The quantitative benefit of isolated, segmental, pharmacomechanical thrombolysis (ISPMT) for iliofemoral venous thrombosis
Article Outline
- Abstract
- Patients and methods
- Technique
- Results
- Discussion
- Author contributions
- Acknowledgment
- References
- Copyright
Background
Early thrombus removal in patients with iliofemoral deep venous thrombosis (IFDVT) reduces postthrombotic morbidity. Preserving valve function and relieving venous obstruction prevents deterioration of quality of life and loss of economic potential. The preferred method for treating IFDVT is catheter-directed thrombolysis (CDT). Recently, isolated segmental pharmacomechanical thrombolysis (ISPMT) has emerged as a treatment option for patients with extensive IFDVT. The purpose of our study is to determine whether there are advantages to using ISPMT as the primary treatment for patients with iliofemoral IFDVT and, if so, to quantify those advantages relative to CDT.
Methods
Forty-three patients with IFDVT were treated with percutaneous CDT between May 2003 and June 2007. Twenty-one patients (27 limbs) were treated with CDT and 22 patients (25 limbs) were treated with ISPMT ± CDT. Demographics, extent of thrombus, procedural details, and thrombus resolution were recorded.
Results
Treatment time (55.4 vs 23.4 hours; P < .0001) and dose of rt-PA (59.3 vs 33.4 mg; P = .0009) were decreased and overall lytic success (60% vs 80%; P = .0016) increased with ISPMT. Adjunctive venoplasty and stenting, complications, hospital length-of-stay (LOS), and intensive care unit LOS were similar between groups.
Conclusion
ISPMT offers more effective thrombus removal in less time and with a reduced dose of thrombolytic agent. However, decreased treatment time did not translate into decreased hospital or ICU stay. Longer-term follow-up is required to determine whether improved thrombus resolution translates to better functional outcome and reduced postthrombotic morbidity.
Iliofemoral deep venous thrombosis (IFDVT) is associated with significant postthrombotic morbidity if treated with anticoagulation alone.1, 2, 3 There is an increasing body of evidence suggesting that thrombus resolution is associated with markedly improved outcomes. Studies of acute deep venous thrombosis (DVT) treated with anticoagulation alone show that when spontaneous clot lysis occurs, especially if it occurs early, patients are likely to retain normal valve function.4
A multicenter trial of venous thrombectomy plus arteriovenous fistula vs anticoagulation alone in patients with IFDVT showed significant benefit in those randomized to operative venous thrombectomy.5, 6, 7 A case-controlled cohort study of catheter-directed thrombolysis for IFDVT demonstrated significantly better quality-of-life (QOL) at 16 and 22 months in patients successfully lysed compared to patients treated with anticoagulation alone.8 A small randomized trial of catheter-directed thrombolysis vs anticoagulation alone for IFDVT confirmed these prior observations.9 The Eighth American College of Chest Physicians' (ACCP) consensus conference recommendations for the management of venous thrombo-embolic disease recognized the potential benefit of a strategy of thrombus removal in patients with IFDVT, and such a strategy is now recommended for good-risk patients.10
Catheter-directed thrombolysis (CDT) has been used for years in patients with extensive DVT with good results; however, the dose of lytic agent required for treatment was large and treatment times long.11, 12, 13 With advancing technology, mechanical techniques have been integrated with catheter infusion approaches to lytic therapy with varying success and enthusiasm. A number of studies have evaluated pharmacomechanical techniques compared with percutaneous mechanical techniques alone, demonstrating improved outcomes with combined therapy.14, 15, 16, 17
A promising new technique is isolated, segmental, pharmacomechanical thrombolysis (ISPMT), which isolates the segment of thrombosed vein to be treated between two occluding balloons.18, 19 A small dose of lytic agent is infused into the target segment and the intervening catheter spins at 1500 revolutions per minute (RPM) for 15 to 20 minutes. Following aspiration of the liquefied and fragmented thrombus, the vein is re-evaluated and re-treated, if necessary. If the vein is cleared of thrombus, the catheter is repositioned to treat another thrombosed venous segment.
During the past 28 months, we have used this technique for the initial management of patients with IFDVT. The purpose of the study is to quantify the benefit of adding ISPMT to a strategy of thrombus removal compared to CDT without ISPMT in patients with IFDVT.
