Developing a 21st Century Neuroscience Workforce: Workshop Summary (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

5 Enhancing Training to Support Translational Research Key Highlights Discussed by Individual Participants • Undertranslation, overtranslation, and “pseudotranslation” are all common pitfalls in the translation of basic neuroscience discoveries (Ferrini-Mundy and Landis). • Many trainees do not understand the drug development pipeline, which creates inefficiencies in identifying and validating targets, and translating discoveries into treatments (Yocca). • Industry is increasingly turning to academia for help with identifying drug targets, validating those targets, and developing new pipelines (Yocca). • Training programs need to educate students about the full end-to-end process of drug discovery, development, and translation, even if in- dividual students are not necessarily involved in translational re- search (Yocca). NOTE: The items in this list were addressed by individual participants and were identified and summarized for this report by the rapporteurs. This is not intended to reflect a consensus among workshop participants. A central aspect of the neuroscience enterprise is translating basic science discoveries into therapies that can be used to treat humans. Joan Ferrini-Mundy and Story Landis described three common fallacies sur- rounding translation: undertranslation, overtranslation, and “pseudo- translation.” Frank Yocca, vice president of Neuroscience iMed at AstraZeneca Neuroscience, discussed the pharmaceutical industry’s re- 57 

58 DEVELOPING A 21st CENTURY NEUROSCIENCE WORKFORCE cent transition to targeting rare, gene-linked neurological and psychiatric diseases and their continued interest in partnering with academia to make new discoveries. He explains how neuroscientists can be trained to work in collaborative translational teams and gain the skills necessary for translational science. James Barrett, professor of pharmacology and physiology at the Drexel University College of Medicine, and Anthony Ricci, professor at the Stanford School of Medicine, described programs at their institutions dedicated to training students and postdoctoral re- searchers in translational science. UNDER-, OVER-, AND PSEUDO-TRANSLATION Ferrini-Mundy stated that neuroscientist trainees have to be mindful about both undertranslations—failure to translate promising discoveries from the lab into clinical therapies—as well as overtranslations— misguided attempts to use neuroscience discoveries to explain or solve every human problem. Meanwhile, Story Landis cautioned against the temptation to artificially generate a translational component into basic neuroscience research projects—a process she calls “pseudo-translation.” Such interventions are unlikely to ever be applicable to patients, said Landis, yet investigators propose these types of studies under the as- sumption that granting agencies such as NIH will not fund research that does not have translational relevance. The phenomenon is so widespread that Landis launched a program when she was the director of NINDS to encourage the submission of grant applications for more purely basic research, as described in Chapter 1, which she said has deep intrinsic value as the basis for the discovery and development of treatments (Landis, 2014). She added that trainees need to be taught the importance of basic research as well as how to design translational studies of real significant value. THE PHARMACEUTICAL INDUSTRY’S PIVOT IN TRANSLATIONAL NEUROSCIENCE According to several participants, there are numerous gaps in neuro- science expertise around translational science, and training students to have a greater understanding and knowledge in furthering innovative therapeutic development will be critical. The development of each new 

ENHANCING TRAINING TO SUPPORT TRANSLATIONAL RESEARCH 59 drug targeting a neurological disorder is a complex endeavor (see Figure 5-1), spanning, on average, 10 to 15 years and requiring an investment of $1 billion to $2 billion, said Barrett. Given the high costs and risk (less than 10 percent success rate) in developing central nervous system drugs, in addition to the general challenges in translational neuroscience (see Box 5-1), the pharmaceutical industry has subsequently reduced its neu- roscience research and development spending over the past decade (Abbott, 2011; Miller, 2010). As a result, Yocca noted that some pharmaceutical companies down- sized their neuroscience division, often having to significantly decrease the workforce (e.g., AstraZeneca’s neuroscience division decreased from more than 700 scientists in the late 2000s to approximately 50 today). Atul Pande, chief medical officer and executive vice president of Tal Medical, Inc., emphasized that the impact of this withdrawal in neurosci- ence research and development will be felt progressively over the com- ing years, which could affect trainees and postdoctoral researchers who would like to pursue a career in industry. The problem, according to Yocca, has not been a lack of commitment, but rather, most drugs fail in phase II trials because they are found to be ineffective. The major chal- lenge in developing effective treatments is a general lack of bi- omarkers—biological signatures that indicate the progression of a disease, he added. For example, no biomarkers are known to exist for many large-market neurological diseases such as Parkinson’s disease and Alzheimer’s disease. With no way to quantify disease progression or to stratify patients in- to various disease stages, Yocca noted that it is difficult to determine the effect candidate treatments are having in patients. He cautioned that many drugs might show efficacy in some people, but there is variability that cannot be explained without a way to stratify patients, making de- velopment of such drugs risky. 

