Creation and Governance of Human Genetic Research Databases

Cre

atio

n a

nd

Go

ve

rna

nc

e o

f Hu

ma

n G

en

etic

Re

se

arc

h D

ata

ba

se

s

Creation and Governance of Human Genetic Research Databases

SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH

BIOTECHNOLOGY INNOVATION SCIENCE HEALTH BIOTECHNOLOGY INNOVATION SCIENCE HEALTH BIOTECHNOLOGY INNOVATION

INNOVATION HEALTH SCIENCE BIOTECHNOLOGY HEALTH INNOVATION SCIENCE BIOTECHNOLOGY HEALTH

HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY


BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION

INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY HEALTH





HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH

HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH BIOTECHNOLOGY













SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY

BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION

INNOVATION HEALTH SCIENCE BIOTECHNOLOGY

HEALTH SCIENCE BIOTECHNOLOGY INNOVATION

SCIENCE

OECDPUBLISHING

Scientists have known for years that complex diseases, including cancer, heart disease, stroke, and diabetes, arise from a complex combination of lifestyle, environmental, genetic and random factors. Large-scale study of populations may contribute significantly to science’s understanding of the complex multi-factorial basis of diseases and to improvements in prevention, detection, diagnosis, treatment and cure. As a result of developments in biotechnology and bioinformatics, the opportunity to store and analyse increasingly large amounts of genetic data have rendered possible the creation of large-scale population databases. Genetic research, involving the use of such databases containing human genetic and genomic data, information, and biological samples, is thus becoming increasingly feasible.

More recently, the databases being developed include data, information and biological samples from whole populations. Large-scale population databases which contain a significantly broader range of information about individuals also raise a number of issues and concerns. While some of these are not new, the increasing breadth and scope of such databases amplifies them. Moreover, the combination of a broader set of genetic data and personal information in these databases raises new issues about the use of such information, especially in a non-clinical or non-research context. In addition, as such databases will increasingly be international in scope and cover populations from numerous jurisdictions, new sets of questions will arise.

The OECD organised a workshop in order to begin the process of considering, at the international level, policy challenges associated with the establishment, management and governance of human genetic research databases. This report provides an overview of the complex issues that were discussed at that workshop and which need to be considered or addressed, in recognition of the significant contribution that human genetic research databases could play in translating scientific advances into innovation in health.


www.oecd.orgISBN 92-64-02852-8 93 2006 09 1 P-:HSTCQE=UW]ZWY:

The full text of this book is available on line via these links:http://www.sourceoecd.org/scienceIT/9264028528http://www.sourceoecd.org/governance/9264028528

Those with access to all OECD books on line should use this link:http://www.sourceoecd.org/9264028528

SourceOECD is the OECD’s online library of books, periodicals and statistical databases. For more information about this award-winning service and free trials ask your librarian, or write to us at [email protected].

OECDPUBLISHING

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

Creation and Governance of Human Genetic

Research Databases

001.fm Page 1 Monday, October 23, 2006 12:46 PM

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

The OECD is a unique forum where the governments of 30 democracies work together toaddress the economic, social and environmental challenges of globalisation. The OECD is also at

the forefront of efforts to understand and to help governments respond to new developments andconcerns, such as corporate governance, the information economy and the challenges of an

ageing population. The Organisation provides a setting where governments can compare policyexperiences, seek answers to common problems, identify good practice and work to co-ordinate

domestic and international policies.

The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic,

Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea,Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic,

Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission ofthe European Communities takes part in the work of the OECD.

OECD Publishing disseminates widely the results of the Organisation’s statistics gathering andresearch on economic, social and environmental issues, as well as the conventions, guidelines and

standards agreed by its members.

© OECD 2006

No reproduction, copy, transmission or translation of this publication may be made without written permission. Applications should be sent to

OECD Publishing [email protected] or by fax 33 1 45 24 99 30. Permission to photocopy a portion of this work should be addressed to the Centre français

d'exploitation du droit de copie (CFC), 20, rue des Grands-Augustins, 75006 Paris, France, fax 33 1 46 34 67 19, [email protected] or (for US only) to

Copyright Clearance Center (CCC), 222 Rosewood Drive Danvers, MA 01923, USA, fax 1 978 646 8600, [email protected].

This work is published on the responsibility of the Secretary-General of the OECD. The

opinions expressed and arguments employed herein do not necessarily reflect the official

views of the Organisation or of the governments of its member countries.


FOREWORD – 3

CREATION AND GOVERNANCE OF HUMAN GENETIC RESEARCH DATABASES – ISBN-92-64-02852-8 © OECD 2006

Foreword

Under the leadership of the Japanese and Canadian governments and under the aegis of the Working Party on Biotechnology (WPB), the OECD Directorate for Science, Technology and Industry (DSTI) organised a Workshop on Human Genetic Research Databases – Issues of Privacy and Security. While human genetic research databases hold much potential for improving science’s understanding of disease and complex, multi-factorial conditions, they also raise numerous challenges. The OECD’s involvement and work in this field arises from the need to address such challenges.

The Workshop brought together leading experts from numerous jurisdictions, international organisations, diverse backgrounds including academia, government, industry and research labs, as well as from different fields including genetics, law, medical, ethics, philosophy and economics. The aim of this Report is to begin the process of considering, at the international level, the policy challenges associated with the broader subject of the establishment, management and governance of human genetic research databases. Ultimately, this Report aims to provide an overview of the complex issues that need to be considered or addressed, in recognition of the significant contribution that human genetic research databases could play in translating scientific advances into innovation in health.

We are especially grateful to the Japanese and Canadian governments for their generous sponsoring of the Workshop and support. This Report was written by Ms. Christina Sampogna, of the Biotechnology Division, OECD. It draws upon the experts’ background papers, commentaries and presentations, recordings of the Workshop, the Rapporteur’s report, as well as additional research. As this is a rapidly evolving field, with each human genetic research database continuously reviewing its policies and approaches, the information contained herein is up to date to August 2006.

