Surgical or endovascular repair of thrombosed dialysis vascular access: Is there any evidence?
Article Outline
Introduction
Endovascular and surgical strategies have been used to manage patients with thrombosed vascular access for hemodialysis. We analyzed the evidence to see whether endovascular or surgical treatment has the best outcome in terms of primary success rate and long-term patency.
Methods
We performed a systematic literature search of endovascular and surgical repair of thrombosed hemodialysis vascular access. The analysis included meta-analysis, randomized, and population-based studies of thrombosed arteriovenous fistulae and grafts.
Results
One meta-analysis and eight randomized studies on the treatment of arteriovenous graft thrombosis were identified. Studies conducted before 2002 demonstrated a significantly better primary success rate and primary and secondary patencies of surgical thrombectomy vs endovascular intervention. After 2002, similar results of both techniques have been reported. Only population-based studies on the treatment of thrombosed autogenous arteriovenous fistulae have been published, showing similar outcome of surgical and endovascular intervention in terms of primary success. The long-term primary and secondary patencies are slightly better for surgical treatment, but this concerns only forearm fistulae.
Conclusions
The outcome of endovascular and surgical intervention for thrombosed vascular access is comparable, in particular for thrombosed prosthetic grafts. Surgical treatment of autogenous arteriovenous fistulae is likely to have benefit compared with endovascular means. Definitive randomized trials are needed to provide the level 1 evidence to resolve this latter issue.
Functional vascular access is a prerequisite for adequate hemodialysis treatment in patients with end-stage renal disease (ESRD). The major drawback of vascular access is a high incidence of thrombotic occlusion, caused by obstructive stenoses that are initiated by hyperplastic lesions of the vessel intimal layer. Surgical or endovascular thrombus removal and repair of stenotic lesions remains the major goal toward access salvage and prolongation of patency rate. The strategy to choose for endovascular or surgical intervention remains obscure, however, and evidence from the literature may possibly elucidate the potential benefits and outcomes of both techniques.
Methods
A systematic literature review from 1995 to 2009 using a Medline search of English language publications was performed. Keywords used were vascular access, hemodialysis, thrombosis, arteriovenous fistula, and arteriovenous graft. For this review in particular, large population-based, randomized, and meta-analysis studies were included. Small patient studies and case reports were excluded. Because of potential differences in treatment methods and outcomes in arteriovenous (AV) grafts and fistulae (AVF), these types of access were analyzed separately.
Results
We identified 78 studies. There were 42 studies on endovascular (n = 37) and surgical (n = 5) intervention for graft thrombosis, of which eight randomized studies compared endovascular therapy vs surgery. One meta-analysis on the treatment of AV graft thrombosis was identified. There were 36 studies on autogenous fistulae thrombosis treatment, either by surgery (n = 6) or percutaneous techniques (n = 30). None of the studies on fistula thrombectomy were randomized. The studies on fistulae appeared to be published more recently than the graft studies, indicating the growing interest and strategy to create predominantly autogenous access in ESRD patients in the United States and Europe.
The meta-analysis by Green et al1 published in 2002, included seven randomized studies (1987-1999) comparing surgical vs endovascular graft thrombectomy in terms of primary patency at 30, 60, 90 days, and 1 year. Calculation of the relative risk and risk difference showed at all time periods for the aggregated data a superiority of surgical thrombectomy vs endovascular intervention. These differences can be mainly ascribed to the relatively low initial technical success rate of endovascular thrombectomy of 76% vs 85% for surgical intervention. In the randomized studies included in this meta-analysis, different techniques for endovascular thrombectomy were used, comprising pulse-spray infusion of urokinase,2 mechanical thrombectomy,3, 4 or a combination of both.5, 6, 7
More recently published studies, including a randomized study, showed markedly improved technical success rates (mean, 92%; range 79%-100%) of endovascular intervention for graft thrombosis. The randomized study by Uflacker et al8 reported initial thrombectomy success rates of 79% for endovascular and 77% for surgical treatment. Graft patency at 30 and 90 days was 79% and 75% for the endovascular and 73% and 68% for the surgically treated patients (P = NS). The endovascular long-term patencies seem to have improved compared with older studies. The administration of tissue plasminogen activator (tPA),9 a more potent thrombolytic agent, and the use of newer thrombectomy devices10, 11, 12 may have contributed to this better outcome of endovascular treatment (Table I).
