Conquering Cancer: Engineering Solutions to the Problems of Cancer
HAMMOND: Welcome back, everyone. We're entering our second session-- Engineering Solutions to the Problems of Cancer. I'm Paula Hammond. I'm actually a chemical engineering faculty member at MIT, and a member of the Koch Institute for Integrative Cancer Research. And as an engineer, I'm very excited about this session that we're about to enter because it represents how cancer cell biologists, chemists, engineers can get together and actually think differently about the problems of cancer.
Now, in this session-- the session before, we heard about the milestones that have led to significant advances in cancer, and which led to pathways between cancer biologists, engineers, and medical doctors to begin collaborative efforts. In this session, we will address ways in which engineers and scientists from across fields and disciplines come together to address critical challenges in cancer therapy.
Engineers bring a new set of tools, and a new way of looking at problems that are posed by biologists. And these tools and approaches include the ability to design functional materials, from a nanometer length scale, from the very smallest of things, down to the molecular scale, and combine clever and chemistry with physical phenomena, via smart delivery approaches, but they can also design up to micro and macro scopic lane scales to design new kinds of devices that can help us address cancer.
Ultimately, what's interesting is that engineering can also reveal things about biology that would be difficult to uncover without those tools. And we'll also address how modeling and simulation and understanding of engineering systems can help us address cancer solutions. So I decided to use just one very simple example to illustrate this. And this is actually some work that was done as a large collaboration between Mass General Hospital and MIT, and it involves investigators, which were-- who were interested in understanding how they could capture metastatic tumor cells in cancer patients who have already had cancer.
The idea is to be able to find the potential for metastasis, and then address it. However it's very difficult to find that one in a billion cells that is traveling through the bloodstream and is the metastatic cell that is targeted. In this engineering solution, a microchip was developed in which blood could be streamed through the microchip, and when this blood is streamed through, it runs past millions of these tiny posts, these micron scale posts. And these posts are decorated with antibodies that actually recognize a very specific component on the tumor cell surface. So these posts can actually bind to, and ultimately capture, those very few metastatic cells in the microarray.
Now what was interesting about this problem was that it required chemistry to be able to define what's on the surfaces of those posts, and then to determine how we can attach these anti-bodies to those surfaces. It required engineering. It actually required a very careful design of how these microfluidic channels were developed so that we could get fluid through without disrupting those cells, and yet, have a rapid throughput.
And on the other hand, of course, it involved very critical cell biology to understand how we can actually pick up these metastatic cells. Now the development's moving onward, and the team looking at this problem is interested in how they can capture those metastatic cells, and study them, and examine the nature of these metastases in patients.
Finally, a second example. This is simply a slide that I often use to describe how we can understand ways to generate materials that act as this kind of smart bomb for cancer that you saw illustrated in Bob Langer's video. We learned about these things because engineers who listen to cell biologists learned that cancer cells can overrepresent specific molecules on their surfaces, called receptors. And those large protein molecules will bind with or recognize other small proteins.
And we can use that as a way to target these smart bombs. So this is just a micrograph in which a biologist has labeled some of these receptors with these black dots here-- these are gold nanoparticles used to show the presence of these receptors on cancer cell surfaces. If we can understand that, then we can generate particles that are decorated with these receptors. We can understand how they arrange. We can actually use that as a way to address cancer.
Now I'm going to introduce the illustrious panel that we have, which is-- they're actually going to address several examples of how engineering and the life sciences converge to actually create, in some cases, new areas of discipline. Our panel begins with Sangeeta Bhatia. Sangeeta is the John and Dorothy Wilson Professor of Health Science and Technology and Electrical Engineering and Computer Science at MIT and a Howard Hughes medical investigator.
So in this case, we actually see an example of electrical engineering, biomedical engineering, and material science coming together. In this case, Sangeeta Bhatia is particularly interested in micro to nano scale platforms for understanding and diagnosing and treating human disease. She is one of the nation's most promising young professors in science and engineering, as recognized by the David and Lucile Packard Foundation. And she's also, not only worked in industry, but has been involved in two biotechnology startups.
Our second speaker is Dr. Joe DeSimone. And Joe is actually coming from UNC Chapel Hill to join us today. He is the Chancellor's Eminent Professor of Chemistry at UNC, and the Keenan Professor of Chemical Engineering at NC State University. As Bob Langer alluded to earlier today, Joe is actually the director of a center for cancer nanotechnology excellence. One of the very few in the nation that spearheads the efforts of material science toward cancer.
Joe has received several major awards, including the NIH Director's Pioneer Award, and a very critical and important award that we have here at MIT, the Lemelson MIT prize for invention. He's also the founder of the nanobiotech company, Liquidia Technologies. And he's a member of the National Academy of Engineering and the American Academy of the Arts and Sciences.
Finally, we have our own Dr. Doug Lauffenburger. And Doug is the fore professor of bioengineering and the chair of the Department of Biological Engineering at MIT. As the chair, he actually was also one of the founders of this department. And his philosophy about convergence of biology and engineering is one of the reasons that we're here today to be able to talk about how we can pull these disciplines together to address critical diseases such as cancer.
In his work, he has studied the fusion of engineering and molecular cell biology. In particular, systems biology. And he brings the perspective of the power of modeling and systems approaches to address cancer. Doug had served as a consultant as a scientific advisory board member for several companies, including AstraZeneca, and Johnson and Johnson, and Merrimack Pharmaceuticals. And he is also a member of the National Academy of Engineering and the American Academy of Arts and Sciences.
Finally, we have a graduate student represented, who will be addressing our questions. Rebecca Ladewski, who is a member of the Koch Institute for Integrative Cancer Research, and is very excited to help handle your questions today. When you address questions-- this is an interactive session-- so after the talks, we will be taking some of the questions that we've gathered from the audience beforehand.
If you have a question now, raise your hand, and we have graduate students and post-docs from the Koch Institute who will be looking for your index cards. And we'll pick them up. And we'll get them to Becky so that she can ask questions. I'd also like to remind everyone that after the session we will have lunch. And I believe it is in the DuPont Athletic Center. So after we convene, there will be lunch available. It is part of your registration. So I'm going to turn things over to our first speaker, Sangetta Bhatia.
BHATIA: Hi, everybody. So you may have heard the old adage that getting an education at MIT is like drinking water from a fire hose. One of our favorite educational tools is the pop quiz. So you have five seconds to answer this question. If I can get it up on the screen. Which of the following diseases claimed the most lives in 2002? A. HIV/AIDS, B. Tuberculosis, C. Malaria, D. Cancer. Worldwide. OK.
So in 2002, cancer surpassed all three disease categories combined. Actually, has persisted in claiming human lives. This is what we call a confidence building question. Given the nature of this symposium today. But my point really is that the face of cancer is changing. Over the next 10 or so years, 70% or more of these new cases will be identified in the developing world. You can imagine that the infrastructure that's associated with the diagnosis and treatment of these cancers is enormous. And we're really going to have to change the way we think about diagnosing and treating this disease to meet that challenge.
