Strategies for Management

of Ischemic Steal Syndrome

Paul Nash Suding, MD, and Samuel Eric Wilson, MD

Constructing vascular access for hemodialysis causes changes in blood flow to the
extremity, which can lead to distal ischemia. Ischemic steal syndrome is manifested by
pain; weakness; pallor; and, in severe cases, ulceration and tissue loss. Severe ischemia,
requiring reintervention, has an incidence of 4%, although some degree of ischemia
causing pain or parasthesias occurs in 10% to 20% of patients following access construc-
tion. Pathophysiology may be on the basis of inadequate arterial collateral inflow due to
occlusive disease, particularly involving the medium-sized vessels, or high flow in a fistula
exceeding the inflow capacity in the absence of intrinsic occlusive disease of the inflow
arteries. Predicting steal remains difficult, although certain patient characteristics and
preoperative techniques can help identify those patients in whom arteriovenous fistulas
have an increased risk of causing steal. Patients with diabetes, multiple access procedures,
and constructions based on proximal arteries are more prone to ischemia. Ultrasonography
and digital-brachial indices measured by photoplethysmography or Doppler techniques
have been used to predict fistulas that are more likely to cause ischemia, but these fall short
of reliability. Operative techniques for correcting steal include arteriovenous fistula liga-
tion, percutaneous transluminal angioplasty, banding or restrictive procedures, and distal
revascularization interval ligation or modifications of this technique. Operative intervention
for ischemic steal syndrome successfully resolves ischemia in 80% to 95% of patients.
Some patients can have persistent pain despite healing of ulceration.
Semin Vasc Surg 20:184-188 © 2007 Elsevier Inc. All rights reserved.

I SCHEMIC STEAL SYNDROME is a complication of vascu-

lar access surgery and remains a persistent problem and a
cause of litigation. Recent work has reexamined the physiol-
retrograde flow will be present in the artery just distal to the
fistula in 80% to 90% of all AV fistulas. Reversal of distal
arterial flow usually occurs when the diameter of the fistula
ogy, preoperative evaluation, and treatment of ischemia. Al- opening exceeds the diameter of the inflow artery. Distal
though several procedures provide definitive treatment, pulses are frequently diminished or absent. Despite these
there is no consensus on reliable prediction of steal. In this physiological changes, clinically evident ischemia may not be
article, we examine the pathophysiology of steal, detail meth- present.
ods for preoperative recognition, and review the results of Arterial steal after AV fistula construction may be based on
procedures to correct ischemia. inadequate collateral perfusion secondary to occlusive dis-
ease or because of very-high fistula flow. Frequently, both
Pathophysiology factors contribute to symptomatic steal. Pathological changes
in the arteries of patients in renal failure, particularly when
Differentiating between the hemodynamic changes expected the cause is diabetes, can contribute to ischemia. Diabetics
after arteriovenous (AV) fistula construction and pathological suffer medial calcification of the vessels, which leads to ob-
steal remains difficult. Construction of an AV fistula leads to struction of flow in medium and small distal arteries. Lack of
decreased resistance to blood flow and increased cardiac out- collateral blood flow increases the incidence of ischemia fol-
put. Blood flow can reverse in the distal vascular bed so that
lowing access surgery. Interestingly, Yeager et al,1 in a recent
review of digital ischemia, found finger gangrene in some
hemodialysis patients even though the AV fistula was located
Department of Surgery, University of California, Irvine, Irvine, CA.
Address reprint requests to Samuel Eric Wilson, Department of Surgery,
in the opposite arm. Hemodialysis patients, particularly dia-
University of California, Irvine, 101 City Drive South, City Tower, Suite betics, may have intrinsic vascular disease contributing to
810, Irvine, CA 92868. E-mail: [email protected] ischemia, regardless of access-construction techniques. In fe-

184 0895-7967/07/$-see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1053/j.semvascsurg.2007.07.009
Management of ischemic steal syndrome 185

