Our nation has invested heavily in basic science research for more than six decades. And not without payback. In medicine, basic science research has led to a staggering array of discoveries about how genes work, how proteins interact, how cells grow and die and how diseases begin.
But many who work in biomedicine believe it is time to expand our investments to support translational medicine.
The term translation is often bandied about. Here, it refers to the taking of basic science discoveries, judging their relevance to human health, and investigating whether they can be used in humans for disease treatment, diagnosis, management or prevention. The final handoff of translational research is—almost always in this country—to the companies that commercialize new therapies.
This kind of disease-directed investigation is gaining traction at academic institutions, including HMS. Translational research has been a central feature of the HMS Strategic Planning Initiative. Prior to this process, the School launched programs such as the NeuroDiscovery Center, which takes a translational approach to studying neurodegenerative disease, and, more recently, Harvard Catalyst, the Harvard Clinical and Translational Science Center (see “Harvard Catalyst: Ways and Means...”).
The motivation for this movement arises from concerns that the existing bio-medical research model is not fit to seize today’s great opportunities. New genomic, imaging and informatics technologies are making it possible to study the relevance of discoveries to human disease as never before. Yet at the same time, funding sources, training programs and institutional infrastructures are not ideally organized to support the collaborative, large-scale, patient-focused investigations these tools enable.
William Crowley, HMS professor of medicine and director of clinical research at Massachusetts General Hospital, has been at the forefront of talking about new models for biomedical research for 15 years. He and James Gusella, the Bullard professor of neurogenetics at MGH, coauthored a commentary in the October 2009 flagship issue of Science Translational Medicine. The article identifies the barriers that trip up translation in the old model and describes a vision for a new way forward.
Lost in Translation
It was 1945 when Vannevar Bush defined the current model for biomedical research. His vision led to the creation of the National Institutes of Health and its tradition of funding individual investigators.
Over time, however, this model has also led to an ever-widening gap that Crowley and others call the “Valley of Death.” Depending on the point of view, the valley divides academy and industry, basic and applied science, bench and bedside. “There are lots of little deserts to die in, unfortunately,” said Adrian Ivinson, director of the Harvard NeuroDiscovery Center.
One of the first places discoveries fall through the rickety footbridge across this chasm is in earlystage translational research that aims to connect a scientific advancement to human health.
Take Gusella and colleagues’ discovery of the Huntington’s disease gene as an example. The first animal models of Huntington’s made after the discovery used a portion of the huntingtin protein to mimic end-stage brain degeneration seen in humans. Even though these models did not initiate the disease the same way or show the same patterns of nerve cell death seen in humans, they remained popular for years. “Scientists were trained to look at disease as pathology as opposed to disease as a process that leads to pathology,” said Gusella. One way to shift thinking toward the latter, he said, is to take information that comes from human patients and use it to help design experiments.
“The bench-to-bedside interface is no longer unidirectional but bidirectional,” said Crowley. So much so that “basic inquiries can begin with patients with disease states.” Patient data, which allowed for the discovery of disease genes for Huntington’s, Lou Gehrig’s, Alzheimer’s and other diseases, has pointed scientists to new pathways to study. Prior to these discoveries, said Crowley, “no one was studying these relevant pathways.”
This new approach, in which the patient becomes partner rather than subject, represents one defining change in the new model Crowley and Gusella present. It is an important shift because, in the new model, the patient becomes a key driver, a valuable guide and a primary beneficiary. “The speed, efficiency, efficacy and overall productivity of drug discovery has been severely limited by not having coherent access to patient information that can be used to guide development of treatments to disease,” said Gusella.
Outgrowing the Cottage Industry
Another reason translational science stumbles is cultural. “At some point, we have to shift from a cottage industry”—personified by individual investigators making independent discoveries— “to an enterprise,” said Crowley.
This “cottage” model still makes sense for basic science (see “Novel Blood Thinners...”). “It’s like a potting bed with a hundred seeds,” said Ivinson. “You’ve no idea where the next shoot is going to come up,” he said. “If you have $100 million to spend, it’s probably a good idea to spend $1 million on each person because you don’t know which study is going to work out.”
But when it comes to translational research, this approach has weaknesses that Ivinson’s NeuroDiscovery Center is working to remedy.
Ivinson’s team, for example, is working to identify biomarkers of Alzheimer’s disease. The only way to definitively diagnose Alzheimer’s today is to inspect the brain after death. A biomarker would not only allow for earlier diagnosis, it would allow for more effective and informative clinical trials of Alzheimer’s drugs. Identifying a biomarker, however, requires the inspection of a large number of clinical samples, something no individual researcher has access to.
“We thought we could just go to everyone involved in neurology research and pool all the patients they’ve got into one database,” said Ivinson. “But each collection is different enough that we’re comparing apples and oranges.” Even the plastic collection tubes a lab chooses can dramatically alter results of subsequent protein screens.
Rather than a cottage approach, in which each lab uses its own proprietary protocols, Ivinson is working on taking a more enterprise approach that involves cooperation and control every step of the way. He also outsources repetitive “factory” tasks, such as genetic analysis, to the bidder with the highest quality and lowest price.
A consortium of investigators, led by Ivinson, Clemens Scherzer, HMS assistant professor of neurology at Brigham and Women’s Hospital, and Brad Hyman, the John B. Penney Jr. professor of neurology at MGH, are working on building this data and sample repository for biomarker research. They are well on the way to having a collection of 2,000 consistent patient samples of DNA, RNA, plasma, serum, cell lines and now cerebral spinal fluid that will be available to “anyone with a decent idea,” he said. “Biomedical research shouldn’t be about an individual lab or a department or an institution. It should be about the goal.”
Fumbled Handoffs
On the far side of the Valley of Death sit the patients, clinicians and pharmaceutical companies waiting for new ideas to reach them. Few make it across. Even worse, the far-side ledge seems to be moving farther away as pharmaceutical companies cut their scientific staffs and academic centers shy away from conflicts of interest.
Academic medical centers such as the Harvard-affiliated hospitals could be the perfect institutions to fill this widening gap, said Crowley. “Or not.” Academic medical centers have a confluence of basic scientists, clinicians, clinical investigators, patients and longitudinal records. “But on the flip side,” he said, “they are also slow, expensive and are filled with regulations and internal conflicts that make them very difficult for industry to deal with.”
At the NeuroDiscovery Center, Ivinson is thinking about how to narrow this widening gap. “We need to connect our pipeline with theirs. Ours is a wide funnel. We put a lot of ideas in and a few drip out.” By engaging in a constant dialogue with industry, Ivinson hopes to refine that funnel in a way that will help ensure that someone will be waiting on the other end to catch those precious droplets and, if all goes well, improve patients’ lives.
Crowley, who lectures on this subject to young investigators frequently, says “interregnums” such as this one are especially tough on young investigators. It is harder for them to find mentors. They must learn and invent new tools and infrastructures. And funding models for new approaches to research lag
At the same time, he said, “I wish I were younger. What these guys are going to see is a spectacular opportunity for translational investigators.”
For more information, students may contact William Crowley at wcrowley@partners.org, James Gusella at gusella@helix.mgh.harvard.edu or Adrian Ivinson at adrian_ivinson@hms.harvard.edu.
Conflict Disclosure: The authors declare no conflicts of interest.
Funding Sources: N/A
