Freeman, Mason Wright MD; Dervan, Andrew P. MD
Our understanding of human biology has increased tremendously for the last several decades,1-5 yet the pace at which these discoveries have translated into new therapies for patients has been frustratingly stagnant.6 Universities and academic health centers (AHCs), as the major recipients of public investment in biomedical science, have an obligation to translate new knowledge into applications that confer human benefit. However, translating fundamental discoveries into practical applications takes many expensive, often highly regulated steps, and these activities have generally not been a major focus of academic institutions. Challenges in engaging universities and AHCs in translational research include building the appropriate infrastructures for human investigation, training and stabilizing the careers of young scientists and physicians interested in the requisite work, educating academic investigators about the regulatory requirements inherent in successful therapeutic discovery and development, and finding more efficient ways to match good ideas with adequate funding resources.
The purpose of this article is to examine the early-stage drug development process and evaluate the role that academia could play in it. Because interest in early-stage drug development grows among academic investigators, the need for more integrated partnerships among the academia, government, and industry has become increasingly apparent.
GROWING INTEREST IN TRANSLATING DISCOVERIES INTO NEW THERAPEUTICS
In response to the observation that breakthroughs in basic science were not resulting in new therapies for patients, the National Institutes of Health (NIH) catalyzed an effort to transform the way biomedical research is conducted and launched the NIH Roadmap for Medical Research in 2004.7 Part of this effort was to create the Clinical and Translational Science Awards (CTSA) program. With nearly 50 of these awards distributed to date and a goal of 60 centers to be funded by the year 2012, academic research centers in the United States are increasingly being viewed as essential players, and potential partners, in pharmaceutical innovation. Although the CTSA centers could potentially form a very powerful network of clinical investigators, only a few of the centers are truly focused on creating an environment in which early therapeutics development could take place.8 At a time when the pharmaceutical and biotechnological communities are looking for help and are prepared to outsource work they formerly performed in-house, the mismatch between industry needs and academic resources to meet those needs is glaring. The desirability of the academy participating in this work, from a scientific and medical perspective, seems obvious, but there are cultural and political barriers that must be addressed if meaningful, ethical, and productive partnerships are to be formed.
ROOTS OF TODAY'S TECHNOLOGY TRANSFER AT UNIVERSITIES
To understand the needs and limitations of early-stage drug development in academic settings, one must understand the history of such drug development efforts. In the 1970s, although scientific discovery based on federally funded research blossomed, there was a feeling that taxpayer funded research was not being leveraged efficiently to create new medicines/technologies. As a result, Senators Birch Bayh of Indiana and Bob Dole of Kansas in 1980 sponsored the University and Small Business Patent Procedures Act, commonly known as the Bayh-Dole Act.9 It allowed academic entities and nonprofit/small businesses to retain their rights to inventions made during the course of research supported by federal research funds at these institutions. Previous to this legislation, those discoveries were considered in the public domain, free for anyone to use. Without ownership rights, biomedical discoveries that had commercial potential were frequently not pursued either by the academy or industry because the investments needed to bring those discoveries to the clinic were viewed as unrecoverable in the absence of intellectual property rights protection. In addition, universities and AHCs were unlikely to make such investments because they lacked capital to do so and had no way to manufacture, market, or distribute products arising from the inventions of their faculty. After passage of the Bayh-Dole Act, patents based on federally funded research were controlled by the originating academic institutions, and the discoveries covered by these patents could then be licensed to industrial partners possessing the capital and expertise to transform them into commercially valuable products.
The securing of patent rights for academic institutions came with several very important strings attached. If universities were to take advantage of this new privilege, they had to enable others to use the discovery in noncommercial ways that meant granting the federal government a nonexclusive license to the invention. The bill also codified the right of the inventor(s) to share in the financial benefits that accrue from the invention, although it was the university that would file and hold the patent. The specific benefit that faculty inventors would receive was not mandated in the bill, so different universities created different royalty sharing schemes. On average, universities have given 40% of the royalties to the inventor(s) themselves, another 16% to the inventor's laboratory, and 26% to the hospital or university. The remaining royalties went to the technology transfer offices and other administrative bodies in the university system.10 At our institution, the Massachusetts General Hospital (MGH) at Harvard Medical School, there is a 4-way split: a quarter of the royalty will go to the hospital, a quarter will go to the department in which the work was conducted, a quarter will go to the laboratory in which the work was conducted, and a quarter will go to the inventors (as listed on the patent), collectively.
