Evaluation of Postoperative Fistula: Controversies, Tips and Pitfalls - SD

Introduction

My name is Mark Lockhart.
I am associate professor of radiology at the University of Alabama at Birmingham.
My main focus of research and interest are vascular ultrasound, hemodialysis imaging, and gen urinary radiology.

Welcome for the next 30 minutes we'll be discussing the evaluation of the postoperative hemodialysis fistula.
We'll discuss some of the controversies, certain tips for enhancing the examination and potential pitfalls to evaluate for.

I have no disclosures related to this presentation, and I'd like to thank my co-investigators as listed below.

Reasons for Evaluation

The reason that we do these evaluations is because of the emphasis that has been given to fistula improvement in a national campaign by the NIH and Medicare.

A fistula first program has been instituted in this program.
There is additional emphasis on the creation of hemodialysis fistulas as opposed to hemodialysis grafts for hemodialysis.

The reason that this has been done is there is literature which suggests that once a fistula is usable, fistulas survive longer with fewer complications than grafts, and also fewer interventions are necessary to maintain a patent fistula.

However, despite this added emphasis, still only 40% of US patients dialyze with a hemodialysis fistula.

Reasons for Fistula Failure

So what are some of the reasons that the fistulas will tend to fail?

There's increased failure rates in female patients, blacks, the elderly, patients who are obese, which limit access to the vessels, and then also patients with vascular comorbidities.

Types of Fistulas and Anatomic Variations

When we're looking at the different types of fistulas, first of all, I think we should go over the different anatomic variations.

A brushes amino fistula is a anastomotic communication between the radial artery and the cephalic vein.
There can be an upper arm cephalic vein fistula, which communicates between the brachial artery and the cephalic vein, and then you can have an upper arm basilic vein transposition, and I'll go into this in more detail.

The first example that is shown is a forearm cephalic vein fistula, which takes an inside anastomosis of the cephalic vein at the wrist as shown into the radial artery.
This allows for vascular access within the forearm draining vein once it's completed.

An upper arm cephalic vein fistula is an inside anastomosis between the cephalic vein or a venous branch, such as the medial cubital vein into the brachial artery, slightly above the antecubital fossa.

A basilic vein transposition is a bit larger procedure where the multiple branches from the basilic vein are ligated, and then the basilic vein is freed up and tunneled beneath the skin more laterally.
This is because the basilic vein typically is very medial, and this location is difficult for hemodialysis.

So what will happen is the basilic vein will be transposed laterally and superficial, and then an end to side anastomosis much like the other two types is performed.

Postoperative Evaluation Techniques

When we're evaluating these after the surgery, you need to understand some of the different techniques that are used to optimize the evaluation.

The patient should be either sitting or slightly reclined and in a comfortable position, their arm should be laid out on an instrument stand.
Usually with the palm up we use a large amount of ultrasound gel, almost as a standoff, so that we are not putting undue pressure on the draining vein.

In the initial overview, we will look at the location of the vessels, we'll evaluate the feeding artery and the draining vein.
It makes it much easier to do this study if you have the surgical note so that you understand what vessels were anastos to create the fistula.

And you'll notice many of my notes about the tips are in yellow.
And you'll notice for a long exam such as this, maybe up to 30 minutes, you really should focus on having the patient in a comfortable position for the post-op ultrasound technique.

We begin with gray scale imaging in the transverse and longitudinal planes.
We will then follow with color doppler ultrasound to evaluate for aliasing in any area of stenosis.
There may be turbulent or increased velocity flow, and this will often wrap around into different colors, creating a disorganized or turbulent appearance.

Any area that we find of stenosis will then be evaluated with spectral doppler using peak systolic velocity in diastolic velocity, and measuring a ratio of the peak velocity to an area of inflow upstream approximately two centimeters from the stenosis.

For all of our velocity measurements, you should maintain a Doppler angle of less than 60 degrees.
To limit the artifact of variability, we will routinely evaluate the anastomosis and any areas of concern for the evaluation.

The frequency of the transducer is important.
We like to use as high frequency as transducer as we can due to the high spatial resolution in terms of looking for stenosis or for thickening.
However, you must realize that high frequency transducers may not allow detection of high flow velocities, and you may get more aliasing.

Controversies in Postoperative Fistula Evaluation

So when we're talking about the evaluation of the post-op fistula, there are several areas which are a bit controversial that we'll discuss.

