Combined regional thrombolysis and surgical thrombectomy for treatment of iliofemoral vein thrombosis
Article Outline
Objective
In at least half of patients with iliofemoral deep vein thrombosis post-thrombotic syndrome develops when only anticoagulant therapy is given. We combined thrombolysis, applied under ischemic conditions,with surgical thrombectomy to restore patencyand valve function. The technique and the short-term and long-term results in 2 patient series are reported.
Methods
A catheter was inserted into a foot vein of the thrombosed leg, and the limb was excluded from the circulation with a pneumatic cuff placed on the thigh with the patient under general anesthesia. Urokinase (0.5 million–3 million IU) and heparin were infused and allowed to act for 30 minutes while the pelvic axis was cleared with a Fogarty catheter through an inguinal venotomy. The external iliac vein was then clamped and the cuff removed. Thrombi that detached from the wall were flushed out with reactive hyperemia and squeezed out with manual leg compression. The blood was retrieved, washed, and transfused back into the patient. Various additional procedures were performed to secure outflow. Two patient series are reported: 1 with 12 consecutive patients and 1 with 21 patients who were successfully treated 6 to 10 years previously. Follow-up data were obtained for all patients after 1 year and for 18 of 21 patients after 6 to 10 years. Patency and valve function were assessed with duplex scanning or venography. Studies of blood coagulation and the kinetics of urokinase were performed in 5 additional patients.
Results
Vein patency and valve function were restored in all consecutive patients. At 1 year none of the 33 patients had had recurrence, and none showed clinical signs of post-thrombotic syndrome. At 6 to 10 years 3 of 18 patients had experienced another venous thromboembolism, but none in the treated leg. Sixteen legs were asymptomatic without compression therapy, and 2 had venous claudication. Coagulation studies showed a trace concentration of urokinase and a mild decrease in fibrinogen in the systemic circulation. The concentration of urokinase in blood collected from the treated leg was only 1% of that infused.
Conclusion
Regional thrombolysis combined with surgical thrombectomy is relatively easy to perform and seems safe. Vein patency and valve function were restored, and post-thrombotic syndrome was prevented. Additional procedures to overcome pelvic vein obstructions were required in 11 of 33 patients (33%). The procedure should be tested against standard anticoagulation therapy in patients with acute iliofemoral thrombosis.
The immediate risks of acute deep vein thrombosis (DVT) are effectively reduced with anticoagulation therapy.1 However, resolution of thrombi is slow and often incomplete. Remaining thrombi harbor viable thrombin, which may cause recurrences when anticoagulant is withdrawn.2, 3, 4 They also obstruct outflow and impair valve function, with ensuing venous reflux.5, 6, 7 Thus half of patients with a first proximal DVT sustain post-thrombotic syndrome, with clinical signs and symptoms manifested as soon as 6 months.8, 9 To prevent these severe late sequela strategies have been developed to dissolve or remove the thrombi, including systemic thrombolysis, surgical thrombectomy, and catheter-directed thrombolysis. These methods all bear intrinsic risks, such as pulmonary embolism, major bleeding, early recurrence, and technique-related morbidity. To date, the role of these techniques has not been clearly established.10, 11, 12, 13, 14
We developed an intervention that combines regional thrombolysis, applied under ischemic conditions, with thrombectomy, with a primary success rate of 80%.15, 16
This report includes 2 series of patients who underwent this treatment, 1 with consecutive patients and 1 with patients treated successfully 6 to 10 years previously, and a study of the possible mechanism of action of the technique.
Methods
Criteria for patient selection
Patients with severely symptomatic, extensive proximal DVT were considered for combined thrombolysis and thrombectomy when it seemed likely that full restitution could be achieved with the intervention and when they had no contraindications. Patients with malignancy or other severe concomitant disease and patients with unknown duration of symptoms were excluded. Thus candidates typically had a short history of symptoms (<1 week), a first episode of DVT, and descending rather than ascending iliofemoral DVT (Table I). Diagnosis was made with ascending venography or color-coded duplex ultrasound scanning. Involvement of the vena cava was not routinely checked. Written informed consent was obtained from all patients.