I would argue today that convergence of engineering and life sciences is one way to approach that. So let's just think about diagnosis for a second. So this is diagnosis here in Boston. This is a virtual colonoscopy suite. So you might imagine that typically diagnosis is done by a blood test, by an imaging test, or by a biopsy. And the infrastructure associated with this is enormous. Not just from the equipment perspective, but also from the trained personnel that have to administer the exams, and also interpret those exams.
So when one starts to think about how might we transform diagnosis for cancer, one attractive idea is to use this road map. So this is a road map of the computer revolution. This is 1 transistor to 100 million over the course of 50 years, of a set of scientists and engineers working together. And this is the revolution of miniaturization. So we and others have been thinking about how one can apply this miniaturization revolution to cancer diagnosis and therapy.
So I want to give you some examples. So here's an example applied to a cervical cancer screening, which is normally done by a pap smear. So what you see here is-- normally a pap smear is a sample of cells from the cervix that's observed under microscopy by a trained pathologist. Just the microscope piece of that whole process is $7,000 or more. One of our alum, an HST alum, Rebecca Richards-Kortum, who is now at Rice, has miniaturized this entire microscope into this $10 part. It fits on the end of a fiber optic. So what she's done by doing this is moved the process from taking patient cells to the microscope from taking the instrument to the patient. And all along the way, reduce the cost enormously.
So when you think about taking this diagnostic paradigm even further, what about the idea that one could do non-diagnostic tests, completely non-diagnostic tests? So this is a dream at the moment. But the dream would go something like this. That we could do urine biomarkers screening for cancer diagnosis. Just want to give you a snapshot of how one can envision that that could be enormously cost effectively done.
This is a new field that's emerging. It's called paper diagnostics. So this is a postage stamp size of paper. This device is three layers of paper. You can see there are channels here that are micro fabricated on double-sided adhesive tape. OK. And the liquid flows in this device by capillary wicking. And it turns out there's a fascinating field of microfluidics just related to how fluids move in paper, in these lateral flow devices. So you can envision that this is a modern cancer-based assay, much like a pregnancy test. OK. So that's a vision of how we might think about transforming diagnosis.
How do we think about therapy? So at the moment what we have at our disposal for therapy, as you've heard, is surgery, chemotherapy, and radiation. And patients need to be hospitalized during this time typically. This Is enormous, economic, and personal hardship as they go through this process. So we start to think about why this is and how one might be able to change it.
One of the reasons that this is is that most of the non-smart drugs that Jackie [? Lee has ?] alluded to this morning, most of the drugs in our cancer armamentarium are actually just poisons. They poison rapidly dividing cells. And, therefore, there are enormous side effects. If you look at quantitatively at what fraction of drug gets into tumors at the moment, it's a less than 1% of the injected dose.
So as you've heard from Bob Langer and others this morning, one notion is that if we could simply take these cytotoxic or poisonous drugs more directly to the tumor, we could increase our therapeutic effect and decrease the side effects. And one way to do that-- to try and do that is to use nanotechnology. So this is an artist's rendition of how this might work. You would administer an intravenous drug to this patient. It needs to be nanoscale for a couple of reasons.
First of all, it needs to escape the filtration organs of the body. So the liver and the spleen and the kidney are the natural filters of the body. And their job is actually to filter things out. So you need to escape those. And furthermore, you'd like to get into the tumor. And it's been shown by Rakesh Jain and many others, that you need-- to get into the tumor, you need things to be at the nanoscale. Smaller than 400 nanometers. Preferably smaller than 50 nanometers. So we need to be quite small.
So one of the things that's important about this process-- about taking the payload deep specifically to the tumor is that one would like to decorate the nanoparticle payload with some kind of molecular zip code. So you might ask how one would identify such a molecular zip code. This is the way we do it in our laboratory. We've been working with many years for scientist named Erkki Ruoslahti, who pioneered this method. He's at the Burnham Institute in California.
And this is a screening approach to identifying molecular zip codes that you would decorate these nanoparticles with that carry that drug payload. And in this approach what you do is you take a mouse. You grow a human tumor in that mouse. And you inject a library of viruses. These viruses are called bacteria phage. You inject a library of 10 billion different bacteria phage. And each bacteria phage has 400 copies of a different peptide sequence, a different zip code on its surface. So you inject these in the mouse, and then you then harvest the bacteria phage, the 55 nanometer bacterial particles now, that have homed in on a tumor.
And then you can just sequence them and find out what the coatings were that conferred homing when they were displayed in multiple copies on a nanoparticle that was about 55 nanometers in size. OK. And when you do this sequence, you can sequence them here-- just a free peptide has been delivered to this mouse. It has a fluorescent molecule on it. You can see that it's homing in on the tumor, and then the excess is coming out here in the mouse bladder. So this is a way to identify molecular zip codes.
What kinds of interesting particles might you put them on? These are particles that we developed with collaborators at UC San Diego. This is Michael Sailor's group. He's a chemist. And these are what we call nanoworms. So these are magnetic nanoworms. They're made of clusters of iron oxide or rust, and decorated encased in a coating of dextrin, a sugar. And on their surface, they carry those molecular zip codes that I just showed you-- the peptide sequences-- as well as a drug payload.
So we were interested in not just exploiting the fact that you could decorate these with a drug payload, but thinking about the fact that the particles are magnetic. They have some interesting properties. So one thing that you can do with a magnetic material is inductively heat it. So we designed these nanoparticles. So this is a cross-section now of that magnetic nanoworm that I showed you on the last slide. And now it's decorated with drug. And the drug is attached to the surface of the magnetic nanoworm with a heat sensitive bond.
And what happens then when you apply about 500 kilohertz of radio frequency electromagnetic energy, that bond melts, and the drug is delivered on demand. So the vision is that these nanoparticles would sit embedded, homed in the tumor, and that you would release the drug of interest at a particular point in the patient's course of cancer therapy. So here is this working in the lab. Here we're pulsing these nanoparticles with power. And you can see the drug is coming off in solution.
For those of you who have molecular biology backgrounds, you might be interested to know that the heat sensitive bond that we used was actually two complimentary strands of DNA. And of course, the temperature sensitive melting of DNA has been well worked out in the fields of microbiology and chemistry. So here are these particles now in a mouse tumor. And you can see, this is another point I'd like to make, that these are actually visible on MRI.
So this is another nice feature of the magnetic nanoparticle is that it can act both as a diagnostic, and here when you-- these are advancing on their-- and when you apply an RF field you can see that you get on demand drug release into the tumor. So this is an emerging field of nanotechnology that people are calling theranostics, therapeutic and diagnostic nanoparticles in one. This is just one of many, many sort of flavors of these that are emerging in the literature. OK.
So one thing that we'd like to do, in addition to release drug on demand, is actually get more payload to the tumor. And we started thinking about ideas for getting much more than 1% to the tumor, which even in the scenario, is actually how much we got there. We came up with this idea that we should cooperate. So here we have two different nanoparticles that have separate jobs. The first nanoparticles job is to act as a beacon, to act as a tumor homer, to have that molecular zip code, to go in and find the tumor, and it has really nothing else to do other than to just send a signal, I found a tumor. Here it is.
And the second nanoparticles job then is to carry the payload. Whether it's the drug or the diagnostic, or multiple different drugs. And so now here they cooperate, where the first nanoparticle is sending a signal, and the second nanoparticle is carrying the payload. What we'd like to do further is to have an amplification system where you have just a few nanoparticles that get in, but then you could recruit lots more nanoparticles with the payload.