TABLE OF CONTENTS – 5


Table of Contents

Executive Summary ............................................................................................................9 Human genetic research databases...................................................................................10 Establishment of a human genetic research database ......................................................10 Data and sample collection and management ..................................................................12 Database management and governance ...........................................................................14 Commercialisation considerations ...................................................................................16 Résumé ..............................................................................................................................19 Bases de données de la recherche en génétique humaine ...............................................20 Établissement d’une base de données de la recherche en génétique humaine ................21 Collecte et gestion des données et échantillons ..............................................................23 Gestion et gouvernance des bases de données ................................................................25 Considérations relatives à la commercialisation .............................................................27 Chapter 1. Introduction....................................................................................................29 1.1. Context ....................................................................................................................29 1.2. Overview of issues ..................................................................................................30 Chapter 2. Human Genetic Research Databases ...........................................................35 2.1. Examples of Human Genetic Research Databases .................................................36 2.1.1. The Personalised Medicine Research Project (“Marshfield Project”) ......36 2.1.2. CARTaGENE (“CARTaGENE”) .............................................................36 2.1.3. Genome Database of the Latvian Population (“Latvian Project”) ............38 2.1.4. Icelandic Health Sector Database (“Icelandic HSD”)...............................39 2.1.5. Estonian Genome Project (“Estonian Project” or “EGP”) ........................40 2.1.6. The United Kingdom Biobank (“UK Biobank”).......................................41 2.1.7. Translational Genomic Research in the African Diaspora (“TgRIAD”)...41 2.1.8. GenomEUtwin (“GenomeEUtwin Project”) .............................................42 2.1.9. The International HapMap Project (“HapMap Project”) ..........................43 2.1.10. COGENE (“COGENE”) ..........................................................................44 2.1.11. P3G – Public Population Project in Genomics (“P3G”) ..........................44 2.2. What is a human genetic research database?..........................................................44 Notes ..............................................................................................................................46 Chapter 3. Establishment of an HGRD ..........................................................................55 3.1. Nature and Scope of Database................................................................................55 3.1.1. Ensuring representativeness of populations ..............................................55 3.1.2. Involvement of children and protected adults ...........................................56 3.1.3. Nature of database.....................................................................................56 3.1.4. Intended purposes and uses of the database ..............................................58 3.2. Funding of a database.............................................................................................59 3.2.1. Private sector funding ...............................................................................59 3.2.2. Public-private partnership .........................................................................60 3.2.3. Public sector funding.................................................................................60

6 – TABLE OF CONTENTS


3.3. Legal structure........................................................................................................61 3.3.1. Legal context for databases .......................................................................61 3.3.2. Human rights statutes ................................................................................64 3.3.3. Jurisdictional aspects of population databases ..........................................64 3.4 Privacy and confidentiality.....................................................................................65 3.5. Public engagement in the establishment of a population database.........................68 3.5.1. Explaining the purpose of the population database...................................68 3.5.2. Approaches to engaging the public ...........................................................69 3.5.3. Approaches to ongoing communication strategies....................................71 3.5.4. Implications of communication strategies.................................................72 Notes ..............................................................................................................................74 Chapter 4. Data and Sample Collection and Management ...........................................83 4.1. Data and samples....................................................................................................83 4.1.1. Nature of genetic information ....................................................................83 4.1.2. Type of data and samples ..........................................................................84 4.2. Ownership of data and samples..............................................................................88 4.3. Consent...................................................................................................................89 4.3.1. Nature of consent ......................................................................................90 4.3.2. Consent for population databases..............................................................90 4.3.3. Children’s and protected adults’ consent ..................................................93 4.3.4. Renewed consent and re-contact ...............................................................94 4.4. Right to withdraw...................................................................................................95 4.5. Results ....................................................................................................................96 4.5.1. Results back to database............................................................................96 4.5.2. Results back to participants.......................................................................97 4.6. Education and training of data collectors and researchers......................................99 4.6.1. Education and training of data collectors ..................................................99 4.6.2. Education and training of researchers .....................................................100 Notes ............................................................................................................................101 Chapter 5. Database Management and Governance ...................................................105 5.1. Management and governance of databases ..........................................................105 5.1.1. Legislation and regulation of HGRDs.....................................................105 5.1.2. Role of ethics and oversight committees.................................................107 5.1.3. Management of HGRDs..........................................................................109 5.1.4. Powers, compliance and enforcement .....................................................110 5.2. Security of databases ............................................................................................111 5.2.1. Custody of code registry .........................................................................111 5.2.2. Limited data release ................................................................................111 5.2.3. Limited data access .................................................................................112 5.2.4. Data encryption .......................................................................................113 5.3. Access to population databases ............................................................................113 5.3.1. Principles underlying access to HGRDs .................................................113 5.3.2. Purpose and conditions of access to HGRDs ..........................................115 5.4. Demise of database...............................................................................................117 Notes ............................................................................................................................119

TABLE OF CONTENTS – 7


Chapter 6. Commercialisation Considerations ............................................................123 6.1. Intellectual property..............................................................................................124 6.1.1. Intellectual property generally ................................................................124 6.1.2. Intellectual property and population databases .......................................124 6.2. Commercialisation ...............................................................................................126 6.3. Benefit sharing .....................................................................................................127 Notes ............................................................................................................................130 Chapter 7. Conclusions...................................................................................................131 7.1. Policy themes arising from the workshop ............................................................132 7.1.1. Is genetic information special?................................................................132 7.1.2. Public perceptions ...................................................................................132 7.1.3. Public trust ..............................................................................................133 7.1.4. Human rights norms and existing legal frameworks...............................134 7.1.5. International harmonisation.....................................................................134 7.1.6. Protection of identifiable information .....................................................134 7.1.7. Linkability ...............................................................................................135 7.1.8. Revisiting basic principles ......................................................................135 7.1.9. Commercialisation policy .......................................................................136 7.2. Future areas of work.............................................................................................136 Notes ............................................................................................................................138 Glossary ...........................................................................................................................141 Workshop Agenda ..........................................................................................................143 List of Workshop Participants.......................................................................................147 Bibliography ....................................................................................................................153

EXECUTIVE SUMMARY – 9


Executive Summary

The development of biotechnology and bioinformatics affords the opportunity to store and analyse increasingly large amounts of genetic data. Genetic research involving the use of databases containing human genetic and genomic information, sometimes alone or in combination with other personal or medical information, has thus become increasingly important. More recently, the databases contemplated and being developed for genetic research are quite different in nature and larger in magnitude. Many of these emerging databases focus on and include data, information and biological samples from populations. These population databases, also referred to as human genetic research databases (HGRDs), may contribute significantly to science’s understanding of the complex multi-factorial basis of diseases (genetic and non-genetic components) and therefore to improvements in detection, prevention, diagnosis, treatment and cure. Such databases may contribute significantly to the identification of genes associated with disease, an understanding of the frequency of genetic variants in particular populations, and to a better understanding of the reasons for drug reactions (both positive and negative) and reactions to other environmental factors.

These databases also raise a number of issues and concerns. While some of these are not new, the increasing breadth and scope of such databases amplifies them. Moreover, the combination of a broader set of genetic data and personal information in these databases raises new issues about the use of such information, especially in a non-clinical or non-research context. In addition, such databases are increasingly international in scope, covering populations from numerous countries, which raises new sets of issues.

Despite pressing concerns, there is limited international guidance on the establishment, governance and management of human genetic research databases. While certain institutions, such as UNESCO and the Council of Europe, have developed instruments which focus on the use of genetic data, these instruments do not address the multitude of issues raised by such databases. The underlying motivation for the OECD’s involvement in this field is the need to address the development, use and access to population databases containing genetic data and personal/medical information. The aim of this Report is to begin the process of considering, at the international level, the policy challenges associated with the broader subject of the establishment, management and governance of human genetic research databases. Ultimately, this Report aims to provide a summary of the complex issues that need to be considered or addressed, in recognition of the significant contribution that human genetic research databases could play in translating scientific advances into innovation in health.

10 – EXECUTIVE SUMMARY


Human genetic research databases

The linchpin question is the determination of which databases ought to be considered a human genetic research database. There are numerous types of different databases including nucleotide sequences, sequences variations, mutation sequences, gene expression, gene loci, protein structures, as well as model organisms and diseases (i.e. pathology databases). Many of the databases currently being undertaken share the characteristic of collecting data, information and biological samples with the aim of allowing research for one or multiple diseases and conditions. However, the linchpin question is much broader. The consideration is whether only databases that contain data, information and biological samples for one or more “populations” or for large subsets of one or more populations ought to be considered HGRDs.

