My name is Bob Langer and I'd like to now go over the second lecture, which is Drug Delivery Technology: Present and Future, but I should briefly summarize what I went over in my first lecture. And, in that first lecture, I discussed the fact that controlled release systems offer long-term drug release with release rates primarily determined by the system itself, and I went over some different ways that one could achieve that, and I in particular went over how one might design certain polymer systems or pump-based systems to do that. Now what I'd like to do is actually go back in time and tell you how I got involved in this area in the first place, and then talk to you about some of the systems we developed and even some of the ones we're developing for the future. So, when I got done with my doctorate it was 1974 and I was in chemical engineering, and most of my friends at the time went into the oil industry, there was a gas shortage then and they had lots of jobs, but I got a lot of job offers from the oil companies but I wasn't very excited about it, and I was looking for a way to try to use my chemically engineering background to either help in education or human health, and I was very fortunate that Judah Folkman, who was a surgeon, offered me a job in his lab on something very, very different, but that I felt was incredibly exciting.
And I though I'd start out and just show you a picture, actually from the New York Times, in 1971, of Dr. Folkman's vision of how tumors grow, and what he proposed is that a tumor cell would somehow be created and it would grow to a 3-dimensional mass, and it would never get larger than say about a millimeter cubed, because it ran into a nutrition problem. Cells in the center would die because they couldn't get nutrients or get rid of wastes. Well, what he said is that somehow the tumor is able to solve that problem because the tumor would create a chemical signal which he called TAF, tumor angiogenesis factor, that would diffuse to the surrounding blood vessels which normally didn't do anything, but when the TAF was there it would cause them to multiply and grow, and grow right to the tumor, and that would cause a second phase of growth, which you see here.
And, in that second phase of growth, the tumor is vascularized and that solves the nutrition problem for the tumor. It gets bigger and bigger and ultimately can spread through those blood vessels, a process called metastasis, and eventually kill. Dr. Folkman's idea, which is I didn't realize but was very controversial at the time, was that it you could stop the blood vessels from growing, achieve anti-angiogenesis, maybe that would be a whole new way of thinking about stopping cancer. So, when I came to his lab in 1974 there was no such thing as an angiogenesis inhibitor, and he asked me to isolate, actually, what would become the first of these. How do you think about a problem like that? Well, we kind of broke it up into two parts. First, where could you find something that might stop blood vessels from growing, and one of the things that we thought about was cartilage, which is in your nose and your knee, and cartilage doesn't have blood vessels.
So, I was able to get some cartilage from the little rabbits we had in the lab, but I couldn't get that much. So then, I started thinking, you know, well where can I get more? …you know, and I found a slaughterhouse that had cows and I got some of their bones, but still I could only get a couple bones. So, I found out, where do all of the cow bones in the Northeast go, and they go it turns out to a slaughterhouse, to some meatpacking places in south Boston. So I made an arrangement with them to get all their bones and I'd bring them back to the lab and I would process them, meaning that I would scrape meat off of the bones, I'll actually just show you one. Here's a bone and if you look at the top of it, that's where the cartilage it. And so I'd scrape the meat off the top, which I did in this case, and then I'd slice off the cartilage, and then I'd put it through various extraction and purification procedures so that at the end of several years I maybe had 50 or 100 different what are called fractions that I wanted to study and test to see if they would stop blood vessels from growing.
But, that then brings up the second problem. How do you study something like blood vessel growth? And, if you look back at the history of medicine, one of the biggest challenges whenever somebody comes up with a new factor or substance is often finding a bioassay, a way to study it. And there was no such bioassay, really, for studying angiogenesis, so we had to create one. And, one of the things that we thought about was, as we started to think about creating them, was… one of the big issues was that almost everywhere you go in the body or any organism there are background blood vessels, so we wanted to find a place where there weren't, and so we thought about the eye of a rabbit.
And, it turns out that Michael Gimbrone had shown that if you put tumors, certain types of tumors like B2 carcinomas in the eyes of rabbit, they will cause, over about a 2-3 month period, blood vessels to grow from the edge of the cornea, the limbus, to the tumor. So, we thought we could take an ophthalmic microscope and actually measure the length of the longest blood vessel, but the problem was, if we wanted to now find an inhibitor, we had to also put the inhibitor in the eye and the inhibitors would quickly diffuse away. So, we thought a way to solve that is to have what we call a controlled release polymer that could take any of the things we isolated from cartilage, all of which were fairly large molecules, and deliver them to the eye and to the tumor and to the blood vessels over this 2-3 month period.