So we designed a nanoparticle system that has these attributes. And it has, as the signaler, these nanoparticles here. So these are gold nanomaterials. They have a very interesting property that was first described by El-Sayed at Georgie Tech. And that is that they have shape dependent excitation with light. OK. So these materials have very mobile electrons on their surface. This is called a surface plasmon effect. And depending on the shape of the surface, you can get resonant excitation of that material. OK.
So as they get elongated, you get resonant excitation of these, or they act like antenna, for different colors of light. And in particular, what we want for human tissues is antennas for light in the near infrared. This is what we call the optical window. This is a nice wavelength of light that doesn't scatter and doesn't absorb. And you can all do this sort of experiment at home. Put your remote control right up next to your finger, and you'll see that red light penetrates tissues. OK.
So we'd like these nano antenna to be sensitive to this near infrared light. And they're going to act as our beacons. OK. So this is the scheme. They go into the tumor. We shine light that's near infrared. They trigger an amplification mechanism in the human body. In this case, in the mouse body. The amplification mechanism that we wanted to trigger in this case was to create a little bit of clot, a little bit of injury. This is a whole enzymatic cascade in your blood.
And now this second nanoparticle, its job is to be a clot seeker and carry a payload. So you send a signal within a nanorod. You amplify the signal with the human bodies-- or the mouse bodies' enzymatic cascade, and then you carry a payload with a clot seeking decoration. OK. And when we use these togetherr-- so this is how we envision that. You would inject the gold nanorods. Then you would inject the cargo, the payload particle, and shine the light. And trigger this amplification.
Does this work? So is there mouse experiments now? And so here what you're looking at are nanorods in these two animals. And this is a thermographic image of the nanorod heating in the flank of that animal in the tumor. And if you use a payload, in this case, which is an imaging payload, which are those magnetic nanoworms that I showed you before, you get 150 times more nanoworms in that tumor than you did before.
So for every nanorod that you get in, you get 150 diagnostic nanoparticles. Here we've made them fluorescent so that you can see them. But they're magnetic, so you can see them by MRI. If you use a therapeutic payload, if you use a standard chemotherapeutic drug, like doxorubicin, which is an FDA approved chemotherapeutic, you get 35,000 times more drug than nanorod. So for every nanorod you get in, you get 35,000 drug molecules.
And when you do-- this is a mouse model of breast cancer-- when you do a therapeutic study, you can see that when these nanoparticles communicate, that you can suppress tumor growth as compared to even the same two nanoparticles that are administered when they're not communicating or cooperating. OK.
So this is sort of hopefully a snapshot of our vision for the future. And the last thing I'd like to point out is just that I think that we have to keep in mind this challenge. This is what's called the global accessibility map. This was actually published by the European communities in 2008. And what they've mapped here, worldwide, is access to health care by how long it takes to get to a doctor. And if you just look here, it takes for over 2 billion people in the world, it takes a day or longer to get to a physician. So if one billion of those are in remote regions, and one billion are actually slum dwellers that are in our cities.
So we should start to think about advancing cancer technology and advancing diagnosis and advancing therapy. This is a huge challenge for us for the next generation of scientists and engineers to think about how to make these more affordable and more disseminatable. So I would argue that our future students have lots of work to do. So thanks for your attention.
MODERATOR: The next speaker is Joseph DeSimone.
DeSIMONE: Great. Thanks very much. It's great to be back up here in Boston. It's not quite golf weather up here yet, but it's good to be back. So our session titled Engineering Solutions to the Problems of Cancer. I thought I would sort of give you an overview-- and that slide's not showing quite like it is on the screen as it is up here. I wander if the guys in a back know that. So the screen up here looks very different than the one we're looking at here. And we can't even do a chalk talk either. Right. Well at least the four of you can see it.
AUDIENCE: Tell some jokes, Joe.
DeSIMONE: Yeah, I'll have to. Yeah.
MODERATOR: They're adjusting [INAUDIBLE].
DeSIMONE: OK. Great. Would you like someone else to go instead? They're working on it. All right. Terrific. OK. Great. So I have looked at thinking about choosing a couple of unmet needs in cancer. Certainly one related to delivery to poorly vascularized tumors as an area. The opportunity and need for preferential delivery and release of chemotherapies in an enhanced dosage range at the site a disease. I think a little about cancer vaccines as well as cancer preventions. And so at very high level, I thought I would just scream through these with you. And let me first start on the poorly vascularized tissue issue.
So certainly we're all very familiar with a variety of different solid tumors. There was a discussion earlier about those that have a terrific highway or access through angiogenesis. But there's a number of other tumors, especially pancreatic cancer, that is poorly vascularized. And so the opportunity for thinking about treating in a focal or local way with device assisted approaches. And thinking about an area of interventional oncology.
And as a little bit of context, I had a great fortune of partnering with Bob Langer, and we developed a fully bioabsorbable drug eluting stent for treating coronary heart disease. This technology, ultimately bought by Guidant, and now it's part of Abbott. And we have about 1,000 people now that have a bioabsorbable stent in them. And we just got CE mark approval in Europe.
But this introduced me to the whole area of interventional cardiology and the opportunities for catheter based delivery of devices. Separately, got involved in a technology using iontophoretic approaches for delivering drugs and particles to poorly vascularized tissues. And it's in the eye.
And a company up here in Massachusetts, EyeGate Technologies, and so thinking about bridging these two very different approaches of interventional oncology and iontophoretic approaches, we are developing a catheter-based approaches in an interdisciplinary approach that develops catheters. This particular catheter is a four lumen catheter electrode that is able to generate electric fields in combination with a counter electrode to use iontophoresis or electrophoresis to drive drugs into poorly vascularized tissues.
In some activity with small molecules, we began looking at pancreatic cancer. And in this particular case-- the animal model is a canine-- where we were driving drugs directly into the pancreas. The counter electrode's placed on like a Bovey patch on the outside of the animal. And we have the ability to drive a huge amount of drug, in this particular case, gemcitabine, directly into the pancreas with current. Then-- and shown in red without current. And this shown in green is the amount of drug that gets in through the standard parenteral delivery approaches.
So we have a massive increase in drug concentration and a key for us is all of the drug that we add is delivered locally. We see no gemcitabine circulating systemically. And so this is the opportunity for focal delivery of drugs. And not only small molecules, but also working on nanoparticles in a really interesting approach for thinking about biologics. And as was stated earlier, there's a very small fraction of biologics that actually get where we want them to go. And the opportunity of infusing a tumor with biologics using catheter-based approaches or fragments of antibodies would be really interesting.
And we've also now moved into small animal studies. And we've miniaturized our device. And we now have these four lumen iontophoretic delivery devices working with a variety of different animal models. Both seen a graph as well as gem models. So we're pretty excited about where this opportunity can go. Beyond just pancreatic cancer, which is our prime focus, we also are working on some transdermal approaches with the triple negative breast cancer as well as melanoma and a whole host of other cancers and a range of drugs that could be delivered focally in this manner.