First of all, size is not enough for hemodialysis fistulas.
There are many other factors that contribute to whether it's a usable or a fail fistula.

There's lack of standardization in some of the thresholds and criteria that are used.
Also, most of the evaluation that we do is a gross anatomic evaluation, whereas there are histologic characteristics which may have an effect.

And then we'll briefly talk about whether routine screening of fistulas in asymptomatic patients is acceptable.

Size Is Not Enough

So first of all, discussing whether size is not enough.

It's been shown that for fistulas that are created, if you look at the vessels prior to the surgery, there's no difference in size of the vessels.
When you're comparing the mature to the immature fistulas, this is once you reach the size threshold of 2.5 millimeters in the vein, or two millimeters in the artery, as has been previously shown.

So even though these criteria may be met for the size, 50% of fistulas may fail to mature.
This is due to a number of issues such as stenosis, branches of the draining vein, whether or not the artery is able to deliver adequate blood flow to allow the sheer forces to dilate up the vein.
And then whether or not there are other factors that may prevent the vein from adequately dilating.

Postoperative Ultrasound Criteria for Sonographic Maturity

When we're looking at the postoperative ultrasound criteria for sonographic maturity, minimum vein diameter is one of the criteria.

So size is not enough, but it is still important.
If the draining vein is greater than four millimeters diameter, these patients were adequate for dialysis in 89% of cases when they were less than four millimeters, only 44% were graphically or clinically mature.

We also look at the volume blood flow of the fistulas, and if the volume blood flow is greater than 500 milliliters per minute, fistulas were adequate in 84%, whereas if less than 500, less than half of these were actually clinically adequate.

Be sure and look for the reason.
If you see a minimum blood flow less than 500, so if you are looking and the flow is low, but there's no stenosis, you need to look closer a, to find a stenosis or b to look for something that's limiting the outflow, such as a central venous stenosis.

Now, what happens when both the diameter and the blood flow are considered together?
If both criteria are positive, then 95% of these will be adequate for dialysis.
Whereas if neither criteria are met, so the diameter is less than four millimeters and the velocity is less than 400 milliliters per minute, then only 33% were adequate for dialysis.

Whereas these criteria were based on a research study.
There are differing criteria from the National Kidney Foundation or the K dokey guidelines, which basically have a six millimeter diameter, six millimeter depth, and 600 milliliter per minute thresholds.

This is different from our four millimeter diameter, five millimeter depth and 500 liter per minute thresholds.
To our knowledge, these criteria were basically developed on a consensus panel and are not actually shown by research data.

If we were to use these criteria to our patient population, nearly one third of our successful fistulas would not have qualified for fistula formation by their guidelines.

Imaging Findings: Normal Mature Fistula

So now we'll switch over to try to look at some of the imaging findings of a normal mature fistula and compare that to some abnormalities in the artery, which is two centimeters cranial to the anastomosis.

In this case, we see arterial wave forms with peak systolic velocity of 215 centimeters per second.
We'll use this to compare to the anastomosis, which is 260 centimeters per second, so there's no gradient between the two.

Notice that the artery has a nice black window showing that this is fairly laminar flow without turbulence or aliasing.

The draining vein at the anastomosis has a slightly small diameter, 2.7 millimeters, but as you get into the draining vein in the cau a forearm, it rapidly dilates up to a usable fistula diameter of greater than four millimeters, in this case 7.4 millimeters.

Also notice that the depth to the skin is very small, less than five millimeters deep.

Looking at the volumetric flow, we are over 500 millimeters per minute as shown here, 524.

So this would be something that we would call a sono graphically mature fistula.

Always we try to look at the artery distal to the anastomosis to see whether or not there is still flow going to the hand.
And in this case, you'll notice even though we are below the anastomotic site, the flow is coming toward the transducer, which is toward the patient's head.

So there's reversal of the distal arterial flow in this example.

Immature Fistula

Now, comparing this patient to an immature fistula, the diameter of this fistula is 2.9 millimeters, so less than the four millimeter cutoff.
It is not too deep.
But notice when we look at the volumetric flow, we only have a volume flow of 267 millimeters per minute.

So when I try to show you that we can call a fistula mature or immature by ultrasound, the next question should come up.
Well, how does this correlate to the actual clinical access of the fistula?
If we are saying it's sonographic mature, does this predict whether or not the patient's fistula can be used for dialysis?