A final, and most notable, codicil in the act is the provision that the academic institution will lose its rights to a patent if it does not "actively" attempt to commercialize it. Senators Bayh and Dole wanted to make sure that research discoveries were actually being turned into economically viable products that benefited the society. (To date, this right has not, to our knowledge, ever been acted upon by the government.) So in today's environment, where academia-industry relations are strained by high profile, if rare, incidents of misconduct,11 many believe the idea of the hospitals' or universities' devoting effort in commercializing research discoveries runs counter to the values of the academy. It is therefore important to remember that the federal government expects and demands that academia commercialize its scientific discoveries if the latter's intellectual property rights are to be retained. The issue is therefore not whether to commercialize but rather how to do it. The key in managing that process is to recognize the inherent conflicts between academic discovery and product development and ensure that those conflicts are dealt with in a manner as transparent as possible.11
It is worth noting that some of the most esteemed universities in the United States are the most prolific patent filers (Table 1).12 These universities and hospitals take seriously the business of patenting the inventions of their faculty in an effort to further their research missions and to provide the society the benefits of their work.
Although the chief benefit to an academic institution of the commercialization of its inventions is almost certainly the enhancement of the institution's reputation as a major contributor to the well-being of the society, some discoveries do generate very substantial revenues. Although rare, several institutions have earned half a billion dollars or more from 1 invention. For example, Northwestern University sold its royalty rights to the anticonvulsant/pain reliever Lyrica (pregabalin) for $700 million, and New York University's Remicade (infliximab) generated more than $650 million. Massachusetts General Hospital's fusion protein technology that led to the development of the antiinflammatory medication, Enbrel (entanercept), earned the hospital more than $630 million, and Emory University's Emtriva (emtricitabine), a reverse transcriptase inhibitor used against human immunodeficiency virus, returned $525 million.13 None of these universities or hospitals actually sold the products themselves, but rather their discoveries led pharmaceutical companies to develop commercial products whose sales generated the royalties mentioned. It is notable that these innovations span the gamut of medicinal therapeutics. The aforementioned list includes a small molecule, a recombinant DNA protein, and a monoclonal antibody-all of which differ in the processes needed to take them from laboratory discovery through to the clinic. In addition to therapeutic molecules, universities also generate medical devices or enabling technologies (e.g., Baylor developed an antimicrobial coating for medical devices that makes them less susceptible to infection). Two of the biggest patents ever rewarded to universities were actually for methods that enable commercially important processes. One such patent was for the method of calcium phosphate transfection of mammalian cells, developed at Columbia University, that allows scientists to engineer cell lines that can produce substantial quantities of a recombinant DNA molecule in a cell culture system. The recombinant DNA revolution, itself, was the result of groundbreaking work by Boyer and Cohen14 that was patented by the University of California, San Francisco and Stanford University.
Although these discoveries provide dramatic examples of the financial rewards of patenting faculty inventions, they are by no means typical. Only 0.6% of almost 21,000 active licenses in the year 2000 generated royalties in excess of $1 million.15 In a report quoted by Thursby and Thursby,16 the average income per license was $66,645, and 43% of licenses earned no royalties at all. However, it is the perception or, perhaps more accurately, the hope that any given patent could be worth millions that has frequently slowed collaborative efforts between universities and industrial partners, while each side tirelessly negotiates to ensure their organization is best served by the licensing agreement.
WHAT TYPE OF DRUG DEVELOPMENT HAPPENS IN UNIVERSITIES TODAY?
What often happens with innovation in research laboratories is that when an investigator makes a discovery, that discovery is deemed to have market value by the university, and a provisional patent is filled on the inventor's behalf before the discovery is made public (usually a 30- to 60-day delay for writing the patent). That patent is converted to a full patent within 12 months by submitting a formal application and, if granted by the US government, can then be licensed either exclusively to a partner to further develop the technology or licensed nonexclusively to many partners to all use. Sometimes, the inventor becomes involved as a key scientific member of the team that further develops the ideas in the patent, other times not. Importantly, investigators should be aware that the US patent system (in contrast to the rest of the world) uses a first-to-invent standard, rather than a first-to-file (the paperwork) standard. The moral is thus to keep good laboratory notebooks and records of your thinking and work, so that if someone else files a patent on the work you think is yours, you can show documentation of your priority.