So looking at some other unmet needs, the opportunity for systemic delivery and nanoparticles is a big opportunity. Our particular focus is-- can we harness the manufacturing tools of the computer industry, using lithography and offshoots of lithography, to make new medicines and new vaccines. And that's a big focus for our lab. And again, a lot of light space in bridging two very disparate fields. This is not to imply we're going to make babies in a clean room, but to bridge a couple of interesting areas.
And so we've developed a particle molding technique. We call PRINT. Particle replication in non-wetting templates. But basically it starts with silicon wafers that are patterned using standard lithographic techniques. We make molds of those. And then we, in a roll-to-roll process, we fill those molds with liquid, shown in red, that are precursors to medicines and vaccines. In a templated manufacturing approach, we're able to harvest those products. This is a classic roll-to-roll or film-based process where we can then isolate individual particles where their size, shape is defined by the template, and then a chemistry is defined by what we put in there.
And so this is very different than a kinetically driven self-assembly approach, where one often has challenges with chemical composition control. And so this has now been extended to a roll-to-roll process and just as important, as we had said, vision without execution is just a hallucination. This is now a GMP manufacturing compliant process. And so this is a nano manufacturing platform. Not a particular drug, but a platform that's now GMP compliant.
And so these are the range of particles that we're able to fabricate, where they are clearly achieving sizes and shapes that are quite unusual relative to kinetically driven objects, but also controlling the chemical composition in some really complex ways by controlling interfacial tension. And some of the things that we're interested in doing is trying to understand the interdependent role that, not only science plays, but shape plays, in a variety of different processes, either cellular uptake processes or bio distribution processes.
Understanding and engineering and antisocial escape, understanding endosomal size and shape based on the size and shape of the particle. In fact, some of these hex knots actually generate toroidal endosomes with a very high surface to volume ratio. We start seeing patterns of how macrophages will take up particles of different shapes to try to avoid the REF's, or both in the airway as well as in perenteral delivery.
So a lot of really interesting opportunities. And then combining that with unique chemistries, we can design chemistries that once particles are in the cell, we can take advantage of a unique cellular chemistry to might, for example, trigger the degradation of the particle to have a Trojan horse. And these are examples of green particles designed to take advantage of the intracellular pH. And that can then fall apart.
And once these particles enter cells-- it sort of looks like Bob's movie, a little bit-- but basically now, we take advantage of the fact that the particles are single molecules, because they're networks. As a network starts breaking down, we create molecules. A colligative property as ismodic pressure, water floods the endosome, ruptures the endosome, and dumps the contents of sitosol. And so we can do that with a range of chemistries.
PRINT is now just a tool to allow us to make particles of any chemistries that we want. And the way we think about things is we have rigid particles made out of materials that are similar to bioobservable stent or bioobservable suture, but now in a nanoparticle. Or we have soft carriers. And let me just walk you through some examples of this. Some PLGA-based particles.
PLGA is probably the most translatable matrix because the FDA has a lot of comfort with it. And these are examples of particles of controlled size and shape made using PRINT that now are loaded up with a chemo agent such as docetaxel. Because of the molding process, we can now adjust the chemical composition of these particles from 0% drug all the way up to 40% drug. So these are solid state solutions of PLGA and docetaxel in ranges of chemical composition that are actually very difficult to get to through other kinetically driven processes. And what that allows us to do then is look at toxicity, or site of toxicity.
These are the anti carriers shown here in black. And then as we increase-- this is plotted against drug concentration-- as we vary the drug concentration, we're getting to the point of almost increasing the IC 50 value-- or decreasing it by a factor of 10. And so dramatic improvements. And then when you go invivo, this is an SKOV xenograph model. We're actually seeing a classic depot effect through the EPR effect of up to 20 times more drug in the tumor relative to the taxotere approach.
And so this is about putting more drug on target using these nanoparticle therapies. And this is a non-targeted particle. So this is just taking advantage of the depot effect of size issues. And we're just beginning now to go into efficacy studies, and this is-- so this is some unpublished data, just for three doses of drug, both taxotere control as well as our PRINT particles. You see the piece scores here. But with taxotere, you can extend the median survival out to six days and then we have dramatic improvement once we start using the nanoparticle approach.
And so we're on to a whole series of studies now using this type of approach. So beyond the rigid carriers, we think a lot about the soft hydrogel carriers. And we're very interested in areas of Mechano biology and Subra Suresh, a former dean of engineering here, was a big proponent of this area and studied this a lot. And these are examples of particles that we've just published and PNA-- apologize-- it slid off the screen here-- but this is where we've systematically varied the modulus of the particles and their ability to deform as a way of navigating biological barriers much in the way that red blood cells do, and much in the way that cancer cells do.
And so these are our red blood cell mimics, same size, same shape, and same deformed ability as a red blood cell. And then we actually then go invivo, and we use intravital microscopy to monitor-- these are large particles. These are six microns. But they are now deformable. And we can look at the biodistribution and elimination half life, and as we drop the modulus down by dropping the amount of cross flanker, we're getting into circulation times that are now getting out to four days for a particle that's six microns. But we achieve this by lowering the modulus, and going very small-- or staying large but going very soft.
And so the strategy now is-- now that we understand the ability to dramatically increase circulation times with softness, we can now begin doing this with particles of different sizes, different shapes, getting down to 80 nanometer by 90 nanometer particles. But they're not rigid like the PLGA particles, but they're, in fact, much softer.
And doing this in combination with targeting, this is example of Herceptin, which is easy to conjugate on to these particles, with HER2 positive cell line. This is 100% isotype control, and then we systematically add Herceptin, and we see really good uptake in these cells, but the isotope controlled barely have any uptake. And then we go to HER2 negative cell line. And again, we have very low uptake.
So the idea is can we start combining these approaches. If you now can conjugate antibodies on the particles, you can now start getting on the order of 10 to fifth chemotherapy molecules per antibody. Very different than a molecular conjugate that might only have six chemo agents per antibody. And that's a great opportunity for doing that. So the challenge is-- these very soft hydrogels don't hold onto chemotherapy agents very well, and so we have to use a pro drug strategy.
And so here we use a silyl ether linkage group that allows us to conjugate molecules like camptothecin. And we can now conjugate these with different linkers depending on these substantial insulin silicon and you get very different release rates. So it's the interplay between kinetics of distribution combined with a kinetics of drug release. And we're working with a whole host of other drugs that we can now, in a plug and play approach, combined drugs deformability and targeting, and this is the direction that we're going.
One of the big holy grails in drug delivery is RNA interference approaches. And when we use rigid particles, we don't have very good bio distribution with some of those relative to the soft ones. This is some intertumeral delivery of SIRNA in a xenograph prostate model, but one of the directions we're going now is that we have a pro drug strategy with SIRNA that only relates to SIRNA in an intracellular environment.
And so we have these disulfide bridges that get cleaved, and you can see there's none of the SIRNA release from particles in PBS, but it presents a little bit glitothyon we can release the SIRNA. And so we can get some really potent ways of delivering SIRNA onto these deformable hydrogen carriers. And this is a direction and combination where we're trying to use some of the chemotherapy agents that might induce DNA damage, and some of the RNA interference that might inhibit DNA repair, for some combination therapies.
