- Email: [email protected]
Regulatory considerations for gene-therapy strategies and products
189
commerce and regulation through the R & D companies, rather than through in-house research. More ideas originate in R & D companies, and if a particular research direction appears to be a cul-de-sac, it can be easily dropped by not renewing the agreement. If, on the other hand, the research looks promising, further agreements would be drawn up, or even purchase of the company would be considered. Sandoz, Merck, Parke-Davis (a division o f Warner-Lambert) and Green Cross Corporation have invested millions of dollars in GTI, Vical and Viagene, respectively, to develop clinical applications of gene therapy. Other drug-company giants are scouting for deals. It is hard to tell, at present, which company will be first to bring gene therapy to the market, and whether it will be a commercial success. GTI, the pioneer in the field, has been involved in more than a dozen clinical trials and has acquired much experience. However, the other companies that also have a strong scientific environment could become important competitors. Firms such as TargeTech or Vical, which are develop-
ing non-viral vectors, might be first to commercialize gene-therapy products, since gaining regulatory approval is likely to be easier. However, even when a gene-therapy protocol has been shown to be safe and effective, and has regulatory approval, commercial success will depend on the market size, and on the willingness of the company to meet the costs. Many governments are currently re-evaluating the cost o f health-care systems and the problems of financing expensive new treatments. The costs o f gene therapy must, however, be compared with those for current treatments. A costly, but definitive cure for a disabling genetic disease may, in the long-run, be cheaper than lifelong palliative treatment - the cost o f current treatment for Gaucher's disease is estimated at US$450 000 annually. For some, even if not all, the diseases which are currently being targeted, gene therapy is likely to prove commercially as well as medically beneficial. Reference 1 Anderson, W. F. (1992) Science256, 808-813
Regulatory considerations for gene-therapy strategies and products Nelson A. Wivel The current view of human gene therapy in the USA is the product of a fairly prolonged, detailed analysis of the relevant issues. At the time of the Asilomar conference in 1975, the discussion centered on the principal problem facing molecular biologists from the world's leading laboratories - the establishment o f acceptable conditions for conducting recombinant D N A experimentation; human gene therapy was not even a consideration. However, in response to this meeting, the National Institutes of Health (NIH) established an Office of Recombinant D N A Activities and developed the ' N I H Guidelines for Research Involving Recombinant D N A Molecules '1. The issue o f human gene therapy became the focus o f public discussion in the early ] 980s. Under President Carter, a commission was formed, and in 1982, the report 'Splicing Life' (1Ke£ 2), was published. In response to this report, the N I H Recombinant D N A Advisory Committee (1KAC) established a working group on human gene therapy that carried out a thorough study o f the science and ethics of this techN, A . Wiuel is at file Q[fice of Recombinant D N A Activities, Department of Health and Human Services, National Institutes of Health, Bethesda, M D 20892, USA. © 1993, Elsevier Science Publishers Ltd (UK)
nology. The document, 'Points to Consider in the Design and Submission o f Protocols for the Transfer o f Recombinant D N A into the Genome o f Human Subjects' (Re£ 3), was developed over the period from ]984 to 1986, and was adopted by the R A C in 1986. Periodic revisions were included as the technology for gene therapy became defined more precisely. In ] 989, the first human gene-transfer protocol was approved, and in ]990, the first human gene-therapy protocol was approved. As of March 1993, 42 protocols have been approved by the NIH. (See Tables 1 and 2 for a list of approved protocols.) C o n t e x t o f the N I H review o f gene therapy The Points to Consider is the principal document used by the K A C in its consideration of human genetransfer research proposals. Within this framework, there must be concern for the clinical benefit to the research subject, concern regarding the informed participation of research subjects, concern about fair subject selection, and the need for special biosafety precautions when there is genetic uncertainty. For example, certain types of vectors used to deliver genes pose a small risk of causing cancer. An assessment o f the risk of the treatment versus the benefit to the patient, the scientific validity of the experiments, and TIBTECHMAY 1993 (VOL 11)
190
commerce and re2ulation Table 1. Human gene-therapy protocols - retrovirus vectors Disease
Gene inserted
Adenosine deaminase deficiency
Adenosine deaminase (ADA)
Advanced cancers
Tumor necrosis factor (TNF) Interleukin 2 (IL-2) Interleukin 4 (IL-4) Multi-drug resistance (MDR)
Renal-cell carcinoma
Interleukin 2 (Ib2)
Melanoma
Interleukin 2 (IL-2)
Ovarian cancer
Herpes simplex virus thymidine kinase (HSV-TK) Multi-drug resistance (MDR)
Hypercholesterolemia
Low density lipoprotein (LDL) receptor
AIDS