For many investigators, the process often ends with the filing of the patent. Occasionally, however, the investigator will pursue the development of the invention in the small start-up company that has licensed his or her invention from the university. Alternatively, the investigator may simply continue to do research at his or her home institution that is related to the original discovery but not focused on its actual commercialization. The latter approach makes sense for most investigators because the process of drug development is long, expensive, and highly regulated and requires specialized expertise not typically found in academia. Too often, glib discussions about the ease with which academic institutions might become dominant contributors to the therapeutic pipeline do not take into account how expensive and time consuming this process really is.
SPECIFIC REQUIREMENTS IN EARLY-STAGE DRUG DEVELOPMENT: A LONG, COMPLICATED PATH
As shown in Table 2 taken from the Food and Drug Administration (FDA) Web site, the drug development process is complicated and time consuming. Also, if you are a young investigator with a new faculty position, spending the 11 to 16 years it takes to develop 1 medicine from bench to clinic may not be a wise career decision because you will probably be off the faculty for lack of research productivity before the new medicine reaches its first patient. Yet, though support for full-scale drug development is difficult to envision in academia, there are aspects of drug development that academia can do and do very well.8
DISCOVERY AND TARGET SELECTION
Drug discovery begins with identifying the target for the intended drug. A good target is one where the mechanism of action in human biology is well defined and for which one hypothesizes that engaging the target by activation or inhibition (e.g., through the addition of your drug) will have a desired effect on that system (and will not have undesirable effects on other systems in the body). Then, one sets up an assay (model system) in the laboratory to manipulate that target and to measure the biologic effect. Academic investigators do this regularly. However, making this testing of potential drug compounds more efficient and high throughput is the forte of industry. Industry also has large libraries of molecules that can be tested in these assays. With recent efforts of groups such as the NIH (with PubChem), academic researchers have gained more access to large libraries of potential compounds, although even these libraries are typically not as sophisticated or large as those available in industry.
In continuing development, what often is misunderstood by many academics is that, once a compound that acts on a given target is found, it is only the starting point. The initial screening process commonly identifies compounds whose affinities for the target are low (i.e., requiring micromolar concentration ranges for the desired effect). But for safety and specificity in humans, one seeks compounds with 1000 times that affinity (nanomolar concentrations for effect). Industry's approach in solving this problem is to use teams of 15 or 20 medicinal chemists who make hundreds or thousands of modifications to the original compound to find ones that have the desired profile with higher affinity. Academia has enormous chemistry expertise, but it is focused on cutting-edge concepts in chemical synthesis rather than on this type of applied high-throughput chemical modification and screening. The latter chemistry is often being outsourced by industry to contract chemistry laboratories in China or India because of the lower labor costs in those countries (which, however, are rising rapidly). In some cases, the candidate compound is not a small molecule but rather a protein. For these biologics, proteins can typically be made in a cell culture system. However, it can take $10 to $15 million to get a master cell line system to produce the grams or kilograms of protein needed for large-scale testing. All told, the medicinal chemistry and manufacturing challenges of drug discovery are formidable obstacles to drug development in academia. Most NIH grants are not budgeted to cover this type of work because they are intended instead for the biologic discovery work that has led to the target identification concept. So, partnering with industry, both for financing and manufacturing expertise, can be critical at this stage of drug development.
Another aspect often not fully appreciated in the academic community is that, to test a drug in humans, one needs to have substantial compound characterization and animal-based safety data. Before safety studies can be done in animals, it is critical to show that the compound (drug) being tested (1) can be made in a reliable and consistent way; (2) has few impurities, and that those impurities it does have are not biologically important; and (3) has a long shelf life. None of this work plays to the usual strengths of university laboratories. Thus, to efficiently accomplish these tasks, it makes sense to outsource this work to contract research firms who do this all the time. These processes cost $350,000 to $600,000, a large sum for any single investigator. To then also show that the compound does not cause chromosomal defects in cells, an additional study costing approximately $100,000 is needed. Other safety work that is generally required before going into a human trial involves in vivo toxicology studies, conducted in 2 different animal species for a relatively brief period (2-4 weeks). These studies give insights into the dose-toxicity relationships and point investigators toward the tissues most likely to bear a toxic insult in humans. In an effort to assure the FDA that the first patient who gets the drug will not die of a cardiac arrhythmia (due to the activation or inhibition of a critical cardiac ion channel), ion channel studies are also typically done, and these can cost up to $200,000. Depending on the target human population, additional safety studies may be required, such as respiratory or kidney safety studies. Finally, longer in vivo toxicology studies may be required because the animal exposure to the drug must last at least as long as the planned length of exposure that people will be subjected to in the first-in-human trial. If the only appropriate model for the toxicology work is a primate (e.g., if the biologic therapeutic only cross-reacts with primate proteins or the small-molecule drug only inhibits primate enzymes), the costs of toxicology studies can quickly escalate to more than $1 million for a 4-week study. All told, the safety studies for a new drug can cost $1 to $2 million, a sum out of reach for most NIH-sponsored investigators, and are all just a prelude to putting the drug into a human for the first time.
