Douglas S Diekema
Children are important participants in genetics research, and in genetics research involving children, IRBs must address the primary issues of privacy, confidentiality, informed consent, and the return of results. While these issues are not different than with other kinds of research, when these issues involve children they can present differently and are be more challenging to manage. The work and actions of investigators from Arizona State University who performed research beginning in 1990 on Havasupai Tribe serves as a case study for these issues. Issues involved with family studies, large population studies, data and tissue banks, and socially identifiable populations will be discussed.
Children are important participants in genetic research, and IRBs are increasingly faced with the difficult questions that arise with research involving genetics, especially when children are involved. The human subjects issues raised by genetic research are not necessarily different from those raised by other kinds of research, but they present in different ways and may prove more challenging to manage. In one sense, the human genome is simply a rich database of personal information, similar to a patient’s medical record. Issues of privacy, confidentiality, informed consent, and return of results remain the primary issues that IRBs must tackle when confronted with protocols involving genetic information.
To illustrate some of the complexities raised by genetic research, I would like to begin with a case study. Beginning in 1990, investigators at Arizona State University (ASU) collected more than 200 blood samples from members of the Havasupai tribe in an effort to describe genetic variants that might contribute to the increased incidence of diabetes among tribe members.  Over the course of the following 10 years, those “banked” blood samples were also used by other investigators from a number of diverse disciplines to examine DNA variants linked to schizophrenia, alcoholism, metabolic disorders, and the geographic and anthropologic origins of the Havasupai people.  This work resulted in more than 20 academic papers, including several that might be considered stigmatizing for members of the tribe (links to alcoholism and schizophrenia and one study that suggested high rates of inbreeding). [3,4] Particularly vexing to tribal leaders was a paper using DNA analysis to suggest that the Havasupai ancestors had crossed the Bering Strait and migrated from Asia. This “story” told by their blood directly contradicted the story told by tribe elders that taught that the Havasupai had originated in the canyon in which they lived and had been appointed as its guardian.  The schizophrenia studies may have been performed despite knowing that such studies would offend the Havasupai. ASU Anthropology professor John Martin recruited Dr. Markow to do the genetic research.
When she asked if the project could be expanded to include schizophrenia, one of her research interests, he informed her the Havasupai would likely not be interested. According to court records, Markow nonetheless prepared a grant application to study schizophrenia, and the grant was approved. 
It is worth noting that most IRBs would have categorized all of the research involving blood samples collected from the Havasupai people as minimal risk, and while the original study required IRB review, it probably qualified for expedited review under the federal regulations. Furthermore, it is very possible that some or most of the follow-up studies could have been performed under an exempt determination, since they used existing specimens, as long as the data was recorded in such a manner that subjects could not be identified directly or through identifiers linked to the subjects.
Despite the “minimal risk” nature of the study, however, it raised significant ethical concerns. First, the original consent process was likely inadequate. [1,2,6] Whether those providing consent were meaningfully informed about how their blood would be used and that it would be banked is disputed. [5,6] Many tribe members had not graduated from high school and English was not their first language, raising questions about how well individuals understood what they were being asked.  It is doubtful that any of them understood the kind of information that might be obtained from DNA samples extracted from their blood. The consent process also did not cover the specific future uses of the blood samples, though it did refer to the intention “to study the causes of behavioral/medical disorders.” Second, while the study may have been considered minimal risk, published results did significant harm to the community and its members through stigma and the disruption of tribal beliefs. 
Defenders of the study point out that some of the research performed was important to understanding health problems that existed within the tribe, and that data from published surveys reveal that many people have no problem with their existing specimens being used for any scientific purpose. They also argue that setting too high a bar for consent to the use of DNA, tissue, blood, and data will ultimately interfere with important research. While that may all be true, it is also important that research be performed in a way that minimizes risk, and that responsibility is one of the primary roles of the IRB.
As IRBs consider protocols that involve the use of genetic information, they must be aware of the potential issues that can arise in genetic research. While I do not have time and space to provide a comprehensive examination of these issues, I would like to highlight some of the important considerations that IRBs should be discussing.
Family-based studies typically begin with identification of an index case. Enrollment of other family members is often of importance in addressing research questions, and it is usually accomplished through the assistance of the index case patient or the patient’s parent contacting relatives or releasing contact information to investigators. Family studies pose risks of loss of privacy and coercion. The risk to privacy is more likely and potentially more harmful than in other kinds of studies, since enrollment of family members may lead to unwanted disclosure of personal information about one family member to others within the family. For rare disorders, especially those with recognizable manifestations, just the publication of a family pedigree may increase the identifiability of family members to others. Finally, in genetic studies, it is not uncommon for matters unrelated to the disease under study to be inadvertently discovered and disclosed—for example, the disclosure of misassigned paternity.