There are diverse other dimensions to the linchpin question. One dimension is consideration of whether biobanks/tissue banks that are being amassed, especially within the private sector, should be included in the definition of HGRDs. Another key dimension is whether “human genetic research databases” should be considered to include databases that also contain personal, medical or other data and information. This determination includes consideration not only of the type of biological samples or information that will be collected and stored within the database but also the determination of the source of that information. Genetic data have been broadly defined as “all data, of whatever type, concerning the hereditary characteristics of an individual or concerning the pattern of inheritance of such characteristics within a related group of individuals.” Genetic data do not necessarily include information derived from DNA or RNA specimens; they can be inferred from family history, medical records or phenotype. Genetic data may be distinguished from personal genomic data which have been defined as “detailed personal data derived from analysis of DNA specimens.”

Moreover, consideration should be given to whether HGRDs should include databases that contain personally identifiable information or only databases that contain data and information that cannot be associated with or result in the identification of individuals. Personal data have been defined as “any information relating to an identified or identifiable individual”. Genetic data may be collected from individuals in different manners, each associated with varying degrees of identifiability. This complex issue is related to those raised by anonymisation and linkability of data, which are considered separately.

Establishment of a human genetic research database

At the inception of an HGRD, a crucial issue is the nature and scope of the database. Also critical to the scientific legitimacy of such a database is ensuring that the sample population chosen is genetically representative of the population it is to serve. Ideally, all groups should be included in a given study. However, owing to financial and practical constraints, this is not always feasible. An important consideration is the selection of criteria that will ensure a careful and rigorous selection process which results in a representative database.



Another issue pertinent to the nature and scope of the database is whether or not children should be allowed to participate in genetic studies. Some have argued that young children should be excluded from genetic studies. This, however, would severely obstruct research on genetic diseases which occur early in life. Others have argued in favour of including children, but have often tied their support for inclusion to the issue of consent. Consideration should be given to whether or not children should be included in population database studies, and if so what are the appropriate safeguards.

The intended uses of the population database should also be determined at inception. This determination is especially important in order to ensure that appropriate information can be provided to the participants in the project. One of the obvious objectives in establishing HGRDs will be to carry out research. However, one issue will be whether the specific nature of the intended research may be determined or determinable at the time that the HGRD is established or at the time of the collection of biological samples, data and information. The degree to which this determination may be made will have implications for issues pertaining to consent, to communication with the community and to building public trust. Another set of determinations pertain to whether the biological samples, data and information collected in a database could or should be allowed to be employed for other purposes. Therein, a key consideration is whether HGRDs should be allowed to be used for non-scientific/medical research purposes. Examples of other purposes for which the contents of a database could be employed include the delivery of clinical genetic services, law enforcement, insurance, legal actions and identification purposes (e.g. for military or civil).

Different funding structures are available for the establishment of HGRDs: for-profit (i.e. private undertaking), not-for-profit (i.e. public undertaking) or mixed model (i.e. public-private partnerships). For example, the Icelandic Health Sector Database was envisaged as a for-profit endeavour, the Estonian Genomic Database was conceived as a mixed model, and the United Kingdom Biobank database was established as a not-for-profit undertaking. What criteria should be employed to determine whether an HGRD should be a public, private or mixed- model undertaking?

The collection of a large number of data and information about a given individual raises numerous privacy and confidentiality issues. Privacy is generally considered to mean the right to be left alone. In the context of genetic research, it could also mean the right not to know genetic information. Genetic information obtained in the research context raises unusual privacy concerns because it has the potential to generate information and knowledge beyond that which was originally sought, and because it raises the possibility that researchers obtaining the information might be obligated, in some situations, to provide that information back to the patients who contributed DNA samples.

Confidentiality connotes the notion of a professional keeping information private once it has been revealed by a client, and is part of a fiduciary relationship. In the context of genetic research, the concern is most often related to keeping genetic information that is collected in a research setting from third parties such as health insurers or employers. However, in the computer age, there are now also concerns about violation of databases or the sale of genetic information for marketing purposes.

Privacy and confidentiality issues may vary with the type of database. Database systems built for the purpose of finding answers to only one or a limited number of research problems may raise different considerations from broad population databases. In the case of specific research databases, the manner in which data are collected, stored and



accessed will be more targeted and more limited. In such cases, the storage of the information may be simpler (e.g. not connected to an external network). However, in the case of an HGRD, the information collected and stored in the database may be more broadly accessed. Conversely, the information contained in specifically-targeted databases may be more easily linked back to identifying information and therefore enable identification. Large population databases, by their very size, may reduce this type of risk. An important consideration is the principles that should be established to ensure that the participant’s privacy and confidentiality are respected.

Another factor that influences the issue of privacy and confidentiality is the nature of the biological samples, data and information collected. Identified samples pose the most direct challenges to privacy and confidentiality. Thus, many researchers have chosen, or have been required, to use coded, unlinked or unidentified samples. Even though direct identifiers have been removed from coded samples, they may remain identifiable and thus still present privacy and confidentiality concerns. Researchers often find coded samples more valuable (even critically so, in certain types of research) than unlinked or unidentified samples because the linkage with identity provides a way to follow up with individuals in longitudinal studies. Therefore, it has been advocated that this type of linkage cannot, and should not, be completely prohibited. To ensure privacy and confidentiality of databases with coded material, technical and procedural computer security (including control and monitoring of access to, and transport of, data) may be essential. An important element is the measures that should be undertaken to ensure the protection of the data and information contained within databases.

In establishing a population database, public support is essential, given that participation is voluntary. This implies that the collection of biological samples, data and information and the inclusion in the HGRD of these depend on the consent of the donor. It may be difficult to estimate participation rates in projects where benefit is indirect, long-term and at the population level, especially in resource-poor communities or populations with different beliefs, cultures or languages. In order to build public trust, it will be important to bridge the distance between the research community and participants. However, the manner in which public support is elicited may vary considerably. In seeking to establish an HGRD, the methods employed to engage the public, the information that should be provided, and the manner for communicating it most effectively are important factors.

Data and sample collection and management

The cornerstone of a human genetic research database is the data, information and biological samples collected and stored therein. At the outset, one important consideration is whether the data, information and biological samples should be unidentified (i.e. anonymous), unlinked (i.e. anonymised), coded (i.e. linkable or identifiable), or identified. Such considerations may have implications for issues of privacy and confidentiality as well as participants’ involvement. Each of these approaches has advantages and disadvantages which must be assessed in light of the HGRD’s objectives and purpose. For example, anonymous data may raise the least risk for breach of privacy but may not be as valuable for researchers, especially for longitudinal studies.

