4
Discovery Research for Rare Diseases and Orphan Product Development

I could see there was a transformation of cancer treatment on the horizon thanks to breakthroughs in biochemistry and genomics. I wanted to be part of that, which is why I was a physician-researcher… . By the late 1980s, C.M.L. [chronic myeloid leuke-mia], though rare, was a cancer that scientists knew a lot about. We knew, for instance that a chromosomal abnormality existed in every C.M.L. patient. We knew that this abnormality created an enzyme that caused the uncontrolled growth of cancer cells… . If you want to develop targeted chemotherapies, C.M.L. is the disease to study. We know the most about it—and, if we can figure out a way to block this enzyme, we can turn off the cancer switch.

Interview with Brian Druker (Dreifus, 2009)


The research undertaken by Brian Druker and his colleagues and predecessors offers a classic example of the foundation that basic research builds for the subsequent development of therapies for rare diseases. Breakthroughs in biochemistry and genomics, as well as advances in computational tools, have transformed the process of research and drug development. The process begins with basic laboratory studies that reveal the molecular mechanisms of disease, which related to a chromosomal abnormality in the case of chronic myelogenous leukemia (CML). This foundation leads to the discovery of biomarkers for rare conditions and the discovery of potential biological targets on which drugs can act. The target in CML is a rogue enzyme created by the mutated chromosomes, which triggers uncontrolled cell growth. Once a target is defined, the process shifts from basic research



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
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 111
4 Discovery Research for Rare Diseases and Orphan Product Development I could see there was a transformation of cancer treatment on the horizon thanks to breakthroughs in biochemistry and genomics. I wanted to be part of that, which is why I was a physician- researcher. . . . By the late 0s, C.M.L. [chronic myeloid leuke- mia], though rare, was a cancer that scientists knew a lot about. We knew, for instance that a chromosomal abnormality existed in every C.M.L. patient. We knew that this abnormality created an enzyme that caused the uncontrolled growth of cancer cells. . . . If you want to develop targeted chemotherapies, C.M.L. is the disease to study. We know the most about it—and, if we can figure out a way to block this enzyme, we can turn off the cancer switch. Interview with Brian Druker (Dreifus, 2009) The research undertaken by Brian Druker and his colleagues and prede- cessors offers a classic example of the foundation that basic research builds for the subsequent development of therapies for rare diseases. Breakthroughs in biochemistry and genomics, as well as advances in computational tools, have transformed the process of research and drug development. The pro- cess begins with basic laboratory studies that reveal the molecular mecha- nisms of disease, which related to a chromosomal abnormality in the case of chronic myelogenous leukemia (CML). This foundation leads to the discovery of biomarkers for rare conditions and the discovery of potential biological targets on which drugs can act. The target in CML is a rogue enzyme created by the mutated chromosomes, which triggers uncontrolled cell growth. Once a target is defined, the process shifts from basic research 

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS to the discovery of a therapeutic approach. Imatinib mesylate (Gleevec), the drug discovered by Druker, specifically deactivates the enzyme target in CML. It was approved by the Food and Drug Administration (FDA) in 2001 and is now used not only for CML but also for other rare cancers. Increased knowledge of kinase inhibitors (of which imatinib was the first) is supporting the development of more potent, second-generation drugs for CML that may also be less susceptible to resistance (Sawyers, 2010). Today, as a result of scientific and technological innovations, much of the basic research initially undertaken with CML could be done more quickly, inexpensively, and easily. For example, identification of the genetic cause of conditions that are clearly inherited used to involve speculative approaches and laborious analytical tools. The sequencing of the human genome has spawned an array of rapid and relatively inexpensive DNA analysis tools that have the potential to foster more targeted and efficient therapeutics development for rare diseases. Advances in the scientific un- derstanding of disease mechanisms likewise are helping researchers focus more efficiently and effectively on potential therapeutic targets. As a result, the future holds the promise of continued innovation that will further ac- celerate biomedical research to the benefit of patients with rare as well as common diseases. As discussed in Chapter 1, research on rare diseases can illuminate dis- ease mechanisms and therapeutic opportunities for more common diseases. Box 4-1 briefly summarizes several additional examples of rare diseases research that have yielded broader knowledge. Many of the same approaches and techniques are used to study both rare and common diseases, but research on rare diseases faces some spe- cial barriers and constraints. One is the sheer number of rare diseases, an estimated 5,000 to 8,000. Many of the challenges stem from the low prevalence that is the defining characteristic of rare diseases. Particularly for extremely rare conditions, the small numbers of affected individuals means a dearth of biological specimens, which severely limits studies of disease mechanism and etiology. Small numbers also constrain epidemio- logic research and clinical trials as highlighted in Chapters 3 and 5. Other challenges include the limited funding for research and a limited number of investigators committed to the study of rare conditions. The basic research tools available to investigators have advanced dra- matically over the past 20 years, with new approaches continuing to evolve, both in the laboratory and from the use of computational biology. Along with new and better tools, models for supporting discovery research have also undergone a transformation in recent years. This chapter briefly ex- amines the implications for rare diseases research of a number of current research strategies for both target discovery and therapeutic discovery. The next chapter focuses on product development, particularly from the per- spective of companies and their academic and government collaborators

