Potential Conundrums of Hypervascular Lesions - HD

Introduction

Good afternoon everybody,
and thank you again.
It's such a great honor to be here, to be invited by Paul.
I'm speaking really in front of all the great gurus
of contrast ultrasound, many of them who taught me how
to do contrast ultrasound.
But after receiving Paul's email, this was me
because of the topic I was given.
And really I had to ask to extend the topic
because to include really what things have been shown,
all the classical features of all the benign
and malignant lesions, what causes me most confusion,
and this is because really as David had alluded to earlier,
that in 5% of your contrast cases, you may not be able
to come up with a diagnosis.
And we know that in any imaging modality,
not all lesions demonstrate classical enhancement
characteristics.
What may help is patient history,
and you will have anticipated findings prior
to doing a contrast study based on the history.
And that can be or may not be helpful
and sometimes can be detrimental.
But what you do have as well,
apart from contrast ultrasound,
are the other complimentary imaging of CT and MRI.
So what I've done, and I was preparing this thought, was,
it's quite a difficult topic
and I thought perhaps I could divide it up into things
that cause me confusion into those,
that incidental findings in patients who come up
with an incidental ultrasound scan
and we just pick it up lesion in the liver to those
who have known history of malignancy
and then what I call the overlap lesion
because it's really difficult
to tell whether it's benign or malignant.
And then those with chronic or diffused liver disease.

Incidental Findings

So let's start off with incidental findings.
And a lot of us on our routine ultrasound list will pick up
the incidental echogenic lesion.
And this patients normally have no history of malignancy
or chronic liver disease.
And hence Peter showed us many examples
of beautiful hemangiomas.
We were actually looking at this echogenic lesion here,
and further up in the liver,
we actually saw a second larger lesion,
which we couldn't appreciate.
So on gray scale.
And you can see these have classical features
of a hemangioma.
Again, focal nodular hyperplasia can be very difficult
to see on gray scale,
but they exhibit the typical features of
central enhancement from the center outwards with
that classical spoke wheel appearance of these vessels.
And most importantly, they do not wash out
in the late phase.
However, there are some different lesions which don't always
behave like what has been set out in the textbooks
and it can be particularly difficult.
For example, I just want you to concentrate
your eye here in this lesion.
Here is hypervascular.
Perhaps it's filling from the center outwards,
and if we go to the B mode, we can actually see it then.
It's quite a difficult area you can see is
the diaphragm of the liver.
And this we classified as a high flow hemangioma
because it didn't wash out in the late phase.
And also it seemed to fill
with a slightly nodular enhancement
from the center inwards.
But if you are struggling
with whether it's filling from the center outwards
or outside inwards, then one key characteristic
of contrast enhanced ultrasound compared to CT MRIs
that you can go back and review these video clips.
And time and time again,
I find myself slowing the video clip down, just making sure
where this large lesion is enhancing from.
And if you slow it down slow enough,
you can actually see a feeding vessel in this lesion
and is filling from the center of the lesion outwards
with a nice structure.
And there you go, just slowing it down.
And you can see this was clearly an FNH,
but if you're still struggling,
there are other tools which you can use.
Something called parametric imaging.
So there are tools on the ultrasound scanner which you can
actually color code the arrival times of the microbubbles.
And for example, in this same lesion, you can see red is
where it starts to arrive earliest.
And you can actually see a better pattern whether it's
filling from the center outwards
or a haphazard pattern.
And so there are tricks
and tools which you can help with deciding
how the filling is of for these hypervascular lesions.

