Preoperative Radiological Assessment for Vascular Access

1. Introduction 

In many centres patients presenting for primary access proceed directly to arteriovenous fistula (AVF) formation after careful physical examination and if there is a satisfactory radial pulse and suitable forearm veins there is a good chance of success. Clinical guidelines provided by NKF-DOQI in 1997 state that imaging is only necessary in certain patients. Venography is indicated where there is a suspicion of central vein stenosis or trauma, and in patients with multiple previous access attempts.1 Ultrasound or MRI is suggested in certain complex cases or in predialysis patients with minimal residual renal function where there is a risk of iodinated contrast causing acute renal failure. Rarely, angiography or arterial duplex are indicated when arterial pulses in the access location are diminished.

Despite these guidelines up to one third of access procedures will either fail or not mature well enough to be useful for dialysis. More widespread use of preoperative imaging could reduce the rate of primary fistula failure and reduce unnecessary surgery. Traditional contrast venography, which for many years has been the gold standard, has several limitations including relative invasiveness, risk of allergic reaction, contrast nephropathy and cost. These drawbacks prevent its generalised introduction. More recently, there is increasing evidence that preoperative duplex ultrasound cannot only increase the utilisation of native AVF for dialysis access, but also allow proper selection of a target vessel with adequate luminal diameter to improve dialysis access outcome.2 Ultrasound is being increasingly utilised in many centres, as it is readily available and cheap but its major limitation the relative inability to assess central vein patency so that contrast venography may still be needed if a central venous stenosis or occlusion is suspected. In those patients in whom iodinated contrast is contra-indicated, magnetic resonance venography (MRV) and carbon dioxide venography have recently become available as alternatives. MRV can also give excellent anatomical detail of peripheral arm veins although its relative lack of availability and high cost will probably prevent its widespread introduction.

This article briefly describes the current literature on preoperative imaging prior to vascular access formation. Duplex ultrasound is rapidly emerging as an essential investigation and its widespread introduction in the future will have major resource implications.

2. Duplex Ultrasound Scanning 

2.2. Venous study 

This can be performed at the same time as the arterial study. The arm is scanned proximal to distal, with and without a tourniquet. If there is a significant proximal arterial or venous narrowing, or abnormality, which would jeopardise the success of the AVF, the examination is discontinued. Veins should be thin walled, vary in size with respiration, collapse completely on compression with the transducer, and augment with distal compression. Essential parameters which are measured include vessel depth, internal diameter with and without the tourniquet, continuity with the deep system and the presence of any stenosis or thrombosis. Veins should dilate by approximately 50% by the use of a tourniquet. There should be respiratory variation in the subclavian vein and, if indicated, the venous flow rate can be recorded during unforced inspiration. Recordings of depth and diameter (with and without a tourniquet) should be recorded on a chart, which can be taken with the patient to theatre (Fig. 1).

View Large Image Download to PowerPoint

Fig. 1. A suitable chart for recording readings during a preoperative venous duplex examination.

2.3. Arterial study 

The arm is again scanned proximal to distal. Internal diameters are recorded at different levels, as well as calcification, and abnormal arterial wall thickening. Recordings are made on another suitable chart (Fig. 2).

View Large Image Download to PowerPoint

Fig. 2. A suitable chart for recording readings during a preoperative arterial duplex examination.

The arterial waveform should also be evaluated both proximally and distally and, in particular, at the site of the proposed AVF. Peak systolic velocities should be recorded. The normal waveform should be triphasic high resistance flow with no evidence of dampening which could indicate a proximal stenosis. Following AVF construction blood flow through the feeding artery is increased and the peripheral resistance is decreased because of diversion to the low resistance venous circulation. The Doppler waveform is changed from a triphasic high resistance waveform, to biphasic low resistance waveform with increased diastolic flow. Reactive hyperaemia (RH) simulates the decreased vascular resistance after AVF construction and the Doppler waveform during RH can be used as a test for the ability of an artery to sustain the increased flow3 (Fig. 3). RH can be induced by opening a fist, which has been clenched for 2min. The change in the waveform can be recorded by measuring the resistive index (RI) as well as changes in peak systolic and end diastolic velocity.

View Large Image Download to PowerPoint

Fig. 3. Radial artery waveform during clenched fist (A) and on release showing a hyperaemic response (B).