Correlation to Clinical Access

So to this end, we evaluated 95 patients who on clinical palpation by a trained nurse, were deemed to be clinically immature.
We went to see whether or not ultrasound would be of benefit in these patients.

In this group, 29%, so nearly a third actually were mature by sonographic criteria, and in these patients, 91% were subsequently usable for dialysis by the hemodialysis nurses.

In the 71% that were sonographic immature patients who refused to undergo salvage only had a 31% usable rate, whereas the patients who underwent salvage of an abnormality such as correction of stenosis or ligation of competing veins, 78% eventually were able to dilate was the fistula.

Abnormalities in Clinically Immature Patients

So what was the variety of abnormalities that we saw in these patients that were clinically immature?
And then also graphically immature stenosis was present in about a third of patients, and then nearly a half or approximately a half had a competing branch within the first 10 centimeters, and then about a third were too deep, which can be helped with superficial.

If you notice, these numbers do not add up to a hundred percent because some patients had multiple lesions.

Stenosis

When we're evaluating the different types of abnormalities, stenosis is the one that is most commonly brought to mind for stenosis.

The anastomosis is the most common site that it occurs.
We've shown previously that if there is a ratio of the anastomosis to the feeding artery, two centimeters approximately a ratio greater than 3.0 suggests a 50% stenosis.

So you need to really pay close attention to the anastomotic region, and this includes two to four centimeters of the feeding artery and the first two to four centimeters of the draining vein.
This is the sort of most common area to look for stenosis.

Now, if you find a stenosis, don't panic.
You still need to look at the rest of the exam because there could be multiple lesions.
And if this patient's going for a salvage procedure, they would want to correct them all at the same time.

This is an example of a fistula with an anastomotic stenosis.
Notice that there is turbulent disorganized flow in the anastomotic region.

When you look at the feeding artery two centimeters from the anastomosis, we have a peak systolic velocity of 123 at the anastomosis.
Notice this very turbulent flow loss of the black window underneath the waveform and a peak systolic velocity of 593 centimeters per second.

So this is greater than a three to one ratio, and this would be suggestive of anastomotic stenosis.

When we move to the draining vein, we do not use the three to one ratio for 50%, but instead the criteria have suggested that a two to one ratio is consistent with a 50% stenosis.
For a three to one ratio, it was more suggestive of a severe stenosis greater than 75%.

Draining Vein Stenosis

When you see dilated veins in the draining vein, look to see whether they're very tortuous, look to see whether there's a good flow volume.

The reason being, if you have a draining vein stenosis, the veins will often try to bypass this stenosis through collateral vessels.
So you could be seeing a high pressure collateral trying to bypass the stenosis, and you'd want to try to pick this up.

This is an example of a draining vein stenosis.
Again, the anastomotic region looks pretty good, 51 centimeters per second.

As we look at the anastomosis, there is some turbulence in aliasing, but no gross significant stenosis.

Then we look at the artery, two centimeters cranial to the anastomosis, and this really has no ratio between the feeding artery and the anastomosis.

However, as we start following the vessel out, and this is in a longitudinal view, five centimeters from the anastomosis, we start seeing thickening and irregularity of the draining vein.

So what would we do next?
We'd put color on, we see that the vein is patent, however, there is some narrowing and irregularity.

And when we put a velocity gate on this region, the peak systolic velocity is over 200 centimeters per second.

So what would we do next?
We would look two centimeters upstream from this level, and we see that the peak systolic velocity is only 40.

So this gives us a five to one ratio.
So this is easily consistent with a draining vein stenosis.

And in this case, just to confirm, we give the volumetric flow measurement, which shows a volume flow of 177 milliliters per minute.

So again, an additional criteria that this is not clinically or sonographic mature.

Sometimes we like a, when that's a long segment stenosis such as this, we'll use an extended field of view just to show the length of the abnormality, which I think shows nicely in this image.

Draining Vein Occlusion

Now, draining vein stenosis should be differentiated from draining vein occlusion as we see here.

When you have echogenic material filling the draining vein, this is a good hint.
You do compression, it won't compress, and then if you put a spectral gate on, you may get some artifact, but really no adequate flow.

And then this is just a nice video clip showing the puls utility of the clot as it's being sort of pushed upon by the vascular flow.

