IntroductionI am greatly honored to receive the MRS

Von Hippel Award, and I am also excitedto see biomaterials as a field recognized bythis award. In this presentation, I wouldlike to talk about how I became involvedin the field of biomaterials and about someof the research my group and I have con-ducted to understand and create biomate-rials that could be useful in various areasof medicine.

First, let me provide a bit of motivationfor improved drug delivery systems.Whenever, we, as patients, take drugs,whether by swallowing pills or taking in-jections, the level of the drug in our blood-streams usually starts out very low, thenrises, and then goes down again. We thentake the drug again, and the same thinghappens. At the peaks, those drug levelscould be toxic, while in the valleys, the drugmight not be effective at all. Yet almost alldrugs are administered in this problem-atic way. More than 100,000 deaths everyyear can be attributed to people taking pre-scription drugs in the correct way, oftenbecause of that very fact.1,2 That figure, bythe way, is four times the number of deathscaused annually by AIDS in the UnitedStates. We and others wanted to find amethod of administering drugs in such away that they would go into the desiredrange and stay there, something that very

few drugs do. Accomplishing that wouldbe a major step toward preventing sideeffects. Controlling drug delivery couldalso lead to new medical therapies, as Iwill later demonstrate.

Polymer-Based Drug DeliverySystems

During the past 25 years, a whole newfield of materials-based drug delivery sys-tems has emerged. These polymer-basedsystems have had a dramatic impact onsafe and effective drug delivery. One of theearliest developments, for example, was anitroglycerin patch used to treat angina.This patch, approved in the early 1980s, isa thin polymer system containing nitro-glycerin. Placed on the skin, it delivers thedrug over a 24-hour period.

Systems like these can be used also todeliver drugs for much longer time periods.One of the more familiar systems, the con-traceptive Norplant, was approved in theUnited States in 1991 and is now in use inmore than 50 countries around the world.Norplant is composed of little tubes ofsilicone rubber no bigger than matchsticks.The drug diffuses out from the center ofthese tubes for over 2,000 days, or fiveyears, after which the device is removed.

In 1980, controlled drug delivery sys-tems were virtually nonexistent. Last year,

the drug delivery field saw nearly 100 mil-lion people worldwide using polymer-based systems. Annual sales in the UnitedStates alone for these systems are about$30 billion; it’s probably double that on aworldwide basis.

Before I describe how these polymer-based systems were developed andevolved, let me tell you a bit about how Igot involved in this field. I received myScD degree in chemical engineering fromthe Massachusetts Institute of Technology(MIT) in 1974. At that time, somewhat liketoday, we were involved in an energy crisis.As such, many chemical engineers endedup getting lucrative jobs with oil compa-nies. I received four job offers from Exxonalone. One in particular made an impres-sion on me—it was a job offer from Exxonin Baton Rouge, Louisiana. At the inter-view, they told me that if I—or anyone, forthat matter—could improve the yield ofoil by 0.1%, it would be worth billions ofdollars to them.

While flying back home to Boston thatnight, I realized that I didn’t want to dothat. Rather, I wanted to see if there was afield where I could apply my chemical en-gineering background to some endeavorthat would help people’s health or educa-tion. I applied to different places for teach-ing positions, but didn’t get any of thosejobs. Then I started applying to hospitals.One of my colleagues suggested that Iwrite to a surgeon named Judah Folkman,which I did. Dr. Folkman later called andtold me about a problem he had beenworking on—figuring out how blood ves-sels grow in the body. He wanted to see ifthere was some way to isolate a substancethat could stop this blood vessel growth,something he called an “angiogenesis in-hibitor.” When I started working with him,angiogenesis inhibitors were only theoret-ical, and many people did not agree withDr. Folkman’s concept. Moreover, this areaof blood vessel growth was difficult tostudy. We realized that in order to solvethis problem, we would not only need toisolate an angiogenesis inhibitor, which isoften in the form of a large molecule, butwe would also need to find an assay. Wechose the eye of a rabbit, because there arenormally no blood vessels in the eye. Weput a tumor in the eye that would mimicwhat happens in the human body. Overtime, blood vessels grew from the edge ofthe cornea to the tumor. At that point, wewanted to stop those blood vessels, but todo that we needed a controlled-releasepolymer that could deliver the differentmolecules I was isolating. Since there wereno such systems at that time, I decided to tryto develop one. I can trace my interest inthe drug delivery field back to this project.

MRS BULLETIN • VOLUME 31 • JUNE 2006 477

Biomaterials forDrug Delivery andTissue Engineering

Robert Langer

AbstractThe following article is an edited transcript based on the Von Hippel Award presentation

by Robert Langer of the Massachusetts Institute of Technology on November 30, 2005, atthe Materials Research Society Fall Meeting in Boston. Langer was honored with MRS’shighest award for his “pioneering accomplishments in the science and application ofbiomaterials in drug delivery and tissue engineering, particularly in inventing the use of ma-terials for protein and DNA delivery, and for his achievements in interdisciplinary researchwhich have generated new medical products, created new fields of biomaterials science,and inspired research programs throughout the world.”

Keywords: biological, biomedical, microstructure, tissue.

www.mrs.org/bulletin

That was 32 years ago. It took 30 years,from that initial study in 1974 until 2004,before the first angiogenesis inhibitor, adrug made by Genentech called Avastin,was approved. Medicine, unlike many ofthe areas in which materials researcherswork, is extremely slow in terms of movingfrom concept to production. Yet, for thosewho follow that field, the angiogenesis in-hibitor has been a spectacular success andcould become the most successful anti-cancer drug of all time.

