Stuart H. Orkin, M.D. Arno G. Motulsky, M.D. Co-chairs December 7, 1995
The Panel finds that:
1. Somatic gene therapy is a logical and natural progression in the application of fundamental biomedical science to medicine and offers extraordinary potential, in the long-term, for the management and correction of human disease, including inherited and acquired disorders, cancer, and AIDS. The concept that gene transfer might be used to treat disease is founded on the remarkable advances of the past two decades in recombinant DNA technology. The types of diseases under consideration for gene therapy are diverse; hence, many different treatment strategies are being investigated, each with its own set of scientific and clinical challenges.
2. While the expectations and the promise of gene therapy are great, clinical efficacy has not been definitively demonstrated at this time in any gene therapy protocol, despite anecdotal claims of successful therapy and the initiation of more than 100 Recombinant DNA Advisory Committee (RAC)-approved protocols.
3. Significant problems remain in all basic aspects of gene therapy. Major difficulties at the basic level include shortcomings in all current gene transfer vectors and an inadequate understanding of the biological interaction of these vectors with the host.
4. In the enthusiasm to proceed to clinical trials, basic studies of disease pathophysiology, which are likely to be critical to the eventual success of gene therapy, have not been given adequate attention. Such studies can lead to better definition of the important target cell(s) and to more effective design of the therapeutic approach. They often can be carried out in appropriate animal models. Pathophysiologic studies may also suggest alternative treatment strategies.
5. There is a clear and legitimate need for clinical studies to evaluate various aspects of gene therapy approaches. Although animal investigations are often valuable, it is not always possible to extrapolate directly from animal experiments to human studies. Indeed, in some cases, such as cystic fibrosis, cancer, and AIDS, animal models do not satisfactorily mimic the major manifestations of the corresponding human disease. Clinical studies represent not only practical implementation of basic discoveries, but also critical experiments which refine and define new questions to be addressed by non-clinical investigation.
6. Interpretation of the results of many gene therapy protocols has been hindered by a very low frequency of gene transfer, reliance on qualitative rather than quantitative assessments of gene transfer and expression, lack of suitable controls, and lack of rigorously defined biochemical or disease endpoints. The impression of the Panel is that only a minority of clinical studies, illustrated by some gene marking experiments, have been designed to yield useful basic information.
7. Overselling of the results of laboratory and clinical studies by investigators and their sponsors--be they academic, federal, or industrial--has led to the mistaken and widespread perception that gene therapy is further developed and more successful than it actually is. Such inaccurate portrayals threaten confidence in the integrity of the field and may ultimately hinder progress toward successful application of gene therapy to human disease.
Based on these findings, the Panel recommends the following:
1. In order to confront the major outstanding obstacles to successful somatic gene therapy, greater focus on basic aspects of gene transfer, and gene expression within the context of gene transfer approaches, is required. Such efforts need to be applied to improving vectors for gene delivery, enhancing and maintaining high level expression of genes transferred to somatic cells, achieving tissue-specific and regulated expression of transferred genes, and directing gene transfer to specific cell types. To stimulate innovative research, the Panel recommends the use of interdisciplinary workshops, specific program announcements in these areas, and the use of short-term, pilot grants for testing new ideas and for encouraging investigators from other areas to enter the field of gene therapy.
2. To address important biological questions and provide a basis for the discovery of alternative treatment modalities, the Panel recommends increased emphasis on research dealing with the mechanisms of disease pathogenesis, further development of animal models of disease, enhanced use of preclinical gene therapy approaches in these models, and greater study of stem cell biology in diverse organ systems.
3. Strict adherence to high standards for excellence in clinical protocols must be demanded of investigators. Gene therapy protocols need to meet the same high standards required for all forms of translational (or clinical) research, whatever the enthusiasm for this (or any other) treatment approach.
4. To enhance the overall level of research in this area, the Panel recommends that NIH support broad interdisciplinary postdoctoral training of M.D. and Ph.D. investigators at the interface of clinical and basic science. Mechanisms for physician training in this area might include use of career development awards based on a program announcement in gene therapy.
5. Investigators in the field and their supporters need to be more restrained in their public discussion of findings, publications, and immediate prospects for the successful implementation of gene therapy approaches. The Panel recommends a concerted effort on the part of scientists, clinicians, science writers, research advocates, research institutions, industry, and the press to inform the public about not only the extraordinary promise of gene therapy, but also its current limitations.
6. NIH has already provided an appropriate initial investment in gene therapy. Future gene therapy research should compete with other forms of biomedical research for funding under stringent peer review. Only with fair, yet critical, peer review will high standards be met and maintained. The Panel specifically does not recommend special gene therapy study sections, expansion of existing center programs in gene therapy, or expansion of the recently funded core vector production program. To ensure that the level of support remains appropriate, the NIH investment in this field should be reexamined periodically.
7. To enhance the contribution of industry to the field, the Panel recommends that NIH encourage collaborative arrangements between academic institutions and industry that complement NIH-supported research, and also implement mechanisms that facilitate the distribution and testing of vectors and adjunct materials for use in clinical studies.
8. In an effort to improve gene therapy research and reduce duplication of effort, the Panel urges better coordination and scientific review of such research throughout the NIH Intramural Program. In addition, NIH Institute Directors should resist pressures to include gene therapy research in their portfolios (either Intramural or Extramural) to "round out" their programs or compete with other Institutes. Instead, they should include such research only when there are compelling scientific reasons to go forward. Institute Directors should take the lead, where it seems appropriate, to focus efforts on improvement of diagnosis and understanding of disease pathogenesis and await further developments in vector technology before expanding clinical gene therapy programs.
Since genetic material is the putative therapeutic agent, some observers view gene therapy as qualitatively different from other forms of treatment. Seen from a broader perspective, however, somatic gene therapy reflects a natural progression in the application of biomedical science to medicine. In altering the genetic material of somatic cells, gene therapy may correct the underlying specific disease pathophysiology. In some instances, it may offer the potential of a onetime cure for devastating, inherited disorders. In principle, gene therapy should be applicable to many diseases for which current therapeutic approaches are ineffective or where the prospects for effective treatment appear exceedingly low.
Five years have elapsed since the first patients received gene modified cells at the NIH. Since then, the field of gene therapy has attracted increased attention in scientific, medical, media, and lay circles. As of June 1995, 106 clinical protocols involving gene transfer were approved by the NIH Recombinant Advisory Committee (RAC). Indeed, a total of 597 subjects have already undergone gene transfer experiments. Currently, NIH provides approximately $200,000,000 per year for research related to gene therapy. Industrial support of gene therapy research has grown steadily, such that it now is estimated to exceed that of the NIH and underwrites a major proportion of approved clinical protocols. With this high level of current activity the young field of gene therapy is the focus of attention and scrutiny as a frontier of modern medicine.
To advise Dr. Harold Varmus, Director of the NIH, the Panel to Assess the NIH Investment in Research on Gene Therapy (see Appendix A) heard presentations from NIH Institute Directors, basic researchers, and clinical investigators from academic and federal institutions and from the private sector (see Appendix B). The Panel also reviewed recent basic and clinical research in gene therapy.
Panel members are unanimous in recognizing the extraordinary potential, in the longterm, of gene therapy for managing and correcting human disease. Integrating efficacious and workable gene therapy procedures into the health care system would signal a major development in medicine, comparable to past milestones, such as the introduction of aseptic techniques, antibiotics, vaccines, and tissue transplantation. The realization of this long-term goal requires proper development of its scientific underpinnings and validation of its utility to patients with carefully designed, controlled, and evaluated clinical trials.
Although expectations have been great-fueled by the escalating enthusiasm of some investigators, industrial sponsors, and members of the media-it must be recognized that clinical efficacy in human patients has yet to be clearly established for any gene therapy protocol. This sobering reality highlights the challenge of bringing this, or any other, complex technology to clinical practice. Typically, many years are required before new therapies are proved successful. For example, transplantation of bone marrow and other organs--now an accepted therapy for lifethreatening diseases-required more than two decades of development during which frequent failures often provoked widespread skepticism. At this early stage in the development of gene therapy, the Panel considered the following issues:
The types of diseases under consideration for somatic gene therapy are diverse, and have many different underlying causes. Accordingly, the rationales and strategies for treating particular diseases are varied. To assess gene therapy's prospects and status, we, therefore, distinguish among major disease categories.
Although "gene addition" is the simplest strategy for somatic gene therapy, several practical difficulties need to be addressed. Particularly important among these is the need in many instances to deliver the appropriate gene to a specific cell type or tissue. Other challenges includes gaining access to the relevant cell type for correction, assessing the total fraction of cells in a tissue that need to be corrected, achieving the level of expression required for correction, and regulating expression of the added gene once it is transferred into appropriate target cells.