So, one of the big challenges, then, became to try to develop such a polymer… and they didn't exist, and in fact Dr. Folkman, he was on the board of the one company, ALZA, working in this area, and he went out to ask them if they could help us. But they said no, they said that large molecules can't slowly diffuse through solid polymers. It's kind of like saying, could any of us walk through a wall? In fact, the literature said similar things, that the use of polymer matrices has been virtually restricted to small molecules.
The only thing I really had going for me is I just didn't know any of that, so I went ahead and tried to do it anyhow. I experimented in the lab and I… kind of almost Edisonian-like… and I actually found over 200 different ways to get this to not work. But eventually, I was able to make little microspheres, those shown here and one shown here and then the other's cut in half, and we were able to show that by making these the right way we could actually get release, this is from a paper in Nature in 1976, for over 100 days, for really any molecule. And that enabled us to start to think about doing these bioassays and to do controlled release as well. Later on, one of the challenges was to get constant release, and we worked out some ways, using some engineering models where we could predict certain shapes or drug distributions, where we could get constant release, and here's an example of that.
When I first presented some of this work, people were very, very skeptical about it. I remember giving a talk at a major meeting in 1976, and I practiced this talk for many weeks before because I was a very young guy, I was a postdoc and there were all these very famous polymer chemists and engineering in it, and when I got done with the talk I actually felt I did alright, but what happened was all these older scientists, when I got done, they came up to me and they said, "We don't believe anything you said." They were just very skeptical that you could release these large molecules.
But what happens, of course, in science is the key is whether people reproduce what you do, and it turned out that over the next couple of years a number of groups did, and the question shifted to how could this happen. So, to understand the way this happened, I had a graduate student, Rajan Bawa, when I was at MIT, and we cut thin sections through the polymer with a cryomicrotome. Here, for example, is one of those thin sections. It's a 5 micron thin section of a polymer we used called ethylene-vinyl acetate copolymer. And, if you had a molecule that was 300 molecular weight or greater, it would not be able to diffuse from one side of this to the other. So, how could the molecules get through? Well, now we cut a second section. This section has a red drug in it, actually a red protein, myoglobin, and this is cut before any release has taken place.
And we see what we call, in this case, a phase separation. You see the red myoglobin chunks here, and then you see the white polymer here, and you see that throughout. So, this is what happened before any release. Now, let's say you released it for a year and then you cut a thin section. What you'd see is left behind… where the drug was, are these pores, and these pores are large enough so that molecules even millions of molecular weight can get through, but what happens is… we did a lot of serial sectioning and also scanning electron microscopy… and what happens is it turns out that these pores are interconnected, they have tight constrictions between them, and they're incredibly winding and tortuous, so it takes a really long time for the molecules to get through them.
One way, when I try to explain this to people when I give lectures around the world, is I sometimes say it's kind of like driving a car through Boston. Boston has, what we call in chemical engineering terms, which I'm a chemical engineer, is Boston and these structures have what we call a very high tortuosity. And, if you have a high tortuosity, that you can use to slow release down, and over the years our graduate students and postdocs have worked out ways to create all kinds of these porous tortuous structures and to even develop mathematical models to predict how to make these, and so you can make these last anywhere from days to years, or any time in between.
