Steal syndrome complicating hemodialysis access procedures: can it be predicted?

The development of steal syndrome distal to an arteriovenous fistula (AVF) created for hemodialysis access remains a significant clinical problem. This study was undertaken to determine the role of intraoperative noninvasive testing in the prediction and management of steal syndrome following arteriovenous fistula creation. First, in order to determine a threshold digital/brachial index (DBI) for patients at risk for steal syndrome, we performed a retrospective review of patients who had had the DBI measured and who developed symptoms (steal syndrome) following AVF creation. This was followed by a prospective evaluation of the ability of the DBI to predict which patients undergoing AVF surgery would develop steal syndrome. A DBI of <0.6 identifies a patient at risk for steal syndrome. Intraoperative DBI cannot be used to predict which patient will develop steal syndrome; however, if revision is indicated, the DBI should be increased to >0.6. Failure to accomplish this puts the patient at risk for continued steal syndrome.

Subfascial perforator vein ablation: comparison of open versus endoscopic techniques.

PURPOSE: To compare the outcomes and complications of open (OSPS) versus endoscopic subfascial perforator surgery (SEPS) for treatment of chronic venous insufficiency. METHODS: Data were retrospectively collected on 25 patients who underwent 27 SEPSs from February 1996 to August 1997 and from 22 patients who underwent 29 OSPSs between March 1978 and May 1993. Outcomes were evaluated for postoperative complications, ulcer healing, recurrence, and venous dysfunction scores on the last follow-up for the SEPS group and at 1-year follow-up for the OSPS group. RESULTS: The 2 groups were similar in age, sex, history of previous venous surgery, healed or active ulcers, etiology, deep venous incompetency, pathophysiology, and venous refill times. Eighteen (90%) of 20 active ulcers in the SEPS group healed with recurrences in 5 (28%) limbs at 7.5 +/- 5.4-month follow-up. All 19 ulcers in the OSPS group healed, with recurrences in 13 (68%) limbs at 35 +/- 35-month follow-up. Clinical venous dysfunction scores showed significant improvement following SEPS (10.0 +/- 3.6 to 5.4 +/- 4.1, p < 0.001) and OSPS (10.0 +/- 3.2 to 6.7 +/- 3.6, p < 0.001) with no significant difference between groups. Both groups also had significant improvement in anatomical and disability scores. There was no postoperative mortality in either group. The OSPS group had significantly more wound complications (45%) than the SEPS group (7%) (p < 0.005). The hospital stay and readmission rate for wound problems were also higher in the OSPS group. CONCLUSIONS: The early outcome showed equal improvement in clinical venous dysfunction scores in the 2 groups, but with significantly fewer complications in the SEPS group. Although the long-term durability of the endoscopic approach has not been determined, the short-term results would favor SEPS for treatment of severe venous insufficiency when perforator incompetence is a significant component.

Duplex directed caval filter insertion in multi-trauma and critically ill patients.

This study was undertaken to determine the safety and feasibility of inferior vena cava (IVC) filter insertion at the bedside using duplex imaging in multi-trauma and/or critically ill patients. From February 1996 to August 1997, 53 multi-trauma and/or critically ill patients, who were in the intensive care unit and referred for an IVC filter, were prospectively evaluated for possible duplex directed caval filter (DDCF) insertion. Screening IVC duplex scans were performed in all patients. Satisfactory ultrasound visualization in 46 patients (87%) allowed attempted DDCF insertion. All procedures were percutaneously performed at the bedside using Vena Tech IVC filters. The results from this series showed that DDCF insertion can be safely and rapidly performed at the bedside in multi-trauma or critically ill patients. The procedure is dependent on satisfactory visualization of the IVC by duplex ultrasonography, which was possible in 45 out of 53 (85%) patients. Insertion at the bedside substantially reduces the procedural cost and avoids the need for transport, radiation exposure, and intravenous contrast.

Endoleak after aortic stent graft repair: diagnosis by color duplex ultrasound scan versus computed tomography scan.

