High incidence of venous thrombosis after surgery for abdominal aortic aneurysm
Article Outline
- Abstract
- Objective
- Materials and methods
- Results
- Discussion
- Conclusion
- Author contributions
- Acknowledgment
- References
- Copyright
Objective
The incidence of venous thromboembolism (VT) after aortic abdominal aneurysm (AAA) surgery is imprecisely reported. On one hand, thromboprophylaxis has improved, on the other hand, AAA patients have become older and/or present worse comorbidities. Herein, we prospectively analyzed the incidence of VT in a continuous series of patients operated on for AAA repair and looked for predictive factors.
Materials and Methods
Between January 1, 2005, and December, 31, 2006, 193 consecutive patients (177 men and 16 women), mean age 73 (range, 47-93) underwent elective AAA repair, 137 open (71%) and 56 endovascular (29%), in our institution. Thromboprophylaxis consisted of thigh-length compression bandages or stockings, early mobilization, and a daily subcutaneous injection of low-molecular-weight heparin (enoxaparin 40 mg per day). Patients with renal insufficiency or aged over 80 were given unfractionated heparin (5000 IU twice a day). Heparin was started between day 1 and day 5 (median = day 1) after surgery, according to the prescription of the surgeon. A bilateral lower limb duplex venous compression ultrasonography scan using 3 to 7.5 MHz transducers was systematically done before and after surgery in each patient. Two groups were considered: group 1 with postoperative VT (n = 17) and group 2 without (n = 176). The 17 patients with VT were compared with 51 patients randomly chosen among the 176 patients without VT. Different characteristics such as venous risk factors, preoperative antithrombotic treatment, anatomical features of the AAA, and perioperative data were studied.
Results
Seventeen patients (8.1%) were diagnosed with postoperative VT (15 asymptomatic deep vein thrombosis [DVT] and 2 symptomatic pulmonary embolisms [PE]). VTE tended to be more frequent in open (10.2%) than endovascular (5.3%) repair (P = .28). Among perioperative data, delay to thromboprophylaxis was related to bleeding complication (P = .05) and blood transfusion (P = .02), and tended to be longer in VT patients (1.7 +/− 1.4 vs 0.9 +/− 0.9 day; P = .09).
Conclusion
Despite systematic prevention with heparin, surgery for AAA repair induces a high incidence of postoperative VT. This series mandates for vigilance about VT, with particular attention to the patients who received transfusion with fresh frozen plasma. Even though this series is one of the largest ever published on this topic, the rarity of the events calls for confirmation with a larger prospective study.
Venous thromboembolism (VT) frequently occurs in hospitalized patients and especially after surgery. Some major surgical procedures are known to be associated with a high risk of developing VT. For example, in major orthopedic surgery 65% of patients developed VT in the control arm without thromboprophylaxis.1 Through a tremendous effort in prevention, the use of routine pharmacological and mechanical devices led to a threefold reduction in the incidence of deep venous thrombosis (DVT) and pulmonary embolism (PE) after elective hip and knee joint surgery.2, 3 In a recent prospective study, a VT incidence of 1.34% (95% confidence interval [CI], 1.04–1.64) was observed after surgery for hip fracture, despite the use of thromboprophylaxis.4 There are fewer reports concerning VT following vascular surgery, and in our practice we noticed that aortic abdominal aneurysm (AAA) repair generated some postoperative venous thrombotic events. At the time of surgery, vascular patients in general, and AAA patients in particular, tend to be older with more debilitating comorbidities, leading to a higher risk for venous complications. Despite this, no recommendations have been clearly established.
Objective
To prospectively analyze the incidence of VT in a continuous series of patients operated on for AAA repair and to look for predictive factors in order to improve the perioperative prevention of venous thrombosis.
Materials and methods
Between January 1, 2005, and December 31, 2006, 193 consecutive patients underwent elective AAA repair in our department, 137 open (71%) and 56 endovascular repairs (29%). There were 177 men and 16 women. The mean age was 73 years, ranging from 47 to 93.
During surgery, all patients received a heparin bolus of 0.5 mg/kg (50 IU/kg) before clamping. Biological monitoring by activated coagulation time (ACT) was performed in case of surgery longer than 2 hours and additional bolus when needed. Heparin was neutralized with protamine (ratio 1:1) at the end of the operation.