So now, we were able to go back and try to address the problem, the angiogenesis problem, because now we have these polymers that could deliver molecules of any size and we also were able to make these polymers in a way that would not cause irritation to the eye, which was also a big challenge. So this was the assay I mentioned we wanted to create. The tumor is there and the polymer, and what we did is we put different fractions, we probably did close to 2000 eyes, and when we did…
Most of them didn't work. There were all different fractions that we isolated and most of them didn't work. I should also almost all but one didn't work. And I'll just show you some pictures of what they look like. So, this is from a paper we wrote in Science in 1976 with this what's called rabbit corneal pocket assay. If you didn't have the cartilage-derived inhibitor, that's what I call CDI, if you didn't have it… over, this is about 9 weeks after the start of the experiment, you get a sheet of blood vessels growing from the bottom of the eye over the polymer to the tumor. You can actually see the tumor… where the tumor is it's a little bit cloudy. And, if you looked at this eye, or any eye like it 2-3 weeks after this, what would happen is it would be…
The tumor would be 3-dimensional. It would be out of the orbit of the eye… we sacrificed the animals before that, but nonetheless you see the rapid blood vessel growth. In contrast, if you look at the next panel where we put the CDI in the polymer, the cartilage-derived inhibitor, notice how the blood vessels are lower: they avoid the polymer, they don't grow into the tumor. This is at exactly the same time, and it turns out that about 40-50% of the time, the tumors on the right will never grow, whereas 100% of the time the tumors grew on the left, and like I say, we did hundreds, thousands of eyes over the years to look at this.
So, that actually enabled us to isolate the first angiogenesis inhibitor. It did a couple of important things, I like to think. One, is that we did develop an assay that people could use in the future for all future angiogenesis inhibitors. Secondly, I like to think that this really established there were angiogenesis inhibitors that were chemical, and they did exist. And then we had this first one. Now, what happened is of course it took… this was just the start. It took the work of many companies, particularly Genentech and others, to really move this field forward, so it's wasn't… you know, many years later. So, it wasn't until 2004 when the first angiogenesis inhibitor got approved, and this is just a list of angiogenesis… and it's not even a complete list, that have gotten approved since 2004. Avastin, which is a Genentech drug, is one of the biggest, most widely-used biotech drugs in history, but there are many others as well and they've been, as we can see in this slide, been used for all kinds of cancers. And, not just cancer, but many people have different…
Of what's called… an eye disease called macular degeneration, where you get blood vessels growing into the back of the eye causing hemorrhage. And, before this, the only way to treat them was to use lasers to do what's called photocoagulation. Now, you can use these inhibitors like Eylea or Lucentis or Macugen to actually stop the blood vessels from growing and even reverse macular degeneration. What's happened is angiogenesis, this whole area has become a quite large area. Now, about 20 million patients have been treated with angiogenesis inhibitors and the FDA has said that there are four kinds of ways of treating cancer: angiogenesis treatment, chemotherapy, surgery, and radiation, and sometimes these are used together.
But it also seemed to me that not only might this be useful… the controlled release systems for angiogenesis, but they might be useful… studies… but they might be also useful in their own right, for delivering all kinds of drugs. And, as a proof of principle, Larry Brown, one of my graduate students, just took a molecule, insulin, and again I'm simplifying this, it was actually his whole doctoral thesis, but he put insulin in these pellets, designed a certain way, and was able to get three months release of a fairly large molecule. So, we thought, both Dr. Folkman and I, that this might be, you know… really open up the door to all kinds of new delivery systems. So, I was working at Children's Hospital when I started this work and Dr. Folkman said to me one day, he said, "Bob, we should file for a patent on this." And, it's interesting, I'd never had a patent before at the time, and Dr.
Folkman actually said the entire Children's Hospital never had a patent at the time. So, we worked with a lawyer and filed the patent, and five years in a row the patent examiner turned it down and, you know, we felt he didn't understand it, but it really didn't matter. The patent examiner was the one calling the shots. So, the lawyer said to me around 1982, we started this process in 1976, the lawyer said to me, he said, "Bob, you know, you're wasting a lot of money for the hospital, you should just give up." But, I don't like to give up, so I was thinking, how could we convince the examiner, you know legally of course, that the patent… you know that this was novel. And I thought, you know, when I first started talking about this work everybody told me it was impossible, it couldn't work. I remember getting my first nine grants turned down. There was just this enormous skepticism about whether it could work, and I wondered whether anybody ever wrote anything down.
So, I actually did what's called the science citation search, meaning that I could go back to our original paper, which we wrote in Nature in 1976, and see who wrote stuff about it and what they said, and it was actually fascinating. I found a number of quotes, but this one in particular was very useful, and I'll just read this. It was by five of the leading polymer scientists in the world, and what they said, and they were describing this field is, "Generally the agent to be released is a relatively small molecule with a molecular weight no larger than a few hundred. One would not expect that macromolecules, e.g. proteins, could be released by such a technique because of their extremely small permeation rates through polymers. However, Folkman and Langer have reported some surprising" … that surprising word is a very good word for a patent examiner… "have reported some surprising results that clearly demonstrate the opposite." So, I showed this to our lawyer and he was very excited.
