- Email: [email protected]
Improved patency after axillofemoral bypass for aortoiliac occlusive disease
Improved patency after axillofemoral bypass for aortoiliac occlusive disease Russell H. Samson, MD, David P. Showalter, MD, Michael R. Lepore Jr, MD, Deepak G. Nair, MD, Douglas A. Dorsay, MD, and Ricardo E. Morales, PA-C, Sarasota, Fla
ABSTRACT Objective: Axillofemoral bypasses (AxFBs) have been used since 1962 to treat aortoiliac disease. In the past, reported patency rates (37%-76%) for these extra-anatomic grafts have been inferior to those for anatomic aortic grafting. Reported low survival rates after AxFB (40%-50%) have confirmed that these procedures have been used primarily in patients at high risk for complications from aortofemoral bypass. However, modern medical and anesthesia management, preoperative scanning, donor artery preparation, postoperative graft surveillance, and graft technology may improve outcomes after AxFB, possibly supporting expansion of its use. We therefore report our last 15-year experience with AxFB. Methods: Ring-reinforced, 8-mm expanded polytetrafluoroethylene grafts were used in all cases. The cross-femoral limb of axillobifemoral bypass (AxBFB) grafts was preconstructed. Heparin was administered intraoperatively, with protamine reversal. Loss of primary patency was defined as graft thrombosis of part or all of the inserted graft. Five-year primary patency rates were calculated by Kaplan-Meier analysis. Results: Between February 1991 and June 2016, a total of 161 grafts were inserted (85 AxBFBs and 76 axillounifemoral bypasses [AxUFBs]) in 91 male and 70 female patients (median age, 72.6 years; mean age, 73 years; range, 41-94 years). Indications for treatment were rest pain (49.6%), ischemic lesions (26%), claudication (22.3%), failed prior revascularization (9.3%), infection (3.7%), and dissecting aneurysm (1.2%). Reasons for performing AxFB rather than aortofemoral bypass were hostile aorta (44.1%), high risk (19.2%), prior failed reconstruction (12.4%), advanced age (8.7%), infection (4.3%), hostile abdomen (4.3%), aortic dissection (0.6%), and morbid obesity (0.6%). During follow up, 63 patients died, 17 within the first year; but only 3 patients died within 30 days of surgery (performed to treat an acute aortic occlusion). The 5-year survival rate was 55%. Five-year patency rates were 83.7% for all procedures, 81.8% for AxBFB, and 85.5% for AxUFB; the difference between AxBFB and AxUFB was not significant. Conclusions: Our data indicate that AxBFB and AxUFB performed with the use of modern protocols and technology may render them an acceptable valid primary intervention in patients in whom endovascular treatment has failed or is unlikely to offer long-term success. The simplicity of performing these grafts and their low mortality and morbidity lend their application to surgeons with limited open aortic experience. Because AxUFB and AxBFB have similar patency rates, AxBFB should be reserved for bilateral indications. (J Vasc Surg 2018;-:1-8.) Keywords: Axillobifemoral grafts; Axillofemoral grafts; Axillounifemoral grafts; Long-term patency; Long-term survival
Severe aortoiliac occlusive disease (AIOD) may be managed with aortofemoral bypass (AoFB), with extraanatomic bypass (axillofemoral bypass [AxFB], either axillounifemoral bypass [AxUFB] or axillobifemoral bypass [AxBFB], and femorofemoral bypass), or endovascularly (angioplasty and stenting). AoFB has traditionally been considered the standard surgical treatment because of its effectiveness in relieving symptoms and its reported 5-year patency rates of about 80% or higher.1,2 However,
From the Sarasota Vascular Specialists, Affiliated with Florida State University. Author conflict of interest: R.H.S. is a consultant for W. L. Gore & Associates. Correspondence: Russell H. Samson, MD, Clinical Professor of Surgery (Vascular), Florida State University Medical School, 600 N Cattleman Rd, Ste 220, Sarasota, FL 34232 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2018 Published by Elsevier Inc. on behalf of the Society for Vascular Surgery. https://doi.org/10.1016/j.jvs.2018.01.061
alternatives to AoFB have long been sought, especially for older and high-risk patients, because of the considerable operative and postoperative morbidity and hospital stay associated with the procedure, which requires opening of the abdomen.1,3,4 Although the use of endovascular therapy for AIOD has expanded dramatically since its introduction in the 1990s,5,6 systematic reviews and meta-analyses have found that the 5-year patency rates associated with this approach (60%-80%) remain lower than those achieved with surgical revascularization.1,7 Therefore, patients in whom angioplasty and stenting are used exclusively may require repeated procedures. The history of AxFB, which was first described almost simultaneously by Louw8 and Blaisdell and Hall,9 has been characterized by a considerable discrepancy in outcomes obtained.10,11 Early series generally had poor outcomes, with 4- or 5-year primary patency rates being 50% or lower in several studies10,12-14 and the highest rates being about 70%.15 Patient survival rates were also low, although it has often been observed that this finding is probably related primarily to the population of patients 1
2
Journal of Vascular Surgery
Samson et al
---
in whom AxFB has generally been used (ie, poor candidates for open aortic surgery because of age, cardiopulmonary disease, renal insufficiency, infection, hostile aorta or abdomen, and other high-risk conditions).3,11,14,16,17 In the 1990s, reports of better outcomes with AxFB began to appear. For example, in the study by Taylor et al18 of 184 patients who underwent AxUFB or AxBFB using externally supported, 8-mm-diameter expanded polytetrafluoroethylene (ePTFE) grafts, the 5-year primary patency rate was 71%. Passman et al19 retrospectively compared 108 patients who had undergone AxFB with 139 who had AoFB and observed a 5-year primary patency rate of 74% in the AxFB group and 80% in the AoFB group, a difference that was not significant. The patient survival rate was lower in the AxFB group (45% vs 72%), primarily because it contained a larger proportion of high-risk patients. The rate of major postoperative complications, however, was significantly higher in the AoFB group (9.2% vs 19.4%; P < .05). To our knowledge, a large AxFB series with a follow-up time of 5 years has not been reported since that of Passman et al. We here describe a series of AxFB procedures that began in 1991, well within the “modern era” of vascular surgery. We hypothesized that our outcomes would indicate a contemporary sustaineddand possibly enhanceddimprovement in AxFB outcomes.
METHODS Study design. The study was a retrospective evaluation of prospectively collected data obtained from a clinical experience between February 1991 and June 2016. Consent of the patients for the procedures was obtained per standard protocol. However, because this study was not performed as part of an ongoing clinical investigation or trial, consent for inclusion in the ongoing investigation was not necessary, nor was Institutional Review Board approval required. Four surgeons were involved in inserting the grafts, although the majority were performed by R.H.S. and D.P.S. Demographic data, risk factors, indications for surgery, previous therapy for vascular disease, and reasons for choosing axillary flow were obtained from the enrolled patients’ medical records and entered into a database for analysis (Atrium Medical Inc, Nashua, NH). Noninvasive and clinical preoperative workup. All patients had routine physiologic studies of the lower and upper extremity arteries. Most patients also had diagnostic arteriography, either transfemoral or through brachial puncture, to plan surgery but also to exclude a possible endovascular or open aortic procedure. Computed tomography angiography (CTA) evaluation of the aorta was obtained in low-risk patients to evaluate the status of the aorta to determine whether
2018
ARTICLE HIGHLIGHTS d
d
d
Type of Research: Retrospective, single-center, cohort study Take Home Message: Axillofemoral bypasses were performed in 161 patients, with three deaths within the first 30 days. The 5-year survival and patency rates were 55% and 83.7%, with no difference between axillofemoral and axillobifemoral bypasses. Recommendation: This study suggests that axillofemoral bypass can be used as a primary treatment of aortoiliac disease with good 5-year patency rates.