HGRDs raise issues with respect to ownership of the data, information and the biological samples collected. Irrespective of whether the database is a private, public or a



mixed-model endeavour, the issue of who should be able to claim title to the data, information and biological sample arises. The ownership issue includes not only consideration of the immediate question of proprietorship but also longer-term implications, for example for the database’s functioning or any commercialisation being contemplated. Equally important is the issue of remuneration for the provision of the biological samples, data and information. Whether or not to remunerate participants, beyond simple reimbursement of basic expenses, is another important issue. It includes consideration of whether such activity is permitted pursuant to the applicable national or regional law, as well as whether it may affect the credibility of the HGRD, the researchers, and collectors as well as its representativeness.

Informed consent is one of the most complex issues for human genetic research databases. Informed consent has become the pillar for protecting autonomy in research involving human subjects. Within the medical/scientific field, informed consent generally presumes the ability to indicate clearly to the participant the use and purpose of the particular research activity. While this is feasible for purpose-specific research, the very nature of HGRDs renders the provision of this type of information difficult. Therefore, the issue becomes what constitutes informed consent within the context of an HGRD given that the purposes for which the data, information and biological samples are collected and the uses for which they may be employed, usually, may be described only in a general manner. Many have queried whether the traditional model of informed consent is applicable in the context of HGRDs or whether a new model/paradigm should be developed. For example, some authors have advocated blanket consent. Others have advocated a general consent for limited purposes, with the undertaking to return to the participant should the proposed uses go beyond those limited purposes.

Additional consent questions raised by HGRDs include children’s consent and renewed consent. In situations where a determination has been made about the involvement of children in genetic studies and HGRDs, consideration of the consequences of their involvement and the manner in which to obtain consent is primordial. For young children, obtaining consent may involve obtaining the consent of the parents. For older children, consent may involve a variety of approaches depending on their level of development and understanding. Issues that arise in the context of renewed consent include returning to the participant to obtain consent for new uses in the context of the HGRD. For example, the question arises of whether a new consent is required when existing databases are converted to HGRDs.

Re-contacting participants, in any of the above-mentioned situations, raises practical difficulties (e.g. the person may have died or moved away) but also more complex issues, such as whether or not the person wished to be re-contacted. This question involves consideration of a principled approach to re-contacting participants. As well, consideration should be given to whether participants ought to be provided with information pertaining to the issue of re-contacting prior to obtaining their initial consent.

Having voluntarily contributed to an HGRD, participants may, at some point wish to withdraw from the study and have their biological samples, information and data destroyed. One issue is whether HGRDs’ should adopt a policy pertaining to participants’ right to withdraw their data, information and samples and related issues, including whether this is possible and if, so under what circumstances. In some cases, it may be possible for participants to withdraw their data, information and biological samples throughout the duration of the HGRD. In others, depending on the manner in which the HGRD is established, it may be possible to withdraw data, information and biological



samples only prior to their anonymisation. Moreover, the right of withdrawal may entail various options. For example, if the participant’s data have been included in information provided to a third party, it may not be possible for these to be selected or withdrawn.

In many epidemiological studies, information or results are provided back to either the database and/or the participant. However, given the breadth of an HGRD, the question is whether or not the providing back of results is feasible or desirable. First, there is the issue of whether or not results from users of the data and samples should be returned to the database. While the providing of results back to the database may enrich it overall, quality assurance of the results provided and included in the database is an important consideration. A second set of issues pertain to the providing back to the participants results derived from the use of the data and biological samples contained in the HGRD. Given the breadth and purpose of HGRDs, the issue is whether it is feasible to contemplate a policy of providing results back to participants and the value of doing so, especially if provided outside of a clinical context.

The training of health-care professionals and researchers will be important to the success of an HGRD. Health-care professionals responsible for the recruitment of participants and collection of the data (i.e. interviews, questionnaires, medical examination, drawing of blood, transfer of the information collected) may not be familiar with genomic research. Thus, a policy for training such health-care professionals may be important. For example, such a policy may include a protocol explaining their role to general practitioners, outlining the information they may divulge to the participants, and guiding them in the appropriate handling of contentious situations.

Database management and governance

The governance of databases involves numerous operational, technical and legal issues, including consideration of applicable legislation and regulation, the role of ethics and oversight committees, the powers, compliance and enforcement attributes granted to the HGRD, the security features of the database, access to the database and the demise of an HGRD.

In the establishment of an HGRD the question arises of whether the database should result from a statute or be independent of an act of parliament. For example, the Estonian Genomic Database and the Icelandic Health Sector Database would be the creation of an act of parliament. Conversely, the UK Biobank and the Canadian CARTaGENE initiative were established independent of specific legislation but are subject to a number of existing legislation. There are advantages and disadvantages in establishing a database via an act of parliament versus a softlaw instrument, such as a memorandum of understanding.

A review of most HGRD initiatives reveals that they require some form of oversight committee, but the composition and formation vary across projects. The role, function and nature of the oversight committee in the governance of an HGRD should be addressed. The composition of the oversight committee, especially whether it should be multi-disciplinary, the term for an individual’s appointment, a policy or approach for determining which issues should be submitted for the consideration of the oversight committee are all important issues. For example, the Ethics Committee of the Estonian Genomics Database ensures compliance with ethical guidelines, and any requests on behalf of projects seeking access to the Database may be submitted for its consideration.



An important set of issues for HGRDs concerns the powers, the ability to ensure compliance and the mechanisms for enforcement of decisions. Such powers, or the lack thereof, may have implications for ensuring that privacy and security policies are respected, and for ensuring that a private entity to which commercialisations rights are granted respects these rights and does not act in an abusive manner. For example, since the Icelandic database would operate by licence, the power to revoke that licence is a means of effecting compliance. If any of the provisions of the operating licence or the enabling legislation are violated, the Minister can issue a written warning, and set a deadline by which action must be taken. Inaction or intentional and gross negligence may result in revocation of the licence.

Given the potential for misuse of data and samples collected in HGRDs, the security of the database is primordial. This is both a legal and a technical question. Given the objective of such databases, considerations of the best methods for ensuring security, ensuring that access occurs only in the permitted manner, and ensuring that access to the data and samples is not stifled are significant. One method proposed is through custody of a code registry. This method would be most valuable for protecting the confidentiality of data in databases wherein links are maintained between data and personal identifying information. In this approach, the custodian could be a person who is under the duty of non-divulgation of confidential information, such as a medical doctor. An additional security feature of this approach may be the use of stand-alone computers for handling personal identifiers and other personal information, including health data, so as to reduce the risk of unauthorised access of networked computers.

Another approach to ensuring security and protecting privacy is to limit the amount or type of data released or accessible to researchers using the HGRD. This may involve a combination of legislation and technical solutions. For example, it could include limiting the bin size or ensuring that data and information are not made available to researchers if the data sample does not involve more than a certain number of individuals. Similarly, privacy and confidentiality may be protected by limiting or monitoring access to the data. A simplified way of doing this is to allow only researchers with approved passwords to access a database. A more sophisticated version of this strategy is to use rule-based control of access to data, instead of, or in addition to, human intervention. This method would allow different users to access different categories of data and information according to their roles. An alternative is to allow a very limited set of analysts to query the primary data directly. In this scenario, “outside” researchers would be allowed to query the data only indirectly, through these analysts, and would only receive summarised answers to their queries (e.g. means, P-values, etc.).