Hygromycin phosphotransferaseherpes simplex virus thymidine kinase (HyTK) Rev MIO
Neuroblastoma
Interleukin 2 (IL-2)
Primary and metastatic brain tumors
HSV-TK
Non-small-cell lung carcinoma
Antisense k-ras/p53
Gaucher's disease
Glucocerebrosidase
the safety parameters that need to be addressed are of foremost importance. While it is assumed that the benefits to the patient must exceed the risks of the therapy, it should be recognized that there are certain experiments in which the transduced gene is used as a marker, which have no therapeutic benefit. In such cases, it should be realized that any potential benefits will accrue only for succeeding patients if the biological questions have been answered successfully by the gene-marking studies. In those cases in which gene therapy is the goal, the patient selection criteria heavily influence the risk:benefit ratio. The stage of the disease, the potential for therapeutic outcome, and the prevention of a life-threatening condition are key factors that have to be taken into consideration. In addition, existing treatments and their efficacy, or the lack thereof, have to be carefully evaluated when considering gene therapy. 'In their evaluation of proposals involving the transfer of recombinant D N A into human subjects, the R A C will consider whether the design of such experiments offers adequate assurance that their consequences will not go beyond their purpose '4. In effect, this is another way of saying that there must be sufficient prior knowledge to conclude that the probable benefits of the therapy outweigh the possible risks to the patient and the known benefits of available alternative treatments. In assessing the scientific validity of a protocol, it is essential that there is a sound rationale for the treatment of the disease in this manner. Is the intent to
TIBTECHMAY1993(VOL11)
compensate for a gene mutation by adding a normal copy of the gene, or is the intent to modify a gene product by altering the endogenous gene? The scientific hypothesis should be based on a sound idea and the results should reveal useful data. Carrying out a precise evaluation of the results is exceedingly important. Does one anticipate a partial or complete response? Will any useful information be derived if the treatment is unsuccessful? What are the criteria for discontinuing treatment if toxicity occurs? Much of the scientific validity for a given research proposal is derived from the quality of the preclinical data. The characterization of gene function, including determining the normal levels of expression, and the specificity for cells and tissues, is of critical importance. The toxicity of the gene product is a principal consideration, as overexpression or inappropriate expression could create a clinical emergency. In vitro assays of the delivery system and gene function are essential. It is necessary to establish that the same target cells that will be transduced in the patient can be transduced, that sufficient levels of expression can be maintained for a sufficiently prolonged period to achieve a clinical effect, and that the transduction efficiency is adequate. The use of animal models is exceedingly important, although it is recognized that a precise model, such as the Watanabe rabbit for familial hypercholesterolemia, is relatively rare. T o meet the safety criteria, a thorough characterization o f the delivery system, including a full description o f the methods and reagents to be employed for gene delivery, is required. A complete nucleotide sequence analysis, or a detailed restriction enzyme map of the total construct, including the regulatory elements such as promoters, enhancers, polyadenylation sites, and replication origins, is also required. If using a viral delivery system, it is necessary to determine whether or not stable integration of the virus vector has occurred, and if there are common or random insertion sites. Since retroviral vectors integrate at random sites in the genome of the host cell, there is the possibility o f insertional mutagenesis. In the case of adenovirus vectors, host immunity and its effect on gene transfer should be considered; furthermore, there is a definite possibility of rescue of the replicationdefective vector by wild-type adenovirus. For example, a patient who still has their tonsils is likely to have adenovirus in the crypts of that lymphoid tissue, and thus has a definite source of rescue o f the crippled adenovirus vector. When a viral vector is used in an ex vivo transduction procedure, there are inherent risks involving both the target cells and the virus. For example, the efficiency of viral-vector infection is increased by exposing the target cells to a small volume of viral supernatant, thus limiting the number of cells that can be involved in the transduction step. The transduced cells must be expanded exponentially to provide sufficient numbers for patient treatment and the steps in this expansion process increase the possibility for contamination by adventitial microorganisms. Some o f the approved protocols have used a non-viral
191