The possibilities for gene transfer as a treatment for common multifactorial diseases are vast. The precise approach needs to be assessed in each instance by considering how specific gene products influence cellular physiology. We can expect many different, sometimes speculative, strategies to be proposed. Each will need to be judged in comparison with conventional treatment approaches.
The vast majority of mutations that contribute to cancer are somatic, i.e., present only in the neoplastic cells of the patient. The introduction into cancer cells of a gene that might alter or inhibit the malignant phenotype is an appealing concept. It is based, in part, on experimental data showing that introduction of normal copies of tumor suppressor genes (e.g., p53 or Rb) into cancer cell lines in vitro restores normal growth properties.
Daunting hurdles must be overcome if gene correction strategies are to achieve a meaningful clinical outcome. First, some cancers arise following mutations in which the gene product has a dominant effect. Hence, transfer of a normal copy of the gene into an affected cell would have little, if any, impact. Second, the number of cells within a clinically detectable cancer is large (>10^9), and the mutation rate within them is so high that mutations in the introduced gene will arise in at least a subset of cells, inactivating its function and resulting in subsequent reoutgrowth of cancer cells. Third, present technologies allow gene transfer to only a subset of cells within a detectable, local tumor mass. Finally, the major dreaded complication of advanced local cancer is distant metastasis, and current means for transferring DNA do not provide feasible strategies for reaching cells that have spread widely in the body.
Because of these formidable problems, other-more indirect-gene therapy approaches to the treatment of cancer are being considered. Included among these are transfer of genes for cytokines or other immunemodulatory products to cancer cells either outside the body (ex vivo) or directly into the patient (in vivo) in an attempt to stimulate immune recognition of not only the genemodified cancer cells, but also cancer cells that have not received the gene situated elsewhere in the body. In some instances, tumorinfiltrating lymphocytes or other immune effector cells have also been transduced in an attempt to increase their specificity and/or reactivity against tumor cells. Although several of these strategies show promise in mouse models, none has demonstrated efficacy in humans.
A second general approach to the treatment of localized cancers, including brain and liver tumors, involves in vivo delivery to cancer cells of genes encoding viral or bacterial enzymes involved in the conversion of nontoxic prodrugs to their active molecules. In one approach the thymidine kinase gene from herpes simplex virus into cells is transferred into cells, rendering them more susceptible to the drug ganciclovir. Finally, genes that provide enhanced resistance to conventional chemotherapeutic agents are being transferred into bone marrow cells, which are then used to reconstitute the bone marrow of patients before treatment with intensive, and otherwise lethal, chemotherapeutic regimens.
In vaccination trials, modified HIV genes are introduced directly into infected individuals following ex vivo treatment of target CD4 or precursor cells, typically with retroviral vectors that express genes encoding antiviral products. Several such products are being tested: mutant proteins that inhibit virus replication; antisense RNA that blocks translation of HIV gene products or causes destruction of the HIV genome; ribozymes that attack HIV RNA at specific unique sites; "decoy" RNAs that efficiently compete for binding of viral proteins; and singlechain antibodies that prevent key HIV enzymes from functioning. Although these approaches block HIV replication in cell culture systems, serious obstacles to their practical application remain. Most importantly, it is not yet known what cell types to target, much less how they will be isolated, treated, and returned to the patient. Furthermore, it is unknown whether resistant mutants-the major obstacle to successful drug therapy-will also present a serious problem. Nevertheless, the pursuit of gene therapy remains an active area of acquired immunodeficiency syndrome research, and one that also promises to provide important insights into HIV pathogenesis.
Several different systems are in use or under consideration for somatic gene transfer (see Table 1). These include DNA (either naked or complexed), RNA viruses (retroviruses), and DNA viruses (adenovirus, adenoassociated virus [AAV], herpesvirus, and poxvirus). Experience is more extensive with retroviral vectors than with other viruses or nonviral DNA. Each vector system has perceived advantages and disadvantages which influence their selection for current or projected clinical applications (see Table 1). Unfortunately, none of the available vector systems is entirely satisfactory, and many of the perceived advantages of vector systems have not been experimentally validated. Until progress is made in these areas, slow and erratic success in applying gene transfer methods to patients can be expected.
The basic biology of retroviruses is the best understood of the vector systems used for gene transfer experiments. Accordingly, retroviruses are employed in the majority of clinical protocols (see Table 2). Among their advantages are efficient entry into dividing cells and integration of the transferred genetic material into the host genome without concomitant introduction of viral genes. Retroviruses would appear to be most suitable for permanent correction of genetic diseases. A major disadvantage of retroviruses is that they infect and integrate only dividing cells. Other problems include cumbersome preparation and relatively low titer, size constraints on inserted genes, difficulties in controlling or ensuring expression, and the potential for genetic damage due to random integration in the host genome.
The adenovirus vector system has found advocates more recently. Among its advantages are high titers and levels of expression, relative ease of handling, efficient infection of many types of human cells, and capacity to infect nondividing cells. Major disadvantages include its relatively high immunogenicity and the complexity of its genome. Despite the widespread belief that adenovirus does not integrate into the host genome, experimental evidence for this assertion is lacking. The persistence and expression of adenoviruses in vivo in somatic gene therapy situations are under investigation in several laboratories.
Experience with other DNA viral systems is less extensive. A major perceived strength of AAV is integration at a specific site in the infected cell genome, a finding confirmed thus far only for the wildtype virus. Research with AAV and herpesvirus has been impeded by the lack of suitable helper cell lines for preparing large amounts of pure, recombinant virus. Poxviruses appear most suitable for vaccination.
Direct administration of DNA or DNA complexes (e.g., liposomes) in vivo is in its infancy. The ease of preparation and virtually unlimited size of constructs for gene delivery make this approach attractive. The lower efficiency of gene transfer (compared with viruses) and the absence of mechanisms for specifically maintaining the introduced DNA within the cell are major disadvantages. However, the use of naked DNA for in vivo vaccination appears feasible and highly promising.
Rather than delivering a particular gene to all cells ex vivo or to a specific tissue in vivo, it appears preferable to target gene transfer to a particular cell type. In principle, this might be accomplished by incorporating ligands for cell surface receptors into viral envelopes or DNA complexes. However, such strategies have not yet reached clinical application.
Of the vector systems studied to date, retroviruses appear to be most suited for delivering genes to host cells in a stable form due to the efficient integration of retrovirally transduced genes. Studies of yeast cells have defined many of the components necessary for maintaining chromosomes within cells. In principle, the development of artificial human chromosomes as vectors might allow for maintenance of transferred genes without the problems resulting from random insertion of foreign sequences into the host genome. Several laboratories are trying to design such vectors. The efficient introduction of these vectors into cells, however, is likely to be a formidable obstacle to their use for gene therapy in the foreseeable future.
These uncertainties point to the relative dearth of wellcontrolled studies of appropriate and sustained gene expression following somatic gene transfer into animals. In many of the published reports in this field, gene expression was monitored by highly sensitive surrogate methods (e.g., cellular resistance to the drug G418 or reverse-transcriptase PCR assay), rather than by direct measurement of the desired protein product by immunologic or enzymatic activity. This practice reflects the generally low absolute level of gene expression achieved in many instances, leading to a reliance on nonquantitative analyses.
How have some of these problems of gene transfer and expression been reflected in gene therapy experiments involving animals and human subjects? Studies of retrovirusbased gene transfer into hematopoietic stem cells provide one perspective. In mice, current protocols permit transfer of genes into a substantial fraction of stem cells following retroviral infection of marrow cells ex vivo. Nevertheless, gene transfer into marrow stem cells of other species (including humans, other primates, and canines) has been much less efficient, with 10% or fewer cells transduced. In clinical protocols to date, the low efficiency of gene transfer is particularly notable. This inefficiency reduces potential benefits of introducing a particular foreign gene, and interferes with efforts to measure expression in vivo. Hence, both the clinical benefit and scientific value of clinical trials are compromised.
Current data are largely inadequate with respect to experimental study of the expression of transferred genes. In mouse experiments, longterm expression of transferred genes has been reported, but the consistency of achieving such results is unknown. Also, the quantitation of levels of gene expression over time has not received adequate attention. In human trials, the extent of gene expression is uncertain. In many instances, the efficiency of gene transfer is so poor that investigators have relied on highly sensitive molecular methods (such as reverse transcriptase PCR) rather than biologically more meaningful protein assays, to evaluate expression in vivo.