Patients with Known Malignancy

Changes slightly in a patient with a known malignancy
and is it a metastasis?
And of course it's very straightforward if it washes
out the multiple lesions.
We've seen loads of examples. But how about this case?
Are these enhancing lesions? Metastasis?
This is the first presentation in a patient who
actually had a CT and multiple focal lesions, which
were assumed to be metastasis.
And you can see hypoechoic lesions.
The liver is typical peripheral enhancement,
but when you look more closely, you can actually see
septi within these lesions itself.
And as we start the portal venous phase, as you scan
through the septi become more apparent.
And these are classical hallmarks of abscesses.
And as we scan through the liver, you can actually see
multifocal abscesses within the liver.
And this was E. coli,
and a patient who had assumed
to be metastasis coming for a biopsy.
Here's another patient with a known colorectal cancer
where contrast ultrasound may be helpful.
This is in a patient who's had several liver resections
and treatments for colorectal cancer and liver metastases.
What I was asked to do was to characterize this,
what was assumed to be a new lesion on follow-up CTs.
You can see this very faint hypodense lesion
in segment two of the liver here,
I'll just show you a little bit more.
She's had part of a metastasis shelled out
in the left liver.
And what you can see on ultrasound as we go back,
it's actually a very faint echogenic lesion
in the liver and also more posteriorly,
a faint echogenic lesion.
Just further up corresponding
to those two lesions I showed you on CT.
And here is the contrast clip of that faint
more anterior hypodense lesion.
Very difficult to see when you switch
to gray scale sometimes
because of the low MI mode to sort
of stay on your lesion itself.
And if you watch more clearly, you can actually see
that you don't have any pattern.
This is the echogenic lesion here.
I've put a cursor on it
and this is its enhancement pattern just
to help me see again, more tools to help you
with these small lesions
to further characterize your lesion.
And of course, in the late phase, I'm waiting to
see whether it washes out and it didn't.
So that left me was is it a new metastasis?
Probably not because it hadn't washed out.
And what could it be?
This lesion further up
posterior was the shelled out lesion.
It showed, it appeared bland throughout the contrast
enhancement phases and it was just a post-treatment site.
But as I dug back, I was wondering why we couldn't
see this probably.
And you can see actually in some
of the late phases, it's very hard to see this lesion.
And when I dug back to her old MRI scans some five years ago
prior to her MRI, you can actually see a faint hyperintense
lesion just high up here.
And this we think is going to be a hemangioma.
It's not washed out. It was probably there before.
It was very small and difficult to see
and it just depended on the phases on CT
that this was picked up.
So we are going to follow that up.
But this can be a conundrum,
but it is the wash out in the late phase
that makes understand that this is likely
hemangioma or benign lesion.
How about this case, where again, it's a patient
with a known malignancy and there was a blush seen on CT
and when I scan through on my gray scale ultrasound,
all I could see was perhaps this echogenic
lesion high up in the liver.
And I thought, oh gosh, it's probably going
to be a hemangioma based on the blush
or perhaps focal nodular hyperplasia.
And we gave contrast again on the B mode
because of its position.
Very, very difficult to see.
But what we did notice was
what was this enhancing hypervascular lesion posteriorly
there with a big vein.
Well, here it is again, very difficult
to see suddenly look suspicious, not typical
of any benign lesion, classical benign lesions that
has been discussed.
And so we switched over to sort
of improved doppler resolution like B flow.
Other manufacturers have similar forms
and you can actually see now with just the improved
resolution that is actually a vascular malformation.
You can actually see the feeding vessels coming in
and out of this lesion itself.
And a big draining vein.