commerce and re2ulation delivery system, i.e. liposomes combined with the direct injection of the mixture into a subcutaneous tumor mass. However, this approach introduces the risk o f targeting inappropriate cells, including those of the germline. Although the ability to control the regulation of gene expression has not been a physiological requirement for the protocols approved thus far, it is an issue that will probably have to be faced in the not-toodistant future. Regulation ofgene expression requires a knowledge o f the normal level, the minimal level required, and the maximal level of gene expression that is likely to be attained. One has to determine if excessive expression or expression in an inappropriate tissue leads to toxicity. There have to be stop criteria to prevent expression in a setting where withdrawal of the gene-vector construct is not possible; thus it may be necessary to add a 'suicide' gene to the system to abort toxic reactions rapidly. A commonly used 'suicide' gene is the thymidine kinase gene derived from the herpes simplex virus. This gene, when transduced into cells, renders them sensitive to the drug gancyclovir, creating the option o f killing the cells quickly, if the need arises. Context o f the FDA review o f gene therapy Like the NIH, the Food and Drug Administration (FDA) has developed a Points to Consider document s for the consideration o f somatic-cell therapy and gene therapy. Although the FDA is a regulatory agency, this Points to Consider document is not mandatory, but rather is regarded as a compendium of pertinent issues that should be considered in this particular area of biotecbnology. There are separate mandatory regulations that pertain to gene therapy. O f particular concern is the characterization of cell populations that are to be administered to patients. If collecting cells from patients, the following information is required: the cell type; the criteria for donor selection; tissue typing; and the procedures used to collect the cells. Cell-culture operations are subject to quality-control procedures that involve testing materials, manufacturing controls, and equipment validation and monitoring. There are clearly defined specifications for culture media and cell-banking procedures: a requirement to test for adventitious agents; to monitor cell identity and heterogeneity; and to test the particular products biosynthesized by the cells. In agreement with the N I H guidelines, pre-clinical testing of the gene-vector construct, using a combination of animal studies and in vitro studies, is required so that safety and efficacy can be established. Retroviral vectors must be tested for replicationcompetent virus, appropriate functioning of the inserted gene must be demonstrated, and in vivo safety testing, using animal models, is often essential to reveal adverse effects of the gene-modified cells or their products. Unlike the NIH, the FDA focuses considerable attention on lot-to-lot manufacturing control and release testing. It is recognized that preparations
Table 2. Human gene-therapy protocols non-retrovirus vectors Disease
Gene inserted
Advanced cancer
HLA-B7 (Cationic liposomes)
AIDS
Rev MIO (Ballistic injection)
Cystic fibrosis
Cystic fibrosis transmembrane conductance regulator (CFTR)(Adenovirus vectors type 2 and 5)
intended solely for individual patients or subjects differ from products that are prepared as large batches; however, the type of testing that is carried out detects lot-to-lot variation, and thus is a measure o f the reproducibility of the procedures. Tests for cell identity, function, viability, and advendtial agents are necessary, and cells that are frozen for subsequent implantation must be tested, by lot, on thawing. Current design and the future At present, there are two independent systems for reviewing human gene-therapy proposals in the USA, one at the NIH, and one at the FDA. The N I H and its K A C function in response to the perceived need for public participation in genetic research policy. As a result, the R A C meetings are completely open to the public, the meetings are announced in advance with provision for public comments, and all the review materials are available to the public, on request. Given the provisions of the N I H Guidelines, only those investigators who receive N I H funding, or who work in institutions that receive N I H funding for recombinant D N A research are required to undergo review by the KAC. However, approval by the FDA is also required. Those investigators and institutions who do not receive N I H funding are exempt from K A C review, but must undergo FDA review. Unlike the R A C , FDA review does not have to be conducted in a public forum; the precedent for this derives from the fact that the FDA has the responsibility for reviewing proprietary information. While the Points to Consider documents developed by the N I H and FDA have been critical to the review of human gene-therapy protocols, undoubtedly they will be revised as experience in reviewing protocols continues to accumulate, and as new scientific developments occur. The potential for flexibility in these documents may be one of their principal virtues as gene therapy moves from its infancy to a form of treatment with demonstrated efficacy. References 1 Federal Register (1986) 51, 16958-16985 2 Splicing Life {1982) President's Commission for the Study of Edlical Problems in Medicine and Biomedical and Behavioral Research, pp. 1-115 3 The revised Points to Consider document (1990) Hum. Gene Ther. I, 93-103 4 Juengst, E. T. (1990) Hum. Gene Ther. I, 425 433 5 Draft of FDA's Points to Consider in Human Somatic Cell Therapy and Gene Therapy 0991) Hum. Gene "lher. 2, 251 256 TIBTECH MAY 1993 (VOL 11)
