Volume 32, Issue 4 , Pages 431-438, October 2006
Fluoropolymer-coated Dacron Versus PTFE Grafts for Femorofemoral Crossover Bypass: Randomised Trial
Article Outline
Objectives
To investigate whether patency of a thin walled 8
mm fluoropassivated Dacron graft was similar to that of a standard 8mm PTFE graft for femorofemoral crossover bypass surgery.
Design
A randomised multicentre clinical trial comparing two vascular grafts with participation of 10 departments of vascular surgery in Denmark, Sweden and Norway.
Patients and methods
198 patients were randomised to PTFE (n=107) or fluoropolymer-coated Dacron grafts (n=91), 63% underwent surgery for claudication, 27% for ischaemic rest pain and 10% for tissue loss. The median follow-up time was 24 months (IQR 19–26 months).
Results
The primary patency rate of the two grafts was similar (log rank test: p=0.35). The primary patency rates (95% CI) for coated Dacron and PTFE grafts were 92% (86–98) and 94% (89–99) at 12 months and 87% (74–95) and 93% (87–99) at 24 months, respectively.
Conclusion
In patients with unilateral iliac artery disease not amenable to angioplasty, the femoral-femoral bypass is durable and effective. No difference in patency was found between the two graft materials (fluoropolymer coated Dacron and PTFE).
Keywords: Femorofemoral bypass, Patency, Fluoropassivated Dacron, PTFE
Introduction
Femorofemoral crossover bypass surgery has been used for nearly 4 decades for extra anatomic reconstructions in patients with unilateral iliac artery disease. When introduced, the crossover bypass was reserved for patients with a high surgical risk as an alternative to aortobifemoral bypass.1, 2 Today it is considered a less invasive alternative when endovascular management is impossible or considered unfavourable.3, 4
There is no solid evidence concerning the best graft material. Experimental data indicate that PTFE may be less thrombogenic and have less affinity towards bacteria, as compared to Dacron.5, 6, 7, 8, 9, 10 Though other studies have indicate equal performance of PTFE and Dacron, PTFE has become the preferred graft for infrainguinal procedures in many centres.11, 12, 13, 14, 15 Specific studies in the femorofemoral position are sparse and so are randomised trials.
A recently developed fluoropolymer coated Dacron graft (fluoropassivated Dacron) aims at combining some of the desirable clinical features of Dacron (ease of handling, freedom of suture hole bleeding and lower price) with those of PTFE (lower thrombogenicity, less intimal hyperplasia and less bacterial affinity).16
The objective of this trial was to investigate whether patency of thin walled 8mm fluoropassivated Dacron graft was comparable to that of a standard 8mm PTFE graft for femorofemoral crossover bypass surgery.
Material and Methods
Study design
The study was a prospective multicentre clinical trial comparing two vascular grafts with participation of vascular surgical units in Denmark, Sweden and Norway. The study was conducted in accordance with the Declaration of Helsinki, with the European Good Clinical Practice Guidelines, and with the European EN540 standards for Clinical Trials. The protocol was approved by the regional ethical committees in Denmark, Norway and Sweden (KF 01-149/99) according to local requirements.
Inclusion criteria
Patients scheduled for elective femorofemoral crossover bypass surgery for uni-iliac occlusive disease. Patients with significant aorto- or donor-iliac artery disease were eligible, provided vascular or endovascular treatment before or in connection with the femorofemoral bypass procedure could secure uninhibited inflow.
Exclusion criteria
Exclusion criteria included: age <18 or >85 years, pregnancy, patients who could not give fully informed consent in writing, patients with serious concomitant medical disease with a life expectancy less than the span of the study, persons mentally not able to understand and give fully informed consent, when adequate follow-up could not be arranged, emergency surgery for trauma, acute thrombosis or embolism and a proximal anastomosis above the inguinal ligament.
Study centres, length of study
Initially 21 centres, expressed interest, but before the first patients were included, 6 centres withdrew, due to logistic or financial reasons. Five other centres withdrew within the first year after they had enrolled either none or a few patients. Any patients from these centres were not included in the study. Therefore we recruited patients from 10 centres (8 Danish centres, 1 Swedish and 1 Norwegian). Recruitment of patients started in January 2000 and was terminated after 34 months in November 2002 having recruited 198 patients. Though the initial recruitment period of 18 months was extended by 16 months we were approximately 100 patients short of the number estimated from the power calculation, due to the participation of only half the initially planned centres.
