CEUS: Opportunities for the Future - SD

Introduction

I am Stephanie Wilson.

I'm a radiologist at the Foothills Medical Center in Calgary
and a professor of radiology at the University of Calgary.

Today I'm going to speak about contrast enhanced ultrasound
and the current opportunities for the future.

This presentation is entitled Contrast Enhanced Ultrasound
Opportunities for the Future.

Historical Perspective on Contrast Enhanced Ultrasound

From a historical perspective,
contrast enhanced ultrasound is performed
with Microbubble contrast agents.

They are unique in that they interact
with the scanning method they oscillate in response
to a low mechanical index
and disrupt when exposed to a high mechanical indexed field.

Imaging with microbubble contrast agents requires
contrast specific imaging techniques
and we will not cover this any further in this presentation.

Now, microbubble contrast agents have been approved
for use in more than 70 countries.

However, there is no approval
for radiology indication in the United States today.

The black box labeling by the FDA in November of 2007
was interpreted as a knee jerk reaction in the face
of inadequate information
and we're hoping for a promise
to rescind this black box labeling in the future.

A cardiology publication published recently in 2008
suggests that Microbubble contrast agents
actually improve patient outcome and do not increase deaths.

Further, an excellent European collective data sets suggests
that complications are extremely rare,
including 23,000 studies performed
by multiple European centers.

Clinical Impact of Contrast Enhanced Ultrasound

Now, the clinical impact
of contrast enhanced ultrasound includes the fact
that there is no risk of nephrotoxicity,
there's no requirement for radiation.

And further the microbubble contrast agents are well
tolerated and easily handled.

Further, they diminish the requirement for CT and MR scan.

So if we consider contrast enhanced ultrasound,
what is the most successful endeavor?

Undoubtedly it is characterization of focal liver masses
and I base this on the published papers
and the known clinical work
and research endeavors which are already done
and are ongoing.

So the two issues are characterization of liver masses
or asking ourselves the question, what is it?

And the second issue of liver mass detection
or asking ourselves the question, is it there?

So what are the clinical situations which cause us to do
an ultrasound for characterization of a liver mass?

Well, the common situations include diagnosis
of an incidentally detected liver mass.

So everyone is familiar with the fact
that when we perform ultrasound, let's say
of the gallbladder or of the kidney,
we might see an incidental unsuspected liver mass.

If these masses are of a sufficient size,
patients are frequently sent for either CT
or MR scan for characterization.

Today in laboratories which perform contrast enhanced
ultrasound, we can do characterization
of the mass at the time of its detection.

Another unique situation is that
of a hypervascular mass found in a young woman.

These hypervascular masses often seen on
outside CT scans are known to represent either
a focal nodular hyperplasia or an adenoma.

Their differentiation is excellently performed
with contrast enhanced ultrasound.

And then we come to the very important question
of differentiation of nodules found on surveillance scans
performed for hepatocellular carcinoma
in high risk patients.

This special topic is not covered in depth in this
presentation but is one of the major
and most important uses for contrast enhanced ultrasound.

And another very important use is for resolution
of masses indeterminate on CT or MR scans
and of course the very important situation of patients
who are referred to us from our clinicians.

And once the clinicians are aware that we are around
and doing this service,
they will very willingly send their patients
for this very safe and effective procedure.

Diagnostic Algorithm for Liver Masses

Now, one of the very first things
that we undertook in our laboratory was
to establish an algorithm for the diagnosis of liver masses.

On contrast enhanced ultrasound as would be expected,
these algorithms are often similar
to those which are developed for the non-invasive diagnosis
of liver masses on CT and MR scan.

And these are based on the enhancement
of the mass in the arterial phase
and the portal venous phase as compared
with the adjacent liver.

Now contrast enhanced ultrasound has the added advantage
of temporal resolution and vessel morphology.

Further, we found early on in our studies
to establish an algorithm
that washout has a high association with malignancy
and that sustained enhancement has a high association
with malignancy.

And in our own study performed around 2000,
we had a high positive predictive value for both
of these outcomes.

