Percutaneous dialysis arteriovenous fistula banding for flow reduction – a case series

Arteriovenous fistulas (AVF) are the preferred method of vascular access for patients on chronic haemodialysis. Common AVF types encountered in practice include the radiocephalic, brachiocephalic and brachiobasilic fistulas (Vascular Access Work G 2006). Occasionally, synthetic polytetrafluoroethylene arteriovenous grafts (AVG) are utilised when there are no suitable veins to form an autologous conduit. Excess shunting through the AVF can result in dialysis-access steal syndrome (DASS) leading to distal ischaemia or high-output cardiac failure when severe. This risk is estimated to be approximately 1% in AVF and between 4 and 5% in AVG (Goel et al. 2006). Symptoms and signs of DASS may include pain, pallor, paraesthesia, weakness and necrosis of the distal extremity and digits in severe cases.

A number of treatment options are available for DASS and traditionally involve surgical procedures such as ligation, distal revascularisation and interval ligation (DRIL) and revision using distal inflow (RUDI) (Gupta et al. 2011). However, procedures such as surgical banding can potentially result in loss of vascular access for haemodialysis all together and the invasive nature of these procedures are associated with a relatively high rate of complications, ranging between 30 and 50% in one surgical series (Leake et al. 2015). Percutaneous endovascular banding, also known as Minimally Invasive Limited Ligation Endoluminal-assisted Revision (MILLER) has been described as a less-invasive fistula reduction technique for DASS (Goel et al. 2006), and in this article, we present a technical review of the procedure and discuss situations where percutaneous banding may have limited success.

We present two cases in which AVF banding was performed for DASS and high-output cardiac failure. Following written informed consent, all procedures were performed under aseptic technique, in a dedicated interventional radiology suite under conscious sedation and local anaesthesia.

In patients with DASS, percutaneous AVF banding or MILLER is a useful and less-invasive alternative compared to surgical ligation where there is the possibility of loss of vascular access, requiring transition to dialysis catheters or pre-emptive creation of a new fistula. The technique for percutaneous AVF banding as described, is relatively straightforward and potentially reversible with angioplasty or stenting, if an absorbable suture is used as in our series (Shukla et al. 2016). Results from larger cohorts show promising results with a primary patency rate of between 75 and 85% at 6 months with primary assisted patency rates of 92% at 1 year (Shukla et al. 2016; Miller et al. 2010). Complications from percutaneous banding are uncommon and mostly relate to fistula thrombosis or need for repeat banding to achieve adequate clinical success (Shukla et al. 2016; Miller et al. 2010). Comparatively, surgical procedures including banding, DRIL and RUDI also demonstrate high fistula preservation rates between 89 and 100%, highest for DRIL procedures, but these were associated with a higher risk of early 30-day complications, particularly for surgical banding where there is an 11% risk of thrombosis and 2.8% infection risk in the early postoperative period (Leake et al. 2015).

The degree of juxtaanastomotic venous outflow reduction is determined by the luminal diameter reduction and is controlled by the diameter of the angioplasty balloon used. Miller and other authors aimed for a band size slightly less than the size of the downstream artery on the pre-reduction angiogram. We used a similar but slightly more conservative estimate in our cases, aiming for a balloon size under half of the juxtaanastomotic venous diameter as the relationship between the band stenosis and arterial diameter can change following redistribution of flow. The risks of AVF thrombosis is higher with greater reduction of flow and in our opinion, it is easier to repeat the banding procedure than salvage a thrombosed fistula. In the first case, the selection of a 5 mm balloon relative to a 13 mm segment resulted in an approximate angiographic stenosis of 75% (Fig. 1i). Banding can also be guided by intraprocedural flow measurements but these require additional flow transducer devices for measurement. Banding to a predetermined luminal reduction of approximately 75% appears to be safe, equally effective and reduces the need for additional costly equipment in most cases (Shukla et al. 2016).

Successful percutaneous AVF banding depends on the exclusion of other confounding causes for DASS symptoms, particularly arterial insufficiency as illustrated in the second case. Often, it can be difficult to confidently exclude distal arterial disease on non-invasive assessment because of the excess shunting and resultant poor distal perfusion prior to banding due to the small calibre of the distal palmar vessels and difficulties with optimal contrast timing on CT angiography. Therefore, upper limb digital subtraction angiography with treatment of underlying arterial stenoses prior to banding is essential for clinical success if there is clinical concern for associated arterial disease (Shukla et al. 2016). In particular, revascularisation strategies should aim to restore adequate inflow to the palmar arch and digits, particularly where tissue loss is of concern.

Percutaneous AVF banding is an effective and safe minimally-invasive technique for treatment of ischaemic DASS with good intermediate-term patency rates, comparable to more invasive surgical procedures. However, there is a need for further studies to assess long-term outcomes including late thrombosis rates.