OCR for page 111
113 DISCOVERY RESEARCH FOR DEVELOPMENT BOX 4-1 Examples of Research on Rare Diseases with Implications for Treatment of Common Conditions Some of the most effective treatments for coronary artery disease (a very com- mon condition) were first established during the study of a rare condition called familial hypercholesterolemia. The disease was ultimately linked to mutations in the gene for the low-density lipoprotein receptor that coordinates the uptake of cholesterol from the blood. This work laid the foundation for the development and use of drugs (specifically, statins) that inhibit the rate-limiting enzyme in choles- terol synthesis, hydroxymethylglutaryl (HMG)-CoA (coenzyme A) reductase, in the lowering of circulating cholesterol and the prevention of coronary artery disease and myocardial infarction (Stossel, 2008). Patients with a rare condition called osteoporosis-pseudoglioma syndrome have loss-of-function mutations in the low-density lipoprotein receptor-related protein-5 (LRP5), while mutations causing rare conditions associated with high bone mass and density produce increased LRP5 function. Subsequent work showed that LRP5 normally inhibits serotonin production in the gut. Inhibition of gut serotonin production has emerged as a promising treatment for common causes of os- teoporosis including the loss of bone mineral density associated with aging and menopause (Haigh, 2008; Long, 2008). Aortic aneurysm is the cause of death in about 1 to 2 percent of individuals in industrialized countries, but its cause is largely unknown and medical treatments are lacking. During the study of Marfan syndrome, a rare connective tissue disorder associated with a high risk of ascending aortic aneurysm and tear, researchers showed that aneurysm development and progression is associated with increased activity of transforming growth factor β (TGF-β), a molecule that instructs cellular behavior. It was subsequently shown that interventions that in- hibit TGF-β, including administration of a neutralizing antibody or the angiotensin II type 1 receptor blocker losartan, could attenuate or prevent many manifesta- tions of Marfan syndrome in mouse models. Responsive Marfan phenotypes included aortic aneurysm, skeletal muscle myopathy, pulmonary emphysema, and degeneration of the mitral valve. This work prompted the launch of the first clinical trial for Marfan syndrome based upon a refined understanding of disease pathogenesis, specifically assessing the efficacy of losartan in attenuating aortic root growth. Alteration of TGF-β activity was subsequently linked to other rare (e.g., Loeys-Dietz syndrome) and common (e.g., bicuspid aortic valve with aneu- rysm) presentations of aortic aneurysm (Jones et al., 2009). Losartan has also proved effective in the treatment of TGF-β-induced myopathy in a mouse model of Duchenne muscular dystrophy (Cohn et al., 2007). who are evaluating and undertaking the complex work needed to transform promising research discoveries into products that are safe and effective for patients in need. All along this continuum from basic research through clinical trials, infrastructure and innovation are needed to accelerate the development of therapies for people with rare diseases. The discussion here

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS focuses on the role of government, industry, academic investigators and in- stitutions, and advocacy groups. Other groups also contribute, for example, organizations such as the American College of Medical Genetics. Both this and the next chapter discuss the infrastructure for rare diseases research and orphan product development and “innovation plat- forms” to encourage and support collaborative work. Such collaboration is needed to bridge the gulf—sometimes referred to as the “valley of death”—between basic research findings and beneficial products, especially in the stages that precede clinical studies of efficacy. Early initiatives to bridge the gulf included public policies such as the Amendments to the Patent and Trademark Act of 1980 (P.L. 96-517, commonly known as the Bayh-Dole Act). That legislation encouraged cooperation among academic institutions, other nonprofit organizations, and small businesses to com- mercialize research discoveries funded by the federal government (Schact, 2007). Efforts continue to successfully engage government, academic, nonprofit, and commercial entities as collaborators in translating research discoveries into safe and effective drugs and medical devices. First, how- ever, must come the discoveries. TARGET DISCOVERY Most rare diseases have a genetic etiology, but the molecular patho- genesis has been defined for a relatively small number of rare diseases. For most of this small group, a specific gene alteration is recognized as respon- sible for the disorder, and for a subset, understanding of the pathogenesis extends to identification of the function of the affected gene product. For an even smaller subset, investigators have described targets such as spe- cific molecules or physiologic pathways that are amenable to therapeutic modification. The next sections discuss some particular areas of research advances and their prospects for increasing understanding of the molecular pathogenesis of rare diseases. Such understanding provides the basis for modern drug discovery. Traditional Genetic Studies Because most rare diseases are caused by defects in a single gene, identification of a mutated gene is the logical starting point for research in most cases. Although the standard approach to mapping the chromosomal location of the gene of interest has used candidate gene analysis or linkage analysis, these methods are inherently slow and often cumbersome. Many factors can limit the utility of genetic mapping studies for rare disorders, notably the lack of large families with multiple affected, surviv- ing individuals. Early death and other disease-related causes of reduced

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT reproduction contribute to this lack as does the general decline in family size associated with economic and social development. Recent technologi- cal advances have enabled researchers to employ genome-wide association studies to identify genetic variation that contributes to the pathogenesis of common disorders, as well as some of the most prevalent rare diseases such as juvenile idiopathic arthritis (see, e.g., Thomson et al., 2010). These studies depend on large patient populations and on an inherent assumption that the predisposing alleles or haplotypes are both ancient and shared among unrelated affected patients, effectively precluding this approach for small patient populations with high locus or allelic heterogeneity. Impaired reproductive fitness, a feature of many rare disorders, imposes allelic het- erogeneity and would therefore implicitly disqualify this approach as a strategy for research on these disorders. Although some critics of genome- wide association studies argue that they have not been terribly informa- tive with regard to individual risk of disease, the studies have highlighted pathways whose relevance to a particular disease had been unsuspected (Hirschhorn, 2009). Fortunately, additional tools for genetic research are now available. Study of Modifier Genes and Epigenetics Variation in secondary genes can alter primary gene effects and related pathways and can attenuate or mask underlying disease predisposition. Studies of these secondary genes are likely to inform the development of novel therapeutic strategies. For many rare and common disorders, there is considerable phenotypic variation among individuals with the same un- derlying primary disease gene mutation. This can be particularly striking when wide phenotypic variation is seen within individual families. For example, in X-linked adrenoleukodystrophy (a metabolic disorder that causes neurological damage), some affected family members have onset of neurodegeneration and death in childhood, whereas others show mild manifestations of disease such as isolated adrenal insufficiency that first manifests in adulthood. Yet other family members may be entirely asymp- tomatic (Maestri and Beaty, 1992; Moser et al., 2009). The study of modifier genes can be facilitated through the use of in- bred mouse strains that often show wide variation in disease severity based upon the genetic background on which a primary disease-causing muta- tion occurs. Animal models also offer the ability to use targeted genetic or pharmacologic perturbations to test focused hypotheses regarding modifier genes and pathways. The identification of modifier genes is of particular value in rare diseases, where diagnosis is already difficult due to the small number of cases. Beyond germline genetic variation, modification of DNA (e.g., DNA

