En Route to Ethical Recommendations for Gene Transfer Clinical Trials
2008; Elsevier BV; Volume: 16; Issue: 3 Linguagem: Inglês
10.1038/mt.2008.13
ISSN1525-0024
AutoresNancy M. P. King, Odile Cohen‐Haguenauer,
Tópico(s)CRISPR and Genetic Engineering
ResumoIn Geneva, Switzerland, on 2–3 April 2007, Clinigene-NoE (the European Network for the Advancement of Clinical Gene Transfer and Therapy) and Consert (program on Concerted Safety and Efficiency Evaluation of Retroviral Transgenesis for Gene Therapy of Inherited Diseases)—two European Union (EU) programs that aim to facilitate the development of sound, safe, and effective human gene transfer—held a joint "think tank" on the ethics of human clinical gene transfer (GT).1Clinigene/Consert The ethics of human clinical gene transfer: en route to ethical recommendations for gene transfer clinical trials Program of meeting held 2–3 April 2007, Geneva, Switzerland <.http://www.clinigene.eu/file_past/23.pdf>Google Scholar The ethics of human GT research has much in common with that of research involving other new biotechnologies,2King NMP Defining and describing benefit appropriately in clinical trials.J Law Med Ethics. 2000; 28: 332-343Crossref PubMed Scopus (199) Google Scholar,3Kimmelman J Recent developments in gene transfer: risk and ethics.BMJ. 2005; 330: 79-82Crossref PubMed Scopus (70) Google Scholar but it has always raised special ethical concerns4Churchill LR Collins ML King NMP Pemberton S Wailoo K Genetic research as therapy: implications of "gene therapy" for informed consent.J Law Med Ethics. 1998; 26: 38-47Crossref PubMed Scopus (47) Google Scholar that arise from the history of the field, public perceptions of genetic manipulation, and the novelty of GT technologies. Continual reexamination of ethical issues as GT science and technology evolve is thus an important task for both the scientific community and its oversight bodies, to be undertaken independently. Thus, the aim of the meeting was to identify special ethical issues that might arise in GT research, with the ultimate goal of generating recommendations from the field. The meeting was organized around three primary questions: (1) why should GT research be pursued?, (2) when is GT research ethically appropriate?, and (3) how should GT research be designed and conducted? This Commentary presents a synthesis of the issues addressed, in light of the current state of translational GT science, from vector design to animal models to first-in-human trials and beyond. GT uses a diversity of methods and encompasses a variety of diseases, ranging from rare inherited disorders, cancer, cardiovascular diseases, and neurodegenerative disorders to metabolic disorders and diabetes. Nonetheless, each application of GT must identify several prerequisite factors: (1) the nature of the putative target cell, (2) the gene of interest, or the therapeutic gene, (3) the ideal route of GT, such as direct in vivo application or by implanting autologous or heterologous genetically modified cells ex vivo, and (4) whether transient or stable expression of the gene is required, and, if stable, whether additional levels of regulation are required. These assessments involve a high degree of uncertainty, and what we think we know can change over time. For example, type 2 adeno-associated viral (AAV) vectors were initially regarded as nonintegrative and nonpathogenic, but there is new evidence that they are integrating.5Donsante A Miller DG Li Y Vogler C Brunt EM Russell DW et al.AAV vector integration sites in mouse hepatocellular carcinoma.Science. 2007; 317: 477Crossref PubMed Scopus (471) Google Scholar The problem of immunogenicity and effective strategies for avoiding immunotoxicity also deserve attention.6Mingozzi F Hasbrouck NC Basner-Tschakarjan E Edmonson SA Hui DJ Sabatino DE et al.Modulation of tolerance to the transgene product in a nonhuman primate model of AAV-mediated gene transfer to liver.Blood. 2007; 110: 2334-2341Crossref PubMed Scopus (187) Google Scholar How should we respond to this high degree of uncertainty in basic GT technology? First, optimism should always be tempered with caution. Second, improved sharing of expertise across disciplines and increased integration of research along translational pathways are essential to reduce uncertainty, to conduct research efficiently with the minimum necessary number of subjects, and to make genuine progress toward safe, effective, and ethically sound GT research and, ultimately, gene therapy. Elements of GT trials that influence ethical assessment and oversight are: (1) novelty and complexity, (2) limited experience and evidence of potential risks and benefits for the subject, (3) optimistic and pessimistic public views of genetic modification, and (4) the fact that risks may extend beyond the participants. Some have argued that issues unique to GT include long-term follow-up and unintentional and deliberate germline modification. Whether these issues are indeed