Table I. Patient characteristics in 2 study cohorts
| Berne cohort of consecutive patients | Lucerne cohort of selected patients | |||
|---|---|---|---|---|
| n | % | n | % | |
| No. of patients (female) | 12 | 11 | 21 | 18 |
| Age (y) | ||||
 Median | 23 | 25 | ||
| Range | 16-44 | 17-61 | ||
| First vs recurrent DVT | 12 | 0 | 19 | 2 |
| Location, left vs right | 10 | 2 | 17 | 4 |
| Mean duration of symptoms (d) | 3 | 3.5 | ||
| Causes of DVT* | ||||
| Idiopathic | 2 | 3 | ||
| Post surgery or trauma | 5 | 5 | ||
| Use of oral contraceptive pill | 6 | 7 | ||
| May-Thurner syndrome | 2 | 5 | ||
| Venous anomaly | 0 | 1 | ||
| Known genetic thrombophilia | 2 | 4 | ||
| Other transient risk factors | 2 | 4 | ||
* Multiple causes possible. |
Interventional technique
The operative technique was worked out by Dr Largiadèr in Lucerne and subsequently adopted with slight modifications in Bern by Dr Schmidli and colleagues. The procedure is as follows. A short 12-gauge catheter is introduced into a foot vein in the thrombosed limb. With the patient under general anesthesia the leg is elevated to empty the veins, and a large silicone tourniquet cuff (Medizintechnik GmbH) is placed on the thigh and inflated to 300 mm Hg to exclude the leg from the systemic circulation. Urokinase HS (Medac), the purified enzyme from human urine, is infused into the leg together with 5000 U of unfractionated heparin and up to 1000 mL of saline solution, distributed by means of leg massage, and allowed to act for 30 minutes. Meanwhile the common femoral vein is exposed and the proximal veins are cleared with a Fogarty catheter, with use of standard techniques. The femoral vein is then clamped proximal to the venotomy, and the tourniquet is removed. The thrombi, detached from the vein wall and partly degraded, are flushed out by an enormous flow. The leg is massaged again, from the foot to the thigh, to retrieve residual loose clots. The blood is aspirated, washed, and transfused back into the patient. The venotomy and the skin are closed. Full-dose intravenous heparin, oral anticoagulants, and compression treatment are started. In the case of persistent proximal stenosis or occlusion, additional measures are considered, such as deployment of a pelvic venous stent, creation of an arteriovenous fistula, or insertion of a femorofemoral crossover bypass, either immediately or later as a separate intervention.
Cohort of 12 consecutive patients
This prospective survey of consecutive patients explored the short-term success rate and safety of the procedure. Patients received treatment at the Cardiovascular Centre of the University Hospital of Bern (Inselspital), Switzerland, between May 2000 and February 2003. Demographic and clinical characteristics are listed in Table I, and the extent of DVT is shown in Fig 1. The dose of urokinase was uniformly 0.5 million IU. An arteriovenous fistula was created between a branch of the greater saphenous vein and the common femoral artery in 10 of 12 patients, and was closed after 6 months. In 4 patients an iliac stent (Wallstent) was placed during surgery. Follow-up examination were performed 2 and 12 months after the intervention. Standard color-coded duplex ultrasound techniques were used to assess vein patency and function of the femoropopliteal valves.17
Fig 1.
Extent of thrombosis in 2 study cohorts.