the patient could undergo aortobifemoral bypass or would require an AxBFB graft because of a hostile aorta. The aorta was considered hostile or inoperable if it was so heavily calcified that clamping was considered to be dangerous or impossible. Originally, computed tomography evaluation of the donor axillary arteries was performed only if there was a disparity in blood pressure between the two arms or if the patient complained of arm discomfort. In recent years, CTA evaluation of the donor artery has been obtained in all patients. Surgical technique. All procedures were performed under general anesthesia. All patients received a ringreinforced, 8-mm-diameter ePTFE graft (W. L. Gore & Associates, Flagstaff, Ariz); in those who underwent AxBFB, the femoral crossover graft segment was preconstructed (ie, it did not require a surgical anastomosis to the main limb). Since 2017, the preconstructed AxBFB grafts have been heparin bonded, but to exclude this confounding factor, this review was restricted to grafts performed before 2017. However, from 2007, the AxUFB grafts were heparin bonded using the CBAS technology acquired by W. L. Gore & Associates. The surgical technique warrants description because this may differ between surgeons, institutions, and prior instructions for use. A rolled sheet is placed transversely under the shoulders, and the donor arm is abducted 90 degrees and placed on an arm board (Fig 1). Antiseptic preparation starts at the chin and ends in the midthigh bilaterally. An adhesive povidone-iodine (Betadine)-impregnated plastic drape is applied to the chest wall, abdomen, and thighs to isolate the graft from any contamination from the skin. The femoral arteries are exposed using oblique groin incisions. The axillary artery is exposed by splitting the fibers of the pectoralis major and dividing the pectoralis minor (Fig 1). An extra-anatomic tunnel is created using the 8-mm Scanlan (Scanlan International, St. Paul, Minn) plastic tunneler starting under the pectoralis major. We do not advise tunneling from the groin toward the shoulder because we have heard of anecdotal occurrences in which the tunneler has inadvertently penetrated the chest cavity
Journal of Vascular Surgery Volume
-,
Number
Samson et al
3
-
Fig 1. Artist’s rendition of the positioning of the arm and the incisions required for a right axillobifemoral bypass (AxBFB) graft. The inset demonstrates the location of the anastomosis, which usually requires division of the pectoralis minor muscle for exposure. Furthermore, the graft is anastomosed to the anterior wall of the artery with enough redundancy to allow the patient to fully raise the arm above the head without any untoward traction on the anastomosis.
in such an approach. When an AxBFB graft is being constructed, a 2- to 3-cm counterincision is placed in the lower flank as high up on the flank as possible but still distal enough to allow the short limb of the AxBFB graft to reach from the counterincision to the ipsilateral femoral artery. The vertical limb of the AxBFB graft is then brought through the already inserted tunneler from the counterincision into the axillary artery incision. The femoral limbs of the graft are brought into the femoral incisions using an aneurysm clamp passed subcutaneously from the groin incisions to the counterincision. The configuration then approaches an inverted Y shape rather than the h shape that usually occurs with axillary-femoral-femoral-femoral configurations (Fig 2). Patients are then heparinized (5000 units of heparin for all patients irrespective of weight). The axillary anastomosis is performed first. The incision is placed anteriorly on top of the axillary artery with a lazy C-shaped turn such that if the patient were to elevate the arm directly above the head, there would be no tension on this anastomosis (Fig 1). We do not advise placing the anastomosis on the underside of the axillary artery as described in older texts. These texts assume that the graft comes into the artery in the same plane, whereas it actually dives down from the anterior chest wall to the front of the artery. The ipsilateral femoral anastomosis is performed next but not fully completed. A hydrogel clamp is then placed on the contralateral limb in the counterincision, and the graft is flushed by releasing the axillary artery flow briefly. The ipsilateral anastomosis is completed and flow restored to that side. The graft is never flushed with saline because we believe it may result in subsequent weeping of plasma
Fig 2. Photograph of a patient having undergone a right axillobifemoral bypass (AxBFB) graft. Note the inverted Y configuration due to the crossover of the preconstructed graft occurring in the lower right flank, with the counterincision still pink after the recent surgery.
through the graft pores. The contralateral anastomosis is then completed in similar fashion. We make a point of sewing the distal anastomosis across the common femoral onto the first part of the deep femoral artery or entirely onto the deep femoral artery in the presence of severe common femoral disease or occlusion of the superficial femoral artery. In some patients, we will reconstruct a femoral bifurcation if the superficial femoral artery and deep femoral artery are patent but the common femoral is severely diseased. Protamine reversal is used in all patients. Follow-up protocol. Patients were encouraged to follow up every 6 months for a clinical evaluation, ankle-brachial indices, and color duplex ultrasound evaluation of the graft as well as of the inflow and outflow arteries. However, not all patients complied. Accordingly, color duplex ultrasound and ankle-brachial indices were obtained outside of the regular follow-up if there was any clinical indication of incipient graft failure. If this was confirmed, CTA or diagnostic arteriography was added. Prophylactic anticoagulants were not used unless required for indications unrelated to graft patency. Aspirin was almost uniformly used and clopidogrel prescribed in later years. However, we do not have accurate information as to patients’ compliance
4
Journal of Vascular Surgery
Samson et al
---
2018
because this was a retrospective analysis. Loss of patency (graft failure) was defined as occlusion of any graft limb as indicated by duplex ultrasound scanning, regardless of whether a subsequent intervention restored patency. Data analysis. Descriptive characteristics of the patients were reported as median (range) or number and percentage. Kaplan-Meier analysis was used to calculate graft patency rates to 5 years and patient survival rate for up to 10 years.
RESULTS Between February 1991 and June 2016, a total of 161 grafts were inserted (85 AxBFBs, 76 AxUFBs) in 91 male and 70 female patients (median age, 72.6 years; mean age, 73 years; range, 41-94 years). Thirty-one of the patients had previously undergone procedures, including femorofemoral bypass (9 patients), aortoiliacfemoral bypass (7), infrainguinal bypass (11), iliac stenting (12), and iliac percutaneous transluminal angioplasty (2). Indications for treatment were rest pain (49.6%), claudication (22.3%), gangrene (14.9%), ulceration (11.1%), failing graft (9.3%), infection (3.7%), and dissecting aneurysm (1.2%). In the majority of cases, axillary flow was chosen because the patient had a hostile aorta (44.1%). Other indications included high risk (19.2%), prior failed revascularization (12.4%), advanced age (8.7%), hostile abdomen (4.3%), infected graft (4.3%), acute dissection (0.6%), and morbid obesity (0.6%). During follow-up, 63 patients died, 17 within the first year, of whom 3 died within 30 days of surgery performed to treat an acute aortic occlusion. No patient died intraoperatively. Twenty-seven patients (16.8%) required admission to the intensive care unit after surgery. The mean length of the hospital stay was 6 days (median, 3 days; range, 2-54 days). The mean follow-up time in the series was 27.8 months (maximum, 160 months). Thirty-three were lost to followup, but only two were lost to follow-up within 1 year of surgery. The 5-year survival rate was 55% (Fig 3). Six patients had occlusion of a graft limb, which was treated with placement of a new femorofemoral crossover graft from the vertical limb (three patients), a new AxUFB or AxBFB from the contralateral axillary artery (one each), or an AoFB (one). One patient had a functioning AxBFB removed and replaced by a new graft from the contralateral side because of the formation of a massive seroma involving the entire graft. Three grafts required an inflow procedure to maintain patencydplacement of a stent in the subclavian artery, percutaneous transluminal angioplasty at the proximal anastomosis, or bypass from the carotid artery to the vertical graft limb (one each). No patient required an amputation during the follow-up. The 1- and 5-year primary patency rates for all grafts were 96.3% and 83.7%, respectively (Fig 4). The 1- and
Fig 3. Patient survival up to 10 years after axillofemoral bypass (AxFB). The number at risk at each interval is shown.
5-year rates for AxBFB grafts were 98.4% and 81.8% (Fig 5); those for AxUFB grafts were 93.8% and 85.5% (Fig 6). There was no significant difference between AxBFB and AxUFB grafts in patency at either 1 year or 5 years (P > .5).
DISCUSSION AoFB provides high long-term patency rates, but vascular surgeons remain reluctant to perform it in high-risk patients because of its morbidity, mortality, and relatively long hospital stay.1,2,4,6,20,21 Although, in the past, graft patency after extra-anatomic bypass was disappointing, AxFB is an established alternative to AoFB, especially in patients with graft infection, infected aortic aneurysm, and aortoenteric fistula.1,3,9,11 The recent encouraging outcomes with AxFB,18,19,22 including ours, suggest that an expansion of its role in treating AIOD is warranted. Our series of patients with AIOD in whom AxFB was performed between 1991 and 2016 had one of the highest 5-year primary patency rates (83.7%) ever reported for this procedure. Moreover, the rates after AxUFB and AxBFB were also high (85.5% and 81.8%, respectively) and did not differ significantly. No patient died during surgery, and only three patients died within 30 days of surgery. Furthermore, no patient required an amputation during a mean follow-up of 27.8 months. Accordingly, our results show that excellent outcomes can now be achieved with these extra-anatomic bypasses. Some may suggest that an experience dating to the 1990s infers that modern techniques may not have been used. However, it was then that we adopted most of the techniques that we maintained throughout the rest of this experience. Admittedly, there have been advances in medical therapy and CTA technology that were more commonly used in later years, but we could not identify a specific time to divide the group. Furthermore, only 27 bypasses were constructed before the year 2000,
Journal of Vascular Surgery Volume
-,
Number
Samson et al
5
-
Fig 4. Kaplan-Meier estimate of primary patency for up to 5 years after all axillofemoral bypasses (AxFBs; axillobifemoral bypass [AxBFB] and axillounifemoral bypass [AxUFB]). The number at risk at each interval is shown. The error bars indicate 95% confidence intervals.
