No. 5: July-September 2012



Alex Mauron


Genetic Biobanks

Academic Foresights

How do you analyze the present situation of genetic databases and biobanks?


Genetic biobanks are an increasingly important resource for studying the connexion between genes and disease. Many common diseases are multifactorial since they result from the interplay between genetic risk factors and environmental factors. Genetic epidemiology studies these gene-environment and gene-gene interactions. A particularly powerful methodology consists of genome-wide association studies, in which genetic data from a large number of individuals are compared to health data from these individuals. The development of molecular population genetics, along with the development of large-scale biobanks, has made these approaches increasingly feasible while raising complex and partly novel ethical questions.


Given the heterogeneity of research approaches linked to biobanks, one needs to answer a preliminary question: what sort of collection of biological material and/or data “counts” as a genetic biobank? Even if we set aside human material in anthropological and anatomical collections, or samples collected for forensic purposes (all of which raise their own ethical issues), it turns out that all sorts of sampling activities go on in the world of modern biomedicine. Particularly in the course of routine clinical care, many kinds of human material and/or data need to be obtained and stored. Their main use is to inform proper care of patients. Yet additional uses follow straightforwardly from this care-centred objective: Clinical labs need “old” samples for quality control and equipment calibration; pathologists need to go back to previously obtained biopsies for diagnostic verification, to solve a clinical problem that was not on the radar screen at the time of initial analysis, or even to address an issue of medical genetics that arises many decades later in a patient that happens to be a genetic relative of the initial sample “source”. In addition, these collections may be used for teaching students in a wide range of health-related professions. Finally, many such collections become potential genetic biobanks secondarily, since modern methods of molecular genetics allow the extraction of DNA from most human materials. Therefore, many such collections become potentially useful resources for research in a way that could not necessarily be perceived at the time of sampling. An interesting example is the Guthrie test that is performed routinely on newborns since the nineteen sixties. The blot spots (“Guthrie cards”) and associated identifying data are usually kept as part of medical files. The technology of genetic analysis has progressed considerably and allows to re-analyse these samples with new research questions in mind, an ethically troubling prospect since these samples were obviously taken without parental consent for research, and in fact without any form of consent since Guthrie tests are part of government-mandated newborn screening in many jurisdictions. That does not mean that reusing existing genetic samples for research is necessarily unethical, but it needs a major rethinking of the interests and values underpinning consent and privacy in biomedical research.


Meanwhile, biobanks-based research moves on at a rapid pace and increasingly, biobanks are created de novo with specific scientific objectives and methodological constraints. It is this kind of biobank, set-up as a research tool from the start, that is the most important today and on which much of the present bioethical discussion centres.


In your opinion, how will the situation likely evolve over the next five years?


The philosophically perplexing nature of genetic biobanks is that they typically have multiple identities. They are repositories of (1) “stuff” (DNA, cells, tissues...), (2) genetic and medical data, and (3) potential genetic data. Indeed you can go back to the “stuff” to squeeze out additional genetic information in a way that is largely open-ended, and that makes biobanks data different from traditional epidemiological and public-health databases. This means that the usual understanding  of consent to research that applies when a patient is asked to be included in a specific clinical research protocol, cannot be used as is. Individuals that are approached to donate a genetic sample and medical data will have to consent to future research in a way that is necessarily broader than is the case in traditional research on human subjects. It is true that in theory the scope of consent could be restricted to a particular research project that is presently considered, and the individual would be re-contacted in the future for every new project. In practice, this is a non-starter for many biobanks, on account of the difficulty and expense involved and also because many large-scale biobanks only make sense as long-term investments in a very broadly defined area of research. One traditional answer to this ethical conundrum is the irreversible anonymisation of samples and data. The current ethical discussion is moving away from this solution and is emphasising either “conditional blanket consent” (consent to a widely defined domain of research on condition that every new project be examined by an independent research ethics board) or “multiple choice consent” (participating individuals answer “yes” or “no” to a list of questions aimed at specifying what concerns are important to them). As the consensus moves away from irreversible anonymisation, and biobank management thus involves keeping coded data which are potentially identifiable, issues of confidentiality and privacy are raised. However, increasingly sophisticated technical answers are being developed, so that this may be less of a concern in the future.


Another issue central to current and future discussions about biobank research is the management of incidental findings that are of potential relevance to participating individuals. While this is not unknown in clinical genetics, where the problem is largely managed on a case-to-case basis, it is becoming a central policy issue in the management of biobanks. Every biobank ought to have a defined policy and decide in advance whether it will re-contact participating individuals to share incidental information obtained in the course of research that may be relevant to that person’s health and medical treatment. Whether this is desirable, realistic, or possibly more harmful than non-disclosure is still very much debated.


Many other ethical and legal aspects of genetic biobanks are currently discussed and are sometimes quite controversial. These include issues of ownership of samples, intellectual property vested in research results, and benefit sharing, especially when samples originate from communities in developing countries for the benefit of “first world” researchers and for-profit companies.


What are the structural long-term perspectives?


The first generation of genome-wide association studies using single nucleotide polymorphisms suffered from inherent methodological limitations that have sometimes led to questionable results. These limitations are about to be overcome by rapid progress in DNA sequencing technology. Whole genome sequencing is becoming a feasible option that will give a considerable boost to biobank-based research.  Massively parallel sequencing, often referred to as “next generation sequencing”, will provide another quantum leap in analytical power, opening up a whole range of new possibilities in basic biomedical research. Both in order to harness this technological potential and to cope competently with ethical issues, genetic biobanks will increasingly be set-up and managed as large-scale research facilities. Scientific and technical personnel will increasingly be employed full-time by biobanks, as is already the case for the largest ones. Among other things, this will avoid the conflict of interest arising if biobank managers are also researchers using the facility. Access to biobank samples and data will have to be based on independently assessed scientific quality of projects. Ethical aspects will also need independent expertise. Finally we will increasingly see major biobanks set-up a permanent interface for participating study subjects, who will be invited to interact with the biobank on a long-term basis, if they so wish. This will increase the scientific value of biobanks, by offering the possibility of updating medical information concerning participating individuals.  This may also give a new turn to the ethical discussion, because several issues such as confidentiality, privacy, and coping with incidental findings could be handled more constructively if participants are considered active members of a research community rather than one-off sample “sources”.


-   -   -

The website of UK Biobank provides a good insight in the scientific, managerial and ethical aspects of large-scale biobanking.



Alex Mauron studied molecular biology at the University of Lausanne (PhD, 1978). After a post-doctoral stay at Stanford University he moved to the field of bioethics, and is presently a full professor of bioethics at the medical faculty of the University of Geneva.


© Copyright: click here                                       Join our discussion group on LinkedIn