Nanotechnology and Spectroscopic Instrumentation for Image-Guided Precision Medicine (Shuming Nie)
1 Video View·Sep 5, 2023
Shuming Nie
07/29/15
BioNanotechnology Summer Institute 15
I'd like to use this lecture to share with you some of the things that we've been doing
in the last few years.
And Andrew, perhaps you will see that some of the things are a little bit different.
Certainly it will be very different from the last time when I spoke.
And one of the things I want to talk about is the combination of nanotechnology and advanced
instrumentation, especially photonic instrumentation, for image-guided precision medicine.
Image-guided precision medicine primarily refers to image-guided diagnostics and therapy
and surgery.
So as I go on, hopefully you will pick up some of the interesting ideas.
I think this area of new technology development for quantitative or precision medicine, especially
like Professor Bashir's overview, I pick up a new word, it's the engineering or technology
development for quantitative medicine, but I would add quantitative and precision medicine.
Because quantitative is precise.
Precise is quantitative.
Now let me start to describe the problem.
Intraoperative, during surgery imaging, and precision surgery.
One of the things in quantitative medicine or precision medicine wants to do is this
idea of precision surgery.
Cancer surgery, you know cancer is still a major problem.
It will be a problem for society with us, in my opinion, for the next hundred years
or even more.
It is a major health problem.
And of course other problems are cardiovascular, neurological, these kind of things.
But one of the things we're being focused on is to develop new technologies for quantitative
cancer studies.
One of the things I've worked with surgeons for the last eight years, I have to say that
I learned quite a bit, mainly from the medical needs, the medical problem point of view.
Some of the people say, my colleagues say, these surgeons make too much money and they
are too arrogant.
They don't want to work with us.
No, that's not true.
They are actually very eager to apply engineering technologies to their patients, but they want
to be the first one.
Especially they want to be the first one.
They're very eager.
They will go out of their way to apply new technologies to their patients.
That's their pride.
That's their accomplishment.
That's their job.
So one of the things I learned is that the tumor margin is still a big problem.
If you look at a pig lung, what you're looking at is a fresh, bloody pig lung.
We know that there's a tumor because we implanted a tumor.
But can you tell where the tumor is?
No.
And it's hard to identify where the tumor is.
And also, even if you identify the tumor, you make a resection, you cut it surgically,
how do you know you cut it entirely?
Is there any residue left?
And the other issue, the problem is that how do you know you have cut the lymph nodes?
Whether the lymph nodes contain tumor cells.
So I've worked with a number of cancer surgeons.
After their surgery, the first thing they do is to meet the family.
That's their happiest moment, is to meet the patient's family.
I'll be waiting outside the door.
But if you hear the surgeons during that thing, they always say, okay, the tumor is removed
completely and the patient is cured.
But as a matter of fact, in 40% of the chances, statistically, there's still residue tumor
left in the body.
And that residue tumor is going to come back as recurrence.
On the other hand, can you imagine that a cancer surgeon would tell the family, oh,
I'm sorry, I'm not sure 40%, there's maybe a 40% chance the tumor is not cut out, so
you should get ready to come back in about one to two years.
They cannot say that, right?
The hard truth, the fact is that there is 40% probability the same cancer is going to
reoccur and needs to do resection one more time.
And that's a major problem.
So we're trying to address that.
The idea of trying to address actually is like this.
Here's a real human pancreatic cancer specimen.
This patient being operated at Emory Hospital by one of our surgeons.
I think we believe we saved his life.
I'm not sure.
I don't have the scientific evidence to follow up, so you're too short.
Here's the idea is that the surgeon made the cut.
This is the tumor.
Then that's the margin.
That's where you cut it.
Don't look at that piece yet.
Imagine that piece is still inside the patient's body.
This is the piece that's first being cut.
Then he immediately brought this sample to us and said, can you use your spectroscopy
nanoparticles?
Look at it, whether this boundary is clean.
When we look at it, wow, it's all kinds of tumor cells.
The optical signals are all glowing.
We said no.
Then the surgeons are very action-oriented.
In a way, they like engineers, like to work on things.
Without asking further questions, he went back into the room, cut another piece.
That's the second piece cut.
Surgeons are paid more because they have to make decisions very quickly.
In a few minutes, they just have to do it.
And then he brought another piece to us.
This time, a smaller piece.
You look at it again.
Then we look at the margin here.
It's clean.
It's perfect.
It's clean.