Fig 5. Kaplan-Meier estimate of primary patency for up to 5 years after axillobifemoral bypass (AxBFB). The number at risk at each interval is shown. The error bars indicate 95% confidence intervals.
with only 3 failures before censure due to death or loss to follow-up. No randomized trial has compared AoFB with either AxFB or angioplasty and stenting or AxFB with the endovascular approach, although several retrospective comparison studies have been reported.14,15,19,23-26 The results of investigations of this nature can be difficult to assess because of differences in populations of patients, including indications for treatment, variations in surgical techniques, and other factors. This problem is especially relevant to studies of AIOD treatment because high-risk patients are more likely to undergo AxFB than AoFB,19,24 and outcomes of AxFB have improved in recent decades. Interestingly, despite this improvement and the marked increase in use of endovascular therapy for AIOD, no study has compared AxFB with the endovascular approach. In their 2010 systematic review of 19 studies of endovascular AIOD therapy published in 2000 to 2009, Jongkind et al7 observed that the hospital stay ranged from 1 day to 4.8 days, the 30-day mortality rate from 1.2% to 6.7%, and the morbidity rate from 3% to 45%. For the eight studies in which 5-year primary patency was reported, the rate ranged from 60% to 86%. A 2013 systematic review and meta-analysis of 28 studies of endovascular treatment of AIOD (1625 patients) that were reported between 1989 and 2010 found that overall, the mean length of hospitalization was 4 days; the complication and 30-day mortality rates were 13.0% and 0.7%, respectively; and the pooled 5-year primary patency rate was 71.4%.2 Therefore, our outcomes with AxFB have been either similar or superior to those reported for endovascular therapy. Despite its popularity, the endovascular approach has some disadvantages. Placement of a stent in a heavily
calcified vessel can be challenging or impossible,19,22,25 and arterial perforation or dissection can occur.2,7 Loss of primary patency often results from stent thrombosis or in-stent stenosis.1 In addition, we and others16,21,22 are concerned that a dependence on endovascular therapy may result in a decrease in the number of surgeons competent to perform open proceduresdat a time when the aging of the U.S. population has substantially increased the number of patients requiring revascularization. The shift from AoFB to endovascular therapy has already resulted in a decline in AIOD operative experience among vascular surgery trainees.4 Expanded use of AxFB, a procedure that is simpler, easier to learn, less invasive, and less likely to have surgical complications than AoFB, may be a valid alternative surgical option for less experienced aortic surgeons. Several factors may have contributed to the improved outcomes with extra-anatomic bypass in our study and other recent series,18,19 including better anesthesia management, new antiplatelet and statin protocols, more specialized training and board certification of vascular surgeons, and improvements in duplex ultrasound for graft surveillance. Routine use of preoperative CTA to evaluate the aorta and the donor and recipient arteries may be especially critical. Unsuspected inflow disease has been reported to affect 16% to 25% of candidates for AxFB.10,27 Inflow disease that is detected by preoperative duplex ultrasound scans or CTA can now be treated endovascularly before AxFB is performed, thereby improving the results of the operative procedure.27 Similarly, duplex ultrasound detection of a subsequent hemodynamically significant stenosis in the inflow or outflow of these grafts can also allow preventive endovascular treatments that will prolong patency.
6
Journal of Vascular Surgery
Samson et al
---
Fig 6. Kaplan-Meier estimate of primary patency for up to 5 years after axillounifemoral bypass (AxUFB). The number at risk at each interval is shown. The error bars indicate 95% confidence intervals.
Several studies have evaluated whether the material, type (supported vs not supported), or configuration (AxUFB vs AxBFB) of the graft affects outcomes of AxFB. Dacron grafts were used in early AxFB series,3,12,28-30 but grafts made of ePTFE are now the most widely employed conduits.11,24 Studies that compared the two materials with respect to AxFB outcomes, including the only randomized trial to address this issue,31 have generally reported no significant differences in patency.28,32 However, externally supported grafts, such as the ring-reinforced ePTFE grafts we use, have been found to have a patency advantage over nonsupported grafts,18,20,29,33 possibly because they resist compression or kinking that leads to occlusion.18,20 Whether AxUFB or AxBFB provides better patency has long been debated. Some early studies supported the idea that an AxBFB configuration would enhance patency, presumably by increasing the flow rate in the main graft limb.13,14,17,28 More recent studies, however, including ours, found no difference in patency between AxUFB and AxBFB.10,12,22,34 In addition, Martin and Katz22 observed that 60% of graft occlusions after AxFB occur in the femorofemoral crossover limb. We agree with Martin and Katz that such findings indicate that AxBFB should be reserved for use in patients with bilateral disease. For patients who do require AxBFB, preconstructed grafts, like those we use, may offer not only enhanced convenience and a reduction in operating time but also a patency benefit related to improved flow dynamics resulting from the absence of a surgical anastomosis between the axillofemoral and femorofemoral graft limbs. This possibility has not been formally studied, but it merits investigation. Heparin-bonded ePTFE grafts (CBAS, Gore Propaten Vascular Grafts; W. L. Gore & Associates) have been used for >10 years in patients requiring a prosthetic
2018
femoropopliteal, femorocrural, or tibial artery bypass.35-39 Although the long-term performance of these grafts has not surpassed that of saphenous vein, they have provided promising results in patients without suitable vein, even in below-knee (BK) applications. Lösel-Sadée and Alefelder37 reported a 5-year primary patency rate of almost 50% for femorocrural bypasses. An analysis of data from the Italian Propaten Registry (1025 patients, 815 BK procedures) revealed cumulative 5- and 9-year primary patency rates of 49% and 38%, respectively.38 No randomized studies of standard ePTFE vs Propaten grafts have been reported, but we39 retrospectively compared the two prostheses in a series of 252 patients who underwent either above-knee (192 patients, 87 Propaten grafts) or BK femoropopliteal bypass (60 patients, 42 Propaten grafts). Propaten grafts had significantly higher 5-year patency rates overall and for above-knee and BK bypasses separately (74.5% vs 56.2% [P ¼ .036], 85.2% vs 59.3% [P ¼ .028], and 59.6% vs 0/undeterminable [P ¼ .016]). Whether heparin bonding would further enhance the patency of preconstructed ePTFE AxBFB grafts is unknown, but the possibility is intriguing and we inserted the first one worldwide early in 2017. The limitations of this study include its nonrandomized design and retrospective nature resulting in 33 patients being lost to follow-up. Although one may suspect that the indication for these bypasses may affect patency, there were too few occlusions to allow an evaluation of this premise. We also have insufficient data as to the status of the runoff arteries to make any comment about the effect of runoff on graft patency. However, as mentioned in the Methods section, we always ensure patency through at least the deep femoral artery. Furthermore, since 2007, the preconstructed AxUFB grafts have been heparin bonded. Our prior experience with heparin-bonded PTFE grafts in the femoropopliteal location suggest that CBAS heparin-bonding improves patency, and this may be responsible for the superior AxUFB patency rates in comparison to earlier studies. However, 3 of 42 of the heparin-bonded grafts failed, whereas 2 of 34 non-heparin-bonded grafts ultimately failed. This suggests that at least in this series, heparin bonding was not responsible for improved patency.39
CONCLUSIONS Although AoFB remains the “gold standard” surgical revascularization procedure, our data indicate that AxBFB and AxUFB performed with the use of modern protocols and technology may render them an acceptable valid primary intervention in patients in whom endovascular treatment has failed or is unlikely to offer long-term success. An axillary-based extra-anatomic graft should be especially considered if a direct aortic procedure is likely to be complicated or high risk. However, we recommend CTA documentation of adequate
Journal of Vascular Surgery Volume
-,
Number
Samson et al
7
-
inflow and prior endovascular correction of significant stenoses that could compromise patency. The simplicity of performing these grafts and their low mortality and morbidity lend their application to surgeons with limited open aortic experience. Since AxUFB and AxBFB have similar patency rates, AxBFB should be reserved for bilateral indications. W. L. Gore & Associates assisted in the statistical analysis of the raw data. We thank Renée J. Robillard, MA, ELS, for editorial assistance.
