Embolization Therapy: Principles and Clinical Applications, 1 Ed.

Dysfunctional Hemodialysis Accesses

Dheeraj Rajan • Ricardo Yamada

In 2006, the last update of the Kidney Disease Outcomes Quality Initiative along with the Fistula First program reinforced the concept of arteriovenous fistula (AVF) as the preferred hemodialysis (HD) access, given its longer durability, less risk of infection, and less overall cost compared with arteriovenous grafts (AVGs).1,2 Unfortunately, AVF primary failure can reach up to 53% according to one series.3Therefore, multiple strategies have been implanted to improve outcome, including better patient selection, refined surgical techniques, and access surveillance. Recently, endovascular techniques have been successfully used to improve AVF development, including collateral veins embolization. In addition, primary failure is not the only obstacle encountered among HD accesses, and percutaneous embolization has also been applied to solve other access-related problems such as arterial “steal syndrome” and traumatic fistula formation adjacent to main AVG.


Anatomic Considerations

The ideal AVF is the one that can be easily cannulated, providing sufficient blood flow with satisfactory long-term patency. Objective parameters have been established to evaluate those characteristics, and they include minimal flow of 600 mL per minute, diameter bigger than 6 mm, and location approximately 6 mm from the skin surface.4 To achieve these goals, quality of the vessels being used is a key factor. The minimum diameter before AVF creation for arteries should be 2.0 mm and for veins 2.5 mm, as shown by Wong et al.5 who demonstrated that fistulas created with vessels smaller than 1.6 mm were associated with early failure. Also, too calcified arteries and/or preexisting focal venous/arterial stenoses can prevent fistula maturation.

The most common and preferred AVF is the radiocephalic fistula, which is created just above the wrist. The cephalic vein is cut and the portion going up to the forearm is anastomosed to the side of the radial artery in an end-to-side anastomosis. The distal end directed to the hand is then ligated. The second most used AVF is the brachiocephalic fistula, created just above the elbow. Again, this is an end-to-side anastomosis, in which the proximal end of the cephalic vein is anastomosed to the side of the brachial artery. The distal cephalic vein is ligated. Finally, the least common AVF is the brachiobasilic fistula, in which the end of the basilic vein is anastomosed to the side of the brachial artery. Because the basilic vein has a deep course, it needs to be mobilized to the subcutaneous tissue where it can be easily accessed.


Basically, fistula creation bypasses a high-resistance vascular bed, diverting blood flow toward a low-resistance venous circuit, which is now receiving “arterialized” inflow. In response to this new environment, progressive arterial and vein dilation are expected over time to accommodate this high blood volume in this new low-resistance system. Initially, vein dilation results from the increased blood pressure within it. In addition, a second component plays an important role and that is called beneficial vascular remodeling.6,7

In this process, the increased blood flow induces an increased longitudinal shear stress against the vein wall, and this promotes endothelial cell quiescence and secretion of anti-inflammatory and anticoagulant agents. This will, ultimately, result in vein dilation and reduction in neointimal hyperplasia. On the other hand, reduction in both blood flow and longitudinal shear stress leads to endothelial cell activation and release of inflammatory and procoagulant factors, which will result in neointimal hyperplasia and vasoconstriction, an opposite process called negative vascular remodeling. Another important part of this process is the degree of medial hypertrophy, which is determined mainly by the transmural pressure. It has been demonstrated that increased intraluminal pressure activates smooth muscle cells, cytokine expression, and production of extracellular matrix components.8 The final result is thickening of the muscular layer of the vessel wall. Therefore, fistula final lumen diameter is determined by the combination of three components: vein dilation, neointimal hyperplasia, and medial hypertrophy. Most of the times, lumen loss due to intimal hyperplasia and medial hypertrophy is compensated by vein dilation, and the final diameter is large enough to promote the necessary blood flow for an efficient dialysis.

