Contrast Enhanced Ultrasound: The Hammersmith Experience - SD

Introduction

I am David Cosgrove.

I work at Hammersmith Hospital in London.

My privilege and pleasure to present to you a talk on liver contrast ultrasound.

Essentially this is the experience of our hospital Hammersmith Hospital.

History and Routine Use of Contrast Agents

Our story of working with contrast agents actually goes back quite a long way to 1996 and onwards, where we started with phase three trials of Leave Vista.

We were fortunate enough with so view, which is the main agent that we use now throughout Europe, to be involved in the phase one trials, which means normal subjects, as well as the phase two and phase three trials.

That began in 1997.

Since then we've worked with several other agents.

Importantly now, contrast studies are completely routine in our department, and we use them on a daily basis for problem solving.

Hospital is quite complex with three units within it, Hammersmith, where I mainly work chairing cross in St. Mary's.

Overall we do about 20 contrast studies per week, excluding research studies.

Most of these are done in dedicated sessions.

We dedicate certain half days each week to contrast studies and try to pull patients into that just for logistical reasons.

It's easier to know that you're expecting to do contrast studies and have the right equipment and the right gear, and possibly an additional person to help with the injections, but some are done as need be.

This is particularly important for the situation where you discover an incidental lesion that you weren't expecting to find, which happens quite commonly.

It's very nice to be able to do a contrast study, solve the problem immediately then and there, rather than the other scenario, which would involve telling the patient that they've got something wrong, but we don't know what it is, leaving them anxious and having to wait perhaps a week or maybe longer for a CT or MR scan appointment.

The ability to fit a patient in, as it were and do a contrast study as the need arises is very important and helpful for us.

That's the pattern that we are working in.

Focus on Liver Applications

In general, contrast ultrasound is most useful in the heart and the liver, and I'm going to focus as per per my title on the liver for today.

There are really two families of problems that we study here, focal and diffuse liver diseases.

We'll take them in that order.

For focal liver lesions, there are three aspects that we need to look at, which are part of our daily practice staging of patients with a known malignancy characterization of liver lesions found on any imaging modality.

Thirdly, monitoring response to treatment either by drug treatment or by interstitial ablation therapy.

Contrast Agent Phases in the Liver

It's important to understand that for the liver, contrast agents offer us three different phases.

The first two are pretty obvious, the arterial and portal phases.

Here are the times written up.

The arterial happens really very quickly, and then disappears over 15 or 20 seconds, when the portal phase takes over.

This interesting sinusoidal phase goes under a variety of names.

Some people call it the late phase, some people call it the Cook for cell phase.

This lasts for several minutes after the injection's been given and long after the contrast has cleared from the blood pool.

This is in a sense a kind of targeted imaging, but one that happened by chance rather than by design.

Many agents now are targeted specifically to different metabolic processes, and we were just lucky that the agents that we use and all of them do this including sono view that we use, they linger in the liver.

The exact mechanism is subject of some debate.

Some people, some agents are phagocytose by the reticular endothelial system, the cofa cells within the the liver sinusoids, probably that isn't the case with sono view.

It probably just dwells because of the very large vascular space that the liver sinusoids offer.

So it's probably almost a blood volume e effect.

But it is crucial.

I'll show you first of all an example of a contrast study of a normal liver and kidney to show her the contrast arrives in the kidney and the hepatic arteries at about the same time.

Notice the clock there just gone 15 seconds when they arrived, and now the liver and the kidney are of about the same echogenicity as the contrast is settling within the liver.

Then if I flip to an image at two minutes odd, you see that the liver has retained the contrast, but the kidney has now washed out along with the liver va vasculature here and here.

This long late phase in the liver is very important for us.

I'll explain a little bit more about why that's so important.

Staging of the Liver

Now let's look at the role of contrast in staging of the liver.

We're after here mainly detecting metastases, and we use the sin sinusoidal phase, and we have an option of to work with a high MI or a low mi.

Initially, it was the high MI that we used, and I'll show you some examples of that.

The basic principle of the late liver phase is that malignant tumors do not have the normal sinusoidal structure and function, and so they do not retain contrast like the normal liver around it does.

By a comparison, most benign solid lesions do retain that sinusoidal function.

They do show this late phase, of course fluid filled lesions like cysts, abscesses, and so on don't have the sinusoidal structure.

Just a very few solid lesions.