A common method for enhancing data privacy is encryption, most often either one-way or public-key. This method may also be used in conjunction with the methods described above. Since encryption can be fairly easily implemented, it is assumed that data transferred to and from research databases will be encrypted in some way. However, it must be recognised that encrypted data can also be decrypted, so encryption should not be relied upon as the only means of privacy protection.

Given that the fundamental purpose of HGRDs is to foster research, access to the database raises crucial issues. Key questions for consideration include who should have access to the database (e.g. only researchers, public- and/or private-sector researchers), the manner in which access should be given (directly versus via an internal researcher), whether access should be free or for a fee (who should pay the fee and what should it be), and to what should access be given (e.g. to the whole database, to parts and which ones,



only to certain data and then only in an anonymised manner). Another key issue is the purpose for which access should be granted.

While there are few examples of HGRDs that failed or were terminated (for example, the Tongan database that was to be established by Autogen Limited), the possible demise of an HGRD should be considered as early as possible, especially when establishing its governance structure. Consideration should be given to clearly determining the consequences of its demise. This could include, for example, whether all of the data and the samples are to be maintained or destroyed and whether participants should be notified of the demise of the database. If the database is operated by a private undertaking, consideration of whether provision should be made for a government to retain the right to have the database handed over to them or whether a government should reserve at least a right of first refusal. Such considerations will be influenced by the applicable legislation. For instance, many countries have enacted legislation prohibiting the sale of human tissue or materials. The consequence of the application of such statutes would need to be taken into account.

Commercialisation considerations

HGRDs raise issues with respect to commercialisation, including intellectual property, the actual commercialisation of the database and/or it components and benefit-sharing.

Intellectual property broadly refers to the legal rights which result from intellectual activity in the industrial, scientific, literary and artistic fields. One set of issues raised by HGRDs pertains to the intellectual property rights (IPRs) that may arise as a result of research employing data or samples accessed from a database. In such circumstances, questions arise of who owns the invention and who is under an obligation to ensure that the relevant IPRs are protected. Another important consideration is access to a follow-on innovation developed using data and samples from the database. A policy that would balance follow-on access while permitting a return on investment would also be an important consideration. Intellectual property issues may also arise with respect to the database itself, including database rights, where they exist, copyright protection for the software and other rights that allow the database to operate effectively.

With respect to commercialisation, the first consideration is whether or not it is desirable to commercialise the database and/or whether commercialisation is in line with the participants’ expectations. If commercial exploitation of the database is undertaken, consideration should be given to the manner in which this should be carried out. For instance, commercialisation could be on an exclusive or non-exclusive basis. If the commercial exploitation were to be undertaken on an exclusive basis, it would be essential to consider how to ensure fair access to the database and how to ensure compliance with competition law. The question of whether or not the database could be sold or transferred for consideration would also be important.

The issue of benefit sharing is a complex one with many aspects that vary depending on the structure of the database. For instance, in the context of an HGRD established as a public-private partnership or as a private undertaking, consideration should be given to whether the government should be entitled to some compensation from the private entity and the form of such compensation. In the case of the Icelandic Health Sector Database, for example, the Licence Agreement would have required payment to the government of



i) an annual fixed fee, earmarked for the promotion of health care and R&D; and ii) 6% of the profits of the Licensee, capped at ISK 70 million a year. If monetary compensation is the option favoured, the use made of these sums would be an important consideration. Moreover, compensation may also take a non-monetary form, such as technical or scientific support. Benefit-sharing also raises the issue of whether or not the participant should be entitled to individual benefits arising from the database. For example, whether participants would be entitled to share in the profits of a successful invention developed using samples, data and information contained in the database. A further issue is whether participants should be given the right to access other, non-monetary benefits, such as the products developed as a result of outside research but involving data and samples from the HGRD.

RÉSUMÉ – 19


Résumé

Les progrès de la biotechnologie et de la bioinformatique nous permettent aujourd’hui de stocker et d’analyser des quantités croissantes de données génétiques. Les recherches en génétique qui mettent à profit les bases de données contenant des informations génomiques et génétiques humaines, utilisées seules ou en conjugaison avec des informations personnelles ou médicales, prennent donc de plus en plus d’ampleur. Les bases de données récemment conçues et développées pour les besoins de la recherche génétique marquent un tournant évident, qu’il s’agisse de leur taille ou de leur nature. Bon nombre de ces nouvelles bases privilégient, et donc accumulent, des données, informations et échantillons biologiques à l’échelle de populations. Ces bases de données populationnelles, appelées également « bases de données de la recherche en génétique humaine » (BRGH) peuvent apporter une précieuse contribution à l’étude des origines complexes des maladies multifactorielles (composantes génétiques et non génétiques) et, partant, au progrès en matière de dépistage, de prévention, de diagnostic, de traitement et de remède. Elles peuvent être très utiles pour identifier les gènes associés à certaines maladies, connaître la fréquence des variants génétiques dans certaines populations, et mieux comprendre les raisons des réactions (positives et négatives) aux médicaments et aux différents facteurs environnementaux.

Ces bases de données posent cependant un certain nombre de questions et problèmes. Si ces questions ne datent pas d’hier, elles prennent aujourd’hui une nouvelle dimension du fait de l’ampleur et du contenu toujours plus vaste de ces bases de données. De plus, le croisement de différentes données génétiques et d’informations à caractère personnel dans ces bases soulève de nouvelles questions concernant l’usage de ces informations, notamment hors du cadre de la recherche et des études cliniques. Par ailleurs, un nombre croissant de bases de données ont une envergure internationale puisqu’elles couvrent des populations de plusieurs pays, ce qui soulève encore d’autres questions.

En dépit de l’actualité de ces questions, il existe assez peu d’orientations internationales concernant l’établissement, la gouvernance et la gestion des bases de données de la recherche en génétique humaine. Certaines institutions, telles que l’UNESCO et le Conseil de l’Europe, ont mis au point des instruments relatifs à l’utilisation des données génétiques, mais ceux-ci n’abordent pas les problèmes que soulèvent ces bases. Si l’OCDE a décidé de s’engager dans ce domaine, c’est parce qu’elle considère qu’il est nécessaire d’aborder la question du développement, de l’utilisation, et de l’accès aux bases de données populationnelles contenant des données génétiques et des informations personnelles/médicales. L’objet du présent rapport est d’amorcer la réflexion internationale sur les défis que posent aux pouvoirs publics l’établissement, la gestion et la gouvernance des bases de données de la recherche en

20 – RÉSUMÉ


génétique humaine. Le rapport se propose en outre de récapituler les questions complexes qu’il conviendra de prendre en considération et de traiter, compte tenu du rôle de premier plan que pourraient jouer ces bases de données pour mettre le progrès scientifique au service de l’innovation dans le domaine de la santé.

Bases de données de la recherche en génétique humaine

La question essentielle est d’abord de déterminer quelles bases doivent être considérées comme des BRGH. Il existe en effet plusieurs types de bases, notamment sur les séquences de nucléotides, les variations de séquences, les séquences de mutation, l’expression génétique, les loci génétiques, les structures protéiques, et également les organismes et maladies modèles (bases de données sur les pathologies). Par ailleurs, bon nombre des bases de données qui voient le jour actuellement ont en commun de collecter des données, informations et matériel biologiques qui serviront à la recherche sur de multiples maladies et conditions. Cependant, la question se pose en réalité en termes beaucoup plus larges puisqu’il faut tout d’abord se demander si les BRGH doivent n’inclure que les bases de données contenant des informations à l’échelle de « populations » ou de larges sous-groupes de population.