unique to GT or simply appear more controversial in this research is a key question. Review bodies—institutional review boards (IRBs) in the United States, ERBs or REBs in the EU countries—may lack experience and expertise in the science of GT research, which increases the likelihood that oversight could fail to address the most salient issues while giving excessive attention to less important aspects of GT research protocols. This issue is not unique to GT research but is perhaps a good example of the need to improve IRB/ERB education and coordination in multicenter trials, especially those that are multinational.7Churchill LR Nelson DK Henderson GE King NMP Davis AM Leahey E et al.Assessing benefits in clinical research: why diversity in benefit assessment can be risky.IRB. 2003; 25: 1-8Crossref PubMed Scopus (36) Google Scholar,8King NMP RAC oversight of gene transfer research: a model worth extending?.J Law Med Ethics. 2002; 30: 381-389Crossref PubMed Scopus (32) Google Scholar For example, one large GT study was set up in nine countries to enroll 250 patient-subjects. Despite the similarity of the application package in each country, there were major differences in each country's approval process, which was anticipated to take a maximum of 180 days per country. The time for response ranged from 3 months in some countries to 14 months in the Czech Republic, where, because expertise in GT review was lacking, ultimately the trial could not go forward. The principal obstacle to its approval was not the assessment of the trial design itself but biological safety, environmental protection, and ethical issues. Increasing the expertise of the ERB/IRB and establishing ways to facilitate access to thorough and validated information are necessary measures to ensure timely review based on good science and appropriate attention to subject safety and ethics. Clinigene-NoE has compiled a general database on vectors and their use (http://www.clinigene.eu/gtref). Collection and sharing of unpublished preclinical data from phase I trials with high-risk medicines would be helpful, as would collection and sharing of data (safety, toxicology) from trials that have been stopped. However, economic and competitive considerations remain, as well as the reluctance of both industry and journal editors to publish negative results, which is yet another barrier to data sharing. It is widely acknowledged that when investigators provide oversight bodies with clear and thorough information on which to base their review, especially review of novel and rapidly changing technologies like GT, those oversight bodies become better educated in the science and are better able to evaluate the ethical and scientific justification for a given research protocol. This is especially important when first-in-human studies are proposed; study rationale should describe the fundamental background of animal/preclinical studies within the framework of the literature and provide arguments for moving into the clinic. Europe lacks a Recombinant DNA Advisory Committee (RAC)–like structure with independence and accountability. In the United States the RAC has had long-term success in leading the field and improving the quality of GT research, by means of publicly accessible data. This includes the development of GeMCRIS, a comprehensive federal database of clinical GT trials (http://www4.od.nih.gov/oba/RAC/GeMCRIS/GeMCRIS.htm), besides the general database at http://www.clinicaltrials.gov. A first step in the EU may be exchange of GT trial reviews among national ERBs. A further step, ideally, would consist of consolidating evaluation data in common. More centralized expert review in the EU countries would be more efficient and less time-consuming, and should also circumvent the relative lack of a critical mass of expertise in individual countries. The RAC has also developed a Web-based guidance document on informed consent in GT research,9National Institutes of Health NIH Guidance on Informed Consent for Gene Transfer Research <.http://www4.od.nih.gov/oba/RAC/ic>Google Scholar which is intended to assist investigators and IRBs. This guidance provides suggested language about the purpose of the study, potential direct and societal benefits, and surrogate end points. Consent forms must be explicit in explaining the difference between research and treatment, and in distinguishing hope from reasonable expectations for patient-subjects. Recommendations include presentation of potential benefit to society as the sole or primary goal of clinical research. In 2006 Clinigene-NoE created a centralized EU database designed to include all clinical GT trials. It contains minimal but essential information, and it omits confidential data. The UK Gene Transfer Advisory Committee (GTAC) database, Orphanet, Euregenethy and the Swiss database are included so far. There are more than 600 entries with links to publications and a database search engine. In addition, there are plans to