Outcome of Stenting in the Lower-Extremity Venous Circulation for the Treatment of Deep Venous Thrombosis

Abstract

Background. Venous stents are being increasingly used as an effective tool during percutaneous endovenous intervention (PEVI) in the treatment of deep venous thrombosis (DVT). There is a paucity of data on the rate of in-stent restenosis (ISR), stent thrombosis and other outcomes of stents placed in the venous system. The purpose of this study was to describe our experience with stents placed in the iliac and femoropopliteal veins for high-grade venous stenosis during PEVI for DVT. Methods and Results. Over a period of 28 months, 287 stents were placed in the iliac and femoropopliteal veins of 133 consecutive patients who had presented with acute severe DVT and venous stenosis. Patients with presumed DVT on venous duplex underwent repeat venography and intravascular ultrasound (IVUS) evaluation of their stents. There were 14 patients with presumed DVT who underwent repeat venography and IVUS evaluation of their stents at a mean follow-up of 27 ± 4 months. Stent thrombosis was found in 4%, but there was no ISR due to neointimal proliferation. There were no stent fractures, extrusions or perforations. Conclusions. We conclude that the natural history of stenting in the venous circulation is fundamentally different from that seen in the arterial system. Venous stenting is associated with a possibly nonexistent clinically significant ISR rate and a low stent thrombosis rate. In symptomatic patients who develop stent thrombosis, the symptoms are usually mild and appear early after stent placement. Stent thrombosis is not associated with significant sequelae and is amenable to re-do PEVI.

VASCULAR DISEASE MANAGEMENT 2010;7(12):E233–E239

Key words: deep venous thrombosis, PEVI, percutaneous endovenous intervention, thrombus, venous stents, venography, venoplasty

Introduction

Venous stents are being increasingly used as an effective tool during percutaneous endovenous intervention (PEVI). The major indication for PEVI has been venous stenosis and external compression, which often leads to deep venous thrombosis (DVT).^1–4 Stents placed in the peripheral arterial circulation are plagued by in-stent restenosis (ISR) and less commonly by acute thrombosis.^5–9 The former is chiefly due to neointimal proliferation, which accounts for high rates of stent closures, especially in smaller arteries.^7,9 Stent thrombosis, however, is less frequent and associated with an acute and severe presentation which may lead to high rates of mortality and morbidity. There is a paucity of data on the rate of ISR, stent thrombosis and other outcomes of stents placed in the venous system. The veins are thinner and there is potential concern about adverse outcomes including perforation, extrusion and fracture. This study describes our experience with stents placed in the iliac and femoropopliteal veins for high-grade venous stenosis during PEVI for DVT.

Materials and Methods

From February 2007 to June 2009, 287 stents were placed in the iliac and femoropopliteal veins of 133 consecutive patients who had presented with acute severe DVT and were found to have venous stenosis on ascending venography. Informed written consent was obtained from all patients and the study was approved by the institutional review boards of the participating centers. The indication for stenting was persistence of > 80% diameter stenosis after thrombolysis and balloon venoplasty or to reconstruct a conduit in an occluded segment with distorted anatomy. The patients’ clinical characteristics are shown in Table 1. The affected veins were divided into five segments based on the anatomical involvement of DVT: inferior vena cava, right and left iliac veins and right and left femoropopliteal veins. These patients were followed up prospectively for a mean of 27 ± 4 months (range, 12–39 months). There were 10 patients (with 18 stents) who were lost to follow-up. The technique for PEVI has been previously described.^10,11 In brief, a micropuncture needle was used under ultrasound guidance and access to the popliteal vein obtained. A 6–8 French sheath was subsequently placed through which venography and intervention were performed. Stent placement was performed at a mean 26 ± 8 hours after presentation. All patients had been started on parenteral anticoagulation with enoxaparin or heparin plus warfarin without interruption of anticoagulation during the procedure. Eighty-four patients had received thrombolytic therapy as part of the thrombectomy regimen or regional thrombolysis. All patients received warfarin post PEVI for a minimum of 3 months and aspirin 81–325 mg daily for 6 months. All had received a prophylactic inferior vena cava filter at the index PEVI with 18 of them (14%) having them subsequently removed. Patients receiving a stent in the femoropopliteal venous segments received clopidogrel 75 mg daily for 2–4 weeks. Patients underwent venous duplex sonography (VDS) at 1, 6 and 12 months post procedure, and every 6 months thereafter, and at anytime DVT was suspected. If there was recurrence or presumed recurrence of DVT on VDS, venography was performed. In such cases, intravascular ultrasound (IVUS) was performed with interrogation of the stent site. DVT was said to have recurred if new non-compression of a previously compressible venous segment had developed on ultrasound evaluation. For previously abnormal segments, a > 5 mm increase in the diameter of the thrombus during full compression was required. For stented segments that were not amenable to compression, absence of flow on color-Doppler imaging and absence of venous spectral waveform were required. Post-thrombotic syndrome (PTS) was defined according to our previously published classification.¹² The patients were fully educated on the signs and symptoms of DVT and PTS. They were instructed to contact the study center immediately if any of these conditions occurred. Patients were evaluated for the signs and symptoms of recurrent DVT and PTS on each follow-up for VDS or sooner if indicated.