Cette question comporte de nombreux autres aspects. Par exemple, les biobanques et les collections de tissus privées constituées par beaucoup d’entreprises privées doivent-elles être considérées comme des BRGH ? Les bases de données contenant des informations personnelles, médicales ou autres, doivent-elles être considérées comme des bases de données de la recherche en génétique humaine ? Il s’agit de définir non seulement le type de matériel biologique ou d’informations qui sera collecté et stocké, mais aussi les sources de ces informations. Les données génétiques ont été schématiquement définies comme « les données, de tous types, concernant les caractéristiques héréditaires d’un individu ou le mode de transmission de ces caractéristiques dans un groupe d’individus apparentés ». Les données génétiques ne comprennent pas nécessairement des informations obtenues à partir d’échantillons d’ADN ou d’ARN ; elles peuvent être tirées de l’histoire familiale, des dossiers médicaux ou des phénotypes. Elles doivent être distinguées des données personnelles sur le génome qui ont été définies comme des «données personnelles détaillées tirées de l’analyse d’échantillons d’ADN ».

Il importe encore de se demander si les BRGH doivent inclure les bases de données qui contiennent des informations sur des personnes identifiables ou seulement les bases de données contenant des données et informations ne pouvant être associées à un individu ou utilisées pour le retrouver. Les données personnelles ont été définies comme « toute information concernant un individu identifié ou identifiable ». Les données génétiques peuvent être recueillies auprès des individus de différentes manières, plus ou moins identifiables. L’anonymisation et la possibilité de mettre en relation les données, posent un problème complexe qui sera examiné séparément.

RÉSUMÉ – 21


Établissement d’une base de données de la recherche en génétique humaine

Il importe, avant de mettre en place une BRGH, de définir sa nature et son champ d’application. Il est en outre nécessaire, pour assurer sa validité scientifique, de veiller à ce que l’échantillon de population choisi soit génétiquement représentatif de la population à laquelle elle doit servir. Idéalement, tous les groupes devraient être inclus dans une étude donnée. Cependant, cela n’est pas toujours possible pour des raisons financières et pratiques et il importera donc d’établir les critères qui permettront d’assurer un processus de sélection rigoureux et précis pour obtenir une base de données représentative.

Toujours en ce qui concerne la nature et le champ couvert par la base de données, il importe de décider si les enfants doivent ou non participer aux études génétiques. Certains estiment qu’il faut en exclure les jeunes enfants. Un tel choix risque cependant de freiner sérieusement la recherche sur les maladies génétiques intervenant dans les premières années de la vie. D’autres préconisent au contraire d’inclure les enfants mais en insistant sur l’idée de participation consentie. Il convient de déterminer si oui ou non les études des bases de données populationnelles doivent comprendre les enfants, et dans l’affirmative, de définir des garanties appropriées.

Il importe également de déterminer, dès sa création, à quoi va servir la base de données. Cela est extrêmement important si l’on veut pouvoir fournir aux sujets participants les informations dont ils ont besoin. La recherche sera évidemment un des principaux objectifs de la création d’une BRGH. Toutefois, il faut se demander si la nature spécifique des recherches à entreprendre est déterminée ou déterminable au moment de la création de la BRGH ou de la collecte des échantillons biologiques, données et informations. Cela aura des conséquences pour les questions de consentement, de communication avec la communauté et de mise en confiance du public. Il faudra aussi décider si l’on pourra, ou devra, autoriser des utilisations secondaires des données, informations et échantillons biologiques collectés dans la base de données. Dans ce cadre, il sera indispensable de décider si la BRGH pourra être utilisée à d’autres fins que la recherche scientifique/médicale. Parmi les utilisations secondaires il y aurait éventuellement la prestation de services de génétique clinique, l’exercice des pouvoirs de police, l’assurance, les actions en justice et l’identification (par exemple, militaire ou civile).

Les bases de données peuvent être envisagées selon différents modèles : à but lucratif (entreprise privée), sans but lucratif (entreprise publique) ou structure mixte (partenariat public-privé). Il existe actuellement des bases de données fondées sur des structures très diverses. Citons l’exemple de la Biobanque islandaise (Icelandic Health Sector Database) qui avait été envisagée comme une entreprise à but lucratif, la Biobanque estonienne (Estonian Genomic Database) qui a été conçue comme une base de données mixte, et la BioBank du Royaume-Uni qui est un projet à but non lucratif. Quels critères appliquer pour déterminer si la BRGH sera une entreprise publique, privée ou mixte ?

La collecte d’un grand nombre de données sur chaque individu pose de nombreux problèmes de protection de la vie privée et de la confidentialité. Le respect de la vie privée s’entend généralement comme le droit d’être « laissé en paix ». Dans le contexte de la recherche génétique, il pourrait plutôt s’agir du droit de ne pas connaître l’information génétique (droit de ne pas savoir). Les informations génétiques obtenues dans le cadre de la recherche posent des problèmes inhabituels de protection de la vie

22 – RÉSUMÉ


privée parce qu’elles permettent de générer des informations et des connaissances qui vont au-delà de leur finalité initiale, et aussi parce qu’il existe un risque que les chercheurs qui obtiennent ces informations soient obligés, dans certaines situations, de les communiquer aux patients qui ont fourni des échantillons d’ADN.

L’idée de protection de la confidentialité renvoie à la notion de conservation par un professionnel d’une information privée confiée par un client, et à celle de rapport de confiance. Dans le contexte de la recherche génétique, il s’agit le plus souvent de faire en sorte que les informations génétiques recueillies dans le cadre d’un travail de recherche ne soient pas accessibles à des tiers, notamment aux assureurs et aux employeurs. Toutefois, à l’ère de l’informatique, le risque de piratage des bases de données, ou de vente des informations génétiques à des fins commerciales, ne peut être exclu.

Les questions de protection de la vie privée et de confidentialité dépendent du type de base de données. Les bases constituées pour apporter des réponses à un seul ou à un nombre limité de problèmes scientifiques ne poseront pas les mêmes difficultés que les bases concernant des populations entières. Dans le premier cas, la collecte, le stockage et l’accès aux données seront plus ciblés et plus limités. Le stockage des informations risque d’être plus simple (sans connexion à un réseau externe, etc.) Les informations recueillies et engrangées dans le cas d’une BRGH risquent en revanche d’être plus largement accessibles. D’un autre côté, les informations contenues dans des bases de données bien ciblées pourraient permettre plus facilement de retrouver une information identifiante et donc d’opérer une identification. Les grandes bases populationnelles devraient réduire ce type de risque, du simple fait de leur taille. Il importe de réfléchir aux principes à mettre en place pour assurer le respect de la confidentialité des données et de la vie privée des sujets participants.