No problem.
No more tumor cells.
If you look at this schematically, that's the first cut.
You cut here.
See, that contains a lot of tumor.
That would be the residual tumor left.
If we had followed his standard procedure without our optical spectroscopy detection,
he would stop right here with that piece left in the body.
If that's the case, for sure, the tumor is going to come back.
Then he made another one that's clean.
The patient is still alive.
Everything is fine.
I don't have the scientific evidence.
I think we made a difference.
That's one of the things.
The other thing that to broadly do this work, we developed an integrated, multi-channel
system for image-guided surgery.
That is pancreatic cancer.
This is the clinical trial at Emory Hospital.
We're also doing another clinical trial for lung cancer at the University of Pennsylvania.
This is specifically for the lung cancer because we're supported by this NIH director's transformative
award.
I think it is transformative.
Transformative award.
The idea is that we build this multi-channel system.
We have four channels, essentially.
The first one is anatomic, just the color video.
We're going to look at the anatomy.
The second channel is that we have a near-infrared laser beam, very much like this laser pointer,
but it's in the near-infrared because we want it to penetrate better blood and tissue.
Optical scattering for light is always a problem.
It's going to limit your penetration depth, but still in the near-infrared, the penetration
is better than in the visible.
For that reason, the red light is always ... It's called red because the wavelength is red,
because you can see better when it's fog and that kind of thing.
The third channel is a near-infrared, either fluorescence or surface-enhanced Raman channel.
It's coded in a pseudo-color.
This is not the real color because we can use a computer to code the near-infrared signal
in any color we want.
The fourth channel is a spectroscopy one because we can analyze that optical signal by a spectrometer
and then disperse the signal versus wavelength.
If you plot the intensity versus wavelength, that's called spectroscopy because if you
do this one, you can subtract the background so well, you can tell the spectroscopic shape.
Believe that when we do spectroscopy, we're able to detect just a few tumor cells in the
body.
Ultimately, academically, we want to do a single cell.
Clinically, I'm not so sure, but academically, it's very intriguing if you're able to just
identify a single cancer cell.
Not only cancer, maybe a stem cell or other kind of rare cell in the body by optical imaging.
That's a very powerful concept that could open a lot of things.
That's something I want to pursue.
Maybe another transformative, who knows?
I think it's transformative, able to detect and manipulate individual rare cells in the
body through this kind of instrumentation.
So then by computing software, we combine all of the things.
So that's the composite image.
So the surgeon will see this in the screen and can identify or highlight the tumor here.
You mark that region.
I think Irvine asked whether we have agents that can detect specific type of tumor.
My answer is that we actually want several types of agents or contrast agent nanoparticles.
One would be able to broadly detect any tumor.
The second would be able to target some of the receptors, tell us what kind of tumor
is it.
And still a third one would be able to highlight the nerves, because you don't really want
to cut the nerves.
Roger Tsien, UC San Diego, he published a paper at PNAS.
He identified some of these peptides able to bond to the nerves during surgery.
I haven't seen any follow-up paper yet by him.
Obviously, I've heard that he got some real health problems.
But even better, not only the nerve, some of the vital blood vessels, the arteries,
if we can highlight that during surgery, that would be beneficial.
So for that kind of application, you really need a multicolor, multiplexing strategy.
So the opportunities, in my opinion, for engineering is really limitless.
One of the key components that we've developed in the last few years is this handheld spectroscopic
device.
And we have a nickname, we call it spectral paint.
Essentially, it's very much like a paint-like kind of device with optical fiber.
So the surgeon or you, your hands, can hold it during surgery, scan through the cavity,
the surgical surface, you look for any residue tumor, or lymph nodes, or anything that is
suspicious on the surface of the cut.
So this actually overcomes the penetration problem, it's because during surgery, you
already cut it open.
You're looking at the surface you already cut.
So if there's any tumor cells, it's exposed.
So you can ask that, what happens if I have a tumor buried deeply behind the surface?
Well, you're out of luck, almost out of luck, unless that tumor nodule is within about a
centimeter in deep.
It turns out that by using 785 nanometers, if a tumor is buried about a centimeter or
so, we can still pick that up.
That I have to say is a lot better than our eyes.
Our eyes cannot even see a millimeter.
If you look at a blood tissue, we don't know anything behind it.
Because the eye is based on visible signals, the near-infrared is still a little bit better.
There are some people who claim that by using optical tomography, they can do as deep as
10 centimeters, maybe.
