During the course of the workshop, several topics consistently arose during the presentations and subsequent discussions. Many of these topics also emerged in the final session of the workshop, a panel session that featured discussants William Rutter, chief executive officer of Synergenics, and workshop speakers Geoffrey Duyk, Daniel Geschwind, and Todd Sherer. This final chapter of the workshop summary highlights the major themes and identifies specific strategies for improving the efficiency and effectiveness of the translation of genomic science (see Box 6-1).
Tension in connecting basic and translational sciences exists because of the difference between their inherent missions. For example, many neuroscientists are interested in learning about how the brain works to gain new knowledge, not necessarily because the information would help treat a disease, Geschwind noted. Understanding the basic science is critical, and that should not be overlooked in the discussion of translational research, he said. However, the current system is designed to incentivize novelty through the publication of papers in scientific journals, and that does not necessarily translate to social benefit, such as improved treatments for patients. The replication of results has value, and, in fact, the scientific method relies upon it, but “the incentives just aren’t aligned” to do this sort of work, said Geschwind.
A system in which the important replication of data is rewarded with academic promotion may be what is needed to improve interactions along
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 39
6 Strategies for Change During the course of the workshop, several topics consistently arose during the presentations and subsequent discussions. Many of these topics also emerged in the final session of the workshop, a panel session that fea- tured discussants William Rutter, chief executive officer of Synergenics, and workshop speakers Geoffrey Duyk, Daniel Geschwind, and Todd Sherer. This final chapter of the workshop summary highlights the major themes and identifies specific strategies for improving the efficiency and effective- ness of the translation of genomic science (see Box 6-1). Aligning incentives for change Tension in connecting basic and translational sciences exists because of the difference between their inherent missions. For example, many neuro scientists are interested in learning about how the brain works to gain new knowledge, not necessarily because the information would help treat a disease, Geschwind noted. Understanding the basic science is critical, and that should not be overlooked in the discussion of translational research, he said. However, the current system is designed to incentivize novelty through the publication of papers in scientific journals, and that does not necessarily translate to social benefit, such as improved treatments for patients. The replication of results has value, and, in fact, the scientific method relies upon it, but “the incentives just aren’t aligned” to do this sort of work, said Geschwind. A system in which the important replication of data is rewarded with academic promotion may be what is needed to improve interactions along 39
OCR for page 39
40 EFFICIENCY AND EFFECTIVENESS OF GENOMIC SCIENCE TRANSLATION BOX 6-1 Pathways to Improving Genomic Science Translationa • unding milestone and outcome-driven research is key to supporting the cur- F rent state of technology with a more efficient biomedical research system. (Liu) • s a starting point for the systemwide integration of genomic science, general A practitioners could have the most impact for embracing genomics and bio informatics as decision support tools. (Huntsman) • esearch goals and incentives need to be aligned for the effective translation R of genomic science. (Geschwind) • niversities should consider including collaborative benchmarks as part of the U criteria for academic promotion to align tenure as an incentive for translational research. (Liu) • cientists should make better use of the high-quality data and resources that S are already publicly available and should be encouraged to share their own data for reuse to accelerate translation. (Butte) • he sharing and reuse of data should be incentivized through funding and T publication mechanisms. (Butte) • ompetition among teams, as opposed to individuals, can encourage multi- C disciplinary, collaborative efforts and be an effective tool for achieving grand translational challenges. (Burke) • ncorporation of multidisciplinary systems biology approaches to integrate I g enetic and phenotypic data will yield rapid progress for understanding com- plex diseases. (Geschwind) • atient advisory groups can convene stakeholders to help focus research on P projects that are most likely to provide patients with treatments today. (Furlong) • any advocacy groups are now participating in research, but enlistment of M these groups needs to be frictionless, and the information needs to be aggre- gated broadly to reduce the risk of fragmenting diseases so much that use of a systems-level approach to translation is not possible. (Terry) • y managing innovative risk-sharing research models, advocacy groups can B build evidence that is more likely to produce treatments that result in private- sector investment and development. (Sherer) • rogress in bringing products to market can be accelerated if the risk of invest- P ment is improved by adjustment of the health care ecosystem to adapt to new technology, improved globalization, and restructured regulation and reimburse- ment policies. (Duyk) • ew business models built on global networking, investments in disease pre- N vention diagnostics, and continued technological development will have a trans- formative effect on translational medicine and cost-effectiveness. (Scott) • ata will be made freely available if patients are given the right to own and D control their genomic data; enforcement strategies will not be necessary to encourage data sharing because patients will realize the value of sharing their information. (Scott) a The statements, recommendations, and opinions expressed here are those of the indi- vidual presenters and participants and are not necessarily endorsed or verified by the Institute of Medicine, and they should not be construed as reflecting any group consensus.