Patients and methods
Forty-three consecutive patients with symptomatic IFDVT in 52 limbs who were judged to be appropriate candidates for CDT or ISPMT form the basis of this study and were retrospectively reviewed following institutional review board approval. From May 2003 to October 2005, 27 limbs (13 right, 14 left) in 21 patients were treated with traditional CDT. Since October 2005, 25 limbs (7 right, 18 left) in 22 patients were treated with ISPMT using the Trellis (Bacchus Vascular, Santa Clara, Calif) catheter, followed by adjunctive CDT when necessary to clear persistent thrombus. Six patients were treated with ISPMT alone for IFDVT and 16 had ISPMT with CDT. Catheter-directed thrombolysis was always used to clear thrombus distal to the popliteal vein catheter entry site if present. Recombinant tissue plasminogen activator (rt-PA) was used in 93% of patients and urokinase in the remaining 7%.
The diagnosis of IFDVT was established with venous duplex ultrasonography scan. Phlebography confirmed the diagnosis in all patients. Most patients were evaluated for the presence of a thrombophilia, and most had spiral CT scans of the chest, abdomen, and pelvis to evaluate for pulmonary emboli and underlying thoracic, abdominal, or pelvic pathology which might be associated with their thrombotic disorder.
Preoperative thrombus burden was quantified by using a modification of the venous scoring method reported by Mewissen, et al,11 who calculated a thrombus score for seven venous segments. We included two additional segments (tibial and profunda femoris), for a total of nine venous segments. The thrombus score was 0 for a thrombus-free (patent) segment, 1 when there was a <50% luminal reduction by thrombus and/or stenosis, 2 for a >50% luminal reduction, and 3 for an occluded segment. The total thrombus score was determined by totaling the scores for the nine vein segments. Overall lytic success was calculated by subtracting the final thrombus score from the initial thrombus score and dividing by the initial thrombus score.
Patient demographic and laboratory data were recorded, including age, gender, admitting diagnosis, medical history, hospital length of stay (HLOS), intensive care unit length of stay (ICU-LOS), hemoglobin, hematocrit, and creatinine level. Additionally, extent of thrombosis, treatment time, dose of lytic agent, thrombus resolution, and adjunctive therapy were recorded. All complications were documented, including bleeding, cardiopulmonary, and renal complications. Bleeding complications were defined as follows: Major, any evidence of hemodynamic instability, serious distant bleeding, large hematoma, or patients receiving blood transfusions associated with any of the aforementioned conditions; Minor, puncture site bleeding, brief hemoptysis, <15% drop in hemoglobin or hematocrit, blood transfusions due to hemodilution, and minor bleeding at a prior operative site. Cardiopulmonary complications were defined as documented acute myocardial infarction with enzyme elevation (troponins, creatine phosphokinase-MB isoenzyme [CPK-MB]) and electrocardiogram (EKG) changes. Renal complications were defined as a doubling of baseline creatinine or need for hemodialysis.
Intensive care monitoring was used for all patients while receiving CDT. Systemic anticoagulation with heparin and vitamin K antagonist (VKA) followed lytic therapy in all patients and was achieved with intravenous unfractionated heparin targeting an activated partial thromboplastin time (aPTT) >2.0 × control or enoxaparin 1 mg/kg subcutaneously every 12 hours. Length of oral anticoagulation was determined on an individual basis. Leg elevation, elastic compression, and intermittent pneumatic compression were integral components of therapy both in the hospital and following discharge.
Technique
Ultrasound-guided access was obtained through an uninvolved common femoral or popliteal vein, depending on the location of the thrombus. If popliteal vein thrombus distal to catheter entry or infrapopliteal thrombus existed, the posterior tibial vein was cannulated for catheter-based lysis. After baseline phlebography, the thrombosed segment was traversed with a 0.035-inch guidewire. Catheter-directed thrombolysis was performed by placing a multiside-hole catheter for intrathrombus infusion. The rt-PA was diluted to 0.01-0.02 mg per mL of saline solution and the infusion rate was 1 mg per hour. Serial phlebography was obtained to monitor resolution, generally after each run of the Trellis catheter, after other forms of mechanical intervention, and every 12-16 hours of CDT.