Definitions for stent patency status. To avoid confusion in nomenclature, we used six pre-defined terms based on IVUS and venography findings to describe the patency status of the deployed stents, and the mechanism and severity of the pathologic process. These terms consisted of: 1) stent stenosis; 2) stent occlusion; 3) stent thrombosis; 4) external compression on the stent; 5) ISR; and 6) presence of minimal thrombus within the stent. Stent stenosis was defined as a loss of ≥ 50% but 20% of stent diameter as assessed by IVUS. For example, if the inner diameter of a 10 mm stent was reduced to ≤ 8 mm due to neointimal proliferation, ISR was said to have occurred. Stent thrombosis was considered to have developed if there was thrombus within the stent involving ≥ 50% of the cross-sectional area (more than two quadrants) on IVUS imaging and extending for a minimum of 1 cm in length. This could have potentially led to stenosis or occlusion depending on flow status as described above (Figure 3). Therefore, stent thrombosis merely defined the presence of thrombus within the stent. If the volume of thrombus was such that flow was entirely stopped, occlusion was said to have occurred. Minimal thrombus was said to be present if there was normal flow on venography, but on IVUS imaging thrombus was seen in less than two quadrants (IVUS analysis. The IVUS catheter was a Visions PV 0.018 F/X on an S5 imaging platform with virtual histology (VH) and Chromaflo software package (Volcano Corp., Rancho Cordova, California). This system would allow for accurate edge detection and flow determination, thereby separating the stent from the endoluminal neointimal proliferation boundary and from any thrombotic mass if present (Figure 4). Electronic calipers were used for measurement of desired distances. The accuracy of the VH program for border detection, tissue composition and differentiation between thrombus and harder tissues has been previously described.^13,14 Therefore, with the use of the S5 imaging system with VH capability, the dimension of endoluminal tissue in-growth from the stent struts (neointimal proliferative dimension) could be accurately measured (Figure 4B). If this distance reached the predefined threshold of 20% of the stent diameter, ISR was said to have been present.

Statistical analysis. All calculations were performed using SPSS version 18 (SPSS, Inc., Chicago, Illinois). Continuous values were expressed as mean ± standard deviation.