In 1974, one of my goals was to see if wecould release different molecules that Imight isolate from cartilage, the tissue I wasthen studying. A lot of the molecules werelarge. At that time, there were no systemsfor delivering these kinds of molecules forlong time periods and in a way that wouldbe safe in the human body. So the work Iwas doing was somewhat basic—to helpin angiogenesis research. However, fromthe standpoint of potential practical im-pact, something happened that I could nothave anticipated. That was the advent ofbiotechnology and genetic engineeringwhere, for the first time, it became possi-ble to create large molecules such as pep-tides and proteins in a commercial way.But these molecules faced serious deliverychallenges. Swallowing them did notwork because they were too large andwould be destroyed by enzymes or acid inthe stomach or intestines. They were alsotoo large to use in a patch. If you tried in-jecting them, they were quickly destroyedby enzymes. Delivering any of these mole-cules on a chronic basis would require away to deliver them in an unaltered form,and yet protect them from harm.

When we started this work, the conven-tional wisdom in the field was that it couldnot be done. Scientists felt it was not possi-ble to slowly release large molecules frombiocompatible polymers.

In VitroResultsAgainst this background, I began work-

ing in the laboratory to see if I could comeup with a way to make tiny systems thatcould deliver molecules for long times. Aftertwo years of experimentation, I had foundmany different unsuccessful methods.

Finally, I discovered one way to make itwork. My students and I took hydrophobicpolymers like ethylene vinyl acetate copoly-mer or lactic glycolic acid copolymer anddissolved them in organic solvents, usu-ally at low temperatures like �80°C. Weadded the proteins to them and slowlydried off the solvent. This is how we cre-ated small microspheres or even nano-spheres (Figure 1).

We published in Nature3 that you coulduse this approach to release molecules of

almost any size, from 14,000 MW to aquarter of a million MW. These moleculescould be released for more than 100 daysin vitro. Although the release rates werenot constant in those initial studies, welater developed ways to ensure constantrates. For example, we were able toachieve a constant release of albumin forover 50 days.

At first, our concepts were not well ac-cepted, and some of our colleagues werevocal in their lack of support. For example,as a postdoc, I had not realized that beinga member of the scientific communitymeant giving a lot of talks. In 1976, I wasasked for the first time to give a lecture—at the Midland Macromolecular Sympo-sium in Midland, Michigan. I was 28 at thetime, addressing distinguished elder poly-mer chemists and engineers. To make agood impression, I had practiced my talkdiligently for weeks. After giving the lec-ture, which I felt I had delivered reasonablywell, I hoped that these older scientistswould want to encourage a young scientistlike me. When I stepped off the podium,however, a group of them came up to me

and said, “We do not believe anything youhave just said.” That was my introductionto the way in which scientists sometimestreat other scientists. It was not until threeyears later that different groups began torepeat what we were doing. Then the issueshifted from “This can’t possibly occur!”to “How does this happen?”

To answer the question about how thishappens, Rajan Bawa, one of my students,began to use the cryomicrotome, a low-temperature cutting instrument, to cutthin sections through one of the modelpolymers we used—ethylene vinyl acetatecopolymer (Figure 2). We did a permeabil-ity study on a 5-μm section of one of thesemodel polymers and found that if the

478 MRS BULLETIN • VOLUME 31 • JUNE 2006

Biomaterials for Drug Delivery and Tissue Engineering

Figure 1. (a), (b) Polymer-basedmicrospheres for drug deliveryapplications. The diameter of themicrosphere in (a) is 1000 μm. (FromReference 6.)

Figure 2. (a)–(c) A 5-μm-thick section ofdepleted polymer initially cast withbovine serum albumin. Pores in (c) are3–100 μm. All micrographs are shownat the same magnification. See text fordetails.

molecules were 300 MW or greater, theywere not able to diffuse from one side ofthe section to the other. To understand themechanism of release better, we then put areddish protein, myoglobin, in the poly-mer matrix. We observed a phase separa-tion (Figure 2b). Next, we began therelease process.

After the release had gone on for a year,we cut a thin section and found that whatwas left behind where the myoglobin hadbeen were pores (Figure 2c). Since thesepores had not existed before the additionof the protein, they clearly had been madeby the formulation process. We then did alot of imaging and even scanning micros-copy studies and found that these poreswere interconnected. They had a lot ofvery tight constrictions between them andwere incredibly winding and tortuous,which accounts for the very long time ittook for the protein to get through. I some-times like to explain this long permeationprocess by comparing it to driving a carthrough Boston, which is often very tortu-ous. We then discovered that by adjustingthese pore structures, we could actuallymake the systems last anywhere from aday to more than three years, or any timeperiod in between.

In VivoResultsOne of my goals in doing laboratory

work has been to move beyond just con-ducting the work and publishing the re-sults to the point where I could apply thatwork to helping people. After our experi-ments worked successfully in vitro, wewanted to move to the next logical step,which was getting the drug release systemto work in animals. For this experiment, Iworked with one of my graduate stu-dents, Larry Brown.

The model system we chose first wasdiabetes. Our goal was to deliver insulin,which is a large molecule, to diabetic ratsin order to lower their blood sugar. Aftermaking the rats diabetic, we made smallplastic pellets, no bigger than aspirintablets, containing insulin and implantedthem into the rats. We found that the pel-lets could lower the rats’ blood sugar from400 mg/dl, which is diabetic, to 100 mg/dl,which is normal, for over 100 days. I willreturn to this concept later to explain howwe may be able to adjust the dosage. Mypoint here is that this experiment showedus that we could maintain biological activ-ities for long time periods in vivo.