When using ISPMT, the Trellis catheter was advanced over a guidewire through an 8 Fr sheath to treat the selected segment. Proximal and distal balloons, which can be inflated to fit blood vessels from 5-16 mm (and can be oversized), were inflated and 3-6 mg of rt-PA was infused between the balloons in 6-8 mL of saline solution (Fig). The wire between the balloons was activated and rotated at 1500 RPM for 20 minutes. Formed and fragmented thrombus was aspirated with a 20-30 cc syringe. Repeat segmental phlebography was performed prior to deflation of the balloons, thereby reducing (or avoiding) systemic exposure of the lytic and contrast agents, which was monitored with fibrinogen levels and creatinine.
Fig.
Vein segments treated with ISPMT using the Trellis catheter (Bacchus Vascular, Santa Clara, Calif). Arrows show inflated proximal and distal balloons.
Statistical analysis
Mean lytic success was compared in the two treatment groups by Student t test. The proportion of patients in groups – complete lysis (>74%), moderate lysis (50-74%), and minimal lysis (<50%) – was also compared among the treatment groups using Fisher's Exact Test.
Results
The etiology of DVT was apparent in most of our patients. Thrombophilia evaluations were incomplete in some due to anticoagulation and the issue of whether results would alter recommendations for care. Thrombophilia was identified in 28% of patients (24% CDT, 32% ISPMT ± CDT) and included AT III, factor V Leiden, and factor VIII. Twenty-one percent of patients were found to have a malignancy (24% CDT, 18% ISPMT ± CDT), including lymphoma, prostate, lung, and metastatic cancer. Other etiologies included May Thurner syndrome, retroperitoneal fibrosis, and Crohn's disease, and 9% of patients were thought to be idiopathic (Table I).
Table I. Etiology of DVT
| Etiology | CDT (n = 21) | ISPMT ± CDT (n = 22) | Total (n = 43) |
|---|---|---|---|
| Coagulopathy | 5 (23.8%) | 7(31.8%) | 12 |
| Surgery | 7(33.3%) | 3(13.6%) | 10 |
| Cancer | 5(23.8%) | 4(18.2%) | 9 |
| Idiopathic | 3(14.3%) | 1(4.5%) | 4 |
| Pregnant | 0(0.0%) | 1(4.5%) | 1 |
| Anatomic | 1(4.8%) | 6(27.3%) | 7 |
| Total | 21 | 22 | 43 |
The two treatment groups were similar except for the distribution of gender. The CDT group had significantly more males, compared to the ISPMT ± CDT group. Interestingly, the 6 patients who were able to be treated by ISPMT alone were all female (Table II).
Table II. Patient demographics and type and location of DVT
| CDT (n = 21) | ISPMT ± CDT (n = 22) | ISPMT alone (n = 6) | |
|---|---|---|---|
| Age | 52.6(17.6) | 44.0(21.0) | 36.2(16.8) |
| Onset of symptoms | 12.7(17.4) | 14.6(19.0) | 18.5(28.2) |
| Male gender⁎ | 14(66.7%) | 7(31.8%) | 0(0.0%) |
| Acute on chronic | 11(52.4%) | 13(59.1%) | 2(33.3%) |
| Caval involvement | 11(52.4%) | 7(31.8%) | 0(0.0%) |
| Femoropopliteal | 16(76.2%) | 12(54.5%) | 2(33.3%) |
| Bilateral iliofemoral | 6(28.6%) | 3(13.6%) | 0(0.0%) |
| Tibial | 12(57.1%) | 9(40.9%) | 2(33.3%) |
⁎P < .5. |
Total thrombus scores were 13.1 ± 5.7 vs 11.7 ± 4.9 (P = .34) for CDT vs ISPMT ± CDT, respectively. Caval involvement was 52% vs 32%; bilateral involvement was 29% vs 14%, femoropopliteal 76% vs 55%, and tibial 57% vs 41% for CDT vs ISPMT ± CDT patients, respectively. None of these differences were significant.
There were no significant differences in pre-treatment thrombus scores (P = .5). Overall lytic success, defined as the amount of thrombus removed after lytic therapy, was highest in the ISPMT ± CDT group (80%), and lowest in the CDT group (60%). The increase in lytic success (approximately 20%) among the ISPMT ± CDT group compared to CDT alone was significant (P = .002) (Table III).