PURPOSE: The purpose of this study was to compare the accuracy of a color duplex ultrasound scan (CDU) to a computerized axial tomography scan (CT) in the diagnosis of endoleaks after stent graft repair of abdominal aortic aneurysms. METHODS: The Endovascular Aneurysm Clinical Trial Core Laboratory records were reviewed from 117 concurrent CDU and CT studies that were performed in 79 patients who were implanted with the Endovascular Technologies stent graft device between December 1995 and January 1997. All of the studies were interpreted by the Core Laboratory as having the presence or the absence of an endoleak or as being indeterminate because of technical factors. Of the 117 videotaped CDU studies available for reexamination, 100 were reassessed for technical adequacy on the basis of the following criteria: a satisfactory imaging of the aneurysm sac and of the stent graft with gray scale, and both color and spectral Doppler scan evaluation for endoleak outside the endograft and within the aneurysm sac. RESULTS: Of the 117 studies, 103 CDUs (88%) and 114 CTs (97%) were recorded as having the presence or the absence of an endoleak and 14 CDUs (12%) and 3 CTs (3%) were indeterminate. For the studies that were recorded to have the presence or the absence of an endoleak, the sensitivity, the specificity, the positive and the negative predictive values, and the accuracy of CDUs as compared with CTs were 97%, 74%, 66%, 98%, and 82%, respectively. Of the 100 CDU videotaped studies available for review, the following results were seen: (1) 93 CDUs had satisfactory B-mode images, (2) 76 had satisfactory color Doppler scan images to evaluate for endoleaks, (3) 55 had color Doppler scan assessment of the entire abdominal aortic aneurysm sac for endoleak, and (4) 27 had spectral Doppler scan waveform confirmation of suspected endoleaks. Only 19 CDU studies (19%) with all 4 criteria for complete assessment of endoleak were performed. CONCLUSION: Although most of the CDU studies were technically suboptimal, the CDUs reliably identified endoleaks with an excellent sensitivity and a negative predictive value as compared with CT scans.

A comparison of surgery for neurogenic thoracic outlet syndrome between laborers and nonlaborers.

OBJECTIVE: To determine factors of outcome following surgical intervention for neurologic thoracic outlet syndrome (NTOS). METHODS: In a retrospective study of patients surgically treated for NTOS, outcome was evaluated by postoperative symptoms and the ability of patients to return to work. RESULTS: Good, fair, and poor results were obtained in 26 (48%), 21 (39%), and 7 (13%) patients, respectively. The best predictor of a good outcome was occupation. Nonlaborers were more likely to have good outcome (21 of 32, 66%) when compared with laborers (5 of 22, 23%; P = 0.0025). Only 6 of 20 (30%) laborers were able to return to their original occupation compared with 17 of 26 (65%) nonlaborers (P = 0.036). CONCLUSIONS: Laborers with NTOS are less likely to have a good result from surgical intervention, are unlikely to return to their original occupation, and may require retraining for a non-labor-intensive occupation if they cannot return to their original work.

Long-term outcome after early infrainguinal graft failure.

PURPOSE: To determine the long-term outcome and prognostic factors after early infrainguinal graft failure (< 30 days). METHODS: Retrospective analysis of limb salvage data, patency data, and prognostic risk factors in 112 new infrainguinal bypass grafts from 1985 to 1995 that occluded within 30 days of operation. RESULT: Thirty-six femoropopliteal and 76 femorotibial/femoropedal arterial bypass ("index") procedures were performed for rest pain (50%), tissue loss (31%), or disabling claudication (19%). In 103 patients, an immediate additional revascularization ("takeback") procedure was performed at the time of early graft failure. Life table analysis of the takeback procedures for threatened limbs (n = 84) revealed limb salvage rates of 74%, 54%, 40%, and 31% at 1 month, 1 year, 3 years, and 5 years, respectively. The 1-month limb salvage rate (threatened limbs) was 12% (1 of 8) in patients who were not taken back for revascularization and 33% (4 of 12) in patients who had undergone more than one takeback procedure within 30 days. The secondary graft patency rates for the takeback procedures (n = 103) were 70%, 37%, 27%, and 23% at 1 month, 1 year, 3 years, and 5 years, respectively. Univariate and life table analysis revealed that patients who were given anticoagulation medication after the index procedure (before graft thrombosis) or patients who had undergone previous ipsilateral leg revascularization had significantly lower rates of limb salvage and graft patency (p < 0.05). The limb salvage rate was also significantly worse in patients who had single-vessel runoff compared with those who had multiple-vessel runoff (p < 0.01). Thrombectomy and revision or complete graft replacement had a better secondary patency rate than thrombectomy alone (p < 0.05). Autogenous vein grafts had better outcome than polytetrafluoroethylene-containing grafts, but statistical significance was not achieved. No significant differences in limb salvage or graft patency rates were found between femoropopliteal versus femorotibial/femoropedal bypass grafting, age, gender, previous inflow surgery, diabetes, hypertension, smoking, or cardiac, renal, or pulmonary disease. CONCLUSION: The long-term limb salvage and graft patency rates after takeback revascularization procedures for early graft failure are poor. Despite poor outcome, a single takeback procedure appears warranted in all patients. Multiple takeback procedures, however, do not appear to be justified, especially in patients who are given anticoagulation medication after the index bypass procedure, repeat leg bypass procedures, or if there is no potential for graft revision.