He said, "I'm gonna fly down to Washington and show it to the examiner." And he did, and the examiner said, he said, "I had no idea". He said, "I'll tell you what, I will allow this patent if Dr. Langer can get written affidavits from these five authors that they really wrote that quote." So, I did that. I wrote each of them, and they were all kind enough to write me back that they really did write that, and then we got this very broad patent issued, and that was the first one in the history of Children's Hospital and then the hospital would license that out to other people and, today, many companies have developed all kinds of products based on either the patent or these ideas, and these are just a few of them shown here. For example, if somebody has certain peptides that you might want to take, l ike leuprolide acetate… people have not figured out ways to give it orally or by skin patches because the molecule is just so big. If it's injected it's destroyed right away. So now, what happens is it's put in little microspheres, just like I showed you, that are injected under the skin and actually deliver the drug for four months.
And there are many other too. This is just pictures of different ones. There's systems that can deliver anti-schizophrenic drugs for several weeks. There's systems that can deliver drugs to treat alcoholism for a month, to treat narcotic addiction for a month, so you give the injections once a month, to treat type 2 diabetes, for where you give an injection once a week, and this is really just the start of this. There are many, many others that end up affecting the lives of tens of millions of patients around the world. So, so far what I've done is I've gone over, now, how you can take systems like this, deliver them at steady rates or maybe slightly decreasing rates over long periods of time. And so, what these systems allow you to do is it allows you to control the level of the drug and the duration of the drug, and generally these are given intramuscularly or subcutaneously.
But we want to even go further, and now I'd like to sort of turn to nanotechnology and even some of the systems for the future, or at least that I hope will be the future. So, could you actually make these… use a lot of the same principles… but make these even smaller, so that you could put them in the bloodstream so they'll be able to go around the bloodstream and find their way to particular cells that you want them to go to. How could you do this? So, what we published, this was about 20 years ago, it was one of the earliest papers on medical nanotechnology, is that the challenge is if you make a particle that you want to put drugs in, and you inject it into the bloodstream, almost always what will happen quite quickly is macrophages, cells in the body, will eat those particles.
So, what we had to do was figure out a disguise for those particles, the nanoparticles, so that that wouldn't happen, and of course you need to get past the macrophages. In a way, you can think of the macrophages as kind of like the guardian. If you could get past the macrophages then maybe you can get to the cells you want, if you can figure out the right other things to add to the nanoparticle. So, what we did is we made this disguise. We picked a substance called polyethylene glycol that we could add to the outside of the nanoparticles, and our thinking was is that takes up a lot of water, and if the cell sees water, well, it's used to seeing water and it doesn't eat water up.
So, that was our hope, that we could disguise these particles, you know, to do that. And then what we did is Omid Farokhzadb, who was a postdoctoral fellow in our lab, actually now is a clinician and associate professor at Harvard Med. School, took it even one step further. We not only put the PEG on the nanoparticles, but he put targeting molecules on that might go to a tumor, for example. Examples could be antibodies or aptamers or things that could target things.
I realize sometimes I don't always explain this perfectly, but I was fortunate that about a year or two ago Nova, the TV show, they came to our lab and they filmed some of what we did, and they made this video that explains it much, much better than I do, so I thought I'd… I've gotten their permission to use that video, so I thought I'd use it and show it to you because I think it explains pretty well how this kind of technology works. So, let me just go to that video. "He starts with a nanoparticle of anti-cancer drugs. That gets incased in a plastic that releases the drug over time. That, in turn, gets a special wrapping that disguises the package as a water molecule, to fool the body's immune system. And, last but not least, the address where it should be delivered, a key that will only fit the lock of cancer cells." I should say that a lot of the clinicians I work with tell me it doesn't blow the cell up quite that way, but I think it gives you an idea of what's happening.
And, then, what we did… we actually, Omid and I got involved in even helping set up a company that created a whole manufacturing plant to make nanoparticles, which was a huge challenge, and this is just a picture of that plant, and then moved it from test tubes to small animals to large animals to humans, where it is now. And, it's been interesting and exciting, the compound… one of the first compounds is one called BIND-014, which is basically putting Docetaxel, a common anti-cancer drug with some side effects, normally, in the nanoparticles.