OCR for page 39
STRATEGIES FOR CHANGE 41 the basic science and translational pathways. The difficulty is that some journals have policies disallowing publication of reanalyzed data. Realign- ment in all areas is needed to line up goals and incentives for the effective translation of genomic science, from analysis and sharing and replication of data to grant proposal evaluation, publication, and academic promotions, Geschwind noted. Redefining Criteria for Tenure Although it was recognized that tenure is not the only mechanism for promotion that should be considered for change, the idea of reevaluat- ing the criteria used to determine tenure was discussed often during the workshop. As a starting point, universities could define a specific process and goals for achieving tenure that could include a collaborative research piece, Liu said. A workshop participant pointed to institutions that have changed incentives for tenure. For example, the University of Wisconsin system altered its tenure criteria to require that faculty engage in outreach to the community, and the University of Georgia system has done something similar. With strong leadership to put new incentives in place, changes are possible, he said. Important stakeholders to be included as part of the dis- cussion to effect these changes would be university presidents and professor unions, Duyk said. Participants also discussed the fact that professional societies are key to shifting incentives, as they can make recommendations to convey what is important to the societies’ membership. Catalyzing Change Through Funding, Collaboration, and Goal Setting As a first step to improve the translation of basic science, Geschwind suggested that if funding agencies, such as the National Institutes of Health (NIH), were to demand data sharing or collaborative multidisciplinary research, behaviors would change. “People will follow the money,” he said. To incentivize data sharing, funding agencies must recognize that it is a priority, in the sense that funding needs to be specifically allocated for sharing so that there is no excuse not to share the information. To accom- plish this, NIH and other organizations could convene short-term working groups to achieve implementation milestones in no more than a year to enforce the requirement that all funded proposals participate in data shar- ing, Geschwind said. Advocacy groups have a lesson on resource sharing to teach scien- tists, as many patients voluntarily share their medical information and in so doing have changed how data sharing is thought about, Geschwind observed. The advocacy groups would be essential for this process because
OCR for page 39
42 EFFICIENCY AND EFFECTIVENESS OF GENOMIC SCIENCE TRANSLATION they could emphasize the importance of sharing information for those who really need to see research results turn into clinical applications quickly, Geschwind said. Advocacy organizations are particularly well suited for determining who is working on particular projects and they also can play a major role in the reorientation of incentives, Liu said. Regulatory agencies could also encourage data sharing. Rutter called attention to a recent report of the President’s Council of Advisors on S cience and Technology that recommended accelerated approval for some therapeutics, thereby allowing patients and physicians to decide if they would like to take greater risks (PCAST, 2012). Rutter suggested that “as a trade-off for the accelerated approval, mandatory data sharing would be a condition” to provide full transparency of the data. A recent Institute of Medicine consensus study recommended making omics-based datasets available and more transparent (IOM, 2012). Encouraging the sharing of data would be a more efficient way to develop drugs, said Rutter, in addi- tion to the mechanisms being used by organizations today. Although sharing of data and resources is needed, it is a difficult bal- ance to achieve because tension also exists between collaboration and com- petition. Although some competition is needed, multidisciplinary efforts are often needed to achieve grand goals, Geschwind said. Competition among teams, as opposed to individuals, can be a very effective tool for making progress, Burke said. Duyk cited the Human Genome Project as a successful example of teams setting specific, clear goals. In addition to incentivization of indi- viduals, “appropriate incentives [should be given] to institutions that are creating a culture of teamwork and an effective mechanism involving a large team that is moving a translational research agenda,” Burke said. To address the issue of working in teams and how sharing among universi- ties and industry could be of benefit to both, Sherer pointed to a tension between top-down versus bottom-up management of the research enter- prise. Today, medical academic research is largely a bottom-up free-agent enterprise. The establishment of teams of experts to provide some top-down guidance for defining goals and providing clear, time-dependent milestones that can be held accountable to the taxpayers and others who fund the research would be beneficial, he said. Even though many systematic changes are needed, these will take time. It must not be overlooked that those with diseases now need accelerated development of treatments in the short term, Sherer said. He asked, “If we have human data, do we even need a transgenic animal model,” for example? “Can we figure out some way to do this faster to get results more quickly? That’s something we shouldn’t miss as we are making recommen- dations that might change 20 or 25 years from now.” The global infectious disease community provides case studies of how change can occur more
