Outcome of Infragenicular Prosthetic Grafts with and Without Distal Anastomotic Modication – A Retrospective Single Center Study

Background: Peripheral revascularization with below-knee bypass (REBEL) utilizing prosthetic grafts demonstrate inferior patency and amputation-free survival rates, as compared to venous conduits. Because of improved hemodynamics, adding a venous modication to distal anastomoses in REBEL with prosthetic grafts is assumed to improve both patency and amputation-free survival. The aim of this study was to assess the performance of prosthetic grafts with and without venous modications of the distal anastomosis (MOD) and compare the results with venous conduits, in terms of patency rates, amputation prevention and survival. Methods: The present single center retrospective, non-randomized observational study involved 221 REBEL in 201 patients. Data was collected by viewing hospital records and follow-up clinical and duplex ultrasound examinations (up to 48 months). For analysis, the patients were divided in 3 groups; REBEL using prosthetic grafts with MOD (PGY) or without (PGN) anastomotic modication and REBEL using plain venous conduits without anastomotic modications (VG). Primary patency, assisted primary patency, secondary patency, amputation-free survival and overall survival was analyzed. Results: The median follow-up was 27 months. The most common site of distal anastomosis in PGY were the crural arteries (77.8%) and 50% of PGY patients had already undergone vein-harvesting for peripheral revascularization. Average duration of surgery for PGY was 284 minutes, >70 minutes more as compared to PGN (p=0,001). In PGY a statistically insignicant tendency to decreased primary and secondary patency rates as compared to PGN was observed at 12 and 36 months. PGY displayed a signicant lower amputation-free survival rate at 12 and 48 months as compared to PGN (62.9% vs 87.6% and 43.4% vs 69.75%; p=0.038). Overall survival for the three groups was not different (p=0.375). Conclusion: To summarize, modication of the distal anastomotic site using venous cuff techniques did not result in improved patency or amputation prevention. Nonetheless, from a technical view of facilitating the creation of anastomoses between relatively rigid, large bore prosthetic grafts and calcied delicate small diameter infrapopliteal arteries, anastomotic modelling offers potential anatomic and hemodynamic advantages, thus improving outcome on long term follow up. However, prospective randomized studies are required to corroborate this hypothesis.


Introduction:
Since the rst successful creation of a bypass graft for the treatment of peripheral arterial occlusive disease (PAOD) by Jean Kunlin in 1948, techniques for peripheral revascularization have come a long way (1). Notwithstanding the recent evolution and implementation of endovascular procedures and strategies calling for the rst line use of interventional recanalization, surgical revascularization by means of bypass grafting still bodes its merits, particularly for complex lesions of the super cial femoral (SFA), popliteal and crural arteries (2). For revascularization of below knee arteries (REBEL), the ipsilateral greater saphenous vein (GSV) remains the conduit of choice (3,4). However, prosthetic grafts still play an important role in cases where suitable autogenous veins are not available. Yet, diameter and compliance mismatch as well as technical di culties to surgically connect a relatively rigid graft to a delicate, calci ed, low ow crural artery pose a signi cant obstacle to satisfactory mid-and long-term outcome. Modelling of the distal anastomosis using various cuff-techniques was thought to simplify the surgical procedure and improve patency of alloplastic bypass grafts. Since the outcome of cuff techniques still remains unsettled, the present study was devised to assess the function of prosthetic grafts with and without modi cation of the distal anastomosis (MOD) as compared to the use of plain venous conduits.

Materials And Methods:
This single center, retrospective, non-randomized, observational study (approved by the ethics committee of the medical council in Hessen) was carried out in patients undergoing REBEL for the treatment of PAOD or peripheral aneurysms (PA) from April 2009 to April 2013. This retrospective analysis involved 221 REBEL in 201 patients. Patients undergoing REBEL after trauma or embolic episodes without PAOD or PA and patients with distal anastomosis proximal to the P3 Segment of the popliteal artery were excluded from the study. Hospital records were reviewed to collect the data. Follow-up was planned and performed prospectively for up to 48 Months by means of clinical examination and/or by phone and was aided by duplex examination (GE LogiQ7, GE Healthcare, USA). A written consent was obtained from all patients or their legal representatives before enrolment into the study.