commerce and regulation through the R & D companies, rather than through in-house research. More ideas originate in R & D companies, and if a particular research direction appears to be a cul-de-sac, it can be easily dropped by not renewing the agreement. If, on the other hand, the research looks promising, further agreements would be drawn up, or even purchase of the company would be considered. Sandoz, Merck, Parke-Davis (a division o f Warner-Lambert) and Green Cross Corporation have invested millions of dollars in GTI, Vical and Viagene, respectively, to develop clinical applications of gene therapy. Other drug-company giants are scouting for deals. It is hard to tell, at present, which company will be first to bring gene therapy to the market, and whether it will be a commercial success. GTI, the pioneer in the field, has been involved in more than a dozen clinical trials and has acquired much experience. However, the other companies that also have a strong scientific environment could become important competitors. Firms such as TargeTech or Vical, which are develop-
ing non-viral vectors, might be first to commercialize gene-therapy products, since gaining regulatory approval is likely to be easier. However, even when a gene-therapy protocol has been shown to be safe and effective, and has regulatory approval, commercial success will depend on the market size, and on the willingness of the company to meet the costs. Many governments are currently re-evaluating the cost o f health-care systems and the problems of financing expensive new treatments. The costs o f gene therapy must, however, be compared with those for current treatments. A costly, but definitive cure for a disabling genetic disease may, in the long-run, be cheaper than lifelong palliative treatment - the cost o f current treatment for Gaucher's disease is estimated at US$450 000 annually. For some, even if not all, the diseases which are currently being targeted, gene therapy is likely to prove commercially as well as medically beneficial. Reference 1 Anderson, W. F. (1992) Science256, 808-813
Regulatory considerations for gene-therapy strategies and products Nelson A. Wivel The current view of human gene therapy in the USA is the product of a fairly prolonged, detailed analysis of the relevant issues. At the time of the Asilomar conference in 1975, the discussion centered on the principal problem facing molecular biologists from the world's leading laboratories - the establishment o f acceptable conditions for conducting recombinant D N A experimentation; human gene therapy was not even a consideration. However, in response to this meeting, the National Institutes of Health (NIH) established an Office of Recombinant D N A Activities and developed the ' N I H Guidelines for Research Involving Recombinant D N A Molecules '1. The issue o f human gene therapy became the focus o f public discussion in the early ] 980s. Under President Carter, a commission was formed, and in 1982, the report 'Splicing Life' (1Ke£ 2), was published. In response to this report, the N I H Recombinant D N A Advisory Committee (1KAC) established a working group on human gene therapy that carried out a thorough study o f the science and ethics of this techN, A . Wiuel is at file Q[fice of Recombinant D N A Activities, Department of Health and Human Services, National Institutes of Health, Bethesda, M D 20892, USA. © 1993, Elsevier Science Publishers Ltd (UK)
nology. The document, 'Points to Consider in the Design and Submission o f Protocols for the Transfer o f Recombinant D N A into the Genome o f Human Subjects' (Re£ 3), was developed over the period from ]984 to 1986, and was adopted by the R A C in 1986. Periodic revisions were included as the technology for gene therapy became defined more precisely. In ] 989, the first human gene-transfer protocol was approved, and in ]990, the first human gene-therapy protocol was approved. As of March 1993, 42 protocols have been approved by the NIH. (See Tables 1 and 2 for a list of approved protocols.) C o n t e x t o f the N I H review o f gene therapy The Points to Consider is the principal document used by the K A C in its consideration of human genetransfer research proposals. Within this framework, there must be concern for the clinical benefit to the research subject, concern regarding the informed participation of research subjects, concern about fair subject selection, and the need for special biosafety precautions when there is genetic uncertainty. For example, certain types of vectors used to deliver genes pose a small risk of causing cancer. An assessment o f the risk of the treatment versus the benefit to the patient, the scientific validity of the experiments, and TIBTECHMAY 1993 (VOL 11)
190
commerce and re2ulation Table 1. Human gene-therapy protocols - retrovirus vectors Disease
Gene inserted
Adenosine deaminase deficiency
Adenosine deaminase (ADA)
Advanced cancers
Tumor necrosis factor (TNF) Interleukin 2 (IL-2) Interleukin 4 (IL-4) Multi-drug resistance (MDR)
Renal-cell carcinoma
Interleukin 2 (Ib2)
Melanoma
Interleukin 2 (IL-2)
Ovarian cancer
Herpes simplex virus thymidine kinase (HSV-TK) Multi-drug resistance (MDR)
Hypercholesterolemia
Low density lipoprotein (LDL) receptor
AIDS
Hygromycin phosphotransferaseherpes simplex virus thymidine kinase (HyTK) Rev MIO