AUTHOR CONTRIBUTIONS Conception and design: RS, ML Analysis and interpretation: RS, DS, DN, DD, RM Data collection: RS, RM Writing the article: RS Critical revision of the article: RS, DS, ML, DN, DD, RM Final approval of the article: RS, DS, ML, DN, DD, RM Statistical analysis: Not applicable Obtained funding: RS Overall responsibility: RS
REFERENCES 1. Clair DG, Beach JM. Strategies for managing aortoiliac occlusions: access, treatment and outcomes. Expert Rev Cardiovasc Ther 2015;13:551-63. 2. Indes JE, Pfaff MJ, Farrokhyar F, Brown H, Hashim P, Cheung K, et al. Clinical outcomes of 5358 patients undergoing direct open bypass or endovascular treatment for aortoiliac occlusive disease: a systematic review and metaanalysis. J Endovasc Ther 2013;20:443-55. 3. DeLaurentis DA, Sala LE, Russell E, McCombs PR. Twelve year experience with axillofemoral and femorofemoral bypass operations. Surg Gynecol Obstet 1978;147:881-7. 4. McPhee JT, Madenci A, Raffetto J, Martin M, Gupta N. Contemporary comparison of aortofemoral bypass to alternative inflow procedures in the Veteran population. J Vasc Surg 2016;64:1660-6. 5. Angle N, Dorafshar AH, Farooq MM, Ahn SS, Gelabert HA, Quiñones-Baldrich WJ, et al. The evolution of the axillofemoral bypass over two decades. Ann Vasc Surg 2002;16: 742-5. 6. Upchurch GR, Dimick JB, Wainess RM, Eliason JL, Henke PK, Cowan JA, et al. Diffusion of new technology in health care: the case of aorto-iliac occlusive disease. Surgery 2004;136: 812-8. 7. Jongkind V, Akkersdijk GJ, Yeung KK, Wisselink W. A systematic review of endovascular treatment of extensive aortoiliac occlusive disease. J Vasc Surg 2010;52:1376-83. 8. Louw JH. Splenic-to-femoral and axillary-to-femoral bypass grafts in diffuse atherosclerotic occlusive disease. Lancet 1963;1:1401-2. 9. Blaisdell FW, Hall AD. Axillary-femoral artery bypass for lower extremity ischemia. Surgery 1963;54:563-8. 10. Ascer E, Veith FJ, Gupta SK, Scher LA, Samson RH, WhiteFlores SA, et al. Comparison of axillounifemoral and axillobifemoral bypass operations. Surgery 1985;97:169-75. 11. Schneider JR, Golan JF. The role of extraanatomic bypass in the management of bilateral aortoiliac occlusive disease. Semin Vasc Surg 1994;7:35-44.
12. Eugene J, Goldstone J, Moore WS. Fifteen year experience with subcutaneous bypass grafts for lower extremity ischemia. Ann Surg 1977;186:177-83. 13. LoGerfo FW, Johnson WC, Corson JD, Vollman RW, Weisel RD, Davis RC, et al. A comparison of the late patency rates of axillobilateral femoral and axillounilateral femoral grafts. Surgery 1977;81:33-40. 14. Johnson WC, LoGerfo FW, Vollman RW, Corson JD, O’Hara ET, Mannick JA, et al. Is axillo-bilateral femoral graft an effective substitute for aortic-bilateral iliac/femoral graft? An analysis of ten years experience. Ann Surg 1977;186:123-9. 15. Onohara T, Komori K, Kume M, Ishida M, Ohta S, Takeuchi K, et al. Multivariate analysis of long-term results after an axillobifemoral and aortobifemoral bypass in patients with aortoiliac occlusive disease. J Cardiovasc Surg (Torino) 2000;41:905-10. 16. Nguyen KP, Perrone KH, Rahman A, Azarbal AF, Liem TK, Mitchell EL, et al. The role of axillofemoral bypass in current vascular surgery practice. Am J Surg 2016;211:968-71. 17. Rutherford RB, Patt A, Pearce WH. Extra-anatomic bypass: a closer view. J Vasc Surg 1987;6:437-46. 18. Taylor LM Jr, Moneta GL, McConnell D, Yeager RA, Edwards JM, Porter JM. Axillofemoral grafting with externally supported polytetrafluoroethylene. Arch Surg 1994;129: 588-95. 19. Passman MA, Taylor LM, Moneta GL, Edwards JM, Yeager RA, McConnell DB, et al. Comparison of axillofemoral and aortofemoral bypass for aortoiliac occlusive disease. J Vasc Surg 1996;23:263-71. 20. Harris EJ Jr, Taylor LM Jr, McConnell DB, Moneta GL, Yeager RA, Porter JM. Clinical results of axillobifemoral bypass using externally supported polytetrafluoroethylene. J Vasc Surg 1990;12:416-21. 21. Bredahl K, Jensen LP, Schroeder TV, Sillesen H, Nielsen H, Eiberg JP. Mortality and complications after aortic bifurcated bypass procedures for chronic aortoiliac occlusive disease. J Vasc Surg 2015;62:75-82. 22. Martin D, Katz SG. Axillofemoral bypass for aortoiliac occlusive disease. Am J Surg 2000;180:100-3. 23. Bunt TJ. Aortic reconstruction vs extra-anatomic bypass and angioplasty. Thoughts on evolving a protocol for selection. Arch Surg 1986;121:1166-71. 24. Hertzer NR, Bena JF, Karafa MT. A personal experience with direct reconstruction and extra-anatomic bypass for aortoiliofemoral occlusive disease. J Vasc Surg 2007;45:527-35. 25. Kashyap VS, Pavkov ML, Bena JF, Sarac TP, O’Hara PJ, Lyden SP, et al. The management of severe aortoiliac occlusive disease: endovascular therapy rivals open reconstruction. J Vasc Surg 2008;48:1451-7. 26. Burke CR, Henke PK, Hernandez R, Rectenwald JE, Krishnamurthy V, Englesbe MJ, et al. A contemporary comparison of aortofemoral bypass and aortoiliac stenting in the treatment of aortoiliac occlusive disease. Ann Vasc Surg 2010;24:4-13. 27. Calligaro KD, Ascer E, Veith FJ, Gupta SK, Wengerter KR, Franco CD, et al. Unsuspected inflow disease in candidates for axillofemoral bypass operations: a prospective study. J Vasc Surg 1990;11:832-7. 28. Burrell MJ, Wheeler JR, Gregory RT, Synder SO Jr, Gayle RG, Mason MS. Axillofemoral bypass: a ten-year review. Ann Surg 1982;195:796-9. 29. Schultz GA, Sauvage LR, Mathisen SR, Mansfield PB, Smith JC, Davis CC, et al. A five- to seven-year experience with externally-supported Dacron prostheses in axillofemoral and femoropopliteal bypass. Ann Vasc Surg 1986;1: 214-24.
8
Journal of Vascular Surgery
Samson et al
---
30. Ray LI, O’Connor JB, Davis CC, Hall DG, Mansfield PB, Rittenhouse EA, et al. Axillofemoral bypass: a critical reappraisal of its role in the management of aortoiliac occlusive disease. Am J Surg 1979;138:117-28. 31. Johnson WC, Lee KK. 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;30:1077-83. 32. Christenson JT, Broomè A, Norgren L, Eklof B. The late results after axillo-femoral bypass grafts in patients with leg ischaemia. J Cardiovasc Surg (Torino) 1986;27:131-5. 33. el-Massry S, Saad E, Sauvage LR, Zammit M, Davis CC, Smith JC, et al. Axillofemoral bypass with externally supported, knitted Dacron grafts: a follow-up through twelve years. J Vasc Surg 1993;17:107-15. 34. Schneider JR, McDaniel MD, Walsh DB, Zwolak RM, Cronenwett JL. Axillofemoral bypass: outcome and hemodynamic results in high-risk patients. J Vasc Surg 1992;15: 952-63. 35. Daenens K, Schepers S, Fourneau I, Houthoofd S, Nevelsteen A. Heparin-bonded ePTFE grafts compared with
36.
37.
38.
39.
2018
vein grafts in femoropopliteal and femorocrural bypasses: 1- and 2-year results. J Vasc Surg 2009;49:1210-6. Neville RF, Capone A, Amdur R, Lidsky M, Babrowicz J, Sidawy AN. A comparison of tibial artery bypass performed with heparin-bonded expanded polytetrafluoroethylene and great saphenous vein to treat critical limb ischemia. J Vasc Surg 2012;56:1008-14. Lösel-Sadée H, Alefelder C. Heparin-bonded expanded polytetrafluoroethylene graft for infragenicular bypass: five-year results. J Cardiovasc Surg (Torino) 2009;50:339-43. Pulli R, Dorigo W, Piffaretti G, Castelli P, Griselli F, Dorrucci V, et al. A decade of arterial bypass results with the Gore Propaten Vascular Graft: long-term clinical results from more than 1000 cases in the multicenter Italian Registry. Ital J Vasc Endovasc Surg 2014;21:101-7. Samson RH, Morales R, Showalter DP, Lepore MR Jr, Nair DG. Heparin-bonded expanded polytetrafluoroethylene femoropopliteal bypass grafts outperform expanded polytetrafluoroethylene grafts without heparin in a long-term comparison. J Vasc Surg 2016;64:638-47.