Table I. Endovascular treatment of thrombosed grafts
| First author, year | No. | Modality | Technical success | Patency (1 yr) | |
|---|---|---|---|---|---|
| Primary | Secondary | ||||
| Falk, 2001 | 62 | Pulse-spray thrombolysis tPA | 92% | 34% (6 mon) | … |
| Sofocleous, 2002 | 68 | Pulse-spray thrombolysis tPA | 94% | 44%(6 mon) | 72%(6 mon) |
| Cooper, 2003 | 17 | Pulse-spray thrombolysis urokinase | 94% | 21%(4 mon) | … |
| Kolakovski, 2003 | 61 | Percutaneous thrombectomy + stent | … | 8% | 25% |
| Uflacker, 2004 | 109 | Mechanical thrombectomy | 79% | 75%(3 mon) | … |
| Bittl, 2005 | 105 | Mechanical thrombectomy | 95% | … | 23% |
| Bakken, 2007 | 114 | Pharmacomechanical thrombolysis | 95% | … | … |
| Kakkos, 2008 | 207 | Mechanical thrombectomy | 93% | 29% | 62% |
| Lai, 2009 | 32 | Mechanical thrombectomy | 94% | 70%(3 mon) | … |
The outcome of surgical treatment of thrombosed grafts depends on whether additional graft revision with patchplasty or graft interposition has been done. Without additional graft revision, a dismal 1-year primary patency of 10% vs 27% for thrombectomy and additional revision has been reported.13
Intervention studies on thrombosed autogenous AVF have predominantly appeared after the year 2000. A few have compared endovascular with surgical repair, but none was randomized. The largest series is from Turmel-Rodrigues et al14 and reports on 93 percutaneous declotting procedures in 73 patients with 56 forearm radiocephalic or ulnobasilic AVF and 17 upper arm brachiocephalic or basilic AVF. Manual thrombus aspiration with or without urokinase was applied, and stents were only used on indication for resistant stenoses or venous rupture. Technical success was higher in forearm than upper arm AVF (93% vs 76% in upper arm AVF), and 1-year primary and secondary patencies were 49% and 81%, respectively (for upper arm AVF, 9% and 60%).
Comparative results can be found in the literature with usually a high technical success rate of 76% to 96%, although low primary patencies of 18% to 70% are reported with the need of multiple reinterventions to achieve acceptable cumulative patencies15, 16, 17, 18 (Table II). Upper arm AVF do worse compared with forearm AVFs. Venous rupture and arterial emboli are major complications and can usually be treated by endovascular means with good results.
Table II. Endovascular treatment of thrombosed fistulae
| First author, year | No. | Location AVF | Modality | Technical success | Patency (1 yr) | |
|---|---|---|---|---|---|---|
| Primary | Secondary | |||||
| Overbosch, 1996 | 24 | 24 forearm | Mechanical thrombectomy | 89% | 32% | … |
| Turmel-Rodrigues, 2000 | 73 | 56 forearm | Thromboaspiration ± urokinase | 93% | 49% | 81% |
| 17 upper arm | 9% | 50% | ||||
| Haage, 2000 | 54 | 50 forearm | Mechanical thrombectomy | 89% | 27% | 51% |
| 4 upper arm | ||||||
| Schon, 2000 | 20 | … | Mechanical thrombectomy + tPA | 92% | … | … |
| Liang, 2002 | 42 | 37 forearm | Angioplasty ± urokinase | 93% | 70% | 80% |
| 5 upper arm | ||||||
| Rajan, 2002 | 25 | 19 forearm | Mechanical thrombectomy + urokinase, tPA | 73% | 24% | 44% |
| 6 upper arm | ||||||
| Bittl, 2005 | 39 | … | Mechanical thrombectomy | 87% | 23% | … |
| Shatsky, 2005 | 62 | 24 forearm | Thromboaspiration ± tPA | 87% | 18% | 69% |
| 36 upper arm | ||||||
| 2 leg | ||||||
| Moossavi, 2007 | 49 | 23 forearm | Mechanical thrombectomy | 96% | 51% | 84% |
| 26 upper arm | 47% | 62% | ||||
| Jain, 2008 | 41 | 21 forearm | Mechanical thrombectomy | 76% | 20% | 54% |
| 20 upper arm | ||||||
| Wu, 2009 | 48 | 48 forearm | Mechanical thrombectomy | 96% | 44% | 89% |
The outcome of surgical revision of thrombosed AVF is summarized in Table III. The studies report thrombectomy with or without patchplasty, but the creation of a new, more proximally located AV anastomosis seems to be particularly advantageous.19, 20, 21 The primary success rate of this latter technique is 90% (range, 80%-100%), with high 1-year primary and secondary patencies of 74% and 87%, respectively. Most studies on surgical thrombectomy of AVF concern forearm accesses, and the results are comparable with the outcome of endovascular treatment in terms of primary success rate (90% vs 89%), but 1-year primary (74% vs 40%) and secondary patency rates (87% vs 72%) are higher.
