Mid-term results of endovascular treatment for symptomatic chronic nonmalignant iliocaval venous occlusive disease
Article Outline
Background
The goal of this article is to present clinical and patency results of endovascular treatment of nonmalignant, iliocaval venous obstructive disease and to discuss the evolution of technical details.
Methods
From November 1995 to June 2004, 44 patients (female-male ratio, 3.9:1; left-right lower limb ratio, 8.6:1; median age, 42 years; range, 21-80 years) had treatment for chronic disabling obstructive venous insufficiency with iliocaval stenosis or occlusion. The clinical class of CEAP was 2 in 11 limbs, 3 in 31, 4 in 4, 5 in 1, and 6 in 1; etiology was primary in 32 patients, secondary in 10, and congenital in 2. Anatomic involvement included superficial veins in 16 patients and perforator veins in 11. Obstruction was associated with superficial reflux in 4 patients, deep reflux in 13, and both in 13. Ten patients had occlusion. All procedures were performed in the operating room with perioperative angiography and angioplasty with or without self-expanding stent implantation. Venous clinical severity and disability scores were obtained before and after treatment. Patency and restenosis were evaluated by duplex Doppler ultrasonography.
Results
No perioperative death or pulmonary embolism occurred. The technical success rate was 95.5% (two recanalization failures), and two (4.5%) perioperative stent migrations occurred. One early thrombosis (2.4%) was treated by thrombectomy and creation of an arteriovenous fistula. One late death and one thrombosis occurred. Restenoses were found in five patients and were all treated successfully (four needed iterative stenting). Median follow-up was 27 months (range, 2-103 months). Median venous clinical severity score improved from 8.5 to 2, and median venous disability score improved from 2 to 0. Cumulative primary, assisted primary, and secondary patency rates of the venous segments at 36 months were 73%, 88%, and 90%, respectively, in intention to treat. The survival rate was 100% at 12 months and 97.3% at 60 months.
Conclusions
Endovascular treatment of benign iliocaval occlusive disease is a safe and efficient minimally invasive technique with good mid-term patency rates. Moreover, it improves cases with obstruction only, as well as cases with associated reflux and obstruction. Primary stenting should always be performed by using self-expanding stents deployed under general anesthesia to avoid lumbar pain. In case of failure, the endovascular procedure does not preclude further surgical reconstruction.
Iliocaval chronic occlusive lesions are very challenging therapeutic conditions that influence the patient’s quality of life and ability to work. Different surgical techniques (mainly bypasses) have been used in the past.1, 2, 3, 4 These procedures are invasive and have given variable, but globally disappointing results, mainly for complex lesions, with patency rates ranging from 54% to 88%.1, 2, 3 Balloon angioplasty and stents offer a new and very attractive approach to such lesions. Although we have used endovascular techniques since 1995,5, 6 the goal of this study was to review our results and to discuss the evolution of the initial technical choices according to our experience and the literature.
Materials and methods
From November 1995 to June 2004, 44 consecutive patients with chronic obstructive venous insufficiency were admitted (intention to treat) for endovascular treatment of symptomatic chronic nonmalignant iliocaval obstructive lesions (stenoses or occlusions). The first nine cases were already included in a previous article.5 All patients had previously received appropriate medical treatment, including compressive stockings, but remained disabled. Exclusion criteria were factors that limited life expectancy to less than 5 years, such as severe chronic obstructive pulmonary disease, cardiac failure, and malignant disease.
There were 35 women and 9 men (female-male ratio, 3.9:1; median age, 42 years; range, 21-80 years). Previous pregnancy was found in 28 patients. The thrombophilia test performed in patients with a history of deep venous thrombosis (DVT; n = 18) or pulmonary embolism (n = 3) included protein C and S, antithrombin III, factor II and V mutation, homocystinemia, and circulating anticoagulant antibodies. It found two cases of factor V mutation, one case of factor II mutation, and one case of factor XII mutation. Moreover, one patient had autoimmune thrombopenia. Twelve patients had a history of varicose vein surgery (10 long stripping of the greater saphenous vein, including perforant vein ligation in 2; 2 greater saphenous vein high ligation), 2 of urologic surgery for retroperitoneal fibrosis, 3 of hysterectomy, and 2 of pelvic radiotherapy (absence of residual tumoral mass).