Results

There were 164 stents (57%) in the iliac veins, 37 (13%) in popliteal and 86 (30%) in femoral veins. The type of stent used is shown in Table 2. The mean stent diameter was 9.3 ± 3.2 mm and length 192 ± 44 mm. All stents had been postdilated with a balloon 1–2 mm above nominal stent size at or near burst pressure. Fourteen patients underwent repeat venography 9 days to 22 months after PEVI based on a VDS that was diagnostic or suggestive of DVT. There were 3 patients with symptomatic DVT, 5 with asymptomatic DVT and 6 with VDS suggestive of DVTs that were found to have minimal or no thrombus on IVUS. Surprisingly, with the predefined definition, no ISR was noted in any patient. Stent thrombosis, however, was noted in all 8 DVT cases: 6 with stent occlusion and 2 with stent stenosis. The stents involved consisted of 1 Viabahn (W.L. Gore & Associates, Flagstaff, Arizona), 4 Absolute (Abbott Vascular, Santa Clara, California) and 3 Protégé (Covidien, Mansfield, Massachusetts) stents. External compression was noted in 2 patients with stent occlusion. These were in the mid and distal femoral veins. Stent thrombosis was invariably due to extension of thrombus from other adjacent sites into the stent. Specifically, it did not occur in the stent area alone. In all 8 patients, DVT had occurred on therapeutic anticoagulation. Three of the patients were on aspirin and 1 on additional clopidogrel. In all patients with DVT, stent thrombosis was due to a new or untreated high-grade stenosis outside the previously deployed stents. Despite stent thrombosis, no new fixed stenotic lesion had developed within any stents. Six of the 8 patients with DVT had retained an inferior vena cava filter from the index PEVI and the other 2 had not. Minimal thrombus was found in 4 other patients with suspected DVT on VDS. There were no stent fractures, extrusions or perforations at a mean follow-up of 27 ± 4 months. All DVT cases underwent successful repeat PEVI. Angiojet thrombectomy (Medrad/Possis, Inc.,Warrendale, Pennsylvania) was used in 4 patients, the Trellis device (Covidien, Mansfield, Massachusetts) in 2, and local thrombolytic therapy with tPA for 20–24 hours in 4 patients (2 had a combination of treatments). Streamline flow was re-established in all, with major thrombus dissolution. Anticoagulation was not withheld during the procedure. The mean neointimal proliferative dimension for all non-covered stents was 0.25 ± 0.10 mm. This figure did not reach the threshold of 20% reduction in stent diameter (due to tissue in-growth) to be regarded as ISR. No significant difference for the neointimal proliferative dimension was noted between the non-covered stents, although this value was zero for Viabahn stents. External compression was noted in 4 patients: 2 with occlusion and 2 with normal flow. Clinically, development of PTS was noted in 6/123 patients at a mean period of 392 ± 55 days after index PEVI. Four of the PTS cases were mild and 2 were moderate according to our reported definition.12 One of the PTS cases was from the 8 patients with DVT.