Table III. Surgical treatment of thrombosed fistulae
| First author, year | No. | Location AVF | Modality | Technical success | Patency (1 yr) | |
|---|---|---|---|---|---|---|
| Primary | Secondary | |||||
| Oakes. 1998 | 29 | 29 forearm | Proximal reanastomosis | 80% | 69% | 89% |
| Morosetti, 2002 | 26 | 17 forearm | Thrombectomy ± proximal reanastomosis | 82% | 93%(6 mon) | … |
| 9 upper arm | 66% | 84% | ||||
| Mickley, 2003 | 30 | 30 forearm | Proximal reanastomosis | 100% | 80% | 95% |
| Ponikvar, 2005 | 268 | … | Thrombectomy ± proximal reanastomosis | 93% | 75% | 77% |
| Georgiadis, 2005 | 59 | 59 forearm | Thrombectomy ± Graft interposition | 95% | … | 85% |
| Palmar, 2006 | 10 | 3 forearm | Thrombectomy | 70% | 51% | 69 |
| 7 upper arm | ||||||
| Lipari, 2007 | 32 | 32 forearm | Proximal reanastomosis or graft interposition | 84% | 73% | 88% |
Discussion
Functional vascular access is of utmost importance for the treatment of ESRD patients who need long-term intermittent hemodialysis treatment. Timely intervention of a thrombosed access is therefore mandatory to re-establish the connection to the artificial kidney. The strategy of whether to use surgical or endovascular means for this purpose remains obscure. Evidence from the literature is lacking and should be analyzed with caution. Implementation of guidelines has resulted in a shift from grafts towards the use of autogenous fistulae for vascular access, particularly in the United States. This feature can be noticed when looking at the access issues that have been published. Before 2002, numerous studies on endovascular and surgical intervention of thrombosed grafts were published, whereas more recently, the treatment of thrombosed fistulae in particular has been referenced.
The meta-analysis of randomized trials of endovascular and surgical prosthetic graft intervention supports the use of surgical techniques for graft thrombectomy.1 This could be mainly attributed to the better initial success rate of surgery. When looking into more detail at these studies, however, several flaws may influence the outcome. For instance, different thrombectomy devices were used, ranging from simple manual thromboaspiration by means of Fogarty catheters, rotational devices to AngioJet rheolytic catheters. Also, in some studies additional thrombolysis with urokinase was performed. These differences in treatment modalities make a fare comparison quite difficult.
One of the randomized studies was a multicenter study with large differences in technical success rate (25% to 100%) between centers, indicating the lack of experience with pharmacomechanical thrombectomy in some centers.4 In addition, the exact surgical procedure in most of the studies was not defined, and it was unclear whether completion angiography and repair of stenoses with patchplasty or graft interposition, or both, was performed. All these issues may negatively influence the outcome of endovascular as well as surgical techniques. Therefore, any reasonable comparison between both techniques in a meta-analysis can hardly been done.
Recent studies report improved success rates of endovascular intervention for thrombosed grafts that are comparable with surgery. This may be attributed to better mechanical devices and the administration of potent thrombolytics (tPA). A major drawback is the high costs associated with the use of mechanical devices.
The incident rate of thrombotic occlusion of autogenous AVF is considerably lower compared with prosthetic graft fistulae. In the past, less attention has been paid to the management of thrombosed fistulae. Since 2000, an increasing number of publications on endovascular and surgical treatment have appeared in the literature.
The mechanism of thrombotic occlusion of the fistula is a progressive stenosis, usually located at the site of the AV anastomosis. In radiocephalic fistulae, side branches may persist to drain the cephalic vein, even when thrombosis has occurred at the anastomosis. A simple reanastomosis of the cephalic vein to the more proximally located radial artery can salvage the access. The primary success rate of this technique is 88% and the 1-year primary patency of 74% is higher compared with thrombectomy alone or combined with graft interposition. Proximal reanastomosis can be performed on an outpatient basis under local anesthesia. The outcome of surgery for upper arm fistulae is worse compared with forearm fistulae.