Our subsequent goal was to see if wecould get this drug release process towork in humans. For the next 10 years, weexperimented with various systems in thelaboratory and published a lot of papers,but our work did not seem to be moving

forward clinically. At that point, I realizedthat in order to achieve our goals, weneeded to develop creative ways to workwith pharmaceutical companies. Al-though initially these companies wereonly vaguely interested in our work, MITeventually licensed patents to many ofthem. Also, by then I had come to believethat one solution was to start our owncompanies to develop these systems, andthat was, in fact, what we did.

Today, products based on the drugdelivery research done by us and othersare used by many people. Examplesinclude microspheres for treating advancedprostate cancer, endometriosis, precociouspuberty, dwarfism, schizophrenia, alco-holism, and other diseases.

For all of these cases, the drug maycome out at a constant rate or a decreasingrate. There is no way to increase the drugrelease rate or to regulate it after the re-lease process starts. So, we are alwaysthinking, how could we do better?

Drug Delivery MicrochipsAbout 12 years ago, I was watching a

television show on PBS about how thecomputer industry made chips, and Istarted thinking that chips might be a goodway to carry out drug delivery. I calledMichael J. Cima, who is one of my col-leagues and a professor of materials scienceand engineering at MIT, and asked for hishelp in developing a way to test the idea.We solicited the help of a student, JohnSantini, who began developing a drug de-livery chip first as a summer project andlater for his PhD thesis. Santini was able tocreate a chip prototype with tiny wells intowhich drugs could be placed (Figure 3).These chips offered myriad delivery possi-bilities. The wells might contain differentdoses of the same drug, or a variety of dif-ferent drugs —literally, a pharmacy on achip. The wells were hermetically sealedand then covered with gold. We found away to dissolve the gold by selectively ap-plying one volt of electricity in the pres-

ence of a small amount of chloride ion. Theconcept was to store these drugs for liter-ally any desired period and then triggertheir release by selectively dissolving thegold on any of the wells. To our great ben-efit, the gold had no toxic effects. James M.Anderson, one of our collaborators and aprofessor in the Department of Pathologyat Case Western Reserve University, hasshown that you can dissolve 16,000 timesthe amount of gold we used withoutharming animals.

One of the first chips that Santini madewas about the size of a U.S. dime. It had 34wells on the top and 34 wells on the bot-tom, shown as white holes on the devicein Figure 4a and the black holes on the de-vice in Figure 4b, with each well capableof carrying a different drug or differentdoses of the same drug. This early chip isjust one example of the many differenttypes we have made. We have fabricatedchips that are thin, like our early chip,chips that are thicker, and chips shapedlike cylinders. We have made chips com-posed not only of silicon and gold, butalso of many other materials. My formerstudent Dr. Amy Grayson has made chipsout of totally degradable materials, for ex-ample.

As I mentioned, the chips work by re-leasing the drugs that are stored in thewells. When we apply one volt selectivelyto a single gold-covered well, in about 10seconds the gold begins to dissolve andthe drug beneath it is released.

Santini experimented first with releasingthe same drug in different amounts and atdifferent times. He put varying levels offluorescein in different wells and triggeredthe release at intervals: one and a halfdays, two and a half days, four and a halfdays, and five and a half days (Figure 5a).

Then, Santini tested the pharmacy-on-a-chip concept. He triggered the release ofone drug from one well at 25 hours, adifferent drug from a different well at37 hours, the same drug as that released at 25 hours but from a different well at50 hours, and that same one as that re-leased at 37 hours but from a differentwell at 60 hours (Figure 5b).

The times and levels in these experimentswere arbitrary, but they demonstrated thepossibilities offered by the chip releasesystem. Santini now heads MicroCHIPSInc., a company that develops these chips,and his firm is continually advancing theircapabilities. They are currently conductinganimal trials on chips that can be controlledremotely using telemetry. After implantingthe chip in the animal, researchers applyradio frequency in a specific way that canremotely open any of the chip wells, in aconcept similar to a remote-controlled

Biomaterials for Drug Delivery and Tissue Engineering

MRS BULLETIN • VOLUME 31 • JUNE 2006 479

Figure 3. Prototype of a drug deliverymicrochip.

For example, in 1967 at the National In-stitutes of Health, some clinicians and en-gineers wanted to make an artificial heart.They started by asking, what object has agood flex life, like a heart? The answer theycame up with was a ladies’ girdle. Theythen determined that since the girdle wasmade of a polyether urethane, they wouldmake an artificial heart from that material.Today, 39 years later, we find that theartificial heart is still made of that samematerial—polyether urethane. Yet, whenblood hits the surface of the artificial heart(the ladies’ girdle material), it can form aclot. That clot can then go to the patient’sbrain and cause a stroke, which could re-sult in death.

Similarly, dialysis tubing was originallymade of sausage casing. The vasculargraft, which is an artificial blood vessel,was developed by a Texas surgeon whosearched for possible materials in a cloth-ing store, based on what fabric would beeasiest to sew. He chose Dacron. Of thetwo materials chosen for breast implants,one is a lubricant (silicone) and the other isa material used for stuffing mattresses(polyurethane).

Medical Materials Designed to Order

Against this background, we and othersbegan thinking that we needed to find amodel for solving medical problems otherthan to search for materials in everydaysettings. As a chemical engineer, I believedthat researchers could take an engineeringdesign approach, asking the question, whatdo we really want in a biomaterial from anengineering standpoint, from a chemistrystandpoint, and from a biology standpoint?Having answered that question, we couldthen synthesize the materials from firstprinciples.