Table III. Analysis of thrombus burden and lytic outcome. (Top) CDT vs CDT ± ISPMT; (Bottom) subset analysis of CDT vs CDT ± ISPMT vs ISPMT only
| CDT (n = 21) | ISPMT ± CDT (n = 22) | P value | |
|---|---|---|---|
| Patient thrombus burden and lytic outcomes | |||
| Per limb | |||
| Pre-treatment thrombus burden | 13.1(5.7) | 11.7(4.9) | .34 |
| Post-treatment thrombus burden | 5.7(4.4) | 3.0(3.7) | .0183 |
| Overall lytic success | 0.6(0.2) | 0.8(0.2) | .0016 |
| Complete lysis | 3(11.1%) | 7(28.0%) | .077 |
| Moderate 50-74 | 16(59.3%) | 16(64.0%) | |
| Minimal <50% | 8(29.6%) | 2(8.0%) | |
| Per patient | |||
| Treatment time (hours) | 55.4(21.3) | 23.4(21.5) | <.0001 |
| Lytic dose | 59.3(26.2) | 33.4(24.7) | .0009 |
| CDT (n = 21) | ISPMT + CDT (n = 16) | ISPMT (n = 6) | |
|---|---|---|---|
| Outcomes showing ISPMT-alone patient group | |||
| Per limb | |||
| Pre-treatment thrombus burden | 13.1(5.7) | 12.1(5.1) | 10.7(4.4) |
| Post-treatment thrombus burden | 5.7(4.4) | 3.1(4.1) | 2.7(1.9) |
| Overall lytic success | 0.6(0.2) | 0.8(0.2) | 0.7(0.2) |
| Complete lysis | 3(11.1%) | 6(31.6%) | 1(16.7%) |
| Moderate 50-74 | 16(59.3%) | 12(63.2%) | 4(66.7%) |
| Minimal <50% | 8(29.6%) | 1(5.3%) | 1(16.7%) |
| Per patient | |||
| Treatment time (hours) | 55.4(21.3) | 29.3(21.3) | 4.7(4.6) |
| Lytic dose | 59.3(26.2) | 39.6(24.4) | 13.8(13.9) |
Treatment time and lytic dose were also significantly different between the two treatment groups (P < .01). The average treatment time for CDT was 55 hours which was significantly higher than ISPMT ± CDT (P < .0001). Lytic dose was higher in the CDT group compared to ISPMT ± CDT (59 units vs 40 units, P = .012). Although other adjunctive therapy (rheolytic thrombectomy, ultrasound-accelerated lysis, etc) was used more often in CDT patients, the difference was not significant. HLOS (P = .69) and ICU-LOS (P = .92) were similar in both groups.
Venoplasty was performed in 90% of CDT patients and 87% of ISPMT ± CDT patients, and stents were used in 45% of CDT patients and 59% of ISPMT ± CDT patients. None of these differences were significant. There was no difference in the number of inferior vena caval filters used in each group; 52% of CDT patients had vena caval filters placed, with 64% of those filters placed prior to the current hospital admission, and 54% of ISPMT ± CDT patients received caval filters. Caval filtration was used in patients with nonocclusive clot in their cava and at the discretion of the treating physician.
A ≥15% drop in hemoglobin/hematocrit (Hgb/Hct) occurred in 19% of patients in both groups. Bleeding complications, most of which were mild, were no different between groups (23% CDT vs 19% ISPMT ± CDT, P = .99). The need for blood transfusions was similar in both groups. Two major complications occurred, one in each group. One CDT patient developed acute renal failure because of hemoglobinuria and dye load following adjunctive rheolytic thrombectomy, and one ISPMT ± CDT patient developed a symptomatic puncture site hematoma requiring blood transfusions.
Discussion
This analysis quantifies the benefit of ISPMT ± CDT vs primary CDT. Significantly more thrombus was removed with ISPMT in less than half the time and with 39% less plasminogen activator. Since post-treatment anticoagulation is standardized for both treatment groups, the reduction in persistent thrombus burden may result in less postthrombotic morbidity and risk of recurrence with ISPMT.
A weakness of this study is its retrospective design. Its strength, however, lies in the fact that it is a single-center experience, with the same physicians using both treatment modalities, practicing similar techniques, using the same plasminogen activator, and attempting to achieve maximal thrombus resolution.
