RAC Oversight of Gene Transfer Research: A Model Worth Extending?

The acronym also serves as a homonym for an instrument of medieval torture, a meaning that is wryly invoked both by investigators and sponsors who appear before it and by members after long meetings.

Others have also addressed the question whether the Recombinant DNA Advisory Committe (RAC) oversight model, or aspects of it, should be more broadly applied. See, e.g., Walters L. , “The Oversight of Gene Transfer Research,” Kennedy Institute of Ethics Journal, 10 (2000): 171–74;.

Rainsbury J. , “Biotechnology on the RAC — FDA/NIH Regulation of Human Gene Therapy,” Food and Drug Law Journal, 55 (2000): 575–600.

The consolidated and harmonized Common Rule was published at 56 Fed. Reg. 28,012 (June 18, 1991); the codification most familiar to institutional review boards (IRBs) and others involved in research oversight are the Department of Health and Human Services regulations at 45 C.F.R. Part 46 (2001).

The Food and Drug Administration (FDA) regulations corresponding to the Common Rule appear at 21 C.F.R. Parts 50 and 56. Key FDA drug development regulations are also found at Parts 312, 314, and elsewhere. The FDA's human subjects regulations are substantially similar to the Common Rule, but in addition, the FDA has a very hands-on relationship with research sponsors in the long process of drug development. See, e.g., the overview provided in N. Plant, “Adequate Well-Controlled Clinical Trials: Reopening the Black Box,” Widener Law Symposium Journal, 1 (1996): 267–97.

Clinical gene transfer research (GTR) may also be subject to additional local review in several forms. If it is cancer research (as most of it is), there may be a local oncology protocol review committee. If there is a general clinical research center affiliated with the institution at which the research will take place, the general clinical research center's review committee must also review the research, if any part of it will take place in the general clinical research center. And some institutions have established their own human gene transfer review committees.

The National Institutes of Health's (NIH) guidelines are available on the Office of Biotechnology Activities website, at <http://www4.od.nih.gov/oba/rac/guidelines/guidelines.html> (last visited September 17, 2002). They were first published in 1976, and have been amended many times.

See, e.g., Rogers M. , Biohazard (New York: Knopf, 1977);

Berg P. , “Asilomar Conference on Recombinant DNA Molecules,” Science, 188 (1975): 991–94;.

Swazey J.P. , “Risks and Benefits, Rights and Responsibilities: A History of the Recombinant DNA Research Controversy,” Southern California Law Review 51 (1978): 1019–67.

Statement of Walters LeRoy Dr. , Director, Kennedy Institute of Ethics, before the Senate Subcommittee on Public Health, Committee on Health, Education, Labor and Pensions, Feb. 2,2000.

The Points to Consider in the Design and Submission of Protocols for the Transfer of Recombinant DNA Molecules into One or More Human Research Participants is Appendix M of the NIH guidelines (supra note 6). Appendix M was added to the NIH guidelines in 1990.

Rainsbury , supra note 2; National Reference Center for Bioethics Literature, Scope Note 24, Human Gene Therapy, at http://www.georgetown.edu/research/nrcbl/scopenotes/sn24.html> (last updated March 2002).

See Walters Palmer , Id.; Rainsbury, supra note 2.

The Office of Biotechnology Activities (OBA) is the NIH office that staffs the RAC. It was formerly called the Office of Recombinant DNA Activities; in recent years it has been expanded and also staffs two other federal advisory committees, on xenotransplantation and genetic testing.

The history of this debate is detailed in the sources cited supra notes 8 and 8.

Rainsbury , supra note 2. Eighteen-year-old Jesse Gelsinger was the first subject to die as a direct result of gene transfer.

See Stolberg S.G. , “The Biotech Death of Jesse Gelsinger,” New York Times Magazine, Nov. 28, 1999, at 137–51.

That is, it functions something like a single IRB established to provide overarching guidance to local IRBs for a multicenter study. In this respect, RAC is a kind of precursor to the current interest in centralizing some aspects of the oversight of multicenter trials.

See Walters , supra note 8.

Id.

Statement of Patterson Amy Dr. , Executive Director, OBA, before the Senate Subcommittee on Public Health, Committee on Health, Education, Labor and Pensions, Feb. 2, 2000, at <http://www4.od.nih.gov/oba/rac/documents1.htm> (last visited September 17, 2002).

in recent studies of infants and young children with one form of severe combined immunodeficiency ( Hacein-Bey-Abina S. , “Sustained Correction of X-Linked Severe Combined Immunodeficiency by ex Vivo Gene Therapy” N. Engl. J. Med., 346 (2002): 1185–93); and in a few studies of squamous cell carcinoma of the head and neck (the only GTR that has as yet reached Phase III trials).