Cohort of 21 patients treated 6 to 10 years previously
The purpose of this study was to assess the long-term outcome in those patients who had undergone successful combined thrombolyis and thrombectomy. From a list of patients who received treatment at the Kantonsspital in Lucerne, Switzerland, between January 1993 and August 1996 we searched for patients who fulfilled the following criteria: symptoms for no more than 5 days before treatment, and complete restitution of the venous system documented on normal postoperative ascending and descending venograms. In 2 patients, however, persistent occlusion of the common iliac vein was noted on the venogram, which was initially interpreted as normal. Patients with incomplete documentation were excluded. We identified 21 patients, and invited them for a follow-up visit with a vascular physician by the end of 2003 or interviewed them by telephone if they could not come to the clinic. Demographic and clinical characteristics are listed in Table I, and extension of the DVT is shown in Fig 1. The dose of urokinase varied in these patients, from 0.5 million to 3 million IU. Femorofemoral bypass grafts were inserted, with use of the contralateral greater saphenous vein as a conduit in 4 patients and a polytetrafluoroethylene graft in 1 patient. Pelvic venous stents (Wallstent) were placed in 2 patients.
Assessment of haemostasis and kinetics of urokinase
Parameters of coagulation were studied in 5 additional patients. These patients received an infusion of 3 million IU (∼43 mg) of urokinase. Three blood samples were collected from the femoral vein, the first immediately on looseningthe tourniquet, the second 2 minutes later, and the last before closure of the venotomy. Four samples were obtained from the systemic circulation through a central venous catheter before, duringn and at the end of the intervention. The following assays were performed: fibrinogen, with clotting time technique; plasminogen, with an amidolytic assay; d-dimer, with an enzyme-linked immunosorbent assay (ELISA); and urokinase, with an ELISA (American Diagnostica). The concentration of urokinase was measured in milligrams per milliliter, with 1 mg equal to ∼70,000 IU.
Results
Early results and safety (Table II)
The results of the 2 patient series are combined, because the main intervention was the same: regional thrombolysis, applied under ischemic conditions, combined with surgical thrombectomy. In all patients thrombi in the proximal veins were removed. An example of how the thrombi usually appear in the operating field is depicted in Fig 2. In each center 2 patients underwent repeat operation because of early proximal recurrence. Placement of an iliac stent or a bypass procedure was performed in conjunction with the second thrombectomy. Before discharge all patients were examined with duplex scanning or ascending and descending venography. Remaining crural thrombi were found in 1 patient, and iliac vein occlusion persisted in 3 patients. In 1 of these patients a stent was successfully placed later, and 2 patients were prescribed long-term anticoagulation therapy. No deep venous reflux was documented in any patient.
Table II. Interventional procedures and immediate outcome
| Berne cohort of consecutive patients | Lucerne cohort of selected patients | |
|---|---|---|
| No. of patients | 12 | 21 |
| Operating time (min) | ||
| Median | 110 | ND |
| Range | 55-240 | ND |
| Hospital stay (d) | ||
| Median | 6 | ND |
| Range | 4-10 | ND |
| Primary success | 12 (100%) | 21 (100%) |
| Early recurrence or repeat procedures | 2 | 2 |
| Additional interventions | 14 | 7 |
| Femorofemoral venous bypass | 0 | 5 |
| Stent placement | 4 | 2 |
| Arteriovenous fistula | 10 | 0 |
| Transfusion requirement | 0 | ND |
| Hematoma requiring drainage | 0 | 6 |
| Systemic bleeding | 0 | 0 |
| Clinical pulmonary embolism | 0 | 0 |
| Other complications | 0 | ND |
| Postoperative assessment | ||
| Thrombi remaining in proximal veins | 0 | 0 |
| Thrombi remaining in crural veins | ND | 1 |
| Pelvic vein still occluded | 0 | 3 |
Fig 2.
Usually, the thrombi appearing in the operating field are only partly degraded. Here, they are assembled to demonstrate the completeness of the thrombectomy. The impressions of the valves are depicted in the insert.
No major technical problems or severe adverse events were encountered in either center. No systemic bleeding was observed, and no clinical suspicion of pulmonary embolism was raised. Local hematoma requiring drainage occurred in Lucerne, where urokinase was often used at a higher dose.
Outcome 1 year after the intervention (both cohorts; Table III)
None of the 33 patients experienced clinically apparent recurrence or a new venous thromboembolism within the first year, even though anticoagulation therapy was stopped in most patients. Clinical signs of post-thrombotic syndrome were absent in all but 1 patient. In this patient post-thrombotic syndrome was ascribed to an earlier DVT. Leg compression had been prescribed for 6 months or less in most patients.