Table 1 Incidence of Ischemic Steal Syndrome

Ischemia Symptoms Ischemia Requiring
First No. of Not Requiring Operative
Author Year Study Design Patients/Procedures Operation Intervention
Knox8 2002 Retrospective review 1,138 patients over 6 Not described 55 (4.8%)
years
Valentine5 2002 Prospective review 72 patients over 3 years Not described 14 (19%)
Papasavas6 2003 Prospective review 35 patients 6 (17%) 3 (9%)
Lazarides3 2003 Retrospective review of 569 procedures Not described 24 (4.2%)*
proximal procedures
Davidson4 2003 Prospective review 325 procedures on 217 4 (1.2%) 16 (4.9%)
patients
Meyer7 2002 Retrospective review 1,253 patients over 5 Not described 21 (1.7)%
years
Morsy2 1998 Retrospective review 409 procedures on 352 Not described 13/299 (4.3%) of AV
patients over 5 years grafts, 2/110 (1.8%)
AV fistulas
Abbreviation: AV, arteriovenous.
*Describes 28 cases but 4 patients were referrals and not part of the original 569 procedures.

male patients who have relatively small arteries, a high-flow Diagnosis

fistula or graft (⬎750 cc/minute) can exceed the capacity of the
feeding arterial system, even in the absence of arterial inflow Steal syndrome remains a clinical diagnosis based on paras-
disease. thesias, pain, and ulceration and tissue loss. Common com-
plaints are parasthesias, pain, and hand stiffness. Physical
exam will show pallor, diminished sensation, and, ultimately,
Prevalence ulceration and gangrene. The radial pulse is usually absent;
although in some patients, ischemia of fingers may be present
Steal syndrome will develop in 2% to 20% of patients with with a palpable radial pulse. Compression of the shunt often
AV fistulas or grafts (Table 1).2-8 Most recent retrospective relieves symptoms temporarily and augments the distal
studies show operative intervention for steal in about 4% of pulse. Increased heart rate and blood volume or symptoms of
patients after vascular access surgery. Prospective studies congestive heart failure are not usually caused by the AV
typically diagnose symptomatic steal in 15% to 20% of pa- fistula in hemodialysis patients, probably because of the
tients. The difference likely represents bias from the study blood-volume adjustment that occurs at each dialysis.
design, in that follow-up interviews with patients are more Sidawy et al,9 present a useful summary for describing
likely to elicit ischemic symptoms. Despite advances in pre- ischemia reflecting the spectrum of symptoms that are
operative evaluation and surgical techniques, this incidence helpful in assessing patients following access construction
has been largely unchanged for the last 20 years. Chronic (Table 2). The decision to intervene depends upon the se-
hemodialysis patients remain at risk for ischemic complica- verity of symptoms and the physical findings. Severe isch-
tions in the absence of an AV fistula. emia and threat of tissue loss will typically require operative
Recent work continues to show early and late presentation intervention.
of symptoms of steal. AV grafts typically develop steal imme-
diately following surgery. Ischemia caused by autologous fis-
tulas presents later as the vein matures and dilates, allowing
Anticipating Steal
increased blood flow. Lazarides et al3 report the median time Predicting patients who will suffer ischemia remains difficult,
to recognition of symptoms as 2 days in AV graft group com- although the risk of steal is anticipated to be greater in dia-
pared to 165 days for autologous fistulas. betic patients and those with multiple previous vascular ac-

Table 2 Classification of Steal Symptoms in Patients with Arteriovenous Shunts9

Grade Symptoms Treatment
Grade 0 No steal None
Grade 1, mild Cool extremity with few symptoms but steal demonstrable None
by flow augmentation with access occlusion
Grade 2, moderate Intermittent ischemia only during dialysis/claudication Intervention sometimes needed
Grade 3, severe Ischemic pain at rest/tissue loss Intervention mandatory
186 P.N. Suding and S.E. Wilson