See also Churchill L.R. , “Genetic Research as Therapy: Implications of ‘Gene Therapy’ for Informed Consent,” Journal of Law, Medicine & Ethics, 26, no. 1 (1998): 38–47.

Most GTR uses delivery vectors to introduce the genetic material, though some studies use naked DNA. Vectors can be made of altered viruses or of other materials, like fat particles. Each vector used in GTR therefore needs to be individually standardized. Because most vectors are viral, there are so many in use, and each is so different, this standardization is a considerable undertaking.

See discussions of vector standardization guidelines for retroviruses and adenoviruses in RAC meeting minutes, December 2000, at <http://www4.od.nih.gov/oba/RAC/meeting.html> (last visited September 17, 2002). The National Gene Vector Laboratories are instrumental in developing standards, at <www.ngvl.org> (last visited September 17, 2002).

See extensive discussion in the minutes of the December 1999

RAC meeting, at <http://www4.od.nih.gov/oba/RAC/meeting.html> (last visited September 17, 2002).

Id. at comments of Pilaro Anne Dr. , FDA, in RAC meeting minutes, December 1999.

Stephenson N. , Zodiac (New York: Bantam Books, 1995).

Sources of concern range from the original Asilomar moratorium on recombinant DNA research to discussions of germ-line interventions, both deliberate and inadvertent. See, e.g., the many sources cited in the Human Gene Therapy Scope Note, supra note 10.

Possible explanations for this shift are structural, financial, scientific, and social. Cancer is a problem that affects many people and is high in public consciousness. The apparatus of oncology research is large, prominent, and experienced in attracting research funding and managing clinical trials. And burgeoning knowledge in areas relating to cancer control mechanisms, such as the role of the immune system and of various genetic mutations, has helped lead to many forms of GTR in cancer: Gene-based vaccines, the introduction into tumors of genes that can be killed by antiviral agents, and studies using tumor suppressor genes are just a few examples. As of May 31, 2002.

there were 332 cancer trials in OBA's database, out of 480 total clinical trials of interventions considered potentially therapeutic. This total excludes marking studies and studies using healthy normal subjects. If all human studies are included, the percentage of cancer studies is about 63 percent of the total; data at <http://www4.od.nih.gov/oba/rac/documents1.htm> (last visited September 17, 2002) (enumerated in the last two pages of the Protocol List).

For some insight into the perspective of oncology research, see Miller M. , “Phase I Cancer Trials: A Collusion of Misunderstanding,” Hastings Center Report, 30, no. 4 (2000): 34–43.

Focusing GTR on monogenic diseases could be viewed as a vital component of the ongoing effort to develop effective interventions for patients with orphan diseases. Moving GTR to more common diseases and disorders with multifactorial causes, like cancer, HIV infection, coronary artery disease, or diabetes clearly makes a promising technology more widely available. At the same time, however, it does two additional things: It greatly increases the investment of money and research infrastructure for GTR; and it helps to focus public and policymaking attention on research involving expensive, cutting-edge technologies as a primary solution for problems that can also be addressed by attention to prevention, public and environmental health and health education, lifestyle, and the complex relationships among genes, environment, and expression.

See, e.g., RAC meeting minutes, March 1999, December 2001, and March 2002

for discussions of inadvertent germ-line effects, at <http://www4.od.nih.gov/oba/RAC/meeting.html> (last visited September 17, 2002).

See also RAC, NIH report, Prenatal Gene Transfer: Scientific, Medical and Ethical Issues, NIH Pub. No. 00-4720 (released in January 2000 following the January 1999 Gene Therapy Policy Conference on the topic), available at <http://www4.od.nih.gov/oba/RAC/meeting.html> at “January 7–8, 1999, Prenatal Gene Therapy: Scientific, Medical, and Ethical Issues-Full Report.”

National Institutes of Health, “Financial Conflicts of Interest and Research Objectivity: Issues for Investigators and Institutional Review Boards,” #OD-00-040, June 2000, at <http://grants2.nih.gov/grants/guide/notice-files/NOT-OD-00-040.html> (last visited September 17, 2002);

Office for Human Research Protections, “Draft Interim Guidance: Financial Relationships in Clinical Research: Issues for Institutions, Clinical Investigators, and IRBs to Consider When Dealing With Issues of Financial Interests and Human Subject Protection,” at <http://ohrp.osophs.dhhs.gov/humansubjects/finreltn/finmain.htm> (last visited September 17, 2002);.