Hull, et al,20 suggested that improvement in clot burden should be considered by regulatory authorities as a meaningful surrogate outcome measure for treatment trials. Residual venous thrombus is an important risk factor for recurrent DVT. Prandoni, et al,21 performed a prospective cohort study following acute DVT in patients with venous ultrasound scan and showed that recurrent venous thrombosis was more frequent in patients with residual thrombus. Agnelli, et al,22 performed a clinical outcomes based meta-analysis of randomized trials of thrombolysis and showed that patients with a lower residual clot burden have a lower frequency of recurrence. Since thrombin generation is related to clot burden23 and is required for recurrent thrombosis, any technique that reduces residual thrombus burden will likely have a meaningful clinical benefit to patients. Achieving this endpoint quicker and with less plasminogen activator is likely to improve the safety of the technique. As our experience with ISPMT developed, patients with traditional contraindications to lytic therapy, such as surgical patients and pregnant women, were considered and treated with catheter-based interventions.
The amount of plasminogen activator used in the ISPMT patients overestimates the amount of systemic exposure, since most of the plasminogen activator solution infused during ISPMT is aspirated. In a larger comparative experience, one would expect to see fewer bleeding complications with ISPMT. Only one major bleeding complication occurred with ISPMT, which was a popliteal puncture site hematoma.
The 2008 ACCP consensus conference on the management of venous thromboembolic disease recommends CDT as a treatment option in good-risk patients with IFDVT.10 They go on to suggest that the addition of mechanical techniques to CDT may offer benefits to patients. Many physicians who treat IFDVT using a strategy of thrombus removal incorporate mechanical techniques during thrombolysis. However, it appears that percutaneous mechanical techniques alone (without a lytic agent) are inadequate for successful thrombus resolution in most patients.
Vedantham, et al,14 evaluated several mechanical thrombectomy devices used as an adjunct to CDT. He concluded that mechanical thrombectomy alone was inadequate and that pharmacologic thrombolysis significantly improved outcomes. Bush, et al,16 drew the same conclusions when they demonstrated more effective thrombus extraction after adding a lytic agent to rheolytic thrombectomy. Lin, et al,17 formalized their 8-year experience when they reviewed 98 catheter interventions for acute DVT. Forty-six of those were catheter-directed thrombolysis alone and in 52 of them pharmacomechanical techniques were used. Unlike our observations, they showed a significant reduction in ICU stay and HLOS with pharmacomechanical techniques and a reduction in blood transfusions. Interestingly, 60% of their patients undergoing pharmacomechanical thrombolysis did not undergo phlebography.
Kasirajan, et al,15 reported their preliminary observations with rheolytic thrombectomy demonstrating that mechanical thrombectomy alone was inferior to the combination of pharmacomechanical thrombolysis. Parikh, et al,24 recently reported a multicenter experience of ultrasound-accelerated thrombolysis using the EKOS LySus catheter (EKOS Corp, Bothell, Wash). In 53 cases treated, complete lysis (≥90%) occurred in 79% after a mean infusion time of 22 hours and a 4% major complication rate. Multiple lytic agents were used across the eight centers included in this analysis. When attempting to evaluate the benefit of the new technique, they compared their observed results to a historical report by Grunwald and Hofmann,25 not to the CDT experience of the participating investigators.
O'Sullivan, et al,18 reported results of ISPMT alone for the management of acute DVT. Complete thrombus resolution was observed in only 14%, which is similar to our results (Table III). However, limiting intervention to ISPMT alone underestimates its value in the overall treatment scheme of patients with extensive DVT. ISPMT alone will leave larger thrombus burdens than if used as an initial approach to thrombus removal, with CDT or other techniques added as needed. Combining pharmacologic and mechanical techniques results in better overall lytic success than the use of any mechanical or lytic technique alone, a finding observed by a number of other authors.
It is now apparent that a strategy of thrombus removal, when successful, reduces postthrombotic morbidity.26, 27 We anticipate that recurrent venous thromboembolic events also will be reduced in successfully treated patients, since residual thrombus burden is less.
The use of ISPMT improves lytic success and requires less time and a smaller dose of lytic agent. It is a safe and effective approach even in patients with traditional relative contraindications for CDT. Therefore, ISPMT should expand the proportion of patients offered a strategy of thrombus removal.
Author contributions
This research was supported in part by the Conrad and Caroline Jobst Foundation, Toledo, Ohio.