Submitted Dec 27, 2017; accepted Jan 31, 2018.
ABSTRACT Objective: Axillofemoral bypasses (AxFBs) have been used since 1962 to treat aortoiliac disease. In the past, reported patency rates (37%-76%) for these extra-anatomic grafts have been inferior to those for anatomic aortic grafting. Reported low survival rates after AxFB (40%-50%) have confirmed that these procedures have been used primarily in patients at high risk for complications from aortofemoral bypass. However, modern medical and anesthesia management, preoperative scanning, donor artery preparation, postoperative graft surveillance, and graft technology may improve outcomes after AxFB, possibly supporting expansion of its use. We therefore report our last 15-year experience with AxFB. Methods: Ring-reinforced, 8-mm expanded polytetrafluoroethylene grafts were used in all cases. The cross-femoral limb of axillobifemoral bypass (AxBFB) grafts was preconstructed. Heparin was administered intraoperatively, with protamine reversal. Loss of primary patency was defined as graft thrombosis of part or all of the inserted graft. Five-year primary patency rates were calculated by Kaplan-Meier analysis. Results: Between February 1991 and June 2016, a total of 161 grafts were inserted (85 AxBFBs and 76 axillounifemoral bypasses [AxUFBs]) in 91 male and 70 female patients (median age, 72.6 years; mean age, 73 years; range, 41-94 years). Indications for treatment were rest pain (49.6%), ischemic lesions (26%), claudication (22.3%), failed prior revascularization (9.3%), infection (3.7%), and dissecting aneurysm (1.2%). Reasons for performing AxFB rather than aortofemoral bypass were hostile aorta (44.1%), high risk (19.2%), prior failed reconstruction (12.4%), advanced age (8.7%), infection (4.3%), hostile abdomen (4.3%), aortic dissection (0.6%), and morbid obesity (0.6%). During follow up, 63 patients died, 17 within the first year; but only 3 patients died within 30 days of surgery (performed to treat an acute aortic occlusion). The 5-year survival rate was 55%. Five-year patency rates were 83.7% for all procedures, 81.8% for AxBFB, and 85.5% for AxUFB; the difference between AxBFB and AxUFB was not significant. Conclusions: Our data indicate that AxBFB and AxUFB performed with the use of modern protocols and technology may render them an acceptable valid primary intervention in patients in whom endovascular treatment has failed or is unlikely to offer long-term success. The simplicity of performing these grafts and their low mortality and morbidity lend their application to surgeons with limited open aortic experience. Because AxUFB and AxBFB have similar patency rates, AxBFB should be reserved for bilateral indications. (J Vasc Surg 2018;-:1-8.) Keywords: Axillobifemoral grafts; Axillofemoral grafts; Axillounifemoral grafts; Long-term patency; Long-term survival
Severe aortoiliac occlusive disease (AIOD) may be managed with aortofemoral bypass (AoFB), with extraanatomic bypass (axillofemoral bypass [AxFB], either axillounifemoral bypass [AxUFB] or axillobifemoral bypass [AxBFB], and femorofemoral bypass), or endovascularly (angioplasty and stenting). AoFB has traditionally been considered the standard surgical treatment because of its effectiveness in relieving symptoms and its reported 5-year patency rates of about 80% or higher.1,2 However,
From the Sarasota Vascular Specialists, Affiliated with Florida State University. Author conflict of interest: R.H.S. is a consultant for W. L. Gore & Associates. Correspondence: Russell H. Samson, MD, Clinical Professor of Surgery (Vascular), Florida State University Medical School, 600 N Cattleman Rd, Ste 220, Sarasota, FL 34232 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2018 Published by Elsevier Inc. on behalf of the Society for Vascular Surgery. https://doi.org/10.1016/j.jvs.2018.01.061
alternatives to AoFB have long been sought, especially for older and high-risk patients, because of the considerable operative and postoperative morbidity and hospital stay associated with the procedure, which requires opening of the abdomen.1,3,4 Although the use of endovascular therapy for AIOD has expanded dramatically since its introduction in the 1990s,5,6 systematic reviews and meta-analyses have found that the 5-year patency rates associated with this approach (60%-80%) remain lower than those achieved with surgical revascularization.1,7 Therefore, patients in whom angioplasty and stenting are used exclusively may require repeated procedures. The history of AxFB, which was first described almost simultaneously by Louw8 and Blaisdell and Hall,9 has been characterized by a considerable discrepancy in outcomes obtained.10,11 Early series generally had poor outcomes, with 4- or 5-year primary patency rates being 50% or lower in several studies10,12-14 and the highest rates being about 70%.15 Patient survival rates were also low, although it has often been observed that this finding is probably related primarily to the population of patients 1
2
Journal of Vascular Surgery
Samson et al
---
in whom AxFB has generally been used (ie, poor candidates for open aortic surgery because of age, cardiopulmonary disease, renal insufficiency, infection, hostile aorta or abdomen, and other high-risk conditions).3,11,14,16,17 In the 1990s, reports of better outcomes with AxFB began to appear. For example, in the study by Taylor et al18 of 184 patients who underwent AxUFB or AxBFB using externally supported, 8-mm-diameter expanded polytetrafluoroethylene (ePTFE) grafts, the 5-year primary patency rate was 71%. Passman et al19 retrospectively compared 108 patients who had undergone AxFB with 139 who had AoFB and observed a 5-year primary patency rate of 74% in the AxFB group and 80% in the AoFB group, a difference that was not significant. The patient survival rate was lower in the AxFB group (45% vs 72%), primarily because it contained a larger proportion of high-risk patients. The rate of major postoperative complications, however, was significantly higher in the AoFB group (9.2% vs 19.4%; P < .05). To our knowledge, a large AxFB series with a follow-up time of 5 years has not been reported since that of Passman et al. We here describe a series of AxFB procedures that began in 1991, well within the “modern era” of vascular surgery. We hypothesized that our outcomes would indicate a contemporary sustaineddand possibly enhanceddimprovement in AxFB outcomes.
METHODS Study design. The study was a retrospective evaluation of prospectively collected data obtained from a clinical experience between February 1991 and June 2016. Consent of the patients for the procedures was obtained per standard protocol. However, because this study was not performed as part of an ongoing clinical investigation or trial, consent for inclusion in the ongoing investigation was not necessary, nor was Institutional Review Board approval required. Four surgeons were involved in inserting the grafts, although the majority were performed by R.H.S. and D.P.S. Demographic data, risk factors, indications for surgery, previous therapy for vascular disease, and reasons for choosing axillary flow were obtained from the enrolled patients’ medical records and entered into a database for analysis (Atrium Medical Inc, Nashua, NH). Noninvasive and clinical preoperative workup. All patients had routine physiologic studies of the lower and upper extremity arteries. Most patients also had diagnostic arteriography, either transfemoral or through brachial puncture, to plan surgery but also to exclude a possible endovascular or open aortic procedure. Computed tomography angiography (CTA) evaluation of the aorta was obtained in low-risk patients to evaluate the status of the aorta to determine whether
2018
ARTICLE HIGHLIGHTS d
d
d
Type of Research: Retrospective, single-center, cohort study Take Home Message: Axillofemoral bypasses were performed in 161 patients, with three deaths within the first 30 days. The 5-year survival and patency rates were 55% and 83.7%, with no difference between axillofemoral and axillobifemoral bypasses. Recommendation: This study suggests that axillofemoral bypass can be used as a primary treatment of aortoiliac disease with good 5-year patency rates.