Overlap Lesions

Just moving on now to the overlap lesion
and that really is adenomas
and enhanced beat has really discussed this at length,
but this is really a problematic lesion
with a contrast and enhanced ultrasound.
And I'd just like to illustrate this.
This is a patient with again
a known malignancy breast cancer.
And we saw this lesion on a background
of a fatty liver hanging off segment six.
As we start the clip, you can see that it enhances slowly.
She had a poor cardiac output.
There is some hypervascular to it,
slightly more than the adjacent liver.
And as time progresses, it starts to begin to wash out
compared to the adjacent liver,
but perhaps not
to the same extent we would expect of metastasis.
So there is still some bubbles hanging around this lesion.
We did postulate an adenoma and we biopsied
and this came back as a hepatic adenoma
and not a metastasis.
So there are some features which could help you
with a diagnosis of hepatic adenoma in the EFSUMB guidelines.
And there are the one of the main features is perhaps
that it can be, most of them can be hyper enhancing
or iso enhancing.
And in the majority
of them they do occasionally demonstrate some very,
very mild washout, perhaps not just much as
what you would see with metastasis.
Again, I'm just going to very, very quickly go through this
cause hospitals covered this.
What they can be with hepatic adenoma is
that they're multifocal and can hemorrhage
and in a small proportion have a malignant potential.
So you're always going to have
to probably biopsy these lesions again,
histologically, there's a big conundrum.
There have been confusions by histopathologists from differentiating these hepatic adenomas
from well-differentiated HCCs.
And it's even seen from FNHs,
but they do have other tricks up their sleeve
with immunohistochemistry, which does help them
with this diagnosis.
So as you can see, histologists are having difficulty
with it, just looking at it.
Then the imaging characteristics from a hepatic adenoma
compared to well differentiated HCC can be very
confusing as well.
So let's look at some examples.
Here is a hepatic adenoma biopsy proven,
see a large hypoechoic lesion hypervascular,
it's called slight rim to it as described and
but it enhances in a slightly peripheral manner
but with some structure to it.
And in the late phase you can actually see
that there is no significant washout.
And that was just a flash and reperfusion again to look at
its arterial enhancement.
What about this? We've seen
that it's enhancing hypervascular doesn't wash out.
Is this a hepatic adenoma? Is this an FNH?
Well, this is one we've seen earlier
and this turns out to be an FNH
because it's got filling from the center outwards.
And this is one case where you perhaps you could go back
to your video clip to actually assess
where it's filling from to get a better assessment of
what sort of lesion it could be.
And in this case it was just an incidental finding.
How about this case? Here is a hyperechoic lesion
and you can see again, perhaps unlike the FNHs,
it enhances more from the periphery with the not
so organized structured from the center outwards as FNH
and you don't have that central spoke wheel pattern.
So I just wanted to draw your attention to one
of the papers in a large series
of adenomas actually written by an Italian group.
And two of the speakers from today
and tomorrow will be here from this paper, Vita Canani
and Merko Donofrio.
And again in 25 lesions.
What they showed was
that these lesions showed centripetal early enhancement,
isoechoic of mild washout in the late phase.
So very similar characteristics sometimes
to well differentiated HCCs, as you'll see
in the guidelines.
One thing that can occur with these are
that they can be multiple.
And this is a case provided by Proi
and Cosgrove, you can actually watch this hypervascular
lesion, but there are multiple focal lesions elsewhere.
And this is a case of multifocal adenoma.
I just borrowed this slide from a paper.
Very nice radiographics paper about hepatic adenoma.
There can be several of them.
One advantage of CT
and MRI compared to contrast ultrasound is it can pick up
the hemorrhage within the lesion itself.
But I think all have difficulty at finding which one has
the malignant transformation in these lesions.
Again, on CT they tend to be hypervascular.
Again, the sub classification,
this was a very nice paper from a JR 2012
and really there were three recognized types based on some
immunochemistry
and one HNF1 alpha inactivated is associated
with steatosis and inflammatory HCAs consultative associated
with amyloidosis and other risk factors.
But it is the beta-catenin ones
that have increased risk of malignancy.
Now, this group had over nearly 50 patients with adenomas, of which only four
of them were beta catenin.
So they couldn't look at its imaging characteristics.
And I just wanted to draw to you its features.
This is the HNF1 alpha inactivated HCA
and versus the inflammatory type.
And you can see in the majority they tend
to be slightly more hyperechoic on B mode
on the inflammatory
and also a bit more on the HNF1 alpha.
So not very helpful. But with enhancement patterns in the late phase,
you tend
to have those in the inflammatory side washing out a lot more than those that
with HNF1 alpha, which tend
to be a little bit more isoechoic, but there is overlap.
Again, there's only 50 cases, so
but there are some slight distinguishing features
between the types.