We started by studying a particularclass of materials—synthetic degradablematerials. Initially, the only family of poly-mers approved by the FDA was suturematerials made from polyesters. They dis-played bulk erosion, which meant thatthey started out as a solid, then got spongy,and then fell apart. Instead of distributinga drug uniformly, these polymers couldbreak up or burst, releasing the drug intothe patient with possibly fatal results if thedrug were a toxic substance like insulin oran anticancer drug.

Looking at this problem from an engi-neering standpoint, we and others thoughtthat it would be desirable to have surfaceerosion instead of bulk erosion. We wantedto structure the polymer so that it wouldbecome increasingly thinner, avoiding acatastrophic breakup and drug dumping.Starting from scratch, we first thought

480 MRS BULLETIN • VOLUME 31 • JUNE 2006

Biomaterials for Drug Delivery and Tissue Engineering

Figure 5. Controlled drug delivery from a microchip: (a) results from the controlled release ofa single compound and (b) results from the controlled release of multiple compounds.(Reprinted by permission from Reference 7.)

Figure 4. (a) Top and (b) bottom view of a drug delivery microchip. This chip is 15 mmwide. (Courtesy of Paul Horwitz, Atlantic Photo Service.)

�

garage door. Santini’s group has been ableto remotely trigger drugs in animals manytimes over a six-month period, very repro-ducibly. They hope to be able to apply thisconcept to humans within the next fewyears. The triggering device, which mightsomeday be embedded in something likea wristwatch or a Blackberry, could beprogrammed by the patient or the doctorand changed when necessary.

While that concept may be attainable inthe not-too-distant future, it also suggestsideas that are more remote. For example,along with microprocessors and powersources, biosensors could be put on thechips to detect signals in the body, such asglucose levels, and provide feedback totell the chip how much drug to deliver.

Medical Materials from Everyday Objects

These drug delivery systems are allbuilt around the idea of engineering mate-

rials to do things they have never done be-fore. That has been one of the primarypursuits of my research. Another facet ofour work has been investigating how ma-terials find their way into medicine, andcreating newer, possibly better, materials.Being a chemical engineer, I had oncethought that experienced chemical engi-neers or chemists were the driving forcefor bringing materials into medicine. Butthe closer I looked into that theory, the lessI found it to be true. Rather, medical doc-tors were the ones who identified prob-lems in their field and, urgently wantingto fix them, came up with materials solu-tions. They would search their surround-ings —for example, their homes or localstores —to find objects that closely resem-bled the organ or tissue they wanted to fix.They would then adapt that material foruse in the human body. While that prac-tice has admittedly resulted in some solu-tions, it has also created problems.

about what catalyst we wanted to use tocause the polymer to degrade. We deter-mined that an enzyme would not workoptimally because everyone has differentenzyme levels, which means that the proc-ess might not be reproduced from personto person. We realized then that excesswater is something everyone possesses,and decided to use water as our catalyst.Next we looked at the problem from achemistry viewpoint. Using water as a cat-alyst, how could we create the conditionsfor surface erosion? We would want tomake the monomers hydrophobic to keepthe water out, while guaranteeing that thepolymer would still dissolve fast enough.A solution would be to have the bondsconnecting those monomers be water-labile. A lot of chemical calculations later,we estimated that the anhydride bondwould be the right one.

Our next step was to choose the rightmonomer. Here, we took a biological ap-proach and sought advice from chemistsand toxicologists. Michael A. Marletta, anoutstanding chemist who now heads theChemistry Department at the Universityof California, Berkeley, but who was thena professor at MIT, helped us determinewhat monomers would be safe. We chosecarboxyphenoxy propane, which is acopolymer, and sebacic acid, and tried tomake polymers out of these substances(Structure 1).

The results were gratifying. Not onlydid we achieve the surface erosion wewere seeking, but we also acquired theability to control the degradation rate. Forexample, by using varying amounts of se-bacic acid—say, from 0%, to 15%, to 55%,to 79%—we attained degradation ratesthat ranged from two weeks to three orfour years. The result was that we couldsimply dial in the monomer ratio and tar-get the material for different degradationtime periods.

Polymer-Based LocalChemotherapy

Based on this success, we thought wemight be able to facilitate other medicalapplications. An interesting example of oursubsequent efforts was the brain cancer

work we did with neurosurgeon HenryBrem. Dr. Brem visited my laboratory in1985 when he was a young doctor juststarting his medical career at the JohnsHopkins University. At that time, he waslooking for a way to treat brain cancer inits worst form—glioblastoma multiforme.This disease is uniformly fatal; regardlessof the type of treatment patients receive,they normally die within a year.

In addition to that, the chemotherapydrugs normally used to treat this cancerare extremely toxic. One of these drugs is BCNU, or 1,3-bis(2-chlorethyl)-1-nitro-sourea. In brain chemotherapy, the drug isadministered to the patient intravenouslyand travels throughout the entire body,with devastating side effects to the liver,the kidney, and the spleen.