Family members may also find the recruitment process intrusive and awkward. Because recruitment often involves identifiable family members, the potential for coercion becomes a real possibility. Family members already enrolled in the study may wish to ensure full participation of all family members and pressure relatives to enroll, or may allow the researcher to contact relatives they know to be reluctant to participate. It may be difficult in some families to assure that the identity of those members who choose not to participate remains unknown to the rest of the family. Learning that some members did not participate may cause strain on what had previously been good relationships within the family, with the dissent of individuals being perceived as a sign of disloyalty or lack of concern. Minors, in particular, may feel less free to dissent in the face of family or parental pressure. 
Large Population Studies and Data/Tissue Banks
Several ethical concerns exist for participants enrolled in large population studies, and these studies must include appropriate procedures for informed consent and privacy protection. The value of large data repositories is that they allow the linkage of genetic data with other health-related data in an ongoing manner. When these repositories are created, the specific questions they might be used to address are frequently unknown or undefined and will evolve with time. Obtaining consent that allows participants to understand how their data will be used can be extremely difficult, as illustrated by the Havasupai studies.
Two general approaches to informed consent have been suggested for the banking of genetic samples for future use: periodic re-consent and “blanket consent.” Periodic re-consent allows the participant or her surrogate to be updated regarding specific uses of the data within the database as they arise. This process has the further advantage of allowing pediatric patients to participate more fully in the assent and consent process as they grow older and to provide a legally valid consent upon reaching the age of consent. Periodic consent is most practicable in situations where the research plan involves the ongoing collection of additional data from enrolled participants. For data repositories derived from large numbers of participants where ongoing collection is not occurring, periodic consent poses significant disadvantages because of the administrative burdens and costs involved in tracking and contacting participants for a re-consent conference.
Thus, blanket consent, where the participants are informed that potential future uses of their specimens and data may encompass a broad array of topics and studies that cannot be further specified, has become more commonly used. IRBs must be aware that while this method may be the only practicable way to allow the use of repository data in some cases, it must rely on less than fully informed consent from participants.  The nature of this problem was vividly illustrated by the studies performed using Havasupai blood. Mello and Wolf have suggested a model of blanket consent called “tiered” consent which allows participants to choose from several options at the time samples are collected. This would allow a participant to decide whether to provide general permission for any future use, permission only for future uses related to the original study topic, or a requirement that the participant be re-approached for specific consent for any future use different from the original study or study topic. 
In any case, any form of broad consent should include as much detail as possible about potential future research or commercial use, the possible risks that may arise from any future research, and the mechanism for review of any future research use.  Importantly, blanket or tiered consent from parents on behalf of minors poses a very difficult problem in that the consent effectively “expires” when the child turns 18. Certainly for data that remains identifiable, a mechanism for obtaining the consent of minors when they become legal adults must be considered. For participants who cannot be reached after turning 18, it may be reasonable to seek a waiver of consent from the institutional review board. 
The protection of data confidentiality is essential for large data sets. Information in data sets containing linked records that combine genome sequence data, health records, demographics, and other potentially sensitive information must be carefully protected from unauthorized disclosures, and IRBs should assure that protections are adequate.
When sufficient information is collected, identification may be possible from the combination of data elements describing a particular research participant, even when identifiers are appropriately coded and protected.  If substantial amounts of individual genome sequence are included in a database, a participant could theoretically be identifiable on the basis of the sequence data alone, through matching with a second comparative sample. While the extent of risk for these kinds of potential disclosures currently seems quite small, the risk could increase in the future. Consent forms should include these risks, and participants enrolled as minors should be informed about these possibilities and potential risks when they become adults. IRBs should carefully consider whether participants will be allowed to have their samples and data withdrawn from a repository. If possible, there should be a mechanism for withdrawal, and when not possible, the consent form should clearly state that this is not an option.
Socially Identifiable Populations
The research studies performed with specimens collected form the Havasupai tribe members illustrate risks inherent in studying members from small, readily identifiable populations. These unique risks include:
- Research findings may create unintended harms to the ethnic, religious, and social well-being of individuals within socially identifiable or isolated communities. Studies suggesting higher rates of inbreeding and exploring the geographic origins are two examples of this kind of potential harm.
- Research involving socially identifiable populations creates the potential for individual and group stigma: Genetic studies that identify genetic predispositions in a certain ethnic group, for example, may reinforce negative stereotypes, create misconceptions about people belonging to those groups, and impact marital, adoption, and child-custody opportunities. This is particularly true for genetic risks related to psychiatric problems or undesirable and criminal behaviors.