Patients with Chronic or Diffuse Liver Disease

Then moving on then
cause adenomas can be difficult to differentiate from
well differentiated HCCs, what about those patients
with chronic liver disease?
Hepatitis C and cirrhosis,
and we know that they're predisposed
to primary liver cancers and they all develop nodules
and progress to large HCCs.
Eventually with advanced stages when it's in this region, it's quite
most modalities are quite good at coming up
with a diagnosis and being giving a high probability of the diagnosis.
But it is in this range here, those dysplastic,
low grade nodules that can be difficult
and pose a problem not only
for contrast enhanced ultrasound, but also for CT and MRI.
Here's the some guidelines on the cirrhotic liver
of regenerative plus minus dysplastic nodules.
More often than not they tend to be an iso enhancing
to the adjacent liver
and they do not wash out, occasionally they say
that it can be slightly hypo enhancing.
This was one which we're still following up of this patient
with no liver cirrhosis.
You can see that it is enhancing nearly
to the same degree as the adjacent liver,
but in the late phase, you see at five minutes,
it really retains the contrast on its own.
Just wanted to put this up to just say with the HCCs,
they also occur in the non-cirrhotic livers
so they can occur without significant
chronic liver disease.
Just something to have in the back of your mind.
And I just wanted to put up that
we've seen some lovely examples.
This is perhaps one which isn't so typical.
You can see it's a large lesion.
It's indenting and bowel making this capsule
of liver bulge out haphazardly.
There's some abnormal vessels in the early arterial phase.
It was perhaps central and peripheral nodular perhaps,
but no, and this area here doesn't fill at all
and it's probably the necrotic center of the HCC.
So to have to watch very carefully watch the video clips again,
to not confuse it with a hemangioma.
And sometimes again, watching those video clips in
that arterial can be very helpful.
Here is a patient with no significant chronic liver disease
and the question was, is it focal nodular
hyperplasia in this lesion?
That was what was raised by the other imaging modalities.
And you actually watch, there is a central vessel,
this is a VRI mode where the contrast is overlaid
on the B mode itself.
And as you watch this video clip again,
you can actually see the fill is actually from more from the
periphery, then from the center outwards,
but it has got some structure to it.
And this is as biopsied to be an HCC.
So it shouldn't confuse you at the end once you look at the early arterial phase
and see where it actually fills from.
So very, very useful to have that tool.
This is a case provided to me by Dr. S from the royal free who's in the audience today.
Cirrhotic liver can consist a nodular outline, cirrhosis,
a lesion that bulges out
and again, not so dissimilar
to the lesion I showed you earlier.
Perhaps this enhancement looks a bit peripheral nodular
and it doesn't all enhance completely,
but its pattern isn't quite so classical for the hemangiomas
that we have seen.
Although this was mooted as a possible diagnosis and even CT
and MRI couldn't take this further
and were more favoring hemangiomas.
But when it was biopsied, this came back
as a hepatocellular carcinoma.
So really I think RA piscal has really discussed non-invasive diagnosis of HCC
and the guidelines available
and the controversies amongst it
and summarized that really, really nicely
for my simplistic mind.
Anybody with cirrhosis and a hypervascular lesion
or a background of significant chronic liver disease
and I see it, it should be suspicious for HCC
until proven otherwise.
And soon we will have the LI-RADS guidelines
to help us with it.
I just summarize what all those other guidelines perhaps hinge on.
Mostly CT and MRI, perhaps some
of them using contrast enhanced ultrasound to diagnose HCCs.
But it's also on size.
And what you do with these lesions is based on the size of
that lesion, whether they're hypervascular or not.
Typically it's less than one centimeter
and this is where CEUS will have more of advantage compared
to the other imaging modalities is looking at the vascular
pattern is that they suggest you follow it up
or in some cases you may hopefully not need
to biopsy, but you may eventually biopsy
some of these lesions.
So perhaps I need some help with this case, which
I stuck my neck out a little bit
because this is the patient we've been following
with cirrhotic Hep C for many, many years.
And you can see it's a very heterogeneous
liver irregular outline.
We picked up this echogenic lesion in the liver,
hadn't never seen before.
Looked a bit suspicious.
Is this a regenerative nodule or is this an HCC?
Well, here is the contrast clip. This is the lesion.
It's hypervascular,
it's got abnormal vessels filling from all
around in my book.
On a background of liver cirrhosis.
This was very, very suspicious
for HCC in the late phase.
You can see it washes out perhaps not as much
as we would expect from a metastasis.
There is some retention of contrast.
What we did find as well is another lesion elsewhere in the
liver, which had washed out.
And when we gave a second dose,
it also enhanced the same degree as the original lesion.
So this was clear cut multifocal HCCs in the liver
in a patient we'd been following
for liver cirrhosis for many years.
This was just a vessel in the way there.
But what I didn't tell you was
that he also had quite advanced prostate cancer,
which had recurred
and it had recurred in such a way
that it also invaded the rectal wall.
And this is the MRI scan
of this highly invasive rectal cancer.
And until today I thought prostate cancer
with liver metastases were quite uncommon,
although hence Peter has shown us two examples today,
particularly prostate mets
to the liver in a background of cirrhosis was even rarer.
And in the MDT I was asked, I stuck my neck on
and said these were multifocal HCCs
but of oncologists, they said they wanted some tissue,
he wasn't the easiest patient to biopsy.
Can I just have a show of hands of those
who have actually biopsied this lesion based on all those
perhaps four biopsy.
Can I have a show of hands? See, no,
not many people, a few at the back.
I feel a little bit better. Diagnosis
of HCC, stick your neck out and MDT say no.
I think these are HCCs. You don't need a biopsy, can I?
A few. Well I did
and I had my arm twisted to biopsy.
This lesion, which you can see on the gray scale,
was already difficult to see.
I had to give contrast to help me localize it.
Here finally is my biopsy needle to
that most anterior lesion we had.
He was in poor in a poor
performance state as well, this patient.
And then at the next MDT I was eating my own words
because it came back as a prostate carcinoma
and metastasis from prostate carcinoma.
So I'm afraid you can't win from all this series.