Now today we've taken our own data
and that of multiple other investigators
and developed this schematic for looking at an algorithm
for diagnosis of commonly encountered masses
eliminating hepatocellular carcinoma from this schematic.

And so you can see here the type
of mass in the left hand column, the appearance
of the mass in the arterial phase in the middle column,
and then the appearance
of the mass in the portal venous phase.

So let's start looking at the right hand column,
the portal venous phase.

What's immediately obvious is
that the metastases shown in the bottom row
are always washed out
or appear black in the portal venous phase,
whereas the lesions on the first three rows representative
of benign tumors tend to be enhanced to some extent.

In the portal venous phase.

Then let's look at the arterial phase appearance
and here you'll see that there's considerable variation in
some of the tumors and that
this variation allows unique
and discriminatory features of most of these masses.

So let's look at these one by one.

Hemangioma

If we consider hemangioma a very common mass first
in the arterial phase, we're looking
for peripheral nodular enhancement
with centripetal progression
of the enhancement in either the arterial
or the portal venous phase.

Once enhanced, these tumors tend
to stay enhanced in the portal venous phase.

So if we look at this patient,
we can see on these three frames these are sequential frames
showing first of all enhancement
around the periphery of the mass.

Then there are nodules of enhancement,
the peripheral nodular enhancement
with centripetal progression
of the enhancement into the center of the mass.

This is a classic hemangioma.

We would not require any further investigation for
confident diagnosis of this lesion.

Let's look at another patient.

An oncology patient
who has only a portal venous phase CT scan.

On this scan,
an unsuspected mass was shown in the liver.

This mass is indeterminate on
that portal venous phase CT scan.

When we go to the ultrasound, we can confirm the presence
of this small mass on this long axis view looking at the
left lateral lobe of the liver
and you can appreciate the proximity of this mass
to the hemi diaphragm and the adjacent heart.

Now when we look at sequential frames from a contrast
enhanced ultrasound study taken in the same projection,
you can see the early marginal filling,
the peripheral nodules
and then the centripetal progression of the enhancement.

And this is of course even more beautifully shown on this
short A VI clip, which shows the classic features
of hemangioma
and you can note the large amount
of contrast in the adjacent heart
and that does not in any way interfere with the
clear characterization
of this very small mass again requiring no further imaging
for its confident diagnosis as an insignificant
and benign tumor.

Now what about this as an example of
where this comes into play in clinical investigations.

So here's a patient who's at risk
for hepatocellular carcinoma
and on a CT scan we can see a hypervascular
tiny nodule on the surface
of the liver in the arterial phase
and in the portal venous phase.

Now when we do a contrast enhanced ultrasound
of this tiny nodule,
you can see easily the peripheral nodules
and then the centripetal progression of the enhancement
of this tiny mass.

And when we look at sequential frames,
we can appreciate this marginal enhancement peripheral
nodules with centripetal progression.

Again, a benign lesion in this case a rapidly filling he
angio explaining this indeterminate mass shown
on the CT scan.

Focal Nodular Hyperplasia and Adenoma

So let's go on now to the frequent dilemma
that poses itself in young asymptomatic females
where a hypervascular mass is shown on CT scan.

So the question that is posed is,
is this an focal nodular hyperplasia or is it an adenoma?

So contrast enhanced ultrasound has the unique ability
to show vessel morphology
and the direction of filling of the mass.

So if we consider first focal nodular hyperplasia,
these lesions are homogeneously hypervascular in
the arterial phase.

They are also characterized by stellate vascularity
and centrifugal filling in the portal venous phase,
they show sustained enhancement often
with an unenhanced scar.

So if we look at sequential frames taken from a large mass
on a woman with such a mass, we can see first of all,
relatively avascular mass
and then we rapidly show the stellate vascularity the
homogeneous enhancement.

And in the portal venous phase the sustained enhancement
with the non enhancing central scar.

And when we look at the beautiful a VI clip,
we see the hypervascular mass, the patients requested
to suspend respiration,
we do MIP imaging showing the beautiful stellate vessels
with the centrifugal filling classic
for focal nodular hyperplasia, another woman
with a very subtle mass on the baseline scan,
barely visible marked by the arrows.