We recently encountered a case of sarcoid with liver granulomas that also showed as a defects in the late phase.

Here's an example of what we're talking about, a liver in a patient with a known malignancy, and these are the metastases showing as punched out dark areas against the enhanced background of the liver.

This is the catchphrase to remember in this late phase for staging, black is bad.

It's really as simple as that.

We are looking for dark areas within the liver.

In practice what we do is scan at three or four minutes.

Whether you need to look in the arterial phase as well depends on the particular problem that you're looking at.

You have to sweep and look back at your images, particularly if you're working with a high MI over 0.4.

Because this is a destructive technique.

You only get one chance at looking in one location.

There are a number of ways to use the high MI methods, the simplest being just to use color doppler in a conventional way.

The lower MI methods, which have been developed more recently, are all multi pulse methods using different ways to alter a pattern of the pulses.

Typically pulse inversion or amplitude changes.

These work at very low mis and we'll look at some examples of that in a minute.

The high MI methods with this rather arbitrary threshold of 0.4 work best with the old fashioned air based agents.

They are destructive, so they're not a real time technique.

Intermittent imaging is required, but it does have the advantage of being exquisitely sensitive, and it's believed that even a single microbubble can be detected by this technique.

It's very very sensitive indeed.

Let's look at some examples of that.

Here's a patient with a fairly obvious echogenic lesion that's suspicious for a metastasis here.

We've simply turned on the C doppler.

We've used a very high PRF and because we're not really looking at movement, and you see that the lesion does not pick up the contrast in this late phase, whereas the liver itself does, but you'll notice extra defects there and there and there, which were not seen on the gray scale scan.

As well as the obvious metastasis, there are other metastases that were not visible on the gray scale, which are shown up by this technique by comparison, here's a patient with a benign lesion.

It's actually a little difficult to see, but it's in this region here.

It's a focal nodular hyperplasia.

Here's what it looks like on the high mi late phase scan.

This is at one minute and 40 seconds.

You can see that the lesion itself has picked up contrast much in the same way as the surrounding liver leaving behind the classical central scar that's seen in about 25% of FNH lesions.

Here are some results of a study that we performed looking at the ROC curves, where the further up towards this top left hand corner the curve goes, the better the result.

You can see that the ultrasound using one of these high MI modes with contrast is much better than the diamonds, which show the unenhanced ultrasound results.

That's result is supported in another study that we were involved in.

This looks at the additional lesions seen over and above conventional ultrasound when contrast ultrasound was used.

In this study, CT contrast CT, of course, was used as the reference, and in 30% of patients additional lesions were seen.

The white bars indicate those which were seen on the contrast CT and the pink ones, those which were not.

Now everybody thinks that a gold standard are references, they be all an end all.

Therefore at the time, and I can't really argue otherwise, at the time, these lesions were classified as being false positives.

But as a matter of fact, it's quite possible that there were real lesions that were simply missed on ct.

Current experience suggests that that is the case.

In a few of these patients, we had an additional standard, for example, Mr. Intraoperative ultrasound or histology from biopsies.

The gray bar here indicates the effect of having the this additional study.

You can see that in this the high mi ultrasound with contrast, which is this bar here, was actually better than of course than reference, better than than standard ultrasound and better than the than ct.

It seems clear that contrast ultrasound is approximately as good as ct.

A more recent study from Italy supports that, this shows the sensitivity and specificity of conventional ultrasound contrast and ultra enhanced ultrasound and contrast ct.

You can see that contrast ultrasound comes out about the same as contrast ct.

That seems to be the general consensus.

Dr Cue from Italy's results are shown here as an ROC curve, and you can see that the CT and contrast ultrasound curves are virtually super imposable.

We think that contrast ultrasound performs as well as contrast ct, probably MR is a little bit better gadolinium, mr, of course.

Characterization of Focal Liver Lesions

Turning now to characterizing focal liver lesions, there are really several levels of information that we might like to get from it.

First of all, distinguishing benign from malignant.

Of course, it's the base level and the most important information that we might need, but we'd also quite like to be able to be a little bit more specific about the particular types of benign focal liver lesions that we're looking at.

For these, we use a combination of the arterial and the late phase can also use the portal venous phase, but it probably doesn't add a lot of additional information.

In the arterial phase, we are looking at the spatial and the temporal patterns of the filling, and therefore the low MI methods are essential.

If you try to use high MI methods for this, you destroy the bubbles and destroy the time sequence.