But the signal they get, the number of photons they get, is so small, they basically have
to wait an entire month to get an image.
Theoretically, yes, it's possible.
In practice, you don't get enough throughput, enough signal intensity for you to do this
in real time.
So at this time, a centimeter is almost the limit we can penetrate into the tissue.
Of course, you can look at it another way.
And inside this fiber optic pain device, the design is very simple.
Just two fibers.
One is to deliver the near-infrared laser beam.
The other is to collect the signals and transmit that back to a spectrometer.
So to initially design this, it cost us about $25,000.
And now, there's a company in China that can do this for $200.
Essentially, it makes it disposable.
It's a single use.
It's like your razor blade.
Because you don't want this one to touch one patient, then you transfer it to another patient,
the contamination, the infection.
So essentially, this thing can be developed as a single use.
And the actual system is not too complicated.
Essentially, if we start here, see, we deal with lung cancer.
We put the patient here on the table.
So that's the white field.
That's a lens.
Look at the entire surgical field of view.
And that's the spectroscopic pain device that lets you do spectroscopy.
And typically, we have two versions.
One version is that we would intravenously inject the contrast agent into the patient
after about four to five hours.
Then you can put the patient on the table.
Then the surgeon can look at the tumor.
Or we have another version.
The patient can drink the contrast agent in water in the morning.
Tell you a little bit about that.
And then you put the patient on the table, and you illuminate that with this light source.
The tumors would be emitting light.
So it's really a very interesting idea.
You design this agent to make the tumor glow in the dark, so to speak.
So the tumor literally emitting light.
And our goal of this is whether we can push the sensitivity and the specificity all the
way to a single tumor cell in the body.
That would be a science paper, I believe.
Maybe another science paper.
And the first version, the prototype of this is this one.
It's built by one of my former students, Michael Mancini.
And he has moved on to start a company.
And he's the VP for research of this company to commercialize this.
This looks, it's the prototype, a laboratory prototype.
With engineering design, with arts, that kind of thing, they can make this very appealing.
So that's not my job anymore.
My talent, I don't have the talent in the artistic design to make it curve, that kind
of thing.
I would say that, I'm going to come back to this again, that the devices, the instrumentation,
the photonics, optical, I don't think that's the bottleneck.
That's my personal opinion.
If we have the money, we have the resources, there are companies can do it very well for
us.
If we just give them the design, that kind of thing, so you see the Apple iPhone, the
iPad can be made in China very nicely.
That's a very complicated device, much more so than this.
I don't think the devices, as long as we have the right design, the design needs PhDs, really,
but not the fabrication.
What's really going to be the limiting factor in the future is the tracers, is the contrast
agent that we can inject into the body, or we can let the patient drink.
That would go to where we want it to go, bound to the tumor, and highlight the tumors.
That's going to be a major challenge to require molecular engineering and design.
I believe that's also where the most value would come from.
One of the things I think very exciting is this activatable fluorescent probes.
The idea is that if you have some of these organic dyes or fluorophores encapsulated
into nanoparticles, they will be self-quenching each other.
It's energy transfer among themselves, self-quenched.
They can circulate in the blood, but they will remain dark or silent.
Once they get in the tumor, either activated by lower pH, by hydrolysis, or by enzymes,
degradation, you activate the process.
This is a nanoparticle of endocyanin green.
You just encapsulate that.
In the nanoparticle state, it's almost completely quenched.
When you release that, the signal goes up by about 100-fold.
I think this kind of smart tracers or nanoparticles that stays silent, it can be therapeutic drugs
as well.
When they're circulating in the blood, they're not active.
But only when they get into your target, especially into the tumor cell, boom, they suddenly become
active.
I think a lot of engineering design will be required for that.
The other thing is this multifunctional, or people call it theranostic nanoparticles,
because it's therapy and diagnostic for integrated imaging and therapy for the same particle.
You have a targeting ligand where you would embed both the imaging agent and the therapeutic
agent together.
So that would give you ability to track where your therapeutic drug would go.
Sometimes a nanoparticle therapeutic is not working.
You don't know what's the problem.
It is often that the therapeutic drug is not being released.
They get into the tumor cell, but it's not released to the cytoplasm.
They're essentially being degraded inside the organelles.
So you do want some kind of information to see what's the problem.
One other thing I want to tell you about is this activatable nanoparticles, both for fluorescent
imaging and therapy, based on a type of copolymers.