A basic understanding of the pathophysiology of disease is therefore highly relevant when designing gene therapy strategies. Besides understanding how a mutation leads to disease, it is important to determine which cells of the body are suitable targets for effective therapy. Disorders resulting from the deficiency of a circulating protein (e.g., clotting factors VIII or IX in hemophilia) might be corrected by expression of the relevant gene in skin or muscle cells, even if the protein is normally made in liver, as long as it is secreted into the bloodstream. In many other situations, expression of a transferred gene is required in a particular tissue. For example, correction of primary hemoglobinopathies, such as sickle cell anemia and Cooley's anemia, necessitates precisely regulated expression of globin chains in developing red blood cell precursors. For cystic fibrosis, which is due to loss or malfunction of a membrane protein (CFTR), it is relevant to ascertain which, and how many, cells of the lung need to express a normal CFTR gene.
Study of disease pathogenesis may sometimes lead to the development of highly effective new therapies, as illustrated by now classic research on the biochemical basis of hypercholesterolemia. Elucidation of feedback regulation of cholesterol biosynthesis led directly to the testing of HMGCoA reductase inhibitors as cholesterol lowering drugs. These agents, which are in use worldwide, have been shown to be effective in preventing cardiovascular disease. In the current climate, where the cloning of a new disease gene is often viewed principally in the context of gene therapy, the discovery of these drugs might not have been made.
Animal models for genetic diseases have arisen spontaneously in a variety of species (e.g., mouse, cat, dog). Using new methods to mutate genes in embryonic stem cells, mice with engineered alterations in any given gene can be produced. Numerous mouse strains with mutations in genes relevant to human diseases have already been created in this manner, and also by injection of human genes into fertilized mouse eggs. In some instances, mice with such mutations exhibit a phenotype similar to that seen in humans (examples: chronic granulomatous disease, hemophilia A, spinocerebellar ataxia1). In others, the effects of specific mutations in the mouse appear more severe than in humans (examples: ADA deficiency, Gaucher's disease).
Unfortunately, however, mouse models often do not faithfully mimic the relevant human conditions. For example, hypoxanthine phosphoribosyltransferase deficiency associated with LeschNyhan disease in humans is benign in mice due to the presence of an alternative metabolic pathway. Mice with mutations in the CFTR gene do not exhibit the pulmonary effects of cystic fibrosis seen in man, but rather suffer from severe gastrointestinal obstruction. Studying the differences between human diseases and animal model phenotypes may provide insights into disease pathogenesis that may, in turn, be exploited either by gene therapy or pharmacological approaches. Animal models for many cancers and for HIV infection have also been developed. In these instances, the relevance of animal models to human disease appears less certain than in typical singlegene disorders.
Despite potential phenotypic differences between human patients and animal models of disease, the study of animal models for the design of gene therapy approaches in a preclinical setting is important and should not be undervalued. As additional genes leading to human diseases are isolated, and gene targeting and transgenic technologies generate more mouse models of various human diseases, we should anticipate an increasingly productive use of such models to elucidate disease pathophysiology, possibly leading to gene therapy approaches.
Confidence in current approaches to somatic gene therapy would rise if a genuine genetic deficiency in an animal were unequivocally corrected. Although genetic defects in animals have been corrected by introducing transgenes into the germline (or by interbreeding with transgenic animals), somatic gene transfer has not permanently corrected a genetic disease in an animal (e.g., a mouse model of a singlegene disorder).
It is unlikely that a single vector will prove optimal for all gene therapy approaches. We, therefore, urge the NIH to support wideranging research in vector development and allied areas. An understanding of the behavior of vectors and the fate of DNA introduced into somatic cells will require basic efforts in virology, cell biology, immunology, and the chemistry of DNA complexes. These efforts should also include novel approaches to the selective inhibition of gene function including, but not limited to, the continued development of antisense and ribozyme strategies.
2. To facilitate interdisciplinary efforts to develop optimal vectors, the NIH should consider several strategies, including workshops and program announcements, to stimulate discovery, interchange, and collaboration among scientists in diverse areas.
3. The Panel finds that very little research effort is focused on understanding the mechanisms that govern maintenance or shutoff of gene expression following gene delivery in gene therapy experiments. Available data are largely anecdotal. We urge the NIH to give high priority to basic research to elucidate how recipient cells, and particularly stem cells, handle and express foreign DNA sequences.
4. The Panel urges expanded NIH research into the biology of stem cells in diverse organ systems, as such cells are particularly favorable recipients for permanent correction of monogenic disorders. Specific topics include identifying and enriching stem cells from various organs, targeted transfer into and expression of genes in stem cells, the discovery of growth factors required by stem cells, and methods for selectively modifying genes in stem cells.
5. In the enthusiasm to begin human gene therapy trials soon after gene discovery, important aspects of disease pathophysiology, cell biology, and biochemistry have often been underemphasized. Better elucidation of these aspects will reveal the nature of the target cells within a tissue that need to receive the transferred gene, potential difficulties in achieving gene transfer into the appropriate cells or tissue, and features of the relevant protein that may be critical for its function in vivo. This increased focus on basic mechanisms of pathophysiology should also foster alternative efforts to develop pharmacological approaches to disease management. We recommend that the NIH vigorously support basic research into molecular mechanisms that produce disease. The present enthusiasm for molecular approaches to therapy, no matter how justified, must not lead to neglect of biochemical and pathophysiologic mechanisms at the tissue and organ level, which may lead to novel therapeutic insights.
6. We recommend that NIH provide continued and expanded support for the development and study of those animal models of disease that faithfully reflect the corresponding human disorders. These models should strengthen the preclinical scientific basis for gene therapy protocols. This approach will often be more costeffective than attempting to perform similar studies in humans.
Although widely referred to as "clinical trials," gene transfer protocols to date are in truth smallscale clinical experiments. Such exploratory studies are meant to test the feasibility and safety of administering particular vectors and to evaluate the effects of expressing specific gene products. Because these studies have not been designed to measure efficacy, they do not include sufficient controls to evaluate the true merits of gene therapy or compare this approach with conventional approaches to the same disease.
Only a few of these clinical studies are designed well enough to address fundamental biological questions. Most notable are several elegant gene marking studies investigating the cellular origin of tumor recurrence and other recent studies comparing the relative survival of cells of HIV-patients simultaneously infected with different retroviruses meant to inhibit HIV replication. These well designed studies greatly increase the information that may be extracted from careful clinical experiments involving only a few patients.
Upon reviewing the status of clinical protocols approved for gene transfer the Panel made several observations:
In view of these and other difficulties, the Panel considered the appropriateness of clinical studies of gene therapy at this time. The consensus view of the Panel is that clinical studies are warranted for several important reasons-precisely those that distinguish basic and clinical investigation:
Many of the issues faced in bringing gene therapy to clinical practice are encountered when any recent discoveries are applied to the management of disease. The success of such endeavors (often termed "translational research") relies on the quality of the underlying science, the care with which clinical protocols are designed, the melding of different disciplines and strategies into a cohesive approach, and the capacity of investigators to bridge science and medicine. Research at the interface of frontier science and patient care is challenging, and requires that investigators have broad training and biological perspective. For this and other fields of clinical investigation to succeed, high standards of experimental design and robust methods for evaluating clinical outcomes are needed. In the Panel's judgment, many clinical gene therapy studies thus far have not met these standards.
2. The Panel endorses efforts to develop broad, interdisciplinary training programs in clinical (or translational) research (see below). Training of clinical investigators with broad experiences in biomedical and clinical activities, including biostatistics, will benefit not only the immediate field of gene therapy, but also other areas of translational research.
3. The Panel urges gene therapy investigators and their sponsors--be they academic, governmental, private, or industrial-to be more circumspect regarding the aims and accomplishments of clinical protocols when discussing their work with the scientific community, the public, and the media.
We cannot predict when the clinical benefits of gene therapy will be realized. The Panel senses that the public has little understanding of the enormous challenges in the field, and may believe its day has already come, or is at least imminent. Raising such false hopes threatens public support, particularly if effective therapies for more common disorders are not quickly delivered, and may encourage patients and their families to make unwise decisions regarding their treatment options. Scientists, clinicians, scientific journalists, and the press need to devote more attention to responsible, public education regarding the current status and prospects for gene therapy.
2. The Panel recommends a concerted effort on the part of scientists, clinicians, science writers, research advocates, research institutions, and the press to inform the public regarding not only the great promise of gene therapy but also current realities. This program of education needs to stress that some time will be required to develop the science of the field and to translate these advances to clinical practice.
3. The Panel urges those who care for patients and provide advice regarding treatment and reproductive options to present the current capabilities of the gene therapy field in an honest and restrained manner. Otherwise, patients and their families may fail to utilize more conventional therapies from which they may receive substantial clinical benefit or choose reproductive options based on unrealistic expectations of curative gene therapy.
Resources currently exist at many institutions for the performance of clinical studies related to gene therapy. The NIH-supported general clinical research centers (GCRCs) represent a highly appropriate resource for the community.
2. For clinical studies, the Panel urges that investigators make efficient use of NIHsupported GCRCs. These centers have been established to promote research at the interface of clinical and laboratory sciences and are well suited for use in human clinical trials.