Thromboprophylaxis consisted of thigh-length compression bandages or stockings, early mobilization, and a daily subcutaneous injection of low-molecular-weight heparin (LMWH) (enoxaparin 40 mg per day).5 The regimen of 40 mg daily has been recommended for high-risk situations of venous thrombosis by the latter American College of Chest Physicians (ACCP) consensus, the only mentioned study with LMWH after vascular surgery concerned enoxaparin 40 mg (4000 IU).6 Patients with renal insufficiency or those over 80 years of age were given unfractionated heparin (UH) (5000 IU, twice a day). When needed, UH was chosen in a moderate dose to prevent VT but also postoperative bleeding. Anticoagulation was started between day 1 and day 5 (median = day 1) after surgery, according to the prescription of the surgeon (Fig). Mechanical prophylaxis consisted of compression bandages (Bandage Biflex 16, Thuane, France), in case of varicose veins (20% of the patients), and stockings (TED, Codiven, United Kingdom) otherwise, without crossover. The compression program was applied by nurses: bandages 20 to 25 mm Hg and stockings 12 to 15 mm Hg. Compression devices were worn continuously during the whole hospital stay. Mobilization consisted of bed to chair at day 1 and ambulation as of day 2.
Fig.
Timing of heparin onset in the postoperative period of the whole series of AAA patients.
Venous duplex scan was performed during the systematically scheduled ultrasonic arterial examination the day before surgery and once before discharge (median time: day 2 after endovascular repair and day 6 after open surgery) or anytime in case of clinical suspicion of DVT. It consisted in a bilateral lower limb duplex venous compression ultrasonography scan using 3 to 7.5 MHz transducers by a team of experienced practitioners from the Department of Angiology. The venous ultrasonography examination used high definition imaging equipment and consisted of imaging the venous segments without and with compression. The inferior cava, bilateral iliac, femoral (common, superficial, and profunda), and popliteal veins were checked in the supine position and the bilateral calf veins including posterior tibial, peroneal, gastrocnemial, and solear, in the sitting position. Venous thrombosis was defined as the presence of endoluminal material with a diameter on compression >5 mm.7
The study population was divided in two groups: group 1 with postoperative VT and group 2 without VT. The 17 group 1 patients were compared to controls chosen among patients from group 2. Three controls were randomized and chosen for each case (51 controls for 17 cases). Regarding anticoagulation prevention, among the 17 VT cases, 12 received enoxaparin 40 mg, and 5 preventive UH. In group 2, 38 patients were treated with enoxaparin 40 mg, 8 with preventive UH, and 5 with curative UH (patients on oral anticoagulant [OAC] for atrial fibrillation). Patients were kept on aspirin during the surgery period and those who received clopidogrel were switched to aspirin in order to decrease the risk of bleeding.
Different characteristics were studied: previous episode of DVT, concurrent malignant disease, history of cardiovascular events, overweight (body mass index >25), preoperative treatment with antiplatelet drugs (aspirin, clopidogrel), or anticoagulants (coumadin), anatomical features (maximum diameter of the AAA, and the presence of thrombus in the aneurismal sac), perioperative data (operative time, clamping time, estimated blood loss, use of blood products (packed red blood cells [PRBC], platelets, fresh frozen plasma [FFP], postoperative data (first day of LMWH prescription).
The high incidence of thrombotic events led us to look for risk factors and to bring additional data compared to previous studies. Our study is only explanatory, as no factor was known sufficiently to allow computation of a sample size on a power basis. The data were evaluated by χ2 analysis for categorical data and the Mann-Whitney U test for continuous data. In the tables, data are expressed with mean values and standard deviations (SD). A P value < .05 was regarded as statistically significant.
This investigation has been included in a research program focused on AAA, and was approved by the ethics committee of our academic hospital (reference: AOI 2003). No patient was asked to sign an informed consent form because neither additional examination nor extra blood sampling was requested for this study.