The surgical procedure and the graft
For the purpose of this study a femorofemoral bypass was defined as the insertion of a bypass graft described by inverse “U” shape with both anastomoses to the common, the profunda or the superficial femoral artery of the right and left groin, respectively.
The prostheses used were thin walled fluoropassivated Dacron grafts (a knitted Dacron prosthesis in which the surface was treated with a fluoropolymer before sealing with gelatin (Sulzer Vascutek Ltd, Inchinnan Renfrewshire, Scotland, UK)) and thin walled expanded polytetrafluouroethylene (PTFE) grafts (W.L. Gore & Associates Inc., Flagstaff, Arizona, USA or Impra Inc., Tempe, Arizona, USA). In all cases 8mm externally supported grafts were used. The costs of the grafts were comparable (∼ €740 in 2005) (Personal communication, Vingmed Denmark, Dec 2005). Patients were enrolled consecutively and randomisation took place in the operating theatre after surgical exposure and only when the possibility of the intended bypass procedure was confirmed. Assignment to either a PTFE or a fluoropassivated Dacron graft was determined using sealed envelopes using computerised block randomisation.
Endpoints and patency evaluation
Primary uncorrected graft patency was the primary outcome. Procedures performed for disease beyond the graft and its two anastomoses were not considered as corrections to improve primary graft patency. Secondary patency, limb survival, perioperative complications and effect on symptoms were secondary endpoints.
Patency assessment was based on the following objective findings: Accepted imaging techniques (duplex, angiography or MR-angiography) or Doppler signal with handheld Doppler at two points directly over a superficially located graft in addition to maintenance of ankle-arm index improvement or maintenance of achieved improvement in ankle-arm index no more than 0.10 below the highest post operative index or direct observation at operation or post-mortem examination.
If patency was restored after occlusion by thrombectomy, thrombolysis or transluminal angioplasty, and/or problems with the graft itself or one of its anastomoses requiring revision, but not a redo-procedure the graft was considered as having secondary patency.
Patients were classified preoperatively and during follow-up as being asymptomatic, suffering claudication or suffering critical limb ischaemia.17 Each limb was considered separately.
Sample size and study period
This study was designed as a non-inferiority study, i.e. that patency of the fluoro passivated Dacron graft was not inferior to that of the PTFE graft. Assuming a 2-year patency rate of PTFE of 75% (based on the available literature when initiating the study in 19993, 18) and an attrition rate over a 2-year period of 25% (20% mortality and 5% lost to follow-up), the numbers required for 80% power (type two error: 0.2) and 95% confidence (type one error: 0.05) were 150 patients in each group. Based on the number of interested centres and the number of femorofemoral procedures performed in those centres in the year 1998, we estimated that it would take 18 months to recruit 300 patients.
Follow-up
All patients were evaluated at least on the day after surgery and on the day of discharge. Thereafter all patients were assessed at 1, 12, and 24 months for the current status of symptoms, pulses and ankle pressures and the graft was examined by hand held Doppler or duplex-ultrasound, as appropriate. If graft occlusion was suspected a duplex examination or arteriography was performed. One primary investigator at each centre was asked to fill in the follow-up forms at the time of each patient review. Extraordinary visits were performed in case of any clinical or subjective deterioration.
Statistical analyses
Analysis was by intention to treat. When appropriate, the χ2-test or Fischer exact test was used for comparison of categorical data, and the Mann-Whitney U-test for continuous data. Kaplan-Meier statistics were used for analysis of graft failure and survival, and differences between groups were assessed by the log-rank tests. Multivariate analysis was performed with Cox proportional-hazards models with stepwise variable selection. Analyses were performed by the SPSS statistical software package (SPSS Inc, Chicago, Illinois, USA).
Results
Patients were randomised to PTFE (n=107) or fluoropassivated Dacron (n=91) (see flowchart, Fig. 1). In total 124 (63%) patients underwent surgery for claudication, and 72 (37%) for critical limb ischaemia (CLI). The baseline characteristics are presented in Table 1. The patients in the two groups did not differ significantly with respect to the listed co-morbidity.
Fig. 1.
