Outline

Acknowledgements 

Statistical analysis was provided by Stanley Azen, PhD, of the University of Southern California, School of Medicine, Division of Biometry.

Back to Article Outline

Discussion 

Dr. Robert B. Rutherford (Silverthorne, Colo.). In a single institution prospective randomized trial, Drs. Katz and Kohl were unable to demonstrate a protective effect of Dextran 40 on the patency of autogenous infrainguinal bypass grafts. Over 90% of the grafts were greater saphenous vein grafts, and just over 50% were to the popliteal artery. Runoff was not characterized. In view of this, I am not surprised at the outcome and agree with their conclusions that Dextran 40 should not be routinely used for infrainguinal bypass grafts with saphenous vein grafts.

In fact, we came to the same conclusion in a multicenter prospective randomized trial reported in the Journal of Vascular Surgery over 13 years ago. However, we deliberately studied only “difficult” distal bypass grafts because we wanted enough cases at risk for early thrombosis to show an effect. Therefore, we entered vein grafts only if they had poor runoff or terminated in crural arteries. Even so, the thrombosis rate for both control and Dextran 40 vein grafts was identical and only 2.6%. However, in the remaining cases, consisting of polytetrafluoroethylene or umbilical vein grafts, or in those requiring adjunctive procedures, such as endarterectomy, the 1-week thrombosis rate for controls was 36% compared with only 12% for those receiving Dextran 40. In our follow-up of this trial, the patency rate difference in favor of those with polytetrafluoroethylene or umbilical vein grafts receiving Dextran 40 was statistically significant to 32 months.

Admittedly, a significant amount of the initial benefit of Dextran 40 was lost between 1 week and 1 month because, presumably in the face of technical imperfections or poor runoff flow, it only delayed inevitable thrombosis. Yet, when we closed the trial at 195 cases, rather than the 156 we reported, the advantage for Dextran 40 was still statistically significant at 1 month, without graft stratification.

Therefore, I still feel Dextran 40 is beneficial in any scenario in which the risk of early thrombosis is high. Our control thrombosis rate was close to 21%, three times higher than in the present trial, which is why I think we were able to show some benefit.

Finally, there is concern about the risk of Dextran 40. With the exception of infusion protocol violations, this was not observed in our trial because patients with a history of congestive heart failure and renal disease were specifically excluded. However, in later practice, I adopted the same protocol as used in this trial to avoid the dangers of bolus administration, although I still administer it for 4 days.

I enjoyed reading the manuscript and feel we should thank the authors for demonstrating that the routine use of Dextran 40 in infrainguinal vein grafts is not justified. However, I hope this is not misconstrued to infer that it is of no benefit in other more difficult distal bypass grafts. In closing, I would ask the authors if they agree with this or if they have abandoned Dextran 40 in their practice entirely.

Dr. Steven G. Katz. Dr. Rutherford, I thank you for your comments. I believe that we are in agreement as to the use of dextran in patients undergoing infrainguinal bypass graft surgery. Although we have abandoned the use of dextran in patients undergoing autogenous reconstructions, we continue to use it routinely in patients undergoing prosthetic bypass graft operations.

Dr. William C. Krupski (Denver, Colo.). I have two statistical concerns about your study. First of all, you analyzed by procedure. So you are counting patients twice, and there may be responders and nonresponders to Dextran 40. On one procedure, the patient may have gotten Dextran 40 and on the other procedure, not have gotten Dextran 40. How did you deal with that?

The second question I have is a concern that there is a 21-patient difference between the control group and the experimental group. In a randomized prospective study, it seems that those numbers should be closer together. I am concerned that you did not analyze this by intent to treat.

Dr. Katz. Those are excellent questions, Dr. Krupski. We did randomize with intent to treat. The only exclusions from randomization were patients in whom dextran administration would be hazardous, and these exclusions were made before the randomization process was begun. Because we did not block randomize, our patient groups would not be expected to be equal, and my statistician assures me that these results are well within the expected parameters.

We considered the question of whether to include patients undergoing bilateral reconstructions in our series. Because these operations were performed, for the most part, on opposite extremities often months to years apart with different conduits, we believed that they behaved as independent rather than dependent variables and thus included them in our study groups.

Dr. Roy Tawes (Burlingame, Calif.). When you started your lecture, you mentioned hypercoagulable states. I noticed in your age distribution that there were many young patients, and indeed Drs. Flanigan and Donaldson demonstrated that people in their 40's and early 50's who undergo infrainguinal bypass grafts do have a higher incidence rate of hypercoagulable states.