60 DE EVELOPING A 21st 2 CENTURY Y NEUROSCIEN NCE WORKFORCE FIGUR RE 5-1 The steeps involved inn developing a drug, from pree-exploration too clinicall development.. NOTE: DT = develo opmental therap peutics, PI = pphase I, PII = phase II, PIIII = phase III. I SOURC CE: James Barrrett presentatio on, Drexel Uniiversity, Octobber 29, 2014. BOX 5-1 Neuroscien nce Presents Several Tran nslational Cha allenges entific Scie • Biologica al complexity and a nonvalida ated targets • Poor preeclinical modells • Challengge of the blood d/brain barrier • Direct ex xamination of drug d exposure e and target en ngagement nical Clin • Patient recruitment r • Patient heterogeneity h • Disease is advanced when w symptom ms appear • Capturingg therapeutic effects e on cliniccal scales with high variabilityy Low w productivity • Long cyc cle times • High cos sts • Low prob bability of succ cess SOURCE: Frank Yocca presenttation, AstraZe eneca Neurosccience, Octoberr 28, 2014. 

ENHANCING TRAINING TO SUPPORT TRANSLATIONAL RESEARCH 61 As a consequence of the dismal success rate in developing drugs for neurological disorders, increased regulatory incentives offered by the Rare Disease Act, and the chance to leverage discoveries of rare disease mechanisms into treatments of more prevalent diseases, neuroscience translation has pivoted from big diseases such as depression and schizo- phrenia to smaller diseases, according to Yocca. Many pharmaceutical companies, including AstraZeneca, are now employing smaller teams to develop treatments for rare neurological diseases that had been largely overlooked in the past due to low potential profit margins (this transition is summarized in Box 5-2). AstraZeneca’s general approach to drug dis- covery, which adheres to the “Five Rs,” is summarized in Figure 5-2. Yocca also said that the search for neurological drugs is being modeled after drug development in oncology in which pharmaceutical companies are pursuing smaller neurological diseases that have a genetic basis. See Box 5-3 for the list of novel approaches to translation that Yocca shared with workshop participants. BOX 5-2 Opportunities for Changing the Approach to Training in Translational Neuroscience • Large internal teams working on literature targets and follow-on approaches  Small internal teams collaborating with academic and biotech part- ners working on genetically driven innovative targets • Limitations driven by rigid disease strategies  More opportunistic approaches to find tractable targets regardless of disease state • Template approaches  Smart discovery and development strategies (translational focus) • Focus on larger diseases driven by peak year sales  Focus on smaller, genetic-based diseases driven by “line of sight” and return on investment SOURCE: Frank Yocca presentation, AstraZeneca Neuroscience, Oc- tober 28, 2014. 

62 DE EVELOPING A 21st 2 CENTURY Y NEUROSCIEN NCE WORKFORCE FIGUR RE 5-2 The “Five Rs” to drug g discovery annd developmentt. NOTE: CDTP = Co ontinuing Day y Treatment P Program; DTPPP = Diphtherria- Tetanu us-Pertussis-Polliomyelitis; PD D = Pharmacoddynamics; PHC C = Personalizzed Health Care; PK = Phharmacokineticcs. SOURC CE: Frank Yo occa presentatiion, AstraZeneeca Neurosciennce, October 228, 2014, adapted a from Cook C et al., 201 14. BOX 5-3 Novel Approaches A to t Neuroscien nce Translatiion • Large-scale unbiased app proaches to da ata collection and analysis and DNA seq quencing • Optogenetics s to focus on circuits c involve ed in diseasess • Identification of biochemic cal pathways iinvolved in dissease patho- genesis • Akin to cancer, mutations within cells m may be proving g to be more important to therapy t than the t cell of orig in • The best way y to determinee convergent p pathophysiological mecha- nisms lies in starting with genetic g discovveries • Multilevel analysis to eluciidate the caussal pathway frrom mutation to behaviorall disorder • Reprogramm ming skin cells from patients into functionaal neurons af- fords us the opportunity o to develop cellu lar disease models 