Table III. Findings 1 year after intervention
| Berne cohort of consecutive patients | Lucerne cohort of selected patients | |
|---|---|---|
| No. of patients | 12 | 21 |
| Recurrence | 1 | 0 |
| Persistent pelvic vein occlusion | 1 | 2 |
| Deep venous reflux | 0 | 0 |
| Duration of anticoagulant therapy (mo) | ||
| <3 | 0 | 5 |
| 3–6 | 11 | 6 |
| 6–12 | 1 | 8 |
| ND | — | 2 |
| Use of medical compression stockings (mo) | ||
| None or ≤ 1 mo | 0 | 8 |
| ≤3 | 0 | 5 |
| ≤6 | 12 | 3 |
| ≤12 | 0 | 3 |
| ND | — | 2 |
Outcome 6 to 10 years after intervention (Lucerne cohort)
Data from 18 of 21 patients (86%) were obtained on average 8 years after treatment, and 15 patients were examined by a physician. Three episodes of venous thromboembolism (17%) were documented, but none affected the initially treated limb. Two patients had cutaneous signs of chronic venous hypertension (CEAPC4), however, preexisting in 1 patient and located on the contralateral leg, which was affected during follow-up. Venous claudication was present in 2 mjn patients, with occluded pelvic veins and an occluded femorofemoral venous bypass, respectively. Since the operation, 8 of 18 patients have had no serious illnesses or injuries, but 10 had diseases including diabetes (2 patients), polyarthronosis, inoperable heart disease, aneurysm of the thoracic aorta, sarcoidosis, vasculitis with renal transplantation, and psychosis, and 2 patients had multiple pregnancies with normal outcomes.
Hemostasis parameters and kinetics of urokinase
In the systemic blood circulation a decrease in fibrinogen was found, from 3.3 g/L (range, 2.3-4.8 g/L) before the start of the operation to 1.55 g/L (range, 0.3-2.6 g/L) at the end of the operation. No further decrease in fibrinogen was observed. The d-dimer concentration increased, from 6.6 ng/mL (range, 3.3-11.6 ng/L) before the operation to 13.0 ng/mL (range, 9.3-20.0 ng/L) at the end of the operation.
In the blood collected from the leg on loosening of the tourniquet the concentration of fibrinogen was markedly reduced (mean, 0.38 g/L; range, 0-0.8 g/L). Plasminogen was also lowered, but not totally consumed (mean, 21%; range, 14%-29%). A few minutes later, just before closure of the venotomy, the average concentration of fibrinogen returned to 1.42 g/L (range, 0.4-1.9 g/L) and plasminogen level was 53% (range, 31%-82%).
Measurements of urokinase in blood are depicted in Fig 3. A small amount of urokinase (mean concentration, 1.83 mg/mL; range, 0.1-7.8 mg/mL) was found in systemic blood during the procedure. After restoration of the circulation a 10-fold increase was found, which, however, was still 100 times lower than the concentration measured in blood coming from the leg. But even this concentration was only about 1% of that infused into the ischemic limb. About 43 mg (corresponding to 3 million IU) of urokinase was infused. If we assume a blood volume in the leg of 1 L and an even distribution of the drug, the calculated intravascular concentration was about 43,000 mg/mL. However, the peak concentration in the first blood coming from the leg averaged only 470 mg/mL.
Fig 3.
Urokinase concentration in blood collected from the systemic circulation and the treated leg. The thrombosed leg was excluded from the systemic circulation by means of a tourniquet placed on the thigh, and infused through a foot vein with 3 million IU of urokinase (∼43 mg). Systemic blood was collected from a central venous catheter before (1) and during (2) the infusion, at closure of the venotomy (5), and at the end of the operation (6). Samples were obtained from the site of the inguinal venotomy immediately on loosening the tourniquet (3), when clots appeared no longer (4), and just before closure of the venotomy (5). Time between collection of samples 1 and 3 was 30 minutes, and between samples 3 and 6 was 30 to 45 minutes. Urokinase concentrations (mg/mL) are plotted on a logarithmic scale. Data are from 5 patients (mean ± 1 SD).