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS methylation, histone acetylation) contributes to rare disorders such as the Prader-Willi syndrome and Angelman syndrome (Adams, 2008), but it may be even more important as a contributory factor in modulating gene expres- sion and, therefore, disease predisposition and severity. These epigenetic modifications are likely acquired as the result of an array of exposures (e.g., prenatal exposure to tobacco smoke) and experiences (e.g., stress). Inves- tigators are now using microarray and sequencing to analyze methylation patterns as biomarkers that can have clinical value. Whole Genome Sequencing, Gene Expression Analysis, and Exome Sequencing Whole genome sequencing provides a complete analysis of the en- tire complement of an individual’s DNA. It can now be used to identify genetic variants associated with rare diseases in individual patients or families (Lupski et al., 2010; Roach et al., 2010). The cost for sequencing has fallen dramatically, but it remains resource intensive and challenging because each exome contains a large number of polymorphisms (variants), only one of which is typically the primary gene alteration (Lifton, 2010; Wade, 2010). Microarray methods, which are used to comprehensively assess which genes are transcribed and which are not active in making proteins, are not diagnostic for genetic diseases. They can, however, be helpful in working out pathways that are dysfunctional in both genetic and acquired rare disorders (Wong and Wang, 2008). Experimental methods to interrupt gene expres- sion in cell culture systems and animal models include the introduction of target-specific microRNAs, a tool that has been used to confirm the role of genes and pathways in the pathogenesis and modulation of disease. Exome sequencing is a promising new approach to the search for disorder-causing genes for rare diseases (Kuehn, 2010; Tabor and Bamshad, 2010). The method focuses on the less than 5 percent of the genome that actually codes for protein. With this method, identification of genes associ- ated with disorders of previously unknown etiology is possible using DNA from as few as two to four patients (Ng et al., 2009; Johnston et al., 2010). This approach provides a particular advantage to rare diseases, given that biological specimens are often scarce. It is expected to accelerate the rate of identification of gene defects for rare diseases. Proteomics and Metabolomics Researchers have made significant progress in the cataloging of genetic variation and its correlation with disease predisposition, initiation, and progression. Parallel initiatives for protein variation are also important.

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT Proteomics is the science of detecting, identifying, and quantifying the products of gene translation and represents another approach to uncovering variation that underlies the pathogenesis of rare diseases. A single gene can generate an array of protein species based upon alternative translational start and stop sites and splicing. The derived proteins can be further diver- sified in relative abundance, structure, and function by posttranslational modifications including phosphorylation, glycosylation, acetylation, and tagging for degradation. Proteomics analyses can detect primary pertur- bations that cause disease (e.g., congenital disorders of glycosylation), pathogenetic or compensatory pathway activation (e.g., the activation of kinases through quantitative analysis of substrates for phosphorylation), and candidate proteins for validation as biomarkers to aid in diagnosis, prognostication, or therapeutic trials (e.g., newborn screening by tandem mass spectrometry or detection of increased circulating levels of cardiac muscle-specific enzymes after myocardial infarction) (see, e.g., Duncan and Hunsucker, 2005; Haffner and Maher, 2006; Suzuki et al., 2009; Van Eyk, 2010). One challenge is that proteomic analysis requires expensive equip- ment (e.g., mass spectrometry) and data analysis tools, which means that this technique is usually centralized in special laboratories. Metabolomics involves the study of the small-molecule metabolites found in an organism. As in proteomics, mass spectrometry can be used As to detect abnormal metabolic products, to diagnose rare diseases, and to understand alterations in relevant biological pathways. An example is the elucidation of a series of synthetic enzyme deficiencies that result in the production of abnormal bile acids leading to serious liver, neurologic, con- nective tissue, and nutritional disorders (Heubi et al., 2007). Systems Biology and Bioinformatics With the aid of translational bioinformatics (Schadt et al., 2005a; Vodovotz et al., 2008), the construction of molecular networks and path- ways relevant to specific rare disorders is increasingly possible. Bioinfor- matic analyses of data from gene expression arrays, proteomics studies, and clinical observations on patients with rare diseases can define signatures of fundamental disease mechanisms (Dudley et al., 2009; Patel et al., 2010; Suthram et al., 2010). Integration of this information with signatures of drug activities or therapeutic responses could intuitively promote discovery regarding the etiology, pathogenesis, and treatment of unclassified or poorly understood disorders (Schadt et al., 2005b). For example, if two diseases show overlapping or identical signatures, established treatments for one might benefit the other. Drugs that show signatures that oppose those seen for certain diseases emerge as candidate therapies. Bioinformatic methods can screen known chemical compounds for structural characteristics that