The initial workup in all patients consisted of color Doppler ultrasonography of the lower extremities and pelvis to identify obstructive lesions and detect superficial and/or deep reflux and collateral pathways. We considered that the lesion was significant when there was a greater than 50% stenosis associated with inflow obstruction: asymmetric flow in the common femoral vein (CFV) and the presence of collateral pathways (the main one being the internal iliac vein with reversed flow). Computed tomographic venography (43 patients) or magnetic resonance venography (1 patient) was performed for ascertainment of the extent of the obstruction, visualization of the collateral pathways, or differential diagnosis of May-Thurner syndrome. Preoperative iliocavography to assess the morphology of the narrowing and visualize the collateral pathways was performed in 15 patients, systematically in the early days of our experience and later only in complex cases (mainly before recanalization). In case of uncertainty regarding the presence or severity of venous stenosis, profile iliocavography always made it possible to effectively assess the size of the venous lumen.
Ten women had symptoms of pelvic congestion syndrome, including five previously treated by embolization, and were re-explored for residual symptoms; only three of them had symptomatic leg disease. Patients with pelvic congestion syndrome only were classified C2. All the other patients had disabling chronic obstructive venous insufficiency and had already been treated with elastic stockings.
The CEAP classification clinical scores7 were as follows: C2 (varicose veins), 11 limbs (10 patients); C3 (edema), 31 limbs (28 patients); C4 (skin changes ascribed to venous disease), 4 limbs; C5 (cicatrized ulceration with skin changes), 1 limb; and C6 (active ulceration), 1 limb. These symptoms were usually unilateral for patients with left iliocaval stenosis, but three had bilateral lower extremity pain that increased in the evening and decreased in the decubitus position. Four other patients had bilateral symptoms associated with bilateral iliac veins and inferior vena cava (IVC) obstructive lesions. All patients with C4 to C6 lesions had a history of DVT except for one, who had retroperitoneal fibrosis. Only two patients classified as C2 had a history of DVT (one ipsilateral leg DVT and one left ovarian vein DVT).
The etiology was primary (May-Thurner syndrome) in 32 patients (73%), secondary in 10 (23%; postthrombotic disease in 8 and retroperitoneal fibrosis in 2), and congenital (hypoplasia) in 2 (4%). The anatomic repartition of the diseased venous segment was deep veins in all patients, but superficial and perforator vein lesions were present in 16 and in 11 patients, respectively.
The pathophysiology was obstruction in all patients. Moreover, duplex scan showed that 30 patients had an associated reflux: superficial reflux only in 4 patients, deep reflux only in 13 patients, and both in 13 patients. All patients with pelvic congestion syndrome had a reflux in the left internal iliac vein. In case of skin damage (6 patients), deep reflux to below the knee was always found. CEAP classification stratification is shown in Table I.
Table I. CEAP classification of 48 limbs (bilateral lesions treated in 4 patients)
| Category | No. patients |
|---|---|
| Clinical grade (all symptomatic) | |
 2 | 10(11limbs) |
| 3 | 28(31limbs) |
| 4 | 4 |
| 5 | 1 |
| 6 | 1 |
| Etiology | |
| Primary | 32 |
| Secondary | 10(12limbs) |
| Congenital | 2(4limbs) |
| Anatomy | |
| Deep only | 17(21limbs) |
| Deep/perforator | 11 |
| Deep/superficial | 16 |
| Pathophysiology | |
| Obstruction only | 14(18limbs) |
| Reflux/obstruction | 30 |
In all cases but one, there were occlusive lesions on the left side. Bilateral iliac and IVC lesions were found in four patients (9%). The left-right lower limb ratio was 8.6:1. Complete occlusion was present in 10 patients (23%) with a mean length of 133 mm. Extensive lesions (more than 1 venous segment) were found 11 times (25%). The homolateral CFV and external iliac veins (EIVs) were involved six times (14%).
All procedures were performed in the operating room under general (n = 24) or local (n = 20) anesthesia. Percutaneous access was performed through the CFV(s) in all patients but five, who had the surgical approach of the CFV (two with a history of DVT involving the CFV and three who had simultaneous greater saphenous vein high resection). After catheterization of the lesion and face with or without profile iliocavography, nadroparin 30 mg/kg, 500 mg of aspirin, and cefamandole for infection prophylaxis were given intravenously. Balloon angioplasty was performed before deployment of self-expanding metallic stents (Wallstent, Boston Scientific-Schneider, Minneapolis, Minn). Completion angiography was performed, and sheaths were retrieved before direct smooth compression. Elastic stockings were put on, and patients were taken to bed until day 1. The postoperative treatment was initially fluindione for 6 months; since 2003, patients have been discharged on nadroparin for 15 days, and antiplatelet drugs (unless there are other indications for fluindione) are present.