Blood flow through the fistula is the major determinant of beneficial vascular remodeling and, in consequence, fistula maturation. Some anatomic problems can prevent adequate flow within the fistula, halting fistula development, and the most common are stenotic lesions and presence of accessory veins.9 According to Beathard et al.,9 outflow venous stenosis is found in 78% of the patients with a failed-to-mature fistula, and presence of accessory veins is found in 46% of patients. In these scenarios, percutaneous balloon angioplasty has been proven to be a useful tool in facilitating fistula maturation and, more recently, accessory vein embolization has become a valuable option.


Nonmatured Arteriovenous Fistula Embolization

It has been confirmed that presence of accessory veins is the second most anatomic factor that prevents fistula maturation.9,10 In the setting of an AVF, accessory veins divert flow from the main venous channel, which in turn reduces the resistance and blood flow within the venous segment above the branch(es). Flow reduction leads to decreased longitudinal shear stress, which triggers negative vascular remodeling, causing exacerbated neointimal hyperplasia and vasoconstriction. Ultimately, AVF maturation may be compromised. According to Turmel-Rodrigues et al.,11 filling of an accessory vein would not be a cause of poor fistula maturation but only a consequence of an underlying outflow venous stenosis, therefore vein obliteration would not be necessary as long as the stenotic lesion is fixed. Nonetheless, Beathard et al.10 in their series of 100 patients described 12 failed fistulas in which isolated accessory veins without outflow stenoses were the only anatomic culprit factors. All of them were obliterated with 100% success rate. Accordingly, other series have shown that accessory vein obliteration was effective in promoting fistula maturation.12,13

Thus, obliteration of those accessory veins has been applied when AVF fails to mature despite venous stenosis correction or when those accessory channels are the only anatomic abnormality. Initially, vein ligation was the only option. For that, after fistulography, accessory vein location is identified and marked on the skin surface. A small incision is made and after fluoroscopic vein visualization, ligation is performed with 4-0 silk and the skin incision is closed with 4-0 sutures. A less invasive technique has been described without skin incision using two 2-0 polypropylene sutures and inserting the needles into the skin adjacent to the vein and advancing them under the accessory vein until they reach the skin on the opposite side. The notches are then set on the skin surface and ligation is confirmed with a fistulogram. With these techniques, successful fistula maturation has been described as high as 100% with the first technique9 and 88 % with the last one.12

Vein ligation is an excellent option for superficial veins as they can be easily reached with a small surgical incision or even transcutaneously. For deeper veins, ligation is not well suited as damage to muscles, nerves, or tendons can occur. For that location, embolization is a better option because these veins can be selectively catheterized and then occluded. On the other hand, coil embolization of very superficial veins can lead to skin irritation/erosion given the close proximity between the embolic agent and cutaneous tissue (Fig. 28.1).

Embolization has been found to be a valid option by Nikam et al.,13 who treated dysfunctional fistulas with coil embolization alone or in combination with angioplasty when necessary, achieving fistula maturation in 10 out of 14 patients. The technique for fistula embolization has been previously described, with the use of pushable fibered coils as the preferred embolic device.14 To prevent distal coil migration, at least 2-mm upsizing should be applied and, ideally, compact packing should be obtained to occlude flow efficiently (Fig. 28.2). Another important technical point is to maintain a safe distance from the point of embolization and the main outflow vein. There should be at least 5-mm distance to avoid outflow vein stenosis due to postembolization inflammatory reaction and also to prevent misplacement of the embolic agent within the main outflow vein (Fig. 28.3).

Other options of embolic devices, taking into consideration a more controlled deployment, are detachable coils and the Amplatzer Vascular Plug (St. Jude Medical, Inc., St. Paul, Minnesota). Powell et al.15reported using the Amplatzer Vascular Plug to occlude a tributary vein that was diverting flow from the cephalic vein and preventing fistula maturation. This device is made of a self-expandable nitinol mesh that expands within the vessel, achieving the necessary wall apposition. This helps secure the device in place, avoiding distal embolization. For that, the manufacturer advises 30% to 50% upsizing to achieve ideal wall apposition and safe deployment. In addition, the delivery mechanism permits deployment, retrieval/reposition, and redeployment of the plug, if necessary. The second generation of the device, the Amplatzer Vascular Plug II, is believed to promote faster vessel occlusion compared with the first generation15 due to increased surface area and thrombogenicity. The device size varies from 3 to 22 mm, which requires up to a 7-Fr sheath or 9-Fr guiding catheter. The latest generation, the Amplatzer Vascular Plug 4, has a low-profile system, allowing delivery of up to a 8-mm plug through a 5-Fr diagnostic catheter.