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS predict desired drug activities that are potentially beneficial for patients with rare diseases. Identification of drugs with overlapping signatures will promote the informed testing and substitution of agents that might show greater efficacy or other desirable characteristics such as reduced toxicity. Through these approaches, it should be possible to identify multiple inter- vention target sites for some disorders. Conversely, studies of biological networks can also identify common pathways for multiple rare diseases that are biologically related. For ex- ample, a more comprehensive understanding of the molecular basis for lysosomal function may provide an opportunity for interventions that are beneficial for an array of lysosomal disorders (Sardiello et al., 2009). More broadly, this capability may open the door for the discovery of single therapies that can benefit multiple rare disorders and, potentially, also more common diseases. The promise of systems biology is built on the availability of molecular and genetic data, combined with the development of valid computational methods for integrating these data into predictive models of disease (Schadt, 2005a). Although most genomic sequences are available in publicly acces- sible databases, many experimental biological data as well as clinical trials data are not collected or stored in a way that ensures broad access to the information. Thus, as discussed later in this chapter, the infrastructure for rare diseases research and product development should include structures and processes for sharing research resources, including data and biological specimens. THERAPEUTICS DISCOVERY Once basic research is performed and findings implicate a specific biological target, which could be an enzyme, a product of a biochemical pathway, an altered gene, an epigenetic mechanism, or a combination of the above, then the search begins for an appropriate therapeutic agent. Sometimes recognition of a molecular defect can point directly to potential therapies. Effective therapies can either inhibit deleterious or excessive functions or restore missing functions, both of which can result from gene mutations. In the former category, for example, the finding that transforming growth factor β (TGF-β) plays a role in the development of aortic root aneurysms in Marfan syndrome led to studies of an inhibitor of TGF-β (angiotensin II type I receptor inhibitor losartan) that are currently in phase III trials (Dietz, 2010). A large number of monoclonal antibodies are available to modulate exuberant immunologic, inflammatory reactions in rare as well as more common diseases. Imatinib successfully treats CML and other cancers by inhibiting tyrosine kinases. Increasingly, small interfering RNAs

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT (siRNAs) are being tested as inhibitory drugs, systemically and by direct instillation into the central nervous system and other tissues (Dykxhoorn and Lieberman, 2006). A few disorders can be treated with “curative” therapies that restore missing functions. Examples of such conditions and treatments include congenital hypothyroidism (replacement of thyroid hormone), bile acid synthetic enzyme deficiencies (oral bile acid therapy), biotinidase deficiency (biotin vitamin therapy), and celiac disease (dietary avoidance therapy). Still, for most rare diseases an obvious and easy therapeutic remedy is elu- sive or beyond current scientific capabilities. As discussed in Chapter 2, for most rare conditions, treatment is limited to symptomatic therapies (see, e.g., Campeau et al., 2008; Dietz, 2010). High-Throughput Screening of Compound Libraries When a potentially relevant target for an identified disease is validated, chemists then mount an intensive search for chemicals that might modify the target or targets. They screen vast compound libraries that are primar- ily assembled and secured within pharmaceutical companies to develop a list of potential “hits” that might some day become a “lead compound” and eventually new medicine, almost always after extensive “medicinal” chemistry to improve various properties of the parent compound and turn it into a drug suitable for testing in humans. This sophisticated process can be divided into three distinct steps: (1) development and maintenance of large compound libraries, (2) specific assay development, and (3) high- throughput screening. Assays are analyses that quantify the interaction of the biological target and the compound that the researchers are investigating. They also might measure how the presence of the compound changes the way in which the biological target behaves. The chemical compounds tested in these assays are maintained in large compound libraries, which may contain more than 5 million chemicals. Products from natural sources such as plants, fungi, bacteria, and sea organisms can be integrated within compound libraries. Most compounds, though, are derived through the use of chemical synthesis techniques, in which researchers create chemical compounds by manipu- lating “parent” chemicals. They might also use combinatorial chemistry, in which researchers create new but related chemical compounds and test them rapidly for desirable properties. Sometimes companies will provide compounds to laboratories for low-volume screening, or alternatively the assay for the molecular target can be provided to a company where it will be optimized for high-throughput screening. Testing the expanding number of available biological targets against thousands or millions of chemical entities requires highly sophisticated

OCR for page 111
0 RARE DISEASES AND ORPHAN PRODUCTS screening methods. Researchers use robotics, for example, to simultane- ously test thousands of distinct chemical compounds in functional and binding assays. Academic researchers with expert knowledge of specific pathways may guide the development of assays in collaboration with indus- try. The chemical compounds identified through this kind of screening can provide powerful research tools that contribute to a better understanding of biological processes. This, in turn, may lead to new targets for potential drug discoveries. The purpose of this chemistry stage is to refine the compound. Hun- dreds and possibly thousands of related compounds may be tested to determine if they have greater effectiveness, reduced toxicity, or improved pharmacological behavior, such as better absorption after a patient takes the drug orally. To optimize the molecules being investigated, scientists use computers to model the structure of the lead compounds and how they link to the target protein—an approach to structure-based design known as in silico modeling (silico referring to the silicon technology that powers comput- ers). This kind of structural information gives chemists a chance to modify lead molecules or compounds in a more rational way. This refinement process is called lead optimization, which may produce a drug candidate that has promising biological and chemical properties for the treatment of a disease. Once a candidate drug (or group of candidates) is developed and its effectiveness in altering the molecular target is verified, then animal stud- ies begin to determine whether the drug can be absorbed through the gastrointestinal tract for oral delivery, whether adequate levels of the drug are achieved in the blood, how the drug is metabolized in the body and excreted, and whether it actually reaches the molecular target defined by the basic research. In addition, if an animal model of the rare disease exists (through genetic alterations), this provides researchers with an opportunity to gather a preclinical proof of therapeutic concept, which can be very im- portant before the compound enters development. This process of drug dis- covery for rare diseases is no different than that for common diseases—the costs and infrastructure required for both are significant. Methodological Approaches to Biologics Discovery For a biologic product (e.g., a specific protein, enzyme, peptide, anti- body, or vaccine), the discovery phase varies considerably from the process for a small-molecule drug described above. It requires different areas of expertise, some of which can be found at academic institutions and others of which are available at biotechnology and pharmaceutical companies.

