The Results of Femorofemoral Bypass Using a Saphenous Vein Graft as an Alternative to PTFE Grafts

Abstract

Purpose

This study aimed to report the results of femorofemoral bypass (FFB) using a great saphenous vein (GSV) graft as an alternative to polytetrafluoroethylene (PTFE) grafts.

Materials and Methods

From January 2012 to December 2021, 168 patients who underwent FFB (PTFE, 143; GSV, 25) were included. The patients’ demographic features and surgical intervention results were retrospectively reviewed.

Results

There were no intergroup differences in patients’ demographic features. In GSV vs. PTFE grafts, the superficial femoral artery provided statistically significant inflow and outflow (P<0.001 for both), and redo bypass was more common (P=0.021). The mean follow-up duration was 24.7±2.3 months. The primary patency rates at 3 and 5 years were 84% and 74% for PTFE grafts and 82% and 70% for GSV grafts, respectively. There was no significant intergroup difference in primary patency (P=0.661) or clinically driven target lesion revascularization (CD-TLR)-free survival (P=0.758). Clinical characteristics, disease details, and procedures were analyzed as risk factors for graft occlusion. Multivariate analysis revealed that none of the factors was associated with an increased risk of FFB graft occlusion.

Conclusion

FFB using PTFE or GSV grafts is a useful method with an approximately 70% 5-year primary patency rate. The GSV and PTFE grafts showed no difference in primary patency or CD-TLR–free survival during follow-up; however, FFB using GSV may be an option in selective situations.

INTRODUCTION

Femorofemoral bypass (FFB) was first introduced in 1952 as an alternative to the direct aortic approach for aortoiliac occlusive disease [1]. Until recently, FFB was a popular surgical technique in cases of unilateral iliac occlusive disease at high risk for open aortic or abdominal surgery. Moreover, FFB is a type of extra-anatomic bypass; therefore, its use is ideal in situations in which abdominal operations must be avoided, such as in the setting of intra-abdominal infection or hostile abdomen due to a previous abdominal operation. Moreover, FFB can be used as a minimally invasive revascularization method in patients with multiple comorbidities when the endovascular approach fails.

Various conduits have been used for FFB. Artificial conduits such as polytetrafluoroethylene (PTFE) and Dacron, autologous grafts such as the great saphenous vein (GSV) and femoral vein, and a cryopreserved femoral vein can be conduits for FFB. Among them, the ideal graft is uncertain, and few studies have compared graft types for FFB.

This study reports the results of GSV vs. PTFE grafts in FFB during the same study period. We analyzed primary patency and clinically driven target lesion revascularization (CD-TLR) rates after FFB and assessed the factors affecting FFB graft occlusion.

MATERIALS AND METHODS

The study protocol was approved by the Institutional Review Board of Daegu Catholic University Medical Center (no. CR-22-134), which waived the requirement for written informed consent owing to its retrospective nature. From January 2012 to December 2021, 168 patients treated at our institution with FFB for unilateral iliac artery occlusion due to atherosclerosis were included. Patients with lesions other than unilateral iliac artery occlusion were excluded. We retrospectively reviewed the patients’ electronic medical records. Information included age, sex, comorbidities, indications for surgery, details of surgery, ankle-brachial pressure index (ABI), graft type, complications, and graft patency.

Patients with FFB were targeted for unilateral iliac occlusion, while good inflow from the contralateral iliac artery was ensured. In FFB surgery at our institution, PTFE grafts are the first-line conduits. A 7-mm ring-reinforced PTFE graft with a thin wall is typically used for reconstruction. A Goretex (W. L. Gore and Associates Inc.) or Advanta VXT (Maquet-Atrium Medical) graft was used.

FFB using autologous GSV grafts was considered in the following situations. First, the GSV diameter should be ≥3 mm, while its length should be secured for FFB. Second, when the diameter of the femoral artery used for inflow or outflow was <4 mm, it was preferentially considered to reduce the risk of graft occlusion due to size mismatch and intimal hyperplasia of the anastomosis site after PTFE grafting. Third, GSV was preferentially considered when the previous FFB with PTFE graft was occluded or the risk of surgical-site infection was considered high.

Most patients underwent FFB under general or spinal anesthesia. High-risk patients underwent surgery under local or regional anesthesia. In FFB, the femoral arteries for inflow and outflow were exposed along the longitudinal inguinal incisions. Suprapubic tunneling involves passing the graft superficial to the inguinal ligament in a gentle C-curve shape using a tunneler or curved Kelly forceps.