ENHANCING TRAINING TO SUPPORT TRANSLATIONAL RESEARCH 63 • Substantially reduce investment risk by concentrating drug devel- opment efforts either on smaller, biologically stratified subsets of patients guided by genetic findings, or on specific circuits and synaptic processes SOURCE: Frank Yocca presentation, AstraZeneca Neuroscience, October 28, 2014, based on materials from Hyman (2012) and Karayiorgou et al. (2012). FOSTERING PARTNERSHIPS BETWEEN INDUSTRY AND ACADEMIA Industry is increasingly looking to partner with academia, which provides a wide variety of expertise in skills necessary for translational science, said Yocca (see Box 5-4). According to Barrett, industry rou- tinely turns to academia to identify drug targets, validate those targets, and develop new pipelines. Indeed, academia now generates the majority of the basic science discoveries that are being translated into new medi- cines (Silber, 2010). By continuing in this role, academic institutions have an opportunity to train students to transition into careers within in- dustry. However, Richard Tsien, professor of neuroscience at the New York University Langone Medical Center, emphasized that even though there are many trainees with the right skills and an interest in working on disease-targeted research challenges remain in matching trainees to the right company. Tsien added that similar to the decreased number of posi- tions in academia industry might not be a secure career option for recent graduates. Yocca said that translating genetic advances into drug discovery and development programs is going to require close collaboration between disease biology experts in academia and the pharmaceutical industry. As an example of the approach to discover targets in gene-linked disorders, Yocca detailed a collaboration between AstraZeneca and the Lieber In- stitute for Brain Development. The researchers look for genes of inter- est—either from patients in clinical studies or by using reverse translation (i.e., back-translation)—to drive RNA sequencing in the search for a transcript associated with illness data or genetic risk. Those transcripts can be used to derive molecular mechanisms of association that can be developed and tested in cell-based models and animal models. One specific example of this approach that Yocca mentioned was using human induced pluripotent stem cell (iPSC)-derived neurons from clinically and 

64 DEVELOPING A 21st CENTURY NEUROSCIENCE WORKFORCE genetically characterized subjects to probe mechanisms associated with genetic risk in neuropsychiatric disorders. Another novel program that highlights the symbiosis between indus- try and academia is Johnson & Johnson’s Innovation Centers.1 These centers in Boston, London, Shanghai, and Silicon Valley, act as regional incubator hubs that bring in entrepreneurs and local start-up companies and support a diversity of new ideas hatched in the labs of local scientists by offering access to costly equipment and services. The purpose of these centers is to fuel innovation and breakthrough science. BOX 5-4 Expertise Needed for Translational Science in Neuroscience • Neuroscientists with expertise in informatics/statistics • Neurobiologists with expertise in genetic manipulations (e.g., Clus- tered Regularly Interspaced Short Palindromic Repeats, or CRISPR) • Cell biologists with expertise in neuroscience and neurodevelopment • Cell biologists with expertise in stem cells • Neurophysiologists with system modeling expertise • Clinicians with expertise in neuroscience and neurodevelopment • Neurodevelopmental processes • Neurophysiologists/neuropsychologists • High-quality clinical studies • Novel pharmacology and repositioning tools • Genetics and patient segmentation • Objective end-points and biomarkers • High-quality diagnostics and patient segmentation • Response biomarkers • Clinical neurophysiologists and clinical psychologists • Functional imaging • High-quality clinical studies • Novel treatment strategies • Patient segmentation • Behavioral analysis, animal models in neuroscience and data cap- ture and data analysis SOURCES: James Barrett presentation, Drexel University, October 29, 2014, and Frank Yocca presentation, AstraZeneca Neuroscience, October 28, 2014. 1 See http://www.jnj.com/partners/innovation-centers (accessed November 3, 2014). 

ENHANCING TRAINING TO SUPPORT TRANSLATIONAL RESEARCH 65 Challenges and Opportunities to Improve Training in Translational Neuroscience One challenge for translational neuroscience is the fact that many trainees do not understand translational science, according to Yocca. That is, they are not fully aware of the drug discovery and development pipeline. Although some trainees go through graduate school without knowing what target validation is, other trainees who actually do target validation may not understand the needs of the next person on the pipe- line. Yocca noted that this lack of the big picture of the overall process, as well as the fact that scientists speak different languages depending on where in the pipeline they do research, creates inefficiencies in the dis- covery and development pipeline. To overcome these challenges, train- ing programs need to educate students about the full scope of translational science, even if individual students are not necessarily in- volved in translational research, he added (see Box 5-5 for an overview of Drexel University’s Master of Science in Drug Discovery and Devel- opment program). Incorporating pharmacology and genetics into gradu- ate school curricula will also better prepare trainees to understand and perform translational research. Yocca pointed out that the resulting trans- lational research will also be enhanced, by providing training in transla- tional science and neuroscience to experts in other fields such as cell biologists and clinicians, neurophysiologists, and clinical psychologists. BOX 5-5 Program Example: Drexel University Drexel University offers a Master of Science in Drug Discovery and Development that provides the rigorous scientific and technical training necessary to facilitate a smooth transition to a productive ca- reer in the biotechnology or pharmaceutical industry. Barrett explained that the primary strength of the program is its integration of the drug discovery and development disciplines as well as emerging disciplines within neuroscience and other biomedical sciences. Below is a list of topics covered in the program’s core courses. Beyond merely bringing together expertise on the laboratory side of drug development, the program engages trainees in the entire dis- covery and development practice by enlisting the participation of the school of medicine, the school of business law, the school of public health, and the department of biomedical engineering. 