Blunt injury of the abdominal aorta: a review.

Injury to the abdominal aorta after blunt trauma occurs much less frequently than injury to the thoracic aorta. Although presentations vary, common themes continue to emerge with each patient. Within a 6-month period, our trauma unit diagnosed and treated two cases of blunt abdominal aortic trauma. Both patients were restrained passengers in motor vehicle crashes with resultant abdominal aortic injuries and demonstrated some of the most common associated injuries. Our two cases bring the number found in the literature to 62 and demonstrate the need for rapid recognition and treatment of this potentially lethal injury. This article is a comprehensive review of the management of abdominal aortic injury from blunt trauma.

A new technique for hemodialysis access surgery: use of the pneumatic tourniquet.

The purpose of this study was to determine whether tourniquet occlusion could be safely used on the upper extremity for vascular control during hemodialysis access surgery. The hospital and outpatient records of 44 patients undergoing 105 hemodialysis access procedures were retrospectively reviewed. In 48 procedures tourniquet occlusion was used for vascular control, whereas in 57 procedures vascular clamps were used. In those procedures in which the tourniquet was used, the mean tourniquet time was 30 minutes and the mean tourniquet pressure was 242 mm Hg. The operative time was significantly less in the tourniquet group as compared to the clamp group (72.5 minutes vs. 84 minutes, respectively; p = 0.029). There was no statistically significant difference in the incidence of nerve injury, bleeding, hematoma, vascular steal, infection, or swelling between the two groups. There were no complications related specifically to the use of the tourniquet. There was no difference in primary patency in comparing the tourniquet control group with the clamp control group (p > 0.5). The use of a pneumatic tourniquet for vascular control during hemodialysis access surgery allows for a faster, technically easier operation with no increase in the complication rate and no effect on primary patency.

Coexistent abdominal aortic aneurysm and renal carcinoma: management options.

Renal carcinoma (RCA) presenting in association with abdominal aortic aneurysm (AAA) is extremely rare, with only sporadic case reports previously described. The management of six cases of AAA and concomitant RCA presenting to a single institution from March, 1991 through December, 1993 was reviewed and management options considered. AAAs ranged in size from 4.5-7.0 cm (mean, 5.6 cm). Three left renal carcinomas were resected via a retroperitoneal approach simultaneous to repair of the AAA. One right renal carcinoma was resected in combination with repair of an AAA through a transperitoneal approach. The fifth case was managed by left nephrectomy, followed by interval aneurysmectomy, and the sixth case was managed by nonsurgical methods because of the presence of widely metastatic disease. Renal malignancies included five renal cell carcinomas and one transitional cell carcinoma. Three patients remain free of disease 8-11 months postoperatively, and one patient had metastatic disease detected 19 months postoperatively. Two deaths have occurred; one due to a massive CVA 1 month following a combined aneurysmectomy and left nephrectomy, and a second due to unknown etiology in the patient managed non-surgically. No peripheral vascular or aortic graft related complications have occurred. The treatment of AAA and RCA should be governed by the size of the AAA, the location of the cancer, and the extent of malignant disease. Simultaneous resection is safe and effective in patients with coexistent AAA and renal cancer. Left sided tumors should be resected via a retroperitoneal approach that also provides excellent exposure for simultaneous AAA resection.(ABSTRACT TRUNCATED AT 250 WORDS)

Intraoperative iliac artery stents: combination with infra-inguinal revascularization procedures.

