Contrast Enhanced Ultrasound to Assess the Tumoral Microvasculature: Potential and Reality

Introduction

My name is Olivier Lucid Arm.
I am an academic radiologist at Lap Petrie Hospital in Paris, France.
And I am going to speak about the use of ultrasound contrast in the assessment of Tumoral microvasculature in clinical practice.

Assessment of the Tumoral Microvasculature

This lecture concerns the assessment of the Tumoral microvasculature.
In clinical practice, two pathways are explored.
The first and the easiest is the path that leads to functional imaging in which contrast and ultrasound has many assets.
The second path, more difficult, leads to the targeted imaging of the vascular bed.

Functional Imaging

So let's start by going toward the functional imaging, which is the most developed technique.
We all know that ultrasound has the advantage of being a bedside, non-expensive and well-accepted technique that can be repeated as much as we want.
And we also know that the contrast agent is composed of micro bubbles with a diameter ranging from two to six microns, almost the size of a red blood cells.
They don't leak outside the vessels, so they are true blood pool agents.
It means that they have the potential to give information about blood flow, mean blood volume, mean transit time, but not permeability surface.
And in addition, they are very well tolerated and they can be used even in case of renal dysfunction.
Finally, contrast enhanced ultrasound is a very sensitive technique to the presence of microbubbles, much more sensitive than CT to iodine and even MRI to gadolinium.

But having said that, how can we assess the micro circulation in clinical practice?
First of all, because it's a bedside technique that can be repeated very often, we can use it in a purely qualitative manner.
It means using a simple visual assessment.
It works very well when the expected results of the treatment should be obvious on the vascular bed of the lesion, like after chemoembolization or radiofrequency ablation.

Qualitative Assessment Examples

Here is an example of an HCC before and after chemoembolization, and you can appreciate before embolization the richness of the vascular bed of the lesion here, and also a second lesion here and 40 hours after chemoembolization the same patient, the same setting, the same contrast.
You can obviously appreciate the rarefaction of the vascular bed, but also the failure of the procedure because of these remaining vessels within the lesion here and here and here.
Indeed, some papers have shown that contrast enhanced ultrasound was a promising technique for chemoembolization, either during the procedure to immediately assess the success of the procedure or after the procedure where contrast enhanced ultrasound seems to be more sensitive than CT in detecting the residual blood supply one week after tests.

Here is another example.
After radiofrequency ablation, thermo ablation necrosis appears as a homogeneous and avascular devascularized area.
Here you can see very easily the persistence of a large vessel crossing the necrosis area.
And it is because this vessel was large enough to dissipate the heat due to its blood flow.
Comparison studies in the literature have shown that contrast enhanced ultrasound had comparable sensitivity to initial incomplete treatment or later local recurrences than CT or MRI.
Thus contrast enhanced ultrasound was proposed to be used alternatively with CT or MRI to follow HCCs after radiofrequency ablation in order to reduce irradiation or costs.

One last example where the expected result on the micro circulation should be obvious.
It's a GIST treated by Imatinib.
And you can see here that the lesion appeared almost completely to totally devascularized except this area where you can still see some tiny little vessels and these vessels could not be seen by CT.

So we can say that contrast enhanced ultrasound is a very convenient tool to assess the effect of treatment when the effect of this treatment on the micro circulation is expected to be obvious, but the most challenging part concerns the treatment where the expected results are more subtle, like most of the targeted therapy, then we need to move toward a quantitative assessment of the vascular bed of the lesion.
This technique is very comparable to DCE MRI or functional CT and consists of studying the variation of the local concentration of bubbles as a function of time by measuring the effect of the contrast on the images.
Fortunately, a linear relation exists between the concentration of the microbubbles within the range and the echopower.

However, the echopower is never displayed as is on the video screen of the ultrasound machine.
Because of its large range of values, the echopower is always log compressed and converted into gray scale according to different transformation laws in order to get a nice clinical picture on the video screen.
So it means that if we want to perform a quantitative analysis, we cannot simply quantify the enhancement obtained on the video image in DICOM because the data obtained will be too far from the true concentration of the micro bubbles.
This is why we need to use dedicated software that will be able to get the result directly here, which is called linear data, or by doing a reversed post-processing and recalculating a signal as close as possible at the true echopower, which is called linearized data.
The first kind of software are usually embedded in the machine and are vendor specific, which may be a limitation in case of multicentric trials with different ultrasound machine.
And the second kind of software works with every kind of machine and it's probably better in case of multicentric trials, but it provides a signal which is not completely equivalent to the true echopower.

Quantitative Techniques

So what kind of results can we get?
There are two major techniques to quantify the micro circulation in clinical practice.
The first one, similarly to CT and MRI consists of injecting the contrast in a short bolus and to get a curve, a time intensity curve here, it'll be an echopower curve as a function of time.
The curve is usually not fitted to a pharmacokinetic model, but described by its wash-in rate by the time to peak or by the mean transit time, which are related with the speed of blood, the maximum intensity which is related with the blood volume and the area under the curve which is related with the blood flow.
However, as always, using a bolus, the shape of the time intensity curve becomes very dependent on the condition of injection on the shape of the bolus.
But, and with ultrasound it's always a problem because of the small field of view.
When you focus on the lesion you want to study, it becomes very difficult to be able to register an input function in a large feeding artery like the aorta, which are usually outside the imaging plane anyway, despite this limitation, this technique has been tested in two major multicentric trials in France.