And then finally, let me just point to some cancer vaccine approaches for the nanoparticle work. Liquidia, our company, has got a product in a clinic today that's a PLGA based carrier for treating influenza vaccine. And it's basically combining your standard influenza vaccine with our nanoparticles, and we see a huge increase in immune response by using this combination approach.
Well now that we can do this, we're in a focus of adding different adjutants like agonists for TLR4, 7, and 8, and combining those approaches as a cargo with a wide range of tumor associated antigens. And so a combination of these nanoparticles, which are now GMP compliant, and then begin adding a variety of different antigens for addressing different cancers. And so that's a focus for us going forward in cancer vaccines.
And then the last thing I'll point to is inhalation. The PRINT particles themselves can address a number of issues related to inhalation products and aerosols. And we have a big focus now in engineering aerosol characteristics where we can use size and shape to change sedimentation rates, and actually get into particles that are auto rotating in low velocity airstreams, tailoring deposition in the airway, and now these particles are 100% drug. And we're interested in combining these in variety of approaches.
And the two that I'll just point to are some interesting things in the literature, where people are interested in focal delivery to the airway of chemotherapy agents. And so we could do that by combining that with our approach. And then there's even some really interesting things in cancer prevention where people are adding chemo prevention approaches for smokers and adding an anti-inflammatory agents. And so one could start thinking about combining these going forward.
So let me just end there. I think the perspectives that you have in the Koch Center of harnessing all these different disciplines is the real driver for innovation. And it's one that we try to strive for in North Carolina as well, even getting NC State students standing under the old well at Chapel Hill. It's kind of hard to do but they'll do that. And then the other vision, without resources, is a hallucination as well. It's a really privilege to have great funding from NIH Department of Defense, NSF, and our company at Liquidia. Thank you very much.
LAUFFENBURGER: My contribution will be to illustrate another dimension of convergence. And it's hinted at here in the title, Designing the Biology not merely of the Box. I hope you'll see what that means as we go along the way. You might think of it as looking under the hood and seeing what one can do inside the biology.
So the convergence that most folks are familiar with and has been pioneered by many individuals and many research and academic units here at MIT is taking the basic sciences of math, physics, and chemistry, putting them in the hands of engineers, whether they come from backgrounds in electrical engineering, chemical engineering, mechanical engineering, analyzing that science, making things out of it, and solving problems in medicine.
And outlined on the left here in some photos are some tremendous accomplishments over the past decades. Drug delivery we've heard some things about. Tissue bone implants. The new types of surgery. Artificial organs. These have been wonderful, wonderful examples of convergence of math, physics, and chemistry based engineering, having impact on the world of medicine. And as I said, pioneered by many folks in many places, but including MIT.
Now the dimension I want to add illustration to here is what I'll call an additional new convergence. And that's a new kind of engineering. That's not rooted specifically in math, physics, or chemistry, but, in fact, is rooted in modern biology. Phil Sharp mentioned the molecular biology revolutions and the genomic biology revolutions. These are the two foundational sciences now for a new kind of engineering.
We're still-- analysis is involved, and making things out of it is involved, but there are new scientific substrates here to be the things that engineers use their thinking and develop their tools on. And surely, major applications to medicine, but also has the prospect for impacting other challenges in society, such as environment, energy, and so forth.
So what does this mean? One way to think about it is that engineers are traditionally good at understanding complex systems, machines, circuits. We know how to analyze them, build them, make them behave in the way we want. Well, what we need to do then is get down to the mechanisms of biology, the tissues level, the cell level, the intracellular level, and view them as machines and circuits, but biomolecular complex machines and circuits. Nonetheless, using our same kind of thinking and aiming to develop the same kinds of capabilities.
So I'm going to show you-- today is just some early fruits of thinking about cancer biology in terms of biomolecular machines and circuits in the ways that engineers might think about them under the hood. I'm going to show you three examples, but all about one piece of biology, in order to make it easy. That we don't have to switch all sorts of different systems.
And this has been mentioned earlier, again in Phil's talk, in Tyler's, and Jackie Leese's as well, one major area of challenge in cancer biology is disregulation of a particular receptor kinase signaling system, EGF receptor system involved in many different types of epithelial tumors. And schematically it looks something like this. That this is a cell, and on the cell's surface are a host of different types of receptors that belong to this family. They get activated by a myriad of different ligands that are out in their environment. They set off cascades of biochemical signals inside the cells that regulate transcription of genes, metabolism, cytoskeleton, and all sorts of phenotypes.
So one can look at this, if you're an engineer, and say, this is what we know how to study. This is a highly multivariate circuit. There's multiple components inside the cell, outside the cell, interacting, and this is something to understand this and figure out how to intervene in it. We can bring our engineering systems analysis whether we think in terms of information processing or kinetics.
At the same time, mapped on top of all this circuitry, is another machine that this is another schematic of the cell-- could be the same cell-- and I had shown you these receptor signals around the surface that are activated by growth factors and so forth in their environment, and they regulate transcription and metabolism inside a skeleton. Well they're not just sitting there statically. All that while, they're moving around the cells to different compartments, interacting in different locations, and different molecular partners. This is a cellular machine that's mapped on top of that cellular circuit, and this can be analyzed, whether you want to think of this in terms of kinetics or mechanics or transport, engineers have thinking tools about understanding what's going on here.
So that's the basic background. So what if we take that background in terms of that biological system, and that thinking in terms of machines and circuits, and ask three questions that are important in cancer biology today. One is we have drugs for this system and many other systems, and they work for some patients and not others. Can we understand what patients a particular drug might work for? So in particular, we're going to talk about a drug that inhibits the signaling of this EGF receptor, draws a big x there, and presumably inhibits things downstream.
There's a whole class of these made by many companies. Iressa, you might have heard of. Tarceva, you might have heard of. This-- I'm going to talk about a project that involves many folks here, and also a company, AstraZeneca that makes some of these. And here's the notion that in a particular type of lung cancer, these kind of drugs can work just splendidly. But only in a fraction of patients. And you'd like to know which patients to use it on, because you've got a really good chance of it working, and you don't want to waste it on people that it's not going to work.
In work that was led by Dan Haber, who's with us today at NMGH, what was found was that there is some gene sequence mutations in the receptor that seem to be more associated than not with the drug working. Depending on the study, it might be as many of 90% percent of the patients for whom the drug worked, had the mutations. Might be as few as 30. What we'd like to know who those are, and we'd like to extend it beyond that.
What we brought as an engineering approach was to take the signalling biochemistry and cell biology going on in those tumor cells when treated with drug or not treated with drug and try to compare and say, can we develop a model for all those things going on? All the different growth factor receptor interactions. All the signaling biochemistry. The machinery of the trafficking moving around the cell. And we're going to determine the parameters that characterize the behavior of tumor cells that are sensitive to drug and tumour cells that are resistant, and we're going to find the key differences. And that's going to tell us what's different beyond the gene sequence.
Two things came out of this combined modeling and experimental exercise that had to do with bio chemistry. The receptor mutants bind to the drugs slightly differently and they bind to ATP, that's involved in the signaling slightly different. But the big surprise was a cell biology, a piece of the machinery popped out. That the other important thing for patients that would respond to this drug, their tumor cells is that machinery somehow got decoupled from the receptor. And so that the receptors never got taken inside and moved to other locations and actually got shut off. OK.
