Principles of Ultrasound Molecular Imaging - SD

Introduction

I'm Lisa Villanova.

I'm a cardiologist from the University of Pittsburgh
and the director of the Center
for Ultrasound Molecular Imaging
and Therapeutics at the University
of Pittsburgh Medical Center.

I'll be talking about the principles
of molecular imaging using ultrasound.

Evolution of Diagnostic Imaging

Contrast.

Diagnostic imaging in medicine has evolved over
decades to first start off
with anatomic definition using imaging
to functional imaging such as with stress testing,
and more recently to the interrogation
of tissue at the molecular level, such
as shown here on a biopsy
or a pet image showing estrogen receptors.

We've been able
to make considerable progress in our ability
to image from the whole body
or whole organ level to physiology,
and now to molecular imaging.

Definition of Molecular Imaging

The definition of molecular imaging is a non-invasive
visualization characterization
and measurement of biological processes at the molecular
and cellular level in humans and in other living systems.

Prerequisites for In Vivo Molecular Imaging

This shows a schematic of some of the prerequisites
to in vivo molecular imaging.

First, this is a cell, and you have targets.
These are the things that you want to image,
and this can be an extracellular protein,
like a transmembrane receptor, cytoplasmic protein
or structure or something in the nucleus.

Then you have a targeting ligand,
which attaches specifically to the target of interest.
And this targeting ligand is coupled to a probe
that can be detected using an imaging strategy.
It could be a nuclear probe,
or in our case, an ultrasound probe, such
as a microbubble molecular imaging.

Research involves identifying targets,
then identifying suitable ligands,
and then testing these in appropriate in vitro
and in vivo models, optimizing the imaging systems.
And finally assessing the coupling
of the molecular probe with the imaging system.

Molecular Imaging Targets

This is just a partial list of molecular imaging targets.

This could include metabolism, blood flow receptors,
receptors, the expression of transgenes,
autonomic innervation, hypoxia, angiogenesis, inflammation,
cell proliferation, apoptosis,
or the trafficking of therapeutically delivered cells.

Targeting Ligands

Now, the targeting ligands
or the molecular probes are
what confers specificity to the target.

These can be antibodies that are attached
to an imaging probe
or the naturally occurring ligand for receptors such as
estradiol, testosterone,
or vascular endothelial growth factor.

The targeting ligand could be a peptide or a small molecule
or an after The concept of ultrasound.

Concept of Ultrasound Molecular Imaging

Molecular imaging is shown on this slide.

This is a schematic of an endothelial cell membrane,
which is expressing some sort of a disease marker, such
as a leukocyte adhesion molecule
as occurs in early atherosclerosis
or heart transplant rejection or other inflammatory states.

A regular microbubble has nothing on the surface
and will transit through the microcirculation unimpeded,
however, a targeted microbubble bears a targeting ligand on
a, on its surface, such as a monoclonal antibody,
which causes it to bind
or to adhere specifically to this disease marker.

And during ultrasound imaging, this will manifest
as a persistent contrast effect as opposed
to a transient contrast effects, which contrast effect,
which would occur when a microbubble just transits
through the microcirculation

Because The ultrasound contrast agents
or microbubbles are exclusively intravascular in location.
The targets that we have for molecular imaging
with ultrasound are endoluminal.
They're function specific,
and it's required that
they have low constitutive expression
and it's should be present in the microcirculation 'cause.
As I mentioned, the microbubbles are
intravascular and location.

Requirements for Targeting Ligand

What about the targeting ligand?
Of course, it must be specific for the endothelial target.
It must have high affinity for the target
form a strong bond,
and it must be able to be attached
to the bubble in sufficient numbers to maximize binding.

Formulations of Targeted Ligands

There's different formulations
of the targeted ligand, as I mentioned.

Antibodies are used, have been used predominantly
for proof of concept studies,
and these have high specificity, strong bonding.
They're large molecules,
but the disadvantage is that they're immunogenic,
which limits the human application of,
of antibodies for targeting.
Ligands. Peptides are good because they're small
and non immunogenic,
and lots of it can be attached to the bubble,
but the pe specificity in binding could be less.

And then another strategy for targeting ligand is
to use a naturally occurring ligand,
which has high specificity and is non immunogenic.

Requirements for Microbubbles

What about the requirements for the microbubble itself?

The microbubble must be stable during in vivo circulation.
That means it must remain acoustically active while
circulating, and also once it's bound to the target,
it must present the targeting ligand optimally
to the endothelial surface.

That means that the orientation
of the targeting ligand must be such
that it will cause the bubble to bind to the active site
of the target.

And often the spacer arm is required to optimize the,
binding kinetics.

Ideal Targeted Microbubble

What is the ideal targeted microbubble?

