T H E   N I H    C A T A L Y S T     J A N U A R Y   – F E B R U A R Y  2008



by Jerry Menikoff
Director, Office of Human Subjects Research

What’s Past is Prologue: OHSR Director Jerry Menikoff beside a poster from a 1998 NIH symposium on "Medical Research Ethics at the End of the 20th Century: What Have We Learned?"

A new year seems the appropriate time to discuss one of the newest and most promising areas of genetic research: genome-wide association studies (GWAS).

In each of these studies, hundreds of thousands of single-nucleotide polymorphisms (SNPs) are evaluated in hundreds of specimens, with the lofty goal of discovering the genetic changes that play a role in many common diseases.

In this piece, I will try to briefly describe some of the NIH procedures that apply to NIH researchers working in the GWAS arena. I will highlight considerations related to the consent of the people whose genomes are being analyzed—among our chief concerns in the Office of Human Subjects Research.

Because the relevant issues are somewhat different for researchers who produce new genomic data and those who use data that have been produced by others, each topic will be addressed separately.

For Researchers Who Produce GWAS Data

If you intend to produce data by analyzing genomes—either from specimens collected here at NIH or from specimens collected by collaborators outside NIH—you should become familiar with the NIH GWAS rules. OER maintains a website that provides a wealth of information about these rules, including specific information for NIH researchers.

Although the rules formally apply to NIH research approved on or after January 25, 2008, many aspects of those rules will apply to your work even if you received approval prior to that date.

NIH does not want the applicable protections to vary substantially merely because a study received approval a few weeks before the official effective date. Analyzing hundreds of thousands of SNPs in a person’s genome, and putting those results in a databank, is an activity that requires appropriate oversight, regardless of the starting date.

The new NIH rules are designed, in part, to maximize the public benefit from the production of GWAS data. Thus, for example, they generally require you to gather and present your data in a way that allows them to be included in the central NIH database for GWAS data, dbGaP.

Appropriate confidentiality protections will have to be in place; in particular, there are special rules relating to informed consent, among them that there must be IRB review of the consent process. Depending on the specific aspects of the study, the IRB could be at NIH or elsewhere.

Also with regard to consent, different considerations apply depending on whether the samples are collected prospectively or already exist.

If you are collecting samples prospectively, then the consent process, including the consent forms, should adequately discuss the special issues relating to GWAS.

Subjects should be made aware of how their genome will be analyzed, what will be done with the results of that analysis (including the fact that the information generated will be put in a database to be broadly shared with other researchers), and of the privacy and other risks relating to those activities.

In designing the consent aspects of your study, you might benefit from looking at the "Points to Consider" document, which was written by NIH to give guidance to IRBs and institutions on consent issues. A link to that document is on the first page of the OER GWAS website mentioned above.

If your study involves genomic analysis of previously collected specimens, there is a good chance that such an analysis was not contemplated at the time of collection. Thus, the original consent forms are unlikely to provide the types of specific information that would be expected in a prospective study.

It will ultimately be up to the IRB to determine whether the original consents are adequate to allow the study to proceed, or whether it might be necessary to renew the consent of the people whose specimens are being used.

For Researchers Who Use GWAS Data Produced by Others

If you want to use data stored in dbGaP for your own research, you will need to go through two steps. First, you need to obtain a one-time approval from your IC that authorizes you to gain access to data in that repository. Once that has been secured, you must then submit requests for the specific sets of data in dbGaP that you want to access.

Those requests will be reviewed by the data access committee (DAC) that has responsibility for the oversight of those datasets. In particular, the DAC will make sure that your proposed use does not violate any use limitations included in the original consent forms.

GWAS research has tremendous potential, but it is complex, and NIH is still in the process of working out the details of implementing these new rules. So, be on the lookout for additional guidance, which should be coming out in the near future. Questions can be sent here.

And if you have specific questions related to protecting the interests of the people whose specimens are being used, don’t hesitate to contact the Office of Human Subjects Research by phone at 301-402-3444, fax at 301-402-3443, or e-mail.


Return to Table of Contents