And here with contrast enhancement,
again a hypervascular mass
where MIP imaging clearly shows the centrifugal filling
and the central stellate vascularity classic four focal
nodular hyperplasia.

Now in contrast to focal nodular hyperplasia,
hepatic adenoma is also a hypervascular mass,
although it is our impression
that these are less hypervascular than FNH in comparison
to the stellate vessels of FNH.

The vessels here are diffuse or chaotic
and the direction of filling is centipede from the periphery
of the tumor to its center.

Furthermore, in the portal venous phase, this is one
of the lesions that is an exception to the rule in that
although one half of the lesions do show sustained
enhancement as expected for a benign tumor,
one half will show washout.

So here's a beautiful example, an asymptomatic young woman
who was scanned with elevated liver function tests.

You can see she has a fatty liver to account
for the elevated tests,
but she also shows a large hypo coic anterior mass.

When we look at the CEUS images,
you can see in the arterial phase a chaotic vascular pattern
followed by homogeneous enhancement at the peak
of arterial phase.

However, in the portal venous phase there is weak washout
which immediately makes us suspicious of a significant
possibly malignant lesion.

Here we can show on the a VI clip the beautiful centipede
filling and in the correct milieu we would suspect
that this would be hepatic adenoma rather than something
like hepatocellular carcinoma or metastasis.

Malignant Liver Tumors: Metastases

So now let's leave those lovely benign lesions
and look at some of the malignant liver tumors.

So metastases of course are very important tumors
as we've mentioned in the portal venous phase,
these lesions will always appear completely washed out
or black and this occurs very rapidly
for all malignant metastatic tumors
in the arterial phase.

Note that there are variable patterns and
although we expected at the outset of our investigations
that most would be hypovascular
more often in fact they are hypervascular transiently
or show rim enhancement again of a transient nature.

And so we learned from our early investigations
of these patients that the timing
of washout is discriminatory between malignant tumors in
that metastasis washout very fast,
often within the timeframe of the arterial phase
and they tend to wash out completely.

So let's look at a representative example.

So here is a CT scan
performed on a middle-aged man from our emergency department
and on this portal phase CT scan we can see an indeterminate
mass marked by the arrow in the left lobe of the liver.

And you can see on the ultrasound
exam performed subsequently again an indeterminate mass
lesion bulging the anterior capsule of the liver.

And on the bottom we can see an arterial
and a portal venous phase image all taken
before one minute showing the transient hyper vascularity
with rapid washout.

So if we look at this on a a, a cine clip,
we can see the mass is instantly hypervascular
as the liver begins to enhance more, we can see already
that this mass is beginning to wash out.

So this hypervascular with rapid wash out classic
for a metastatic lesion patient had an unsuspected
sigmoid carcinoma.

Now what about this patient who's had a previous radio
frequency ablation for liver metastases?

We can see on the baseline scan
that there's abnormality marked
by the arrows in the posterior aspect of the liver.

When we do the contrast enhanced ultrasound,
we can see the completely avascular
radio frequency ablation site
and then two metastases with rim enhancement.

So when we look in the arterial phase,
we can see the rim enhancement of the metastases
and then the portal venous phase, the complete washout
and this complete washout allowed us very early on
to recognize the increased conspicuity
of metastatic lesions when they're looked at in the
portal venous phase.

And so consequently, we
and others were suspicious that
contrast enhanced ultrasound would improve the detection
of metastatic liver disease.

So we participated with multiple other centers
in a multicenter study, early in the
about a decade ago
and this was work was published in radiology in 2003
and we showed that contrast enhanced ultrasound performed in
the portal venous phase improves detection
of metastases over baseline.

This is equivalent to CT scan
and superior in some circumstances.

So to make this point I'd like
to show you a shocking example.

Most are not this shocking
but this is very, very good for demonstrating the point.

So this is a patient who came from our emergency department
and this liver was called normal
and it is a strikingly homogeneous liver on
the gray scale scan.

And so this patient was felt to have a normal liver.

However, on contrast enhanced ultrasound we can see
immediately on injecting the contrast
in numerable hypo coic masses in the liver.