We went on and demonstrated this in a number of different lines. And this is the important part-- that the engineering quantitative analysis of the moving parts of the machinery, moving receptors around along with the circuitry of the cell signalling, identified a way to figure out whether or not the tumor cells would respond to this kind of drug or not. What's shown here is the lower this number, the more responsive the tumor cells are.
And what's shown on this axis is whether they are impaired in taking those receptors inside. And you can even take patients that don't have the mutations, if they have this impairment, the drug will work. You can take tumor cells that don't have the impairment, code drug with another drug that impairs it, and now the drug works. So we now have a strategy based on an engineering analysis of the complex machinery and circuitry that gives us a better idea of when this drug should work and actually how to make it work better.
A second example. What if we aren't content with the drugs that we have right now and we need some additional ones to complement? So same system. But we're going to ask the question-- are there some other targets that in certain cases might be useful for drugging? Well, this is now in breast cancer as opposed to lung cancer. And you're well familiar again talked about Herceptin in cells that over express HER2, which is one of the members of this EGF receptor family. And when they do that, breast tumor cells tend to be much more aggressive. There is this antibody, Herceptin, made by Genentech, that works something like the first case, works splendidly in many patients. But there's many patients for whom it does not work.
In this particular case, we asked the question, let's take the patients for whom it does not work-- is there something that's also deregulated about the cells by having this genetic amplification that could then be drugged and be effective for these cells when Herceptin is not? With Forest White using a very powerful new approach out of an all of chemistry mass spectrometry can measure in these cell circuits dozens scores, hundreds of activities of these circuit biomolecular activities at once. And what one sees is that for these tumors that over express this gene, all sorts of things have changed. It's not easy to just pick out one change and say, that's the one we should drug.
Nonetheless, you come in with engineering computational systems analysis and say, here's data that compare the tumor cells that over express this HER2 gene or don't over express the HER2 gene, and I'm going to measure dozens scores or hundreds of these changes at once. And there's many, many changes. And I don't know which one to pick. But if we use computational modeling to map them against the actual cellular behavior of how aggressive and invasive they are, a computation model can predict a handful that are most strongly associated with the aggressive behavior.
Interestingly, what comes out of these now are molecules that you wouldn't imagine drugging previously. There's one here you probably can't see because the color called SHIP-2. It was always known to be important for diabetes. People have been looking at that as a target for diabetes. This study with Forest White says that it should be important for right breast cancer and some follow up studies, not by our lab, but by others, have now validated as a target in breast tumors for which Herceptin did not work.
Last example. We've talked mainly about small molecules so far. Herceptin is an antibody. Antibodies are very powerful at interrupting a receptors interacting with the growth factors in their environment that activate them and turn on the signaling pathways. And many people, and many companies, are developing antibody drugs. Genentech, I just mentioned with Herceptin, well, is it as straightforward as, you find a receptor, you add an antibody, it blocks the receptor from binding to the growth factor, doesn't activate it, that's really all there is to it? No.
Let's ask the question-- can we find a more effective design principle for making antibody drugs that are better than the way that they're thought about today? This is a collaboration with Dane Winthrop, whose an extraordinary protein engineer, antibody engineer. And here's the notion. This is back up to that circuit diagram. And we can design the typical approach in the field as design antibodies that block the growth factor from binding the receptor, and presumably shuts off everything downstream.
Well, that's the theory. It really doesn't quite work that way, because, again, if you dig into the kinetics and equilibrium thermadynamics, about these kind of interactions between the growth factor, the receptor, the antibodies, who's competing with whom, there's actually a limit to how well you can shut off the cell signaling just by an antibody that blocks the log and bind into the receptor. There's a limit. You can figure it out mathematically by the right kind of analysis of the kinetic modeling.
And the problem is that even though you can block 99%, 99.9% of the receptors, there's still enough signal that can escape and keep the cells viable, alive, proliferating, aggressive, especially because the signaling circuits I told you about are very non-linear, and you can shut off the receptor to very, very low levels and still have very, very powerful signaling downstream. So many, many antibodies don't work even though they block the log and receptor binding very nicely. What else can you do?
Well, let's go back. What else is going on besides just that biochemistry step of the cell surface. There's all this endocytic trafficking machinery going on, and that governs how many receptors are actually there for the growth factor to bind to. So another engineering mathematical analysis and studies with experiments that figure out what all these different rate constants are and an engineering sensitivity analysis on the model to tell you what's the most critical part to interfere with says that whenever the receptors get taken inside, block them from getting recycled back to the cell surface. You're going to shut the whole system down.
Dane figured out that making an antibody that has three different binding sites for three different parts on the receptors, figured out how to evolve them so they'd have the right binding affinities in the different places on the antibody. Now when you put them on cells, they aggregate to the receptors. They prevent the recycling. They down regulate the receptors. Turns out now that these a tri specific antibodies can inhibit the growth of tumors that the best log and blocking antibodies cannot.
I'll show you here xenographed of a particularly difficult type to treat, the colon cancer, colon tumor. And shown here is the growth even with the best log and blocking antibody that you can find that's out in the clinic. Really has no effect. The-- I'll call it the which mab-- that blocks recycling rather than merely just blocking what's at the cell surface, shuts off the tumor growth dramatically, and they can prove in the tumor that it's because it's clustering the receptors, keeping them from going back to the surface. But you would have gotten at this just by looking at the straightforward biochemistry. You have to figure out the many different steps going on, do an analysis of it quantitatively, figure out what was really rate limiting, and what would be the most sensitive part of the biological mechanism for the antibody to aim at. OK.
Let me end by saying, MIT always strives for this not to be ivory tower and sterile, but to go out and have impact. We've heard great examples of that already. Out of this new type of vision of a biology based engineering, now there's affects in many companies, large and small, including-- I'll just mention this one-- Paula mentioned that I'm on their advisory board-- Merrimack locally is a company founded to identify targets, create therapeutics, create the diagnostics, and go into the clinical trials with precisely this fused molecular biology engineering analysis. We will see how they do. They have the things in clinical trials. If it succeeds, it will be a poster child for biology engineering fusion of this kind of convergence we're talking about. Thank you.
HAMMOND: Excellent. So now we're moving to the interactive part of our panel. All four of us will be addressing questions. And we have Becky who is going to present the questions to us.
LADEWSKI: All right. So getting started. You guys are one of the most engineering heavy panels we have today. Is the collaboration among biologists, engineers, and medical doctors hindered at all by a lack of shared background knowledge or a shared language?
HAMMOND: Joe, would you like to take a first stab at that.
DeSIMONE: It's massively-- bottlenecked. Yeah, no, the language issue is a big one and this is a contact sport. I think a lot of people said that earlier. And the vision you guys have about integrating everyone under one roof is really powerful. So yeah, you know I think scientists and engineers often are some of the best problem solvers. They just often don't know where the problems are. And physicians and others tend to own the problems and understand them in a way that, at least for myself, we don't really appreciate as much, and so it's that interplay is really, really, really important.