The following parameters were tabulated (Microsoft Excel, Redmond, WA, USA) and analyzed: Patients' demographics, comorbidities, Fontaine and Rutherford categorization, ankle brachial index (ABI), duplex ultrasound ndings, angiographic ndings, intraoperative details (operative time, localization of the proximal and distal anastomosis, bypass material and diameter, location of the anastomosis and MOD if present) were recorded and tabulated (Microsoft Excel, Redmond, WA, USA). REBEL were categorized based on the type of graft and distal anastomosis modi cations. The three categories were autogenous venous graft (VG), prosthetic graft without MOD (PGN) and with MOD (PGY).
Data analysis: Survival, limb salvage and patency rates were analyzed by Kaplan-Meier-survival curves with log-rank test. IBM SPSS 22.0 (Armonk, NY: IBM Corp.) was used for data analysis Results: The median follow-up was 27 months. In 23 patients the data records were complete for the planned follow-up. Follow-up clinical and duplex ultrasound examinations were performed in 70 patients and 84 bypass grafts. Thirty-two patients were followed up by phone. Sixty-ve patients succumbed during the planned follow-up period. Eleven patients were lost to follow-up. The patients' demographics and comorbidities are illustrated in Table 1. The number of males in VG and PGN was statistically higher (Pearson-Chi2-Test; p < 0,005). Age distribution was comparable in the three groups (Kruskal-Wallis-Test; p = 0,271). There was no signi cant difference in the presence of individual risk factors in the three groups. Hypertension and hyperlipidemia were the most frequently occurring risk factors. All patients underwent preoperative venous mapping with the help of duplex ultrasound. Out of 48 patients with aneurysm, 23 were asymptomatic and 25 were symptomatic. Most of the patients presented with critical limb ischemia (rest-pain n = 44 and tissue loss n = 146). Only 8 patients underwent REBEL for lifestyle limiting claudication. The reasons for opting for a prosthetic graft (n = 62) were; small vein caliber < 3mm (n = 30), vein already harvested for a peripheral (n = 23) or coronary bypass (n = 7) surgery or varicose veins (n = 2). Venous grafts were used in both reversed (n = 147/159) and non-reversed (12/159) techniques. MOD in the PGY group are depicted in Fig. 1. Kruskal-Wallis-Test showed a signi cant difference (p = 0.001) in the distribution of operative time for the three groups. PGN group had signi cantly higher number of REBEL with operative time < 180mins and lower number of REBELs with operative time > 300 mins as compared to the PGY group (Pearson-chi2 Test p = 0.015). This is in coherence with expectations that revascularization with venous graft and MOD would be more time consuming. Comparison of the three groups in reference to secondary patency rate at 36 months, revealed a p-value of 0,058. However, paired-log-rank test showed a signi cant difference in the secondary patency rates ( Fig. 3) between the VG and PGY (p = 0,017) group but this was insigni cant (p = 0,316) between the PGN and PGY group. There was a signi cant difference in the amputation free survival between the three groups at 48 months (Log-rank-test p < 0.001) (Fig. 3). The patients in PGY group showed signi cantly inferior amputation free survival as compared to VG and PGN at 48 months (p < 0,001 and p = 0,038). Discussion: Despite advancements in the eld of endovascular surgery, bypass surgery is an absolute essential part of vascular surgeons' armamentarium. Regardless the superiority of venous conduit (3,4), prosthetic grafts are a viable alternative when the vein is disadvantageous (5), for e.g. in cases of small vein caliber, varicose veins or thrombophlebitis or unavailability because of previous harvesting for coronary or peripheral bypass (6). The initial results of infrapopliteal revascularizations with prosthetic grafts were poor as compared to venous conduits (7,8,9). Several strategies have been utilized to improve the performance and patency of prosthetic grafts. Some of these are: venous modi cations of the distal anastomosis (MOD) (10,11,12,5), heparin bonding at the luminal surface of PTFE grafts (Hb-ePTFE, Propaten; W.L. Gore & Associates, Inc.) (13,14), distal AV-stula (15) and grafts reinforced with rings (16).