It is a particular feature of ultrasound with contrast that because we can work in real time, we have much more information, especially about those early arterial phases, then you can easily get from CT or MR, where you have to do a timed scan a certain number of seconds after the injection.

That's particularly helpful feature for this arterial phase.

The late phase is used in the same way as for staging to look out at washout versus retention of the contrast.

The low MI methods below 0.4 are microbubble preserving methods, and they work best with the per fluro agents, for example, vir, which is what we will use in all the examples that I'll show you, but also with definitive, they work with multi pulse methods whose prime purpose is to suppress the tissue echoes and therefore allow the contrast signals to come through, has an additional effect, which is that using a low MI reduces the amount of tissue harmonics, which can interfere with the suppression algorithms, the multi pulse methods, and as I said, it gives us real time visualization.

This chart is a little bit complex, and I'll show it to you in a slightly different form in a minute.

But essentially if you look at the metastases, you'll see that in the sinusoidal, the late phase, they show no retention of contrast.

HCCs are complex, and the way the better differentiated ones do show some retention.

The less well differentiated, the more malignant ones do not.

They're a little bit complicated, and that's based on the histological variation that we know that they have.

The benign lesions all retain contrast, but they have different patterns in the arterial phase, different amounts of enhancement and different spatial and temporal patterns as to how they behave.

I can illustrate this maybe best by showing you an example.

This was an interesting patient who was known to have non-Hodgkin's lymphoma, and we were asked to stage the patient's liver, and we discovered this lesion in the left lobe.

It looks rather ugly and por, I think we'd we'd have been suspicious of it pec about it, especially with the doppler showing that it's vascularized.

Then there's another lesion, which is a little bit more subtle, and is echogenic, and we dismiss that as probably being a hemangioma.

It didn't show any vascularity.

Here is the first lesion, 40 14, 15 seconds after the injection in the arterial phase.

You can see that there's contrast and appearing around the lesion at 21 seconds.

You may be able to appreciate that the contrast is not uniform, but it's rather nodular and gular in its distribution.

Carrying on a little bit further, we're in at 36 seconds.

Now, this nodularity is accentuated and you can get the impression that the contrast is gradually moving towards the center of the lesion.

Later on here at two minutes, the lesion is almost completely filled.

These are the classical hemodynamic patterns of a hemangioma exactly as is also seen, though less well, I think with CT and Mr using contrast agents.

Much to our surprise, this suspicious looking lesion on the B mode has benign features and wasn't actually a problem at all.

Here's the contrast of the second lesion in the right lobe of the liver lesions up here.

You see that it's not hypervascular in the arterial phase, and as long as we looked, it remained completely devoid of signals, and we of course did scans much later than the 28 seconds where I've terminated this video.

This was biopsied and proved to be a lymphoma deposit.

It's an interesting case because the B mode was really rather misleading.

The contrast told the true story, and that was a metastasis.

Guidelines for Focal Liver Lesions

The European Federation of Ultrasound in Medicine and biology met in 2 0 4, actually took a long time this, and produced a series of guidelines which were updated into oh eight, published actually last year in in 2 0 9.

The guidelines really underlay the complex chart I showed you, but here they are pulled out in a more simplified form.

First of all, looking at non-cirrhotic patients.

This is a the malignant ones here will be metastasis.

You see that the key decision is whether they retain contrast and show sustained enhancement in the late phase.

If they do, they're classified as benign.

If they don't, they're classified as malignant.

Then in the using the arterial phase, which is the yellow boxes down here, you can classify them a little bit further.

If you see, for example, I showed you an example of a focal nodule hyperplasia, hyper enhancing, and on real time you'd see central centrifugal filling and so on for all the different kinds of benign lesions.

Then likewise for the for the malignant lesions.

The sustained enhancement is the benign feature, and then the arterial in the late phase, and then the arterial phase is used for sub classification.

Now, the situation in cirrhosis is a bit different.

'cause here what we're looking for, of course, is a hepatocellular carcinomas.

It's whether or not they're they are hypervascular in the arterial phase, that's the key thing.

If they show that together with a wash out in the late phase hypo enhancement hypo in in the late phase, then that's likely to be a hepatocellular carcinoma and so on for all the other lesions.

Guidelines for Cirrhosis Management

As we progressed in this story of learning how to use contrast agents for ultrasound and how well we could rely on them, there was of course a confidence process involved in convincing, first of all ourselves, the radiologists, and then the referring clinicians.