Iliac artery percutaneous transluminal angioplasty (PTA) can effectively provide in-flow for subsequent distal vascular reconstruction. Iliac artery stents may improve the initial hemodynamics and long term patency of PTA, and thus may be well-suited for combined proximal PTA with distal bypass procedures. This report reviews our preliminary experience with iliac artery stenting in combination with infra-inguinal vascular reconstruction. Thirteen iliac artery stent procedures combined with simultaneous distal revascularization were performed in 11 patients. Ten procedures were performed for limb salvage, two for disabling claudication, and one before planned orthopedic surgery. Distal revascularization procedures included seven femoropopliteal, four femorotibial bypasses, one common femoral endarterectomy, and one thrombectomy of a femoropopliteal bypass. Stent placement was technically successful in all patients. Mean pre-operative ankle-brachial index (ABI) was 0.41 (+/- 0.28), which improved to 0.91 (+/- 0.18) post-operatively (P < 0.0001). Mean systolic iliac artery gradients across the lesions improved from 27.1 (+/- 9.8) mm Hg to 2.7 (+/- 3.4) mm Hg after stent placement (P < 0.0001). Mean follow-up is 5.8 months (range 1-12 months). Two femoropopliteal bypass grafts occluded in the follow-up period. One occlusion was caused by a mid-vein graft stenosis that was repaired with subsequent graft patency. The other graft occlusion occurred in a patient with rest pain who did not require a second bypass procedure, as the ABI increased from 0.3 to 0.7 following stent placement with resolution of symptoms.(ABSTRACT TRUNCATED AT 250 WORDS)

Vein harvest ischemia: a peripheral vascular complication of coronary artery bypass grafting.

Lower-extremity ischemia can lead to impaired healing of saphenous vein excision sites in patients with significant peripheral vascular disease (PVD). Five patients who required infrainguinal revascularization for wound necrosis of the harvest site after coronary artery bypass grafting are described. The male/female ratio was 2:3 with a mean age of 67 (range 45-87) years. The most commonly associated problems were insulin-dependent diabetes mellitus (80%) and congestive heart failure (60%). The saphenous vein was harvested from the thigh and leg in three patients and exclusively from the leg in the others. Manifestations of ischemia ranged from persistent ulceration to complete wound disruption threatening limb loss. Impaired healing was isolated to infragenicular wounds in all patients. Pedal pulses were not detected in any of the affected extremities. Determination of the ankle/brachial pressure indices (ABI) revealed values of < 0.5 in three affected limbs. Non-compressible vessels resulted in falsely raised ABI of > 1.0 in the remaining two limbs; however, Doppler waveform analysis in these patients demonstrated significant PVD. Aggressive wound care and antibiotic therapy were continued for mean of 9 weeks before operative intervention. Infrainguinal reconstruction included femoropopliteal (two), femorotibial (two) and popliteal-tibial bypass (one). Autologous arm and saphenous veins in addition to expanded polytetrafluoroethylene grafts were used effectively. Limb salvage and wound healing were achieved in 100% of the patients without untoward sequelae. It is concluded that unrecognized PVD in patients undergoing coronary artery bypass grafting can lead to significant morbidity. Patients at risk may be identified with a combination of history, physical examination and non-invasive testing.(ABSTRACT TRUNCATED AT 250 WORDS)

Long-term follow-up for recurrent stenosis: a prospective randomized study of expanded polytetrafluoroethylene patch angioplasty versus primary closure after carotid endarterectomy.

PURPOSE: To determine the effect of primary closure (PC) versus expanded polytetrafluoroethylene patch graft angioplasty (PGA) on the incidence of recurrent stenosis (> 50% lumen diameter narrowing) after carotid endarterectomy (CEA), 87 patients undergoing 100 consecutive CEA were prospectively randomized into two groups. METHODS: Forty-four patients underwent 51 PC, and 43 patients underwent 49 PGA. All patients were evaluated after operation by duplex scanning at 1.5, 12, 24, and 36 months. There were no significant differences in the demographic characteristics or operative indications for CEA between the two patient groups. Complete follow-up was achieved in 86% (75/87) of the patients during the 36-month surveillance period. RESULTS: The perioperative permanent neurologic morbidity in the PC and PGA groups was noted to be 4% and 2%, respectively (PC = 2/51 vs PGA = 1/49, p = 0.58). Three additional reversible cerebral ischemic events occurred in the postoperative period (PC = 2/51 vs PGA = 1/49, p = 0.58). Other morbidity included immediate postoperative hemorrhage requiring reexploration (1/51) in the PC group and an infected expanded polytetrafluoroethylene patch requiring removal and replacement with autogenous vein (1/49). Long-term follow-up detected a single patient with significant bilateral restenoses of his primarily closed carotid arteries. None of the patients in the PGA group had restenoses (PC = 2/51 vs 0/49, p = 0.50). In addition, no postoperative dilation of the common or internal carotid arteries or perioperative death was observed. CONCLUSIONS: In patients undergoing CEA, these data demonstrate no significant difference in the perioperative morbidity or mortality between PC and PGA. Use of the patch did not engender patients to patch rupture or aneurysmal degeneration as previously described with vein patch angioplasty procedures. This series supports effective use of either technique to achieve minimal rates of restenosis.