DOSING, PHARMACOKINETICS, AND PHARMACODYNAMICS
If the molecule is deemed safe in the studies described previously, one must now choose a dose. The dose chosen for the human study is capped at the highest (scaled to the body's surface area) safe dose tested in the long-term animal toxicology studies. Often that dose is then reduced by a factor of 10 to create a margin of safety. With healthy volunteers, you start with that dose, given once, in a controlled environment, and measure several physiological parameters, such as liver enzymes, white blood cell counts, vital signs, and cardiac rhythm (as global markers of how the body is reacting to the drug). If there are no serious adverse events, you can give a slightly higher dose to the next set of healthy volunteers, leading to giving multiple doses for several days. This early work of dose escalation is most often done in healthy subjects, but in some circumstances, the FDA will permit patients with the disorder of interest to be the first human recipients of the drug. During the initial study period, other parameters related to the drug are measured such as drug Absorption (how it gets into the body), Distribution (where it goes), Metabolism (how it gets modified and if those metabolites have biologic activity), and Excretion (how it gets cleared by the kidney or liver), in humans. This group of studies, lumped under the rubric of ADME, is critical to the future use and dosing of any experimental agent. Academic investigators rarely have experience in the ADME disciplines. So, although it's exciting to find a molecule or protein that targets a biologic system, the safety and toxicology studies are prerequisites for moving it into the clinic. The common translational medicine catchphrase, "from bench to bedside," glosses over the reality of the ramp that must be climbed to move something from that bench to anyone's bedside. Only with an extensive body of preclinical information in hand, can you then apply for an investigational new drug (IND) approval at the FDA and actually do the human studies that test whether a drug works on the patients for whom it is intended (i.e., a phase IIA proof of concept study to show that the drug leads to a desirable clinical outcome with an acceptable safety profile).
There are a few exceptions to this drug development process that are worth mentioning. A drug that has been approved for clinical trials for another indication can much more easily be repurposed for a new disorder because the qualifying background work has already all be done. In addition, the FDA does recognize that investigator-initiated INDs, targeted for high-risk or orphan populations, can be very valuable. In these circumstances, some abbreviation of the package needed for human studies can be permitted, but the overall path and requirements remain very similar.
Finally, we have been focusing on the academic research contributions in the earlier stages of drug discovery and development. Later-stage clinical trials often are carried out by clinical research organizations, but AHCs also contribute in these later stages of the drug development process. The AHCs have well-characterized patient populations, electronic health records with rich patients' histories, substantial funding from the CTSA to train translational researchers, and, most importantly, the systems biology expertise and the core medical/scientific platforms that enable physician-scientists to elucidate the molecular underpinnings of human biology. With an industry or contract partner to help with the medicinal chemistry, toxicology, regulatory compliance, and financing, academic centers could add greatly to the number of medicines that enters clinical practice every year. For the most part, these 2 types of organizations: academic and pharmaceutical/biotechnological, have extraordinarily complementarity, but forming effective partnerships to exploit these synergies has proven remarkably difficult because of differences in goals, cultures, incentives, timelines, and perceived (and real) conflicts of interest.
THE NEED FOR NOVEL ACADEMIA-INDUSTRY PARTNERSHIPS
As outlined in this article, taking a molecule through early safety and clinical testing is laborious and expensive, and most academic centers do not have the money (nor do its investigators have the patience or expertise) to do this. Many flavors of academia-industry relationships have been tried in the past, but most have been stymied by the differing goals, cultures, and timelines of the partners.17
On the academic side, licensing agreements can be complicated and very time consuming to negotiate because academic institutions are wary of not receiving adequate compensation for the value of their ideas. As a result, some partnerships are never consummated because of negotiation impasses, despite the rarity of the truly lucrative academic discovery. The academy is also engaged currently in a heated discussion about how much of its effort can be legitimately focused on developing commercial products. The right to publish its results, the openness with which academic research needs to be discussed, and the role that students should or should not be playing in applied research projects are all major sticking points in academia-industry partnership discussions. Many institutional leaders also fear that some faculty will transgress on their ethical obligations to their students and patients and put the academy in an embarrassing or legally compromised position. When these issues are added to the differing cultures of the academia and the industry, it is perhaps more surprising that any agreements are negotiated than that so few good ones are.