add nonconfidential trial data from EU projects registered in the European Medicines Agency's EudraCT database since 1 May 2004. The clinical approach to GT starts with the disease, defines the target tissue, and then develops the gene delivery system. The basic research approach, however, is the inverse of that strategy. Diseases are selected on the basis of their favorable characteristics for GT. Pertinent questions include whether a therapeutic gene is available and whether its function is clear, whether the target cell is accessible, whether simple gene regulation is possible, whether it is possible to treat a sufficient number of cells to produce a therapeutic effect, and whether ectopic expression of the gene is harmful. According to specific features of the disease of interest, an "ideal" combination of GT vector, transgene, method, and route of application is chosen that is best predicted to lead to effective treatment. If clinical translation is thought to be urgently needed, less-than-ideal systems might represent a practical and realistic compromise. Basic scientists are generally viewed as more restrictive and deliberate in their approach to developing a line of research than are clinicians: "Basic scientists want to know; clinicians want it to work." This tension is ubiquitous in GT and can be seen in both the ethics and the science. Early thinking about "gene therapy" envisioned repair of disabling mutated genes—so-called "genome editing" or "gene surgery"—rather than the current paradigm of adding multiple copies of the unmutated transgene. These more precise techniques are still unrealized. The potential risks and benefits that can follow introduction of multiple copies of a particular vector–transgene combination vary depending on whether an in vivo or ex vivo approach is used, and whether in vivo GT is targeted or systemic. Bearing these concerns in mind, the role of animal studies should be reexamined, especially the importance of large animal models (LAMs) and nonhuman primate (NHP) models. Animal models may be quite different in terms of toxicity of viral vector-mediated gene expression as well as in terms of immune reactions to different vectors, so NHPs may be needed even when murine data are ample. Biodistribution studies are easier and more informative in LAMs, and long-term follow-up may provide important information needed before human studies are begun. For example, now that what were thought to be nonintegrating viral vectors have been shown to integrate, additional experimental attention and follow-up will be required. If LAM studies can provide information over the long term, the principle that some questions can be answered only in human studies may require reconsideration in some cases. It would therefore be useful to develop a list of criteria for when LAMs are needed, including a catalog of reasons to use NHP models, because the use of NHPs in research must be thoroughly justified. Small-animal results are not always confirmed in LAM and NHP studies, and the latter models sometimes do not adequately represent human disease. Human xenografts have been elegantly used in immunodeficient mice in some instances.10Karnoub AE Dash AB Vo AP Sullivan A Brooks MW Bell GW et al.Mesenchymal stem cells within tumour stroma promote breast cancer metastasis.Nature. 2007; 449: 557-563Crossref PubMed Scopus (2513) Google Scholar Nevertheless, that animal studies are not fully predictive of the clinical fate of products is by no means unique to GT, as demonstrated by the serious adverse event with TGN 1412 monoclonal antibody.11Goodyear MD Further lessons from the TGN1412 tragedy.BMJ. 2006; 333: 270-271Crossref PubMed Scopus (27) Google Scholar Progress has also been made in the in vitro assessment of vector toxicity using organ cultures12Kolodkin-Gal D Zamir G Edden Y Pikarsky E Pikarsky A Haim H et al.HSV-1 preferentially targets human colon carcinoma: the role of extra-cellular matrix.J Virol. 2015; (e-pub ahead of print 31 October 2007)PubMed Google Scholar and human cell lines. There is an urgent need for preclinical assays predictive of adverse immune reactions, because these are by far the most acute and life-threatening adverse events recorded so far in first-in-human trials. Of course, even great improvements in in vitro and animal studies will not perfectly model effects in humans; uncertainties and risks of harm will always accompany clinical GT research, as with other biotherapies and xenobiotics. In the first human GT trials, primary concerns were the safety not only of the patient-subject but also of contacts, care providers, the general population, and future generations. These concerns persist, but they have changed and diminished over time. The 2007 International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use conference in Rotterdam was entirely dedicated to vector shedding, and a report has just been released.13International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use