Teneille Brown and I had the pleasure of attending a conference at Stanford last Friday titled “Biobanking, Bioethics and the Law” and moderating the last session. We will post a link to a video of the conference when it is available. A brief description of some of the presentations follows:

Session on “The Scientific Promise of Biobanks”

Scott Vandenberg (UCSF) discussed the determination and implementation of best practices, including technical and operational standards to ensure quality and reproducible analysis, high-quality annotation of biospecimens, peer review, ethical and privacy compliance, the use of informatics, and communication with participants. Some questions to consider: Does one size fit all or should best practices be tissue-type specific? How should guidelines be established? Should they be mandatory? How is custodianship determined? Who pays? Who owns what, including intellectual property?

Douglas Levinson (Stanford) presented on Biobanking for Genetic Studies of Psychiatric Disorders. He gave a brief history of psychiatric genetics, from genetic epidemiology (what runs in families?) to genetic linkage studies, mapping of genes using SNPs, and the genomewide association study (GWAS). He also briefly discussed the history of biobanking, with the first samples deposited in 1999. Dr. Levinson described ongoing studies, including the Molecular Genetics of Schizophrenia (and the challenges of recruitment, such as informed consent) and GWAS breakthroughs. He examined consent language that had been used historically for biobanking. He believes that medical sequencing will produce a huge leap in knowledge. Challenges to ensuring confidentiality include mitigating the risk that unethical researchers with access to genomic and identifying data might determine who contributed DNA to a different dataset. In addition, how do we approach consent to future studies? How do we respect privacy, while allowing people autonomy to consent?

Russ Altman (Stanford) discussed the Genotype to Phenotype Database that relates genetic information to observable traits. This information helps researchers better understand how people respond to certain drugs (pharmacogenetics and pharmacogenomics). He offered the example of Warfarin, which is a blood thinner that is extremely difficult to dose. Two genes explain much of the variability. Trials are ongoing using demographics and genetics to set dose, reduce side effects and improve outcomes. Pharmacogenetics holds the promise of providing focused treatment, reducing adverse effects, saving drugs that are effective in subpopulations, and better understanding drug interactions. He discussed a Warfarin Pharmacogenetics biobank that evolved based on the need for a globally derived and globally relevant dosing algorithm. The biobank researchers planned genotype-guided random clinical trials, consisting of 21 research groups from 11 countries on 4 continents (rare to have this much open collaboration internationally). Annotation and DNA tissue created a usable bank, which included informed consent for sharing. By using information from the database, the predictive dosage was more likely to be accurate than by using clinical data or a fixed dose approach. By using genetic data, racial variants in dosage response don’t matter. More studies of this type are needed to make pharmacogenomics a reality. Large population based studies, though challenging to set up, are needed for pharmacogenetics to have the greatest chance for success.

Rex Chisholm (Northwestern) presented on Biobanking, Electronic Health Records and Genomics: the NUgene experience. He addressed how natural gene variation affects drug response. He discussed the NUgene Project, in which participants have provided broad consent to allow for future genetic research and the different uses of the data, such as using it to evaluate eMR data and perform bioinformatics research. The informed consent obtained in this research allows for sharing with third-parties, including commercial entities, and the option of recontact. He described the eMERGE Network, which is a group of universities with eMR data linked to their DNA databases. Issues for consideration include when do these types of studies qualify as human subjects research (e.g. if the data is completely unidentified, is it still human research?), what is the role of commercialization, and how do we balance benefits and privacy?

Q&A: How do you determine when/whether to withdraw or use tissue that cannot be renewed (e.g. pediatric brain tumor tissue, as opposed to DNA)? Generally, consent doesn’t address under what circumstances to allow withdrawal. The biobank researchers should be acting as gatekeepers for the uses, and require maximizing usefulness of tissue (share RNA produced). In addition, after an embargo period for publication, the genotypes should become part of the sample.

How do we reconcile privacy with advancing research? A nationwide eMR system would be a good start, standardizing the structure of the data sets and enabling patients at any hospital to have their information readily available. Some safeguards include using data consistent with consent and requiring that researchers not attempt to identify the patients associated with any sample. Many issues with eMR focus on the difficulty of funding, as it is highly costly to coordinate the records. What are we most interested in protecting–is it privacy or non-discrimination based on genetic data? If the risk of an unethical researcher and any attendant harm is minimal, how should we be required to disclose to allow for informed consent? But, are the risk and subsequent harm really minimal, or should we consider any breach of confidentiality significant?