And when we sweep the liver in the portal venous phase,
you can see that that very normal appearing liver on the
baseline scan is virtually riddled with tiny
metastatic deposits.

This of course, as I've said is a very striking example
but it makes the point
that when you're doing contrast enhanced ultrasound,
it is essential on all patients to sweep the liver looking
for occult lesions if you have done a contrast injection.

Now a more common situation is beautifully shown here
where we're looking at an unenhanced CT scan
and an unenhanced ultrasound taken in a similar projection.

Now I'm sure you'd all agree with me that you'd never call
that liver normal on ultrasound
and also with windowing we are suspicious
that the CT scan is also abnormal
but neither one is diagnostic really of anything.

However, when we look at those livers in the portal venous
phase, they're absolutely concordant
and both show that the livers are virtually
riddled with cancer.

And so what does this show to me?

Well it really shows me that both CT
and ultrasound require contrast agents for both detection
and characterization of liver masses.

Hepatocellular Carcinoma

Now let's go from there briefly to malignant liver tumors
to hepatocellular carcinoma.

Now we're not going to cover this in any detail only
to mention so the classic feature
for hepatocellular carcinoma on contrast enhanced studies
including ultrasound is that
of arterial phase hyper vascularity
and portal venous phase washout.

Now what's important about this washout is that
in comparison to metastases which wash out completely
and rapidly, hepatocellular carcinoma often washes out
slowly and very weakly.

Furthermore, there are many variations
to the classic appearance of a hepatocellular carcinoma
and these include in the arterial phase iso vascularity
or even hypo vascularity of the tumor
and in the portal venous phase slow
or even no washout whatsoever.

And so therefore we had to have a whole additional schematic
for the possibilities
for nodules found on surveillance ultrasound.

And so this includes
of course on the top line the typical appearance arterial
phase, hyper vascularity with washout
and then you see the arterial phase variations
and the portal venous phase variations including slow
or no washout.

Furthermore, there are other nodules we see on these
surveillance scans representative of regenerative nodules
and dysplastic nodules
with their imaging characteristics shown there.

Now more detailed presentation on this is beyond the scope
of this presentation but is another very challenging
and exciting area for CEUS.

So let's look at a classic example.

Here's a patient with an indeterminate hypo coic mass on the
baseline scan marked by the arrows.

When we look at this, with contrast enhanced ultrasound,
you can see a classic hypervascular mass in the arterial
phase on the right hand image you can see taken at four
and a half minutes that there is weak
incomplete washout of that mass.

This is a classic hepatocellular carcinoma.

Furthermore, I'd like
to show this example which shows the benefit not only
of recognizing atypical appearances
but also to appreciate the multimodality approach
to the diagnosis of hepatocellular carcinoma.

So I'm showing just a single image from an MR scan
but this patient had actually both CT
and MR both of which reported a nodular cirrhotic liver
negative hepatoma surveillance.

The patient was sent subsequently for ultrasound
and we can see not very much in the way of multi nodularity.

Rather we see a single hypo coic nodule on the baseline
image bottom left And
with contrast enhanced ultrasound bottom right we can see
that the liver has enhanced
and that that nodule is hypovascular.

So this is clearly a variation.

We recommended biopsy in our belief
that this would either be a dysplastic nodule
or well differentiated hepatocellular carcinoma
and this is a well differentiated hepatocellular carcinoma.

Now other indications, for evaluation
of the liver include radiofrequency ablation
and intraoperative ultrasound
and these are not covered in depth in this presentation.

Agreement of Contrast Enhanced Ultrasound with CT and MR

Now I'd like to go on here to talk about the agreement
of contrast enhanced ultrasound with CT and MR scan.

So with Peter Burns we addressed this in a prospective trial
earlier in the two thousands
and as here you can see a beautiful hepatocellular
carcinoma with a more hypervascular higher grade
focus of cancer.

And you can see the beautiful complete agreement
and the similarity
between the contrast enhanced ultrasound image on the left
with the MR image on the right hand side.

Another beautiful example shows a hemangioma
with peripheral nodular enhancement and bridging.

And again you can see the beautiful agreement
between the ultrasound and the CT scan.