Neuroblastoma
Interleukin 2 (IL-2)
Primary and metastatic brain tumors
HSV-TK
Non-small-cell lung carcinoma
Antisense k-ras/p53
Gaucher's disease
Glucocerebrosidase
the safety parameters that need to be addressed are of foremost importance. While it is assumed that the benefits to the patient must exceed the risks of the therapy, it should be recognized that there are certain experiments in which the transduced gene is used as a marker, which have no therapeutic benefit. In such cases, it should be realized that any potential benefits will accrue only for succeeding patients if the biological questions have been answered successfully by the gene-marking studies. In those cases in which gene therapy is the goal, the patient selection criteria heavily influence the risk:benefit ratio. The stage of the disease, the potential for therapeutic outcome, and the prevention of a life-threatening condition are key factors that have to be taken into consideration. In addition, existing treatments and their efficacy, or the lack thereof, have to be carefully evaluated when considering gene therapy. 'In their evaluation of proposals involving the transfer of recombinant D N A into human subjects, the R A C will consider whether the design of such experiments offers adequate assurance that their consequences will not go beyond their purpose '4. In effect, this is another way of saying that there must be sufficient prior knowledge to conclude that the probable benefits of the therapy outweigh the possible risks to the patient and the known benefits of available alternative treatments. In assessing the scientific validity of a protocol, it is essential that there is a sound rationale for the treatment of the disease in this manner. Is the intent to
TIBTECHMAY1993(VOL11)
compensate for a gene mutation by adding a normal copy of the gene, or is the intent to modify a gene product by altering the endogenous gene? The scientific hypothesis should be based on a sound idea and the results should reveal useful data. Carrying out a precise evaluation of the results is exceedingly important. Does one anticipate a partial or complete response? Will any useful information be derived if the treatment is unsuccessful? What are the criteria for discontinuing treatment if toxicity occurs? Much of the scientific validity for a given research proposal is derived from the quality of the preclinical data. The characterization of gene function, including determining the normal levels of expression, and the specificity for cells and tissues, is of critical importance. The toxicity of the gene product is a principal consideration, as overexpression or inappropriate expression could create a clinical emergency. In vitro assays of the delivery system and gene function are essential. It is necessary to establish that the same target cells that will be transduced in the patient can be transduced, that sufficient levels of expression can be maintained for a sufficiently prolonged period to achieve a clinical effect, and that the transduction efficiency is adequate. The use of animal models is exceedingly important, although it is recognized that a precise model, such as the Watanabe rabbit for familial hypercholesterolemia, is relatively rare. T o meet the safety criteria, a thorough characterization o f the delivery system, including a full description o f the methods and reagents to be employed for gene delivery, is required. A complete nucleotide sequence analysis, or a detailed restriction enzyme map of the total construct, including the regulatory elements such as promoters, enhancers, polyadenylation sites, and replication origins, is also required. If using a viral delivery system, it is necessary to determine whether or not stable integration of the virus vector has occurred, and if there are common or random insertion sites. Since retroviral vectors integrate at random sites in the genome of the host cell, there is the possibility o f insertional mutagenesis. In the case of adenovirus vectors, host immunity and its effect on gene transfer should be considered; furthermore, there is a definite possibility of rescue of the replicationdefective vector by wild-type adenovirus. For example, a patient who still has their tonsils is likely to have adenovirus in the crypts of that lymphoid tissue, and thus has a definite source of rescue o f the crippled adenovirus vector. When a viral vector is used in an ex vivo transduction procedure, there are inherent risks involving both the target cells and the virus. For example, the efficiency of viral-vector infection is increased by exposing the target cells to a small volume of viral supernatant, thus limiting the number of cells that can be involved in the transduction step. The transduced cells must be expanded exponentially to provide sufficient numbers for patient treatment and the steps in this expansion process increase the possibility for contamination by adventitial microorganisms. Some o f the approved protocols have used a non-viral
191