The first one included 137 patients with one kind of cancer, a colorectal cancer with liver metastasis treated by only one type of treatment, the bevacizumab, but explored with different type of ultrasound machine.
Here is an example of a liver metastasis before and after treatment.
The second multicentric trial included 539 patients with different type of cancer treated by different type of treatment, but explored with only one type of machine, which was an Aplio from Toshiba.
Here is an example of an HCC before and after contrast.

Interestingly, the initial analysis of the first trial found that at baseline, if the time to peak of the lesion was lower than 60% of the time to peak of the liver in other words, if the speed of the blood within the lesion was very high, then the overall survival after treatment was significantly increased.
Similarly, the initial analysis of the second multicentric trial found that the time to progression was significantly increased if the area under the curve decreased more than 40% after one month compared to the baseline.
But of course, these last result must be stratified according to the different tumors type and treatment that is still ongoing.

The second major technique is very original and specific to contrast enhanced ultrasound.
It's called destruction replenishment technique, and it consists of injecting the microbubbles continuously by infusion instead of a bolus.
Then the concentration of the microbubbles in the vascular bed reach a nice equilibrium at the steady state.
And because the microbubbles are fragile, they can be destroyed instantaneously.
If we switch the ultrasound machine into a destructive mode, which is a normal clinical output power, then you create a known negative rectangular input function.
And as soon as you switch back the ultrasound setting into a non-destructive mode, the bubbles start to replenish the cleaned area following a very simple kinetic model of the first order, which is an exponential rising toward the plateau.
And the plateau is related with a fractional blood volume.
While the initial slope of the exponential is related with a fraction of blood replaced per second, then the product of both parameters reflects the relative blood flow.

This technique is very sensitive to the motion of the probe because of course, if you move the probe during the replenishment phase, then you will go into an area where you did not destroy the bubbles.
Then your replenishment curve will be completely wrong.
So that's why this technique was mainly used with animal models because the animal can be anesthetized and the probe can be fixed.
But some authors now promote to use this technique in human because they allow us to free oneself from the injection condition.
And it is very important because the variability of the injection is a real problem in real life because we inject only 2.4 ml of microbubbles.
So with a such low amount of bubbles, it is very difficult to be reproducible.
That's why the destruction replenishment methods start to be used in clinical studies.
And two papers recently published found that the decrease of perfusion assessed with the product of the parameter A and beta we just saw appear earlier than the morphological change during antiangiogenic treatments.

Targeted Imaging

Now let's go toward the direction of the targeted imaging to assess the microcirculation.
We know that microbubbles can be easily targeted against a variety of targets.
The only thing is that these targets must be located on the vessel wall because bubbles cannot leak outside the vessels.
This is an interesting technique because as we already said, contrast enhanced ultrasound is a very sensitive technique to the presence of a small amount of bubbles.
So making this technique probably the second most sensitive technique after PET for targeted imaging.

Many feasibility studies on animal model have already been conducted, like on this example of a renal cancer on mice where we used microbubbles coated with antibodies targeted against integrins and the antibodies were binding to the microbubbles through the avidin biotin complex.
The bubble alone, non-targeted, did not remain within the tumor after five minutes, it was the red curve while the targeted bubbles remain stuck in the tumor for more than 10 minutes.
This was the blue curve.
So we know that it works and it works in animal model, but what about eventual translation in human being?

It is doable not with the avidin biotin bubbles coated with antibody because they could induce allergy in human but with low allergenic targeted bubbles.
That includes in its shell a peptide having a high affinity to VEGF2, which is over expressed in tumoral vessels.
These bubbles are made by Bracco and it's called BR55 and is consequently usable in human.

In a preliminary rat model of prostate cancer explored with BR55, it has been found that BR55 could differentiate tumor from normal prostate based on angiogenesis.
Here is the prostate and here is the tumor 2 seconds after injection of BR55.
You can appreciate the global enhancement of the prostate and the tumor due to the circulating bubbles.
But after 10 minutes, all the unbound bubbles left the prostate and only the tumor that binding the targeted microbubbles remain strongly enhanced as also shown on this time intensity curve.
And this correlated very well with the immunostaining that showed a high expression of VEGF2 on the prostate cancer and no VEGF2 in the normal prostate.

Now a phase one concerning prostate cancer is actually ongoing with BR55 in human.
And here are some preliminary results like with rat BR55 accumulated in the cancer zone in human and it correlated well with the histopathologic and the immunostaining that demonstrated a moderate VEGF2 expression in that prostate cancer lesion.
Though the conclusion of this phase one study is that BR55 was able to bind to VEGF2 in humans inducing a pronounced enhancement greater than 10 minutes.
And of course it was safe and well tolerated in this phase one study.

Conclusion

So in conclusion, we can say that contrast enhanced ultrasound is a very powerful technique to visualize the micro circulation because it's a very sensitive technique that uses blood pool agent.
It is very useful at bedside in clinical practice already to qualitatively assess a devascularizing treatment like chemoembolization or radiofrequency ablation for example.
But it has also a great potential to quantify the microcirculation mostly in preclinical research, but more and more in human and now new targeted microbubbles usable in human are arriving.
If you want to know more in very practical way about the use of contrast enhanced ultrasound and the recommendation in the early stage clinical trials, you can read this paper recently published in European Radiology.