I would like to ask two questions. Focusing on your failures, they were equal in both your control and Dextran 40 group. Were there any hypercoagulable patients? I intuitively would expect Dextran 40 to be of more benefit in these patients than patients failing mainly just because of poor outflow. Did they all have poor outflow?

As for my second question, I think there are enough data in the literature that indicate that Dextran 40, used selectively as Dr. Rutherford indicated in these difficult cases, is inferior to heparin with conversion to Coumadin long term in hypercoagulable states. Would you comment on that?

Dr. Katz. We only had a few patients who were in the younger category. The age range was broad, but the average age was about 72 years. We did not specifically identify patients with a hypercoagulable state. As to commenting on whether switching from heparin therapy to Coumadin is more beneficial in high-risk patients than Dextran 40 conversion to Coumadin, I cannot answer that question.

Back to Article Outline

References 

1. Schoenfeld NA, ODonnell TF, Bush HL, Mackey WC, Callow AD. The management of early insitu saphenous vein bypass occlusions. Arch Surg. 1987;122:871–875
   • View In Article
   • MEDLINE
2. Stept LL, Flinn WR, McCarthy WJ, Bartlett ST, Bergan JJ, Yao JST. Technical defects as a cause of early graft failure after femorodistal bypass. Arch Surg. 1987;122:599–604
   • View In Article
   • MEDLINE
3. Taylor LM, Phinney ES, Porter JM. Present status of reversed vein bypass for lower extremity revascularization. J Vasc Surg. 1986;3:288–297
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (831 KB)
   • CrossRef
4. Nehler MR, Moneta GL, Edward JM, Yeager RA, Taylor LM, Porter JM. Surgery for chronic lower extremity ischemia in patients eighty or more years of age: operative results and assessment of postoperative independence. J Vasc Surg. 1993;18:618–626
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (862 KB)
   • CrossRef
5. Belkin M, Conte MS, Donaldson MC, Mannick JA, Whittemore AD. Preferred strategies for secondary infrainguinal bypass: lessons learned from 300 consecutive reoperations. J Vasc Surg. 1995;21:292–295
   • View In Article
6. Brown PS, McCarthy WJ, Yao JST, Pearce WH. The popliteal artery as inflow for distal bypass grafting. Arch Surg. 1994;129:596–602
   • View In Article
   • MEDLINE
7. McDaniel MD, Pearce WH, Yao JST, Rossi EC, Fahey VA, Green D, et al.  Sequential changes in coagulation and platelet function following femorotibial bypass. J Vasc Surg. 1984;1:261–268
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (738 KB)
   • CrossRef
8. Bergentz SE. Dextran prophylaxis of venous thromboembolism. In:  Bergan JJ,  Yao JST editor. Venous problems. Chicago: Yearbook Medical Publishers; 1978;p. 529–540
   • View In Article
9. Ahlberg A, Nylander G, Robertson B, Cronberg S, Nillson IM. Dextran in prophylaxis of thrombosis in fracture of the hip. Acta Chirurgica Scandinavica. 1968;387:83–85
   • View In Article
10. Aberg M, Hedner V, Bergentz SE. Effect of dextran on factor VIII and platelet function. Ann Surg. 1979;189:243–247
    • View In Article
    • MEDLINE
    • CrossRef
11. Aberg M, Bergentz SE, Hedner V. The effect of dextran on the lysability of ex vivo thrombi. Ann Surg. 1975;181:342–345
    • View In Article
    • MEDLINE
    • CrossRef
12. Foster J, Killen D, Jolly P, Kirtley J. Low molecular weight dextran in vascular surgery: prevention of early thrombosis following arterial reconstruction in 85 cases. Ann Surg. 1996;16.3:764–770
    • View In Article
13. Rutherford RB, Jones DN, Bergewtz SE, Bergguist D, Karmody AM, Dardik H, et al.  The efficacy of dextran 40 in preventing early postoperative thrombosis following difficult lower extremity bypass. J Vasc Surg. 1984;1:765–773
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (862 KB)
    • CrossRef
14. Kitziger KJ, Sanders WE, Andrews CP. Acute pulmonary edema associated with the use of low molecular weight dextran for prevention of microvascular thrombosis. Journal of Hand Surgery [Am]. 1990;15A:902–905
    • View In Article
15. Burgos-Calderon R, Figueroa JE. Acute oliguric renal failure associated with low molecular weight dextran. Bol Assoc Med P R. 1972;64:1–4
    • View In Article
16. Thomas JM, Silva JR. Dextran 40 in the treatment of peripheral vascular disease. Arch Surg. 1973;106:138–141
    • View In Article
    • MEDLINE