Discussion

In this study, stent occlusion and stenosis occurred in 8/123 patients at a mean period of 423 ± 55 days post index PEVI. They were predominantly due to stent thrombosis with acute thrombus originating from sites outside the stent and extending into it. There was no ISR in any patient. This is an intriguing finding and suggests that neointimal proliferation is not as active in veins as it is in arteries. Despite our low threshold in defining ISR (> 20% diameter stenosis due to neointimal proliferation), not a single incidence of ISR was observed. For ISR in the arterial system, at least a 50% diameter reduction is required. In the superficial femoral artery, the ISR rate may reach as high as 37% at 1 year.⁵ This figure well exceeds 50% in the tibial artery during the same period.7 We had previously demonstrated a high patency rate for stents placed in the venous circulation and suggested that low venous pressure flow with less turbulence and trauma to the vessel wall may be contributing factors.^1,11 On a microscopic level, the venous media and adventitia are thinner with less myocytes, fibroblasts and other cells which usually contribute to the hyperproliferative process. In describing venous stent closure, some authors have not distinguished between ISR due to neointimal proliferation versus thrombosis, and hence all stent closures were regarded as “ISR”.^3,4 Even in patients who underwent IVUS imaging, this distinction was not addressed.³ Neglen et al reported the development of “ISR” in 80% of stents placed for chronic, non-malignant obstruction at 42 months.³ Their definition of ISR was different from ours and based on single-plane venography. They hypothesized that “reaction to stenting of the vein is possibly a cellular response rather than formation of thrombosis”.³ This assumption was based on the fact that thrombolytic therapy was not effective in the subgroup of patients receiving this form of treatment. Our results do not support their assumption. ISR, according to our definition, was not present in any of the 14 patients studied with IVUS. In other words, no significant in-growth of tissue within the stent struts was noted on IVUS. However, we do believe that external compression is an important factor in impeding flow and ultimately contributing to stent thrombosis. In non-tumor situations, external compression may be due to an ongoing fibrotic force constantly opposing stent expansion and leading to recoil. Based on IVUS and venographic images, two major and distinct patterns of DVT could be identified. One category encompasses patients who have a distorted venous anatomy with obliteration of the normal architecture, which we called “venosclerosis” (Figure 5^11,12). In these patients, stent expansion to its nominal diameter could not be achieved even with high-pressure balloons at the index intervention. In other words, a strong fibrotic force exists possibly due to tissues that have replaced the normal veins and their surrounding structures, which would oppose full stent expansion. This process usually affects diffuse venous segments, well beyond the stent length. The extent of fresh thrombus is minimal, and consequently, no appreciable response to thrombolytic therapy occurs. This process was probably the pattern that was reported by Neglen as being unresponsive to thrombolytic therapy.³ However, in such cases when the patient is symptomatic, even the partially expanded stents result in improvement in flow and patients’ symptoms. The other category consists of patients, usually those with acute first-time DVT, in whom the venous anatomy and architecture are preserved (Figure 6). Thrombectomy is highly effective in these situations, and if there are isolated high-grade fixed lesions, they could easily be treated with full expansion of the stent. There is also a large in-between spectrum with features of both categories. When veins exposed to systemic arterial pressure are stented, such as those seen in saphenous vein grafts used in coronary or peripheral artery bypass surgery, neointimal proliferation is the dominant cause of ISR.^15,16 The pattern of ISR is similar to that seen in arteries. Rates as high as 37% of stents placed in coronary saphenous vein grafts can develop ISR at 1 year.¹⁵ For saphenous vein grafts used in the peripheral arteries that undergo stenting, closure rates of 63% have been reported at 1 year, mainly due to ISR.¹⁶ This lends further support to the notion that pulsatile arterial pressure may be a requirement for development of neointimal proliferation. In our study, neointimal proliferation per se was minimal and certainly not a cause for stent closure. DVT occurred in 8/123 patients, with the majority (5/8) being asymptomatic. The symptoms in the 3 symptomatic patients were mild, developing within the first 35 days following the index PEVI. Stent thrombosis was related to incomplete treatment of in-flow or out-flow lesions and not neointimal proliferation. It was uniformly seen as an extension of DVT from other sites into the stent. There were no stent fractures or wall perforations despite the relatively thin venous wall. This finding is consistent with other reports and is probably due to higher venous compliance.^2–4 One can therefore assume that it is not the “padding” effect of the vessel that protects against stent complications, but vessel compliance. There were no differences in outcome between balloon or self-expandable stents. For stenting of the popliteal vein, we used Viabahn and Viatorr stents (W.L. Gore & Associates, Flagstaff, Arizona) with the theoretic assumption that these stents may better endure stress forces at the knee joint. Indeed, they were found to be exceptionally resistant to complications. Venographic images post stenting were “smoother” with Viabahn stents than with others (Figure 7). Tissue in-growth as assessed by virtual histology was nonexistent in these stents. Stent thrombosis was independent of stent type, diameter and length. IVUS imaging demonstrated minimal thrombus in 4 asymptomatic patients not diagnosed on venography. On VDS, this finding was considered an “old thrombus” without any obstruction to flow. DVT was the mode of presentation for stent closure. The cause was believed to be development of new or untreated lesions obstructing flow at other sites. The symptomatic DVT cases occurred early and were therefore suggestive of being secondary to untreated segments rather than new stenotic lesions. All DVT cases responded well to further modalities of PEVI. There was no symptomatic pulmonary embolism, ulcer formation, stent infection, limb loss or mortality with DVT recurrence. One patient had developed cellulitis of the popliteal fossa requiring re-admission and treatment with antibiotic therapy 5 days after the index PEVI. She recovered without deeper extension of the infection and with no sequelae. The development of PTS was 4.9% at a mean period of 392 ± 55 days post index PEVI. Most cases of PTS were mild. We have recently shown that PTS can be significantly reduced with early PEVI.^12,17

Conclusion

We conclude that the natural history of stenting in the venous circulation is fundamentally different than that seen in the arterial system. Venous stenting is associated with a very low and possibly nonexistent clinically significant ISR rate. Stent thrombosis, however, occurs in 4% of cases and is not due to ISR. It does not occur independently and is usually an extension of DVT in the adjacent venous segments with high-grade stenosis (in-flow or out-flow obstruction). External compression is an important factor in preventing full stent expansion, which is seen in extensive venosclerosis and may contribute to stent closure, usually in the presence of some degree of thrombosis, but not ISR. In this scenario, the role of local thrombolytic therapy is limited. Contrary to its arterial counterpart, stent thrombosis is often asymptomatic, and even in symptomatic cases, the symptoms are usually mild and appear early after stent placement. It is not associated with significant sequelae and is amenable to re-do PEVI.