I'm not sure you can see this very well.
It's a copolymer of this polypropylene acetate.
It's a very common polymer structure.
Then you conjugate that to polyethylene glycol, the PEG, make it a copolymer.
These polymers, because it's amphiphilic, are going to self-assemble into nanomicelles
on the order of 20 to 50 nanometers.
The interesting thing is that these kind of particles depend on the hydrophobic and the
hydrophilic forces that you engineer into it.
It can fall off or assemble very sharply.
The standard, that's the transmittance that measures the assembled state.
When the particles are assembled, they become cloudy, I'm going to show you.
When they are dissociated into individual polymers, they're completely transparent.
You can manipulate, control that assembly process with just a 0.1 pH unit.
The standard acid-base titration requires almost one pH unit.
Here you can engineer this by a cooperative effect.
It's a very amazing effect that a tiny change in the pH is going to trigger the entire assembly.
In terms of therapeutic drugs, if the organelle, if the pH inside the tumor cell is just slightly
acidic, exactly at the same value, the right value, it's going to trigger the sudden release
of the entire therapeutic drug.
This is a nanoscale phenomenon.
In my opinion, it's an emergent behavior.
You won't see this on the macro scale.
You won't see this on the molecular scale.
If you have multiple, hundreds of molecules working cooperatively, you start to see this effect.
It's like the hemoglobin, there are four bonding sites.
One site bonds, the other site has a higher bonding affinity.
Exactly the same effect.
In terms of the entropy, enthalpy, the kinetics, that would have to be another lecture.
Hopefully when we publish this you can read it.
This is a wonderful nanoscale emergent behavior you won't see otherwise.
Let me see, I'd like to say the pioneering work for this class of material was actually
published by Ji Ming Gao out of UT Southwestern.
I quote directly his figure, and you can see that you can engineer some of these nanoparticles
to be fluorescent, exactly at the pH you want.
So that's 7.0 or 6.5, and it's going to turn on and off by just 0.1 pH unit.
So here's our own work.
You can see that this is the self-symboled emergent nanoscale behavior.
It's cloudy.
It's because all these nanoparticles interact.
It's opaque, almost opaque, 7.4, 6.2 pH.
You see that 6.2 is completely cloudy, pH 6.1 is becoming almost clear.
If you go to 6.0, it's completely clear.
But in terms of the fluorescent, this is the light scattering.
If you look at the fluorescent, the fluorescent is actually a drug, it's doxorubicin.
How much is being released is that 6.4, you don't have that, it's still encapsulated,
and here it's getting released.
So with that, let me go back a little bit with another type.
So here is the endocyanin green, which we talked about.
In nanoparticles, the fluorescence can be completely quenched.
One of the good things about endocyanin green is that it's FDA approved for human applications.
This is a broad class of agent.
It doesn't target any receptor of the tumor.
It's only based on the leaky vasculatures, the EPR effect.
So this would be an agent to target a broad range of tumors, but not specifically binding
to any receptors.
By itself, it's very useful, because you can use this for many, many tumors.
And one good thing, again, it's FDA approved, when you inject it into the blood, it bonds
to the albumin immediately to form a 5-10 nanometer particle, then that takes advantage
of the enhanced permeability and retention effect.
And the bad thing is that it is an old dye.
It's been approved for roughly 30, 40 years.
So there's a lot of work trying to improve upon this.
Another class of dye is based on a...
This is essentially the endocyanin green, based on the EPR effect, the leaky vasculatures.
No targeting whatsoever.
No active targeting.
The other class I want to talk about is based on the metabolic activities of the tumor cell.
Very much like the glucose, the PET imaging, but in a different way, in an optical imaging
way.
It starts with a very small, chemically modified amino acid by the name of 5-amino levulinic
acid, ALA.
If you look at this, it's less than 200 daltons molecular weight.
This thing you can drink, the patient can drink, it's going to get absorbed through
the GI system so quickly, and after drinking this water, you can put the patient on the
table, operating table, one hour later.
The tumor already would be emitting light.
The reason the tumor is emitting light is that the tumor needs a lot of nutrients.
So this amino acid acting as a nutrient, so the metabolically active tumor has a higher
uptake rate for this chemically modified amino acid, just the same way as glucose.
More importantly is that when this ALA gets into the tumor cell, it is a co-factor in
the biosynthesis of protoporphyrin.
Protoporphyrin is the porphyrin without the metal, without the iron.