To avoid these side effects, Dr. Brem andI developed a concept that we call localchemotherapy. In this case, the idea isbuilt around allowing a neurosurgeon tooperate on the patient to remove as much ofthe tumor as possible, which is the stan-dard procedure; then, prior to closing thepatient, the surgeons lines the surgicalcavity with a polymer containing BCNU.Normally, BCNU has a lifetime of only12 minutes. Our goal was to extend thislifetime by placing the drug in a polymerthat would prevent it from being de-stroyed right away. What Dr. Brem andother neurosurgeons at Johns Hopkinswanted was a degradable polymer thatwould not accumulate in the brain, onethat had the surface erosion characteristicsthat would prevent a sudden drug release,and one that, based on their animal stud-ies, would last for a month. Because wecould target different drug release timeframes by changing the copolymer com-position, we were able to develop a drugdelivery system that would protect theBCNU from degradation and not accumu-late in the brain. The resulting systemallowed doctors to produce high concen-trations of the drug in the brain, wherethey wanted it to be, and low concentra-tions in the rest of the body, where itwould cause harm.

This polymer drug delivery system re-ceived a cool reception when we tried to

get funding to support its development.Like other professors who do research, weneed to raise money to support our proj-ects. In my case, I usually write grants tothe National Institutes of Health. Thesegrants are then reviewed by my colleagues,who are also professors.

When I wrote grant requests for this re-search, the response was quite negative.As an example, our first grant in 1981 wasreviewed by a lot of chemists, who assertedthat we would never be able to synthesizethe polymers. However, Howie Rosen, oneof my graduate students at the time, didmanage to synthesize the polymers for hisgraduate thesis. Rosen ultimately becamepresident of the ALZA Corp., now a sig-nificant part of Johnson & Johnson, andvice president of Gilead Sciences, also a very successful company. He was re-cently elected to the National Academy ofEngineering.

After Rosen had accomplished thepolymer synthesis, we sent the grant backfor another review, and received thisreply: The grant should still not be funded be-cause the polymers will react with whateverdrug you put in. Again, with the help ofseveral resourceful postdocs—KamLeong, now a professor at Duke University,and Robert J. Linhardt, who is Constella-tion Chair and Professor in the Chemistryand Chemical Biology Department atRensselaer Polytechnic Institute—wewere able to work out ways to prevent thisreaction.

Then we returned the grant for anotherreview, which came back with the com-ment that although the work was better,the polymers were fragile and would belikely to break because of their lowmolecular weight, which was 7000. Thistime, two other postdocs in our laboratoryskillfully moved the project forward. AviDomb, who is chair of medicinal chemistryat Hebrew University, found the right cata-lysts, reaction times, and temperature con-ditions to increase the molecular weight ofthe polymers to a quarter of a million, atwhich point they were definitely not fragileany more. Edith Mathiowitz, who is nowa professor of medical science and engi-neering in the Artificial Organs Laboratoryat Brown University, was also very helpful.

After the grant had gone back for an-other evaluation, the reviewers said thatdespite the merit of the research, it couldnot be funded because new polymers werenot safe to test on people. Another graduatestudent, Cato T. Laurencin, then was ableto show that the polymers are safe, sup-porting Marletta’s theories. Laurencin isnow University Professor and Chair ofOrthopedic Surgery at the University ofVirginia and last year was elected to the

Biomaterials for Drug Delivery and Tissue Engineering

MRS BULLETIN • VOLUME 31 • JUNE 2006 481

Structure 1.

Institute of Medicine of the NationalAcademy of Sciences.

These kinds of negative reviews contin-ued until 1996, when the FDA finally ap-proved this treatment—the first time inmore than 20 years that the agency hadapproved a new treatment for brain cancer,and their first time ever to approve theconcept of polymer-based local chemo-therapy. This could not have been achievedwithout the help of our excellent graduatestudents and postdocs, of whom I am ex-tremely proud.

I would like to describe in broad termsthe brain cancer surgery during which thisdrug delivery system is implanted. Duringthis operation, our BCNU-containingwafers, which are about the size of a dime,are inserted into a human brain (Figure 6).Seven or eight wafers are usually insertedbefore the surgeon closes the brain.

The final phase of clinical testing for thisprocedure, which for human patients iscalled the phase three trial, monitored thesurvival rates for patients treated with thedrug release system and for a control group,

which represented the best conventionaltreatments. At the end of a year, 63% in thetreated group had survived, while only19% in the control group had survived. Atthe end of two years, 31% in the treatedgroup had survived versus 6% in the con-trol group (Figure 7). The BCNU drug de-livery system was approved by the FDAand has been in use for some years nowfor recurrent glioblastoma. Three yearsago, it was also approved for primaryglioblastoma. Today, the system is one ofthe common ways of treating this disease.We are still working with the Johns Hop-kins group to find better drugs to use withthe system.

The principle of polymer-based localchemotherapy is continually expanding.Since 1996, when we received FDA ap-proval for the brain cancer drug deliverysystem, other groups have used this con-cept to develop similar systems. A drugrelease system for spinal cancer has beenapproved, and clinical trials are takingplace for other kinds of cancer.

Yet, the area where polymer-based localchemotherapy has had the greatest impactis not in cancer at all, but in interventionalcardiology. Another of my former gradu-ate students, Elazer R. Edelman, who isnow a professor in the Harvard–MIT Di-vision of Health Sciences and Technology,has done some outstanding work in thisfield. One of the standard treatments forheart disease is to insert a stent into theblood vessel. Stents are wire mesh tubesthat look something like Chinese fingerpuzzles. During insertion, the stents ofteninjure the blood vessel, causing the smoothmuscle cells to proliferate wildly. About50% of the time, within just six months afterstent placement, these muscle cells prolif-

erate enough to cause a problem calledrestenosis—a blockage of the blood vesselthat can lead to patient death. Even in thebest-case scenario, the surgeon has to reoperate.