NIH has already provided the field of gene therapy with an appropriate start by support of gene therapy centers and specific requests for applications (RFAs). The Panel believes that the current level of research support for this area of biomedicine is appropriate at this time, and suggests that funds for future efforts be allocated on the basis of traditional peer review to ensure that current problems in the field are addressed critically. The adequacy of funding for clinical protocols, particularly outside the NIH campus, has been difficult to assess, since a substantial proportion of support is currently provided by industry. We see no indication that clinical applications in the field of gene therapy are being held back by inadequate financial support.
2. To guarantee sound review of gene therapy proposals, particularly those which include clinical studies, the Panel urges that membership of NIH study sections be broadened so that they are better able to review both basic and applied aspects of projects. This view is in agreement with the recommendations of the committee chaired by Keith Yamamoto that recently evaluated the peer review system at the NIH.
3. The Panel recommends that gene therapy research compete directly with all other forms of therapeutic research for funding. Because different approaches may lead to successful treatment of disease, it would be unwise to focus only on one approach, such as gene therapy, for special support.
4. The Panel opposes the formation of study sections dedicated to the review of proposals in the area of gene therapy. If high standards are to be met, research in this area needs to compete with that in other fields of biomedical science.
5. The field of gene therapy should be reviewed periodically to assess whether the investment by NIH should be increased or decreased.
6. To stimulate truly innovative research, the Panel recommends that several Institutes of NIH pool funds through the R21 grant program for short pilot projects focused in specific areas, including vector design and expression of transduced genes, animal models of disease, and stem cell biology.
7. Although it did not formally evaluate the role of RAC, in evaluation of clinical protocols, the Panel recognizes the need for continued review of the safety of gene therapy by expert scientists.
Industry has important attributes that recommend its active participation in gene therapy. First, industry is skilled in translational research and the development of drug products. Second, it has significant experience in meeting high manufacturing and quality control standards, and maintains a professional staff dedicated to regulatory and clinical issues. Third, a high level of scientific and technical expertise characterizes modern biotechnology and pharmaceutical companies.
Several companies have ongoing research programs developing improved vectors for gene delivery and better systems for expression of foreign genes. Moreover, industry has been the major supporter of many of the approved clinical protocols. It is axiomatic that success for biotechnology or pharmaceutical companies will be equated with the development of FDAapproved, clinically efficacious gene transfer treatments for disease. Industrial efforts will focus where the perceived use of the product is greatest, and likely to yield high profits. Hence, industry will tend to concentrate on common diseases, such as cancer, rather than rare disorders. This imbalance has not been evident thus far, as some companies are studying rare diseases initially, aiming to demonstrate proof of concept. For example, industry is supporting clinical studies of adenosine deaminase deficiency, Fanconi's anemia, and cystic fibrosis. Once clinical efficacy of gene therapy procedures is demonstrated for specific, infrequent disorders, however, it can be anticipated that market forces will drive industry's involvement toward common diseases for which patient populations are large.
Industry is collaborating with academic institutions across a wide spectrum. On the whole, this involvement is healthy and complements NIHsupported research. For example, industrial partners have prepared GMPgrade vectors for many clinical studies at academic institutions. The development of gene therapy as a clinical activity is threatened, however, by potential conflicts among the demands of good science and the goals of academic researchers, clinicians, industry, and its investors. The field is at risk to the extent that the premature initiation of clinical studies and overzealous, uncritical reports of clinical results are used by industry to promote investment and perceived research dominance. Likewise, if the objectivity and integrity of academic investigators associated with specific companies is undermined as they seek to maintain their industrial ties, the field will be jeopardized. Decisions regarding diseases to be treated need to be made by investigators on scientific rather than financial criteria. Although the problems of conflict of interest in the field of gene therapy do not differ substantially from those encountered in other forms of clinical research, the wide publicity given to clinical gene therapy efforts raises the potential stakes for both academic investigators and those at companies.
For the future development of the field it will be important that issues of proprietary control not limit the development of clinical protocols. The Panel heard several presentations that described logistical difficulties encountered in gaining industrial approval to perform clinical studies in which cytokines and other reagents were to be obtained from several, often competing companies. These obstacles would be reduced if mechanisms were developed to facilitate the dissemination of useful materials for clinical trials.
In the opinion of the Panel, it is premature to assess what impact, if any, the licensing of a broad patent to a single company for ex vivo gene therapy will have on the field. The Panel is concerned, however, that broad patents of this kind will ultimately retard implementation of successful gene therapy protocols once they are developed. Additional study of the impact of patents on the development of the field will be necessary.
2. NIH should encourage collaborative arrangements that complement NIHsupported research. Industry can play an important role in providing GMPgrade vectors for clinical testing and in designing clinical trials that meet rigorous criteria for efficacy and regulatory standards.
3. The Panel urges the NIH to develop and implement mechanisms that would facilitate the distribution and testing of adjunct materials (e.g., cytokines) for use in gene therapy.
In their presentations to the Panel, Institute Directors discussed and spoke highly of research programs in gene therapy. Of these presentations two aspects are noteworthy. First, there appears to be little coordination of research across Institute boundaries, such that duplicative efforts are inevitable. Second, much of the research utilized similar, yet inadequate, vector systems, which were tailored to deliver genes to the tissue of each Institute's interest. In these respects, Intramural research does not differ from that taking place elsewhere. Research of this kind, however, is unlikely to provide innovative advances. Institute Directors should be encouraged to support innovative research approaches, whether they be Intramural or Extramural, in whatever field of endeavor, even if this leads to deemphasis of gene therapy research within an Institute. They should resist the temptation to fill the "portfolio" with research that appears "hot" but may lack a strong scientific basis or likelihood of success relative to other areas.
The Clinical Center of the NIH campus is a superb resource for the execution of clinical investigation at all levels. With a new clinical center, currently under development, the NIH would be assured firstrate facilities well into the next century. The NIH Clinical Center and its staff have proved effective over the years in attracting and maintaining a patient base representing a wide spectrum of diseases, including many rare, inherited disorders. As such, it is an excellent resource, both for the Intramural Program and the country. The recent decline in patient occupancy in the Clinical Center is a cause for concern, which is being appropriately addressed. It is hoped that erosion of the excellent patient resource base of the NIH will not occur, so that clinical investigation in the Intramural branch will not be jeopardized.
2. The Panel urges Institute Directors to include gene therapy within their portfolios only when there are compelling scientific reasons. Accordingly, they should resist pressures to include gene therapy (or any other) research to "round out" their programs or compete with other Institutes. Institute Directors should take the lead, where it seems appropriate, to focus efforts on research in gene discovery, diagnosis or disease pathogenesis and await further developments in vector technology before expanding gene therapy programs.
3. The Panel endorses the efforts of the Director of the Clinical Center to develop strategies to maintain the superb clinical base of the NIH Intramural Program.
Grubb, B. R. et al. Inefficient gene transfer by adenovirus vector to cystic fibrosis airway epithelia of mice and human. Nature 371: 802806, 1994.
Knowles, M. R. et al. A double-blind vehicle-controlled study of adenoviral vector mediated gene transfer in the nasal epithelium of patients with cystic fibrosis. New Engl. J. Med. 333: 823831, 1995.
Adenosine deaminase deficiency:
Blaese, R. M. et al. T Lymphocyte-directed gene therapy for ADASCID: Initial trial results after 4 years. Science 270: 475480, 1995.
(For clinical histories of the ADA-deficient patients in this study, see Hershfield, M. S., Chaffee, S., and Sorensen, R. U. Enzyme replacement therapy with polyethylene glycoladenosine deaminase in adenosine deaminase deficiency: overview and case reports of three patients, including two now receiving gene therapy. Ped. Res. 33: S42S48, 1993.)
Bordignon, C. et al. Gene therapy in peripheral blood lymphocytes and bone marrow for ADA-immunodeficient patients. Science 270: 470475, 1995.
Kohn, D. B. et al. Engraftment of gene-modified umbilical cord blood cells in neonates with adenosine deaminase deficiency. Nature Med. 1: 10171023, 1995.
Hypercholesterolemia:
Grossman, M. et al. Successful ex vivo gene therapy directed to liver in a patient with familial hypercholesterolaemia. Nature Genet. 6: 337340, 1994.
(also see Brown, M. S. et al. Gene therapy for cholesterol. Nature Genet. 7: 349350, 1994.)
Grossman, M. et al. A pilot study of ex vivo gene therapy for homozygous familial hypercholesterolaemia. Nature Med. 1: 11481154, 1995.
Marker, DNA vaccination, and other studies:
Rosenberg, S. A. et al. Gene transfer into humans-immunotherapy of patients with advanced melanoma, using tumor-infiltrating modified by retroviral gene transduction. New Engl. J. Med. 323: 570578, 1990.