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT If the defect in a specific rare disease is due to deficiency of a specific protein, then human protein replacement therapy may be a feasible ap- proach. To accomplish this, the replacement protein can either be isolated from other animals or, more commonly, be expressed in microorganisms or plant, nonhuman mammalian, or human cells after introduction of a gene encoding the desired human protein (so-called recombinant expression). This process can be extremely complicated. Some proteins require specific modifications (called posttranslational modifications) that are only accom- plished by specific organisms or cell types. Other proteins require artificial modifications to target them to a specific tissue or cell type or to facilitate their uptake into cells, if that is where their critical function resides. For example, for some lysosomal enzyme deficiency diseases, it is critical to tar- get the replacement protein for uptake in liver or muscle cells, whereas for other diseases, the replacement protein must have different modifications that promote uptake by reticuloendothelial cells (Grabowski and Hopkin. 2003). Not all obstacles have found solutions. Currently, a sizable number of rare diseases that affect the brain present a major challenge since many biologics lack the ability to cross from the circulation into the central nervous system (the so-called blood-brain barrier). Researchers continue to investigate strategies for overcoming this problem (see, e.g., LeBowitz, 2005; Valeo, 2010). Given this complexity, there is no single path to success for biologic therapeutics. Rather, the opportunities and obstacles must be elucidated for each disease, and the approach must be tailored accordingly—a truly daunting task for thousands of rare disorders. Nevertheless, biologics have strong appeal because they have the potential to address the etiologic foundation of a disease process (e.g., through replacement of a deficient protein), to prevent diseases (e.g., with vaccines), or to harness the power of the immune system to achieve target specificity and to diversify the out- put of potential therapeutic agents (e.g., by production of an antibody that neutralizes a deleterious protein). Good examples include clotting factor proteins to treat hemophilia, vaccines to prevent smallpox or measles, and antibodies to treat multiple forms of cancer (Reichert et al., 2005). Restoration of functional levels of missing molecules includes enzyme replacement therapy, available for several lysosomal storage diseases. Among these are Gaucher disease, Fabry disease, mucopolysaccharidosis I and VI, and Pompe disease (Lim-Melia and Kronn, 2009). Enzyme therapy is also employed for one form of severe combined immunodeficiency, adenosine deaminase deficiency (Aiuti et al., 2009). These approaches have required research efforts to express the protein yeast, bacteria, plant, or mamma- lian cell systems at small laboratory scale to provide sufficient enzyme for research studies. Enzyme therapy does not correct central nervous system

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS by an NIH Clinical and Translational Science Award) and seeks to create opportunities for medical and graduate students to build and execute a product development plan for new chemical entities and other discoveries generated by Stanford faculty that are not far enough advanced to attract industry interest. The program provides faculty and industry-experienced mentors for competitively selected projects. One emphasis is product de- velopment targeted to rare and neglected diseases, including through the repurposing of old drugs or the reconsideration of abandoned ideas or proj- ects. Several products have been licensed and are in clinical development, which suggests that the goals of this program—education, stimulation of applied research, and commercialization of intellectual property—are being achieved. This is one example of an innovation platform that could acceler- ate future orphan products development. Among particular needs for clinical-translational investigators in rare diseases is training in trial designs that can be applied to studies of small populations of patients with rare diseases. These investigators will also have to recognize when they need consultants to give them more expert guid- ance. Clinical subspecialists who work with both children and adults with rare diseases should be trained to collect data that will lead to standardized and detailed phenotyping and the elucidation of clinical natural histories, two potentially important contributions to research progress related to rare diseases. Training in systems biology and bioinformatics will also be key for future investigators working in rare diseases areas because these disciplines hold the potential to rapidly advance knowledge and its application to rare diseases. Beyond scientific training, successful investigators must know how to build and sustain productive collaborations and must be comfortable communicating their work to interdisciplinary audiences. Training of young investigators or retraining of experienced inves- tigators to conduct research on specific rare diseases will depend on the existence of productive and funded programs in rare disorders-specific research that can serve as training sites for both basic and clinical research. Thus, adequate funding for rare diseases research is an important first step in establishing training environments. Funding from NIH, other federal agencies, and disease-specific advocacy groups serves the dual purpose of fostering research progress and exposing investigators-in-training or young investigators to the relevant research activities. The federal government through NIH and other agencies provides training grants, which may focus on individuals or programs (an example of the latter is the T-32 grants from NIH). These grants target specialty fellows in relevant medical subspecialties and graduate students or postdoctoral graduate fellows. Some disease-specific foundations also support training and young investigator grants.

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT Targeted career development awards for young faculty are particularly important in promoting and sustaining interest in and activity related to rare diseases. Examples of such awards include the K series grants from NIH and young investigator grants from CFF (e.g., the Leroy Matthews and Harry Shwachman Awards). The Dana Foundation’s competitive grants programs in brain and immunoimaging and neuroimmunology primarily support new investigators with innovative clinical research hypotheses to develop pilot data on brain or spinal cord diseases, most of which are rare. Some of these new investigators have NIH K-08 (or K-23) mentored grants, which provide up to 75 percent of their salaries, and Dana funds support the remaining 25 percent. Both Dana and the NIH training grants support the new investigators’ salaries, and other research-related costs often are supported by the investigators’ institutions. The Burroughs Wellcome Fund offers a postdoctoral fellow-to-faculty transition grant for physician scien- tists, a model for the NIH K99-R00 awards. This approach is particularly effective at establishing early independence for fellows (Pion and Ionescu, 2003), and it could be employed more broadly for researchers in rare dis- eases areas. The committee did not locate any compilation of resources for training related to rare diseases. Thus, it was difficult to judge the current amount of training or its content as a basis for identifying specific gaps. The em- phasis here is therefore more generally on the need for training in basic and translational or clinical research areas that will be relevant to many rare diseases. INNOVATION PLATFORMS FOR TARGET AND DRUG DISCOVERY The high costs and low success rates associated with drug discovery and development, combined with the absence, in the case of rare diseases, of a large market for approved therapies, have stimulated the development of innovation platforms on a number of levels. One typical characteristic of these emerging approaches involves the sharing of the data, biological speci- mens, chemical compounds, and other resources that are needed at various stages to move from discovery to product approval and marketing. Another characteristic is the involvement of funding organizations beyond their traditional roles of supporting research projects and train- ing. Some patient-led foundations have taken on the task of “de-risking” the early stages of drug discovery through early-stage clinical trials, for example, by combining an infusion of philanthropic capital with the de- velopment of research tools and organized access to patients. For example, CFF has assembled drug discovery tools of potential interest to the scien- tific community working on the disease: an antibody distribution program,