If you put the iron into protoporphyrin, it becomes hemoglobin.
So it's in the path to synthesize blood.
It turns out that if you don't put in the iron, the metal, this protoporphyrin is fluorescent,
fluorescent in the red.
You can excite it in the blue, 405 nanometer, perfect, that's the blue diode laser.
I think that's got the Nobel Prize, right?
405, just amazing, highly efficient.
This free porphyrin is going to shift the emission light to about 635 in the red, and
the human eye can even see that.
It turns out that most of the tumor cells have an impaired enzyme, or lack that enzyme,
to place the iron into the center of the porphyrin.
It turns out that to place iron into this porphyrin is so complicated, it requires the
enzyme itself.
A lot of the biological reactions, even though we think it's very simple, it won't happen
by itself.
It requires a catalyst, enzyme.
In this case, the enzyme is called chelatase, it's a chelating enzyme.
It's called chelatase, and without that enzyme, then the porphyrin is going to stay fluorescent.
And that's one of the reasons that the tumors light up.
This is very powerful, being used in brain tumor surgery, especially in Europe, and my
colleagues are doing clinical trials trying to get this clinically approved in the US.
The other thing I would say is that you see that targeting tumors, there are a number
of mechanisms.
One of them is through the leaky vasculature.
That's the vasculature, that's the blood vessel in the tumor, it has this little hole.
So therefore the EPR effect, by using endocyanin green, is based on that.
Once it gets into the tumor mass, the pH inside the tumor, in the tumor microenvironment,
is always slightly acidic.
By designing the pH-sensitive probes, we can target the acidic tumor environment of the
tumor very precisely.
I believe that Jimmy Gao, in Dallas, he got a lot of funding from the state of Texas.
He has a company trying to commercialize this thing.
We're also doing this.
I think this opens a lot of new opportunities to design these advanced nanoparticles, respond
very specifically to a particular environment in the tumor.
Another class of nanoparticles I want to talk about, the last class, is our surface-in-hand
Raman nanoparticles, I call SERS nanoparticle tags.
Hopefully you will see why I'm very excited about this.
The design of this is that if you have a metallic nanoparticle, for example gold, colloidal
gold, there's a lot of free electrons.
That's caused the surface plasma, the plasmonics, that kind of thing.
If you excite the free electrons with a light beam at the resonance frequency, and then
the electromagnetic field at the surface, too polarized, it can be a circle, can be
a rod, any shape, but on two opposite surfaces of the particle, you have a much more amplified
electromagnetic field.
If you happen to have a molecule residing at that surface, then this molecule is going
to experience a much amplified electromagnetic field.
Because the scattering signal, or spectroscopic signal of the molecule, depends on the local,
not the bulk, but the local electromagnetic field, and that's the way to concentrate
the energy to amplify the spectroscopic signals of that molecule.
It's very similar to the lightning rod effect.
You have a point, you can focus the electromagnetic field to such a degree it can ionize air.
That's why you see the glow.
It's basically the same thing.
It's the concentration of electromagnetic field at the tip.
And to design this actually requires a lot of time.
You have to make this biocompatible, so therefore we use a sulfur-modified polyethylene glycol
to put a layer of this biocompatible polymer, and then on the outside of the polymer we
have a peptide, or antibody, or other molecule that can recognize specific receptors on the
tumor cell, or the blood vessels.
This ligand on the outside provides the targeting ability.
You think this is so easy.
If you look at it now, it's not so difficult.
But actually, the surface-enhanced Raman effect, SERS, was first recognized by my old
boss, Richard P. Van Dyne, at Northwestern University.
I have to give the credit where the credit is due.
He first published a paper in 1977.
At that time, the idea of amplifying the spectroscopic signals by using colloids, at that time there
was no particle.
They don't call it a nanoparticle.
They call it a colloid.
In chemical engineering, colloid on surfaces is a well-established field.
There's no such thing as nanoscience.
So he used colloids and electrochemistry to observe this phenomenon.
It was so out of line with the thinking at that time.
He told me that his manuscript was rejected by several magazines.
First Science rejected it, Nature rejected it, then he submitted it to the Journal of
the American Chemical Society, JAX rejected it.
No one would believe it.
Then he submitted it to Analytical Chemistry.
I see a number of things there in Analytical Chemistry, rejected.
And finally, he has no choice, he submitted it to a Dutch journal, Journal of Electro-Analytical
Chemistry.