Brenner,M. K. et al. Gene-marking to trace origin of relapse after autologous bone-marrow transplantation. Lancet 341: 8586, 1993.
Rooney, C. M. et al. Use of gene-modified virus-specific T lymphocytes to control Epstein-Barr-virus-related lymphoproliferation. Lancet 345: 913, 1995.
Nabel, G. J. et al. Direct gene transfer with DNA-liposome complexes in melanoma: expression, biologic activity,and lack of toxicity in humans. Proc. Natl. Acad. Sci. (USA) 90: 1130711311, 1993.
Dranoff, G. et al. Vaccination with irradiated tumor cells engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates potent, specific, and longlasting anti-tumor immunity. Proc. Natl. Acad. Sci. (USA) 90: 35393543, 1993.
Ulmer, J. B. et al. Heterologous protection against influenza by injection of DNA encoding a viral protein. Science 259: 17451749, 1993.
Recent review articles:
Friedmann, T. The promise and overpromise of human gene therapy. Gene Ther. 1: 217218, 1994.
Yu, M., Poeschla, E., and Wong-Staal, F. Progress towards gene therapy for HIV infection. Gene Therapy 1: 1326, 1994.
| System | Advantages | Disadvantages | Accumulated Experience | Current or Projected Application |
|---|---|---|---|---|
| Retrovirus | Efficient entry. Efficient, predictable, and stable integration into host cells. Biology is well understood. Slight immunogenicity*. No viral genes in vector. | Low titer. Limited insert size. Infection limited to dividing cells. Expression difficult to control and stabilize. Potential for genetic damage*. Expensive, complex to prepare and validate. | Extensive | Marker studies. ex vivo treatments, particularly for AIDS and cancer. Vaccines. |
| Adenovirus | Efficient entry into most or all cell types. High titers. High level of expression. (In principle) no integration of DNA*. Can infect stationary cells. | Vectors contain many viral genes. Highly immunogenic, stimulating both B and T cell responses. Unsuitable for stem cells. Factors controlling tropism poorly understood. Generation of replication competent virus. | Moderate | Localized in vivo treatments: cystic fibrosis, muscular dystrophy, cancer. |
| Adeno-Associated Virus | Integration at specific sites*. | Requires replicating adenovirus to grow. No helper cell line. Specific integration probably does not occur in absence of viral genes. Very limited insert size. | Moderate | Similar to adenovirus. |
| Herpesvirus | High titers. Neurotropic*. | Complex construction. No packaging cell lines. | Slight | Neurologic disorders. |
| Poxviruses | High titers. Large insert size. High expression. | Highly immunogenic. Similar to adenovirus and herpesvirus. | Moderate | Localized, transient in vivo treatment. |
| Naked DNA | Easy to prepare in quantity. High level of safety*. Virtually unlimited size. No extraneous genes or proteins to induce immune response. Lack of integration*. | Very inefficient entry, uptake into nucleus. No mechanism for persistence or stability. | Moderate | Topical applications, mechanical and accessible (skin, vascular, pulmonary, endothelial cells). |
| Facilitated DNA (e.g., liposomes) | Same as DNA. More efficient uptake than DNA. Protected from in vivo Targetable to specific cell types*. | Targeting not yet achieved. No mechanism for persistence or stability. Inefficient entry. | Slight | As for naked DNA. |
* Denotes theoretical advantage or concern, but one that has not yet been adequately tested.
| System | # of Protocols | Percentage |
|---|---|---|
| Retrovirus vectors | 76 | 71.7 |
| Adenovirus | 15 | 14.2 |
| Adeno-associated viruses | 1 | 0.9 |
| Cationic liposome complex | 12 | 11.3 |
| Plasmid DNA | 2 | 1.9 |
| Category | Disease/Disorder | # of Protocols | Percentage |
|---|---|---|---|
| Inherited Monogenic Disorders |
Total ADA deficiency Alpha-1-antitrypsin Chronic granulomatous disease Cystic fibrosis Familial-hypercholesterolemia Fanconi anemia Gaucher disease Hunter syndrome |
20 1 1 1 11 1 1 3 1 |
18.9 0.9 0.9 0.9 10.4 0.9 0.9 2.8 0.9 |
| Infectious Diseases |
Total Human inmunodeficiency virus-1 |
8 8 |
7.5 7.5 |
| Acquired Disorders | Total Peripheral artery disease Rheumatoid arthritis |
2 1 1 |
1.9 0.9 0.9 |
| Cancer (byapproach) | Total Antisense Chemoprotection Immunotherapy/ex vivo Immunotherapy/in vivo Pro-drug/HSV-TK/ganciclovir Tumor suppressor gene |
51 2 4 23 7 11 4 |
49.1 1.9 3.8 21.7 6.6 10.4 3.8 |
| Marking Protocols | 25 | 23.6 | |
| All Studies | 106 | 100.0 |
Data from Debra J. Wilson, Executive Secretary, Subcommittee on Data Management, Office of Recombinant DNA Activities, NIH
Appendix A
Panel to Assess the NIH Investment in Research on Gene Therapy
Panel Members
Stuart H. Orkin, M.D. (Co-Chair) Haig H. Kazazian, Jr., M.D.
Howard Hughes Medical Institute Department of Genetics
Harvard Medical School University of Pennsylvania
Division of Hematology-Oncology School of Medicine
Children's Hospital Philadelphia, PA 19104-6145
Boston, MA 02115
Arno G. Motulsky, M.D. (Co-Chair) Thomas J. Kelly, M.D., Ph.D.
Medicine-Medical Genetics Department of Molecular Biology
University of Washington, RG-25 Johns Hopkins University
Seattle, WA 98195-0001 School of Medicine
Baltimore, MD 21205
Richard Axel, M.D. Robert J. Lefkowitz, M.D.
Howard Hughes Medical Institute Howard Hughes Medical Institute
Columbia University Duke University
New York, NY 10032 Durham, NC 27710-0001
David Botstein, Ph.D. Bernard Moss, M.D., Ph.D.
Department of Genetics National Institute of Allergy
Stanford University and Infectious Diseases
Stanford, CA 94305-5120 National Institutes of Health
Bethesda, MD 20892
John M. Coffin, Ph.D. Thomas A. Waldmann, M.D.
Department of Molecular Biology National Cancer Institute
Tufts University School of Medicine National Institutes of Health
Boston, MA 02111 Bethesda, MD 20892
Pamela B. Davis, M.D., Ph.D. Huda Y. Zoghbi, M.D.
Pulmonary Division Department of Pediatrics
Case Western Reserve University Baylor College of Medicine
School of Medicine Houston, TX 77030-3411
Cleveland, OH 44106-2624
Eric R. Fearon, M.D., Ph.D. Executive Secretary
Division of Molecular Medicine Judith H. Greenberg, Ph.D.
and Genetics National Institute of
University of Michigan General Medical Sciences
Medical Center National Institutes of Health
Ann Arbor, MI 48109-0638 Bethesda, MD 20892
Uta Francke, M.D. Program Assistant
Department of Genetics/Beckman Center Janice C. Ramsden
Howard Hughes Medical Institute Office of the Director
Stanford University Medical Center National Institutes of Health
Stanford, CA 94305-5428 Bethesda, MD 20892
Appendix B
In the aggregate, NIH invests nearly $200 million annually in programs supporting and overseeing gene therapy research. Despite enthusiastic interest and early signs of safety and biological feasibility, however, evidence for therapeutic benefit to patients is meager. Moreover, opinions vary as to what gene delivery systems will prove effective over the long term, and there are unsettled questions as to which diseases are appropriate targets for gene therapy during this phase of its development.
The mandate for the Panel to Assess the NIH Investment in Research on Gene Therapy is to review broadly the gene therapy research enterprise, considering (i) current and proposed investments by NIH centers and institutes in gene therapy and related disciplines, (ii) developments affecting gene therapy in the wider community of academic, government, and industrial laboratories, and (iii) evaluation of the NIH investment in the context of other support for gene therapy research, particularly from the U.S. biotechnology industry and also from outside the United States.
From this comprehensive review, the panel is expected to devise a set of recommendations on NIHsponsored gene therapy research-not a rigid plan-to be presented at the meeting of the Advisory Committee to the Director, NIH, in December 1995. The recommendations are expected to help in NIH budget and program planning for FY 1997 (and, to a limited extent, FY 1996) by addressing specific questions, including the following:
The National Heart, Lung, and Blood Institute (NHLBI) ($53 million); the National Cancer Institute (NCI) ($10 million); the National Institute of Allergy and Infectious Diseases (NIAID) ($16 million); and the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) support the largest efforts in gene therapy research, with seven other institutes sponsoring smaller programs. In addition, the National Center for Human Genome Research, in cooperation with researchers from several other institutes, is developing basic and clinical research projects strictly as part of its intramural program.