Session on the “Legal and Ethical Issues Related to Biobanks”

David Resnik (NIEHS, NIH) presented “Exploitation and the commercialization of human biological materials.” The main objections to exploitation are the desire to prevent commercialization of human tissues, lack of or inadequate consent, harm to individuals and inequitable outcomes. Actions can be wrongful, but morally justified. Other considerations include freedom to consent, autonomy, and social benefits. Powerful interests can skew bargaining power. Courts vary in looking at tissue donation as a gift and as property, though arguably if a donation is a gift, it used to be the donor’s property. Radical solution: make patenting of cell lines illegal, but that would negatively impact medical research and clinical medicine. Another solution: status quo, but improved informed consent. Does informed consent even exist where the negotiating power of the patients is so diminished compared to that of the researchers? Another solution: Enact best practices or legal protection for human biological materials. His suggestion: Recognize human subjects have ownership interests in tissues.

Kelly Ormond (Stanford) discussed “What do participants understand about biobanks?” She presented background on DNAbiobanks. Consent styles: Opt in, Opt Out, Presumed consent (Iceland). Issues: Why do people participate, what do they understand about participation and recontact. Interviews of 200 NuGene participants have been conducted. Main responses for why people participate: altruism, potential personal or family benefit. Rates of participation: hypothetical participation (~90%), actual participation where primary or specialist encourages enrollment (~70-90%), actual participation as a result of “cold call” recruitment (~15%). She discussed the challenges of informed consent, focusing on confidentiality protections and disclosure of results, and various approaches to recontact and reconsent (including expectations regarding results). She also highlighted the need for clarification of components for autonomous decision making and recontact considerations.

Christina DeHayes (GC, Asterand) provided a commercial prospective. She discussed what a commercial biorepository looks like, as well as the types of laws that her company are required to follow. Asterand follows the common rule, although they are not required to do so. They are required, however, to follow the UK Human Tissue Act. She discussed the framework for global operation under which Asterand operates. Collections for commercial companies are different from those for non-profits (e.g. post-mortem collection, consent requirements, ownership, limited withdrawal, addressing international variations).

Ellen Wright Clayton (Vanderbilt) presented on “BioVU: Vanderbilt’s Approach to Biobanking.” The Federal Office for Human Research Protections (OHRP) determined that de-identified previously collected medical records and biological samples do not involve “human subjects.” Vanderbilt created synthetic derivative of eMR for over 1.5 million patients. Vanderbilt collects DNA from residual blood samples from outpatient areas. Although there is no requirement to obtain consent, Vanderbilt allows patients to opt out and has widely publicized its program. Currently, they have 50,000 samples. In terms of oversight, they have numerous Advisory Boards (community based and otherwise), and an IRB overseeing their actions. She also explained how Vanderbilt investigators, who don’t currently have access to the biobank, would need to go through the BioVU Access Request Process. Assessment is ongoing to evaluate the efficacy of their procedures. Challenges include expansion into pediatric populations: children don’t have blood drawn as frequently and the amount of blood drawn is generally smaller. Including pediatrics is important because children have diseases that affect them differently from adults and childhood diseases may significantly affect their lives as adults. In addition, at Vanderbilt, it is challenging to recontact after data has gone through the one-way hash.

Robert Cook-Deegan (Duke) discussed “People, Data, and Materials in Genomic Research: Altruism Meets Commerce.” He began with coverage of genetic patents, namely media reports of genetic patents for certain types of diseases and cases in policy reports. Most intellectual property protection in this area depends on genetic test discoveries. Distinctive features in this space as opposed to patented drugs: most IP (~75%) is owned by the universities doing the research. Currently, there does not appear any need to show clinical utility or expensive trials to prove safety and efficacy. This could change if FDA and CMA require data for approval or reimbursement, which would increase the need for patents. Myriad has patents BRCA1 and BRCA2 and aggressively enforces them. Volume matters: Myriad didn’t sue the Canadian government despite its refusal to allow monopolist pricing. Conclusions: Patents matter, but more so in the US. Coverage and reimbursement matter. Patents interact with complex health and payment systems both in US (Medicare) and abroad. Donors are altruists, while companies try to commercialize via property rights and financial incentives. Universities are stuck in the middle: university licensing of IP rights requires the university to act as a mediator.

— Brenda Simon