It was a Dutch journal.
Have you ever read anything in that journal?
No.
That got published.
It turned out the phenomenon was real.
So therefore, his best work is still this work.
That was 1977.
I just want to inject this historical thing a little bit.
It's interesting.
After a while, you look back, science, sometimes it's quite interesting.
1977, I still contact him, I said, Rick, you're still my old boss.
He says, no, I'm not old, but still your boss.
I said, it's fine.
He's a wonderful scholar, just very deep, published a lot of outstanding papers.
Well, see, 1977, exactly 20 years later, 1977, see, I was young one time, I was young.
I just started my faculty appointment at Indiana University with my first graduate student,
Stephen Emery, exactly the same name as Emory University.
I asked him, I said, is there any connection?
He said, no, there's no connection with Emory University.
Two of us, that time I insisted I should be the first author.
It's because there's some risk in this paper.
I don't want to ruin his career.
This is my, about 200 papers, so this is my only first author paper as a faculty member.
It was this one, because again, it was very risky.
We proposed, actually, we reported that the enhancement factors,
is 10 to the 14th to 10 to the 16th fold.
Whereas Van Dyne, in 1977, reported it's 10 to the 6th fold.
You know, 10 to the 6th fold, that upset a lot of people.
You could not get published anywhere except a Dutch journal.
That's 10 to the 6th.
20 years later, we're saying it can be 10 to the 14th, 10 to the 16th.
And we published in Science.
I told Steve that whenever it's wrong, forget about his PhD or my tenure.
We'll just go to France to join the cold fusion guys.
That time, you know, we found that we give explanation.
Actually, we have this data.
I want you, this is historical.
If I think about this, I feel very happy, still happy.
You see that was submitted 1996, November 1st, accepted December 1996.
The quickest acceptance I've ever seen.
And it was accepted without revision.
I tried to add a little bit.
They said, no, we have to publish this.
Because the two reviewers, actually three reviewers,
agreed with each other, this is such a wonderful work.
We don't know which two are not.
But the scientific, it really says that.
But however, you must let the scientific community
to evaluate, to judge it.
Because we had this data for a year.
I just would not dare publish it.
I asked Stephen Emery to show this is wrong,
to show there's something wrong with it.
So he spent a whole year almost in the lab.
Couldn't prove that we were wrong.
So then we got published.
And then just a year later, Louis Spruce.
Let me find where is Louis Spruce.
I want to show his picture.
Oh, here, Louis Spruce.
At that time, he quit his Bell Lab job.
Because the Bell Lab, they missed the internet.
They missed the wireless.
Still based on their old telephone,
they cannot compete anymore.
They're going to lay off some of these people.
And he started his faculty appointment
at Columbia University.
He was looking for something to do.
Then he found our paper.
And he was the first one, two years later,
repeated everything that we did.
Published his major paper in JETS,
Surface Enhanced Raman Spectroscopy
of Individual Rhodamine 6E, the same molecule,
molecule on large silver nanocrystals.
He shows wonderful results, including the blinking effect.
Because the light emitted by the particle
is no longer continuous.
It can on and off, on and off.
So I got really worried.
Because Louis Spruce, he has never been wrong.
He knows both theory and experiment very deeply.
Before this, his main contribution
was the quantum dots.
He synthesized the first quantum dot
and theoretically explained it.
At that time, Aleph Satos and M.G. Burwinde,
Aleph Sato at Berkeley, M.G. Burwinde at MIT,
were two of his postdocs.
So for that reason, when the Cavalier Prize in nanoscience
was first announced, he was the first recipient.
If there's going to be a Nobel Prize in quantum dots,
quantum nanocrystals, it's not going to be Aleph Satos.
Well, I should not say this.
But I think it's likely to be Louis Spruce.
So it turned out that he reproduced everything
that we did.
It's this work that confirmed our results.
Then I feel much happier afterwards.
And it still takes a lot of work to develop this.
In 2003, my student Bill Deering was with me.
And we developed a core cell structure.
We can coat the gold nanoparticles
with a layer of polymer or with silica, stabilize it in water.
And we'll just take one tiny particle out of that.
A single particle, we get a very nice spectrum.
So it is this kind of work that gave me the confidence
we should have no problem in detecting
a single cell in the body.
Because the signal to noise is absolutely there.
It's no problem.
Again, it takes a few years more.
In about 2008, we developed this.
Finally, this nanostructure is biocompatible.
