The first line of defense against misuse of genetic information garnered from research will always be the researchers themselves. In designing and carrying out their experiments, scientists try to make sure that their subjects are protected, and in general they do a good job. But because individual researchers cannot be expected to spot all the potential pitfalls of a project, universities, research hospitals, government labs, and many corporate labs supply a second line of defense: the institutional review board, or IRB.
"Any research that is federally funded or that is being done for submission to the FDA must by law be reviewed and approved by an IRB," noted Leo Whelan of the legal department at Mayo Clinic. And in general, Whelan said, the IRB approach seems to work well, allowing local oversight of research that can be tailored to local conditions. "It has been very effective in determining what type of review is appropriate, determining what the risks are, addressing those risks, and advising the researchers on appropriate ways to handle the confidentiality and privacy issues, which often have to be tailored to the research. For example, if you are involving minors in genetics research, a whole host of new issues come up. When you have a family linkage study as opposed to a study dealing with individuals, when you are dealing with patients from your own institution versus a population of samples coming from a variety of different institutions—these pose very different sets of issues."
Despite the apparent effectiveness of IRBs at Mayo, however, many workshop participants thought that the IRB system as a whole needed improvement if it is to be effective in protecting the subjects of tomorrow's increasingly powerful and sophisticated genetic research.
One major concern is that IRBs are only as good as their members, but many
institutions put too little effort into supporting and compensating IRB members, which makes it difficult to retain the people best suited for the job. "Where more and more responsibility is being put on IRBs, you have to remember these are non-paid slots, voluntary, and a lot of time is required," said Pearl O'Rourke from NIH's Office of the Director, who recently moved from the University of Washington, where she served on an IRB. While there, O'Rourke found that many clinicians were hesitant to serve on the IRB because it represented 8 hours a month of their time that weren't billable hours. "I also found that as a physician I could sway any vote because community members and others [on the IRB] did not have a clue about genetic information. I also felt that in terms of the local interests as soon as you have NIH or any federal moneys on a grant there is the bias that you have to okay this project because it is good for the institution." If IRBs are to be effective, she said, institutions must find some way to reward or ''validate" service on them. ''It cannot just be this volunteer, non-paid, 'Oh, yeah, I can do it' thing."
Besides attracting and keeping good members, IRBs must find some way to keep up-to-date on the areas of research that they are considering, and they do not seem to be doing a good job of that now. "Professor Mildred Chou at the University of Pennsylvania and I have published a study of IRBs in the Journal of Investigative Medicine focusing on a genetic issue," said Paul Billings. We concluded that most IRBs were seemingly poorly informed about much that was going on at the national level or in states about the social, ethical and legal issues surrounding genetics." One potential solution, Billings suggested: "It might benefit IRBs substantially to have some sort of network of communication, and sharing of information, that apparently is not present currently. IRBs should be informed of research findings, board reprints, commissions and other sources relevant to their deliberations."
Finally, David Korn argued that although the local character of IRBs may have advantages, it also carries a cost. "There may be two or three thousand IRBs extant in the country right now, and every one of them is on its own in trying to interpret a particular research proposal and determine what standards it needs with regard to authorization or this or that or the other." This lack of standardization means that similar research projects performed at different institutions may be evaluated quite differently and placed under quite different constraints. The result is that researchers never quite know what to expect.
"We hear all kinds of scare stories from institutions around the country where research proposals involving issues of genetic information and data privacy have been sitting in the local IRB for up to 2 years while these people sit and have talmudic discussions about whether it can or it can't or what should be done and what shouldn't be done," Korn said. "Mr. Whelan's IRB may be an excellent, well-functioning IRB. But many of them are troubled by these issues and don't really know where to go. So, I think that a federal standard that provided some clarity would be useful."
The closest thing to a federal standard on genetics research now available is a set of IRB guidelines published by the Office for Protection from Research Risks (OPRR) in 1993. It was intended to serve as an educational tool for researchers and members of IRBs, said William Dommel and Melody Lin of OPRR.
The OPRR's process is not entirely satisfactory, Korn said. "Those guidances almost have the effect of rules. Once they are issued IRBs tend to pay very, very careful attention to them." But because the guidelines were not intended to serve as policy, they did not go through a formal rule-making process with public input and debate. "I am not trying to make more work for the government," Korn said, but it would be useful if OPRR or perhaps some other agency would go through such a formal process, with input from the public, to set out policies that IRBs across the country should follow.
Sheri Alpert of George Mason University explained her concern about the IRB review process. "The current regulations only require an IRB to look for risks when information is identifiable to an individual or affects an individual. This does not protect the privacy of an identifiable group that may be ideal for the study of a particular genetic disease." The group may not want to be associated with, or known as high probability carriers for a specific "disease gene." Simply being recognized as a member of that group has the potential to affect their privacy, whether as an individual they carry the gene or not. "I would say the regulations need to be strengthened or re-examined as more and more genetic analysis occurs that is specifically identifiable to an ethnic or social group that has a genetic dimension."
But if IRBs have their shortcomings, they are clearly much better than no review at all—which is precisely what a great deal of genetic research in the United States may face. "Well over half of all the dollars spent in the United States in the pursuit of genetics-based research is spent today by private industry," mainly by pharmaceutical and biotechnology companies, said Barbara Handelin, a private genome research consultant. "Those research protocols that are not done in collaboration with an academic institution and not funded by federal moneys are not subject to IRB review." Drug companies are required to get IRB approval for any research that will be submitted to the FDA in support of a new drug, but that still leaves a great deal of private-industry research uncovered by IRBs. "Mostly what ends up in an FDA filing is preclinical and clinical research," Handelin said. "Basic research typically does not end up in an FDA filing.''
Non-profit organizations may pose a similar problem, said Vicky Whittemore of the National Tuberous Sclerosis Association. "As a non-profit we give out a significant number of research dollars, as do many other non-profits. We require our researchers to submit IRB approval for our applications, but I don't know how widespread that is among non-profits, and it raises the issue
of how much clinical research is actually being done on patients with non-IRB approved protocols. I don't know."
If private companies and non-profit organizations are to conduct genetic research, some workshop participants said, they should hold their researchers to the same standards that federally funded scientists are held to, and that implies review by IRBs or something comparable. "I think that it is incumbent upon the industry to do its own homework," Handelin said, "and I would hope that that can be done in the absence of new and burdensome regulations, that industry gets it together to take a much clearer look at the need for external review of research protocols that are being conducted by and funded by companies. I think that this industry actually should be held to a higher ethical standard than many other industries because we deal in and are given the privilege of dealing in some of the most vulnerable aspects of our society."