Postoperatively, all patients without bleeding risk were prescribed antiplatelet agents. Graft patency and surgical results were checked by arterial blood flow function tests (ABI and photoplethysmography) and duplex ultrasound within 1 week postoperative. Graft surveillance after discharge was performed using the arterial blood flow function test at 3 months, 9 months, and annually thereafter. In addition, ultrasound or computed tomography (CT) angiography was performed at 6 months, 1 year, and annually after FFB. Loss of primary patency was defined as the occurrence of any of the following conditions: 1) any occlusion during follow-up; 2) peak systolic velocity ratio >2.4 on duplex ultrasonography; 3) ABI decreased by more than 20% or by more than 0.15 and CT- or digital subtraction angiography–confirmed stenosis >50%; and 4) any CD-TLR. CD-TLR was defined as any revascularization procedure of previously placed FFB performed when the physician considered retreatment for the target lesion necessary or when the patient complained of exertional limb discomfort or claudication within the follow-up period [2]. Postoperative wound complications were classified and graded according to the recommended standards for reports dealing with lower-extremity ischemia [3].

Statistical analyses were performed using SPSS (version 23.0; IBM Corp.). Continuous data are presented as mean and standard deviation. Categorical data are reported as counts and percentages. Student t-test was used to compare the mean values of continuous variables between the two samples, while chi-squared test was used to compare qualitative variables. The cumulative patency and CD-TLR free survival rates were calculated using Kaplan–Meier survival analysis. The factors influencing FFB graft patency were tested using uni- and multivariate analyses with Cox proportional hazard model. Statistical significance was set at P<0.05.

RESULTS

1) Patients’ clinical characteristics

A total of 168 patients who underwent FFB were included. Of them, 143 received an artificial PTFE graft and 25 received an autologous GSV graft. The mean age at the index operation was 71.5±9.1 years (PTFE group) and 72.3±8.5 years (GSV group) (P=0.687). There was also no intergroup difference in sex and comorbidities, suggesting no difference in the patients’ clinical characteristics. Claudication was the main indication for FFB in both groups (72.0%; P=0.998). In the GSV group, 16.0% (4/25) of cases were redo FFB; this rate was significantly higher than that of the PTFE group (6/143) (P=0.021) (Table 1).

Table 1.

Clinical characteristics

Variable	PTFE graft (n=143)	GSV graft (n=25)	P-value
Age (y)	71.5±9.1	72.3±8.5	0.687
Sex, male	125 (87.4)	20 (80.0)	0.320
Hypertension	82 (57.3)	14 (56.0)	0.900
Diabetes	36 (25.2)	6 (24.0)	0.900
Cerebrovascular accident	35 (24.5)	4 (16.0)	0.354
Ischemic heart disease	26 (18.2)	4 (16.0)	0.793
Hyperlipidemia	5 (3.5)	2 (8.0)	0.299
Chronic renal failure	4 (2.8)	0 (0.0)	0.397
Smokers	74 (51.7)	13 (52.0)	0.965
Symptom
Claudication	103 (72.0)	18 (72.0)	0.998
CLTI	40 (28.0)	7 (28.0)
Redo bypass	6 (4.2)	4 (16.0)	0.021*
Preoperative ABI	0.35±0.23	0.24±0.29	0.072
Posoperative ABI	0.75±0.20	0.65±0.26	0.064

Values are presented as mean±standard deviation or number (%).

PTFE, polytetrauoroethylene; GSV, great saphenous vein; CLTI, chronic limb-threatening ischemia; ABI, ankle-brachial index.

*Statistically significant at P<0.05.

2) Disease and procedural details

As for lesion site, the left iliac artery was affected 10% more often than the right iliac artery; however, the proportion of lesion locations was similar between groups (P=0.893). Notably, the frequency of FFB using the superficial femoral artery (SFA) as inflow or outflow was significantly higher in the GSV than PTFE group (P<0.001 for inflow and outflow). SFA was used as an inflow or outflow vessel in 107 of 286 (37.4%) anastomoses in the PTFE group vs. 38 of 50 (76.0%) anastomoses in the GSV group. The operation time was longer with GSV than PTFE grafts (237.2±84.1 min vs. 162.6±77.1 min, P<0.001). Both common iliac artery (CIA) and external iliac artery (EIA) occlusions occurred at frequencies of 47.6% and 56.0% in the PTFE and GSV groups, respectively. Isolated CIA occlusions occurred in 16.1% of patients in the PTFE group vs. 12.0% in the GSV group, while isolated EIA occlusions occurred in 36.4% of patients in the PTFE group vs. 32.0% in the GSV group. The lesion characteristics of the iliac artery did not differ between groups (P=0.721) (Table 2).