Table 3 Results of Distal Revascularization and Interval Ligation Procedure

First Author Year No. of Patients Patency
Diehl16 2003 13 90% at 6 months, 78% at 12 months, and 71% at 24 months
Knox8 2003 55 83% at 12 months, and 71% at 48 months
Korzets17 2003 11 90% and 80% at 12 and 24 months
Sessa18 2004 18 94% at 12 months

cess operations.1,4,10 Arterial occlusive disease in critical ar- persistent pain symptoms despite improvements in hemody-
eas, such as the palmar vessels or collateral vessels, leads to namics and healing of ulcers.
steal, as retrograde flow diverts blood away from the hand Before or during operation for ischemia, we often employ
and collateral filling is inadequate. Multiple prior procedures angiography. While the cause of ischemia is most frequently
also remain an independent risk factor, due in part to the high flow through the fistula, combined with inadequate col-
more proximal location of the new AV shunt and the loss of lateral circulation, proximal stenosis of the inflow artery can
arterial inflow. also lead to steal.13 In this instance, resolving the stenosis by
Several measurements may be obtained preoperatively to percutaneous transluminal angioplasty may obviate the need
predict increased risk of steal help. Photoplethysmography, to modify to the graft itself. This may be readily accomplished
ultrasonography, and angiography are useful to document in the interventional or operating suite without interfering
the hemodynamics of ischemia. Decreased preoperative fin- with the fistula by performing standard percutaneous trans-
ger pressures or digital-brachial indices (DBI) can suggest luminal dilatation with addition of a stent if there is a ⬎30%
those patients that later develop steal syndrome. Digital-bra- residual stenosis.
chial index is the ratio of brachial to digital blood pressures, Banding, partial suturing, application of a clip, or other
measured on the fingers via photoplethysmography or Dopp- methods to restrict shunt diameter remain accepted proce-
ler technique. While there is no absolute value for the index dures for reducing blood flow in the graft, although one must
that predicts those who will develop steal, indices ⬍1.0 sug- be careful because thrombosis frequently complicates this
gest increased risk. Setting an arbitrary level of preoperative approach. We narrow inflow simply by placing a medium-
DBI at ⬍0.8 provides a sensitivity of 29% and a specificity of sized hemoclip on the vein adjacent to the arterial anastomo-
93%.5 Another smaller series in which a postoperative DBI ⬍ sis while monitoring the digital or radial artery Doppler sig-
0.6 was selected to predict steal showed sensitivity of 100% nals. A test narrowing of the vein helps in deciding the
and specificity 76%.6 Large changes between preoperative position of the clip. Beware that loss of the anastomotic thrill
and immediate postoperative also suggest steal. Ultrasonog- indicates too much flow restriction and subsequent throm-
raphy is useful for assessing the etiology of steal and the bosis is likely. Polytetrafluoroethylene bands14 can also be
effectiveness of modification of the fistula. Preoperative ul- positioned circumferentially adjacent to the arterial anasto-
trasonography has been shown to be useful for analyzing new mosis. Meyer et al7 describes a series of seven patients treated
constructions and developing operative treatments. The rel- with banding in whom blood flow was measured intraoper-
atively infrequent occurrence of steal syndrome compared to atively, and the modification relieved ischemia. No mention
lack of maturation and thrombosis, however, places the em- was made of thrombosis rate, the main criticism of banding
phasis of postoperative evaluation on the adequacy of con- procedures. Goel et al15 report a modification of the banding
struction rather than prevention of steal.11 procedure using an endovascular approach. A ligature was
placed over the shunt near the AV anastamosis after insertion
of an angioplasty balloon to maintain a standard 4 or 5-mm
Operative Treatment diameter. All shunts remained patent in the study follow-up
Treatment for ischemia includes fistula or graft ligation, in- periods of 3 to 11 months.
flow reduction, and distal artery ligation along with revascu- For the patient whose access site is the last available or
larization, or bypass of the arterial origin of the fistula with- essential to maintain hemodialysis, reconstruction with by-
out distal interruption. Ligation of the fistula, or division and pass and/or ligation is a proven method. Most experience has
oversewing of the graft leaving a small polytetrafluoroethyl- been obtained with distal revascularization and interval liga-
ene cuff where it was joined to the inflow artery, may be tion (DRIL) (Table 3),8,16-18 although other modifications of
necessary, and can be particularly valuable for diabetic pa- the technique have been reported recently. The operation
tients with severe atherosclerotic changes, because there is consists of a bypass from a proximal position on the inflow
restoration of the preoperative inflow.12 The problem with artery to a site distal to the fistula opening, using a reversed
access ligation is that a site for access has been lost and a new saphenous vein graft. Then the artery is ligated just distal to
site must be constructed. Preservation of the access site, re- the access site. Retrograde flow in the distal artery is stopped
gardless of the technique employed, may not resolve all and perfusion of the extremity is restored with the bypass
symptoms of ischemia. Most series report complete resolu- graft. A suitable length and diameter of autogenous vein is
tion of pain in 70% to 80% and partial resolution in 80% to preferable because prosthetic conduits are more prone to late
95%. It is not uncommon, however, for patients to have thrombosis. One should be cautious in selecting patients
Management of ischemic steal syndrome 187