We inject it into the body.
And then we inject through the tail vent.
And then we use a near-infrared 785 nanometer laser beam
sign on the tumor.
And we are able to pick up the surfacing aroma
signals from the nanoparticles that bond to the tumor.
So we published this in Nature Biotechnology 2008.
This is really the first paper showing
that this surface-enhanced Raman nanoparticles
can work inside the bodies.
Sometimes, if you look back, you will say,
it would be obviously it's going to be that way.
But even before that, before 2008,
there was a mental block.
No one would dare to inject these surface-enhanced
nanoparticles into live animals.
I don't know why.
We tried to do it.
Of course, it didn't work at first.
Because you have to play with the coatings.
But after we learned how to play with the coatings,
it worked immediately.
We detected beautiful signals out of the particles
bound to the tumor.
After we published this, everybody can do it now.
Everybody inject particles to get beautiful signals.
But why did they not do it before us?
There's a mental block.
So sometimes, you have to think about it.
Sometimes, it's not so difficult. You just
have to pass that psychological block to do it.
And now, we have even moved these particles
for image-guided or for spectroscopy-guided surgery.
See that?
We inject the particles into the tail vein.
And that's a tumor there.
We're trying to locate the boundary
by surface-enhanced Raman spectroscopy.
That's the handheld spectroscopic pain.
No problem.
And we found that most of the particles,
because they're fairly large, about 80 nanometers or so,
they're mainly accumulated at the boundary,
at the margins of the tumor.
That's excellent for image-guided surgery.
But it's not good for nanoparticle therapy.
If we look at the tumor center, see that line here,
the green, that's the center.
We see no particles.
So again, in terms of imaging and image-guided surgery,
it's a little bit easier.
As long as you can define the tumor margin, the boundary,
you will be fine.
But for therapeutics, you really have
to penetrate the tumor.
And here, it clearly shows, for these search nanoparticles,
because they are too big, they cannot get into the tumor
interior.
So if you want this one to carry a drug,
it will not be effective.
So we know now, most people agree,
for cancer therapeutics, you have
to reduce the particle size for deeper penetration.
And then, very quickly, we move to dogs.
So this is not using the search particle anymore.
This is using the endosyne E green, a complex with albumin.
And you see that this is a mid-sized dog.
This is the lung.
That's a tumor.
We have our agent, see that?
That's the optical imaging agent.
It's strongly correlated with the tumor.
And after that, this is another way.
Showing the dogs, this naturally occurring lung tumor
In large animals, we consider dogs are large animals.
But other people, when they hear large animals,
they think about horses and cows.
No, we tell them that.
In Georgia, we consider large dogs are large animals.
In Illinois, you have to deal with horses.
You can see, it's much larger than mice.
See, that's the chest open, everything.
So we are using naturally occurring tumors.
I should say that mice, these xenograft models,
mice, excellent for papers.
Publishing papers using mice models, perfect.
But they have very little relevance to the human disease.
We can routinely cure tumors in mice
by using nanoparticle drugs.
But these drugs would have no effect in humans.
And the animals, dogs, with naturally occurring tumors
is a much better model, but it's a lot more expensive.
Think about this, much bigger, a lot more expensive.
But this is something that we have to do.
So with this kind of data, we were confident just two years
ago to move into human trials.
Here shows that we can define the tumor very well
by using optical imaging.
We know exactly where the margin is.
And then we moved into clinical trial.
We have two clinical trials.
One is at UPenn, University of Pennsylvania.
The thoracic surgeon is Dr. Neil Singal.
And he's leading the effort in the lung cancer.
And here, you can see that by CAT scan.
And that's a tumor by PET imaging.
That's a tumor.
And this is the bright field under the room light.
You see a little bit.
And the fluorescence here is correlated.
Again, for this clinical trial, we cannot use the search
particle anymore.
We have to go back to the endocyanin green complex
with albumin, because that's the only thing available,
approved by the FDA for us to use.
The other type of tumor here is also in the chest.
That's the one.
And it light up very well.
And again, another tumor module that's here.
It's correlated with the optical data.
So I want to show you two cases.
You see that this is the actual human case.
It's a massive tumor of the lung.
This is the inflated.
The lung, you can inflate it.
Or you can collapse it.
It depends whether you pump air into it
or you let the air escape.
So this case is with the air.
You can see that it's very full.
And in situ, just in the test, in the chest, you do imaging.
It's a massive tumor everywhere, the signal.