3. Venography 

For many years contrast venography has been the gold standard imaging investigation prior to AVF formation. It has the particular advantage of providing a venous map, which many surgeons find useful. According to DOQI1 venography is mandatory in patients with a history of ipsilateral central vein catheterisation, collateral vein development, oedema or differential extremity size as these findings may indicate inadequate venous drainage or central vein obstruction. Construction of a peripheral fistula in these patients can cause massive arm swelling and poor dialysis. Until recently temporary dialysis catheters were frequently placed in the subclavian veins causing a high incidence of venous stenosis near the catheter insertion site. Surratt17 found moderate or severe subclavian vein stenosis in as many as 40% of patients with a history of a prior or existing line (Fig. 4).

View Large Image Download to PowerPoint

Fig. 4. A conventional venogram showing a tight subclavian origin stenosis (arrow).

Today best practice is to establish an AVF before commencing dialysis and if this is not possible insert a tunnelled right internal jugular line. This has reduced the incidence of central vein stenosis although innominate vein stenosis remains a significant problem.

Venography with iodinated contrast is relatively contraindicated in predialysis patients as contrast nephropathy may precipitate acute renal failure. Duplex ultrasound is an alternative but is poor at assessing central vein stenosis. Contrast enhanced MR venography has emerged as a promising alternative as the small volume of paramagnetic contrast agent used does not compromise renal function. A similar technique to conventional venography is used with contrast injected upstream of the venous territory to be evaluated (Fig. 5).

View Large Image Download to PowerPoint

Fig. 5. A contrast enhanced MR venogram of the right upper arm.

To avoid T2 shortening effects the paramagnetic contrast should be diluted by a factor in the range 1:10–20.18 Non-contrast MRI using ‘Time of Flight (TOF)’ techniques have also been evaluated for preoperative venous mapping as there is also no risk of compromising renal function. Patients are examined prone with the arm under investigation inserted into a surface coil. Laisey et al.19 have demonstrated the technique can be used successfully for venous mapping of the distal upper arm, elbow, and forearm with veins larger than 2mm visualised. The main limitation of non-contrast techniques is their inability to demonstrate proximal or central veins.

Menegazzo et al.20 prospectively compared TOF MR venography and conventional venography in 24 patients prior to fistula formation. MR and venographic results were well correlated for demonstration of superficial veins and assessment of venous diameter. The correlation between MR venographic and surgical findings (κ=0.78) was superior to that between conventional venographic and surgical findings (κ=0.56). These results are encouraging but much more data is needed before MR venographic techniques can be recommended for routine preoperative evaluation.

Two further venographic techniques are available for the assessment of predialysis patients without compromising renal function. Carbon dioxide (CO2) venography has been used successfully to visualise both peripheral and central veins but has significant drawbacks including local pain during injection, overestimation of venous stenoses and the risk of potentially serious complications for example acute right heart failure.21 A costly special injector may also be required. Despite these drawbacks CO2 venography is widely used in several centres, particularly in France.22

MRI contrast agents have also been evaluated using conventional X-ray digital subtraction techniques rather than magnetic resonance. Gadolinium based contrast agents have shown they absorb sufficient energy to be visualised during angiography by digital subtraction and have been used for several years for angiographic interventions where conventional iodine based contrast is contra-indicated. Gadoterate meglumine (Gd-DOTA) has been evaluated in 45 predialysis patients with renal insufficiency undergoing venography because of inconclusive physical examination prior to AVF formation.23 Three sequences were performed on then forearm, arm and chest at 3ml/s using a total of 35ml of Gd DOTA. There was satisfactory visualisation of the forearm and arm veins allowing the identification of 33 forearm cephalic veins, which would allow the creation of an AVF. However, the innominate veins and SVC were relatively poorly visualised which may limit the introduction of the technique.

4. Summary 

•There is increasing evidence that routine preoperative duplex ultrasound reduces the rate of primary fistula failure and unnecessary surgical exploration.

•Ultrasound increases the utilisation of native AVF and allows proper selection of suitable target vessels of adequate diameter.

•A minimum arterial diameter of 2mm is associated with successful fistula formation. Below this diameter the ability of an artery to increase flow and dilate will determine fistula success.

•A threshold for minimal venous diameter is difficult to establish. Most clinical studies use a value of 2.5mm for AVF and 4mm for prosthetic grafts.

•Traditional contrast venography is still an essential investigation if there is suspicion of central vein stenosis. In predialysis patients MR venography is emerging as a suitable alternative.