MOD are advantageous in decreasing the intimal hyperplasia (17), by improving the ow characteristics (18) and reducing compliance-mismatch and shear stress (19) between relatively rigid prosthetic graft and undersized, delicate, calci ed crural vessels. Experimental studies show that the intimal hyperplasia is most prominent at the transition between graft and MOD and least at the recipient artery (20). This spares the artery in an acceptable state for a redo revascularization and may further improve the secondary patency rates (12). The classical MOD were improved with time (21,22). A boot shaped modi cation was described in 2000 and was known as St. Mary's boot (23 administered 160mg of aspirin per day in the postoperative period. Warfarin anticoagulation was only given to patients who received it preoperatively and in redo venous bypass patients. The patency rates can be compared with the help of Kaplan-meier plots (Fig. 2,3). The primary patency rates for PGY vs PGN were 50% vs 74.2% at 12 months and 36.9% vs 50.1% at 36 months respectively. A similar trend was observed with the assisted primary patency rates. The difference between the PGY and femoro-crural reconstructions with Hb-ePTFE and distal vein patch and reported a 1-year primary patency of 75.4%, as compared to 86% for venous grafts. The primary, assisted primary and secondary patency rates at 12 and 36 months for PGN group were comparable to VG ( Table 3). The secondary patency rates at 12 and 36 months in PGY group were signi cantly inferior as compared to VG (59.7% vs 76.1% and 44.6% vs 67.5%; p = 0.017) but were relatively inferior, even though not signi cant as compared to PGN group (59.7% vs 78.5% and 44.6% vs 59.6%; p = 0.0316). A signi cantly higher rate of amputation was observed in the PGY group. Amputation-free Survival in PGY group was signi cantly lower at 12 (62.9% vs 87.6%; p = 0.038) and at 48months (43.4% vs 69.75%; p < 0.001) as compared to the PGN group. The three groups didn't show any signi cant difference (p = 0.375) in overall survival (Table 3, Supplemental Fig. 1).
In SCAMICOS (29) the primary patency at 36 months was 26% with MOD and 43% without MOD in femoro-popliteal reconstructions and 20% and 17% for femoro-crural reconstructions respectively. The amputation-free survival was better for the revascularizations with MOD both for femoro-popliteal and femoro-crural revascularizations, but this was not statistically signi cant. Similar to our study, the SCAMICOS (29) could not demonstrate any bene t of MOD for below knee revascularizations with PTFE. A meta-analysis (30) published in 2012 involving 885 patients with below-knee popliteal and crural bypasses. This metanalysis showed no signi cant improvement for primary patency or limb survival at 3 years for femoro-crural bypasses with venous cuff. In our study, we achieved good patency rates at 1 month (Table 3) for all three groups. These were better for the PGY group as compared to PGN, even though this was not statistically signi cant. This is an evidence of technical success and expertise of our surgeons. These ndings are also in coherence with the reanalysis in the SCAMICOS cohort (31).
Keeping the abovementioned results in mind, we would like to indicate towards the fact that the patients undergoing REBEL with MOD suffer from advanced PAD with prior single or multiple revascularizations (surgical or endovascular The study was approved by the ethics committee of the medical council in Hessen.

Statement of informed consent:
A written consent was obtained from all patients or their legal representatives before enrolment into the study.
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.  Kaplan-meier plots. No signi cant difference for primary patency (p=0.226; left) and assisted primary patency (p=0.074; right) between the three groups. Figure 3 left: Kaplan-meier plots for secondary patency. Signi cant difference (p=0.017) between VG and PGY but no signi cant between PGY and PGN (p=0.316) right: Kaplan-meier plots for amputation free survival.
Signi cant difference between the three groups (p<0.001).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.