And if you take a piece of the lung out from the body,
you collapse it.
It's like that.
Then you can image that again.
It's there.
So for this kind of massive, you don't need fluorescence imaging
to locate the tumor.
It's so obvious.
However, another case here, it's only some residue things.
That's the in situ fluorescence.
Only a few residue spots of tumor.
You do need the optical signals to locate it.
You can take that out.
It's mostly normal.
Just a few tiny spots are due to tumor.
So this is where the case that our technology
is going to save patient lives.
And this is open surgery.
It's really big open in the test.
It's very traumatic.
It's a lot of risk involved.
In the future, we're trying to work on robotic surgery
and also minimally invasive surgery
by using endoscopy or robotics with image guidance.
Because now, the surgeons are pretty good with their hands.
That's really what they are paid for.
They have a good pair of hands.
They can feel things very well.
Plus, their hands don't shake.
If you shake, it's a big problem.
But when you do this minimally invasive surgery,
you're not going to cut the chest.
What you do is open three holes between the ribs.
You put your instrumentation into it.
There's no way you can touch it anymore.
So the guidance, the instrumentation, the imaging
become even more important in this kind
of minimally invasive surgery.
So I think the need for technology
to determine medical application is really broad.
So this is some of the things we're trying to do.
And we already start a clinical trial at UPenn.
Now we're going to start.
We want to enroll 10 patients for minimally invasive surgery
like this.
And this is already approved by the NIH.
They're very happy to see us to do it.
And the surgical room is more like this.
Normally, a lead physician have two fellows work on it.
See, these are the three holes.
So this is the contrast enhanced image
guided cancer surgery under minimally invasive conditions.
So these kind of operations are also very good for the hospital
because they generate a lot of revenue.
This is a very expensive procedure.
They do it fast.
So lastly, I want to say that the miniaturization
of the devices, what we want to do eventually
is to have very tiny devices, not much bigger
than a millimeter or so, which can
be inserted into the channels in robots, in endoscopes.
And also, we want to deal with the pediatric,
the children's, small babies.
So recently, we've been able to miniaturize the devices,
the entire thing, to about 0.2 millimeter, 200 microns,
almost the size of thick hair.
Our thick hair is about 100 micron.
You can insert into these kind of endoscopes.
So let me just conclude that the device development,
there are two parts.
I think it's no problem as long as you have the money.
But the contrast agent is going to be a frontier,
going to continue to be a frontier in terms
of developing optical or other agent that can be injected.
So currently, the endosyne in green is FDA-approved.
That's a broad agent without any specific targeting.
In a way, it's good, because no matter what kind of tumor,
you can always use it.
Otherwise, it's not so good, because it really
doesn't target any specific receptors of the tumor.
And the ALA is pretty good.
It's approved in Europe, but not in the US.
Hopefully, that can be approved in the US soon.
And the search particles offer ultra-high sensitivity
and specificity.
For cutting-edge research, it's perfect.
We're trying to move that eventually
into clinical applications.
The activatable, or I call the smart,
these nanoparticles with cooperative emergent behavior,
I think, is very worthwhile.
And lastly, the cocktail, the combinatorial tracers.
Multiple of these things who can target different things
would be important, too.
And the work is really being supported by a number of things,
mainly through the NIH.
We had, for a while, the NCI Center of Cancer Nanotechnology
Excellence.
I think here, you have the training grant.
We have the research grant.
Our research grant has a training part as well.
And I'd also like to acknowledge the NIH for Bioengineering
Partnership.
And most recently, we have the NIH Director's
Transformative Research Award.
The key personnel include Dr. Sunil Single.
He's a thoracic surgeon at UPenn.
Andrew Yang is a pathologist.
James Provenzali, he's a radiologist.
And Kosta Hadjipanais is the brain tumor surgeon,
which part I did not talk about.
Lily Yan is an oncologist.
And Dongxin is oncologist.
And Mei Wen is actually my wife for quite some time.
We met in 1990, a long time.
She is a computer expert doing bioinformatics.
So she put all of the software together.
I cannot even write a single code of software.
I don't even know how to do it in BASIC anymore.
But obviously, they can do it.
She can put together professional-level packages
of apps, that kind of things.
And Father O'Kerry is our Chair of Medical Oncology.
And so here's just a few people we've been working on.
And lastly, thank you for the invitation.
It's been really my pleasure, and certainly happy
to answer any questions.
Thank you.

