A case of digital ischemia secondary to vascular access (steal) treated with banding
The patient was a 44-year-old male, long-term diabetic and former smoker who had been on dialysis just over a year via a left two-stage transposed basilic fistula placed in a local hospital. He reports some degree of coldness and pain in the hand below the access, and was advised by his local surgeon to “keep an eye on it”. The patient denies having been treated medically nor having had any diagnostic tests. In the weeks before he was scheduled to come to my office, the hand pain became so serious that he was unable to tolerate the full treatment. Unwilling to wait for his appointment in the office, he came to my hospital’s emergency room.
On examination, the patient was a stocky middle-aged male with a medially placed left transposed basilic fistula (image #1). There was a strong thrill without pulsatility, the fistula was easily palpable, and there were signs of regular successful cannulation. The hand was beet red when down, and pale white when raised. There was no radial pulse.
“Banding for steal” image #1
The fistula is a basilic transposition, in this case transposed posteriorly and only up to the upper mid arm. But it was usable, and the complaint was that the patient had burning pain in the hand since the transposition, eswpecially during dialysis. The anastomosis was generous, and I exspected to find a “Firehose fistula” with excessive flows. I suggested a banding of the inflow.
“Banding for steal” image #2 – Non-invasive arterial study
Digital pressures relative to the brachial systolic pressure (140mmHg) are poor with absolute pressures between 29 and 60 mmHg, and indices between .21 and .43 (upper section). This is critical ischemia, usually causing severe pain, motor dysfunction, sensory loss and often tissue necrosis.
These improve significantly when the fistula is compressed, with digital pressures improving to 80 to 112 mmHg, and indices .64 to .80 (lower section). These pressures are still not normal, are consistent with mild to moderate steal, and would initially be treated medically.
“Banding for steal” image #3 – retrograde arterial study
After confirming the diagnosis by showing poor amplitude waveforms distal to the fistula anastomosis and diminished digital pressures, we may request a formal upper extremity angiogram from the arch out to the fingers if the study suggests proximal arterial disease, or we may do an “on table” angiogram from the fistula. Here, the fistula was cannulated in the OR, and a Kumpe catheter run retrograde through the anastomosis and up to the subclavian – the artery is unremarkable, and there is no flow-limiting stenosis.
“Banding for steal” image #4 – retrograde arterial study
Again, the Kumpe catheter was run retrograde through the anastomosis and up to the subclavian for contrast injection – the axillary and brachial arteries seem unremarkable, without flow limiting stenosis.
“Banding for steal” image #5 – retrograde arterial study
The continuation of the retrograde arterial study shows an unremarkable brachial artery, but all the contrast goes into the fistula and none down to the hand. The anastomosis is generous, and the leading portion of the fistula also.
“Banding for steal” image #6 – retrograde arterial study
The trick to visualizing the arterial system distal to the arterial anastomosis in a case of steal is to compress the access outflow – if the blood flow is not allowed to flow freely into the fistula it will go into the native arterial system. Here we see clearly the size mismatch between the modest artery and the wide open anastomosis with the big fistula. The distal brachial artery splits into radial, interosseous and ulnar branches showing evidence of small vessel disease.
“Banding for steal” image #7 - retrograde arterial study
Visualizing the arterial system in the distal forearm and wrist, we see that the radial and interosseous arteries are diminutive, and the ulnar artery is the dominant supply to the hand.
“Banding for steal” image #8a and 8b – inflow reduction
The leading segment of the fistula is exposed, and reduced in size by putting a side-biting clamp on the fistula, cutting off the excluded part of the fistula, and repairing the damage with running suture (below).
“Banding for steal” image #8a and 8b – inflow reduction
The leading segment of the fistula is exposed, and reduced in size by putting a side-biting clamp on the fistula, cutting off the excluded part of the fistula, and repairing the damage with running suture (below).
“Banding for steal” image #9 – applying the band
Flows are measured through the process of applying and tightening the band – looking for an absolute reduction to the fistula in order to drive more flow to the hand, without reducing the flow enough to complicate dialysis, and finding a happy medium – in this case around 1000 cc/min.
“Banding for steal” image #10 – final result
A follow-up study is done, showing flow into the fistula and also down to the hand – balance has been restored between the demand of the fistula and the need of the hand.