Results
No patient was diagnosed with VT in the preoperative period. Five patients had a past history of VT (>6 months). Seventeen patients (8.1%) were diagnosed with postoperative VT consisting of 15 asymptomatic DVT and 2 symptomatic PEs. One patient presented with acute dyspnea and chest pain 2 days after open AAA repair. Diagnosis of PE was confirmed by computed tomography (CT)-scan, and one asymptomatic bilateral distal DVT was discovered on the duplex scan. Even though the patient presented with bleeding complications at day 0 and had received a blood transfusion (3 PRBC and 1 FFP), a daily administration of 40 mg enoxaparin was started as soon as day 1 after surgery. No venous risk factors were highlighted. The second symptomatic patient presented with a PE on the second day following the operation and was diagnosed with bilateral asymptomatic DVT, distal in the right leg, proximal in the left leg. He had undergone open AAA repair with no bleeding or need for a blood transfusion. He had also received a daily injection of 40 mg of enoxaparin, since the day following the operation. This patient was found to have two venous risk factors: prior DVT after previous orthopedic surgery and obesity (body mass index = 34). The 15 asymptomatic VT patients had proximal DVT (femoro popliteal) in 2 patients, distal DVT in 12 patients, and a superficial VT in a saphenous vein.
VT seemed to be more frequent in open (10.2%) than endovascular (5.3%) repair, but the difference was not significant (P = .28). The VT risk factors are presented in Table I. Preoperative data did not allow identification of patients at risk. No significant difference was observed for patients with a history of DVT, concurrent malignant disease, cardiovascular disease, or overweight. Before surgery, 9 patients were on oral anticoagulation and 43 on an antiplatelet drug. Even though our policy was to maintain aspirin and discontinue clopidogrel for 7 days before surgery, antiplatelet drugs were found to have no incidence of thrombotic complications. Anatomic features such as AAA anterior to posterior diameter and the presence of a thrombus in the aneurysm sac were not predictive of VT either. Neither were operative data such as operating time, clamping time, and estimated blood loss. VTE patients seemed to stay in the intensive care unit (ICU) longer than did non-VT patients (3.7 +/− 4.2 days vs 2.6 +/− 4.5 days, P = .6), but the difference was not significant. The delay to LMWH prescription tended to be longer in VT patients, compared to non-VT patients (1.7 +/− 1.4 days vs 0.9 +/− 0.9 days, P = .09). At day 1 after surgery, 65% of the VT patients and 80% of the non-VT patients were already being given LMWH. The delay to heparin prescription was related to a higher blood loss during the first 24 hours (P = .005) and transfusion of blood products, either PRBC or FFP (P = .02) (Table II). Nevertheless, only FFP was statistically related to the occurrence of VT (P = .02). The transfusion intensity of PRBC (P = .07) but not the percentage of transfused patients (P = .40) was associated with VT complications. A very small number of patients required platelet transfusion.
Table I. Perioperative data in patients undergoing AAA procedure with and without postoperative thrombosis⁎
| No thrombosis (51 patients) | Thrombosis (17 patients) | P Mann-Whitney | P χ2 | |
|---|---|---|---|---|
| Preoperative data | ||||
 Age (years) | 72.5(7.2) | 73.2(7.1) | 0.69 | |
| BMI | 26.3(4.5) | 27.9(4.4) | 0.27 | |
| Cardiovascular history (n = 40) | 60.7% | 52.9% | 0.77 | |
| Prior venous thrombosis (n = 5) | 5.9% | 11.7% | 0.59 | |
| AAA anteroposterior diameter (mm) | 58.1(14.6) | 56.1(10.2) | 0.60 | |
| Preoperative antiplatelet treatment (n = 43) | 34.0% | 41.1% | 0.77 | |
| Preoperative anticoagulant treatment (n = 9) | 86.2% | 8.2% | 1 | |
| Surgery | ||||
| Time of surgery (mn) | 167(51) | 161(83) | 0.30 | |
| Time of clamping (mn) | 59.2(24.8) | 67.1(33.1) | 0.59 | |
| Estimated blood loss (mL) | 1300(1100) | 1160+(855) | 0.81 | |
| Time in intensive care unit | ||||
| Number of days | 2.6(4.5) | 3.7(4.2) | 0.60 | |
| >2 days | 25.4% | 41.1% | 0.36 | |
| Delay to start of thromboprophylaxis | ||||
| Number of days | 0.9(0.9) | 1.7(1.4) | 0.09 | |
| >1 day | 20.0% | 35.3% | .32 | |
| Blood product transfusion | ||||
| Red blood cells - % | 43.1 | 48.8 | 0.40 | |
| - nb of products | 1.4(1.9) | 3.3(3.8) | 0.07 | |
| Fresh frozen plasma - % | 9.8 | 35.2 | 0.02 | |
| - nb of products | 0.3(1.0) | 1.2(1.8) | 0.01 | |
| Platelet concentrate - % | 5.9 | 17.6 | 0.32 | |
| Hemogram | ||||
| Hb (g/dL) | ||||
| - preoperative | 13.8(1.5) | 13.8(0.9) | 0.94 | |
| - postoperative | 10.2(1.5) | 9.7(1.3) | 0.14 | |
| Platelets (g/L) | ||||