Flowchart illustrating participants flow through the study.
Table 1. Baseline characteristics of 198 femorofemoral bypass
| Parameter | Fluoro-Dacron | PTFE | |
|---|---|---|---|
| n | 91 | 107 | |
| Age | median (IQ-range) | 71 (65–76) | 68 (61–77) |
| ABI | median (IQ-range) | 0.60 (0.50–0.77) | 0.60 (0.42–0.88) |
| Indication surgery | Claudication | 60 (66%) | 66 (62%) |
| Rest pain | 23 (25%) | 30 (28%) | |
| Ulcer | 8 (9%) | 11 (10%) | |
| Male sex | 51 (56%) | 50 (47%) | |
| Never smoked | 8 (9%) | 15 (14%) | |
| Diabetes | Not using insulin | 9 (10%) | 9 (8%) |
| On insulin | 1 (1%) | 5 (5%) | |
| History of stroke | 2 (2%) | 3 (3%) | |
| Coronary artery disease | 21 (23%) | 30 (28%) | |
| Hypertension | 31 (34%) | 41 (38%) | |
| Renal disease | 5 (6%) | 7 (7%) | |
| Hyperlipidemia | 12 (13%) | 21 (20%) | |
| Pulmonary disease | 23 (26%) | 19 (18%) | |
| Previous amputation | 3 (3%) | 1 (1%) | |
| Duration of operation (min) | median (IQ-range) | 105 (85–140) | 110 (84–138) |
Donor limb inflow was improved in 76 (38%) patients of whom 35 (18%) patients had a preoperative iliac angioplasty (PTA) while 41 patients (21%) had a simultaneous intra-operative procedure performed: iliac PTA in 7, iliofemoral bypass or endarterectomy in 15, femoral endarterectomy in 15 and other procedures in 4. The patients in the two groups did not differ significantly with respect to the number of inflow procedures performed.
The preoperative angiogram allowed runoff assessment in 193 (97%) patients, 86 (45%) of whom revealed occlusion (n=76) or >50% stenoses (n=10) of the superficial femoral artery (SFA). In this subgroup of patients, intraoperative outflow procedures were performed in 39 (45%): a popliteal or crural bypass in 17 patients and thrombendarterectomy of the common and/or profound femoral artery in 22.
30-day results
The 30-day operative mortality was 0.5% (1/198). One 79-year-old female suffered a fatal myocardial infarction during the first post-operative day. However, 21 patients (11%) suffered a peri-operative (within 30 days) complication. Within the first month, 14 (7%) patients experienced infectious complications, equally distributed between the patients groups (Dacron 7 (8%)/PTFE: 7 (7%)): superficial wound infections occurred in 6 (7%) and 6 (6%) while graft-infections developed in 1 (1%) and 1 (1%), respectively. Three patients had thrombosis of graft, 2 had postoperative bleeding necessitating reoperation and 2 patients suffered thrombosis of a femoro-popliteal bypass of the recipient limb.
Additionally 3 graft infections occurred within 2, 3 and 12 months postoperatively, giving a incidence of graft-infections of 2.5% (5/198) and in total 9% (17/198) having infectious complications. All five patients with graft infection had their graft removed but none required amputation.
Follow-up
During follow-up, median 24 months (IQR 19–26 months), there were 20 graft failures (10%) (Dacron: 11, PTFE: 9) and 24 (12%) deaths. Another 6 patients (3%) were lost to follow-up. A further 6 patients had their graft removed due to graft-infection (n=5) or graft-occlusion resulting in a redo procedure (n=1) (Fig. 1). The primary patency rate of the two grafts was not significantly different (log rank: p=0.35). The primary patency rates for fluoropassivated Dacron and PTFE grafts were 92% (95%-confidence interval: 86–98) and 94% (89–99) at 12 months and 87% (74–95) and 93% (87–99) at 24 months, respectively (Fig. 2, Table 2). Pooling the two groups yielded patency rates (95%CI) at 12 and 24 months of 92% (89–96) and 90% (86–95), respectively. Only 4 of 20 occluded grafts were successfully revascularised after thrombectomy or thrombolysis. Analysis of the occluded grafts (n=20) revealed that 12 (10%) had been operated on for claudication and 8 (11%) for CLI. Of those 12 with claudication and graft occlusion, three eventually had a major amputation while nine (75%) remained stable or improved. Among those operated on for CLI, 2 underwent major amputation following graft occlusion, while 6 remained clinically stable or improved. Finally, 2 further patients had a major amputation due to progression of disease. Thus, in total seven (3.5%) patients underwent major amputation (Dacron n=3/PTFE n=4). There were no cases of donor limb amputation.