for this operation, because late occlusion of the bypass graft Conclusion
with the inflow artery now ligated can threaten the extremity.
Absence of a suitable autogenous vein for the operation less- Steal syndrome leading to severe ischemia with rest pain
ens the utility of the procedure. and tissue loss affects ⬍5% of patients after construction
Published results of the DRIL procedure are good. Diehl of dialysis access grafts. No preoperative evaluation accu-
rately predicts steal, although diabetes, multiple prior ac-
et al16 reported a series of 12 patients in whom patency fol-
cess operations, and DBI ⬍ 0.8 suggest patients at in-
lowing DRIL procedure was 90% at 6 months, 78% at 12
creased risk. Ultrasonography and angiography are useful
months, and 71% at 24 months.16 Knox et al8 has comparable
intraoperative tools for studying the hemodynamics of
data with 86% at 12 months, and 80% at 48 months.8 Korzets
steal and to ascertain whether surgical modification has
et al17 report functional grafts of approximately 90% and improved blood flow. Surgical treatment consists of occlu-
80% at 12 and 24 months. Sessa et al18 report patency of sion, restriction, or DRIL procedures, although variations
94% at 12 months, although 39% of the patients died of on revascularization and ligation have been used success-
unrelated causes within 6 months. The excellent patency fully.
rates support some authors’ contention that DRIL should
be used as the primary modification for all ischemic pa-
tients, although many surgeons would still begin with in- References
1. Yeager RA, Moneta GL, Edwards JM, et al: Relationship of hemodialysis
flow restriction or interruption and relocation of the ac- access to finger gangrene in patients with end-stage renal disease. J Vasc
cess. Surg 36:245-249, 2002
Modifications of the DRIL procedure have been proposed 2. Morsy AH, Kulbaski M, Chen C, Isiklar H, Lumsden AB: Incidence and
recently. Minion et al19 describe a technique termed revision characteristics of patients with hand ischemia after a hemodialysis ac-
cess procedure. J Surg Res 74:8-10, 1998
using distal inflow. The fistula is ligated at its origin followed
3. Lazarides MK, Staramos DN, Kopadis G, Maltezos C, Tzilalis VD, Geor-
by reestablishment of the fistula inflow via bypass from a giadis GS: Onset of arterial ‘steal’ following proximal angioaccess: im-
distal arterial source. In effect, revision using distal inflow mediate and delayed types. Nephrol Dial Transplant 18:2387-2390,
lengthens the fistula by repositioning the arterial anasto- 2003
4. Davidson D, Louridas G, Guzman R, et al: Steal syndrome complicating
mosis more distally, thereby reducing the size of the inflow
upper extremity hemoaccess procedures: incidence and risk factors.
artery as well as increasing resistance in the fistula con- Can J Surg 46:408-412, 2003
duit. Their series consisted of only brachial artery shunts 5. Valentine RJ, Bouch CW, Scott DJ, et al: Do preoperative finger pres-
and the reconstructed arterial inflow anastomosis was re- sures predict early arterial steal in hemodialysis access patients? A pro-
spective analysis. J Vasc Surg 36:351-356, 2002
constructed on either the radial or the ulnar arteries. An-
6. Papasavas PK, Reifsnyder T, Birdas TJ, Caushaj PF, Leers S: Prediction
tegrade flow in the inflow vessel was observed following of arteriovenous access steal syndrome utilizing digital pressure mea-
the modification. surements. Vasc Endovascular Surg 37:179-184, 2003
Zanow et al20 describes proximalization of the arterial inflow. 7. Meyer F, Muller JS, Grote R, Halloul Z, Lippert H, Burger T: Fistula
banding—success-promoting approach in peripheral steal syndrome
The arterial supply of the shunt is moved to a more proximal
Zentralbl Chir 127:685-688, 2002