| - preoperative | 228(59) | 212(37) | 0.36 | |
| - postoperative | 140(36) | 125(37) | 0.21 | |
| Fibrinogen (g/L) | ||||
| - preoperative | 3.6(0.7) | 3.7(0.6) | 0.52 | |
| - postoperative | 2.0(0.5) | 1.9(0.5) | 0.80 |
⁎Results are expressed as mean (SD: standard deviation) or %. |
Table II. Relation between blood loss, transfusion, and delay to start of thromboprophylaxis
| Delay to heparin prevention <24 hours | Delay to heparin prevention >24 hours | P Mann-Whitney | P χ2 | |
|---|---|---|---|---|
| Estimated blood loss | ||||
| - total (mL) | 1040(1040) | 1850(821) | 0.005 | |
| - >1 liter | 29.4% | 76.9% | .007 | |
| Transfusion of blood products | ||||
| Red blood cells | ||||
| - % | 39.2 | 75 | 0.02 | |
| - nb of concentrates | 1.3(2.1) | 3.6(3.5) | 0.006 | |
| Fresh frozen plasma | ||||
| - % | 9.8 | 7.5 | 0.02 | |
| - nb of concentrates | 0.3(1.0) | 1.3(1.9) | 0.01 |
There was neither pertinent information obtained from the perioperative hemogram charts, nor argument for a chronic disseminated intravascular coagulopathy. After surgery, an almost 40% drop in hemogram values and in fibrinogen levels due to bleeding and hemodilution were observed in both groups.
Discussion
This study was conducted over a 2-year period in order to meet enough thrombotic events. Even if the corresponding control number was limited, the time necessary to study all of them was out of reach. Our retrospective search for VT risk factors with a case control study was then settled on an objective selection of the controls by randomization (done by a statistician) giving representative controls.
No recommendations concerning thromboprophylaxis have been clearly established for patients undergoing aortic surgery, although this setting could be considered high-risk for VT. These patients gather several risk factors including advanced age, limb ischemia, long duration of surgery, and intraoperative local trauma, including possible venous injury.8 The incidence of clinically overt VT occurring during the hospital stay or requiring re-hospitalization within 3 months after surgery is 2.5 to 2.9%.9 The incidence of asymptomatic VT varies widely, ranging from 2% to 41%, and depends firstly on the diagnostic method used and secondly on the presence or not of prophylaxis.10 In five prospective studies of vascular surgery patients not receiving thromboprophylaxis, the pooled rate of postoperative DVT was 21% (18 out of 86 patients) using contrast venography11, 12, 13 and 15% (15 out of 98 patients) using Doppler ultrasonography scan.14, 15 Asymptomatic DVT after aorto iliac or aorto femoral surgery would be similar to that reported in other abdominal and pelvic procedures.9 There are few studies specifically done on DVT after AAA surgery (Table III).12, 14, 16 Aortic aneurysm resection or aortofemoral bypass appears to confer a higher risk of DVT than does femoro-distal bypass. More recently, Hollyoak et al14 reported the results of a pre-hospital discharge duplex ultrasound (DUS) study of 50 vascular patients without thromboprophylaxis, and showed a higher rate of DVT in aortic surgery patients (41%) compared to patients undergoing peripheral artery surgery (18%). Farkas et al17 randomized patients to receive either a low subcutaneous dose of UH or LMWH following vascular surgery. Using DUS screening at 7 to 10 days after surgery, with venography confirmation of a positive DUS result, DVT was detected in 8% of patients (11 out of 146) who underwent aortic surgery and in 3% of those who underwent femoro-distal bypass (3 out of 87). Fletcher et al18 published a routine DUS study in vascular surgery patients who received prophylaxis with intermittent pneumatic compression associated with low dose unfractionated heparin (LDUH). The incidence of DVT among patients undergoing aortic surgery or femoro-distal bypass was 12% (6 out of 52) and 9% (5 out of 54 patients), respectively. A prospective registry study of vascular surgery in Finland reported 0.9% of symptomatic DVT after aortic surgery and 0.7% after femoro-distal reconstruction. In the present study, the incidence of DVT was found to be 9.5%, 1% symptomatic, and 8.5% asymptomatic, which is in accordance with the literature. Unlike Eagleton et al, we did not observe a significant difference between the two techniques of AAA treatment but endovascular presented less DVT than open repair (5.3% vs 10.2%; P = .28). Nevertheless, both patients who presented symptomatic PE were treated with open surgery. Furthermore, some DVT were limited to the calf, a feature that was not encountered in Eagleton's series, even though all our patients were examined with DUS only. Systematic duplex scan examination is not recommended after surgery, but the question remains unanswered.