Table 2.

Details of the disease and procedure

Variable		PTFE graft (n=143)	GSV graft (n=25)	P-value
Lesion	Left	78 (54.5)	14 (56.0)	0.893
	Right	65 (45.5)	11 (44.0)
Iliac artery occlusion	Combined CIA and EIA occlusion	68 (47.6)	14 (56.0)	0.721
	CIA occlusion	23 (16.1)	3 (12.0)
	EIA occlusion	52 (36.4)	8 (32.0)
Inflow	CFA	103 (72.0)	9 (36.0)	<0.001*
	SFA	40 (28.0)	16 (64.0)
Outflow	CFA	76 (53.1)	3 (12.0)	<0.001*
	SFA	67 (46.9)	22 (88.0)
Operation time (min)		162.6±77.1	237.2±84.1	<0.001*

Values are presented as number (%) or mean±standard deviation.

PTFE, polytetrauoroethylene; GSV, great saphenous vein; CIA, common iliacl artery; EIA, external iliac artery; CFA, common femoral artery; SFA, superficial femoral artery.

*Statistically significant at P<0.05.

3) Clinical outcomes

The length of hospital stay was significantly longer in the GSV (16.8±16.1 days) vs. PTFE group (11.1±9.9 days) (P=0.018). Wound complications included incisional and wound infections that required antibiotic treatment, groin hematoma, or seroma. Most of the wound complications were grade 1 (70.6% [12/17]), while grade 2 or higher complications occurred in five cases. The total wound complication rate was approximately 10%, showing no intergroup differences (P=0.703) (Table 3).

Table 3.

Clinical outcomes

Variable	PTFE graft (n=143)	GSV graft (n=25)	P-value
Length of hospital stay (d)	11.1±9.9	16.8±16.1	0.018*
Wound complication	15 (10.5)	2 (8.0)	0.703
Follow-up duration (mo)	24.6±30.4	25.0±27.4	0.953
Graft occlusion	19 (13.3)	3 (12.0)	0.860

Values are presented as mean±standard deviation or number (%).

PTFE, polytetrauoroethylene; GSV, great saphenous vein.

*Statistically significant at P<0.05.

The mean follow-up duration was 24.7±2.3 months. During the follow-up period, 22 FFBs became occluded (13.1% [22/168]). The mean interval from the index operation to occlusion was 23.2±2.2 months. There was no intergroup difference in graft occlusion rate (log-rank test, P=0.860).

There was no intergroup difference in primary patency rate at 1-, 3-, and 5-year follow-up (92.9%±2.7%, 83.5%±4.4%, and 74.0%±5.9% for PTFE vs. 100%, 81.5%±11.9%, and 69.8%±14.9% for GSV, respectively; P=0.881; Fig. 1A). The freedom from CD-TLR rate did not differ between groups at 1-, 3-, and 5-year follow-up (95.1%±2.2%, 86.6%±4.2%, and 76.9%±5.5% for PTFE vs. 100%, 81.5%±11.9%, and 69.8%±14.9% for GSV, respectively; P=0.894; Fig. 1B). In a subgroup analysis of SFA as an inflow or outflow vessel, the 1-, 3-, and 5-year primary patency rates were 91.5%±3.7%, 75.9%±6.7%, and 72.3%±7.3% in the PTFE group vs. 100%, 77.1%±14.4%, and 77.1%±14.4% in the GSV group (P=0.355; Fig. 2A). The freedom from CD-TLR rate of the SFA as an inflow or outflow vessel did not differ between groups at 1-, 3-, and 5-year follow-up (95.5%±2.7%, 81.2%±6.4%, and 77.3%±7.2% for PTFE vs. 100%, 77.1%±14.4%, and 77.1%±14.4% for GSV, respectively; P=0.591; Fig. 2B).

Fig. 1.

(A) Primary patency rates of polytetrafluoroethylene (PTFE) vs. great saphenous vein (GSV) grafts. The 1-, 3- and 5-year primary patency rates were 92.9%±2.7%, 83.5%±4.4% and 74.0%±5.9% for PTFE and 100%±0.0%, 81.5%±11.9%, and 69.8%±14.9% for GSV. (B) Clinically driven target lesion revascularization (CD-TLR) rates of PTFE vs. GSV grafts. The 1-, 3-, and 5-year CD-TLR rates were 95.1%±2.2%, 86.6%±4.2%, and 76.9%±5.5% for PTFE and 100%±0.0%, 81.5%±11.9%, and 69.8%±14.9% for GSV.