artery using a small-caliber polytetrafluoroethylene graft (4 8. Knox RC, Berman SS, Hughes JD, Gentile AT, Mills JL: Distal revascu-
or 5 mm). This technique was applied in 30 patients with larization-interval ligation: a durable and effective treatment for isch-
distal ischemia. Patency was comparable to DRIL with pri- emic steal syndrome after hemodialysis access. J Vasc Surg 36:250-255,
2002
mary patency 87% at 1 year and 78% at 3 years. The selection
9. Sidawy AN, Gray R, Besarab A, et al: Recommended standards for
of patients who received proximalization of the arterial in- reports dealing with arteriovenous hemodialysis accesses. J Vasc Surg
flow in their series is not clear because some patients with 35:603-610, 2002
ischemia also underwent banding, access ligation, and recon- 10. Sessa C, Pecher M, Maurizi-Balzan J, et al: Critical hand ischemia after
struction of the arterial anastomosis. Balaji et al21 describe angioaccess surgery: diagnosis and treatment. Ann Vasc Surg 14:583-
593, 2000
treatment by simply ligating the brachial artery without re- 11. Brown PW: Preoperative radiological assessment for vascular access.
vascularization. Blood flow distal to the ligation was main- Eur J Vasc Endovasc Surg 31:64-69, 2006
tained by collateral blood vessels. The operation was success- 12. Tordoir JH, Dammers R, van der Sande FM: Upper extremity ischemia
ful, as suggested by the authors, because the patient had a and hemodialysis vascular access. Eur J Vasc Endovasc Surg 27:1-5,
2004
prior forearm fistula that had already encouraged collateral 13. Malik J, Slavikova M, Maskova J: Dialysis access-associated steal syn-
blood vessel development. Lastly, a simple arterial bypass drome: the role of ultrasonography. J Nephrol 16:903-907, 2003
graft from the artery just proximal to the fistulas to a site just 14. Papalois VE, Haritopoulos KN, Farrington K, Hakim NS: Successful
distal to the fistula (without arterial ligation) may be consid- reversal of steal syndrome following creation of arteriovenous fistula by
banding with a ringed Gore-Tex cuff: a new technique. Int Surg 88:52-
ered. 54, 2003
15. Goel N, Miller GA, Jotwani MC, Licht J, Schur I, Arnold WP: Minimally
(Editor’s note: Distal radial artery ligation alone may suffice invasive limited ligation endoluminal-assisted revision (MILLER) for
in treating steal associated with radiocephalic fistulas where treatment of dialysis access-associated steal syndrome. Kidney Int 70:
765-770, 2006
digital compression of the radial artery beyond the fistula sig- 16. Diehl L, Johansen K, Watson J: Operative management of distal isch-
nificantly improves digital perfusion, as demonstrated by emia complicating upper extremity dialysis access. Am J Surg 186:17-
plethysmography.) 19, 2003
188 P.N. Suding and S.E. Wilson

17. Korzets A, Kantarovsky A, Lehmann J, et al: The “DRIL” procedure—a approach to dialysis-associated steal syndrome. Ann Vasc Surg
neglected way to treat the “steal” syndrome of the hemodialysed pa- 19:625-628, 2005
tient. Isr Med Assoc J 5:782-785, 2003 20. Zanow J, Kruger U, Scholz H: Proximalization of the arterial inflow: a new
18. Sessa C, Riehl G, Porcu P, et al: Treatment of hand ischemia following technique to treat access-related ischemia. J Vasc Surg 43:1216-1221, 2006
angioaccess surgery using the distal revascularization interval-ligation 21. Balaji S, Evans JM, Roberts DE, Gibbons CP: Treatment of steal syn-
technique with preservation of vascular access: description of an 18- drome complicating a proximal arteriovenous bridge graft fistula by
case series. Ann Vasc Surg 18:685-694, 2004 simple distal artery ligation without revascularization using intraoper-
19. Minion DJ, Moore E, Endean E: Revision using distal inflow: a novel ative pressure measurements. Ann Vasc Surg 17:320-322, 2003