To solve this problem, scientists at BostonScientific, Johnson & Johnson, Medtronic,and other places coated the stents withpolymers and used these in conjunctionwith anticancer drugs such as Taxol. Froma medical standpoint, the results have beenremarkable. The rate of restenosis hasfallen from 40% to about 3%. From a mar-ket standpoint, these coated stents have hadan equally enormous impact. Althoughthis device was only approved in 2003,sales in 2005 were already about $5 billion.Efforts to improve the stents are ongoing.

Tissue EngineeringI would now like to describe another

area, the delivery of mammalian cellsthrough materials, which our group hasbeen working on for some time. My moti-vation for working on this project goes backmore than 20 years to work we did withBoston surgeon Joseph P. “Jay” Vacanti,who is now chief of pediatric surgery atMassachusetts General Hospital and theJohn Homans Professor of Surgery at Har-vard Medical School. Among his patientswere small children dying of liver failurewhose lives depended on someone elsedying and donating a liver for transplant.Not enough transplants, however, wereavailable to treat these patients. The prob-lem, of course, was not limited to liver dis-ease, but extended to paralysis and toadvanced diseases of nearly any organ ortissue.

Twenty years ago, Vacanti and I cameup with an idea for tissue engineering.This concept involves taking isolated dis-sociated cells from patients themselves orfrom a close relative. These cells include,for example, bone cells, cartilage cells, livercells, intestinal cells, and urothelial cells.Today, they could also include stem cells.If these cells are injected at random, verylittle happens. It turns out that the cells are“smart,” however, and if you put thesecells close enough together, they can or-ganize themselves and create structures. Agroup at UC–Berkeley has shown that ifyou take mammary epithelial cells and putthem close enough together in vitro, theycan form acinae and make milk.

We wondered if we could make tissuesby taking cells and putting them on apolymer template, in three dimensions,with the right media. We would want togrow them outside the body, on the rightmaterials, and then ultimately returnthem to the body to make whatever tissuewe wanted (Figure 8). We would either

482 MRS BULLETIN • VOLUME 31 • JUNE 2006

Biomaterials for Drug Delivery and Tissue Engineering

Figure 6. Polymer-based localchemotherapy. (a) Photograph of awafer containing chemotherapeutic drugBCNU being surgically inserted into ahuman brain. (b) Seven or eight wafersare inserted in the surgical site beforethe surgeon closes the brain.

Figure 7. Results of phase three clinicaltrials of the local chemotherapy wafersin Figure 6, showing the survival ratesfor patients treated with the drugrelease system and a control group thatrepresented the best conventionaltreatments. At the end of a year, 63% inthe treated group had survived, whileonly 19% in the control group hadsurvived. At the end of two years, 31%in the treated group had survived,versus 6% in the control group.

utilize existing FDA-approved polymerslike lactic glycolic acid copolymer, or syn-thesize brand new ones. For example,Denise Barrera, who was in our laboratorythen and who now works for 3M, devel-oped polymers that could attach aminoacid sequences that would be specific forcertain cell types. We would then convertthem into fibers that can be used as sub-strates for cells to create new blood vesselsor other tissues.

Let me now speculate about where thisfield might be in, say, 30 years. A good ex-ample would be plastic surgery. Imagine apatient visiting a plastic surgeon and ask-ing for a new nose. I believe that in 30 or40 years, patients might be able to go to acomputer screen to select whatever nosethey like, or create their own “designer”noses.

V. Prasad Shastri, one of my formerpostdocs and now an assistant professorof biomedical engineering at VanderbiltUniversity, worked out a way to do thatusing a novel fabrication process that couldsomeday be adapted to computer-aideddesign techniques. Shastri was able to fab-ricate a “regular” nose by using a polymer,lactic glycolic acid, and forming it in theshape of a nose that was 97% porous (Fig-ure 9). Pieces of the polymer could then beadded or reshaped to create other kinds (orshapes) of noses that patients might de-sire. The donor cells for the nose could, forexample, be taken from the patient’s earthrough a minimally invasive surgeryprocedure called arthroscopy. The cellswould be placed on a scaffold and grownin a culture until the nose was fully formed.We are now doing this procedure, but not

on noses. Later, I will describe several dif-ferent tissues that our group and othersare growing in this manner.

Shape-Changing Devices forMinimally Invasive Surgery

I want to shift gears just a little and dis-cuss the concept of minimally invasive sur-gery. Surgery often involves the insertionof bulky medical devices into the humanbody, an unpleasant experience. In the noseexample I just mentioned, a patient might

wonder if that would involve an operationto insert the new nose. An advanced areaof medicine, called minimally invasivesurgery, has eliminated some of the medi-cal and emotional trauma associated withcertain operations. Twenty years ago, priorto the development of this type of surgery,having a gall bladder removed was a majoroperation that involved a large incision, aweek’s recovery in the hospital, and twoor three months of recuperation at home.Today, a minimally invasive gall bladderoperation consists of making a tiny incision,inserting medical instruments, pulling thegall bladder out through that tiny hole, anddoing the entire surgery via a TV screen.The patient is out of the hospital in less thana day and back to work in less than a week.

The success of this type of proceduregave me an idea: what if bulky medicaldevices could be inserted through thesetiny holes? That might sound like sciencefiction, but with the help of materials, Ihope it will become a reality. Since the de-vices that might be inserted into the bodyare usually polymer-based, we thoughtabout making devices from polymers thatwould originally be very thin, like a string,outside the body, but convert to differentbulkier shapes once inside the body. Work-ing with Andreas Lendlein, one of mypostdocs at the time and who has subse-quently headed large groups in Germany—first at Aachen and now at PotsdamUniversity—we came up with several waysto do this. We published one method inScience,4 the other in Nature.5

Biomaterials for Drug Delivery and Tissue Engineering

MRS BULLETIN • VOLUME 31 • JUNE 2006 483

Figure 8. Schematic diagram of tissue engineering. Osteoblasts (bone cells), chondrocytes(cartilage cells), hepatocytes (liver cells), enterocytes (intestinal cells), and urothelial cells aredepicted.