There was a break point that I thought was quite interesting, happened in 2 0 5, when the second meeting of the liver associations from the USA and from Europe produced se a series of guidelines.

These were concerned particularly with cirrhosis.

The problem is managing a patient who in whom a focal lesion was seen on surveillance in a on the background of cirrhosis.

This is the guidelines that this so-called Barcelona agreement came up with.

If the lesion is smaller than a centimeter, then it's just a referred for repeat ultrasound at three monthly intervals.

But if it's larger, then it needs to be investigated further.

The further investigation involves a dynamic study with a contrast.

In the original versions, that had to be a contrast CT or mr.

If it was between one and two centimeters, then two of these dynamic studies were needed.

If it was larger, then one would suffice for distinguishing whether whether they were suspicious for HCC or not.

What the two of five guidelines added, which I thought was very interesting, was that of of the contrast studies, the dynamic studies required for this further evaluation, ultrasound was rated as being equal in importance as CT or mr.

Typically these lesions would be seen on the screening ultrasound contrast study would be done on ultrasound.

If that was, if the lesion was larger than a centimeter, it would be followed up in the case of the one to two centimeters with a contrast CT or mr if it's if it's larger than two centimeters, that could that would be the end of the investigatory process and just the ultrasound would suffice.

Monitoring Response to Treatment

Then my third topic is the use of ultrasound of contrast in evaluating the effects of treatment.

For interstitial ablation, this has turned out to be a crucial component of the ablation process.

It is of course absolutely essential that not only the lesion, but a safety margin is ablated, but b mode doesn't distinguish coagulated from normal tissue, and doppler isn't very sensitive.

The conventional ultrasound techniques that we have are really inadequate for this task, and that's a pity because ultrasound is the best way to guide the interstitial ablation, for example, with radiofrequency for heating the tissue.

In the traditional management system, the ultra ultrasound be would be used to guide the procedure, and then the patient would have to be taken out of ultrasound and go to CT and have a contrast CT study.

If viable tissue is found, patient would have to come back to ultrasound and have the additional lesion re ablated under a second ultrasound session.

Obviously a very inefficient technique.

Contrast ultrasounds completely changed and dramatically transformed that process because it depicts the micro vasculature as a as a perfusion defect.

You can see when tissue hasn't been adequately ablated.

The pro procedure can be performed in the same ultrasound session.

There is one little problem, gas from heating of the tissues obstructs the ultrasound beam.

After the R RF ablation has been performed, you need to wait for five, maybe 10 minutes to let the gas dissipate before continuing on.

Here's an example, kindly loaned me by my colleague, professor Ed Lee at the at Hammersmith Hospital, where A-A-A-A-A lesion was ablated.

On the contrast study, you can see that there's a viable corner of the tissue there.

A second ablation was carried out and the lesion was removed there.

This use of contrast has facilitated this whole procedure and it results in cost and time savings apart from being much more convenient for the patient.

Anti-Angiogenesis Treatment Monitoring

The other aspect of monitoring response concerns anti-angiogenesis treatment, which is widely used now for a large number of different tumors.

The traditional way to measure response was to look at the size of the lesion using the so-called resist criteria.

It's a volume of essentially a volume measurement of the tumor.

But this is very crude and especially in this context of anterior angiogenesis treatment, because these treatments don't shrink the tumor as a first result of the of their effect.

The first effect is to damp down the arterial supply to the tumor.

Really measuring the volume of the tumor is an inappropriate way to evaluate these.

The response of the tumor tumor anti-angiogenesis treatment, very slow.

It doesn't show what we primarily would like to look at, whereas contrast agents do this very well.

I'll just show you one example of a series of measurements, and this is kindly lonely by Dr. Lasal from Paris, and she's been heading up a Marty Center French study, which is now completing, looking at exactly this.

Here's a tumor being treated with one of these anti-angiogenesis treatments, and she's made a functional curve of the wash in and wash out of the contrast agent.

Then repeated this, this is at 14 days.

You see, although the tumor hasn't changed in size very much, the vascularity is all but disappeared.

This seems to be a very good way of monitoring the progress of these tumors treated with anti-angiogenesis agents.

Conclusion

In conclusion, microbubbles really have revolutionized liver ultrasound.

They're absolutely essential for focal liver lesions for characterization, for detection, and for monitoring interstitial ablation.

Thank you very much.