Rotterdam, The Netherlands, 30 October–1 November 2007.http://www.ich.org/LOB/media/MEDIA4360.pdfGoogle Scholar Published data on shedding of viral vectors during clinical GT trials are surprisingly limited.14Schenk-Braat EA van Mierlo MM Wagemaker G Bangma CH Kaptein LC An inventory of shedding data from clinical gene therapy trials.J Gene Med. 2007; 9: 910-921Crossref PubMed Scopus (54) Google Scholar However, shedding does not seem to pose a problem thus far, although new vectors could require more stringent monitoring. Insertional mutagenesis poses much greater risks of adverse events, including in the long term. In the severe combined immunodeficiency disorder (SCID) clinical retroviral vector trial in Paris, four cases of leukemia occurred in 8 patient-subjects who benefited from GT, with insertion sites near the LMO2 gene (as well as some other genes). Despite only minor differences between the two protocols, there had been no case of leukemia in the British trial until the recent report of a case in this trial with insertion at the LMO2 site as well.15Hacein-Bey-Abina S Von Kalle C Schmidt M McCormack MP Wulffraat N Leboulch P et al.LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1.Science. 2003; 302: 415-419Crossref PubMed Scopus (3001) Google Scholar,16Deichmann A Hacein-Bey-Abina S Schmidt M Garrigue A Brugman MH Hu J et al.Vector integration is nonrandom and clustered and influences the fate of lymphopoiesis in SCID-X1 gene therapy.J Clin Invest. 2007; 117: 2225-2232Crossref PubMed Scopus (219) Google Scholar,17Clinigene-NoE X-Linked-SCIDs SAE <.http://www.clinigene.eu/SCIDs/index.lasso>Google Scholar Cases of insertional mutagenesis in animals were reported shortly before the French SCID trial incidents were published. Since then, observations of common insertion sites and clonal dominance in both monkey and mouse have been published. There is now general awareness that untargeted insertion may lead to clonal dominance of transduced cells, as well as to the induction of malignancies; this is corroborated by a chronic granulomatous disease clinical trial in which one patient was recently diagnosed with myelodysplastic syndrome (Manuel Grez, oral presentation at the November 2007 meeting of the European Society of Gene and Cell Therapy). Recent development of improved safer vectors (vectors with insulators, termination sequences, and vectors with improved targeting) may reduce this risk, but data based on better in vitro and in vivo assays are needed. However, latency periods for cancer can be very long, and we need to be able to predict oncogenicity. Therefore, the "ethics agenda" of experimental clinical GT should include new technologies of high-throughput DNA screening to make it possible to inventory insertions, searching for common insertion sites. Once this evidence is available, the ethics question that arises is: How great a likelihood of insertional mutagenesis must there be to prevent a trial from going forward? Transparency and good societal understanding are important to answering this question; long-term follow-up will be necessary to adequately assess risk. Inadvertent germline effects may be another consequence of in vivo methods that disseminate vector and transgene throughout the body. Germline modification is currently not permitted in Europe; when the possibility of germline effects is not excluded, the experimental GT will be banned. However, this may constitute an over-reaction to a very low likelihood of harm. Detection of vector in semen may or may not indicate integration of transgene in spermatozoa, according to fractionation studies. Thus far, for example, AAV2 vector has been transiently detected in semen of subjects in hemophilia studies but not in spermatozoa.18Manno CS Pierce GF Arruda VR Glader B Ragni M Rasko JJ et al.Successful transduction of liver in hemophilia by AAV-Factor IX and limitations imposed by the host immune response.Nat Med. 2006; 12 (Supplementary Table 3 ): 342-347Crossref PubMed Scopus (1568) Google Scholar Little information is available about inadvertent effects on oocytes. These findings, even though far from definitive, have raised concerns that inadvertent transmission of germline changes—which could be either corrections or mutations—might accompany the most promising somatic cell GT.19King NMP Accident and desire: inadvertent germline effects in clinical research.Hastings Cent Rep. 2003; 33: 23-30Crossref PubMed Scopus (15) Google Scholar,20Frankel MS Chapman AR Human Inheritable Genetic Modifications: Assessing Scientific, Ethical, Religious, and Policy Issues. American Association for the Advancement of Science, Washington, DC2015http://www.aaas.org/spp/sfrl/projects/germline/report.pdfGoogle Scholar A low likelihood of harmful germline effects is tolerated for teratogenic products that are considered standard treatments for both serious and non–life-threatening conditions, such as cancer,21de Rooij DG van de Kant HJ Dol R Wagemaker G van Buul PP van Duijn-Goedhart A et al.Long-term effects of irradiation before adulthood on reproductive function in the male rhesus monkey.Biol Reprod. 