HAMMOND: I'd like to add one thing, which is interesting, with regard to the Koch Institute, which put biologists and engineers in the same building together. It's really had an impact on my own research group, which is a chemical engineering research group, in being able to parlay and to talk with some biologists, and now my students speak much better biology than I do, which is a big asset. It is a true barrier. If we are able to communicate with each other with greater ease, as Joe pointed out, the engineers will understand a little bit more about how we can use the methods that we have to address some of the challenges, and biologists will begin to understand how these systems can be manipulated. So I think it has a huge impact.
BHATIA: Yeah, I would-- that's fine-- go ahead, Doug--
LAUFFENBURGER: All yours--
BHATIA: I think I would add to what the two other speakers have said which is that I think that communication with physicians and patients is also really important, and trying to understand-- how do you do diagnosis at the bedside? Trying to talk to physicians about what's currently done. In the current health care climate, they actually don't have a lot of time to think about innovation themselves. So trying different ways, experiments of culture, if you will, to figure out how to have that conversation best, I think is going to be important.
We certainly at the Koch have now in clinical investigators within the Koch who are practicing oncologists, which is really important. Here at MIT, we have engineering programs through the agency program where students actually go and wear a white coat, take classes with medical students, learn to use a stethoscope with their own two hands. So I think all of those kinds of experiments of culture are going to be important.
LAUFFENBURGER: I would like to add one different facet, and that is I think beyond language, I think language might be one step, but at least from my perspective, the biggest gulf is that biology is a different type of science, at this point, still from physics and chemistry in the following sense. That in physics and chemistry, there's been decades, centuries of identifying, understanding the variables, how to quantify them, establishing what we'll call laws, so that when one sets up any kind of experiment, or a model, one has a great amount of confidence that one knows what the variables are, and that even more so, one can control most of the variables so that one is undertaking an experiment or developing a model and trying to make predictions, you've got variables under your command. You know what they are. You can control many of them. You can manipulate many of them.
Biology is very different in the following sense. Not that it doesn't follow the laws of physics and chemistry at its most mechanistic level, but that we're still woefully ignorant of what the variables are. OK. The key of genomics in finding the parts lists, and enabling folks in bio chemistry and physics now to dig into them deeply, and start to characterize those variables. That's what's finally putting us on the road to be able to do this sort of systems biology, biological engineering, and so forth.
But it's very different to develop an engineering predictive design viewpoint when one knows that you don't know what most of the variables are, that when you set up an experiment, when you set up a model, when you set up a prediction, what you know about the system is vanishingly small compared to what you don't know, and at the risk of sun like Donald Rumsfeld, that you don't know that you don't know. It's absolutely true.
It's absolutely true in this science, and that to an engineer is a simultaneously, it's uncomfortable, because we're used to engineering and physics and chemistry, we know and control a lot more, but simultaneously it is a comfort zone, because you are one of the parts of the engineering training is to think about complex systems with incomplete information. And that when one models and designs this elements of uncertainty and that's involved. So it's simultaneously a brand new type of thinking to have to jump into, but yet the training actually prepares you pretty well. But it's definitely more than language.
LADEWSKI: All right. So keep the questions coming. We have some more coming up now. OK. Next question. What's next in terms of engineering and biology coming together? What's the low hanging fruit or what's the thing that you would love to understand more that might really advance the field in new directions?
HAMMOND: Doug, would you like to take the first stab at that and you can pass it around?
LAUFFENBURGER: Yeah, I can do that. I think what's frustrating is, like in the examples I showed you today, it's encouraging that we think we can do some things that come from bringing engineering thinking into contact with, when I say fails to evolution as a molecular atomic biology, it's encouraging, but it's daunting because those examples are few and far in between. The areas of biology for which we know enough to do the kinds of things that I showed today, in terms of quantitative analysis, any type of reasonable prediction, is such a tiny portion of biology.
What we need to do is expand that level of information to many, many more realms than small specific areas, receptor signaling, maybe gene transcription, things like that. So I think the big challenge is how to take the data information coming out of the genomic revolution, and translate that into biochemistry and biophysics. That's what's the quantum step needed to really do engineering on it is how to take what's broadly known now in genomics, and translate that into the widest possible biochemistry and biophysics.
BHATIA: So I think in addition one of the big challenges and low hanging fruits is to think more about translation of the inventions and discoveries that are being made that are pouring out of many of our labs and mouse models. Just look at genomics, and you think about-- how do we turn those into patient therapies? We need to innovate around how to down select those and how to move them forward, and how to do de-risk them for commercial markets, how to get health care reimbursement, all those things, so that we can actually get to the bedside.
DeSIMONE: Perhaps, building on that a little bit-- yeah-- clearly major roadblocks related to cost, and regulatory approval. And you know, Sangeeta had some beautiful work in the paper diagnostics as one way of driving down costs, and value engineering things and being able to take new approaches in manufacturing would be really important to make things more cost effective. But, you know, the regulatory hurdles are significant. The whole med tech industry in devices is almost completely driven off overseas now, and you think about the biodurable stent being done all overseas, and there are some real major roadblocks out there. And so I think we need to address this from the patient point of view, and really think a lot about the regulatory and cost issues associated with these emerging technologies.
LADEWSKI: Another question that's kind of related to what Sangeeta was talking about was-- what is the balance right now between high risk, high reward research being done in cancer versus lower risk me two approaches? And does there need to be a change? And how can we foster that if so?
HAMMOND: Sangeeta, would you like to take that?
BHATIA: Well, I guess to echo the wise words of Bob Langer earlier, I think if there-- if you're not limited in resources, you would bet on all horses. Certainly. I think that the trick is we are limited in resources. So I think at the moment, people are thinking about balanced approaches. absolutely need high risk. But the trouble with high risk is not only that you don't know when it will win but you actually, if you look back sort of historically over the course of invention and discovery, you don't actually know where the wins are going to come from. You don't sometimes even know that something is going to be applied in a completely different area. So it might be high risk for x and for y. So then we absolutely need the bread and butter of understanding how these things work, and there's just a lot of heavy lifting in biomedical research. So I think we need both.
HAMMOND: One of the things I think that will actually help to narrow the field is when engineers actually do have meaningful and strong collaboration's with biologists and with people who are actually practicing or in a translational clinical positions, there is much deeper understanding of how to engineer a system. So that it has some meaningful impact. So I have experienced watching the numbers of people in my field and chemical engineering material science dive into the field, but those who go in uninformed are often the ones who just sort of do these less productive efforts, which lead to systems that aren't by all compatible systems that are not really effective in the body. And I think if we can begin to encourage our community to have deep and meaningful interactions across disciplines, we'll have less of the me too and more of a meaningful impact in this work.
LAUFFENBURGER: I think the thing I would add on that is that sometimes it's hard to tell whether something that looks incremental can lead to a dramatic benefit. What I'm thinking of is, in terms of some of these drugs that we've talked about, they work great either on one set of patients, but not on others. Or they will work great on this one type of tumor, but not on others. Or they work great if you can get the delivery right, but not-- what that means to me is that whenever there is a success landed on, even a tiny little segment, it tells you you've done something profound there that really has affected the pathological biology.