Les problèmes de confidentialité et de respect de la vie privée sont aussi liés, dans une large mesure, à la nature des données, informations et échantillons biologiques collectés. Les échantillons identifiés sont les plus directement concernés par la protection de la vie privée et la confidentialité. Pour cette raison, beaucoup de chercheurs ont choisi, quand cela ne leur a pas été imposé, d’utiliser des échantillons codés, non retraçables ou non identifiés. Même si les identificateurs directs ont été retirés des données codées, celles-ci peuvent encore être identifiées et continuent donc de poser un problème de protection de la vie privée et de la confidentialité. De l’avis de nombreux chercheurs, les échantillons codés sont préférables (voire tout à fait préférables dans certains types de recherche pour lesquels ils sont indispensables) aux échantillons non retraçables ou non identifiés parce que les liens avec l’identité du sujet permettent de suivre des individus dans les études longitudinales. En conséquence, il a été constaté que ce type de corrélation ne pouvait et ne devait pas être complètement interdit. Pour garantir le respect de la vie privée et la confidentialité des bases de données contenant du matériel codé, il faudra bien souvent recourir à des procédures techniques de sécurité informatiques (contrôle et surveillance de l’accès et du transport des données, par exemple). Il sera donc essentiel de déterminer les mesures à prendre pour assurer la protection des données et informations contenues dans les bases de données.

L’établissement d’une base de données populationnelle ne peut se faire sans l’adhésion du public sachant que la participation est librement consentie. Cela signifie que la collecte de données, informations et échantillons biologiques et leur stockage dans la BRGH sont tributaires du consentement du donneur. Il sera sans doute difficile de prévoir le taux de participation dans le cas de projets dont les retombées bénéfiques seront indirectes, à long terme et au niveau de toute la population, surtout dans les

RÉSUMÉ – 23


communautés dépourvues de ressources ou chez les populations qui ne partagent pas les mêmes croyances et la même culture ou parlent une autre langue. Pour gagner la confiance du public, il sera important d’établir une passerelle entre la communauté scientifique et les sujets participants. Toutefois, toutes sortes de méthodes peuvent être appliquées pour mettre le public en confiance. Lors de la création d’une BRGH, il sera essentiel de définir les méthodes qui seront employées pour associer le public, les informations qu’il importera de fournir et les options les plus efficaces pour les communiquer.

Collecte et gestion des données et échantillons

Les données, informations et échantillons biologiques collectés et stockés constituent la pierre angulaire de toute base de données de la recherche en génétique humaine. La première question est de savoir si ces données, informations et échantillons biologiques doivent être non identifiés (anonymes), non retraçables (anonymisés), codés (chaînables ou identifiables) ou identifiés. Les choix opérés auront des conséquences pour la protection de la vie privée et de la confidentialité et la participation du public. Chacune de ces formules présente des avantages et des inconvénients qui doivent être évalués au regard des objectifs et de la vocation de la BRGH. Par exemple, les données anonymes réduisent le risque d’atteinte à la vie privée mais présentent moins d’intérêt pour les chercheurs, en particulier pour les études longitudinales.

Les BRGH posent aussi la question de la propriété des données, informations et échantillons biologiques collectés. Que la base de données soit une entreprise privée, publique ou mixte, il conviendra de se demander qui pourra revendiquer des droits de propriété sur les données, informations et échantillons biologiques. Le problème de la propriété nécessite une réflexion sur les choix immédiats mais il a également des implications à long terme, par exemple pour le fonctionnement des bases de données ou en cas de projet de commercialisation. La rémunération des données, informations et échantillons biologiques fournis est un autre aspect important. Déterminer s’il faut ou non rémunérer les donneurs, au-delà du simple remboursement des frais de base, constitue une autre étape importante. Il faut pour cela se demander si la législation nationale ou régionale applicable autorise une telle activité, et si ce type d’approche risque d’affecter la crédibilité et la représentativité de la BRGH, etc.

Le consentement éclairé est l’un des problèmes les plus complexes que posent les bases de données de la recherche en génétique humaine. Le consentement éclairé est devenue la pièce maîtresse de la protection de l’autonomie de la recherche faisant intervenir des sujets humains. Dans le domaine médical/scientifique, le consentement éclairé suppose généralement la possibilité d’indiquer clairement au participant l’utilisation et la finalité du travail de recherche considéré. Si cela est faisable pour les projets de recherche ciblés, la nature même des BRGH fait qu’il est souvent difficile de fournir ce type d’information aux sujets pressentis. Par conséquent, la question est de savoir ce qu’il faut entendre par consentement éclairé dans le contexte d’une BRGH, sachant que le but de la collecte des données, information et échantillons biologiques et l’usage qui en sera fait ne pourront généralement être décrits qu’en termes généraux. De nombreux auteurs se sont demandé si le modèle traditionnel du consentement éclairé était applicable dans le contexte des BRGH ou s’il ne fallait pas plutôt créer un nouveau modèle/paradigme. Certains ont préconisé un modèle de consentement général. D’autres

24 – RÉSUMÉ


ont recommandé un consentement général initial pour certains usages, l’entreprise devant reprendre contact avec le participant pour obtenir son assentiment dès lors que d’autres usages sont envisagés.

Les autres questions soulevées par les BRGH dans ce domaine concernent le consentement des enfants et le consentement renouvelé. Lorsque la participation d’enfants à des études génétiques et à l’établissement de la BRGH a été approuvée, l’étude des conséquences d’une telle participation et des modalités d’obtention de l’assentiment sera primordiale. S’agissant des jeunes enfants, il pourra s’agir de recevoir l’assentiment des parents. Pour les enfants plus âgés, différentes approches pourront être envisagées en fonction de leur niveau de développement et de compréhension. Le consentement renouvelé pose aussi toute une série de problèmes notamment liés à la nécessité de recontacter le participant pour lui faire renouveler son consentement, par exemple pour autoriser de nouvelles utilisations, et à la possibilité de conversion de bases de données en BRGH, qui pourra impliquer, ou non, un renouvellement des consentements.

La nécessité de reprendre contact avec les sujets participants, dans les différents scénarios que l’on vient d’évoquer, ne va pas sans difficultés pratiques (la personne peut être décédée ou avoir déménagé) mais soulève aussi d’autres questions plus complexes (par exemple, la personne souhaite-t-elle ou non être recontactée). Il conviendrait peut-être dans ce contexte de définir les principes à suivre dans cette démarche. Il faudrait en outre se demander si les participants doivent être informés des possibilités de reprise de contact avant d’accorder leur consentement.

Après avoir accepté de participer à une BRGH, certaines personnes peuvent à un moment donné souhaiter se retirer d’une étude et faire détruire leur données, informations et échantillons biologiques. Il importe donc de déterminer si la BRGH acceptera que des sujets participants aient le droit de retirer leurs données, informations et échantillons. Il faudra donc bien préciser si cela est possible et dans quelles circonstances. Dans certains cas, les participants auront la possibilité de retirer leurs données, informations et échantillons biologiques de la BRGH toute au long de sa durée de vie. Dans d’autres, en revanche, selon la façon dont aura été mise en place la BRGH, ils pourront le faire avant que les données soient anonymisées. De plus, le droit de retrait peut revêtir plusieurs formes qu’il importera de définir. Par exemple, si les données d’une personne figuraient dans des informations communiquées à une tierce partie, il risque de ne pas être possible de les retirer ou de les extraire.