Discussion
In patients with extensive proximal DVT severe post-thrombotic syndrome is likely to develop despite standard anticoagulation therapy. Strategies to improve the outcome have been designed, but the heuristic goals of restoring both vein patency and valve competence were not predictably achieved. Furthermore, they require a multidisciplinary team of specialists, are time-consuming, and are associated with important risks such as pulmonary embolism and systemic bleeding.
We developed a relatively simple procedure to restore both vein patency and valve function. Thrombolysis applied to the leg excluded from the systemic circulation was combined with surgical thrombectomy. The 2 patient series reported here provide evidence that the intervention leads to complete clearance of thrombi from the proximal veins and is devoid of severe complications. It was associated with neither pulmonary embolism, because the clots were retrieved, nor systemic bleeding, inasmuch as the thrombolytic agent exerted only minor systemic activity. To prevent recurrence, remaining proximal occlusions were treated with additional interventions, which were different in the 2 cohorts. Inasmuch as the clinical results after 1 year were equally good in both series, we assume that all methods effectively secured the outcome.
The retrospective survey of patients treated long ago resolved 2 issues: in successfully treated patients post-thrombotic syndrome does not develop, and DVT does not recur in these legs. However, some patients may be prone to further venous thromboembolism.
Laboratory studies were performed to elucidate the specific features of the procedure. A decrease in systemic fibrinogen and an increase in the d-dimers was found on loosening the tourniquet. These transient changes were ascribed to the inflow of blood, which contained no fibrinogen but large amounts of d-dimers. We previously found no further decrease in fibrinogen and therefore have no evidence for ongoing systemic thrombolysis. However, that traces of urokinase were found in systemic blood shows that the tourniquet could not prevent some leakage. The tightness of the tourniquet is the key prerequisite for a safe intervention. The most astonishing finding, however, was the low concentration of urokinase measured in the blood coming from the leg. We recovered only 1% of the amount infused. We speculated that urokinase had left the circulation. We rejected the idea that it was degraded to the extent that it would have lost almost all of its immunoreactivity. If urokinase diffused into the surrounding tissue, it is not surprising that it did not exert a systemic effect. Protracted local activity, however, may occur and be responsible for the low frequency of recurrence. Secondary thrombolysis of remaining clots was occasionally documented. On the other hand, ongoing thrombolysis may jeopardize hemostasis. Indeed, most local hematomas occurred late after hemostasis had been secured, and predominantly in patients who received a higher dose of urokinase.
Ischemic exclusion of an extremity was used long ago to reduce blood loss during surgery, and later was found to produce persisting hyperemia.18, 19 The infusion of procaine into the empty veins of an extremity excluded from the circulation was introduced by August Bier in 1908 to achieve regional anesthesia.20 The same technique was used to treat chronic leg ulcers with urokinase,21 and cruropopliteal DVT with recombinant tissue plasminogen activator.22 Studies performed long ago showed that drugs administered in this way diffuse rapidly into areas that would otherwise be difficult to reach, such as clots, ischemic tissues, and skin ulcers.23, 24
In summary, our technique seems to work. Local thrombolysis, applied under ischemic conditions, combined with surgical thrombectomy enables clearing of thrombosed veins, preserves valve function, and retrieves the thrombi. As repeatedly documented, remaining proximal occlusions should be resolved by placement of endovascular stents or bypass. The intervention prevented development of post-thrombotic syndrome, and selected patients enjoyed long-lasting benefit. The technique should be compared with standard anticoagulation therapy in a randomized trial of patients with acute proximal DVT.
