(Illustration by AAD Goudappel)
I have run large organizations but spent little time in medical labs. So when I write about translational medical research, I think about the famous announcement in the London Underground: “Mind the gap.” Translational research is all about filling two organizational gaps. The first gap is between the scientists making discoveries in the lab and the clinicians seeking patient therapies. The second gap comes later—between finding attractive targets for drugs and diagnostics, for example, and getting the financing for clinical trials.
Researchers have estimated that only 14 percent of all scientific discoveries are ever translated into patient therapies, and even successful translations often take years to happen. This gap is a result of many factors. In some cases, there aren’t enough resources to convert promising discoveries into therapies. In other cases, there is insufficient cooperation between scientists doing basic research and physicians working in clinics.
At the same time, the gap is widening between identifying a drug target and securing the funds for the extensive clinical trials needed to obtain approval from the US Food and Drug Administration. These trials have become increasingly expensive—up to $600 million for one drug approval. Because of this large expense and the high risk of failure, drug companies and venture capital firms have become reluctant to finance early-stage drug development.
Because of the urgency of delivering new drugs to combat debilitating diseases, institutions that engage in translational medical research must attempt to close both these gaps—by promoting collaboration between bench scientists and clinical doctors, and by developing drug targets to the point where they will be taken up for clinical trials by for-profit groups. Filling these two gaps is a core mission of the Harvard NeuroDiscovery Center, where I have served for more than five years on the advisory council.
Harvard Medical School established the center in 2001 to focus on neurodegenerative diseases, including Alzheimer’s, ALS (also known as Lou Gehrig’s disease), multiple sclerosis, and Parkinson’s. In order to find therapies for these diseases, the center built a drug discovery lab to test relevant scientific discoveries to determine whether they could be translated into drug targets. Then, in theory, the most promising targets would be licensed to drug companies or venture capitalists to undertake further development and the extensive clinical trials required for drug approval.
The center’s approach to closing this gap is clear and logical, but achieving success has not been easy. The center’s experience illustrates important lessons about what it takes in practice to become successful at translational medical research.
Closing the Gap
Let’s begin by considering the challenge of closing the gap between scientists and clinicians. At the start, the center informally solicited projects for the lab mainly from scientists at Harvard University-related facilities. When the center received a multiyear grant from the US National Institutes of Health, the center was required to conduct a nationwide solicitation for projects, with only one winner per year allowed to come from a Harvard-affiliated institution. More recently, the selection process has been further constrained because some private funders have insisted that the center design projects to address specific diseases.
The lab began with the policy that every new project would require a post-doctoral fellow from the basic research side to work closely with the lab’s staff, who were generally drawn from commercial industry. This seemed to be a sensible policy, because fellows were more readily available and less expensive than medical school professors. Yet the policy proved problematic because the fellows often left after one or two years, making it difficult to carry out projects that usually took several years to complete. In response, the lab pushed hard for more involvement by professors, especially principal investigators with a long-term commitment to translational projects.
More broadly, Harvard Medical School struggled with an organizational question: how to promote more interaction between bench scientists and clinicians. To increase such interaction, the center’s management encouraged more basic researchers to become members of the center. To provide more collaboration between the lab and the clinic, Harvard Medical School recently decided to have the center report jointly to a research scientist heading the school’s neurobiology department and an applied disease expert based in a Harvard hospital.
Getting Discoveries Out the Door
Closing the gap between finding drug targets and financing clinical trials proved even more challenging than getting scientists and clinicians to work together. Although the center tried to generate projects attractive to large pharmaceutical companies, they seemed “to keep moving back the goal posts,” says Adrian Ivinson, the center’s director. The drug companies initially expressed interest in new and promising drug targets. Then they switched to demanding compounds identified by high throughput screening of a large compound library. Next they wanted “lead” compounds that behaved more like the drug targets, requiring a substantial staff of medicinal chemists who could deliver much-improved drug-like compounds. Finally, the companies imposed an additional condition: They needed the center to examine the toxicity of the potential drug and to demonstrate that it was active in animal models of disease.
The escalating demands of Big Pharma held back the commercial take-up of drug targets, but the center was very successful in delivering a set of what I call “tools”—such as a tissue bank, optical microscopy, mouse models, and advanced MRI. These tools could be used by a wide range of scientists and clinicians working in other research organizations. The center not only leased out these tools at relatively low prices but also offered training courses on how to use them.
As the center continued to raise money and expand its activities, I and other advisory council members raised another question: How should we evaluate the success of the center? One council member suggested that the center should measure its success by asking simply whether it had found the cure to a major disease. But that target seemed to be unrealistic and, to a significant extent, subject to conditions outside the center’s control. Even so, it seemed insufficient to say, “We have a smart staff and they work very hard.” What we needed was a framework for evaluating intermediate goals.
Creating a Framework for Evaluation
I assumed that other research institutes and foundations had already developed such frameworks, so I asked a Harvard graduate student to do a survey. What a surprise! Most of the institutions gave rudimentary answers to the survey and wanted to know if we could provide them with a detailed framework. So I decided to take on the challenge. I designed a flexible framework that would ask each institution to define its intermediary objectives and agree on metrics across a shared set of seven dimensions.
The first and second dimensions of the framework were two critical inputs: recruiting talent and obtaining funding. The third and fourth dimensions were the results of the experiments: creating drug targets and validating discoveries. The fifth and sixth dimensions were about spreading the results: disseminating the discoveries to the world and encouraging uptake by third parties. The seventh and last dimension was collaborating, an essential component of the other six dimensions.
I was worried that this framework would seem too systematic for the discovery process—that scientists would maintain that breakthroughs happen more by chance than by a disciplined process. But I was encouraged by the response of the senior scientists at the Harvard Stem Cell Institute who listened to my presentation on the framework. They felt that the institute could benefit by defining more clearly its objectives on each dimension and its metrics for success. In fact, they insisted that the research scientists as well as the board members be involved in the process of setting objectives and metrics.
After four or five years, the board members and the scientists should use the framework to assess how well the institution has met its objectives. At that time, it is important to have the assistance of an independent review committee composed of unaffiliated experts from academia and industry. At our center, independent reviewers were very helpful. They focused the drug discovery lab on a smaller number of projects in areas of its comparative advantage, and they suggested increases in staffing levels in medicinal chemistry and pharmacology.
Since my framework was published, I have been contacted by quite a few institutions that want to adopt a similar approach. Within a few years, we hope we will know a lot more about what works best in organizing translational medical research.
I am concerned, however, by one recent development. The National Institutes of Health has recently established its own National Center for Advancing Translational Sciences (NCATS) with a significant amount of money and staff. NCATS has the potential to drain federal funding for translational research from universities across the country, concentrating the bulk of the scientific work in a large national laboratory. NCATS will have its own full-time staff that will automatically handle most large projects, which may reduce the role of peer review in allocating funding.
I believe that the centralization of translational research into a government entity would reduce the number of scientific breakthroughs, although NCATS may realize some cost savings from economies of scale. What we need in translational medicine is multiple institutions pursuing patient therapies, each in its own way, with defined objectives and specific metrics.
Read more stories by Robert Pozen.