The primary success rates of endovascular treatment of thrombosed fistulae are comparable with surgical revision. Pharmacomechanical thrombolysis or mechanical thrombectomy alone yields similar outcome, with a >90% clinical success rate. One-year primary and secondary patencies are, however, reduced compared with surgery. Tight stenoses, usually at the AV anastomosis, may be resistant to angioplasty with conventional balloons (up to 12 atm), and (ultra) high-pressure (to 32 atm) or cutting balloon angioplasty might be needed. A residual stenosis also may progress in time and result in recurrent thrombosis.
The organization of an acute endovascular service with the availability of trained interventionalists, experienced in pharmacomechanical thrombectomy of vascular access, is a prerequisite for good functional early and long-term patency. In addition, in those centers that lack adequate endovascular services, timely access to the operating theater is mandatory for surgical revision and salvage of the access.
Definitive randomized trials comparing endovascular with surgical intervention for thrombosed autogenous fistulae are desperately needed.
Conclusion
Endovascular and surgical intervention for thrombosed dialysis prosthetic graft fistulae results in comparable early success and long-term primary and secondary patency rates. Surgery yields a better outcome for autogenous arteriovenous fistulae, in particular in the long-term.
References
- . A metaanalysis comparing surgical thrombectomy, mechanical thrombectomy, and pharmacomechanical thrombolysis for thrombosed dialysis grafts. J Vasc Surg. 2002;36:939–945
- . Thrombolysis versus thrombectomy for occluded hemodalisis grafts. Am J Surg. 1994;167:473–476
- . Treatment of thrombosed dialysis access grafts: randomized trial of surgical thrombectomy versus mechanical thrombectomy with the Amplatz device. J Vasc Interv Radiol. 1996;7:185–192
- Comparison of the angiojet rheolytic catheter to surgical thrombectomy for the treatment of thrombosed hemodialysis grafts (Peripheral AngioJet Clinical Trial). J Vasc Interv Radiol. 1999;10:1195–1205
- . Thrombolysis versus surgical thrombectomy for the treatment of dialysis graft thrombosis: pilot study comparing costs. J Vasc Interv Radiol. 1996;7:507–512
- . Prospective randomized comparison of surgical versus endovascular management of thrombosed dialysis access grafts. J Vasc Surg. 1997;26:373–380
- . Endovascular versus surgical treatment for thrombosed hemodialysis grafts: A prospective, randomized study. J Vasc Surg. 1999;30:1016–1023
- . Thrombosed dialysis access grafts: randomized comparison of the Amplatz thrombectomy device and surgical thromboembolectomy. Eur Radiol. 2004;14:2009–2014
- Alteplase for hemodialysis access graft thrombolysis. J Vasc Interv Radiol. 2002;13:775–784
- . Prospective assessment of hemodialysis access patency after percutaneous intervention: Cox proportional hazards analysis. Catheter Cardiovasc Interv. 2005;66:309–315
- Percutaneous therapy to maintain dialysis access successfully prolongs functional duration after primary failure. Ann Vasc Surg. 2007;21:474–480
- . Secondary patency of thrombosed prosthetic vascular access grafts with aggressive surveillance, monitoring and endovascular management. Eur J Vasc Endovasc Surg. 2008;36:356–365
- . Surgical thrombectomy for thrombosed dialysis grafts: comparison of adjunctive treatments. World J Surg. 2008;32:241–245
- Treatment of failed native arteriovenous fistulae for hemodialysis by interventional radiology. Kidney Int. 2000;57:1124–1140
- . Percutaneous treatment of thrombosed primary arteriovenous hemodialysis access fistulae. Kidney Int. 2000;57:1169–1175
- . Salvage of occluded autologous arteriovenous fistulae. Am J Kidney Dis. 2000;36:804–810
- Single-center experience with the Arrow-Trerotola Percutaneous Thrombectomy Device in the management of thrombosed native dialysis fistulas. J Vasc Interv Radiol. 2005;16:1605–1611
- Non-surgical salvage of thrombosed arterio-venous fistulae: a case series and review of the literature. Semin Dial. 2007;20:459–464
- . Surgical salvage of thrombosed arteriovenous fistulas and grafts. Ther Apher Dial. 2005;9:245–249
- Use of short PTFE segments (<6 cm) compares favorably with pure autologous repair in failing or thrombosed native arteriovenous fistulas. J Vasc Surg. 2005;41:76–81
- Outcomes of surgical revision of stenosed and thrombosed forearm arteriovenous fistulae for haemodialysis. Nephrol Dial Transplant. 2007;22:2605–2612
PII: S0741-5214(09)01364-0
doi:10.1016/j.jvs.2009.06.058
© 2009 Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