the patient could undergo aortobifemoral bypass or would require an AxBFB graft because of a hostile aorta. The aorta was considered hostile or inoperable if it was so heavily calcified that clamping was considered to be dangerous or impossible. Originally, computed tomography evaluation of the donor axillary arteries was performed only if there was a disparity in blood pressure between the two arms or if the patient complained of arm discomfort. In recent years, CTA evaluation of the donor artery has been obtained in all patients. Surgical technique. All procedures were performed under general anesthesia. All patients received a ringreinforced, 8-mm-diameter ePTFE graft (W. L. Gore & Associates, Flagstaff, Ariz); in those who underwent AxBFB, the femoral crossover graft segment was preconstructed (ie, it did not require a surgical anastomosis to the main limb). Since 2017, the preconstructed AxBFB grafts have been heparin bonded, but to exclude this confounding factor, this review was restricted to grafts performed before 2017. However, from 2007, the AxUFB grafts were heparin bonded using the CBAS technology acquired by W. L. Gore & Associates. The surgical technique warrants description because this may differ between surgeons, institutions, and prior instructions for use. A rolled sheet is placed transversely under the shoulders, and the donor arm is abducted 90 degrees and placed on an arm board (Fig 1). Antiseptic preparation starts at the chin and ends in the midthigh bilaterally. An adhesive povidone-iodine (Betadine)-impregnated plastic drape is applied to the chest wall, abdomen, and thighs to isolate the graft from any contamination from the skin. The femoral arteries are exposed using oblique groin incisions. The axillary artery is exposed by splitting the fibers of the pectoralis major and dividing the pectoralis minor (Fig 1). An extra-anatomic tunnel is created using the 8-mm Scanlan (Scanlan International, St. Paul, Minn) plastic tunneler starting under the pectoralis major. We do not advise tunneling from the groin toward the shoulder because we have heard of anecdotal occurrences in which the tunneler has inadvertently penetrated the chest cavity
Journal of Vascular Surgery Volume
-,
Number
Samson et al
3
-
Fig 1. Artist’s rendition of the positioning of the arm and the incisions required for a right axillobifemoral bypass (AxBFB) graft. The inset demonstrates the location of the anastomosis, which usually requires division of the pectoralis minor muscle for exposure. Furthermore, the graft is anastomosed to the anterior wall of the artery with enough redundancy to allow the patient to fully raise the arm above the head without any untoward traction on the anastomosis.
in such an approach. When an AxBFB graft is being constructed, a 2- to 3-cm counterincision is placed in the lower flank as high up on the flank as possible but still distal enough to allow the short limb of the AxBFB graft to reach from the counterincision to the ipsilateral femoral artery. The vertical limb of the AxBFB graft is then brought through the already inserted tunneler from the counterincision into the axillary artery incision. The femoral limbs of the graft are brought into the femoral incisions using an aneurysm clamp passed subcutaneously from the groin incisions to the counterincision. The configuration then approaches an inverted Y shape rather than the h shape that usually occurs with axillary-femoral-femoral-femoral configurations (Fig 2). Patients are then heparinized (5000 units of heparin for all patients irrespective of weight). The axillary anastomosis is performed first. The incision is placed anteriorly on top of the axillary artery with a lazy C-shaped turn such that if the patient were to elevate the arm directly above the head, there would be no tension on this anastomosis (Fig 1). We do not advise placing the anastomosis on the underside of the axillary artery as described in older texts. These texts assume that the graft comes into the artery in the same plane, whereas it actually dives down from the anterior chest wall to the front of the artery. The ipsilateral femoral anastomosis is performed next but not fully completed. A hydrogel clamp is then placed on the contralateral limb in the counterincision, and the graft is flushed by releasing the axillary artery flow briefly. The ipsilateral anastomosis is completed and flow restored to that side. The graft is never flushed with saline because we believe it may result in subsequent weeping of plasma
Fig 2. Photograph of a patient having undergone a right axillobifemoral bypass (AxBFB) graft. Note the inverted Y configuration due to the crossover of the preconstructed graft occurring in the lower right flank, with the counterincision still pink after the recent surgery.
through the graft pores. The contralateral anastomosis is then completed in similar fashion. We make a point of sewing the distal anastomosis across the common femoral onto the first part of the deep femoral artery or entirely onto the deep femoral artery in the presence of severe common femoral disease or occlusion of the superficial femoral artery. In some patients, we will reconstruct a femoral bifurcation if the superficial femoral artery and deep femoral artery are patent but the common femoral is severely diseased. Protamine reversal is used in all patients. Follow-up protocol. Patients were encouraged to follow up every 6 months for a clinical evaluation, ankle-brachial indices, and color duplex ultrasound evaluation of the graft as well as of the inflow and outflow arteries. However, not all patients complied. Accordingly, color duplex ultrasound and ankle-brachial indices were obtained outside of the regular follow-up if there was any clinical indication of incipient graft failure. If this was confirmed, CTA or diagnostic arteriography was added. Prophylactic anticoagulants were not used unless required for indications unrelated to graft patency. Aspirin was almost uniformly used and clopidogrel prescribed in later years. However, we do not have accurate information as to patients’ compliance
4
Journal of Vascular Surgery
Samson et al
---
2018
because this was a retrospective analysis. Loss of patency (graft failure) was defined as occlusion of any graft limb as indicated by duplex ultrasound scanning, regardless of whether a subsequent intervention restored patency. Data analysis. Descriptive characteristics of the patients were reported as median (range) or number and percentage. Kaplan-Meier analysis was used to calculate graft patency rates to 5 years and patient survival rate for up to 10 years.
RESULTS Between February 1991 and June 2016, a total of 161 grafts were inserted (85 AxBFBs, 76 AxUFBs) in 91 male and 70 female patients (median age, 72.6 years; mean age, 73 years; range, 41-94 years). Thirty-one of the patients had previously undergone procedures, including femorofemoral bypass (9 patients), aortoiliacfemoral bypass (7), infrainguinal bypass (11), iliac stenting (12), and iliac percutaneous transluminal angioplasty (2). Indications for treatment were rest pain (49.6%), claudication (22.3%), gangrene (14.9%), ulceration (11.1%), failing graft (9.3%), infection (3.7%), and dissecting aneurysm (1.2%). In the majority of cases, axillary flow was chosen because the patient had a hostile aorta (44.1%). Other indications included high risk (19.2%), prior failed revascularization (12.4%), advanced age (8.7%), hostile abdomen (4.3%), infected graft (4.3%), acute dissection (0.6%), and morbid obesity (0.6%). During follow-up, 63 patients died, 17 within the first year, of whom 3 died within 30 days of surgery performed to treat an acute aortic occlusion. No patient died intraoperatively. Twenty-seven patients (16.8%) required admission to the intensive care unit after surgery. The mean length of the hospital stay was 6 days (median, 3 days; range, 2-54 days). The mean follow-up time in the series was 27.8 months (maximum, 160 months). Thirty-three were lost to followup, but only two were lost to follow-up within 1 year of surgery. The 5-year survival rate was 55% (Fig 3). Six patients had occlusion of a graft limb, which was treated with placement of a new femorofemoral crossover graft from the vertical limb (three patients), a new AxUFB or AxBFB from the contralateral axillary artery (one each), or an AoFB (one). One patient had a functioning AxBFB removed and replaced by a new graft from the contralateral side because of the formation of a massive seroma involving the entire graft. Three grafts required an inflow procedure to maintain patencydplacement of a stent in the subclavian artery, percutaneous transluminal angioplasty at the proximal anastomosis, or bypass from the carotid artery to the vertical graft limb (one each). No patient required an amputation during the follow-up. The 1- and 5-year primary patency rates for all grafts were 96.3% and 83.7%, respectively (Fig 4). The 1- and
Fig 3. Patient survival up to 10 years after axillofemoral bypass (AxFB). The number at risk at each interval is shown.
5-year rates for AxBFB grafts were 98.4% and 81.8% (Fig 5); those for AxUFB grafts were 93.8% and 85.5% (Fig 6). There was no significant difference between AxBFB and AxUFB grafts in patency at either 1 year or 5 years (P > .5).
DISCUSSION AoFB provides high long-term patency rates, but vascular surgeons remain reluctant to perform it in high-risk patients because of its morbidity, mortality, and relatively long hospital stay.1,2,4,6,20,21 Although, in the past, graft patency after extra-anatomic bypass was disappointing, AxFB is an established alternative to AoFB, especially in patients with graft infection, infected aortic aneurysm, and aortoenteric fistula.1,3,9,11 The recent encouraging outcomes with AxFB,18,19,22 including ours, suggest that an expansion of its role in treating AIOD is warranted. Our series of patients with AIOD in whom AxFB was performed between 1991 and 2016 had one of the highest 5-year primary patency rates (83.7%) ever reported for this procedure. Moreover, the rates after AxUFB and AxBFB were also high (85.5% and 81.8%, respectively) and did not differ significantly. No patient died during surgery, and only three patients died within 30 days of surgery. Furthermore, no patient required an amputation during a mean follow-up of 27.8 months. Accordingly, our results show that excellent outcomes can now be achieved with these extra-anatomic bypasses. Some may suggest that an experience dating to the 1990s infers that modern techniques may not have been used. However, it was then that we adopted most of the techniques that we maintained throughout the rest of this experience. Admittedly, there have been advances in medical therapy and CTA technology that were more commonly used in later years, but we could not identify a specific time to divide the group. Furthermore, only 27 bypasses were constructed before the year 2000,
Journal of Vascular Surgery Volume
-,
Number
Samson et al
5
-
Fig 4. Kaplan-Meier estimate of primary patency for up to 5 years after all axillofemoral bypasses (AxFBs; axillobifemoral bypass [AxBFB] and axillounifemoral bypass [AxUFB]). The number at risk at each interval is shown. The error bars indicate 95% confidence intervals.