References
- . The socioeconomic effects of an iliofemoral venous thrombosis. J Surg Res. 1977;22:483–488
- . Venous function assessed during a 5 year period after acute ilio-femoral venous thrombosis treated with anticoagulation. Eur J Vasc Surg. 1990;4:43–48
- . Venous claudication in iliofemoral thrombosis: long-term effects on venous hemodynamics, clinical status, and quality of life. Ann Surg. 2004;239:118–126
- . Deep venous insufficiency: the relationship between lysis and subsequent reflux. J Vasc Surg. 1993;18:596–605
- . Thrombectomy with temporary arteriovenous fistula: the treatment of choice in acute iliofemoral venous thrombosis. J Vasc Surg. 1984;1:867–876
- . Long-term results of venous thrombectomy combined with a temporary arterio-venous fistula. Eur J Vasc Surg. 1990;4:483–489
- . Venous thrombectomy for iliofemoral vein thrombosis–10-year results of a prospective randomised study. Eur J Vasc Endovasc Surg. 1997;14:367–374
- . Catheter-directed thrombolysis for iliofemoral deep venous thrombosis improves health-related quality of life. J Vasc Surg. 2000;32:130–137
- . Early results of thrombolysis vs anticoagulation in iliofemoral venous thrombosis (A randomised clinical trial). Eur J Vasc Endovasc Surg. 2002;24:209–214
- Kearon C, Kahn SR, Agnelli G, Goldhaber SZ, Raskob G, Comerota AJ. Antithrombotic therapy for venous thromboembolic disease: ACCP evidence-based clinical practice guidelines (8th ed). Chest. In press.
- . Catheter-directed thrombolysis for lower extremity deep venous thrombosis: report of a national multicenter registry. Radiology. 1999;211:39–49
- Iliofemoral deep venous thrombosis: safety and efficacy outcome during 5 years of catheter-directed thrombolytic therapy. J Vasc Interv Radiol. 1997;8:405–418
- . Catheter-directed thrombolysis for the treatment of acute iliofemoral deep venous thrombosis. Phlebology. 2001;15:149–155
- . Lower extremity venous thrombolysis with adjunctive mechanical thrombectomy. J Vasc Interv Radiol. 2002;13:1001–1008
- . Percutaneous AngioJet thrombectomy in the management of extensive deep venous thrombosis. J Vasc Interv Radiol. 2001;12:179–185
- . Pharmacomechanical thrombectomy for treatment of symptomatic lower extremity deep venous thrombosis: safety and feasibility study. J Vasc Surg. 2004;40:965–970
- Catheter-direct thrombolysis versus pharmacomechanical thrombectomy for treatment of symptomatic lower extremity deep venous thrombosis. Am J Surg. 2006;192:782–788
- . Pharmacomechanical thrombectomy of acute deep vein thrombosis with the Trellis-8 isolated thrombolysis catheter. J Vasc Interv Radiol. 2007;18:715–724
- . Clinical and economic evaluation of the Trellis-8 infusion catheter for deep vein thrombosis. J Vasc Interv Radiol. 2008;19:377–383
- . Quantitative assessment of thrombus burden predicts the outcome of treatment for venous thrombosis: a systematic review. Am J Med. 2005;118:456–464
- Residual venous thrombosis as a predictive factor of recurrent venous thromboembolism. Ann Intern Med. 2002;137:955–960
- . Thrombolysis vs heparin in the treatment of pulmonary embolism: a clinical outcome-based meta-analysis. Arch Intern Med. 2002;162:2537–2541
- Randomized trial of different regimens of heparins and in vivo thrombin generation in acute deep vein thrombosis. Blood. 2002;99:1965–1970
- Parikh S, Motarjeme A, McNamara T, Raabe R, Hagspiel K, Katzen BT, et al. Ultrasound accelerated thrombolysis for the treatment of deep venous thrombosis: initial clinical experience. J Vasc Interv Radiol. In press.
- . Comparison of urokinase, alteplase, and reteplase for catheter-directed thrombolysis of deep venous thrombosis. J Vasc Interv Radiol. 2004;15:347–352
- . A strategy of aggressive regional therapy for acute iliofemoral venous thrombosis with contemporary venous thrombectomy or catheter-directed thrombolysis. J Vasc Surg. 1994;20:244–254
- . Treatment of acute iliofemoral deep venous thrombosis: a strategy of thrombus removal. Eur J Vasc Endovasc Surg. 2007;33:351–360
Competition of interest: none.
PII: S0741-5214(08)01148-8
doi:10.1016/j.jvs.2008.07.013
© 2008 Published by Elsevier Inc.