Table III. Incidence of venous thrombosis (by duplex US) after AAA surgery despite thromboprophylaxis
| Lower limb examination | n | Postoperative anticoagulant prophylaxis | Total DVT (%) | Proximal DVT (%) | |
|---|---|---|---|---|---|
| Farkas15 | duplex US | 75 | UFH/LMWH | 7.5 | 2.7 |
| Fletcher16 | duplex US | 52 | UFH | 11.5 | 6.3 |
| Eagleton14 | duplex US | 50 | UFH(ifpreoperativeOAC) | 6 | 6 |
| Killewich13 | duplex US | 88 | UFH | — | 2 |
| Present study | duplex US | 193 | LMWH | 8.5 | 1.5 |
Different data that could be considered as potential risk factors for VT were assessed. The delay between the end of the operation and the first injection of antithrombotic drugs was found to be significantly longer in the VT group (1.7 days vs 0.9 days, P = .09). This delay could be explained by bleeding (P = .05) and by the need for the transfusion of blood products (P = .02), which made surgeons cautious about prescribing a VT prophylaxis injection the same day. The occurrence of VT was related to the transfusion of FFP, but not to that of PRBC. These data reflect together that intensive transfusion could represent a predictive risk factor of VT, and one explanation may be the delay for the beginning of heparin treatment (administration of FFP when four or more units of PRBC were transfused). No patient was excluded for any reason in this continuous series, even if a prescription delay may have been involved in the VT occurrence. These data could advise us to start thromboprophylaxis as soon as possible after the end of the operation, and as much as possible the same day, even in the presence of bleeding. Other prophylactic methods such as intermittent calf compression and elastic compression stockings should be considered systematically during the whole hospital stay, but were only used intermittently in our series, after ICU discharge. Thrombotic events were observed early after surgery (day 2 for both symptomatic PE and first postoperative week for asymptomatic VT). Further studies should take account of these data. Occurrence of postoperative VT could be related to other causes, such as activation of coagulation by aneurysmal disease (reported chronic disseminated intravascular coagulation). Some of the thrombotic complications may stem from the underlying aneurysmal disease. The development of AAA is known to be associated with local activation of coagulation.19 Leukocyte recruitment is promoted by proinflammatory cytokines and locally produced chemokines. The mural thrombus in the aneurysm sac could also play a role in trapping and releasing leucocytes, platelets, proteases, and plasminogen and its activators.20 A state of activation of both coagulation and fibrinolysis was found in AAA patients before surgery.21, 22 The Cochrane collaboration study recently reviewed published studies about anticoagulant prevention with or without mechanical devices in AAA surgery. Systematic thromboprophylaxis was not recommended but the conclusions were based on only two relevant studies.15, 23, 24 The first study done in the 70s was stopped after only 50 inclusions out of 150 planned because of bleeding complications, but surgical techniques have been improving for the last 30 years. The second study included 98 patients divided in two groups (with or without heparin treatment). Only one thrombotic event (2%) occurred in each group. The low incidence of VT could be explained by the absence of venous examination in the calf. It must be noticed that one person diagnosed a DVT, eventually developed a PE. The 1% incidence of symptomatic PE should not be neglected and further studies with larger population are needed to conclude.
