Gone are the days when investigators postulated that common genetic variants were largely responsible for common disease. In this era of genome-wide association studies, researchers are beginning to realize that rare variants play a bigger role in disease susceptibility than was once thought.
Recruiting enough patients with rare variants for deep-phenotyping studies can be difficult. "By definition, being rare, it's kind of hard to find people that have them," says Laura Beskow, assistant research professor at the Duke Institute for Genome Sciences and Policy. "And so, if by virtue of a previous study you've already identified people who have that variant … then that's going to be the more efficient way to go as opposed to just enrolling huge numbers of people and hoping that some of them have that variant that you're interested in."
"From a scientific perspective, it would make a lot of sense to try to group individuals based on genotype for hypothesis-driven research," adds Amy McGuire, associate director for research in the Center for Medical Ethics and Health Policy at Baylor College of Medicine.
Indeed, genotype-driven research recruitment holds promise for deep-phenotyping investigations aimed at deciphering the clinical significance of genetic variants, Beskow says. If, as a result of a GWAS, researchers find that a particular variant is significantly associated with a particular disease, they might be interested in pursuing follow-up work with participants who carry the allele in question.
"Once researchers decide that they would like to do this kind of recruitment, I think that they — without any ethics person around — themselves realize the difficulties of it," Beskow says.
The ethical issues raised by recruiting study participants based on their genetic information are far-reaching, inextricably context-dependent, and yet to be fully explored. Beskow and her colleagues at Duke have been awarded an NIH Challenge Grant to investigate these issues through a survey of institutional review board members, study coordinators, and participants. They will hold a workshop — to include attendees representing federal agencies and IRB chairs, among others — at the end of the study in May 2011 to draft guidelines for recruiting participants on the basis of their genotype. "My guess is that they're not going to be hard and fast rules," Beskow says.
At the heart of the issue, Beskow says, lies the principle of upholding patients' "right not to know" their genetic information. In a Genome Research paper published online in April, Beskow and her colleagues describe a scenario in which investigators re-contacted the original participants from a study examining variants for risk of epilepsy. In an effort to preserve the patients' right not to know, the study coordinator issued letters soliciting participation in a follow-up study to all members of the cohort, regardless of whether they possessed the genotype of interest. "The idea [is] that just by virtue of getting a letter ... that by itself is telling the person 'you're one of the ones who has this genetic characteristic.' You're not giving them the chance not to know that," Beskow says. "If I find myself contacted by a researcher, even if they told me, 'Look, Laura, really we don't know what this means,' I would give some weight to the fact that ... they're calling me, they must think it means something."
This is not a problem specific to disease risk studies, she says. It applies to just about every genotypic investigation. "Generally speaking, the idea is that researchers might come up with something and think that they should tell people, but as soon as you call them and say 'I've got a finding and was wondering if you'd like to know it,' you've already given people information," Beskow says. "It's very, very tricky."
Figuring it out
Both Beskow and McGuire are unsure of the best way to re-contact study participants based on their genetic information. While Beskow's team examines the issues inherent to individual re-recruitment scenarios, McGuire is considering the potential structural and policy issues that could arise as more and more investigators look to recruit study participants based on their DNA.
McGuire says that re-recruiting participants from "within your own research projects … is one thing, [but] if you're accessing data that's available in a database like dbGaP and you want to go back and contact other people's participants to then recruit them into your study, then it becomes much more complicated." In the latter situation, McGuire says, there is a strong need for an objective third party, or an "honest broker," to facilitate that communication.
"Right now the burden sort of falls on the investigator who collected the original samples," McGuire says. "If this really becomes widespread, I think that's not a very feasible model." Some investigators do not have the time, resources, or motivation to re-contact their participants, while others "have a sense of wanting to protect that community and … may have an interest in sort of vetting what types of research they get recruited into," she says.
Whatever the PI's disposition, the task inevitably requires the transfer of sensitive, identifying information. McGuire suggests that a neutral third party — "governed by very strict rules of confidentiality" — should mediate these transactions. In a 2008 Genome Research paper, she and her colleagues suggested that as a "kind of as a first step, that the NIH -Data Access Committee take on this role," she says, though she notes the potential for participants to be resistant to the notion of "the government having access to identifying information about them."
Beskow and her colleagues at Duke have only just begun to analyze the thousands of survey responses as part of their Challenge Grant. She says that her team will consider the post-workshop guidelines they produce to be an important first step. "It's definitely a matter of putting those guidelines out there and getting feedback on them, doing some additional research with other stakeholders," she says. "There's a lot of work left to be done and data is always good for informing these kinds of policy and ethics questions."