Fig 5. Kaplan-Meier estimate of primary patency for up to 5 years after axillobifemoral bypass (AxBFB). The number at risk at each interval is shown. The error bars indicate 95% confidence intervals.
with only 3 failures before censure due to death or loss to follow-up. No randomized trial has compared AoFB with either AxFB or angioplasty and stenting or AxFB with the endovascular approach, although several retrospective comparison studies have been reported.14,15,19,23-26 The results of investigations of this nature can be difficult to assess because of differences in populations of patients, including indications for treatment, variations in surgical techniques, and other factors. This problem is especially relevant to studies of AIOD treatment because high-risk patients are more likely to undergo AxFB than AoFB,19,24 and outcomes of AxFB have improved in recent decades. Interestingly, despite this improvement and the marked increase in use of endovascular therapy for AIOD, no study has compared AxFB with the endovascular approach. In their 2010 systematic review of 19 studies of endovascular AIOD therapy published in 2000 to 2009, Jongkind et al7 observed that the hospital stay ranged from 1 day to 4.8 days, the 30-day mortality rate from 1.2% to 6.7%, and the morbidity rate from 3% to 45%. For the eight studies in which 5-year primary patency was reported, the rate ranged from 60% to 86%. A 2013 systematic review and meta-analysis of 28 studies of endovascular treatment of AIOD (1625 patients) that were reported between 1989 and 2010 found that overall, the mean length of hospitalization was 4 days; the complication and 30-day mortality rates were 13.0% and 0.7%, respectively; and the pooled 5-year primary patency rate was 71.4%.2 Therefore, our outcomes with AxFB have been either similar or superior to those reported for endovascular therapy. Despite its popularity, the endovascular approach has some disadvantages. Placement of a stent in a heavily
calcified vessel can be challenging or impossible,19,22,25 and arterial perforation or dissection can occur.2,7 Loss of primary patency often results from stent thrombosis or in-stent stenosis.1 In addition, we and others16,21,22 are concerned that a dependence on endovascular therapy may result in a decrease in the number of surgeons competent to perform open proceduresdat a time when the aging of the U.S. population has substantially increased the number of patients requiring revascularization. The shift from AoFB to endovascular therapy has already resulted in a decline in AIOD operative experience among vascular surgery trainees.4 Expanded use of AxFB, a procedure that is simpler, easier to learn, less invasive, and less likely to have surgical complications than AoFB, may be a valid alternative surgical option for less experienced aortic surgeons. Several factors may have contributed to the improved outcomes with extra-anatomic bypass in our study and other recent series,18,19 including better anesthesia management, new antiplatelet and statin protocols, more specialized training and board certification of vascular surgeons, and improvements in duplex ultrasound for graft surveillance. Routine use of preoperative CTA to evaluate the aorta and the donor and recipient arteries may be especially critical. Unsuspected inflow disease has been reported to affect 16% to 25% of candidates for AxFB.10,27 Inflow disease that is detected by preoperative duplex ultrasound scans or CTA can now be treated endovascularly before AxFB is performed, thereby improving the results of the operative procedure.27 Similarly, duplex ultrasound detection of a subsequent hemodynamically significant stenosis in the inflow or outflow of these grafts can also allow preventive endovascular treatments that will prolong patency.
6
Journal of Vascular Surgery
Samson et al
---
Fig 6. Kaplan-Meier estimate of primary patency for up to 5 years after axillounifemoral bypass (AxUFB). The number at risk at each interval is shown. The error bars indicate 95% confidence intervals.
Several studies have evaluated whether the material, type (supported vs not supported), or configuration (AxUFB vs AxBFB) of the graft affects outcomes of AxFB. Dacron grafts were used in early AxFB series,3,12,28-30 but grafts made of ePTFE are now the most widely employed conduits.11,24 Studies that compared the two materials with respect to AxFB outcomes, including the only randomized trial to address this issue,31 have generally reported no significant differences in patency.28,32 However, externally supported grafts, such as the ring-reinforced ePTFE grafts we use, have been found to have a patency advantage over nonsupported grafts,18,20,29,33 possibly because they resist compression or kinking that leads to occlusion.18,20 Whether AxUFB or AxBFB provides better patency has long been debated. Some early studies supported the idea that an AxBFB configuration would enhance patency, presumably by increasing the flow rate in the main graft limb.13,14,17,28 More recent studies, however, including ours, found no difference in patency between AxUFB and AxBFB.10,12,22,34 In addition, Martin and Katz22 observed that 60% of graft occlusions after AxFB occur in the femorofemoral crossover limb. We agree with Martin and Katz that such findings indicate that AxBFB should be reserved for use in patients with bilateral disease. For patients who do require AxBFB, preconstructed grafts, like those we use, may offer not only enhanced convenience and a reduction in operating time but also a patency benefit related to improved flow dynamics resulting from the absence of a surgical anastomosis between the axillofemoral and femorofemoral graft limbs. This possibility has not been formally studied, but it merits investigation. Heparin-bonded ePTFE grafts (CBAS, Gore Propaten Vascular Grafts; W. L. Gore & Associates) have been used for >10 years in patients requiring a prosthetic
2018
femoropopliteal, femorocrural, or tibial artery bypass.35-39 Although the long-term performance of these grafts has not surpassed that of saphenous vein, they have provided promising results in patients without suitable vein, even in below-knee (BK) applications. Lösel-Sadée and Alefelder37 reported a 5-year primary patency rate of almost 50% for femorocrural bypasses. An analysis of data from the Italian Propaten Registry (1025 patients, 815 BK procedures) revealed cumulative 5- and 9-year primary patency rates of 49% and 38%, respectively.38 No randomized studies of standard ePTFE vs Propaten grafts have been reported, but we39 retrospectively compared the two prostheses in a series of 252 patients who underwent either above-knee (192 patients, 87 Propaten grafts) or BK femoropopliteal bypass (60 patients, 42 Propaten grafts). Propaten grafts had significantly higher 5-year patency rates overall and for above-knee and BK bypasses separately (74.5% vs 56.2% [P ¼ .036], 85.2% vs 59.3% [P ¼ .028], and 59.6% vs 0/undeterminable [P ¼ .016]). Whether heparin bonding would further enhance the patency of preconstructed ePTFE AxBFB grafts is unknown, but the possibility is intriguing and we inserted the first one worldwide early in 2017. The limitations of this study include its nonrandomized design and retrospective nature resulting in 33 patients being lost to follow-up. Although one may suspect that the indication for these bypasses may affect patency, there were too few occlusions to allow an evaluation of this premise. We also have insufficient data as to the status of the runoff arteries to make any comment about the effect of runoff on graft patency. However, as mentioned in the Methods section, we always ensure patency through at least the deep femoral artery. Furthermore, since 2007, the preconstructed AxUFB grafts have been heparin bonded. Our prior experience with heparin-bonded PTFE grafts in the femoropopliteal location suggest that CBAS heparin-bonding improves patency, and this may be responsible for the superior AxUFB patency rates in comparison to earlier studies. However, 3 of 42 of the heparin-bonded grafts failed, whereas 2 of 34 non-heparin-bonded grafts ultimately failed. This suggests that at least in this series, heparin bonding was not responsible for improved patency.39
CONCLUSIONS Although AoFB remains the “gold standard” surgical revascularization procedure, our data indicate that AxBFB and AxUFB performed with the use of modern protocols and technology may render them an acceptable valid primary intervention in patients in whom endovascular treatment has failed or is unlikely to offer long-term success. An axillary-based extra-anatomic graft should be especially considered if a direct aortic procedure is likely to be complicated or high risk. However, we recommend CTA documentation of adequate
Journal of Vascular Surgery Volume
-,
Number
Samson et al
7
-
inflow and prior endovascular correction of significant stenoses that could compromise patency. The simplicity of performing these grafts and their low mortality and morbidity lend their application to surgeons with limited open aortic experience. Since AxUFB and AxBFB have similar patency rates, AxBFB should be reserved for bilateral indications. W. L. Gore & Associates assisted in the statistical analysis of the raw data. We thank Renée J. Robillard, MA, ELS, for editorial assistance.