Now here's the hemangioma on ultrasound and CT scan
and when we did our study peripheral nodular enhancement
and the centripetal progression of enhancement
of he angios had the highest agreement at 93%
of all of the observations that we made.

But you can see in many other cases here a young man at risk
for hepatoma with a hypervascular necrotic tumor
complete agreement in the arterial phase
with beautiful washout.

Again, complete agreement in the portal venous phase
and this is a very common occurrence.

Discordance Between Modalities

Now when do we disagree?

Well, discordance is very important
and the explanations for
occasional disagreement are enlightening.

And so the two types of disagreement which I'd like to cover
with you include timing discordance as well
as contrast behavior discordance.

So let's look first of all at timing discordance.

So this is a 44-year-old woman with colon cancer
and we can see a hypo coic mass in her liver at baseline.

In the arterial phase you can see rapidly a hypervascular
mass and you can see that the mass is hypervascular
and that the liver parenchyma has not begun
to enhance at tall.

Now as the liver enhancement progresses, you can see
that very rapidly the liver lesion is completely washed out
or appears black even
before the liver enhancement is complete.

So this lesion is obviously hypervascular transiently.

On contrast enhanced ultrasound, this is another example
of the same pathology showing this very rapid transient
hyper vascularity with washout in the timeframe
of the arterial phase.

If we look at this patient
or other patients like this on ct, we see
that these metastases appear hypovascular
through all phases of enhancement.

And this is related to timing discordance
as the arterial phase image on the CT scan is taken at a
later time as a snapshot in time on the CT scan.

So this discordance is due to timing
and is beautifully shown on this little graph
where we can see the metastasis is the gray line which goes
rapidly up on the left hand side of the image,
whereas the black line representative
of the liver enhancement crosses the
metastasis enhancement
and then the CT arterial phase is taken when the metastasis
has already washed out.

Now the other very important kind of
of discordance is shown in this case of a 29-year-old male
with fructose intolerance
and he had a mass found on outside ultrasound scan performed
for abnormal liver function tests.

So when we look at his original CT scan, we can see
that in the arterial phase he has a heterogeneously
enhancing mass in his liver.

In the portal venous phase, this mass becomes more enhanced,
increasing both in amount of enhancement
and in the intensity of enhancement.

So this is reported as an indeterminate result.

We can't tell whether this is a benign or a malignant tumor
and we certainly don't know its diagnosis.

So after that indeterminate CT scan was performed,
this young man was then sent for an MR scan
and it shows the very same thing,
a heterogeneous mass in the arterial phase
with progressive enhancement in the portal venous phase
again indeterminate result.

So the patient was then sent
for contrast enhanced ultrasound.

And so here we can see the same result in the arterial
phase, a heterogeneous enhancing mass in the arterial phase,
but in the portal venous phase,
the mass washes out completely so that it's black
and this happens very rapidly.

So we suggested this was a malignant tumor,
this was biopsied and this is cholangio hepatoma.

So what is the explanation
for complete washout on ultrasound when both CT
and MR show increasing enhancement of the mass?

And so the explanation for this is another very important
clue to what's going on when this kind
of discordance happens.

So this is related to the different behavior
of contrast agents.

So on CT and MR we know
that there's a well-recognized tendency
for malignant epithelium to allow the contrast agents
to pass through the walls of the blood vessels into the,
adjacent soft tissue.

Therefore, the intensity of the hand
of the enhancement is related to
not only the vascular volume
but also the vessel permeability.

So if we contrast this with the microbubble contrast agents,
we're only talking about the blood volume in
that the microbubble contrast agents are of such a size
that they do not pass through the vascular endothelium.

So the microbubble contrast agents are pure
blood pool agents.

So the explanation for the contrast in the tissue,
of that tumor is related to tissue to
contrast in the tumor interstitium rather than
within the blood pool.

Conclusion

In conclusion,
our talk on microbubble contrast agents shows us
that contrast enhanced ultrasound is excellent
for characterization and detection of focal liver masses.

It has excellent agreement with contrast.
Enhanced CT and MR scan
and occasional discordance shows the advantage
of purely vascular microbubble contrast agents.

Thank you very much.