The NIH intramural program, from which the first several clinical protocols to be approved arose, continues to have a strong focus on gene therapy research. The wide variety of projects on the NIH campus to study disparate diseases, particularly rare disorders; specialized facilities, including stateoftheart human stem cell processing and transfer technology; an emphasis on highrisk, lab bench-to-bedside research at the clinical center; a concerted effort to reinvigorate the intramural program that features stringent staff reviews and a new tenure track system; and recently mandated incentives to encourage technology transfer from federal laboratories to the private sector are some of the reasons behind this focus. Recently, some 100 researchers in the intramural program formed a campuswide interest group.
A variety of funding mechanisms is available for supporting gene therapy efforts through the NIH extramural program. Researchers may submit investigatorinitiated grant applications, usually R01s, or prepare applications in response to RFAs, which invite investigators to submit proposals for projects in NIHspecified research areas. Typically, NIH commits funds for RFAs that it issues, and applications receive special reviews. Nonetheless, RFAs allow considerable latitude for researchers at different institutions to establish innovative arrangements and to set up collaborative networks.
In addition, there is a more formal grant mechanism for forming specialized multidisciplinary research centers at single institutions or among several institutions in a "Centers without Walls" program. Besides these grant mechanisms, the extramural program also can designate areas for competitive proposals to do contract research and development projects, usually with very specific targets. Beyond these standard funding measures, the NIH Director now has discretionary authority to transfer 1 percent of NIH funds for a particular fiscal year into research areas of special interest or need.
Additional research resources supported by the extramural program of the National Center for Research Resources (NCRR) are part of a nationwide research infrastructure that already supports some gene therapy research activities and could be tailored or expanded to support additional efforts. For example, 14 of 75 general clinical research centers, most associated with U.S. medical schools, are conducting gene transfer trials. A biotechnology resource center now at Louisiana State University maintains an extensive, everexpanding database for human genemapping studies. There are seven regional primate research centers where gene therapy animal model studies can be conducted. As part of a new resource, three Institutes (NCI, NHLBI, and NIDDK) will begin supporting in mid1995 one to three national gene vector laboratories, whose establishment is based on a $3.5 million setaside for a joint RFA.
Another important element of NIH's overall involvement in gene therapy research is the role it plays in overseeing policy matters such as the review of clinical protocols. As of May 1995, RAC has recommended approval for 105 human gene transfer protocols, including 77 involving some form of cancer, 19 involving various genetic disorders, and 8 on AIDS. Of this total, 25 are genemarking experiments without any direct therapeutic potential. RAC is now streamlining its review procedures, and full responsibility for several categories of review now resides with FDA.
The NIH Office of Technology Transfer (OTT) serves under a congressional mandate to evaluate research and technology supported by the intramural program and to take appropriate steps to ensure that such intellectual property is further developed. Thus, OTT helps in identifying patentable inventions and filing applications, coordinating the development of cooperative research and development agreements (CRADAs) and material transfer agreements with researchers in industry or at universities, and arranging licensing agreements with industrial partners that seek to develop commercial products. NIH researchers, primarily from NCI and NHLBI, have filed 81 gene therapy-related patent applications (some of them diagnostic developments and others research tools). To date, NIH has completed 22 licenses covering gene therapy-related technologies.
The invited speakers, who may include leading exponents in this field and critics, will be asked to focus generically on an assigned topic, not merely to provide a summary of an individual's particular experiences relevant to the topic. In addition to presenting a stateoftheart summary on the assigned topic, speakers will be asked to outline major problems or challenges relevant to the topic, including infrastructure and administrative matters, and to propose ways of solving some of those problems and encouraging progress in their particular subject areas. Speakers will also be asked to provide the panel with a brief summary of important points they plan to make.
In addition to making a general plan for the panel's next two meetings, panel members began to identify problems to address as they assess the NIH investment in gene therapy research. One issue that the panel will consider, which is not unique to gene therapy research, is how different NIH institutes and centers divide resources between intramural and extramural programs. On average, the intramural program budget is about 11 percent of the overall NIH budget, but there is considerable variation across specific programs and projects. Historically, the first few gene therapy clinical protocols were undertaken by researchers in the intramural program, and there is continued strong interest in pursuing such developments. Is that an appropriate strategy?
This issue is related to a more general question of how institutes and centers coordinate overlapping programs in gene therapy research both across extramural portfolios and in the intramural program. In practical terms, a question for the panel may be framed as follows: Should several institutes and centers focus on a few seemingly tractable genetic disorders, such as cystic fibrosis and Gaucher's disease, simultaneously supporting relatively comparable research approaches? Or should early efforts be directed more broadly and targeted for a much more diverse set of diseases?
Other issues that the panel may consider include the following:
List of Speakers Duane F. Alexander, M.D. John I. Gallin, M.D. Director Director National Institute of Child Health and Warren Grant Magnuson Clinical Center Human Development Wendy Baldwin, Ph.D. Robert A. Goldstein, M.D., Ph.D. Deputy Director for Extramural Research Director Office of the Director Division of Allergy, Immunology, and Transplantation National Institute of Allergy and Infectious Diseases James F. Battey, Jr., M.D. Michael Gottesman, M.D. Director, Division of Intramural Research Deputy Director for Intramural Research National Institute on Deafness and Office of the Director Other Communication Disorders Henning Birkedal-Hansen, D.D.S., Ph.D. Richard J. Hodes, M.D. Director, Division of Intramural Research Director National Institute of Dental Research National Institute on Aging Francis S. Collins, M.D., Ph.D. Claude Lenfant, M.D. Director Director National Center for Human Genome National Heart, Lung, and Blood Research Institute Karl Csaky, M.D. Michael Lockshin, M.D. Medical Officer Acting Director National Eye Institute National Institute of Arthritis and Musculoskeletal Diseases Carl Dieffenbach, Ph.D. Harry L. Malech, M.D. Acting Associate Director Deputy Chief Basic Science Program Laboratory of Host Defenses Division of AIDS National Institute of Allergy and National Institute of Allergy and Infectious Diseases Infectious Diseases Judith Fradkin, M.D. Daniel Rotrosen, M.D. Chief Chief, Host Defense & Inflammation Endocrine and Metabolic Diseases Division of Allergy Immunology & Program Branch Transplantation National Institute of Diabetes and National Institute of Allergy and Digestive and Kidney Diseases Infectious Diseases Maria Freire, Ph.D. Giovanna Spinella, M.D. Director, Office of Technology Transfer Health Scientist Administrator Office of the Director Developmental Neurology Branch Division of Convulsive, Developmental and Neuromuscular Disorders National Institute of Neurological Disorders and Stroke Judith L. Vaitukaitis, M.D. Harold Varmus, M.D. Director Director National Center for Research Resources National Institutes of Health Robert E. Wittes, M.D. Nelson A. Wivel, M.D. Acting Director Director Division of Cancer Treatment Office of Recombinant DNA Activities National Cancer Institute
Several experts believe that, eventually, these two separate vector strategies may converge as researchers try to develop synthetic or semi-synthetic vectors that incorporate the useful features of viruses and chemical agents. Meanwhile, although specific strategies to build useful vectors have strong advocates, no particular vector has emerged as a clear front runner. Each approach has its own problems, and most of them also share problems.
For example, except for AAV, these virus vectors integrate randomly, if at all, in the host cell's chromosomes. Moreover, transduction efficiencies for the virus vectors vary widely--in part reflecting their poor ability to integrate into the chromosomes of resting cells. This problem may even affect HIV, despite a widely held notion that it can infect resting cells. Nonetheless, according to Dr. Richard Mulligan, some of the more recently refined retroviral vectors efficiently transduce non-resting target cells, particularly if they carry appropriate LTR sequences and selectable marker genes or, in some cases, specific promoter-enhancer sequences.
Another general problem is that very little research has been done to incorporate externally controllable gene sequences into viral vectors. For instance, regulated beta-globin gene expression is perhaps the most widely studied prototype. However, when this gene is transduced successfully into human cells growing in tissue culture, its expression cannot yet be properly regulated. Some of these difficulties in attaining gene regulation may arise because of the randomness of integration.
In part because gene regulation questions are unanswered, determining the appropriate dosage levels for viral vectors presents another major challenge. For example, according to Dr. Alan Smith, in clinical trials involving patients with cystic fibrosis (CF), there is a concern that the vector and the CFTR gene product it carries may pose problems if they are delivered in too high doses. Because CFTR is ordinarily effective in cells when present at very low levels, low doses of the transferred gene may be required for effectiveness and may be less likely to induce host inflammatory responses.