Follow-up was scheduled at 1, 3, 6, and 12 months and then annually. It included clinical examination and duplex scanning. Venous clinical severity (VCSS) and venous disability (VDS) scores8 were evaluated before surgery and during the follow-up to assess clinical results. In case of clinical recurrence of the symptoms or of restenosis at color Doppler ultrasonography (>50% of in-stent restenosis with inflow obstruction was considered significant), transfemoral venography was performed. Primary, assisted primary, and secondary patency rates were calculated by using survival analysis with the life-table method.
Results
No perioperative death or pulmonary embolism occurred. All stenoses were successfully treated (Fig 1). Eight recanalizations were performed with success (Fig 2), and two failed (bilateral common iliac vein [CIV] and IVC agenesis and one left femoroiliac chronic thrombosis). All in all, 57 venous segments were treated: 4 IVCs, 42 CIVs, 6 EIVs, and 5 CFVs. The technical success rate was 95.5% but was only 80% for recanalizations.
Fig 1.
Perioperative angiography in a patient with pelvic congestion syndrome caused by May-Thurner syndrome. A, Computed tomographic scan showing left common iliac vein stenosis. B, Angiography: tight stenosis of the left common iliac vein with large transversal collateral pathways. C, After angioplasty and stenting, the collateral pathways are no longer opacified.
Fig 2.
A 32-year-old woman who had had left deep venous thrombosis during pregnancy 2 years before. A, Angiography showing complete occlusion of the left common iliac vein with large transversal and ascendant collateral pathways. B, After recanalization and stenting with an 80-mm-long Wallstent.
Three patients (7.3%) had angioplasty only because left CIV stenoses were found to be moderate at iliocavography during the procedure (50% stenosis), whereas the preoperative duplex scan and computed tomographic scan revealed >50% left CIV compression.
Fifty-eight stents (12-16 mm in diameter and 40-90 mm long) were used: 2 in the IVC, 45 in the CIV, 6 in the EIV, and 3 in the CFV. In one case, a kissing stent was performed in the iliocaval junction, and two CFVs were dilated but not stented. Moreover, six left CIVs were treated with two stents. The mean number of stents for treating the lesions per patient was 1.37.1, 2, 3, 4 The mean length of stented vein was 73 mm (range, 40-220 mm). In two patients, a stent migrated during the procedure. One stent jumped into the retrohepatic IVC: it was pulled down into the infrarenal IVC, where it adopted a transversal position; another stent was deployed in the iliocaval junction. The patient was asymptomatic 33 months later, and the stent did not migrate. Another stent was pulled back from the right atrium down to the left CFV with two Amplatz Goose Neck Snare kits (ev3, Inc, Plymouth, Minn), where it was retrieved surgically. Another stent was accurately deployed 3 months later.
Concomitant procedures were performed in five patients: two left greater saphenous vein high resections (one iterative), one left greater saphenous vein stripping, one left lesser saphenous vein high resection, and one skin grafting (the patient had an active ulcer). One patient had thrombosis 6 days after EIV recanalization; surgical thrombectomy with balloon angioplasty of the stent and creation of an arteriovenous fistula were performed with good results. He was discharged at day 14. Another patient had postoperative bleeding through the puncture hole which was stopped after redoing the compressive bandage.
All the other patients had a postoperative period without complication. One patient stayed 13 days because he had secondary left calf endoscopic perforant vein ligature and left saphenous vein stripping at day 11, and one patient stayed 10 days for skin grafting. The median hospital stay was 3 days (range, 1-14 days). Thirty-two patients were discharged with fluindione (8 patients since 2003 due to history of DVT, which was associated with pulmonary embolism in 2 and hypercoagulability in 3), and 12 were discharged with nadroparin and antiplatelet drugs.
The median follow-up was 27 months (range, 2-103 months). One patient died from carcinomatosis 16 months later with a patent stent.
In case of skin damage, improvement occurred in all patients but one, who underwent angioplasty alone. The ulcerated limb healed and stayed healed until the last follow-up at 38 months. Duplex scan showed that reflux was not modified in four and was improved in two.
One late thrombosis (2.4%) occurred at 15 months, and a Palma procedure was performed, with good results. Restenosis occurred in five patients (13%) after a median of a 9 months’ (range, 4-27 months) symptom-free period and was detected by means of the duplex scan. One patient who initially had CFV and EIV recanalization for postradiotherapy thrombosis had angioplasty only at 13 months to treat a symptomatic 80% in-stent restenosis, with a good initial result, but iterative restenosis nevertheless occurred at 16 months, and a Palma procedure was performed. The four other patients had restenosis (symptomatic in three) of the iliac vein at the caudal side of stents implanted in the CIV; they were treated by iterative percutaneous balloon angioplasty and stenting (Fig 3).