Cho M. , “Policies on Faculty Conflicts of Interest at U.S. Universities,” JAMA, 284 (2000): 2209–14;.

McCrary S.V. , “A National Survey of Policies on Disclosure of Conflicts of Interest in Biomedical Research,” N. Engl. J. Med., 343 (2000): 1621–25;.

Morin K. , “Managing Conflicts of Interest in the Conduct of Clinical Trials,” JAMA, 287 (2002): 78–84.

Wilson James Dr. , then-director of the University of Pennsylvania's Institute for Human Gene Therapy, where the ornithine transcarbamylase deficiency trial took place, and a co-investigator, had a financial interest in a company he founded, Genovo Inc.

which had a stake in the success of the trial's liver-directed gene transfer methodology, developed by his laboratory at Penn. See Nelson D. Weiss R. , “Hasty Decisions in the Race to a Cure?,” Washington Post, Sunday, Nov. 1999; at A01, for a review of the potential financial conflicts of interest and how they were viewed before Mr. Gelsinger's death.

See supra notes 7 and 7.

See, e.g., Report of the RAC, NIH, supra note 29, especially speaker paper, Sugarman J. , “Ethical Questions Related to the Prospect of in utero Gene Transfer Experiments,” at 71–75.

Other entities, such as oncology protocol review committees, general clinical research centers, institutional biosafety committees, local human gene transfer committees, and review committees convened by sponsors, may have scientific but not ethical expertise, or may not be disposed to ask certain scientific questions.

Indeed, the three determinations are nothing more than my own phase I-specific gloss on the federal Common Rule's first two criteria for IRB approval of research, codified in Department of Health and Human Services regulations at 45 CFR 46.111(a), which addresses the IRB's overall responsibility to approve only research that minimizes risks and that demonstrates an appropriate balance between the risks of harm and the benefits from anticipated gains in knowledge.

For example, the nonexistence of a “knockout mouse” model, genetically altered to knock out a gene of interest and therefore mimic a particular human disorder, is not the same as the limitations of the information that can be gained from knockout mice about human disorders; nor is it the same as the knowledge that there is a knockout mouse available but that purchasing the right to use it from the patent holder is costly and will delay the start of a Phase I study in humans.

Choosing the right first subjects posed unexpected challenges in the ornithine transcarbamylase deficiency trial. At least three different possible subject populations were apparently discussed at various times: severely affected newborn infants currently in crisis from an excess of ammonia; severely affected but currently stable infants and young children awaiting liver transplant; and partially affected adults (usually men and women with a late onset form of the enzyme deficiency;

Gelsinger Mr. , who had been diagnosed at 3 years of age, apparently had a spontaneous mutation rather than an inherited form of the disorder). One important consideration in choosing subjects is the ethical preference for first recruiting adults who can make their own decisions about research participation. This consideration is especially powerful in first human trials because of the extreme uncertainty about potential efficacy in an intervention as yet untested in humans, and because of the design and goals of Phase I trials: dose escalation studies designed to elicit and examine toxicities, beginning at low doses unlikely to provide benefit to subjects even if the intervention worked perfectly (which it usually doesn't). Added to this was the concern that the parents of newborns in hyperammonemic crisis would be emotionally and informationally stressed, having just learned of the disorder because their child was gravely ill from it, and being asked to decide quickly about an unprecedented but emergent intervention. Aside from decision-making and consent issues and regulatory limitations on research with children (45 C.F.R. Part 46 Subpart D), all of which favored recruitment of adult subjects first, minimizing harm to subjects appeared to favor enrollment of newborns in crisis, since gene transfer was believed to pose low risks of harm and since these newborns were already seriously ill and already receiving maximal but suboptimal therapy. (Moreover, investigators and regulators alike were greatly tempted to reason that if this gene transfer worked, these babies could be saved.) In contrast, maximizing generalizable knowledge favored recruitment of partially affected, currently stable subjects. It would be very hard to determine, in newborns in acute crisis, which of any effects seen, whether for good or ill, resulted from the disease, the standard treatments, or the experimental intervention.

King N.M.P. , “Defining and Describing Benefit Appropriately in Clinical Trials,” Journal of Law, Medicine & Ethics, 28, no. 4 (2000): 332–43.

See Rainsbury , supra note 2;

statement of Walters LeRoy Dr. , supra note 8.

Most recently, in response to questions raised about GTR oversight after Mr. Gelsinger's death, the Acting Director of NIH convened a working group from the standing Advisory Committee to the Director to critically examine the RAC's role and functions.