Well, the tar ideal bubble should bind quickly
and strongly to the target.
It must remain acoustically active.
Once it's bound, it must bind in proportion to the magnitude
of target expression,
and that way it can detect a range of disease severity.

The microbubble should be detectable as a signal
that's distinct from freely circulating microbubbles,
which constitute noise.

So therefore, typically when we do molecular imaging,
we image after the unbound bubbles have dissipated in order
to detect the signal that's specific for a bound bubble.

Clinical Applications of Ultrasound Molecular Imaging

Some of the potential clinical applications
of molecular imaging using ultrasound
are shown on this slide.

Because micro bubbles are intravascular agents,
we can use them to characterize endothelial phenotype such
as cardiovascular diseases associated
with endothelial dysfunction or inflammation,
or markers of neovascularization or angiogenesis.

Some of the clinical applications include early detection
of disease, such as early atherosclerosis, follow up
of treatment using molecular markers
as surrogate endpoints for assessing response to therapies
and targeted imaging
and delivery of therapeutic agents using ultrasound induced
microbubble destruction or so-called theranostics.

Inflammatory Imaging with Ultrasound

What I'm gonna do is just review some of the data
that have been reported on molecular imaging
with ultrasound in the realm of inflammatory imaging
and in the detection of an angiogenesis.

The initial studies for molecular imaging
with ultrasound were based on the concept
of a leukocyte binding to inflamed endothelium.

This is a schematic of a leukocyte and an activated
or inflamed endothelial cell
that is over expressing leukocyte adhesion molecules such as
ein pectin, vca, and, and icam.

Through interactions between the leukocytes ligands
and these receptors on the endothelial cell.
A leukocyte is led to tether, then slow down roll,
and then firm adhesion.

The tethering
and rolling is mediated by the selectin family
of leukocyte adhesion molecules
and firm adhesion is mediated by ica.

This is a intravital microscopic image
of an inflamed YL in red cream master muscle.

As you can see here, these are individual leukocytes
that are slowing down rolling
and firmly adhering eventually.

Here's a rolling leukocyte here,
and these interactions that slow these cells down
or mediated by these leukocyte adhesion molecules.

And so the idea was to design a microbubble
that would behave like these leukocytes and slow down
and bind to leukocyte adhesion molecules in
the microcirculation.

Targeting ICAM-1

What we initially chose
to do when we first started doing this work is
to target the adhesion molecule ICAM one,
intercellular adhesion molecule one,
because this molecule is over expressed in early
atherosclerotic lesions.

So this is a rat aorta.
This is an aorta from an APOE deficient mouse showing
early fatty streaks.

And you can see that there's colocalization
of ICAM one in the area with early atherosclerotic lesions.

So what we decided to do is make a lipid microbubble that
bears a monoclonal antibody directed against
ICA one on its surface.

And this is a approximately three micron bubble
that we conjugated an antibody
to the shell using biotin Aden bridging chemistry.

So we came in with a strep Aden, which attached
to the biotin on the microbubble shell,
and then we took a biotinylated anti icon monoclonal
antibody and bridge that to the microbubble by attaching it to the strep Aden.

Our initial studies were in vitro in which we took cultured
endothelial cells
and mounted them in the interior
of a parallel plate profusion chamber that had inlet
and outlet ports
through which we can perfuse the microbubbles using
well-defined wall shoe rates
because the fluid dynamic properties
of this chamber well characterized.

And then we could mount this chamber
on an inverted microscope
and visualize the interactions of the microbubbles
with the endothelial cells.

This is some of our initial work,
and we took fluorescently labeled microbubbles shown here in green
and attached either the anti ICAM antibody to the surface
or a control non-specific isotype IgG.

These are micrographs of normal endothelial cells
and activated endothelial cells that are lining
that parallel plate profusion chamber that I just showed.

As you can see here, activated
or inflamed endothelial cells that over express ICAM one
bind these microbubbles shown here
as green dots in large large quantities,
whereas normal endothelial cells minimally bind the
microbubbles and the non-specific control IgG microbubbles
do not adhere to the cells under either condition.

So this showed proof of concept
that these targeted microbubbles could bind into biological
surfaces over expressing the target.

We then took this to an in viva model
to determine whether these binding events could
be visualized.

So we took a red heart transplant model of acute rejection,
which is associated
with very fulminant overexpression of ICAM one.

This is an abdominal heterotopic heart transplant in which
the heart is planted into the abdomen of a,
of a another rat.

So the control group iso graft with no rejection is a Lewis
to Lewis strain heart transplant,
and the acute rejection group is a brown Norway
to a Lewis strain heart transplant.

This shows some images
that we obtained in the short axis view using ultra harmonic imaging.

Shown here on the left are delayed images
after injection of monoclonal
of microbubbles bearing ICA on the surface
or on the right microbubbles bearing non-specific IgG.