Conclusion
In conclusion, surgery for AAA induces a high incidence of postoperative VT, despite systematic prevention with heparin. In our series, the transfusion of blood products was found to be predictive of developing postoperative VT complications, and a tendency was observed with the delay to the first injection of antithrombotic agent. Coagulation activation has been reported in patients presenting with AAA and this setting could also predispose patients to VT but has not been explored in this study. Most events were asymptomatic, but the 1% incidence of symptomatic VT in this study should lead us to optimize thromboprophylaxis after aortic surgery; indeed this 1% incidence is close to the residual incidence of VT after major orthopedic surgery. Even though this series is one of the largest ever published on this topic, the rarity of the events calls for confirmation with a larger prospective study.
Author contributions
We would like to thank Philippe d'Athis for statistical analysis and Philipp Bastable for English language revision.
References
- . Prevention of venous thromboembolism. Chest. 1992;102(Suppl 4):391S–407S
- . Prevention of venous thromboembolism in general surgical patients (Results of meta-analysis). Ann Surg. 1988;208:227–240
- . A meta-analysis of thromboembolic prophylaxis following elective total hip arthroplasty. J Bone Joint Surg Am. 2000;82-A:929–938
- . Venous thromboembolism and mortality after hip fracture surgery: the ESCORTE study. J Thromb Haemost. 2005;3:2006–2014
- . Fondaparinux vs enoxaparin for the prevention of venous thromboembolism in major orthopedic surgery: a meta-analysis of 4 randomized double-blind studies. Arch Intern Med. 2002;162:1833–1840
- Prevention of venous thromboembolism: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133(Suppl 6):381S–453S
- . Diameters of acute proximal and distal deep venous thrombosis of the lower limbs. Int Angiol. 1998;17:260–267
- . The prevalence of risk factors for venous thromboembolism among hospital patients. Arch Intern Med. 1992;152:1660–1664
- Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(Suppl 3):338S–400S
- . Prevention of venous thromboembolism in surgical patients. Circulation. 2004;110(24 Suppl 1):IV4–IV12
- . Investigation of oedema of the lower limb following successful femoropopliteal by-pass surgery: the role of phlebography in demonstrating venous thrombosis. Br J Surg. 1972;59:979–982
- . The incidence of deep venous thrombosis in patients undergoing abdominal aortic aneurysm resection. J Vasc Surg. 1993;18:1037–1041
- . Leg edema following femoropopliteal autogenous vein bypass. Arch Surg. 1972;105:883–888
- . Deep venous thrombosis in postoperative vascular surgical patients: a frequent finding without prophylaxis. J Vasc Surg. 2001;34:656–660
- . A randomized, prospective trial of deep venous thrombosis prophylaxis in aortic surgery. Arch Surg. 1997;132:499–504
- Endovascular treatment of abdominal aortic aneurysm is associated with a low incidence of deep venous thrombosis. J Vasc Surg. 2002;36:912–916
- A randomised controlled trial of a low-molecular-weight heparin (Enoxaparin) to prevent deep-vein thrombosis in patients undergoing vascular surgery. Eur J Vasc Surg. 1993;7:554–560
- . Incidence of deep vein thrombosis following vascular surgery. Int Angiol. 1997;16:65–68
- . An activated state of blood coagulation and fibrinolysis in patients with abdominal aortic aneurysm. Am J Surg. 1998;175:297–301
- Renewal of mural thrombus releases plasma markers and is involved in aortic abdominal aneurysm evolution. Am J Pathol. 2006;168:1022–1030
- Clinical implication of plasma level of soluble fibrin monomer-fibrinogen complex in patients with abdominal aortic aneurysm. J Vasc Surg. 2005;42:200–205
- Influence of intraluminal thrombus on structural and cellular composition of abdominal aortic aneurysm wall. J Vasc Surg. 2003;38:1283–1292
- . Interventions for preventing venous thromboembolism following abdominal aortic surgery. Cochrane Database Syst Rev. 2008;(1):CD005509
- . Low dose heparin in the prevention of deep-vein thrombosis after aortic bifurcation graft surgery. Thromb Haemost. 1980;42:1429–1433
This study was supported by a grant from the University Hospital of Dijon and the Regional Council of Burgundy. Reference: AOI 2003.
Competition of interest: none.
PII: S0741-5214(08)01674-1
doi:10.1016/j.jvs.2008.10.005
© 2009 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