Figure 9. Cartilage tissue engineering: polymer scaffold for a human nose (courtesy ofV.P. Shastri).

Our project was based on using differentstimuli to trigger the shape change. First,we used temperature as a trigger, devel-oping a polymer that at room temperaturewould have one shape, like a string, thenat body temperature would convert to adifferent shape—whatever-shaped medi-cal device was desired. We then experi-mented with light as a trigger, changingthe material from one shape to another atdifferent wavelengths.

To provide a specific illustration of howthis works, I would like to describe somework done by Lendlein and myself. Infact, there is now an innovative companyin Germany called Mnemoscience, whichis developing these applications. First,Lendlein created a device programmed insuch a way that at room temperature it hasthe shape of a string, but when droppedinto body-temperature water, it changesinto a coil like a stent.

Another example, one that has interest-ing applications for surgery, is a self-tyingsuture. An ordinary suture works well inthe case of an arm wound, for instance,where the doctor can easily tie a surgicalknot. If that wound is in the lung or stom-ach, however, and minimally invasivesurgery is needed, tying the knot becomesmuch more complicated. We believed thata solution would be to develop a suturethat would tie itself into a knot with achange in temperature. At room tempera-ture, the suture could be introduced in aloose loop like a lasso, but when it reachedbody temperature, the suture wouldtighten. This concept was demonstratedby dropping the loosely tied suture intobody-temperature water, which causes itto tighten. Because many different kindsof shape changes can be created along theselines, this concept could have a profoundeffect on minimally invasive surgery andtissue engineering, among other areas.

Replacement CartilageTo conclude this talk, I would like to

present three examples of tissue engineer-ing using the principles I have discussed.The first one is a form of replacement car-tilage that we developed in cooperationwith Charles A. Vacanti, who now headsthe Department of Anesthesiology, Peri-operative, and Pain Medicine at Brighamand Women’s Hospital. Every year, aboutone million people need to replace carti-lage in their bodies. Vacanti did some ex-periments with nude (hairless) mice, usingtheir own cells on a scaffold to developnew cartilage that would resist rejection.For example, he reconstructed the skull ofone mouse and the cheek of another. Ifyou opened up the animals to see the re-sults, you would see pure-white glisten-

ing cartilage, with histologically the samecharacteristics as the original cartilage.Since this replacement cartilage is not asmechanically strong as the original, how-ever, we still have work to do before wecould provide, for example, new cartilagefor a human arthritic knee.

However, the development of artificialcartilage for such areas as cosmetic defectsmay be closer to becoming a reality. Vacantisees patients who, for example, do not haveears. Linda Griffith, one of my formerpostdocs and now a professor of mechan-ical and biological engineering at MIT, hascreated an artificial human ear by makinga scaffold in the form of an ear with poly-mer fibers, cartilage cells, and a matrix.Vacanti has not yet put this ear on humanpatients, but he has tried it on rabbits totest its safety.

Replacement cartilage has been tried onhumans, however, in other forms. JayVacanti first used it on a 12-year-old boywho had no ribs covering his heart. Thisboy, like many his age, liked to play base-ball, but he would have been at grave riskfrom something as ordinary as being hit inthe chest by a stray ball. Vacanti operatedon him and gave him a new chest createdon a polymer scaffold using the boy’s owncells.

Replacement SkinA second example is the development

of new skin for burn victims. This researchwas done by Smith & Nephew, which li-censed our MIT patents. Let us consideran actual patient, a two-year-old boy whowas badly burned. The clinicians took apolymer scaffold with neonatal skin fibro-blasts and inserted it into the child at thetime of injury. Three weeks later, new skinhad started to form. Six months later, theburned areas were nearly invisible. This

treatment was approved by the FDA forpatients with skin ulcers and burns.

Spinal Cord RepairA final example is spinal cord repair. We

are still at an early stage in developing asolution for this major problem. ErinLavik, who is now an assistant professorat Yale University, led this work in ourlab. Lavik made a polymer scaffold thatmimicked the human spinal cord. Theouter portion was oriented to provideaxonal guidance, while the inner portionhad large pores seeded with neural stemcells, which are similar to brain stem cells(Figure 10).

We then worked with neurosurgeonTed Tang and stem cell specialist EvanSnyder to test this device in animals. Para-plegic mice that had difficulty moving theirback legs were tested. We divided the miceinto four groups: a treatment group thathad the polymer stem cell scaffold inserted,and three control groups: ones that weregiven a sham operation, ones implantedwith stem cells only, and ones with thepolymer only. We then followed the micefor a year, doing numerous studies.

At 100 days following the operation, we found that in the control groups, themean—about 40 animals—dragged theirlegs, and their paws were splayed in anawkward fashion. This is typical. In con-trast, in the mice that were implanted witha polymer scaffold containing stem cells,we see results that are dramatically differ-ent from those of the control group. Thesemice can now bear their own weight andwalk. Their paws are splayed in a muchmore normal fashion. This treatment as itnow stands, however, is by no means a curefor humans, and a lot of work remains tobe done. Nonetheless, it shows that prom-ising results can be obtained. Hopefully,

484 MRS BULLETIN • VOLUME 31 • JUNE 2006

Biomaterials for Drug Delivery and Tissue Engineering

Figure 10. Schematic illustration of a polymer scaffold for spinal cord repair.

with further research, we and others willbe able to develop improved solutions.