commerce and re2ulation delivery system, i.e. liposomes combined with the direct injection of the mixture into a subcutaneous tumor mass. However, this approach introduces the risk o f targeting inappropriate cells, including those of the germline. Although the ability to control the regulation of gene expression has not been a physiological requirement for the protocols approved thus far, it is an issue that will probably have to be faced in the not-toodistant future. Regulation ofgene expression requires a knowledge o f the normal level, the minimal level required, and the maximal level of gene expression that is likely to be attained. One has to determine if excessive expression or expression in an inappropriate tissue leads to toxicity. There have to be stop criteria to prevent expression in a setting where withdrawal of the gene-vector construct is not possible; thus it may be necessary to add a 'suicide' gene to the system to abort toxic reactions rapidly. A commonly used 'suicide' gene is the thymidine kinase gene derived from the herpes simplex virus. This gene, when transduced into cells, renders them sensitive to the drug gancyclovir, creating the option o f killing the cells quickly, if the need arises. Context o f the FDA review o f gene therapy Like the NIH, the Food and Drug Administration (FDA) has developed a Points to Consider document s for the consideration o f somatic-cell therapy and gene therapy. Although the FDA is a regulatory agency, this Points to Consider document is not mandatory, but rather is regarded as a compendium of pertinent issues that should be considered in this particular area of biotecbnology. There are separate mandatory regulations that pertain to gene therapy. O f particular concern is the characterization of cell populations that are to be administered to patients. If collecting cells from patients, the following information is required: the cell type; the criteria for donor selection; tissue typing; and the procedures used to collect the cells. Cell-culture operations are subject to quality-control procedures that involve testing materials, manufacturing controls, and equipment validation and monitoring. There are clearly defined specifications for culture media and cell-banking procedures: a requirement to test for adventitious agents; to monitor cell identity and heterogeneity; and to test the particular products biosynthesized by the cells. In agreement with the N I H guidelines, pre-clinical testing of the gene-vector construct, using a combination of animal studies and in vitro studies, is required so that safety and efficacy can be established. Retroviral vectors must be tested for replicationcompetent virus, appropriate functioning of the inserted gene must be demonstrated, and in vivo safety testing, using animal models, is often essential to reveal adverse effects of the gene-modified cells or their products. Unlike the NIH, the FDA focuses considerable attention on lot-to-lot manufacturing control and release testing. It is recognized that preparations
Table 2. Human gene-therapy protocols non-retrovirus vectors Disease
Gene inserted
Advanced cancer
HLA-B7 (Cationic liposomes)
AIDS
Rev MIO (Ballistic injection)
Cystic fibrosis
Cystic fibrosis transmembrane conductance regulator (CFTR)(Adenovirus vectors type 2 and 5)
intended solely for individual patients or subjects differ from products that are prepared as large batches; however, the type of testing that is carried out detects lot-to-lot variation, and thus is a measure o f the reproducibility of the procedures. Tests for cell identity, function, viability, and advendtial agents are necessary, and cells that are frozen for subsequent implantation must be tested, by lot, on thawing. Current design and the future At present, there are two independent systems for reviewing human gene-therapy proposals in the USA, one at the NIH, and one at the FDA. The N I H and its K A C function in response to the perceived need for public participation in genetic research policy. As a result, the R A C meetings are completely open to the public, the meetings are announced in advance with provision for public comments, and all the review materials are available to the public, on request. Given the provisions of the N I H Guidelines, only those investigators who receive N I H funding, or who work in institutions that receive N I H funding for recombinant D N A research are required to undergo review by the KAC. However, approval by the FDA is also required. Those investigators and institutions who do not receive N I H funding are exempt from K A C review, but must undergo FDA review. Unlike the R A C , FDA review does not have to be conducted in a public forum; the precedent for this derives from the fact that the FDA has the responsibility for reviewing proprietary information. While the Points to Consider documents developed by the N I H and FDA have been critical to the review of human gene-therapy protocols, undoubtedly they will be revised as experience in reviewing protocols continues to accumulate, and as new scientific developments occur. The potential for flexibility in these documents may be one of their principal virtues as gene therapy moves from its infancy to a form of treatment with demonstrated efficacy. References 1 Federal Register (1986) 51, 16958-16985 2 Splicing Life {1982) President's Commission for the Study of Edlical Problems in Medicine and Biomedical and Behavioral Research, pp. 1-115 3 The revised Points to Consider document (1990) Hum. Gene Ther. I, 93-103 4 Juengst, E. T. (1990) Hum. Gene Ther. I, 425 433 5 Draft of FDA's Points to Consider in Human Somatic Cell Therapy and Gene Therapy 0991) Hum. Gene "lher. 2, 251 256 TIBTECH MAY 1993 (VOL 11)