Fig 3.
Restenosis at 8 months. A, Color Doppler scan showing an angulation at the caudal side of the stent. B and C, Face and profile angiography showing restenosis due to an angulation of the vein at the caudal side of the stent with collateral pathways. D and E, Face and profile angiography after iterative stenting: absence of residual stenosis and regression of the collateral pathways.
Of the three patients who had isolated angioplasty, one improved slightly, and one had no improvement. For the third, the symptoms initially disappeared but recurred at 16 months, and a duplex scan showed left CIV restenosis. The median VCSS and VDS were, respectively, 8.5 (range, 4-18) and 2 (range, 2-3) before surgery and 2 (range 0-9) and 0 (range 0-2) at the end of the follow-up.
Table II and Fig 4 show the patency rates. Primary patency, assisted primary patency, and secondary patency rates were 73.2%, 87.6%, and 89.9%, respectively, at 36 and 60 months. The survival rate was 97.3% at 60 months.
Table II. Results of the published series of endovascular treatment for chronic iliocaval occlusive disease of varying etiology
| Variable | n | Technical success | PP at 12 mo | Late PP | Late SP |
|---|---|---|---|---|---|
| O’Sullivan9 | 20 | 93.9% | |||
| Nazarian10 | 56 | 92% | 50% | 50% at 48 mo | 75% at 48 mo |
| Blättler11 | 14 | 85.7% | 79% | ||
| 45 | |||||
| Neglen12 | 5 | 75% at 36 mo | 93% at 36 mo | ||
| Hurst13 | 18 | 100% | 79% | ||
| Hartung | 44 | 95.5% | 83.6% | 73.2% at 36 mo | 89.9% at 36 mo |
Fig 4.
Cumulative primary, assisted primary, and secondary patency rates of 44 patients admitted for endovascular treatment of chronic iliocaval occlusive disease (intention to treat). Standard error was more than 10% at 30 months for primary patency, at 48 months for assisted primary patency, and at 57 months for secondary patency (the curves are represented as dotted lines when SEM is >10%).
Discussion
Chronic venous occlusive disease, though generally not life threatening, is a severe and disabling pathology that occurs mainly in young patients and restricts their ability to work. Medical treatment, mainly represented by compressive therapy, is applied, but improvement is not always enough to make a normal lifestyle possible. Several teams have reported on studies of angioplasty with or without stenting,9, 10, 11, 12, 13 with good results (Table II). Thus, stenting is the method of choice to treat iliocaval obstructive disease.
In our experience, treatment of isolated May-Thurner syndrome was performed with use of local anesthesia in 20 patients. Although this technique has several advantages for the patient, it is not comfortable as a result of very intense acute and durable back pain from stent implantation. Moreover, treatment of complications (ie, stent migration) with the patient under local anesthesia can be very difficult, whereas general anesthesia offers the possibility of using other approaches. That is why we currently perform such procedures with patients under general anesthesia.
Angioplasty only was used three times on May-Thurner syndrome cases because the stenoses did not look significant. Moreover, one patient was only 22 years old and had only bilateral edema. This last patient was the only one of the three whose condition really improved, but the symptoms reappeared 16 months later with restenosis at the duplex scan. On the basis of this, it can be said that stenting should always be performed.
The choices of stent configuration, length, diameter, and position, as well as postoperative treatment, are without doubt important, particularly in the treatment of left CIV lesions. Self-expanding metallic stents have already been used on venous lesions, with good results.14 We used a balloon-expandable stent only in two cases reported previously.5 In one case, the stent was implanted in a pregnant woman after left femoroiliac venous thrombectomy. Despite excellent immediate results, iterative thrombosis occurred 2 months later, and the stent was completely crushed.
Initially, we used 14-mm-diameter and 30-mm-long stents. Experience has shown that this choice was not the right one. First, two stents jumped because of vein elastic recoil. Moreover, proximal or distal restenoses occurred despite initially excellent-looking results and were believed to be due to underestimated lesion length. This is certainly due to the poor accuracy of angiography, even when face and profile were used to evaluate the extent of the lesions. Intravascular ultrasonography is the best way to estimate them,15 but this device is not available in our institution. We were also reluctant to put the distal end of the stent across the iliocaval confluence. Neglen and Raju16 showed that such a placement was needed and did not induce right iliac vein thrombosis. In our experience, this condition occurred in one case after stenting in a patient with autoimmune thrombopenia despite efficient fluindion treatment. For left CIV lesions, we commonly use stents that are 16 mm in diameter and at least 60 mm long, with the distal end largely protruding into the IVC.