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS primary human epithelial cells harvested from lung transplants, a purified CFTR (cystic fibrosis transmembrane conductance regulator) protein sup- ply, and validated assay services. These efforts of CFF and others are sum- marized in Chapter 5. NIH has also created new internal capacities and new partnership mechanisms for facilitating drug discovery, which are described in this section. Public-Private Partnerships and Other Coordinating Strategies Public-private partnerships have been a standard approach when the needs of the public sector converge with goals of the private sector, prompt- ing the joint provision and management of resources for targeted projects. Examples include the delivery of services or facilities in the energy, trans- portation, education, or urban development sectors. NIH defines a public- private partnership as an agreement for the agency “to work in concert with a nonfederal party or parties to advance mutual interests to improve health” (NIH, 2007, p. 2). Although gifts, clinical research contracts and other contracts, and technology transfer agreements involve relationships with a nonfederal party or parties, NIH does not consider these arrange- ments to be partnerships. Other groups may have more expansive interpre- tations of the concept. The formation of public-private partnerships involving government, industry, and nonprofit organizations has been a successful model for the infrastructure gaps in the area of neglected tropical diseases, which share with rare diseases the lack of commercial incentives for product devel- opment. For example, the multilateral Special Programme for Research and Training in Tropical Diseases (Morel, 2000; Ridley, 2003) and, more recently, the Medicines for Malaria Venture (Ridley, 2002) combine gov- ernment, philanthropic, and industry funding2 and enlist the expertise of an external scientific advisory board to select projects for support. These initiatives coordinate activities between industry and academic centers (e.g., sharing of compound libraries) to discover new molecules for the treat- ment of tropical diseases and shepherd them through the subsequent stages along the discovery-development pipeline, thereby acting as “virtual bio- tech” companies. Projects not meeting specified milestones are dropped and replaced with others, such that each organization manages a portfolio 2 The malaria venture was, for example, initially cosponsored by the World Health Organi- zation, the International Federation of Pharmaceutical Manufacturers Associations (IFPMA), the World Bank, the Dutch government, the Department for International Development in the United Kingdom, the Swiss Agency for Development and Cooperation, the Global Forum for Health Research, the Rockefeller Foundation, and the Roll Back Malaria Partnership.

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT of projects with varying degrees of risk (MMV, 2002, 2003; Nwaka and Ridley, 2003; TDR, 2008). An example of a public-private partnership in the rare diseases area is the Spinal Muscular Atrophy project (http://www.smaproject.org/). Estab- lished by the National Institute of Neurological Diseases and Stroke, the pilot project is a multisite drug discovery and development enterprise that is guided by consultants with academic, FDA, NIH, and pharmaceutical industry expertise. The project focuses on optimizing lead compounds and making them available to researchers for preclinical testing. NIH has initiated several broader programs to support drug discovery for rare diseases. The NIH Chemical Genomics Center (NCGC), which was established as part of the NIH Roadmap, focuses on novel targets as well as roughly a dozen rare and neglected diseases. As described on its web- site, it will “optimize biochemical, cellular and model organism-based as- says submitted by the biomedical research community; perform automated high-throughput screening (HTS); and perform chemistry optimization on confirmed hits to produce chemical probes for dissemination to the research community” (http://www.ncgc.nih.gov/about/mission.html). The NCGC is also building a library of approved drugs so that these compounds can be more easily screened for possible repurposing for new indications; it has undertaken screening related to certain lysosomal storage diseases among other rare conditions (Austin, 2010). The Therapeutics for Rare and Neglected Diseases (TRND) program, which was established in 2009, will collaborate with the NCGC as well as companies and nonprofit patient groups; it thus can be considered a public-private partnership (NIH, 2009b). The program aims to bring promising compounds to the point of clinical testing and adoption for further development by commercial interests. TRND, which had an initial budget of $24 million a year, is expected to ramp up to work on roughly five projects per year. Its first pilot projects involve sickle cell disease, chronic lymphocytic leukemia, Niemann-Pick Type C, hereditary inclusion body myopathy, and the parasitic diseases schistosomiasis and hookworm (Marcus, 2010b). (The NIH Rapid Access to Interventional Development program, which takes projects through preclinical development, is dis- cussed in Chapter 5.) TRND is a much-needed and innovative development. However, its scope (five projects per year) is well below the number of rare diseases that need therapies and have researchers positioned to take advantage of this capability. In addition, extension of services to include access to animal models of rare diseases could facilitate preclinical studies, a frequent barrier to therapeutic development. Expansion of both capacity and geographical distribution of TRND activities could advance therapeutics development for more rare diseases.