Angioscopic evaluation of valvular disruption during in situ saphenous vein bypass.

Several valvulotomes are currently available to achieve valvular disruption; however, studies comparing the efficacy of these endoluminal instruments are lacking. This prospective study evaluates the efficacy and safety of the three most commonly employed valve cutters: the Hall, LeMaitre, and Mills valvulotomes. A total of 30 in situ greater saphenous vein bypass grafts were included in this investigation. Valvular disruption was attempted with either the LeMaitre (11 cases), Hall (12 cases), or Mills (7 cases) valvulotomes. Subsequently, angioscopy was employed to assess the completeness of valvulotomy and to identify vein wall injury. Incomplete disruption of one or more valve complexes was identified in 2 of 12 (17%) grafts in the Hall group, 10 of 11 (91%) grafts in the LeMaitre group, and 0 of 7 grafts in the Mills group (p < 0.01). Intact valve cusps were noted in 2 of 36 (5.5%) valves, 31 of 42 (74%) valves, and 0 of 38 valves after valvulotomy with the Hall, LeMaitre, and Mills instruments, respectively (p < 0.01). A total of three valvulotome-related injuries occurred; two injuries were noted in conjunction with the Hall instrument, one was associated with the Mills valvulotome, and no injuries were detected after use of the LeMaitre instrument (p = 0.33). These data demonstrated a significantly increased incidence of retained valve cusps when the LeMaitre valvulotome was used. No significant difference in the rate of vein wall injury was noted in the three groups. Thus this study suggests that the LeMaitre instrument is not as effective as either the Hall or Mills valvulotomes for achieving valvular disruption.

Successful management of traumatic false aneurysm of the extracranial vertebral artery by duplex-directed manual occlusion: a case report.

A 32-year-old man was transferred to our hospital after a 2.0 by 2.5 cm traumatic false aneurysm of the distal extracranial vertebral artery was noted after a stab wound of the posterior side of the neck. To obviate the need for operative exposure of the distal vertebral artery at the base of the skull, we elected to perform duplex-directed manual occlusion of the lesion. Angiography before and after the procedure, as well as 10-month follow-up duplex ultrasonography, demonstrated satisfactory thrombosis of the false aneurysm without evidence of a residual arterial defect. There was no morbidity associated with the procedure. We conclude that duplex-directed manual occlusion, a new technique recently described for the nonoperative management of postcatheterization femoral false aneurysms, can be applied safely and effectively to false aneurysms in other locations in which the risks and technical difficulties of operative repair render surgery less desirable.

Color-flow duplex-directed manual occlusion of femoral false aneurysms.

PURPOSE: During the past 14 months we conducted a prospective clinical trial to evaluate the efficacy of duplex-directed manual occlusion (DDMO) of iatrogenic femoral false aneurysms (FFAs) as an alternative to standard operative management. METHODS: In all cases DDMO was performed with real-time color-flow imaging while steady, continuous external pressure was applied manually to the neck of the FFA by an experienced vascular technologist for a period of 10 minutes. RESULTS: Ten of the 11 FFAs treated with DDMO in this series were thrombosed successfully, requiring a mean of 30 minutes of compression per aneurysm (three compressions of 10 minutes each). DDMO was unsuccessful in one patient, whose session was terminated because of severe discomfort as a result of the procedure. All 10 patients with successfully thrombosed FFAs are without recurrence at 1-month follow-up color-flow duplex examination, and there has been no morbidity attributable to DDMO. CONCLUSIONS: We conclude that DDMO of postcatheterization FFA can be performed safely and is an inexpensive, effective, nonoperative method of managing such lesions. The precise role of this technique would appear to be as a first-line treatment for uncomplicated iatrogenic FFAs.