2002; 66: 486-494Crossref PubMed Scopus (58) Google Scholar leprosy, and even cystic acne. However, in experimental human GT the possibility of inadvertent germline modification seems to be regarded as far more significant. Because the likelihood of germline modification is greatest at the earliest stages of human development, children with life-threatening diseases are the primary focus of ethical debate about inadvertent germline effects. The low likelihood that children with diseases for which GT is an appropriate intervention could pass on an inadvertently modified gene to their offspring later in life may be a very minor issue, not only for their parents but also from a scientific perspective. In the case of GT it should be easier to detect the presence of specific sequences in offspring, compared with random genomic events following chemotherapy and irradiation. This is why some investigators think that a threshold of acceptably low risk could be set, using data from preclinical animal studies and semen sampling from male humans in GT trials. Improved vector-targeting technology could also ameliorate the risks. If concerns persist, or if inadvertent germline effects are seen in preclinical research, cryopreservation of gametes before the genetic intervention could be done routinely for GT research subjects, as is done in clinical oncology. Semen was banked before the GT procedure in all the men enrolled in AAV GT trials for hemophilia, even though all subjects showed clearing of AAV in semen over time. It is important to ensure that concerns about inadvertent germline effects do not preclude promising research from going forward, if there are ethically acceptable ways to manage those risks, and especially if more can be learned about the risks themselves. In this respect, the US treatment of inadvertent germline effects is somewhat more flexible and permissive than the EU regulations, and harmonization is desirable. There has long been consensus support for a ban on intentional germline transmission. But many patient advocacy groups aren't at all worried about this and have nothing against, for instance, eradicating hemophilia in their descendants. Patient organizations may even exert pressure in favor of this in the future. If it were possible to deliberately modify the germline, which would require perfect gene surgery without nonhomologous background integration, would it be unethical to do so?22Juengst ET Grankvist H Ethical issues in human gene transfer: a historical overview. In: Ashcroft, RE, Dawson, A, Draper, H and McMillan JR (eds). Principles of Health Care Ethics. 2nd edn. Wiley, New York, 2007Google Scholar,23Dresser R Genetic modification of preimplantation embryos: toward adequate human research policies.Milbank Q. 2004; 82: 195-214Crossref PubMed Scopus (13) Google Scholar Those who argue that germline modification is not unethical ask, "Why not treat the whole family, including descendants?" Others point out that germline transmission of an inherited disorder or other undesirable trait can already be prevented simply via preimplantation genetic diagnosis (PGD) and selection and implantation of healthy embryos. Proponents of intentional germline modification (IGM) respond that preventing the manifestation of disease in a patient is not the same as preventing the patient's birth following PGD. However, given that IGM would require PGD first, why should the correction of mutations in an embryo take priority over the implantation of an already existing mutation-free embryo? Perhaps the proponents of IGM are imagining successful gene therapy of every cell in an embryo in utero. If such a combined somatic cell and germline correction were possible, it might be the only ethically acceptable use of IGM. Nevertheless, achieving the modification of every germ cell in an embryo in utero may be something out of science fiction, because somatic cells are the targets. In utero GT has been widely considered and discussed, and there already is evidence for successful in utero cell therapy, such as bone marrow transplantation (BMT). Of course, in utero GT would be far more likely to have inadvertent germline effects, and such research raises distinct ethical issues: the risks of harm posed to the pregnant woman, the problem that offspring will then become research subjects without their consent, and the possibility of unpredictable serious harms.24Dresser R Designing babies: human research issues.IRB. 2004; 26: 1-8Crossref PubMed Scopus (13) Google Scholar,25National Institutes of Health Recombinant DNA Advisory Committee Prenatal Gene Transfer: Scientific, Medical, and Ethical Issues.http://www4.od.nih.gov/oba/rac/gtpcreport.pdfDate: 2000Google Scholar Recent advances in stem cell research, with successful reprogramming of autologous adult stem cells, seem to hold promise for the development of effective cell therapies.26Hanna J Wernig M Markoulaki S Sun C-W Meissner A Cassady JP et al.Treatment of sickle-cell anemia mouse model with iPS cells generated from autologous skin.Sci Express. 