- Individuals belonging to a group with a predisposition to certain genetic traits may be discriminated against because of group membership, leading others to attribute certain traits to the individual simply because of their membership in the group
- Study findings that identify genetic predispositions within the community, particularly if they represent stigmatizing conditions, can lead to intra-community discord over participation in or support for the research by select members within the group.
Research performed on readily identifiable populations can cause significant harm both to that population and individuals who belong to that population. Traditional forms of individual consent fail to protect against many of these harms, in part because of the focus on individual risks and benefits. Research designed to answer questions about readily identifiable populations should trigger a review of potential group harms, and strongly consider involvement of community members or leadership in the review of these projects.
Return of Results of Genetic Studies
The return of results obtained in a research study is a contentious issue, especially when children are involved. [14,15,16,17] IRBs must consider these issues prospectively and assist investigators in creating a plan for return of genetic (and other) results. The return of results is complicated by the fact that many genetic results are of uncertain significance, involve probabilistic determinations of risk, and may not be clinically actionable. The issues to be considered in creating a plan for return of results include:
- Return of results should be strongly considered if results are scientifically valid and reliable, have health implications for the individual, and could inform the use of some intervention that might improve the health outcome of the individual.
- For any result that is clinically actionable, the result should be confirmed in a lab that is CLIA-certified prior to the release of results to a patient or parent.
- In situations where family members or the research participant may also be affected by a clinically actionable health condition, the investigator should inform the participant of this fact and advise them to relay the information to those family members with the recommendation that they also get tested.
- Individual results not scientifically validated or replicated should not be released to participants.
- Results that do not have health implications, or would not inform the use of some intervention to improve the health outcome of the individual, require more careful consideration by IRBs regarding if, and how, results will be released to participants. This is particularly true when the participants are children. In situations involving children, disclosure of results that do not lead to changes in the care of the child potentially interfere with that child’s future right to decide about what they wish to know. As a general rule, individual results should be shared with a young child’s parent only when they are scientifically valid and reliable, have health implications for the child, and could inform the use of some intervention that might improve the health outcome of the child. Older adolescents may be capable of sufficient understanding that exceptions can be made to this rule with the assent of the adolescent.
- The process of sharing genetic results with participants in the research should involve professionals with the necessary expertise required to adequately communicate the meaning of the results and any health implications. 
- The process of sharing genetic results with participants from cultures that differ from those of the investigators should include a consideration of whether those cultural differences might have implications for obtaining consent, handling sensitive or taboo subjects, and honoring family structures and dynamics. Consultation with representatives of such communities may be appropriate when creating a process for disclosure of genetic results. 
- IRBs should review and approve any plan to return results to participants. 
- The consent form should be very clear about any plan for return of results (or intent not to return them).
- A mechanism should exist for reviewing whether unanticipated results of potential medical importance should be shared with participants. The focus of these decisions should be on the welfare of the participant.
None of the ethical concerns that arise in pediatric genetic research is unique. However, certain concerns are more common in genetic research because of the study designs used for gene discovery. In addition, the power often accorded to genetic information in our society generates additional concern when a research study involves collection of genetic information. Issues of privacy, confidentiality, informed consent, and return of results represent the primary ethical concerns that IRBs and investigators must struggle with in designing and reviewing studies involving the use of genetic information.
The Havasupai tribe remained largely unaware of how their blood samples were being used until Carletta Tilousi, a member of the tribe who was a student at Arizona State University, attended a research presentation and asked whether permission had been obtained to use the blood samples for purposes other than diabetes research.  Ultimately, a lawsuit was filed against the university. The case was eventually settled and included monetary compensation, an apology, and the return of blood samples so that they could be properly buried. Carletta Tilousi told Amy Harmon of The New York Times in 2010:
“I’m not against scientific research. I just want it to be done right. They used our blood for all these studies, people got degrees and grants, and they never asked our permission.” 
The author has disclosed no conflicts of interest.
Douglas S Diekema MD, MPH
Director of Education, Treuman Katz Center for Pediatric Bioethics
Chair, Institutional Review Board, Seattle Children’s Hospital
Professor, Division of Bioethics
Professor, Division of Emergency Medicine
Adjunct Professor, Department of Bioethics and Humanities
Adjunct Professor, Internal Medicine
University of Washington School of Medicine
Adjunct Professor, Department of Health Service
1 Sterling RL. Genetic Research among the Havasupai: a cautionary tale. Virtual Mentor. February 2011; 13(2): 113-117. https://journalofethics.ama-assn.org/article/genetic-research-among-havasupai-cautionary-tale/2011-02.