66 DEVELOPING A 21st CENTURY NEUROSCIENCE WORKFORCE The program strives to give trainees real-world experiences, not- ed Barrett. Faculty present students with detailed drug discovery case studies—both failures and successes. Trainees assemble into teams that bring together expertise in medicinal chemistry, project manage- ment, and business. Teams are assigned specific projects and work together to come up with plans for identifying and validating targets. Teams then present their ideas—how to formulate a drug, how to market it, what are the liabilities and risks—to a panel of faculty who judge the viability of the projects. The program also leverages unique features of the Drexel com- munity. In coordination with the school of public health, the program hosts a section on pharmacoepidemiology, which uses epidemiologi- cal data to look at drugs and off-purpose targets. For example, a re- cent epidemiological statistic showed that schizophrenics on long- term antipsychotics have a lower incidence of cancer. The program also works with Drexel’s school of media arts and design to create ways to “gamify” the drug discovery process. Echoing Yocca’s declaration that biomarkers are critical to trans- lation, Barrett said a primary focus of the program is training students to exploit biomarkers in whatever form they can be found, whether they are imaging indicators or genetic markers. Barrett said the program also emphasizes the analysis of behav- ior, which he called “the ultimate expression of psychiatric and neuro- logical disorders” when developing drugs. Single-dose treatments in animal models are often poor indicators of clinical success, but the ability to predict clinical outcomes increases to 70 percent if behavior- al data are collected to generate full-dose response curves. Barrett cited a recent commentary in Nature Neuroscience (Gomez-Marin et al., 2014) that makes the case for behavior being the “foundational problem of neuroscience” and describes opportunities in big behav- ioral data created by innovations in technology. SOURCE: James Barrett presentation, Drexel University, October 29, 2014. Program Example: Stanford University School of Medicine Ricci discussed some of the innovative approaches that Stanford University is taking in training students in translational neuroscience. The overall philosophy of Stanford’s neuroscience program is to create experts and leaders in their respective career track, regardless of whether the track is inside or outside academia. Students are given the freedom to explore whatever topics and technologies they think will best suit their training needs. To accommodate these explorations, the program is asso- 

ENHANCING TRAINING TO SUPPORT TRANSLATIONAL RESEARCH 67 ciated with 25 departments and schools, spanning the biological, physi- cal, and informational sciences. Students are also given the freedom to select faculty advisers from any department of interest and participate in any internships or similar opportunities outside of the program. In the first year, each student selects a neuroscience topic, which is usually, but not always, disease oriented. They then choose three mini- courses—from the list of genetics, translational, behavior, computational, cognitive, systems, neuroanatomy, molecular, cellular, and develop- ment—to explore per quarter. At the end of each quarter, students pro- duce reports on their topic incorporating lessons from the three mini- courses. They also produce a yearly report and presentation incorporating all of the mini-courses. One notable aspect of this program is that transla- tion is considered just another facet of neuroscience. Optional transla- tional courses also offered are Neurobiology of Disease, Current Issues in Aging, Molecular Mechanisms of Neurodegenerative Diseases, and Experimental Stroke. Finally, Ricci noted that Stanford also offers three unique profes- sional development programs that prepare trainees for careers in transla- tional neuroscience: Master of Medicine,2 Biodesign Program,3 SPARK Program.4 SUMMARY Neuroscience is in an era of growth and popularity. Given the scien- tific progress in the field, trainers seek to develop and strengthen training programs to better prepare the 21st century neuroscience workforce. Stevin Zorn concluded the workshop by saying, “No time in our history of neuroscience have we ever been more equipped to make the kinds of discoveries that are needed to understand the brain and the underlying diseases that we don’t [yet] fully understand. Right now is the time for us to energize a new generation of neuroscientists by putting the call out, just like President Kennedy did in 1961, so that we can build our training programs, our neuroscientists, and the field itself, so that it is capable and ready to face these challenges at this unprecedented and exciting time.” 2 See http://msm.stanford.edu (accessed October 29, 2014). 3 See http://biodesign.stanford.edu/bdn/index.jsp (accessed October 29, 2014). 4 See http://sparkmed.stanford.edu (accessed October 29, 2014).