2007; ([online], 6 December)Google Scholar,27Boer GJ Ethical issues in the approaches of restorative therapies for Parkinson's disease.in: Olanow W. Brundin P Restorative Therapies in Parkinson's Disease. Kluwer, New York2006: 13-49Crossref Scopus (4) Google Scholar This scientific development could significantly alleviate ethical concerns about stem cell research and expected therapeutic developments and, in addition, could significantly alter the comparative risks, potential benefits, and ethical issues arising from in utero GT. How serious should a disorder or condition be before GT is considered an appropriate potential treatment? In the United States, trials for erectile dysfunction and hyperactive bladder function have been approved by the RAC. The history of GT has led to general agreement that it ought to be restricted to treatment of serious conditions. However, it seems inevitable that as GT becomes safer, and especially if it shows more effectiveness in disease treatment, it will be tried in less serious conditions. It also seems inevitable that it will be tried in nondisease states. For example, potential treatments being studied for muscular dystrophies or for cachexia could be used to develop antiaging protocols or to enhance athletic ability. This possibility is not specific to GT, but because GT is one focus of the ethical debate, the likelihood of enhancement applications should be considered.28Parens E Enhancing Human Traits: Ethical and Social Implications. Georgetown University Press, Washington, DC1998Google Scholar,29Mehlman MJ Cognition-enhancing drugs.Milbank Q. 2004; 82: 483-506Crossref PubMed Scopus (73) Google Scholar Although we are very far from being able to perform genetic enhancements of human physical and mental performance, some pharmacological, neurological, and surgical enhancements are well established. Assessing and balancing the risks of harm with the potential benefits of a GT intervention can be difficult when treatment is foreseen. When the benefit would be an enhancement, it is even more difficult to compare the value of enhancing an already normal state to the same risks of harm—for example, immunotoxicity, which can be significant in GT. The ethical discussion of future possibilities in GT returns us to several key ethical questions that have yet to be addressed in ordinary, potentially therapeutic GT research. The question whether GT should be tried only in diseases for which there are no good treatment options, or whether it should be tried early to maximize the likelihood of being able to discern evidence of effectiveness, helps to illustrate how the distinction between research and treatment is increasingly becoming blurred in clinical trials—especially in pediatric research. Selection of subjects for GT research includes questions about whether to choose the healthiest or the sickest. It is necessary to design studies to gather the maximum amount of good data at the lowest possible risk to the subjects. When children face hereditary disorders with cellular dysfunction and death at a young age, the question arises as to when an experimental intervention should be tried. In the case of Duchenne muscular dystrophy, parents and patient associations are eager to move research to the clinic, and often offer themselves or their children as experimental subjects. Their willingness to participate in hope of effective treatment is very high, even though a research intervention is not a treatment yet. Such hope will always exist, and scientists would be naive to ignore this. With the initial success of the GT intervention in X-linked SCID patients, and the subsequent identification of delayed adverse events, X-linked SCID is a first example of balancing real risks of harm and real potential benefits in GT research. The treated child subjects, despite their leukemia, had 3 to 5 years of good life. However, potential benefits from GT may have to be reconsidered when allogeneic BMT is available. In one center, investigators reported that maternal donation before a child is 3.5 months old offers a 96% chance of survival; transplant success drops to 50% at 6 months of age. If diagnosis could be made this early, and such dramatic data broadly reproduced in other centers, haplo-identical BMT could be more effective than GT. Perhaps GT should be considered only when there is neither a related donor nor a good-match unrelated donor. The start of the first clinical trial of potentially therapeutic GT, in lymphocytes of adenosine deaminase (ADA)–deficient patients in 1990, was delayed because transduction of human stem cells was still inefficient, no clinical experience with retroviral vectors had been reported, the dise
Referência(s)