These considerations raise a more general and potentially serious problem, namely that viral vectors may carry genes--either their own or the particular recombinant genes they are modified to carry--that elicit host immune system responses. This phenomenon might interfere with the efficacy of gene therapy procedures, possibly curtailing long-term expression of transferred genes and prohibiting repeat administration of the therapeutic agent. Other factors, such as counter selection of the transduced cell by immune or other mechanisms and the randomness of integration, may also contribute to apparent low transduction efficiencies and/or short-lived expression of transferred genes.
Dr. Smith said that cationic lipid vectors are being improved and now perform as much as 500-fold more effectively than naked DNA but are still less effective than is the AV vector in rodent model systems. A potential advantage of cationic lipids is that they can be administered repeatedly to rodents. However, at high doses they induce some focal inflammatory responses, albeit without evidence of eliciting antibodies or provoking T cell activation. Dr. Smith speculated that cationic lipids activate macrophage cells.
Additional advantages and problems associated with specific vector candidates:
In addition, host cell range tends to be narrow, although introduction of genes from other viruses such as vesicular stomatitis virus (VSV) may help in broadening that range. However, the presence of VSV genes may introduce new toxicity problems, leading to damage or killing of the host cell.
Dr. James Wilson said that other approaches to controlling the inflammatory response are being considered, including production of antibodies to block T cell activation, use of agents such as the drug cytoxan to block T cell proliferation, and use of cytokines to reduce or block production of neutralizing antibodies.
Dr. Thomas Shenk said that several AV genes influence tumor formation in animal model systems and malignant transformation of cultured cells. Thus, AV represents a potential problem when modified versions of the virus are used as vectors, even though AV has not been observed to cause human tumors. He also is studying the molecular and cellular events required for AV to recognize, bind to, and penetrate target cells, and to deliver and integrate the genes it carries to the target cell nucleus.
Yet another set of problems entails uncertainties over the target cells for gene transfer procedures. Dr. Arthur Nienhuis noted that several issues may help to account for low overall gene transfer efficiency in clinical settings. These include the phase of the cell growth cycle that a particular target stem cell may be in, the current unavailability of effective cytokines to regulate that cycle, difficulties in stimulating specific viral receptor production by the cell, and problems in improving the transduction efficiency of target cells. Stimulation with cytokines or, alternatively, the introduction of drug resistance markers and subsequent use of the corresponding drug may provide ways of expanding specific transduced target cell populations. However, Dr. Nienhuis cautioned that such approaches are still at a very early, preclinical stage of development.
Results from clinical trials so far are limited. Relatively few patients have been treated; measures of biological response are often not adequately sensitive, except in cases where host inflammatory responses have been reported; the effects observed seem to be erratic; and the reporting of effects so far has been almost entirely anecdotal, rather than in peer reviewed publications.
According to Dr. Ronald Crystal, AV-delivered CFTR genes may be expressed along airways of CF patients as many as four days after being administered; however, that expression is observed in only a low percentage of the patients treated. According to Dr. James Wilson, in other experiments involving CF patients, expression of the CFTR gene is rare, not stable, but also not toxic. Although sustained expression is attained in knock-out mice, efforts to introduce the CFTR gene in other animal model systems tend to induce immune responses directed to vector (AV) genes.
Clinical results are also variable in the few ADA patients who are partaking in gene transfer experiments, according to Dr. Michael Blaese. One youngster has been infused 11 times over 23 months with her own T cells after they were treated with a retrovirus carrying an ADA gene, and ADA+ T cells have persisted for two years following the eleventh infusion. He said there is one copy of vector per peripheral T cell, and a positive signal for circulating mRNA (earlier, that signal was "intermittent"). A complicating factor is that PEG-ADA is still being administered to the patient, albeit in a low dose that was established before she more than doubled in weight.
The results for a second child under the same treatment regime are more ambiguous but apparently less promising. However, Dr. Blaese said that three other children whose cord blood was treated at birth show persistent expression of the vector after more than 12 months following the procedure. In addition, good expression of the ADA retroviral-delivered gene was obtained in vitro from foreskin cells obtained from two of these patients, suggesting that small skin grafts using modified cells might be an effective alternative means of delivering the corrective ADA (or other) genes.
Results from gene transfer experiments involving AIDS or cancer patients are scanty. For example, in some cases the HIV+ member of an identical twin pair develops positive skin responses following a gene transfer procedure, but whether this change will lead to clinical benefits is not yet known.
Dr. Philip Greenberg also refers to "transient" antiviral effects and "proof of concept" in gene transfer experiments involving modified HIV genes in patients with AIDS.
A wide range of clinical experiments involving patients with a variety of cancers is under way. Dr. Blaese said there is some evidence of efficacy, such as tumor shrinkage in patients with glioblastomas. Some of the protocols call for the gene transfer procedure to induce immune system responses against the tumor, according to Dr. Gary Nabel. In some cases, patients appear to go into long-term remission; in other cases, the effects are transient. Partial effects are commonplace in cancer treatment, and gene therapy approaches therefore may find acceptance as a useful addition to the therapeutic arsenal.
Dr. Nabel and Dr. John Mendelsohn pointed out that, in gene transfer experiments involving cancer patients, better measures of biological activity are needed. This need is particularly acute in early tests involving patients with advanced disease when other treatments and other clinical abnormalities make assessment of a single experimental procedure exceedingly difficult.
Responses to the question of whether the field is ready for clinical trials:
The U.S. Patent and Trademark Office (PTO) has issued several broad-based patents covering fundamental gene therapy technologies, including a patent granted to NIH and licensed exclusively to Gene Therapy, Inc., covering ex vivo gene therapy and another patent granted to the University of Michigan and licensed exclusively to Genovo that covers any viral gene therapy vector carrying the CFTR gene, which is impaired in individuals with CF.
Ms. Eisenberg said that these examples as well as other signs indicate this field of biotechnology is likely to be "more littered" with patents than is the earlier emerging field of biotechnology involving the discovery and development of therapeutic proteins.
Ms. Eisenberg attributes this difference to the fact that universities and other research institutions are being even more aggressive now than a few years ago in pursuing patent protection for intellectual property their researchers are developing. The Bayh-Dole Act, which specifies that such institutions may retain ownership in patents arising from federally sponsored research, now provides strong incentives for pursuing patents--raising expectations in the university community that royalties from licensing agreements eventually will become a significant source of revenue.
Although in some noteworthy cases involving biotechnology inventions universities are benefitting from significant royalty payments, there are potential problems to face from the flurry of patent applications being put forth in the field of gene therapy, according to Ms. Eisenberg. Perhaps chief among them is that research teams and clinicians may, in effect, be faced with a series of "toll booths" along the road to developing and implementing effective gene therapy procedures. She says that research groups may be hemmed in and financially pinched if they have to enter into complex cross-licensing agreements or if institutions set royalty requirements at levels that are too high. Additional complications include potential priority disputes between competing "inventors," disagreements over ownership when researchers at several institutions are collaborating on a project, and differences arising because some researchers such as medical geneticists tend not to patent their work, whereas other researchers such as molecular biologists do so.
List of Speakers Kenneth Berns, M.D. Gary J. Nabel, M.D. Deparment of Microbiology Howard Hughes Medical Institute Cornell University Medical College University of Michigan Medical Center New York, NY 10021 Ann Arbor, MI 48109-0650 Michael R. Blaese, M.D. Arthur W. Nienhuis, M.D. National Center for Human Genome Research St. Jude Children's Research Hospital National Institutes of Health Memphis, TN 38101 Bethesda, MD 20892 Ronald G. Crystal, M.D. Thomas Shenk, Ph.D. New York Hospital-Cornell Medical College Department of Molecular Biology New York, NY 10021 Princeton University Princeton, NJ 08544 Rebecca Eisenberg, J.D. Alan E. Smith, Ph.D. University of Michigan Law School Genzyme Corporation Ann Arbor, MI 48109-1215 Framington, MA 01701-9322 Philip D. Greenberg, M.D. James M. Wilson, M.D. University of Washington University of Pennsylvania Seattle, WA 98195 Philadelphia, PA 19104 John Mendelsohn, M.D. Department of Medicine Memorial Sloan-Kettering Cancer Center New York, NY 10021-6094 Richard Mulligan, Ph.D. Somatix Therapy Corporation Alameda, CA 94501 and Massachusetts Institute of Technology Whitehead Institute Cambridge, MA 02124
Some of these investigators criticized current proponents of gene therapy for portraying the field in unrealistic terms and misrepresenting progress as more rapid than it has been. For example, Dr. Joseph Goldstein called for greater realism in the way these researchers present views of their field to the public. He also pointed out that the development of any new therapeutic product is a laborious, time-consuming effort.