Fig. 2.
Primary patency of the two graft materials (PTFE and Fluoropassivated Dacron).
Table 2. Patency data
| Interval start (months) | Number entering interval | Number exposed to risk | Kaplan Meyer cumulative patency at 1, 6, 12 and 24 months. (95%-CI) and {SE}. |
|---|---|---|---|
| Flouropassivated Dacron | |||
| 0 | 91 | 88 | 98% (95–100%) {0.02} |
| 6 | 80 | 80 | 95% (90–100%) {0.02} |
| 12 | 76 | 74 | 92% (86–98%) {0.03} |
| 18+ | 69 | 63 | 87% (74–95%) {0.04} |
| PTFE | |||
| 0 | 107 | 104 | 100% (100–100%) {0} |
| 6 | 97 | 95 | 97% (94–100%) {0.02} |
| 12 | 90 | 87 | 94% (89–99%) {0.03} |
| 18+ | 82 | 73 | 93% (87–99%) {0.03} |
Survival (95%CI) in patients operated on for claudication was 98% (95–100%) and 93% (86–99%) at 12 and 24 months, respectively, as compared to the survival in patients operated on for critical limb ischaemia: 88% (79–97%) and 79% (64–94%) (log rank: p=0.0068; Hazard ration 3.6 (95%-CI: 1.5–8.8) after adjustment for gender and age.
Risk factors for graft occlusion
As patency of the two grafts did not differ significantly, data were pooled for further analysis. Neither indication for surgery (claudication or critical limb ischaemia) did not affect patency (log rank: p=0.518) nor did runoff (0–49% SFA stenosis vs >50% or occlusion of SFA) (log rank: p=0.143) (Fig. 3). Preoperative donor iliac PTA was not associated to patency rate (Log rank: p=0.440). However, endarterectomy of the donor common femoral artery was associated to an increased risk of graft failure (log rank: p=0.011, Hazard ratio 4.1(95%CI 1.4–12.4)).
Fig. 3.
Patency according to the severity of SFA disease.
Symptoms and pressures
Recipient limb ankle-brachial index increased from a median of 0.41 (interquartile range: 0.30–0.54) preoperatively to 0.80 (0.64–0.96) and 0.80 (0.62–0.95) at 12 and 24 months, respectively (Fig. 4). A total of 75% and 77% patients were reported to be free of symptoms at 12 and 24 months respectively (Fig. 5).
Fig. 4.
Development in ankle brachial index (ABI) in donor (●) and recipient (-) leg (mean and interquartile range).
Fig. 5.
Effect on ischaemic symptoms following femorofemoral bypass surgery (CI: claudication intermittence; CLI: critical limb ischemia).
Donor limb ankle-brachial index remained unchanged, from a preoperative median value of 0.82 (0.64–0.99) to 0.82 (0.64–1.00) and 0.82 (0.60–1.00) at 12 and 24 months, respectively (Fig. 4). Preoperatively, 78% reported no symptoms related to the donor limb, compared with 79% and 80% at 12 and 24 months postoperatively. One month postoperatively 17 patients (9%) reported onset or worsening of ischaemic symptoms of the donor limb.
Discussion
Patency and survival
This trial found that a fluoropassivated Dacron graft performed as well as a standard PTFE graft used in the femorofemoral position. Moreover, the crossover bypass seemed durable and effective for treating patients with uni-iliac occlusive disease. In fact, results obtained in patients with critical limb ischaemia equalled those obtained in patients with claudication, although survival was inferior in the former group, as expected19 (Table 3). However, the study was underpowere, as we did not reach the planned number of patients, in spite of a prolonged period of patient recruitment.