Additional Conundrums: Contrast Ultrasound and HCC Screening

Another topic I'd like just to touch on, on conundrums
of these hypervascular lesions is contrast ultrasound
and HCC screening.
Because it is controversial,
we are reliant on the arterial phase
for detecting those HCC lesions.
And as we know well-differentiated HCCs
do not always wash out.
Again, there's no large published data on whether we should
be giving contrast in the setting to look for HCC screening from our own unpublished data
where we followed for a couple of years of a hundred HCV,
biopsy proven liver disease in 30 cirrhotics followed
for two years with six monthly scans to develop HCCs.
But we actually could see the lesion without contrast.
So with all those conundrums, hopefully
when you say, oh my god, what is it?
And the patient's looking at you anxiously just keep
calm and scan on.
And just to say, I put this up, there are lots
of senior colleagues who nice big committee.
I have the fortunate lesion to actually see,
have Paul and David to ask advice from.
So I'd just like to summarize for my simplistic view of
what happens when I see a hyper enhancing lesion in the non,
I divided them into non-cirrhotic liver in the cirrhotic
liver and those with clear washout,
then I'm thinking metastasis most commonly.
And the background HCC, even though sometimes
as we know non-cirrhotic livers can occur
in the less common things.
Cholangiocarcinoma. So lymphoma
in the non-cirrhotic liver incidental findings,
if there's no wash out
and I see a hypervascular lesion, I think of FNH
if it's very hypervascular.
And then in the back of my mind, high flow hemangioma
or even a vascular malformation.
And I would rely on the early video clips
in the arterial phase to help me with that diagnosis.
Internal septa, I think it comes to a talk later on this afternoon are quite classical
features for abscess.
But if the B mode and enhancement pattern are iso
or there is some mild wash out
but not maybe quite enough that you expect for metastasis,
then perhaps you should consider adenoma or HCC.
And inevitably I think you end up biopsying these lesions
with a cirrhotic liver and there is no washout.
Perhaps you can be reassured
that it's a regenerative nodule.
Again, guidelines can be helpful
and it's really based on the size of nodule.
But just wanted to say you have CT and MRI to help you
or the interval scan.

Conclusion

So in conclusion, CEUS is really good I think to help
with the distinguishing benign for malignant lesions
and to help you to also decide which ones you may want
to follow up and which ones you want to biopsy
or which ones you can discard.
It's very helpful to apply the basic principles from the
EFSUMB guidelines, but as we know,
conundrums are inevitable in all imaging modalities.
And as one Rachel just told me when I was training said,
you can only be wrong when they occur.
Patient medical history, very helpful,
but sometimes can confuse you.
But what is very helpful are other imaging modalities which
are complimentary to help you achieve that diagnosis.
So finally I just leave you
with the last conundrum from my talk is do we need
to give contrast in these patients?
Here is a patient with clear liver cirrhosis.
This is B flow
or SMI in the Toshiba,
looking at the improved doppler techniques
that we have a lesion which clearly has an avid vascular
supply with neovascular vessels, clearly HCC, do we need
to give contrast
or how about in this incidental finding in a patient which
just came for a routine scan
and we actually see the nice centripetal spoke wheel pattern of this FNH.
And sometimes we can even produce 3D images showing the
central scar and the feeding vessels from the FNH.
Do we still need to give contrast?
I just pose to you as a part of the conundrums talk.
It finally just leaves me to thank all my colleagues
and also to invite you
to the British Medical Ultrasound Society Annual Scientific
Meeting this year, seventh 9th of December.
As you can see, there's a great committee to exchange ideas
and also a lot big educational program.
Here's the outgoing
and incoming scientific chairs for it.
We also have a lot of fun and this meeting.
So please, I invite you quarterly to come
to join us in York in December.
And finally to thank all my colleagues who've
provided cases for my talk today.
Thank you very much for your attention.