Dr. Goldstein said that some of the diseases now targeted by gene therapy researchers might be treated sooner, by other strategies, if investigators pursued more traditional studies into the pathophysiologic basis of the diseases in question. He cited several examples where this alternative approach has paid off either recently or several decades ago. For instance, prednisone treatment reverses steps in a defective sterol metabolic pathway that otherwise leads to masculinization. In a more recent development, an inhibitor of cholesterol production (lovastatin) overcomes a LDL receptor deficiency and, by lowering cholesterol levels, helps to prevent coronary heart disease.
Dr. Goldstein also referred to several genetic diseases that arise because of protein trafficking abnormalities. In some of those cases, the critical mutations lie outside the functional coding region of the enzyme product and, instead, serve to misdirect nascent proteins, which are transported into the wrong biological compartments. He called for basic research that could provide an alternative means to gene therapy for correcting such defects.
Dr. Irving Weissman and Dr. Goldstein said that studies with animal models deserve greater emphasis than they are receiving by researchers who are moving quickly from basic research to the clinic to test new ideas about gene therapy. This general problem is particularly applicable to several unsolved problems involving stem cells, which are important but elusive targets of many gene transfer protocols in which long-term gene expression is a major goal.
Dr. Weissman pointed out that stem cell biology in humans and mice is essentially equivalent. From studies on mice, investigators have learned that there are three critical subsets of stem cells in bone marrow and that the most desirable subset for gene transfer is the rarest and is very difficult to work with.
A key problem in the use of retroviral vectors is to determine which factors will induce self-renewing stem cells to divide. Without such detailed information that can be applied practically, gene transfer procedures will likely fail because genes will not be integrating into target progenitor cells. Dr. Weissman said that, with such fundamental obstacles to human gene transfers, it may make sense to focus instead on activating genes that are already present rather than on replacing defective or missing genes.
Dr. Victor Dzau pointed out that, for certain clinical conditions including several that affect the cardiovascular system, short-term rather than long-term gene expression may be all that is needed to address specific problems. Moreover, in a rabbit model system, studies indicate that localized high pressure can improve DNA transduction rates, enabling antisense oligonucleotides to block transiently a cell-proliferative response that otherwise may interfere with surgically grafted blood vessels. Experiments indicate that high pressure also enhances the delivery of oligonucleotides into cultured human cells, improving the efficiency of transduction.
Dr. Gerald Crabtree described the use of synthetic, lipid-soluble dimerizing reagents that can be used to bring cellular regulatory proteins into covalent juxtaposition, thereby changing their functional status. For example, with appropriate dimeric reagents, specific transcriptional factors might be modified in such a way that they permanently activate this process, meaning that a transgenic cell produces high levels of the designated gene product. Another potential use of such dimerizing reagents would be to cross-link specific cell receptors to induce apoptosis. Although this approach shows promise and many other applications are imaginable, studies are limited so far to cellular systems and considerable work will be needed before animal model studies can be undertaken.
Dr. Flossie Wong-Staal pointed out that in vitro studies or animal models of AIDS are far from adequate, making it best to go forward rapidly with small, focused clinical trials to test gene transfer procedures. Although the rationale for using ribozyme genes to block HIV gene expression appears sound when tested at the cellular level, many questions, such as the extent to which target cells in patients will be genetically modified and then selected and whether HIV will develop resistance to the ribozyme, can only be addressed through clinical studies.
Dr. Anderson outlined a variety of gene therapy research studies at his institution, suggesting that this locally concentrated diversity of interests and ideas is another sign that this field is healthy and populated with creative young investigators. He also described a long-term project that involves making a series of improvements in a current retroviral-based vector that could extend its half-life in the host circulatory system, increase its efficiency of binding to and entering specific target cells of the host, improve its chances of delivering genes for long-term expression, and eventually lead to a readily injectable gene-delivery product. Efforts to realize these goals are only at the "very beginning."
Other current basic research developments may eventually help solve some of the challenges that investigators conducting human gene transfer protocols now face. For example, Dr. Donald Kohn described efforts to modify the long terminal repeat (LTR) in a retroviral vector now being used in gene transfer protocols as a way of extending the expression of transferred genes after they are delivered to target cells. Hematopoietic cells from mice are providing a valuable model in which to study this problem, and some results indicate that methylation within the LTR correlates with the disappearance of transferred gene expression.
In a model system in which human bone marrow cells are introduced into immunologically deficient nude mice, Dr. Kohn and his collaborators find that the addition of stroma enhances gene transfer in vitro and also extends long-term expression of the transduced genes. The impact of growth factors on these steps is also being evaluated. Dr. Kohn said that, despite the value of this information from experiments in mice, clinical trials are needed to understand in detail how each of these steps work in humans.
One important problem that has come to light from early gene transfer clinical studies is that host immune responses may abbreviate expression of transferred genes. Dr. Paul Tolstoshev described efforts to develop sophisticated vectors that can overcome this problem. Less immunogenic vectors are being constructed for use in conjunction with immunosuppressive agents such as dexamethasone or cyclosporin that can reduce immune system responses, including deleterious inflammatory reactions.
Dr. Anderson pointed out that progress is more likely to be rapid if individual investigators--rather than a central committee--direct research decision making. He also recommended that the development and use of vectors made in NIH-supported specialized laboratories not be restricted to only those researchers whose work is being supported by NIH. He was less certain whether a policy of limiting such vector development to research on orphan diseases should be adopted.
Dr. Barrie Carter pointed out that efforts to begin the first clinical trials and subsequent efforts to test additional gene transfer protocols in clinical settings are driving a great deal of basic research in biology. Although NIH programs provided the fundamental research from which gene therapy derives, industry now furnishes enormous resources to further these developments. He noted that NIH spends about $200 million annually on gene therapy research, and this amount represents less than 2 percent of total NIH research expenditures. He recommended that NIH spending be maintained at this level, concentrating in several program areas such as gene delivery systems, target cell biology, and preclinical models.
Dr. Carter noted that basic and clinical research within the NIH Intramural Program is a valuable component of overall efforts in the field of gene therapy. He also praised the role NIH plays in supporting programs in basic research on viral vectors and at General Clinical Research Centers. However, he questioned the value of NIH setting up new gene vector production facilities, suggesting that industry can do a better job producing vectors. Dr. Tolstoshev noted that companies are conducting a great deal of fundamental research on gene vectors and, in many cases, these vectors are being made available to university researchers for testing and evaluation.
Industry representatives pointed to several technology transfer arrangements that are helpful to them, despite specific obstacles which sometimes arise. For example, Cooperative Research and Development Agreements (CRADAs) are now being used extensively to establish relationships between companies and NIH investigators in the field of gene therapy. Dr. Tolstoshev said that the many CRADAs established between his company, Genetic Therapy, Inc. (GTI), and individual NIH investigators are particularly helpful in leveraging the company's expertise. Other types of agreements, including material transfer agreements and scientific collaborations between industry and university researchers, are providing a major source of funding for this developing field, and that source could grow larger as major established pharmaceutical companies take a greater interest in gene therapy.
Legal and policy difficulties sometimes have made CRADA negotiations drawn out and cumbersome. Dr. Tolstoshev noted that, by eliminating a clause calling for "reasonable pricing" of drugs and other products that may flow from a CRADA, NIH removed what had become an important stumbling block for industry. However, he also said that protracted negotiation of the legal terms of many CRADAs can still be an impediment to efficient technology transfer.
Industry representatives pointed to other important policy issues, including a need for clear-cut patenting policies and the relative value of exclusive versus non-exclusive licensing agreements. Industry representatives said that, in general, licensing agreements granting particular companies the exclusive right to commercialize intellectual property developed by NIH investigators are more likely to provide essential incentives to pursue development than are non-exclusive agreements.
List of Speakers French W. Anderson, M.D. Paul Tolstoshev, Ph.D. Gene Therapy Labs Genetic Therapy, Inc. University of Southern California Gaithersburg, MD 20878 School of Medicine Los Angeles, CA 90033 Barrie J. Carter, Ph.D. Irving L. Weissman, Ph.D. Targeted Genetics Department of Pathology Seattle, WA 98101 Stanford University School of Medicine Stanford, CA 94305-5324 Gerald R. Crabtree, M.D. Flossie Wong-Staal, Ph.D. HHMI University of California, San Diego Beckman Center La Jolla, CA 92093-0665 Stanford University School of Medicine Stanford, CA 94305-5428 Victor J. Dzau, M.D. Stanford University School of Medicine Stanford, CA 94305 Joseph L. Goldstein, M.D. Deparment of Medical Genetics University of Texas Southwest Medical School Dallas, TX 75235-9046 Mark A. Kay, Ph.D. Division of Medical Genetics University of Washington Seattle, WA 98195 Donald B. Kohn, M.D. Division of Research Immunology Bone Marrow Transplant Unit Childrens Hospital Los Angeles Los Angeles, CA 90027