2 Hart S, Sobraske KA. Investigative report concerning the medical genetics project at Havasupai. Arizona State University Law Library. December 23, 2003. In Garrison NA, Cho MK. Awareness and Acceptable Practices: IRB and Researcher Reflections on the Havasupai Lawsuit. AJOB Primary Research. 2013; 4(4): 55-61. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5310710/.
3 Hedrick PW, Black FL. HLA and mate selection: no evidence in Southern Amerindians. Am J Hum Genet. 1997; 61:505-511. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1715963/.
4 Markow T, Hedrick PW, Zuerlein K, Danilovs J, Martin J, Vyvial T, Armstrong C. HLA Polymorphism in the Havasupai: evidence for balancing selection. Am J Hum Genet. 1993; 53:943-952. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1682387/.
5 HAVASUPAI TRIBE OF HAVASUPAI RESERVATION v. ARIZONA BOARD OF REGENTS. Court of Appeals of Arizona, Division 1, Department D. Nos. CA-CV 07-0454, 1 CA-CV 07-0801. Decided Nov 28, 2008. http://caselaw.findlaw.com/az-court-of-appeals/1425062.html.
6 Lowenberg K. The Havasupai case and how to make consent forms better. Stanford Law School. SLS Blog. https://law.stanford.edu/2010/04/23/the-havasupai-case-and-how-to-make-consent-forms-better/. Accessed 20 Mar 2018.
7 Garrison NA. Genomic justice for Native Americans: impact of the Havasupai case on genetic research. Sci Technol Human Values. 2013; 38(2): 201-223. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5310710/#R5
8 Havasupai Tribe and the lawsuit settlement aftermath. American Indian & Alaska Native Genetics Resource Center. National Congess of American Indians. http://genetics.ncai.org/case-study/havasupai-Tribe.cfm. Accessed Mar 20, 2018.
9 Burke W, Diekema DS. Ethical Issues Arising from the Participation of Children in Genetic Research. J Pediatr. (Supplement) 2006; 149 (1): S34-S38.
10 Mello MM, Wolf LE. The Havasupai Indian Tribe case—Lessons for research involving stored biologic samples. New Engl J Med. 2010; 363: 204–207. http://www.nejm.org/doi/full/10.1056/NEJMp1005203. Accessed 20 Mar 2018.
11 Caulfield T. McGuire AL, Buchanon JA, Burgess MM, Danilczyk U et al. Research ethics recommendations for whole-genome research. PLoS Biology. 2008; 6(3): e73 (430-435). http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.0060073.
12 Brothers KB, Holm IA, Childerhose JE, Antommaria AH, Bernhardt BA, Clayton EW, Gelb BD, Joffe S, Lynch JA, McCormick JB, McCullough LB, Parsons DW, Sundaresan AS, Wolf WA, Yu JH, Wilfond BS; Pediatrics Workgroup of the Clinical Sequencing Exploratory Research (CSER) Consortium; Pediatrics Workgroup of the Clinical Sequencing Exploratory Research CSER Consortium. When participants in genomic research grow up: contact and consent at the age of majority. The Journal of Pediatrics. 2016 Jan; 168: 226-231.e1.https://www.ncbi.nlm.nih.gov/pubmed/26477867.
13 Lowrance WW, Collins FS. Identifiability in Genomic Research. Science. 2007; 317(5838):600-602. http://science.sciencemag.org/content/317/5838/600.
14 Jarvik GP. Return of genomic results to research participants: the floor, the ceiling, and the choices in between. Am J Hum Genet. 2014;94(6): 818-26. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4121476. Accessed Mar 20, 2018.
15 Ziniel SI et al. Parents’ preferences for return of results in pediatric genomic research. Public Health Genomics. 2014; 17(2):105-14. https://www.ncbi.nlm.nih.gov/pubmed/24642506. Accessed Mar 20, 2018.
16 Hallowell N, Hall A, Alberg C, Zimmern R. Revealing the results of whole-genome sequencing and whole-exome sequencing in research and clinical investigations: some ethical issues. J Med Ethics. 2015; 41:317-321.https://www.ncbi.nlm.nih.gov/pubmed/25038088.
17 Lakes KD et al. Maternal perspectives on return of genetic results: context matters. Am J Med Genet A. 2013 Jan; 161(1): 38-47. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3535554/. Accessed Mar 20, 2018.
18 Garrison, NA. Considerations for returning research results to culturally diverse participants and families of decedents. J Law Med Ethics. 2015 Fall; 43(3):569-575. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4617668/.
19 Harmon, Amy. Indian tribe wins fight to limit research of its DNA. The New York Times. Apr 21, 2010. https://www.nytimes.com/2010/04/22/us/22dna.html.