Fig. 2.

(A) Primary patency rates of polytetrafluoroethylene (PTFE) and great saphenous vein (GSV) grafts using the superficial femoral artery as the inflow or outflow vessel. The 1-, 3-, and 5-year primary patency rates were 91.5%±3.7%, 75.9%±6.7%, and 72.3%±7.3% for PTFE and 100%±0.0%, 77.1%±14.4%, and 77.1%±14.4% for GSV. (B) Clinically driven target lesion revascularization (CD-TLR) rates of PTFE and GSV grafts using the superior femoral artery as the inflow or outflow vessel. The 1-, 3-, and 5-year CD-TLR rates were 95.5%±2.7%, 81.2%±6.4%, and 77.3%±7.2% for PTFE and 100%±0.0%, 77.1%±14.4%, and 77.1%±14.4% for GSV (P=0.591).

4) Risk factors for graft occlusion

Patients’ clinical characteristics, disease details, and procedures were analyzed as risk factors for graft occlusion (Table 4). The univariate analysis showed that the risk of FFB graft occlusion was higher among younger patients whose symptoms were claudication rather than chronic limb-threatening ischemia and when the preoperative ABI was high. However, the multivariate analysis revealed that none of these factors increased the risk of FFB occlusion.

Table 4.

Univariate and multivariate analysis of the risk factors for graft occlusion (n=168)

Variable	Univariate analysis				Multivariate analysis
	OR	95% CI	P-value		aOR	95% CI	P-value
Sex, male	3.726	0.476-29.136	0.210		1.729	0.168-17.840	0.646
Lesion (left)	0.525	0.211-1.307	0.166		0.958	0.244-3.758	0.951
Inflow (SFA)	1.458	0.582-3.652	0.421		1.539	0.292-8.116	0.611
Outflow (SFA)	2.085	0.803-5.413	0.131		1.638	0.351-7.646	0.530
Claudication	9.660	1.261-74.013	0.029*		2.577	0.253-26.258	0.424
Diabetes	0.632	0.201-1.984	0.431		0.284	0.045-1.805	0.182
Smokers	1.973	0.752-5.172	0.167		2.440	0.561-10.619	0.235
Hyperlipidemia	2.820	0.512-15.520	0.233		1.353	0.071-25.656	0.840
Hypertension	1.096	0.441-2.726	0.843		2.735	0.598-12.514	0.195
Chronic renal failure	1.690	0.180-15.859	0.646		0.015	0.000-16.478	0.239
Ischemic heart disease	1.424	0.480-4.218	0.524		0.735	0.116-4.668	0.744
Cerebrovascular accident	0.295	0.066-1.321	0.110		0.218	0.025-1.878	0.166
Wound complication	0.387	0.049-3.073	0.369		0.215	0.016-2.876	0.245
Age	0.927	0.880-0.975	0.003*		0.965	0.893-1.043	0.375
Preoperative ABI	26.544	2.804-251.261	0.004*		3.247	0.119-88.795	0.485
Posoperative ABI	11.979	0.848-169.198	0.066		1.642	0.067-40.187	0.761
Operation time	0.997	0.991-1.003	0.316		1.010	0.999-1.020	0.066
Admission period	0.955	0.885-1.030	0.232		0.856	0.673-1.088	0.203
GSV graft	1.124	0.306-4.120	0.860		1.038	0.181-5.948	0.967
Isolated occlusion of CIA	1.931	0.584-6.387	0.281		0.625	0.097-4.032	0.621
Isolated occlusion of EIA	1.248	0.451-3.449	0.669		1.837	0.394-8.567	0.439

OR, odds ratio; CI, confidence interval; aOR, adjusted odds ratio; SFA, superficial femoral artery; ABI, ankle-brachial index; GSV, great saphenous vein; CIA, common iliacl artery; EIA, external iliac artery.

*Statistically significant at P<0.05.

DISCUSSION

Despite advancements in endovascular equipment and techniques for unilateral iliac occlusive disease, FFB is effective and safe when an endovascular approach has failed or is deemed impossible and the patient is at high risk for abdominal surgery and general anesthesia. Moreover, FFB can also be useful in patients with intra-abdominal infection, including prosthetic graft infection or infected aneurysm, in conjunction with axillofemoral bypass or infected aneurysm exclusion. As endovascular aneurysm repair (EVAR) for abdominal aortic aneurysm has increased, FFB has become an option for unilateral limb occlusion after EVAR.