Table 3. Primary patency of femorofemoral bypass reported in recent literature
| n | Primary patency Critical Limb Ischaemia/Claudication | Survival Critical Limb Ischaemia/Claudication | |||
|---|---|---|---|---|---|
| 1 year | 2 years | 1 year | 2 years | ||
| Criado 199318 | 110 | 88%/82% | 74%/75% | 90%a | 84%a |
| Johnson 199920 | 340 | 78%a | 68%a | ||
| Purcell 200524 | 144 | 88%/93% | 82%/92% | 86%/99% | 67%/97% |
| Kim 200526 | 192 | 86%/95% | 73%/90% | ||
| Eiberg 2006 | 198 | 90%/95% | 90%/90% | 88%/98% | 79%/93% |
aBoth CLI and claudicants. |
Only a few studies have compared different graft material for femorofemoral grafts. In 1999, Johnson et al. found no difference in patency of PTFE and preclotted knitted Dacron grafts in a randomised trial.20 For many simple Dacron is the gold standard for femoro femoral bypass and a 3rd arm of simple Dacron would have been tempting but unrealistic due to insufficient available patients. In the only recent randomised controlled trial, Johnson et al. reported the 2 year patency of simple Dacron as 68% (n=340, CLI: 72%).20 A retrospective study found a 2 year patency of Dacron of 79% (n=86, 48% CLI).18 When pooling patients who received Dacron and patients who received PTFE, the patency rate of the crossover-bypass in our study (90% at 2y) is comparable with both the unilateral iliac bypass (92% at 2y) and the aorto-bifemoral bypass (91% at 2y) as reported by van der Vliet et al. in an retrospective review of 184 unilateral and 350 aorto-bifemoral reconstructions for obstructive disease.21
Risk factors
Over the years several risk factors for femorofemoral bypass graft failure have been analysed. Indication for surgery, quality of inflow, superficial femoral artery outflow and more recently hypertension have been identified, though results are not consistent.18, 22, 23, 24, 25, 26 In our prospective study no correlation between patency and any of the co-variables listed in Table 1 was observed.
Complications
Ever since the introduction of the femorofemoral bypass deterioration of donor limb haemodynamics causing a “steal syndrome” has been a concern.4, 27 In our study 9% reported new or worsening of ischaemic symptoms in the donor limb, though no significant reduction of ABI was recorded. These new donor limb symptoms represent both “silent” arterial insufficiency of the donor limb becoming apparent when the recipient leg was improved, real haemodynamic steal as well as progression of the atherosclerotic process. These results are comparable with others reporting progression of donor limb symptoms in approximately 10%.28 More than one third (38%) had an additional inflow improving procedure explaining why 14% experienced clinical improvement of both limbs.
Though infectious complications occurred in 9% the frequency of graft-infections was only 2.5%, comparable with results reported in the literature.20, 23 There was no evidence to indicate a higher rate of infectious complications with Dacron grafts as compared with PTFE, corroborating other randomised prospective clinical trials.10, 14, 15, 20 However, the absolute number graft-related complications were low and this randomised study was probably not large enough to detect a difference between Dacron and PTFE.
The major limitation of this trial is the risk of overlooking a smaller difference (type ll error) due to under-recruitment, a fact that weakens our conclusions. Nevertheless, the present study is one of the largest controlled femorofemoral bypass trials obtained within a relative short recruitment-period. In previous trials, comparing two graft materials for femoro-femoral bypass20 and femoropopliteal bypass11, 29 approximately 200 patients were found to be sufficient, according to power calculations.
Because the randomisation was stratified per centre, the imbalance between the sizes of the two arms (91/107) may be a consequence of the unexpectedly low number of patients being recruited in some centres.
Although vascular imaging is the gold standard for determining patency, only two of the participating centres had duplex-ultrasound for routine follow-up of crossover bypass. Instead continued improvements in the ankle-arm index and/or Doppler insonation directly over the graft was considered to indicate patency, both methods that are generally acceptable.17, 20
Although our trial was under-recruited and hence underpowered, we found that fluoropassivated Dacron grafts performed as well as PTFE grafts for femorofemoral cross-over bypass surgery, both having excellent patency at up to 2 years.
Founding
The study was partly financially supported by Sulzer Vascutek Ltd, Inchinnan Renfrewshire, Scotland, UK.