Central Venous Issues

When you see draining vein thrombosis or when you see anything that causes a low volumetric flow, if you haven't identified a proximal abnormality, meaning an inflow or anastomotic or proximal draining vein stenosis or thrombosis, then you should start moving farther along the flow toward the central venous system.

Typically, we will follow the draining vein up the arm and go to the subclavian vein and then try to infer central flow of the brachycephalic veins with our spectral doppler In an arterial graft.

The large volumes of flow that may be there may overwhelm a the central venous ity.
However, in hemodialysis fistulas you would typically expect even with the increased flow that the venous wave form here should go back to the baseline, and sometimes we'll have transient reversal of flow.

In this case, you can say this is very monophasic flow worrisome for central occlusion or stenosis.

We then look at the left subclavian vein and a very similar appearance.
Monophasic flow, no return to the baseline.
This is worrisome for a central occlusion.

If you have the ability to do A MRV, then you can try to evaluate the central veins.

This has become more problematic in recent years because the diagnosis of nephrogenic systemic fibrosis in patients who receive MRI gadolinium contrast with renal failure.

Since most of these hemodialysis patients are in renal failure, there's a likelihood that these patients would not be able to get a gadolinium enhanced MRV in these current days.

But you can see here on this image at least, there is a long segment occlusion of the left brachiocephalic vein and also no flow is visualized in the right cephalic brachiocephalic vein.

When we find something in the central veins, which is an abnormality, whether it's a stenosis or whether it's an occlusion, there have been attempts to try to repair these to improve the blood flow.

The two main options have been stenting of the stenosis or angioplasty with or without stenting.

The problem we run into is both of these groups will tend to have multiple subsequent interventions necessary to maintain the patency.

These slow flow vessels just tend to stenosis or thrombose.
And also with these two techniques, there's really no difference in the primary or secondary patency rates that have been shown for the fistulas once these therapies have been performed.

Large Competing Veins

Now moving away from stenosis or occlusion, we're going to change gears and look at large competing veins.

These are most important in the first 10 centimeters.
So as you come from the anastomosis into that early draining vein, you need to look whether there are large branch vessels which are drawing flow away from the main vein.

This is because the vein needs an adequate amount of flow to create a shear effect to allow the vein to dilate, pay special attention to the codal portion of the arm, or sort of in the stick zone where you would be placing your hemodialysis needles.

Deep Draining Vein

So moving from a large competing vein, we're gonna look at the third abnormality that may occur.
In this case, the draining vein is too deep for consistent vascular access.

Although the diameter of the vein is eight millimeters, which is well above our four millimeter threshold, it is greater than the five millimeter depth that is easily accessible.

Again, this is most important within the area where the hemodialysis needles will be placed.

Just another example in the cephalic vein in the forearm where we have a small draining vein.
But look how deep this is over six millimeters deep, I think most nurses would have a difficult time trying to access this vein.

Salvage Procedures for Abnormalities

So when we have one of these three abnormalities, whether it's a stenosis, a large competing vein or a vein that's too deep, often there are attempts to try to correct the abnormality to improve the likelihood of successful dialysis.

Notice that the AV fistula adequacy rate is typically lower in women.
They have smaller veins and only about 31% are initially successful.

Also, the success rate is worse in the forearm, which also has smaller veins because women start off with a lower success rate.

Salvage procedures actually work better for them.
They may cause improvement in 50% of women, whereas men improve in about 37% notice.

The final overall success rate ends up equaling out fairly closely between men and women.

So when we detect an abnormality of an immature fistula, the salvage success depends on the underlying lesion that's causing it to fail.

The likelihood of fistula ma maturation depends on whether or not you use a salvage procedure.

If there's a stenosis which is not treated, 0% will go on to be clinically usable.
Whereas if the stenosis is treated, 70% can go on to mature.

If there's a large competing vein branch, 83% will subsequently be usable if corrected.

And then if there's a superficial for a deep draining vein, 71% of these will eventually go on to have a useful hemodialysis fistula.

Example of Successful Salvage: Ligation of Competing Veins

This is an example of a success story of a large competing veins that were fixed.

The initial post-op images show a draining vein, two centimeters from the anastomosis, a larger one at four centimeters from the anastomosis, and another one at five centimeters from the anastomosis.