OCR for page 39
STRATEGIES FOR CHANGE 43 rapidly for self-organizing and how new technology can be used to create solutions to immediate issues, Duyk said. Perhaps rapid response teams would be useful in this case, too. Training of the next generation of scientists to operate in an innova- tive culture that connects basic and translational science teams was also discussed. Young people need to hear that leadership and management skills are important, said Duyk. If a young researcher wanted to take three courses in the business school, he or she would probably not be encour- aged to do so, as this would distract from time dedicated to science. “You have to tell people what you think is important. You have to incent people, whether it’s economically or otherwise, to do the right thing. And be care- ful. You can send the wrong messages very easily,” said Duyk. Sherer said that not everyone needs to be an innovator, a leader, a project manager, or a cutting-edge scientist. “You have to figure out what are the opportunities for people and then provide the training for those. Not everyone wants to lead their own lab. They may want to be part of the team.” Innovation in the current research system could also be catalyzed through the funding of fellowship grants and awards. MacArthur and Markey fellowships and awards from the Defense Advanced Research Projects Agency are examples of ways to provide promising researchers and institutions with enough funding and freedom to explore and innovate, Duyk said. NIH funds New Innovator and Pioneer Awards as well. Suc- cess then begets more success, especially because other people in a system tend to follow winners. The distribution of funding over a broad range of recipients will not necessarily allow for the creation of an effective research model that people will follow, Duyk suggested. Similarly, a grand challenges approach can attract talent to problems that need to be solved, especially if this approach outlines a clear problem that needs to be solved and a prize is awarded as an incentive. a better model for industry–academia collaboration Change that would help translate discoveries could also come from achievement of better industry–academia collaboration. Achievement of such collaboration is often a difficult task because of conflicts of interest, Sherer said. Duyk observed that pharmaceutical companies are interested in invest- ing money in academia, because it is cheaper than investing in biotechnology companies for some activities. However, complex conflict-of-interest rules can hinder such investments. The availability of better guidelines for collabo- ration that would address university technology transfer issues would also be useful. Technology transfer offices in universities need staff who can make good decisions and negotiate win-win agreements. The interface between the public and private sectors needs to be more functional, Duyk said.
OCR for page 39
44 EFFICIENCY AND EFFECTIVENESS OF GENOMIC SCIENCE TRANSLATION The model for translational research that the California Institute for Quantitative Biosciences (known as QB3) uses to foster innovation without distracting from the business models of pharmaceutical companies could be considered, said Rutter. QB3 is a joint project between the University of California, San Francisco, the University of California, Berkeley, and the University of California, Santa Cruz, which merged the efforts of the faculties of the three universities. The faculty network brings expertise in a range of disciplines and a central facility provides services in support of bioscience entrepreneurs. QB3 provides laboratory space, professional development training, mentorship, and legal services for those innovators who would like to start a business. Almost 60 companies have emerged from the initiative and are either still part of the network or have expanded. NIH could facilitate the development of similar facilities around core areas of technology and expertise, Rutter suggested. QB3 is a good example of how lean development could be valuable for the translation process, according to Rutter. Relatively small groups could initiate a project and carry it out under low-cost conditions until it becomes robust enough to be supported by venture capital or other groups. Several small companies also could be managed collectively, which would allow oversight by experienced managers in industry or a management group. Such strategies could help address the time-cost differential for development. In addition to thinking about when to develop a research finding for clinical use, it is also just as important to know when to cease a project. Duyk said that researchers are too soft about terminating projects and reorienting their organizations. “We’re about to go through a financial crisis that’s going to take a generation, probably, to resolve. And we should use that as an opportunity to rethink and make some hard decisions.” FINAL WORDS “We have to use those tools and figure out how we can get maximum impact out of the medical process,” said Duyk. “I think we don’t have enough iteration. If every drug development program is a rocket shot to the moon, you’re not going to do a lot of rocket shots to the moon.” To transform the translational pathway, “we have to really think about 21st-century solutions.” Terry added, “The solution is going to be us. . . . It will be somehow resolved within this kind of community, [with] this kind of group of stakeholders.”