Embolization for Steal Syndrome

Steal syndrome occurs when the AVF prevents adequate blood supply to the hand. It is a rare complication and its incidence ranges from 1.7% to 8%, depending on multiple factors, with fistula location being a major factor.16 AVFs at the level of the elbow have higher incidence of steal syndrome compared with the ones located on the wrist. Symptoms and signs include numbing, tingling, pain, and weakness; decreased temperature; and diminished pulses. In severe cases, ulceration can occur. A high index of suspicion should be maintained as those signs and symptoms may be nonspecific and misinterpreted as diabetic neuropathy.

Initially after fistula creation, there is reduction of the distal tissue perfusion, which in turn induces collateral circulation development and peripheral vasodilation. These compensatory mechanisms in combination with a competent ulnar artery and palmar arch permit adequate tissue perfusion in most cases. Unfortunately, in a minority of patients, the hemodynamic changes after AVF creation are not adequately compensated by those mechanisms, leading to insufficient distal blood perfusion to meet the metabolic requirements.

This new hemodynamic environment is part of the complex pathophysiology of steal syndrome, in which the major determinant is the difference between fistula’s resistance and the resistance of the hand’s capillary circulation. Because fistula resistance is lower than the resistance of the distal capillary bed, most or even all blood from the donor artery will be diverted into the AVF, but that should not be an issue, as the ulnar artery and the palmar arch can guarantee adequate hand perfusion (Figs. 28.4 and 28.5A).

However, lack of blood supply can result from either insufficient antegrade blood flow or complete retrograde flow in the artery distal to the fistula.17 The first uncompensated scenario occurs when decreased or absent antegrade flow in the donor artery distal to the fistula is not compensated by a diseased ulnar artery and/or incompetent palmar arch (Fig. 28.5B). A second scenario takes place when fistula’s resistance is so low that not only the whole blood flow from the donor artery is diverted into the outflow vein but also the flow from the ulnar artery, which is directed into the fistula through the palmar arch and distal donor artery in a retrograde fashion (Fig. 28.5C). The last scenario occurs when a proximal stenosis in the donor artery is present, preventing adequate distal flow to the extremity, which is more common in upper arm fistulas (Figs. 28.5D and 28.6A).

Depending on the underlying cause, different treatment strategies should be applied, and arteriography plays an important role in understanding flow dynamics in each particular patient. For example, proximal stenosis in the donor artery can be managed with balloon angioplasty alone (Fig. 28.6). When distal retrograde flow (as shown in Fig. 28.5C) is the underlying problem, surgical or endovascular approaches can be performed.

Initially, surgical strategies aimed to increase fistula resistance by banding, plicating, or lengthening were attempted. Unfortunately, those techniques have been shown to be of little help and were associated with increased risk of fistula thrombosis.18 In 1988, Schanzer et al.19 described a new surgical technique consisting of ligation of the artery distal to the fistula and creation of a distal bypass, the so-called distal revascularization and interval ligation (DRIL) procedure. In consequence, resistance ratio between the fistula and peripheral circulation is decreased.19 This technique has been validated by other authors with series of cases demonstrating symptom resolution and preserved fistula patency.20

Recently, Plumb et al.21 described a less invasive approach using coil embolization to prevent retrograde flow from the distal arterial into the AVF. In their report, a 46-year-old patient with a radiocephalic fistula presented with steal syndrome, and arteriography showed retrograde flow within the distal radial artery as well as in some collateral arterial vessels. The radial artery distal to the fistula and those collateral vessels were embolized with fibered coils during three different sessions, and the patient remained asymptomatic after 6 months follow-up. According to them, the advantage of the endovascular approach was the possibility of collateral circulation embolization, which was contributing to the steal syndrome and would not be suited for surgical ligation. Since then, other case reports have demonstrated use of coil embolization in treating these patients,22,23 accomplishing 100% symptom relief. In addition, Miller et al.22 suggested that coil embolization was preferred over arterial ligation as first-line therapy as it is safe, effective, and a quicker alternative.