Notice in this case, the volumetric flow is only 364 milliliters per minute.

Now, after ligation of the competing veins, notice how the draining vein has dilated.
It's now 6.3 millimeters diameter and we now have a volumetric flow well over 500 milliliters per minute.

Routine Screening of Asymptomatic Patients

Our last topic that I'd like to discuss is the routine screening of the asymptomatic patient with a hemodialysis fistula.

Several studies have been performed where they do frequent repeated ultrasounds to detect abnormalities in an effort to try to pick up an abnormality before it led to thrombosis.

The meta-analysis included 12 different studies.
Four of these were fistula trials.

The screening in these trials did decrease access thrombosis, so less than half of the typical thrombosis rate.

However, even though the thrombosis rate was decreased, it did not reduce the fistula loss.

So this suggests that the thrombosis may be a marker of the fistula loss as opposed to an underlying etiology.

In the eight graft trials that were included in the analysis routine screening did not reduce thrombosis rates or graft loss rates.

Steal Syndrome

Sometimes when you have a very successful fistula, there may be a sumping of flow of the distal artery.

As I showed at the very beginning of the talk, the reversal of flow is taking the blood flow away from the hand or the forearm.

Usually these are asymptomatic.
However, if there are symptoms, there should be a modified sonographic Alan's test where that the flow from the ulnar artery and the radial artery are checked to make sure that there is a patent Palmer arch with collateral supply to the thumb.

If the arch is insufficient and if they're having symptoms, occasionally revision or even sacrifice of the fistula may be necessary.

This is just an example of an artery with a reversal of arterial flow beyond the anastomosis.

Notice we are above the baseline and above the baseline flow is headed toward the patient's head in a reversed fashion.

Pitfalls in Evaluation

Some of the pitfalls that you need to be aware of, I think that it's become pretty standard.

Use that ultrasound prior to fistula creation should be used.
It's been shown that if you use ultrasound prior to fistula creation, there is a higher success rate.

Inadequate sonographer training is another problem, which may limit your ability to do these studies.

Once a stenographer has been trained, the training doesn't really end there.
There should be constant quality control of the stenographers and of the studies that are being performed.

Other limitations which may affect this technique are variability among the physicians in your group.
And then also just the inherent intra observer variability with the studies.

One study that looked at the inter observer variability looked at 24 accesses in successive sessions, and they found out that even within the same session, the variability was 11.6%.

This is larger than the variability for the dilution method, which is 7.7%.

They noted that there was especially high variation if the needle orientation was altered between the sessions.
For the dilution technique.

Microscopic Abnormalities

Moving away from the macroscopic abnormalities that we detect with our typical ultrasound exams, we're going to consider a couple of the more microscopic abnormalities.

Intimal changes in fistulas have been shown.
It's been thought to be due to the increased flow through the fistula.

However, there is some recent research that suggests that there may be undetected intimal disease prior to the fistula creation.

So these may be patients that are predisposed even before they have the fistula placed to failure.

Other questions rise whether the arterial inflow or the venous disease is the most important component to the failure rates.

And then when we find hyperplasia of the artery, it hasn't really been shown whether hyperplasia versus actual plaques and calcifications have differing effects on the outcomes.

Intimal hyperplasia has been associated with stenosis formation.

And then there's also preliminary data that hyperplasia itself is associated with fistula failure.

Once we go beyond the radiologist and the imaging, the clinicians have tried to see whether or not they could use anti-platelet therapy thinning the blood to try to prevent thrombosis in these patients.

In a randomized trial of 877 patients, there was reduced frequency of early thrombosis in those with the anti-platelet therapy.

However, even though the thrombosis rate went down, the proportion of usable fistulas did not significantly change.

There. Again, this is supporting the fact or the idea that early thrombosis may be a manifestation of the failure rather than an underlying etiology.

Even farther away from radiology, we should look at potential for genetic markers.

If hyperplasia is an influencing factor, then we should look at whether or not there are genetic markers for the hyperplasia such as listed below to help predict which patients may have a worse outcome.

Summary

So in summary, hemodialysis ultrasound is very useful for the monitoring of patients with abnormal hemodialysis fistula maturation.

We can detect stenosis competing veins and draining veins that are too deep, whereas screening alone may not be very successful.

The early identification of abnormalities suggests that we can improve these fistulas with early salvage.

Thank you for your attention.