PTFE graft, Dacron graft, GSV, femoral vein, and cryopreserved femoral vein can be FFB conduits. A randomized trial reported no difference in the patency of Dacron grafts and PTFE grafts for FFB [4]. Nguyen et al. [5] reported that femoral veins in FFB (overall cohort of isolated FFB and FFB in conjunction with axillobifemoral bypass) have superior primary patency and similar complication rates to PTFE grafts (3-year patency: 100% vs. 69.8%). In regard to infrainguinal bypass, the procedures are mainly performed using GSV as an autologous conduit. A Cochrane review reported moderate-quality evidence of improved long-term (60 months) primary patency for autologous vein grafts compared to prosthetic materials for above-knee bypass [6]. Although current study is not a true comparative analysis due to the different surgical situations and anastomosis sites, it provides a relatively large number of FFB outcomes using GSV grafts. Although SFA was more frequently used as the inflow and outflow vessels and redo FFB was more common in GSV grafts, there was no significant difference in primary patency and CD-TLR rates between GSV grafts and PTFE grafts. Also, in a subgroup analysis of previously introduced study [5], isolated FFB using PTFE vs. the femoral vein revealed no significant difference in primary patency or secondary patency between them at 1, 2, and 3 years. This finding is consistent with our study, which showed no difference in patency between autologous vein grafts and artificial grafts for isolated FFB. However, in this study, the GSV was used as the autogenous conduit instead of the femoral vein. Therefore, if the GSV is of sufficient size for bypass surgery, it may be the preferred autogenous conduit given the time-consuming nature and higher complication rates associated with femoral vein harvesting. Decades of experience with FFB have been accumulated, and several studies have reported 5-year FFB graft patency rates of 60%-73.3% [7-11]. In our study, the 5-year patency of FFB grafts was 74.0% for PTFE grafts and 69.8% for GSV grafts, and these results are considered similar to other studies.

We noted significantly greater SFA inflow (P<0.001) and outflow (P<0.001) with GSV vs. PTFE grafts. This is because the GSV graft was more commonly used for repeat FFB when the previous PTFE graft connected to the bilateral common femoral artery (CFA) was occluded. If the CFA could not be used as the inflow or outflow due to a previous wound or infection, GSV grafts were preferentially considered as conduits for size matching, particularly when using SFA with a small diameter. In our study, the primary patency and freedom from CD-TLR rate in the subgroup analysis of SFA as an inflow or outflow demonstrated no intergroup difference.

There was no intergroup difference between GSV and PTFE graft in primary patency or CD-TLR rate over 5 years of follow-up. Therefore, PTFE grafts may be preferred for FFB because the operation is relatively simple and requires less time. However, GSV can be useful in cases in which autologous veins could be preferentially considered because of previous PTFE graft occlusion, infection, or difficult surgical access to the wound.

A univariate analysis of the risk factors of graft occlusion in our study revealed that the risk of FFB graft occlusion was increased in patients of younger age, with claudication rather than limb ischemia, and with a higher preoperative ABI. Considering the aforementioned risk factors, it can be inferred that competitive flow may cause higher graft thrombosis rates. However, the multivariate analysis showed that these factors had no effect on patency. Several studies have examined the risk factors for FFB patency. However, in a study regarding risk factor analysis with 124 FFBs, the severity of limb ischemia and outflow status did not affect the immediate or secondary thromboses in the mid-term like the current study [12]. Several limitations exist in this retrospective study. First, the study results were based on a single-center experience with a brief follow-up period and fewer patients, especially in GSV graft. Second, this study is not an actual comparative study. For selecting GSV and PTFE graft for FFB, some factors such as vessel diameter and redo status affected the decision-making. Therefore, selection bias is inevitable. Finally, some critical factors associated with graft occlusions, such as outflow status and postoperative medications, including anticoagulants and antiplatelet agents, could not be presented due to the limitation of data; hence, the study results should be interpreted cautiously.

CONCLUSION

In our study, FFB using artificial or autologous GSV grafts was effective, with a 5-year patency rate of nearly 70%. There was no difference in 5-year primary patency or CD-TLR rates between GSV and PTFE grafts; however, the GSV graft may be the ideal conduit according to the anatomy of the inflow and outflow vessels and the patient’s condition.

Funding Statement

FUNDING None.