Another endovascular technique that has been described to treat steal syndrome is coil embolization of deep collateral drainage veins.24 This is based on the concept that collateral veins impose lower fistula resistance and occluding these collateral channels would increase system resistance, which in turn decreases blood shunt. Kariya et al.24 have successfully demonstrated effectiveness of this approach when dealing with deep accessory veins not suitable for future dialysis access. In their series of five patients, pushable coils were used in four patients and detachable coils were used in one patient based on location and flow velocity, adding more safety to the procedure.24Especially in this high flow velocity environment, one should always consider the use of detachable coils.

In more unusual and extreme situations associated with severe symptoms, based on case-by-case analysis, complete access occlusion is required due to failure of more conservative options or when dialysis is not required anymore.15,25,26 Traditionally, surgical ligation is the preferable option, but in some circumstances, such as presence of extensive ischemic changes and/or arm swelling, surgery is not always possible. In these cases, the fistula can be closed by endovascular methods.

Coils and N-butyl cyanoacrylate (NBCA) have been used for this purpose, and more recently, the Amplatzer Vascular Plug device became another option. Combined use of these embolic devices has also been reported.26,27Regardless of the chosen embolic agent, precise deployment should be done as close as possible to the arterial anastomotic site without extending into the artery to avoid both aneurysm formation in the remnant venous stump and thrombosis of the arterial side. To achieve this goal, some important maneuvers and/or tools can be applied, such as external compression of the outflow vein until complete flow stasis, inflation of an occlusion balloon, and use of embolic agents with controlled delivery system, such as detachable coils and the Amplatzer Vascular Plug. In fact, the vascular plug allows for very precise deployment, which can be performed under road mapping and/or ultrasound guidance,26 with the possibility of retrieval and redeployment in case of inappropriate position.

Traumatic Fistula Embolization

Fistulous connections between the AVG and nearby native veins do not have a well-known incidence, with limited literature reporting an incidence ranging from 0.027% to 9% (Fig. 28.7).28,29 Their formation is related to repeated needle punctures and/or pseudoaneurysm rupture in a setting of high intragraft pressure due to an outflow venous stenosis.30

Most of the time, a graft-to-vein fistula (GVF) is an asymptomatic condition without clinical significance. Nonetheless, in some circumstances, these abnormal communications can result in decreased access flow, decreased HD adequacy, and increased risk of graft thrombosis.30 Clinically, some patients can present with palpable thrill on the arterial limb only and dilated superficial veins on the forearm (or the thigh in patients with groin loop grafts), indicating that the venous limb is likely thrombosed and the flow is maintained through the fistulous connections.28 This situation can be associated with difficult graft cannulation during dialysis, poor clearance, and upper/lower extremity edema.

In such cases, intervention is required, and treatment of the outflow venous stenosis/occlusion, which is almost always present, is mandatory. In fact, balloon angioplasty and/or stent placement might be the only necessary treatment modalities, as normalization of intragraft pressure can result in closure of fistulous communication.31 In situations when angioplasty is not sufficient or presence of the abnormal communication is the only problem, closure of the fistula is required. That can be achieved by surgical ligation as described by Min et al.32 or percutaneous coil embolization as described by Harris.33

Given the rarity of this entity and yet unclear clinical significance, necessity of GVF closure is still not well established. Recently, Margoles et al.34 described a case series of 12 patients with GVF in whom coil embolization was performed in 5 of them. Together with the previous experience in the literature, the authors proposed an algorithm to clarify the need for GVF closure based on presence of stenosis, arm swelling/partial thrombosis, and flow measurements. Basically, patients with symptoms and/or inadequate access flow, who did not respond to angioplasty or who did not present with outflow stenosis, should have their GVF closed.34



• Do not extend the coil to the collateral/vein interface. The inflammatory reaction may cause stenosis of the AVF.