OCR for page 111
0 RARE DISEASES AND ORPHAN PRODUCTS Sharing Biological Data on Disease Mechanisms As discussed earlier in this chapter, arriving at a candidate drug requires extensive basic research into the disease mechanism, identification of poten- tial targets for the drug, and generation of extensive molecular and genetic data. Typically, these mechanistic data are held by intrinsically competitive academic or industry labs that may have interests in protecting publication priorities or intellectual property or both. One consequence is that the data are not collected or stored in a way that ensures broad access to the infor- mation. This, in turn, has slowed the pace of information dissemination and driven up the cost of drug discovery. In recent years, several developments have challenged this approach, particularly in the pharmaceutical industry. One development is the growing recognition that many research questions in human diseases are too complex for any one laboratory or any one company (Duncan, 2009). Another de- velopment is accumulating research that shows that many common diseases actually consist of subsets of disease based on their molecular characteristics, which often determine how individuals will respond to therapy. As a result, what might once have been a potentially large market of patients for a par- ticular therapy is fragmented into a number of small markets. At the same time, companies have seen increasing costs of drug discovery and develop- ment without a corresponding increase in productivity of the industry, as measured by output of new molecular entities (Munos, 2009). Taken together, these trends are stimulating innovation in the form of initiatives to share data “precompetitively” (see, e.g., Stoffels, 2009; Hunter and Stephens, 2010; Marcus, 2010a; but see also Munos, 2010). A recent workshop at the Institute of Medicine explored the opportunities and chal- lenges of such collaborations, some of which involve only private entities (e.g., several companies or advocacy groups and companies) whereas others also involve the public sector (IOM, 2010b). One model of precompetitive collaboration outside the health care arena is the development of the Linux operating system, which involved competitors sharing the benefits of increased productivity resulting from joint, voluntary investments in early-stage research. Given the increasing information richness of biology, similar integration of knowledge about biochemical pathways and networks from a wide range of researchers may spur productivity in the identification of molecular targets for diseases and otherwise advancing discovery research and product development. Exist- ing examples of such efforts to share biological information or technology development resources include the following: • Enlight Biosciences, a private company created in partnership with major pharmaceutical companies to develop enabling technologies that

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT will alter the process of drug discovery and development (Zielinska, 2009); and • voluntary, open-source sharing of biological data through the Sage Commons, an initiative of Sage Bionetworks, a new, nonprofit medical research organization (http://sagebase.org). The second example, Sage Bionetworks, uses data shared by phar- maceutical companies and others to develop computational models that predict potential drug targets as well as potential toxicities (Melese et al., 2009). The data shared with Sage will eventually be publicly available and could be particularly valuable for rare diseases research. For example, the organization has already provided a significant amount of clinical data to the Huntington disease research community. The data were generated in a clinical study of Alzheimer disease in which individuals with Huntington disease were used as controls. Without the Sage resource, these data would likely have remained unknown and unavailable to Huntington disease in- vestigations (Marcus, 2010a). Some rare diseases advocacy groups have made the sharing of re- search data by grant recipients a prerequisite for funding. For example, through its Accelerated Research Collaboration model, the Myelin Re- pair Foundation has insisted that those funded in its collaborations share their research findings with one another without awaiting scientific pub- lication (MRF, 2010). To cite another example, in 2007 the Multiple Myeloma Research Consortium launched a Genomics Portal, through which researchers have unrestricted access to prepublication genomic and other molecular data (Kelley, 2009). This approach is not conventional in academic environments where researchers are rewarded for individual achievement, but by mandating data sharing, the consortium has succeeded in significantly expanding the therapies currently under development for multiple myeloma. Sharing Compound Libraries A related trend is the opening of what were once proprietary company libraries of chemical compounds to investigators interested in the potential of compounds to interact with drug targets across a wide range of diseases. Compounds and information on their structures have typically been gener- ated and held tightly by pharmaceutical companies. In recent years, com- panies have begun sharing compound libraries with researchers working in neglected diseases areas. For example, • Eli Lilly Co. is sharing its compound libraries with researchers seeking therapies for tuberculosis (http://www.tbdrugdiscovery.org/);

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS • Pfizer Inc. has signed an agreement with the Drugs for Neglected Diseases initiative to share its library of novel chemical entities so that investigators can screen it for potential treatments for human African trypanosomiasis, visceral leishmaniasis, and Chagas disease (DNDi, 2009); and • GlaxoSmithKline will share the chemical structures of compounds with potential activity against malaria through websites supported by fed- eral, for-profit, and foundation funding (Guth, 2010). In addition to the compound sharing initiative noted above, Eli Lilly has also established the Phenotypic Drug Discovery Initiative (https://pd2. lilly.com/pd2Web/). The company provides access to a phenotypic assay panel at no cost to external investigators, who can make a confidential compound submission and receive a full data report in return. Promising findings can lead to a collaboration agreement. In the area of neglected diseases, access to compound libraries and chemical structures can significantly lower the threshold for pursuing drug discovery and development. Similarly, the European Rare Diseases Thera- peutic Initiative has worked to bring about such access for academic institu- tions pursuing treatments for rare diseases (Fischer et al., 2005). To address intellectual property concerns, it has been proposed that compounds with commercial value might be accessed using a trusted intermediary, with initial confidentiality about the compound maintained and companies’ reserving the option of first refusal for development (Rai et al., 2008). The experience with these efforts might inform the development of institutional mechanisms to facilitate access to proprietary compound libraries. RECOMMENDATIONS Two critical issues for rare diseases research are the small number of patients available to participate in research on rare diseases and the limited sources of funding for discovery and development of potential therapies for these diseases. It is therefore particularly important to make the best use possible of the information and other products that research generates—whether the research is directed specifically at a rare condition or at a more common condition that potentially has relevance for a rare condition. Making the best use of information and resources has several dimensions that target problems created by current practices. These prob- lems include • institutional and individual interests—economic, reputational, and professional—that can impede collaboration and resource sharing even as they may also stimulate innovation;