As you can see, the rejection heart
demonstrated strong persistent signal from microbubbles
that were directed against IAM one likely due to adhesion
of IAM one to the, to the,
the micro bubbles bearing Ike onto the endothelium,
whereas the control conditions did not show
strong persistent signal.

As you can see here, the rejecting heart is associated
with very strong infiltration
of inflammatory cells compared to the control.

And then immunohistochemical staining shows brown I
strong ICAM staining.

So this is an example of how this technology can be used
for the non-invasive detection of acute organ rejection.

Other Studies on Leukocyte Adhesion Molecules

Other groups have used similar strategies in which
leukocyte adhesion molecules were targeted in this case.

This is from Dr. Kaufman in Dr.
Linder's lab using a
VCA targeted microbubble VCAs over expressed in early
atherosclerotic lesions of a OE deficient mice.

And this is in a imaging high frequency imaging
of the aorta in these, my showing a strong signal
of persistent contrast enhancement from the V cam targeted
bubbles not seen under control conditions.

So this is an example of early atherosclerosis detected
using molecular imaging with ultrasound.

Detection of Ischemic Memory

I'm now going to talk a little bit about the role,
potential role of molecular imaging in the
detection of ischemic memory.

The clinical rationale for this is shown in this slide.

So the challenge of it's very challenging
to make the diagnosis
of acute coronary syndromes in the emergency department.

So the scenario is a patient that presents, for example,
to the emergency room with chest pain
that is maybe a little bit atypical of an acute coronary syndrome,
or it's resolved by the time they
present to the emergency room.

So they say, doctor, I had chest pain earlier this morning.
I didn't worry about it until now, a few hours later.

And so you need the challenges
to determine whether this is coming from the heart
or whether it's non-cardiac in etiology.

The ECG, the e kg is often non-diagnostic,
and the typical serum cardiac biomarkers are not sensitive
for detecting transient ischemia in the
absence of infarction.

And it takes several hours to become positive.

So in the acute setting, this could be on,
this is not that helpful.

Segmental wall motion on echocardiography may be normal,
even though there has been prior ischemia,
if the ischemia has resolved.

And importantly, in patients who've had prior myocardial
infarction, abnormal wall motion may not distinguish
between an old event versus a new ischemic insult,
which is what you want to detect.

So there are molecular signatures,
however, of true myocardial ischemia,
which if detected can help make the diagnosis as
to whether chest pain is due to cardiac ischemia
or a non-cardiac cause.

And in particular, pectin is overexpressed in ischemia reperfusion.

This is from a model of ischemia reperfusion
and pectin.
Overexpression was quantified on histology.

And as you can see here upon reperfusion,
there is a dramatic overexpression of pectin due
to release a preformed pectin from the cytoplasm of
the, the endothelial cells.

So what we decided to do was to make a bubble
that would be targeted to bind to pectin using syl X
as the targeted ligand.

SLU X is the naturally occurring tetra saccharide ligand
for p and e selectin,
and it's has some advantages in
that it's got low molecular weight, it can be synthesized,
it strongly conserved between species.

And we're able to also create a control saccharide,
which is Lewis X, which is not adhere,
which does not adhere to selectins.

So this is a intravital microscopic study showing adhesion of syl X microbubbles to inflamed endothelium.

So this is a schematic of a microbubble bearing syl X
and basically here,
this is TNF alpha inflamed es in rat cream master muscle.

And as you can see here, there is adhesion these bright dots
of fluorescent microbubbles labeled with SLU X
to the inflamed endothelium.

This is these are the control conditions showing non-inflamed endothelium with minimal adhesion
of bubbles, which are pseudo colorized here in red.

And this is inflamed endothelium with the adhesion
of the syl X microbubbles.

So we prove the concept that Sluis x microbubbles can adhere
to inflamed endothelium.

So then can, the question is,
can this be image in a scenario similar to what I mentioned
with a patient presenting with chest pain
to the emergency room?

So what we did was we took a rat model of myocardial ischemia reperfusion.

So we took a rat
and occluded the left coronary artery
to create ischemia.

This is a regular myocardial profusion image showing a risk area during coronary occlusion
and during reperfusion, there is a reflow
to the previously ischemic area.

This is non-targeted microbubble imaging.

So this is an example for of a patient
with transient ischemia.

In the absence of infarction,
after injection of microbubbles bearing silo X are targeted
to to pectin, you can see
that there is persistent contrast enhancement in the area
that used to be ischemic, indicating
that the tissue remembers
or has memory for the prior ischemic event by virtue
of overexpression of pectin
to which the microbubbles are binding,
whereas the control microbubbles do not have any
persistent contrast enhancement.

So not only does the technique show the presence of prime myocardial ischemia,
but it can also show you the extent of myocardium at risk.