• Consider embolization of venous collaterals only when areas of stenosis have been excluded, including the inflow artery within nonmaturing AVFs.

• Err on the side of shorter coil length used rather than longer coils to prevent extension into the vascular access.

• Amplatzer Vascular Plug use should be reserved for deep vein and prosthetic graft embolization.


• To prevent forward migration of a coil, anchoring the leading edge within a branch vein helps secure the location.

• When using embolization coils, to pack the coil properly, rotate the catheter and gently pull back slowly as you pack the coil.


Percutaneous interventions to dialysis accesses such a balloon angioplasty and mechanical thrombectomy are very well-established and widely used tools to increase access functionality and durability. Embolization has become another important tool in the armamentarium of percutaneous techniques in treating different problems related to dialysis access, as shown in this chapter. Proper knowledge of malfunctioning access pathophysiology and hemodynamics as well as familiarity with embolization technique and different available devices are crucial for successful outcomes.


 1. National Kidney and Urologic Diseases Information Clearinghouse. Kidney disease statistics for the United States. National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health Web site. http://kidney.niddk.nih.gov/kudiseases/pubs/kustats/#7. Published 2012. Updated November 15, 2012. Accessed May 13, 2013.

 2. Navuluri R, Regalado S. The KDOQI 2006 Vascular Access update and Fistula First program synopsis. Semin Intervent Radiol. 2009;26(2):122–124.

 3. Allon M, Robbin ML. Increasing arterio-venous fistulas in hemodialysis patients: problems and solutions. Kidney Int. 2002;62(4):1109–1124.

 4. Vascular Access 2006 Work Group. Clinical practice guidelines for vascular access. Am J Kidney Dis. 2006;48(suppl 1):S176–S247.

 5. Wong V, Ward R, Taylor J, et al. Factors associated with early failure of arterio-venous fistulae for haemodialysis access. Eur J Vasc Endovasc Surg. 1996;12(2):207–213.

 6. Dixon BS. Why don’t fistulas mature? Kidney Int. 2006;70(8):1413–1422.

 7. Roy-Chaudhury P, Spergel LM, Besarab A, et al. Biology of arterio-venous fistula failure. J Nephrol. 2007;20(2):150–163.

 8. Lehoux S, Castier Y, Tedgui A. Molecular mechanisms of the vascular responses to haemodynamic forces. J Intern Med. 2006;259(4):381–392.

 9. Beathard GA, Arnold P, Jackson J, et al; for the Physician Operators Forum of RMS Lifeline. Aggressive treatment of early fistula failure. Kidney Int. 2003;64(4):1487–1494.

10. Beathard GA, Settle SM, Shields MW. Salvage of the nonfunctioning arterio-venous fistula. Am J Kidney Dis. 1999;33(5):910–916.

11. Turmel-Rodrigues L, Mouton A, Birmelé B, et al. Salvage of immature forearm fistulas for haemodialysis by interventional radiology. Nephrol Dial Transplant. 2001;16(12):2365–2371.

12. Faiyaz R, Abreo K, Zaman F, et al. Salvage of poorly developed arterio-venous fistulae with percutaneous ligation of accessory veins. Am J Kidney Dis. 2002;39(4):824–827.

13. Nikam M, Popuri RK, Inaba A, et al. Arterio-venous fistula failure: is there a role for accessory draining vein embolization? J Vasc Access. 2012;13(4):498–503.

14. Rajan DK, ed. Essentials in Percutaneous Dialysis Interventions. Dordrecht, The Netherlands: Springer; 2011:323–339.

15. Powell S, Narlawar R, Odetoyinbo T, et al. Early experience with the Amplatzer Vascular Plug II for occlusive purposes in arterio-venous hemodialysis access. Cardiovasc Intervent Radiol. 2010;33(1):150–156.