Non-invasive studies confirmed critical ischemia in the affected hand, and also an element of steal (image #2). In the operating room, an on-table fistulogram and arterial study were performed, showing a large and well-formed fistula with no outflow nor central stenosis. A retrograde arterial study was performed, showing a modest but otherwise unremarkable subclavian, axillary and brachial artery draining almost totally into the fistula (images #3, 4 and 5). Fistula compression arterial injections demonstrated an increasing degree of small vessel disease in the forearm and hand, with the ulnar being dominant (images #6 and 7).
The fistula flow was around 1200 cc/min under general anesthesia, which depresses cardiac function enough to distort the results – usually, we do these studies with regional anesthesia to get a more accurate measure.
The leading end of the fistula was exposed, skeletonized, and reduced in size by applying a side biting clamp, excising part of the fistula, and then repairing the excision site with a running suture line (images #8a and 8b). If better control of the flow is desired, a Teflon tube (a graft segment) can be applied to the fistula and tightened gradually while repeated measurements are done to reach the desired flow rate (image #9).
Finally, a repeat retrograde study is done showing the reduction in the inflow segment, and the spontaneous flow past the anastomosis down to the hand (image #10).
Discussion
Arteriovenous shunts (either grafts or fistulas) are created in many patients with renal failure in order to provide the means for dialysis. They do not occur normally in healthy people. A shunt is considered better than a catheter for dialysis in most people due to the lower rates of infection and damage to the veins around the heart. Years of research have demonstrated the superiority of dialysis via a shunt over a catheter, and a survival advantage.
Yet, there are some drawbacks to shunts. Every shunt represents a short-circuit of the circulatory system. Blood that would ordinarily travel down to the hand (or foot) is diverted into the shunt. In most people, there is more than enough blood for both the shunt and the hand. The artery tends to dilate after the surgery to bring more blood down the arm, and the patient usually notices only occasional “pins and needles”, tingling, or coldness of the hand on the same side as the shunt (the “shunt hand”). We call this “steal”. These symptoms tend to get better with time as the body adjusts to the new circulatory pattern. We recommend that all patients keep their shunt hand warm, avoid smoking, and take low-dose aspirin to reduce these problems.
However, some patients have more significant problems. Steal can be painful. It can interfere with sleep. Numbness of the hand may interfere with activities of daily living. Poor circulation may prevent healing of minor injuries that result in infection of the finger and can even lead to loss of a digit. Steal is usually worse in the winter than the summer, and worse on dialysis. Steal causing significant problems can be managed several ways:
With mild steal (occasional numbness or coldness) we generally recommend a wait-and-see approach, keeping the shunt hand warm with a glove, particularly on dialysis and in cold weather, stopping smoking, and taking low-dose aspirin, assuming there is no contra-indication. Generally, things get better with time.
For moderate to severe steal (persistent numbness, weakness of the hand, or pain) a more aggressive approach is generally needed. If warming the hand does not resolve the symptoms, an addition of Trental (a medication to improve the circulation) may be added. Trental can take up to a month to show effect. We may ask for digital pressures (to see how much blood is getting down to the fingers) or other non-invasive studies done in the hospital. Other options may include branch ligation or banding of the fistula or graft (limiting the blood flow to a single channel or placing a collar to slim down the access at the inflow end, thereby forcing more blood down to the hand); a fistulogram with retrograde brachial angiogram to detect and treat narrowings in the artery just above or below the shunt with balloon angioplasty; an upper extremity angiogram, to detect and treat narrowings in all of the arteries from the heart to the fingers; revision with proximalization, to reattach the shunt to a larger artery further up the arm; or a DRIL procedure to rearrange the circulation by bypassing the shunt with a vein from the leg. Finally, ligation of the fistula or graft and returning to a catheter as the long-term dialysis access may be required. The National Kidney Foundation Dialysis Outcome Initiative guidelines state that up to 10% of dialysis patients may not have good enough veins, circulation or strong heart to have a shunt.
Epilogue
Banding is the most straightforward and easiest inflow reduction procedure. Other procedures include inflow fistuloplasty (as done in this case) or aneurysmoplasty, revision of the inflow with proximalization or distalization (moving the inflow source up toward the axilla, or down to a smaller artery), or the DRIL procedure (distal revascularization, interval ligation – not performed often).