And what happens is, this is a semi-log plot, but if you look at the red dots and the red curve, if you put the drug in by itself, it goes to zero very, very quickly. But, if you put just the same amount of drug in the nanoparticle it lasts for days, so it keeps pounding the tumor. The consequence of that maybe can be seen even better by looking at human pharmacokinetic data, and in particular the thing to focus on might be, let's look at the dose just to make these absolutely equivalent, of BIND-014 at 30 mg and Taxotere, the same drug, also at 30 mg.
Well, what you see when you look at the third panel, the area under the curve, is you could take the same drug but we've changed it dramatically. When you put it in the long-circulating nanoparticle, the area under the curve is 127,280. When you don't put it in the nanoparticle it's 512. So, you get something that's almost 250 times higher when you put it in the nanoparticle, so it's just pounding the tumor, and it's early yet but the consequence of this, this is from some papers we published in Science Translational Medicine, show that there's at least some hints of efficacy.
If you look a the top CAT scan, you look at the patient's lung before and then maybe 42 days after. If you look at the bottom, that's another example where you look at the lung before and 42 days after, and notice that the nodules, that are circled in yellow, go away after this treatment. Now, of course, what's happening is hundreds of patients are being done and we'll get a better feel for where this may work, where it may not work, but I think it's the dawn of a whole new era of using nanotechnology to deliver drugs. This is a small molecule drug. W e're also using nanotechnology in this form and in the form of different lipid systems, to deliver DNA, to deliver siRNA, to deliver mRNA, and all these things I think are a very, very exciting opportunity for the future and, again, I think the problems are still unsolved, but I hope that this is the start of solving some of them and bringing them into patients.
I want to mention one other idea that also might be somewhat futuristic, but I think also will be a part of how drug delivery can change how people do things. I was watching this television show a number of years ago about how they made microchips in the computer industry, and I thought when I watched it, you know, that would be a great way to make a drug delivery system.
Now, of course, I've spent 34 years of my life working on drug delivery systems, so somebody might think, you know, any TV show I say I might think that, and they may be right. But, I just want to show the idea I had. The idea I had when I watched the show was that maybe what you could do is make a chip, but rather than put electrical things in it, you could also put chemical things in it, and what we see in this chip, this is just a schematic and it's a cut-away where we're just looking at… the chip itself is fully whole and I'll show you some in a minute…
But, when you look at it, we have these wells where you could put active substances in. So, you could put different doses of the same substance in, or your could literally, theoretically, have what we call a pharmacy on a chip. You could put multiple drugs in and have them come out whenever you want, and they're really stored in these chips indefinitely. But, notice that the chips have a cover which looks like a gold cover here, it could be gold or it could be a platinum alloy, and they are hermetically sealed and the drugs are underneath them but, as I'll show you, when we apply, by remote control…
We can actually take those covers off and the drug could come out whenever we want to make it do so. Let me just show you some of the work that was done. I did this work with my collaborator Michael Cima at MIT, and we had a very good graduate student John Santini, and we made these chips using techniques that were never used in the pharmaceutical industry, but using techniques that we adapted from microelectronics. And here you see in the top two pictures, which is both a top view and a bottom view of one of these chips, and it's got something like 34 wells in it, and these are tiny little wells, but they don't have to be. They can be bigger or smaller, and they don't have to be short, flat chips. We've actually made sort of cylindrical chips that could be injected into the body and so forth, but just to give you a size idea, here's a United States dime.
Let me just show you how they work. So, here's a well, it's covered with the metal and you can see this, this is a scanning electron micrograph of it, it would actually stay in the body like this for years, but if you come along and just give one volt by remote control, in nanoseconds the cover comes off and you see that happening here. And when the cover comes off the drug comes out. So, this is from another paper we wrote in Nature, where we put different amounts of drug in different wells and the drug comes out at different times. This is in test tubes, in vitro. Along the lines of the pharmacy on a chip idea, we put multiple model drugs in and triggered release at different times and that's shown here.