References
- . The treatment of unilateral iliac artery obstruction with a transabdominal, subcutaneous, femorofemoral graft. Surgery. 1962;52:343–345
- . Cross-over femorofemoral bypass. Am J Surg. 1966;111:216–219
- . The current role for femorofemoral bypass. J Vasc Surg. 1987;6:71–76
- Efficacy of femorofemoral bypass for intermittent claudication. Clinical and hemodynamic assessment. Am J Surg. 1986;152:215–219
- . Experimental evaluation of bleeding complications, thrombogenicity and neointimal characteristics of prosthetic patch materials used for carotid angioplasty. Cardiovasc Surg. 1996;4:746–752
- . Bacterial adherence to vascular prostheses. A determinant of graft infectivity. J Vasc Surg. 1986;3:732–740
- . Bacterial adherence to vascular grafts after in vitro bacteremia. J Surg Res. 1985;38:648–655
- . An in vitro study of the properties influencing Staphylococcus epidermidis adhesion to prosthetic vascular graft materials. Ann Surg. 1987;206:612–620
- . A clinical survey of aortobifemoral bypass using two inherently different graft types. Ann Surg. 1988;208:625–630
- Prospective randomized trial of polytetrafluoroethylene and Dacron aortic prosthesis. I. Perioperative results. Ann Vasc Surg. 1988;2:248–254
- Dacron vs. polytetrafluoroethylene grafts for femoropopliteal bypass: a prospective randomised multicentre trial. Eur J Vasc Endovasc Surg. 2001;22:226–231
- . Graft type for femoro-popliteal bypass surgery. Cochrane Database Syst Rev. 2000;CD001487
- . Is the preferential use of ePTFE grafts in femorofemoral bypass justified?. Ann Vasc Surg. 2001;15:383–387
- A prospective randomized multicentre comparison of expanded polytetrafluoroethylene and gelatin-sealed knitted Dacron grafts for femoropopliteal bypass. Cardiovasc Surg. 1999;7:214–218
- Prosthetic above-knee femoropopliteal bypass grafting: results of a multicenter randomized prospective trial. Above-Knee Femoropopliteal Study Group. J Vasc Surg. 1997;25:19–28
- . Development of intimal hyperplasia in six different vascular prostheses. Eur J Vasc Endovasc Surg. 2000;20:241–249
- . Management of peripheral arterial disease (PAD) - TransAtlantic Inter-Society Concensus. J Vasc Surg. 2000;31:S1–S296
- . Femorofemoral bypass graft: analysis of patency and factors influencing long-term outcome. J Vasc Surg. 1993;18:495–504
- Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study. The Cardiovascular Health Study Group. Arterioscler Thromb Vasc Biol. 1999;19:538–545
- . Comparative evaluation of externally supported Dacron and polytetrafluoroethylene prosthetic bypasses for femorofemoral and axillofemoral arterial reconstructions. Veterans Affairs Cooperative Study #141. J Vasc Surg. 1999;30:1077–1083
- . Unilateral vascular reconstruction for iliac obstructive disease. J Vasc Surg. 1994;19:610–614
- . Durability of femorofemoral bypass grafting after aortouniiliac endovascular aneurysm repair. J Vasc Surg. 2003;38:498–503
- . Femorofemoral bypass grafts: factors influencing long-term patency rate and outcome. Surgery. 2001;129:451–458
- . Critical appraisal of femorofemoral crossover grafts. Br J Surg. 2005;92:565–569
- . Femorofemoral versus aortobifemoral bypass: outcome and hemodynamic results. J Vasc Surg. 1994;19:43–55
- . Factors affecting the long-term patency of crossover femorofemoral bypass graft. Eur J Vasc Endovasc Surg. 2005;
- . Assessment of donor limb hemodynamics in femorofemoral bypass for claudication. Surgery. 1981;90:764–773
- . Donor limb vascular events following femoro-femoral bypass surgery. Arch Surg. 1991;126:681–685
- . Heparin-bonded Dacron or polytetrafluorethylene for femoropopliteal bypass: five-year results of a prospective randomized multicenter clinical trial. J Vasc Surg. 2004;40:924–931
Awarded the prize as the best clinical paper in the Young Vascular Surgeons session at the annual ESVS meeting Sept 2005, Helsinki, Finland.
PII: S1078-5884(06)00220-6
doi:10.1016/j.ejvs.2006.04.018
© 2006 Elsevier Ltd. All rights reserved.
Volume 32, Issue 4 , Pages 431-438, October 2006