See “Advisory Committee to the Director, Working Group on NIH Oversight of Clinical Gene Transfer Research, Enhancing the Protection of Human Subjects in Gene Transfer Research at the National Institutes of Health,” July 12, 2000, at <http://www.nih.gov/about/director/07122000.htm> (last visited September 17, 2002). After the Advisory Committee to the Director Working Group issued its report, the RAC received authorization to expand its membership from fifteen, to add new relevant expertise in areas such as public policy and statistics.

Larry R. Churchill Arlene M. Davis Daniel K. Nelson Benjamin S. Wilfond . The project also includes interviews with GTR investigators, study coordinators, and subjects, as well as review of nearly all consent forms and Points to Consider responses on file with OBA. Co-investigators Churchill, Nelson, and Wilfond conducted the IRB interviews between December 2000 and November 2001. The data presented here are preliminary results only.

Perhaps most common is some IRBs' reluctance to mention autopsy in the consent form, though Appendix M requires investigators to include in the consent form the information that permission for an autopsy of the subject will be requested from the next of kin at the time of the subject's death for any reason, in order to learn more about the long-term effects of GTR. Appendix M's discussion of autopsy requests thus addresses most IRBs' concerns, which include failure to appreciate the need for the information, worry that mentioning death might unduly alarm sick subjects, and fear that an autopsy request might be mistaken for an autopsy requirement; yet unless they read Appendix M, IRBs cannot discern this. Once the specific requirements that Appendix M places on investigators are drawn to the IRB's attention by OBA or the RAC, IRBs are, in my experience, very receptive to making suggested changes in consent forms.

Correspondence from OBA is sent to the principal investigator listed in OBA's files, and copied to the principal investigator's IRB. This means that in multicenter studies, the only IRB receiving correspondence is the IRB at the primary site — and even that IRB does not receive the attachments included in the letter to the principal investigator at the site. IRBs reviewing GTR thus must take additional steps in order to be most fully informed.

NIH, “Financial Conflicts of Interest and Research Objectivity,” supra note 30;

NIH, “Further Guidance on Data and Safety Monitoring for Phase I and Phase II Trials,” #OD-00-038, June 2000, at <http://grants2.nih.gov/grants/guide/notice-files/NOT-OD-00-038.html> (last visited September 17, 2002);

NIH, “Required Education in the Protection of Human Research Participants,” #OD-00-039, June 2000, at <http://grants2.nih.gov/grants/guide/notice-files/NOT-OD-00-039.html> (last visited September 17, 2002).

65 Fed. Reg. 60328–60332 (Oct. 10, 2000).

It is important that IRBs reviewing GTR become accustomed to requesting copies of Appendix M responses — from the principal investigator, OBA, or the IBC. Closer relationships between IRBs and IBCs are desirable as well, because IBCs usually do know about Appendix M, and may routinely review Appendix M responses from principal investigators — but many are unaccustomed to reviewing clinical research, and could learn much from the IRB.

Between 20 and 30 percent of protocols are selected for full public review and discussion, and that percentage is dropping as the number of protocols submitted to OBA for RAC review continues to grow.

Stolberg S.G. , “Agency Failed to Monitor Patients in Gene Research,” New York Times, Feb. 2, 2000, at 19.

66 Fed. Reg. 57970–7 (Nov. 19, 2001). Final Paperwork Reduction Act clearance for adverse event reporting harmonization was given in March 2002.

66 Fed. Reg. 57970–7 (Nov. 19, 2001). The Gene Transfer Safety Advisory Board (GTSAB) will have about fifteen members. Two will be members of the RAC.

Members of the NIH staff will be included, as well as a FDA liaison. The remaining members will be chosen for their relevant expertise (e.g., scientific, clinical, statistical, ethical); ad hoc members will be involved as needed. The GTSAB will meet quarterly in closed session, and will provide summary reports to the RAC and for publication.

The Genetic Modification Clinical Research Information System (GeMCRIS) database is being developed by OBA in collaboration with the FDA and with input from all NIH institutes and centers that deal with gene transfer, as well as the Clinical Center and the National Library of Medicine.

A range of documents, including listings and classifications of all open studies, is available on OBA's website. A limited, Phase I version of the GeMCRIS database is currently accessible at <http://www4.od.nih.gov/oba/rac/clinicaltrial.htm> (last visited September 17, 2002). It provides a range of information about current studies in a truncated searchable form.

Walters , supra note 2.

The appellation comes from Abbey Meyers, a former RAC member and a member of the recently disbanded National Human Research Protections Advisory Committee.

Rainsbury , supra note 2.

Churchill , supra note 20. As Churchill have noted, somatic cell GTR has long been held to raise no new questions — but that does not mean that there are no old questions; in fact, there are many.