But over time, what John and Mike and a little company, Microchips, that we were involved with did, is take this all the way from test tubes, to small animals, to large animals, to humans. And what I'm going to mention now might almost sound like space-age medicine, but we actually did it. What was done is you put the chips in the human body and you can communicate with them over a special radiofrequency called the Medical Implant Communications Service Band. It's been approved by both the FCC and the FDA, and sometimes people think about tampering, I mean… I doubt that that's going to be a problem, and that should be the biggest problem we'd face, but to that extent we even can have a special computer code that we built in that only the patient or doctor could know, if they want to change or administer the dose. Also, what we have, what we built in is in a bidirectional communications link between the chip itself and the receiver.
The receiver, by the way, could be a cell phone, it could be something like this, and it can give you all kinds of information like, did you take the drug? I have to admit, as I've gotten older, that's something I sometimes forget about, so, did you take the drug, the battery life, and so forth. Okay, let me tell you about the clinical trial that we did. We did 8 patients, this was done in Denmark, and thinking about what kind of trial we wanted to do, one of the things that happened as we'd send our grants in is people would keep telling us why our approach wouldn't work, and the biggest reason they told us it wouldn't work is you'd get what's called fibrous encapsulation around the chip and that would mean that molecules couldn't diffuse through that fibrous capsule and wouldn't get into the bloodstream.
So we felt, let's give ourselves a hard test to really see if they're right. Let's give us… let's pick a large molecule, and if that got through certainly we'd expect smaller molecules to get through. So, what we chose was parathyroid hormone, which is a large peptide, and it's used in osteoporosis, and we also chose it because we felt this is a place where someday, maybe, we could even make an impact. In the case of osteoporosis, women are supposed to take parathyroid hormone by injections once a day, but one of the additional problems with this is that the women don't do it. There's actually a 77% dropout rate of the women who have to take these shots. And you can't take it continuously, either, with the little microspheres. Continuous is bad because that'll actually cause bone resorption.
So, you really have to give an injection once a day apparently, and that just has been fraught with problems. So, what was done is the trial is a small office procedure in the doctor's office to do the implant, and the results were very positive. The women preferred this to some of the other methods. You got the same pharmacokinetics, I'll go over what I mean by that in a second and show you some of the data, with less variability, which may not be important in this case but could be important in others, and the three major measures of whether you're treating this disease are Ca, PINP, and CTX, and they were the same as daily injections.
Just to show you the data, on the next to final slide, the top curve is human data where what you see is data for the woman at day 60, 68, 76, and 84. Notice how the points really are pretty much superimposed on top of each other. It's very reproducible. On the bottom right, what you see are pictures of the chip itself. In this case what we did is we made these little chips and in them we also put electrical components, a battery, a power source, and even a computer program. What you can't see on these chips is on the back end of it we actually built in an antenna, we imprinted an antenna right into the back end of the chip, and that's how you can communicate with it. You can communicate with it by, depending on your device, it could be a cell phone, it could be something like this, and so forth. As you can also see from these pictures, the amount of fibrous encapsulation…
It's not zero, we definitely get some, but it's very small and obviously it's small enough so that it has no effect on the release rate of this large molecule and it's also, I'd say maybe 1/20th of what a pacemaker gets. We also did histology, that's actually taking sections of the tissue and seeing whether you got inflammatory cells. And, in the panels on [the left] we see that there are no inflammatory cells over the implant. So, it ended up being very safe, and effective, and now we're moving this project in at least three directions. One is we're making a two-year device.
Two, we've been working with the Gates Foundation, where they've been interested in a… for family planning in the third world, that you could have a contraceptive device that you could turn on and off whenever… a woman could turn on and off whenever she wanted. That can't be done with conventional technology, but with this you can, so we're actually designing a 16-year device that could be turned on and off whenever the woman wanted it to. And finally, what we're doing… one of my colleagues, Michael Cima, he's been putting little sensors in these chips, and someday our hope is we'll be able to sense signals in the human body and then tell the chip how much to deliver in response to those signals.
So, that's what I wanted to largely go over in this lecture. Just to summarize where we are, in the first lecture I've gone over advances in controlled release technology and gave an overview. Here I've given some examples of both current and possibly future technology. And, in my third lecture, I'll go over biomaterials and biotechnology and talk about how one might use… create new biomaterials for drug delivery, and also how one might use biomaterials to help lay the foundation for tissue engineering.