At the beginning of our experience, all patients who had iliocaval stenting were discharged on fluindione for 6 months. Today, only patients treated for acute DVT or post-DVT obstructive lesions and those who have had recanalization receive such treatment. Patients with coagulation anomalies or other reasons for fluindion will also have this treatment. All others are treated with nadroparin for 15 days and aspirin or clopidogrel for 6 to 12 months. Despite Raju and colleagues’ good results with aspirin only,14 we are reluctant to adopt this because of the lack of references on the role of antiplatelet drugs in venous disease.
The improvement of VCSS and VDS confirmed that stenting is an effective way to treat occlusive lesions, as Neglen et al17 stated. Initial clinical presentation found severe venous disease (C4 to C6) only in patients who had the association of obstruction and reflux. All but one had a history of DVT. All these patients had clinical improvement after stenting, and reflux was improved in two cases only but never worsened.
Taking all patients with stents into consideration, thrombosis occurred in two patients (one acute and one late; 5.1%), and restenosis occurred in five patients (12.8%). These results are in line with those of Neglen and Raju,16 who reported a thrombosis rate of 4% and a >50% restenosis rate of 17% at 3 years. In one case, angioplasty only was performed because of in-stent restenosis on a stent implanted for left CFV and EIV recanalization. When restenosis occurred on iliac veins at the end of the stent(s) (four cases), it was always treated successfully by iterative angioplasty and stenting.
Conclusion
Angioplasty and stenting is a safe and efficient minimally invasive technique for treatment of iliocaval occlusive disease. Mid-term patency rates are good, and this technique does not preclude subsequent open operation in case of failure.
References
- . Iliac vein reconstructions to treat acute and chronic venous occlusive disease . J Vasc Surg . 1997;25:673–681
- Surgical reconstruction of iliofemoral veins and the inferior vena cava for nonmalignant occlusive disease . J Vasc Surg . 2001;33:320–327
- . Femoro-femoral crossover grafts (Palma operation) (a long-term follow-up study) . In: Bergan KK , Yao JST editor. Surgery of the veins . Orlando: Grune & Stratton; 1985;p. 241–254
- . New trends in the surgical and endovascular reconstructions of large veins for nonmalignant chronic venous occlusive disease . Curr Opin Cardiol . 1998;13:375–383
- . Treatment of nonmalignant obstructive iliocaval lesions by stent placement (mid-term results) . Ann Vasc Surg . 2001;15:227–232
- . Endovascular treatment of iliocaval occlusion caused by retroperitoneal fibrosis (late results in two cases) . J Vasc Surg . 2002;36:849–852
- . Reporting standards in venous disease (an update) . J Vasc Surg . 1995;21:635–645
- . Venous severity scoring . J Vasc Surg . 2000;31:1307–1312
- Endovascular management of iliac vein compression (May-Thurner) syndrome . J Vasc Interv Radiol . 2000;11:823–836
- . Iliofemoral venous stenoses (effectiveness of treatment with metallic endovascular stents) . Radiology . 1996;200:193–199
- . Relief of obstructive pelvic venous symptoms with endoluminal stenting . J Vasc Surg . 1999;29:484–488
- . In-stent recurrent stenosis in stents placed in the lower extremity venous outflow tract . J Vasc Surg . 2004;39:181–188
- . Diagnosis and endovascular treatment of iliocaval compression syndrome . J Vasc Surg . 2001;34:106–113
- . The clinical impact of iliac venous stents in the management of chronic venous insufficiency . J Vasc Surg . 2002;35:8–15
- . Intravascular ultrasound scan evaluation of the obstructed vein . J Vasc Surg . 2002;35:694–700
- . Balloon dilation and stenting of chronic iliac vein obstruction (technical aspects and early clinical outcome) . J Endovasc Ther . 2000;7:79–91
- . Venous outflow obstruction (an underestimated contributor to chronic venous disease) . J Vasc Surg . 2003;38:879–885
Competition of interest: none.
PII: S0741-5214(05)01271-1
doi:10.1016/j.jvs.2005.08.012
© 2005 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
Refers to erratum:
- Correction