So in this particular case, this is a large area at risk,
and so if the patient had this type of image, you would know
that not only had they been ischemic,
but a large portion of their heart was at risk.

Angiogenesis Imaging with Ultrasound

I'm just gonna finish up now with some work on the use of ultrasound molecular imaging
to detect angiogenesis.

The clinical rationale
for angiogenesis imaging is shown in this slide.

Clinical trials and therapeutic angiogenesis have been
disappointing in that the,
they have not shown the same robust results that were demonstrated in preclinical studies using the
traditional clinical endpoints.

It's been argued that perhaps one reason is
that the clinical methods currently in use are not sensitive
for detecting neovascularization
and patients may differ in their individual responses
to angiogenesis treatment.

Furthermore, neovascularization involves complex molecular
events in vivo detection
of which may demonstrate important molecular responses
and identify patients more likely to respond
to angiogenic strategies.

So the goal here would be to identify molecular markers
of angiogenesis as a surrogate for neovascularization.

In collaboration with colleagues at the University
of Pittsburgh who've identified a tripeptide sequence,
we developed a microbubble that appears to bind specifically
to tumor endothelium.

This tripeptide was found using a peptide display library
panned against mouse squamous cell carcinoma
derived endothelial cells.

And this is after intravenous injection
of fluorescently labeled RRL the tripeptide.

And as you can see here, there is perivascular
location in the sarcoma of a mouse,
and these are cultured cells that are stained
with a fluorescent RRL.

And as you can see, there is adhesion of the RRL
to tumor derived endothelial cells,
but not to control cells.

So what we did was we put this tripeptide,
RRL on a microbubble
and injected it intravenously into a variety
of tumor models in mice, one bearing Sarco clone c sarcoma,
and the other PC three human prostate cancer.

And we intravenously injected microbubbles bearing RRL
or a controlled microbubble bearing glycine,
glycine glycine peptide tripeptide sequence.

As you can see here, the tumor shows strong persistent
contrast enhancement after injection of the
microbubbles bearing the tripeptide sequence the specific for angiogenic tumor endothelium,
which is not seen under control conditions.

And for controlled tissue, we use myocardium showing
that there is no significant adhesion
of either microbubble type to normal tissue.

So this potentially can be used to detect tumor angiogenesis and
or responsive tumors to anti-angiogenic treatments.

Dr. Howard Young poi also has described
a method for detecting angiogenesis targeting
integrins that are over expressed during neovascularization.

And this is a model of rat hind limb ischemia.

And as you can see here, this is after treatment with FGF two.

This is the ischemic limb.

As you can see that there's intense angiogenesis after targeted imaging of
of a treated ischemic limb showing more intense angiogenesis
compared to the non-treated limb.

So an example here of how molecular imaging
of angiogenesis can be used to follow therapeutic responses.

Summary

So in summary, ultrasound molecular imaging is possible using microbubbles that can be targeted
to bind to function specific endothelial epitopes.

The pathophysiologic states
that can be ultrasonically image in vivo
that have been shown in terms of proof
of concept include infl, inflammatory states such
as heart transplant, rejection, ischemic memory,
and atherosclerosis, as well as neovascularization.

Advantages of Ultrasound for Molecular Imaging

There are strong advantages of ultrasound
for molecular imaging compared
to other molecular imaging strategies.

For example, there's relatively simple instrumentation
and it's portable and can be used at the bedside.

Unlike some of the more complex nuclear magnetic
or isotope based strategies,
there is no extravasation of the contrast agents so
that we're able to image
exclusive endothelial targets.

This is not a hotspot imaging technique,
and we don't have the problem of extra cardiac uptake
of the probes
because we can do simultaneous anatomic imaging as well
as molecular imaging so that we can get registration
of our molecular signals superimposed on the anatomic
information, which is not possible using nuclear methods.

Future Work

The future work that remains
for molecular imaging with ultrasound includes the need
to improve the signal
to noise ratio using various strategies such
as increasing microbubble binding,
using multi targeting strategies
or better targeting ligands.

Importantly, imaging systems also need to be optimized
to detect microbubble binding, perhaps using approaches that are more sensitive to detect signals
that are specific to bound as opposed to circulating microbubbles.

I think it would be important to
identify non antibody targeting ligands so that the immunogenicity associated
with antibodies can be avoided.

And importantly, we really need to make efforts
to translate molecular imaging into the clinical arena.

Another important area of future work is
to use these targeting strategies
to implement microbubble mediated ultrasound therapeutics,
such as drug and gene delivery,
and other strategies
for treatment using the unique interaction,
the unique behavior of microbubbles When presented in an
ultrasound field, I'd like to just acknowledge the people in the Center
for Ultrasound Molecular Imaging
and Therapeutics, both past
and present, who have been involved in the data
that I presented in this talk, as well as our funding
agencies.