And it may look incremental to say, well, why did that work on lung cancer, and it has no effect on colon, no effect on pancreatic? What's the difference there? And that can look very incremental, because you're just kind of studying the same thing, but in a slightly different environment, taking it into invivo systems versus invitro and so forth. And yet, as soon as you find that one key, that's now the difference between colon and long-- all of a sudden, things that were very effective up there can rush into a dramatic benefit there. So I think one thing we have to be careful of is sometimes things that look incremental at a certain level, actually can end up having huge payoffs.
DeSIMONE: Sure. Just building on that comment a little bit too. Certainly we need the staying power in certain areas to unlock some of these subtle, but very important differences. But it's, at the same token, perhaps thinking about for the young people, the strategy is often all about being different. There's a business professor up here that talks a lot about that. And the opportunity of bringing very different approaches to areas often can yield new insights and new perspectives that often are distinguishing, and can, in fact, be areas where new resources can flow in ways too, because they are different. And there's a lot of value in thinking about that as well.
LADEWSKI: A lot of you talked about drug delivery in your different talks and getting drugs specifically to the site of the tumor. In fact, a lot of the buzz these days is about drug delivery. At present, do you see that the biggest hurdle in treating patients is designing new drugs or using the drugs that we have in more effective ways through delivery or other means?
HAMMOND: Joe, would you like--
DeSIMONE: Well, certainly drug delivery is a big issue and Rakesh Jain and others have continued to point out the big challenges and the clinic even points it out even further the clinical results. Getting drugs on target in ways that minimize systemic exposures is a big deal, I think. And with advances in imaging, we are detecting cancers earlier and earlier, when they're still localized. But many of our therapies are still systemic. And so the idea is of trying to think about focal delivery, whether sprint early or device assisted, I think there's a lot of value in that. And maybe even triage in drugs in ways that you couldn't do, because they have pharmacokinetic issues in access.
LADEWSKI: What do you mean by triage drugs?
DeSIMONE: Well, if you don't know the drug is what is going to work, because it's brick dust. And being able to get it into a cell, or get it into an organ, or if you have an antibody scaffold that might get renally cleared too quickly, and could you infuse it in areas that you can see if it has efficacy and whether it becomes a delivery issue or there's a fundamental biological issue and mode of action you can start differentiating these by combining some of these approaches.
LAUFFENBURGER: I think this is one of those questions where it's absolutely, not only easy, but right to say, both. I don't think anybody is content with the current tool kits of drugs in terms of the numbers of targets you can hit, in terms of how clean they are, in terms of understanding how to differentiate them, segment them into different patients. We've seen the tip of the iceberg with the Gleevecs and the Herceptins, and the Iressas and so forth. But there is a big iceberg underneath all of that that needs to be gone after, simultaneously with taking the ones that are on the tip of the iceberg, and delivering them better.
HAMMOND: Regarding that iceberg, when I think about the possibility of new drugs, I realize more and more as I'm attending talks from the biologist studying these problems that there are many ways to get at a cancer cell. And someone mentioned cancer cell metabolism earlier this morning in discussing Matt Vanderheyden, one of the faculty members at the Koch, just understanding how cancer cells work and what allows them to thrive gives you several new targets if you can understand them entirely. So that it's going to take some time for us to have that understanding, but as we gain it, we're going to gain new targets, and they could be more meaningful or more effective or be better for early onset, or for later onset, depending on the system.
DeSIMONE: Look at RNA interference. Right. I mean, this is a field that's had a lot of investment. It's on the cusp of not maintaining that investment, and people are backing out, and it's a delivery issue. Right? And so being able to advance, make advances there, is going to be really important for a massively important field.
LADEWSKI: Doug, I'll go ahead and direct this question directly to you. What are the different systems biology approaches to understanding preventing or treating cancer?
LAUFFENBURGER: Well, is Lee Hood here yet? All right. When you see Lee's talk, he will cover them all. So-- so that will be good. Make sure you come back for Lee's talk. It's accelerating, and I think the biggest challenge is what people would call data integration. So everything's moving forward really fast. The ability to sequence broadly whole genome individuals. The ability to transcriptionally profile different tissues. Get gene expression. Do it at the protein level. Now other metabolites and so forth. No single one of these types of measurement tells the whole story.
They're all moving really fast in terms of more coverage faster, cheaper, smaller. To put the whole picture together, the question is how do you integrate them? How do you figure out I got gene sequence information, what does that tell me about the metabolism going on in a cell? And do I need proteomic information to make the link between gene expression, and let's just say, metabolism rate or cytoskeleton? So I think Lee will talk about this afternoon that, in the end, you claim you need to measure everything at all levels then computationally put it all together.
Now that goes back to what I said before about being an engineer. That engineers have for decades and centuries designed things with incomplete knowledge. And all the types of things we've talked about this morning are based on incomplete knowledge. So one of the other keys of systems biology is how can you get the greatest insight and the greatest predictive ability even when you have a very incomplete knowledge. And that's a very interesting engineering challenge, but there are tools to do that. We know it's incomplete. We only are seeing slight snapshots. Nonetheless, there are methods to gain the greatest predictive insights out of that.
LADEWSKI: We have time for one last question. So you guys have all been working in this field for some time. Is there any technology that when it came out changed the way that you viewed cancer as a disease?
HAMMOND: Technology or discovery?
LADEWSKI: Would you like me to rephrase the question? Any discoveries of new processes in the body or anything like that.
BHATIA: So I think one has already been mentioned, which is RNA interference. But I think that the convergence-- we're using this word a lot-- but the identification of pathways at the molecular level was working in parallel with now the opportunity to turn single gene products off. And the knowledge that we've gained over the last 30 years, more and more about molecular machinery, now with the power to actually intervene with certain parts of the circuitry is, for me, enormously exciting. And there are now new technologies that are we hope going into patients, and we'll get sustained investment. And so I think it's an incredibly exciting time for those two things to be coming together.
LAUFFENBURGER: I might actually go back-- since I think I'm the oldest person here-- even to think--
DeSIMONE: I think you are.
LAUFFENBURGER: Good. I got that one right. Even transcriptional profiling, in the following way, that even under the most innocuous conditions, how many things in the cellular wiring gets changed by the slightest perturbation, and in cycling feedback ways. And I think one reason that type of appreciation is going to matter so much for cancer is that even with all the best drugs we have, the thing that comes back and hits you in the face again is resistance. They find a way. They find a way to escape. They find a way to survive by continually rewiring themselves even in response to the perturbations that you give them with a very effective drug. So just seeing how transcriptional profiling under any different conditions just show you the cells are just rewiring themselves in response to anything that happens. They're now a new cell. They're now a new tissue. You thought you were studying and treating this. As soon as you did whatever you did, it's now not this anymore. It's that. And that's a big part of the challenge.
MODERATOR: I think we'll bring this session to a close. I want to thank Paula and Becky and the other panelists.
So as Paula told you, lunch is now served over in the DuPont gymnasium. And if you don't know where that is, it's just a short walk across the courtyard. You can ask anybody out in the hallway. That's for those who are registered. For the presenters and the Koch Institute faculty, we have lunch elsewhere. And I can't remember where so--
AUDIENCE: McCormick Hall.
MODERATOR: McCormick Hall. So we'll meet those folks there at McCormick Hall. Everybody else, please be back here at 1:30 sharp. Thank you.