16. Zibari GB, Rohr MS, Landreneau MD, et al. Complications from permanent hemodialysis vascular access. Surgery. 1988;104(4):681–686.

17. Wixon CL, Hughes JD, Mills JL. Understanding strategies for the treatment of ischemic steal syndrome after hemodialysis access. J Am Coll Surg. 2000;191(3):301–310.

18. De Caprio JD, Valentine RJ, Kakish HB, et al. Steal syndrome complicating hemodialysis access. Cardiovasc Surg. 1997;5(6):648–653.

19. Schanzer H, Schwartz M, Harrington E, et al. Treatment of ischemia due to “steal” by arterio-venous fistula with distal artery ligation and revascularization. J Vasc Surg. 1988;7(6):770–773.

20. Wixon CL, Mills JL Sr, Berman SS. Distal revascularization-interval ligation for maintenance of dialysis access and restoration of distal perfusion in ischemic steal syndrome. Semin Vasc Surg. 2000;13(1):77–82.

21. Plumb TJ, Lynch TG, Adelson AB. Treatment of steal syndrome in a distal radiocephalic arterio-venous fistula using intravascular coil embolization. J Vasc Surg. 2008;47(2):457–459.

22. Miller GA, Khariton K, Kardos SV, et al. Flow interruption of the distal radial artery: treatment for finger ischemia in a matured radio-cephalic AVF. J Vasc Access. 2008;9(1):58–63.

23. Shukla PA, Contractor S, Huang JT, et al. Coil embolization as a treatment alternative for dialysis-associated steal syndrome. Vasc Endovascular Surg. 2012;46(7):546–549.

24. Kariya S, Tanigawa N, Kojima H, et al. Transcatheter coil embolization for steal syndrome in patients with hemodialysis access. Acta Radiol. 2009;50(1):28–33.

25. Bui JT, Gaba RC, Knuttinen MG, et al. Amplatzer vascular plug for arterio-venous hemodialysis access occlusion: initial experience. J Vasc Access. 2009;10(1):5–10.

26. Ozyer U, Harman A, Aytekin C, et al. Application of the AMPLATZER vascular plug in endovascular occlusion of dialysis accesses. Cardiovasc Intervent Radiol. 2009;32(5):967–973.

27. Owens CA, Bui JT, West DL, et al. Use of the Amplatzer Vascular Plug as a coil constrainer during endovascular occlusion of a dialysis shunt. Cardiovasc Intervent Radiol. 2007;30(4):754–756.

28. Standage BA, Schuman ES, Quinn SF, et al. Single limb patency of polytetrafluoroethylene dialysis loop grafts maintained by traumatic fistulization. Ann Vasc Surg. 1998;12(4):364–369.

29. Dousset V, Grenier N, Douws C, et al. Hemodialysis grafts: color Doppler flow imaging correlated with digital subtraction angiography and functional status. Radiology. 1991;181(1):89–94.

30. Haddad NJ, Vachharajani TJ, Van Cleef S, et al. Iatrogenic graft to vein fistula (GVF) formation associated with synthetic arterio-venous grafts. Semin Dial. 2010;23(6):643–647.

31. Kanterman RY, Vesely TM. Graft-to-vein fistulas associated with polytetrafluoroethylene dialysis grafts: diagnosis and clinical significance. J Vasc Interv Radiol. 1995;6(2):267–271.

32. Min SK, Park YH, Lee HH, et al. Iatrogenic fistula between prosthetic haemodialysis access graft and autogenous vein: unusual cause of graft thrombosis. Nephrol Dial Transplant. 2004;19(10):2647–2649.

33. Harris V. Percutaneous coil embolization of an iatrogenic fistula from a polytetrafluoroethylene dialysis shunt to the superficial venous system. J Vasc Interv Radiol. 1993;8:138–141.

34. Margoles HR, Shlansky-Goldberg RD, Soulen MC, et al. A proposed management algorithm for fistulae between hemodialysis access circuits and adjacent veins. J Vasc Access. 2012;13(3):374–380.