References

1. Sharifi M, Mehdipour M, Berkovits A, et al. Low restenosis rate of stents placed in peripheral venous circulation. Catheter Cardiovasc Interv 2010;75:S83–S84.

2. Neglén P, Darcey R, Olivier J, Raju S. Bilateral stenting at the iliocaval confluence. J Vasc Surg 2010;51:1457–1466.

3. Neglén P, Hollis KC, Olivier J, Raju S. Stenting of the venous outflow in chronic venous disease: Long-term stent-related outcome, clinical, and hemodynamic result. J Vasc Surg 2007;46:979–990.

4. Neglén P, Raju S. In-stent recurrent stenosis in stents placed in the lower extremity venous outflow tract. J Vasc Surg 2004;39:181–187.

5. Schillinger M, Sabeti S, Loewe C, et al. Balloon angioplasty versus implantation of nitinol stents in the superficial femoral artery. N Engl J Med 2006;354:1879–1888.

6. Duda SH, Bosiers M, Lammer J, et al. Sirolimus-eluting versus bare metal stent for obstructive superficial femoral artery disease: The Sirocco II trial. J Vasc Interv Radiol 2005;16:331–338.

7. Rand T, Basile A, Cejna M, et al. PTA versus carbofilm coated stents in infrapopliteal arteries: Pilot study. Cardiovasc Intervent Radiol 2005;29:29–38.

8. Bosiers M, Kallakuri S, Deloose K, et al. Infragenicular angioplasty and stenting in the management of critical limb ischaemia: One year outcome following the use of the Multi-Link Vision stent. EuroInterv 2008;3:470–474.

9. Schlager O, Dick P, Sabeti S, et al. Long-segment SFA stenting — The dark sides: In-stent restenosis, clinical deterioration, and stent fractures. J Endovasc Ther 2005;12:676–684.

10. Sharifi M, Mehdipour M, Skloven D, et al. Power pulse spray and Angiojet thrombectomy in massive inferior vena cava and bilateral lower extremity deep venous thrombosis. Vascular Disease Management 2008;5:62–65.

11. Sharifi M, Mehdipour M. Percutaneous therapy of acute on chronic lower extremity venous occlusive disease. Catheter Cardiovasc Interv 2010;75:685–689.

12. Sharifi M, Mehdipour M, Bay C, et al. Endovenous therapy for deep venous thrombosis: The TORPEDO trial. Catheter Cardiovasc Interv 2010;76:316–325.

13. Nair A, Margolis MP, Kuban B, Vince DG. Automated coronary plaque characterization with intravascular ultrasound backscatter; Ex vivo validation. Eurointerv 2007;3:113–120.

14. Murphy EH, Broker HS, Johnson EJ, et al. Device and imaging-specific volumetric analysis of clot lysis after percutaneous mechanical thrombectomy for iliofemoral DVT. J Endovasc Ther 2010;17:423–433.

15. Savage MP, Douglas JS Jr, Fischman DL, et al. Stent placement compared with balloon angioplasty for obstructed coronary bypass grafts. N Engl J Med 1997;337:740–747.

16. Siskin GP, Stainken BF, Mandell VS, et al. Management of failing prosthetic bypass grafts with metallic stent placement Cardiovasc Intervent Radiol 1999;22:375–380.

17. Sharifi M, Mehdipour M, Smith G, et al. Reduction of post-thrombotic syndrome in acute deep venous thrombosis by power pulse spray and Angiojet thrombectomy. Catheter Cardiovasc Interv (abstr) 2009;75:S43.

___________________________________________________________________________

From Arizona Cardiovascular Consultants and A.T. Still University, Mesa, Arizona. The authors report no conflicts of interest regarding the content herein. Manuscript submitted August 19, 2010, provisional acceptance given September 18, 2010, final version accepted September 27, 2010. Address for correspondence: Mohsen Sharifi, MD, Arizona Cardiovascular Consultants, 5440 E. Southern Ave., Suite 104, Mesa, AZ 85206. E-mail: seyedmohsensharifi@yahoo.com