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT • fragmented, proprietary patient registries that have developed in the absence of consistent standards for the creation of accurate, usable information; • fragmented, poorly preserved, and inaccessible biospecimen collec- tions; and • other resources such as biological data and research findings that are not broadly accessible to researchers who may then have to collect that information anew. The committee does not underestimate the diverse barriers to resource sharing and collaboration or the need for creativity and patience in dealing with them. Nonetheless, it believes that the initiatives cited above illustrate promising strategies for either overcoming or coexisting with these barriers. As components of an integrated policy to accelerate rare diseases re- search, several steps can be taken to develop a system that will support the sharing of resources, for example, compound libraries, and discourage the creation of a duplicative infrastructure. In some instances, steps may include the required sharing of research resources, for example, tissue speci- mens and data generated by federally funded or foundation-funded research on rare diseases. What is envisioned is essentially a “research commons” and public-private partnership (or series of partnerships) that has several unlinked or loosely linked elements. RECOMMENDATION 4-1: NIH should initiate a collaborative effort involving government, industry, academia, and voluntary organizations to develop a comprehensive system of shared resources for discovery research on rare diseases and to facilitate communication and coopera- tion for such research. Creating such a system of shared resources for rare diseases research will require a significant developmental effort and commitment of public, commercial, and nonprofit funding and other resources, for example, as- sistance in creating mechanisms for coordination and oversight and model provisions for public access to the information developed with government and nonprofit grant support. Key elements of this system would include, among other possible features, • a repository of publicly available animal models for rare disor- ders that reflect the disease mechanisms and phenotypic diversity seen in humans; • a publicly accessible database that includes mechanistic biological data on rare diseases generated by investigators funded by NIH, private foundations, and industry;

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS • common platforms for patient registries and biorepositories (see Chapter 5); • model arrangements and agreements (e.g., template language on intellectual property) for making relevant portions of compound libraries available to researchers in rare diseases areas; and • further exploration of precompetitive models and opportunities for developing technologies and tools for discovery research involving rare diseases. Given the challenges outlined in this chapter and other parts of this report and given the important role that NIH plays in supporting research on rare diseases, the committee believes that a comprehensive NIH action plan on rare diseases would be useful to better integrate and expand exist- ing work. This plan would take into account developments since the 2000 report of a special panel on coordinating rare diseases research programs within NIH (NIH, 2000). The following recommendation spans all phases of research on rare diseases and orphan products. Thus, it supports not only the discovery research discussed in this chapter but also the product development work and recommendations discussed in Chapter 5. It would likewise encompass research and development involving medical devices for people with rare diseases. RECOMMENDATION 4-2: NIH should develop a comprehensive ac- tion plan for rare diseases research that covers all institutes and centers and that also defines and integrates goals and strategies across units. This plan should cover research program planning, grant review, train- ing, and coordination of all phases of research on rare diseases. The development of an action plan would, at various points, necessarily involve consultation with FDA, advocacy groups, and industry. It likewise would involve consultation with investigators and academic institutions engaged in rare diseases research and product development. The aspects of the plan that involve training would include incentives to attract new and established academic investigators to the study of rare diseases and orphan products and also support investigators currently studying rare diseases. Such a plan could include a loan repayment program for investigators working on rare diseases, the creation of an award for highly innovative proposals for rare diseases, and the broader use of the K99-R00 (Pathway to Independence) awards to attract outstanding new investigators in rare diseases research. Training opportunities through the NIH intramural research programs could also be identified. In addition, the program could include a mechanism for identifying training opportunities (especially in computational science, small clinical trial design, and orphan

OCR for page 111
 DISCOVERY RESEARCH FOR DEVELOPMENT products development) that are particularly useful for investigators of rare diseases. Likewise, the program could support the identification, develop- ment, and replication of successful training models for investigators in rare diseases. For all investigators, the creation of an award similar to the NIH Director’s Pioneer Awards could provide an incentive and a reward for in- novation. It would also draw attention to the opportunities for rare diseases research. These awards are intended to support investigators of outstanding creativity who propose truly innovative and even transforming biomedical research. With respect to the review of proposals for research on rare diseases, the NIH action plan would include the development of guidance for study sections and institute councils. This guidance would, for example, clarify the potential public health relevance of rare diseases research, the range of appropriate methods for studying rare diseases, and the use of alternative mechanisms to ensure expert review of grant applications on rare diseases. Such mechanisms could include appointing special experts on rare dis- eases as primary reviewers to existing study sections, including rare dis- eases experts in the Center for Scientific Review, or creating a study section dedicated to rare diseases grants. More generally, NIH could investigate means of accelerating its decisions about preclinical (and clinical) awards for research on rare diseases. Further, as discussed in Chapter 3, NIH and FDA should continue to cooperate in developing training and guidance to improve the quality of NIH-funded rare diseases and orphan products research and increase the likelihood that the research—including preclinical studies—will provide ac- ceptable evidence for FDA review of marketing applications for drugs and biologics. For example, one element of the action plan could be a focused Request for Applications for natural history studies of rare diseases to help identify therapeutic targets for rare diseases or build the evidence base to support FDA approval of a specific drug being studied with NIH support. The lack of natural history studies has been identified as a problem in Chap- ter 3. Such studies are one focus of the Rare Diseases Clinical Research Network, but this network (as described in Chapter 5 and Appendix E) supports only 19 consortia that study approximately 165 rare conditions. NIH also funds a number of studies outside the network, but the natural history of many more rare diseases remains to be studied. Another element of the action plan would be the development of a systematic, reliable, and comprehensive system for identifying and track- ing public and private funding for rare diseases studies to help highlight gaps and opportunities for public and private research sponsors. As more private foundations and research initiatives are created, the lack of inte- grated information on funding will become a more serious problem and

OCR for page 111
 RARE DISEASES AND ORPHAN PRODUCTS will interfere with the ability of these groups to target their resources and collaborate effectively. The following chapter shifts the focus from basic research to the pre- clinical and clinical development investigations that are required to establish safety and efficacy and otherwise meet regulatory standards for approval of pharmaceuticals and biologics. It concludes with additional recommenda- tions for resource sharing and collaboration.