TRANSLATIONAL MEDICINE GROUP AT MGH AS 1 NEW MODEL FOR ACADEMIA-INDUSTRY PARTNERSHIP
Despite the barriers, there is a growing opportunity for academic centers to get involved in drug development in specific ways that are suited to an academic center's strengths. Janet Woodcock, the director for the Center for Drug Evaluation and Research at the FDA, recently spoke at an NIH forum on best practices for academia, government, and industry relationships, and urged academic institutions to remain committed to the development of enabling technologies such as animal models, new biomarkers, and new ways of assessing the activity of a molecule saying, "I think industry needs to provide academia with support to get that type of workdone."18
Groups such as our own Translational Medicine Group at MGH have begun to forge some new partnerships with industry that we think can serve as models for a path forward. Unlike some of the partnership models that have come before,17 we have hired a group of former industry-based scientists with knowledge and experience of the drug development process to manage the complicated drug discovery tasks outlined previously and oversee work within and outside the academic, scientific community. This managerial role helps alleviate many of the competing tensions between industry's and academia's goals and expectations. For example, we have created templates for negotiating many of the most challenging obstacles that have hindered prior industrial/academic collaborations, and we are hoping this will expedite future partnership discussions. Our academic project managers understand the time pressures that the industry drug developers labor under because they work under those same time constraints for both our internal and external projects. In a new initiative that will be launched in the coming months, we expect to work closely with a team of pharmaceutical scientists to select targets identified by our institutional scientists and then jointly create molecules that can interrogate those targets for potential therapeutic benefit. This relationship will entail scientists sharing information freely between the academy and pharmaceutical company, with prenegotiated intellectual property rights and royalties agreed from the outset to further facilitate the free flow of information. Molecules that emerge from this collaboration can be more easily channeled into the clinic in a timely manner, if appropriate, because the partnership outlines in advance the roles and responsibilities of the academic and industrial partner in advancing drugs into the clinic. At the same time that we are experimenting with new forms of partnership, we are also trying to facilitate the dialogue by becoming an institutional interface for biotechnological/pharmaceutical discussions about a multitude of projects. The goal is to try to ensure that the accumulated experience and wisdom gained from prior interactions are not lost in future partnership negotiations.
The MGH Translational Medicine Group has also created some financial incentives for industry to work with us, and industry, in turn, has accepted that some return from commercially successful endeavors should be rightly returned to the academy for its contributions. Although agreements of this kind are complicated to negotiate, it seems that industry has come to recognize that academia's freedom to publish is a principle that must be honored and academia has learned that an idea's practical value is frequently only realized when commercial resources and know-how are intelligently combined with a novel concept. In initiating collaborative relationships of this kind, we have worked closely within our academic community to build programs where patients can feel that their safety is protected as much as possible, while still enabling discovery of novel treatments for diseases we cannot adequately treat today. This entails careful selection of clinical investigators who are free of financial conflicts and the creation of oversight boards that can monitor pharmaceutical/academy relationships. In the coming years, we expect that core technologies (constantly and rapidly changing in the AHCs), such as whole-genome sequencing, proteomics, and metabolomics, will be integrated with new imaging modalities to enable better monitoring of disease prevalence and response during early-stage clinical trials.
In this process, we are also trying to change cultural attitudes, convincing industry that academia can and will make a meaningful commitment to efficient, timely execution of its contracted work and convincing our academic colleagues that working with industry to make new medicines is not, in fact, making a pact with the devil, but is an important and valuable way for us to help the patients who place their trust in our care. A recent survey suggests that the attitude of academic investigators may indeed be changing. Many investigators now perceive that working with companies is a source of both valuable ideas and collaborations, as well as an enabler of productivity; and 40% of researchers have initiated industrial relationships that have had important nonfinancial scientific contributions to their work.19 It is our hypothesis that by improving academic/industry partnerships on terms that preserve the integrity and independence of the academic investigative community, we will also improve the lives of those who have diseases we now cannot effectively treat.