Vascular Prosthetics: the Gore-Tex ePTFE Stretch Graft.

Following introduction of the Gore-Tex stretch vascular graft in 1991, over 15,000 bifurcated and 35,000 straight grafts have been distributed (Figure I). This novel graft, recommended for arterial and venous reconstruction, is touted to afford significantly superior handling characteristics, kink-resistance and conformability. In addition, the "stretch" feature is thought to confer ease of sizing and anastomotic accuracy. Whether this product of advanced polymer science represents a panacea in the field of vascular grafting or a mere addition to the host of less than ideal synthetic conduits remains to be discerned. The first human implantation of a stretch graft was conducted by one of the authors (R.T.G.) during aortic replacement at the Eastern Virginia Medical School (EVMS). A favorable experience has since been gained by utilization of the stretch graft during vascular reconstruction for aortoiliac, infrainguinal and visceral arterial disease. This review will present the intriguing history of synthetic graft development and recount our experience with the Gore-Tex stretch vascular graft.

The ischemic window: a method for the objective quantitation of the training effect in exercise therapy for intermittent claudication.

Twenty-two patients with intermittent claudication were prospectively enrolled in a 12-week program of supervised, graded treadmill exercise therapy. Severity and distribution of arterial occlusive disease were ascertained by noninvasive determination of segmental lower extremity blood pressures and waveforms. No attempt was made to modify risk factors for atherosclerotic occlusive disease. The exercise-induced reduction of the ankle pressure and its recovery were recorded over time, and the area under this curve, the "ischemic window," represents the severity of the ischemic deficit. Absolute systolic ankle pressure, ankle-brachial index, maximum walking time, claudication pain time, and the ischemic window were measured before and after exercise training in all subjects. Maximum walking time and claudication pain time increased 659% and 846%, respectively, among the 19 patients completing the 12-week program (p = 0.001; p = 0.0002). These patients underwent a mean reduction of 58.7% in the ischemic window after a standardized workload (p less than 0.05), and this correlated with the degree of symptomatic improvement. Absolute ankle pressure and ankle-brachial index were unchanged after exercise training. This study confirms the utility of supervised exercise therapy in the treatment of intermittent claudication. The ischemic window is a useful method for quantifying the ischemic deficit produced by exercise and provides a reproducible means of documenting functional improvement in patients undergoing exercise training.

Mycotic aneurysm of the suprarenal abdominal aorta.

We report on the successful treatment of a patient with a mycotic aneurysm of the suprarenal aorta. The aorta was resected and reconstructed using an in-situ polytetrafluoroethylene graft with a side arm branch to the left renal artery. The use of polytetrafluoroethylene graft for aortic reconstruction after suprarenal mycotic aneurysm resection has not been previously reported. The etiology, bacteriology, diagnosis, and principles of management of mycotic aneurysms of the suprarenal aorta are discussed.

Duplex scanning for the intraoperative assessment of infrainguinal arterial reconstruction: a useful tool?

Duplex scan, arteriography, and graft flow rates were used intraoperatively to assess 56 infrainguinal arterial reconstructions for technical error. Intraoperative duplex scan identified a technical defect or low graft flow velocity in 22 of 56 (39%) grafts. Eleven of the defects were judged to be clinically significant and were corrected. Four of these defects were missed by the completion arteriogram. One technical defect identified by completion arteriography was missed by duplex scan. Fifty percent (5/10) of grafts with an abnormal intraoperative duplex scan which were not corrected occluded within 30 days. Graft flow rates measured by the electromagnetic flowmeter were neither predictive of technical defect nor early graft outcome. Although the sensitivity of arteriography and duplex scan (88% sensitivity for both) were both high for predicting early graft occlusion, the combination of duplex scan and completion arteriography was significantly more accurate (p less than .0001) in predicting early graft outcome than either study alone. Duplex scan identified significant graft defects which were not detected by completion arteriography or graft flow rate measurement. The duplex scan also provided hemodynamic information which was predictive of early graft outcome. The duplex scan can be an important adjunct to completion arteriography for the intraoperative assessment of infrainguinal arterial reconstruction.