ConclusionThroughout all of this research—with

its challenges, setbacks, and successes—Ihave remained incredibly excited about themedical potential of materials. I believe thatwe are only at the tip of the iceberg in ex-ploring and developing materials for alltypes of medical applications. It’s my hopethat scientists throughout the world willmore fully embrace materials science as aresource for creating technologies that canprofoundly relieve suffering and prolonglife.

AcknowledgmentsI want to thank the National Institutes

of Health and the National Science Foun-dation in particular for supporting this re-search over many years. I also want tothank the many students, postdoctoral fel-lows, and collaborators who are respon-sible for this research.

References1. R. Langer, Nature 392 (Supp.) (1998) p. 5. 2. J. Lazarou, B.H. Pomeranz, and P.N. Corey,JAMA 279 (1998) p. 1200.3. R. Langer and J. Folkman, Nature 263 (1976)p. 797.4. A. Lendlein and R. Langer, Science 296 (2002)p. 1673.5. A. Lendlein, H. Jiang, O. Junger, and R.Langer, Nature 434 (2005) p. 879.6. M.V. Sefton, L.R. Brown, and R.S. Langer,J. Pharm. Sci 73 (1984) p. 1859.7. J.T. Santini Jr., M.J. Cima, and R. Langer, Na-ture 397 (1990) p. 335.

Robert Langer is one of14 Institute Professors(the highest honorawarded to a facultymember) at the Massa-chusetts Institute ofTechnology (MIT).Langer has written morethan 870 articles andhas more than 500 is-

sued or pending patents worldwide that havebeen licensed or sublicensed to over 150 phar-maceutical, chemical, biotechnology, andmedical device companies. Langer’s contribu-tions, discoveries, and inventions have beenrecognized with numerous honors andawards, including the Charles Stark DraperPrize (2002), considered the equivalent of theNobel Prize for engineers and the world’smost prestigious engineering prize, from theNational Academy of Engineering; theGairdner Foundation International Award(1996), which Langer is the only engineer tohave received; the Dickson Prize for Science(2002); the Heinz Award for Technology,Economy, and Employment (2003); theHarvey Prize (2003); the John Fritz Award(2003), given previously to inventors such asThomas Edison and Orville Wright; and theGeneral Motors Kettering Prize for CancerResearch (2004). Langer also shared the 2005Dan David Prize of $1 million, from which hedonated several scholarships of $15,000 eachto outstanding doctoral students around theworld. He also received the Albany MedicalCenter Prize in Medicine and Biomedical Re-search (2005), the largest prize in the UnitedStates for medical research, and was elected tothe National Inventors Hall of Fame in 2006.In 1998, Langer received the Lemelson–MITPrize, the world’s largest prize for invention,for being “one of history’s most prolific inven-tors in medicine.” In 1989, Langer waselected to the Institute of Medicine of the Na-tional Academy of Sciences, and in 1992 hewas elected to both the National Academy ofEngineering and the National Academy ofSciences. He is one of very few people everelected to all three U.S. National Academies,and the youngest in history (at age 43) to re-ceive this distinction.

Langer served as a member of the U.S.Food and Drug Administration’s ScienceBoard, the FDA’s highest advisory board,from 1995 to 2002, and as its chair from 1999to 2002. He has also served, at various times,on more than 15 boards of directors and 30scientific advisory boards of such companiesas Wyeth, Alkermes, Mitsubishi Pharmaceu-ticals, Warner-Lambert, and Momenta Phar-

maceuticals. Langer has received honorarydoctorates from the ETH (Switzerland), theTechnion (Israel), the Hebrew University ofJerusalem (Israel), the Université Catholiquede Louvain (Belgium), the University ofLiverpool (England), the University ofNottingham (England), Albany MedicalCollege, the Pennsylvania State University,Northwestern University, Yale University,and Uppsala University (Sweden).

Upon receiving his ScD degree in chemicalengineering from the Massachusetts Instituteof Technology in 1974, Langer became aresearch associate at Children’s HospitalMedical Center of the Harvard MedicalSchool in Boston—a position he still holds.From 1988 to the present, his professionalappointments have included the WhitakerCollege of Health Sciences, Technology, andManagement; the Harvard–MIT Division ofHealth Sciences and Technology; and his cur-rent position at MIT.

Langer can be reached at the MassachusettsInstitute of Technology, Department of Chemical Engineering E25-342, 77 Massa-chusetts Ave., Cambridge, MA 02139, USA; tel. 617-253-3123, fax 617-258-8827, e-mailrlanger@mit.edu, and Web site http://web.mit.edu/langerlab/index.html. �

Biomaterials for Drug Delivery and Tissue Engineering

MRS BULLETIN • VOLUME 31 • JUNE 2006 485

Advertisers in This Issue

Page No.

Chemat Technology, Inc. 437

Global School for Advanced Studies 441

High Voltage Engineering Inside front cover

Huntington MechanicalLabs, Inc. Outside back cover

Janis Research Company, Inc. 465

Kurt J. Lesker Company Inside back cover

MMR Technologies, Inc. 465

NanoInk, Inc. 434

Shiva Technologies, Inc. 438

Sigma-Aldrich 433

ULVAC Technologies, Inc. 440

For free information about the productsand services offered in this issue,

check http://www.mrs.org/bulletin_ads

The Materials Gateway—

www.mrs.org The Materials Gateway—

www.mrs.org

www.mrs.org/bulletin