AUTHOR CONTRIBUTIONS Conception and design: RS, ML Analysis and interpretation: RS, DS, DN, DD, RM Data collection: RS, RM Writing the article: RS Critical revision of the article: RS, DS, ML, DN, DD, RM Final approval of the article: RS, DS, ML, DN, DD, RM Statistical analysis: Not applicable Obtained funding: RS Overall responsibility: RS
REFERENCES 1. Clair DG, Beach JM. Strategies for managing aortoiliac occlusions: access, treatment and outcomes. Expert Rev Cardiovasc Ther 2015;13:551-63. 2. Indes JE, Pfaff MJ, Farrokhyar F, Brown H, Hashim P, Cheung K, et al. Clinical outcomes of 5358 patients undergoing direct open bypass or endovascular treatment for aortoiliac occlusive disease: a systematic review and metaanalysis. J Endovasc Ther 2013;20:443-55. 3. DeLaurentis DA, Sala LE, Russell E, McCombs PR. Twelve year experience with axillofemoral and femorofemoral bypass operations. Surg Gynecol Obstet 1978;147:881-7. 4. McPhee JT, Madenci A, Raffetto J, Martin M, Gupta N. Contemporary comparison of aortofemoral bypass to alternative inflow procedures in the Veteran population. J Vasc Surg 2016;64:1660-6. 5. Angle N, Dorafshar AH, Farooq MM, Ahn SS, Gelabert HA, Quiñones-Baldrich WJ, et al. The evolution of the axillofemoral bypass over two decades. Ann Vasc Surg 2002;16: 742-5. 6. Upchurch GR, Dimick JB, Wainess RM, Eliason JL, Henke PK, Cowan JA, et al. Diffusion of new technology in health care: the case of aorto-iliac occlusive disease. Surgery 2004;136: 812-8. 7. Jongkind V, Akkersdijk GJ, Yeung KK, Wisselink W. A systematic review of endovascular treatment of extensive aortoiliac occlusive disease. J Vasc Surg 2010;52:1376-83. 8. Louw JH. Splenic-to-femoral and axillary-to-femoral bypass grafts in diffuse atherosclerotic occlusive disease. Lancet 1963;1:1401-2. 9. Blaisdell FW, Hall AD. Axillary-femoral artery bypass for lower extremity ischemia. Surgery 1963;54:563-8. 10. Ascer E, Veith FJ, Gupta SK, Scher LA, Samson RH, WhiteFlores SA, et al. Comparison of axillounifemoral and axillobifemoral bypass operations. Surgery 1985;97:169-75. 11. Schneider JR, Golan JF. The role of extraanatomic bypass in the management of bilateral aortoiliac occlusive disease. Semin Vasc Surg 1994;7:35-44.
12. Eugene J, Goldstone J, Moore WS. Fifteen year experience with subcutaneous bypass grafts for lower extremity ischemia. Ann Surg 1977;186:177-83. 13. LoGerfo FW, Johnson WC, Corson JD, Vollman RW, Weisel RD, Davis RC, et al. A comparison of the late patency rates of axillobilateral femoral and axillounilateral femoral grafts. Surgery 1977;81:33-40. 14. Johnson WC, LoGerfo FW, Vollman RW, Corson JD, O’Hara ET, Mannick JA, et al. Is axillo-bilateral femoral graft an effective substitute for aortic-bilateral iliac/femoral graft? An analysis of ten years experience. Ann Surg 1977;186:123-9. 15. Onohara T, Komori K, Kume M, Ishida M, Ohta S, Takeuchi K, et al. Multivariate analysis of long-term results after an axillobifemoral and aortobifemoral bypass in patients with aortoiliac occlusive disease. J Cardiovasc Surg (Torino) 2000;41:905-10. 16. Nguyen KP, Perrone KH, Rahman A, Azarbal AF, Liem TK, Mitchell EL, et al. The role of axillofemoral bypass in current vascular surgery practice. Am J Surg 2016;211:968-71. 17. Rutherford RB, Patt A, Pearce WH. Extra-anatomic bypass: a closer view. J Vasc Surg 1987;6:437-46. 18. Taylor LM Jr, Moneta GL, McConnell D, Yeager RA, Edwards JM, Porter JM. Axillofemoral grafting with externally supported polytetrafluoroethylene. Arch Surg 1994;129: 588-95. 19. Passman MA, Taylor LM, Moneta GL, Edwards JM, Yeager RA, McConnell DB, et al. Comparison of axillofemoral and aortofemoral bypass for aortoiliac occlusive disease. J Vasc Surg 1996;23:263-71. 20. Harris EJ Jr, Taylor LM Jr, McConnell DB, Moneta GL, Yeager RA, Porter JM. Clinical results of axillobifemoral bypass using externally supported polytetrafluoroethylene. J Vasc Surg 1990;12:416-21. 21. Bredahl K, Jensen LP, Schroeder TV, Sillesen H, Nielsen H, Eiberg JP. Mortality and complications after aortic bifurcated bypass procedures for chronic aortoiliac occlusive disease. J Vasc Surg 2015;62:75-82. 22. Martin D, Katz SG. Axillofemoral bypass for aortoiliac occlusive disease. Am J Surg 2000;180:100-3. 23. Bunt TJ. Aortic reconstruction vs extra-anatomic bypass and angioplasty. Thoughts on evolving a protocol for selection. Arch Surg 1986;121:1166-71. 24. Hertzer NR, Bena JF, Karafa MT. A personal experience with direct reconstruction and extra-anatomic bypass for aortoiliofemoral occlusive disease. J Vasc Surg 2007;45:527-35. 25. Kashyap VS, Pavkov ML, Bena JF, Sarac TP, O’Hara PJ, Lyden SP, et al. The management of severe aortoiliac occlusive disease: endovascular therapy rivals open reconstruction. J Vasc Surg 2008;48:1451-7. 26. Burke CR, Henke PK, Hernandez R, Rectenwald JE, Krishnamurthy V, Englesbe MJ, et al. A contemporary comparison of aortofemoral bypass and aortoiliac stenting in the treatment of aortoiliac occlusive disease. Ann Vasc Surg 2010;24:4-13. 27. Calligaro KD, Ascer E, Veith FJ, Gupta SK, Wengerter KR, Franco CD, et al. Unsuspected inflow disease in candidates for axillofemoral bypass operations: a prospective study. J Vasc Surg 1990;11:832-7. 28. Burrell MJ, Wheeler JR, Gregory RT, Synder SO Jr, Gayle RG, Mason MS. Axillofemoral bypass: a ten-year review. Ann Surg 1982;195:796-9. 29. Schultz GA, Sauvage LR, Mathisen SR, Mansfield PB, Smith JC, Davis CC, et al. A five- to seven-year experience with externally-supported Dacron prostheses in axillofemoral and femoropopliteal bypass. Ann Vasc Surg 1986;1: 214-24.
8
Journal of Vascular Surgery
Samson et al
---
30. Ray LI, O’Connor JB, Davis CC, Hall DG, Mansfield PB, Rittenhouse EA, et al. Axillofemoral bypass: a critical reappraisal of its role in the management of aortoiliac occlusive disease. Am J Surg 1979;138:117-28. 31. Johnson WC, Lee KK. 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;30:1077-83. 32. Christenson JT, Broomè A, Norgren L, Eklof B. The late results after axillo-femoral bypass grafts in patients with leg ischaemia. J Cardiovasc Surg (Torino) 1986;27:131-5. 33. el-Massry S, Saad E, Sauvage LR, Zammit M, Davis CC, Smith JC, et al. Axillofemoral bypass with externally supported, knitted Dacron grafts: a follow-up through twelve years. J Vasc Surg 1993;17:107-15. 34. Schneider JR, McDaniel MD, Walsh DB, Zwolak RM, Cronenwett JL. Axillofemoral bypass: outcome and hemodynamic results in high-risk patients. J Vasc Surg 1992;15: 952-63. 35. Daenens K, Schepers S, Fourneau I, Houthoofd S, Nevelsteen A. Heparin-bonded ePTFE grafts compared with
36.
37.
38.
39.
2018
vein grafts in femoropopliteal and femorocrural bypasses: 1- and 2-year results. J Vasc Surg 2009;49:1210-6. Neville RF, Capone A, Amdur R, Lidsky M, Babrowicz J, Sidawy AN. A comparison of tibial artery bypass performed with heparin-bonded expanded polytetrafluoroethylene and great saphenous vein to treat critical limb ischemia. J Vasc Surg 2012;56:1008-14. Lösel-Sadée H, Alefelder C. Heparin-bonded expanded polytetrafluoroethylene graft for infragenicular bypass: five-year results. J Cardiovasc Surg (Torino) 2009;50:339-43. Pulli R, Dorigo W, Piffaretti G, Castelli P, Griselli F, Dorrucci V, et al. A decade of arterial bypass results with the Gore Propaten Vascular Graft: long-term clinical results from more than 1000 cases in the multicenter Italian Registry. Ital J Vasc Endovasc Surg 2014;21:101-7. Samson RH, Morales R, Showalter DP, Lepore MR Jr, Nair DG. Heparin-bonded expanded polytetrafluoroethylene femoropopliteal bypass grafts outperform expanded polytetrafluoroethylene grafts without heparin in a long-term comparison. J Vasc Surg 2016;64:638-47.
Submitted Dec 27, 2017; accepted Jan 31, 2018.