References
- Antithrombotic therapy for venous thromboembolic disease . Chest . 2001;119(suppl 1):S176–93
- . Thrombin activity of fibrin thrombi and soluble plasmic derivatives . J Lab Clin Med . 1983;102:220–230
- Residual venous thrombosis as a predictive factor for recurrent venous thromboembolism . Ann Intern Med . 2002;137:955–960
- . The incidence of recurrent thromboembolism after treatment with vitamin K antagonists in relation to time since first event: a metaanalysis . Arch Int Med . 2003;163:1285–1293
- . The venous stasis syndrome after deep vein thrombosis and pulmonary embolism: a population based study . Mayo Clin Proc . 2000;75:1249–1256
- . The prevalence of chronic lower limb ulceration has been underestimated: results of a validated population questionnaire . Br J Surg . 1996;83:255–258
- . The site of residual abnormalities in the leg veins in long-term follow-up after deep vein thrombosis and their relationship to the development of the postthrombotic syndrome . Int Angiol . 1996;15:14–19
- Randomized trial of effects of compression stockings in patients with symptomatic proximal venous thrombosis . Lancet . 1997;349:759–762
- Prandoni P, Frulla M, Mosena L, Sartor D, Marchiori E, Bernardi E, et al. Below-knee elastic compression stockings to prevent the post-thrombotic syndrome: a randomised, controlled trial. Ann Intern Med 2004;141(4):249-56.
- . Late results of iliofemoral venous thrombectomy . J Vasc Surg . 1997;25:417–442
- . Catheter-directed thrombolysis for lower extremity deep vein thrombosis: report from a national multicenter registry . Radiology . 1999;211:39–49
- . Iliofemoral deep vein thrombosis: conventional therapy versus lysis and percutaneous transluminal angioplasty and stenting . Ann Surg . 2001;233:752–760
- . Catheter-directed thrombolysis for deep vein thrombosis . Vasc Endovasc Surg . 2004;38:1–9
- . The rationale and evidence for the treatment of lower-extremity deep venous thrombosis with thrombolytic agents . Curr Opin Hematol . 2002;9:437–442
- . Stellenwert der Chirurgie 1996 in der Behandlung der akuten Becken-Beinvenen-Thrombose (The value of surgery 1996 in the treatment of acute iliofemoral vein thrombosis) . Gefasschirurgie . 1996;1:91–97
- . Therapeutic concept for acute leg and pelvic vein thrombosis . Acta Chir Belg . 2002;102:356–361
- . Extremity venous examination: technical considerations . In: Zwiebel WJ editors. Introduction to vascular ultrasonography . 4th ed. Philadelphia (PA): Saunders; 2000;p. 311–345
- . Die kontrollierte Kreislaufsperre: Historische Grundlagen und Zukunftsperspektiven (Use of pneumatic tourniquet: Historical basics and future perspectives) . Wien Med Wschr . 1993;7:144–146
- . Persisting hyperemia following prolonged arterial occlusion . Ann Surg . 1968;168:837–843
- . Ueber einen neuen Weg: Localanaesthesie in den Gliedmassen zu erzeugen (A new way to create local limb anaesthesia) . Arch Klin Chir . 1908;96:1007
- . Treatment of resistant leg ulcers by retrograde intravenous pressure infusions of urokinase . Phlebology . 1991;6:13–21
- . Lysoblocktechnik: Eine neue lokale Behandlungsform von Unterschenkelvenen- und arterienverschlüssen (New local therapy for lower leg arterial and venous leg occlusions) . VASA . 1992;21:289–293
- . Retrograde intravenöse Perfusion: Technik und verschiedene Varianten (Retrograde intravenous perfusion: Technique and variations) . Wien Med Wschr . 1993;7:167–170
- . Nuklearmedizinische Untersuchungen zur Gewebekonzentration und h&aynklmodynamische Effekte von retrograden intravenösen Druckinfusionen (Nuclear investigation for tissue harvesting and haemodynamic effects of retrograde intravenous infusion under pressure) . Wien Med Wschr . 1993;7:172–176
Competition of interest: none.
PII: S0741-5214(04)00952-8
doi:10.1016/j.jvs.2004.07.033
© 2004 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