Les résultats et informations tirées des études épidémiologiques sont souvent communiqués aux sujets. Cependant, compte tenu de la taille des BRGH, on doit s’interroger sur la faisabilité et l’opportunité de rendre compte des résultats aux sujets. Premièrement, les résultats obtenus par les utilisateurs des données et échantillons doivent-ils être communiqués à la base de données ? Cela permettrait certes d’enrichir la base de données, mais se pose alors la question de la qualité des résultats fournis. Deuxièmement, en ce qui concerne la communication des résultats aux participants, compte tenu à nouveau de la taille et de la vocation des BRGH, la question est de savoir s’il est réaliste d’envisager une politique de communication des résultats aux sujets et l’intérêt d’une telle démarche, surtout si elle intervient hors du cadre clinique.

La formation des chercheurs et des professionnels de santé jouera un rôle important dans la réussite des BRGH. Les professionnels de santé chargés du recrutement des participants et de la collecte des données (interviews, questionnaires, examens médicaux, prélèvements sanguins, transfert des informations recueillies) ne seront pas forcément

RÉSUMÉ – 25


spécialistes de la recherche génomique. En conséquence, il sera important d’adopter une politique de formation de ces professionnels. On pourrait, par exemple, établir un protocole pour expliquer aux généralistes leur rôle et quelles informations divulguer aux sujets, et les guider dans la gestion de certaines situations délicates.

Gestion et gouvernance des bases de données

La gouvernance des bases de données fait intervenir toute une série de questions techniques, opérationnelles et juridiques, renvoyant notamment à la législation et à la réglementation applicables, au rôle des comités d’éthique et de surveillance, aux questions de compétences, droits et obligations des BRGH, à la sécurisation des bases de données, aux questions d’accès et à la suppression d’une BRGH.

Il importe de déterminer, lors de la mise en place d’une BRGH, si la création de la base de données doit être entérinée ou non par la législation. Par exemple, la Biobanque de l’Estonie et celle de l’Islande ont été approuvées par le parlement de ces pays, alors que la Biobank du Royaume-Uni et le projet canadien CARTaGENE existent indépendamment de toute loi habilitante mais sont subordonnées à différents textes existants. La création d’une base de données sanctionnée par acte parlementaire ou au contraire par un instrument non contraignant de type mémorandum d’accord, présente à la fois des avantages et des inconvénients.

L’examen de la plupart des projets de BRGH révèle qu’ils doivent tous être supervisés d’une façon ou d’une autre par un comité de surveillance dont la composition et la formation varient toutefois selon les cas. En établissant le régime de gouvernance des BRGH, il importe de définir le rôle, la fonction et la nature du comité de surveillance. S’agissant de la composition du conseil du comité de surveillance, on se demandera, par exemple, s’il doit être pluridisciplinaire, combien de mandats les membres du conseil pourront exercer au maximum et quelles stratégies ou approches adopter pour déterminer les questions qui devront être portées à l’attention du comité de surveillance. Par exemple, le Comité d’éthique de la Biobanque estonienne assure le respect des principes d’éthique et chacun est autorisé à s’adresser à lui pour demander l’accès à la base de données.

Les compétences et l’aptitude des BRGH à assurer le respect et la mise en application des décisions sont aussi des points importants à aborder. De l’étendue de ces prérogatives dépendra la possibilité d’assurer le respect des mesures de protection de la vie privée et de la confidentialité, et de garantir que l’entité privée détenant des droits de commercialisation respecte ses engagements et n’outrepasse pas ses droits. Par exemple, la base de données islandaise fonctionnerait sous licence et la possibilité de révoquer cette licence est un moyen d’assurer le respect des dispositions prévues. Si les conditions de la licence d’exploitation ou de l’acte habilitant ne sont pas respectées, le ministre peut émettre un avertissement écrit, et imposer la prise de mesures correctives dans un délai fixé. En cas d’inaction ou de faute lourde intentionnelle la licence peut être révoquée.

Compte tenu du risque d’utilisation abusive des données et échantillons conservés dans les BRGH, la sécurité de ces bases de données est primordiale. Cet aspect a des implications juridiques et techniques. Il importe de définir, connaissant l’objectif de la base de données, les meilleures méthodes qui permettront d’assurer la sécurité, de n’autoriser l’accès que selon les modalités autorisées, et d’assurer que l’accès aux données et échantillons n’est pas entravé. L’une des méthodes possibles consiste à tenir

26 – RÉSUMÉ


un registre des codes. Si les données peuvent être reliées à une information personnelle identifiante, le meilleur système de protection de la confidentialité des données est sans doute de conserver un registre de codes permettant d’anonymiser des données en ayant la possibilité de les retrouver. Le gardien officiellement en charge du registre devra gérer celui-ci en toute confidentialité. La personne occupant cette fonction sera assujettie à l’obligation de non divulgation des informations confidentielles ; il pourra s’agir par exemple d’un médecin. De plus, des ordinateurs autonomes pourraient être utilisés pour gérer les identificateurs individuels et les autres informations personnelles, notamment sur la santé, afin de réduire les risques de piratage sur un réseau.

On peut également choisir, pour assurer la sécurité et la protection de la confidentialité des bases de données génétiques, de limiter la quantité ou le type des données diffusées ou accessible aux chercheurs qui utilisent les bases. Cette option combine des solutions législatives et techniques. Par exemple, on peut décider de faire en sorte que les données/informations ne soient communiquées aux chercheurs qu’à partir d’une certaine masse critique d’individus. On peut aussi protéger la vie privée et la confidentialité en limitant ou en contrôlant l’accès aux données. Une façon simple de procéder est de n’ouvrir l’accès de la base qu’aux chercheurs possédant un mot de passe autorisé. Cette formule existe sous une forme plus sophistiquée consistant à utiliser un système à base de règles pour contrôler l’accès aux données, en remplacement ou en complément de l’intervention humaine. Avec ce système, les différents utilisateurs sont autorisés à accéder à différentes informations selon leurs fonctions. Il est aussi possible de n’autoriser qu’un très petit nombre d’analystes à interroger directement les données primaires. Dans ce scénario, les chercheurs « extérieurs » pourront accéder à ces données seulement de façon indirecte, par l’intermédiaire de ces analystes, et ne recevront que des réponses synthétiques à leur demande (par exemple, moyennes, valeurs prédictives, etc.).

La cryptographie, asymétrique ou à clé publique, peut être utilisée pour renforcer la protection des données. Cette méthode peut être utilisée en conjonction avec les autres méthodes décrites ci-dessus. Le cryptage étant assez facilement réalisable, on peut supposer que les données transférées dans les bases de données et extraites de ces bases seront cryptées d’une façon ou d’une autre. Force est de reconnaître toutefois que les données cryptées peuvent être décryptées, c’est pourquoi le cryptage ne peut garantir à lui seul la protection de la vie privée.

Sachant que l’objectif premier des BRGH est de faciliter la recherche, l’accès aux bases de données revêt une importance primordiale. Un certain nombre de questions doivent donc être résolues : qui doit avoir accès à la base de données (uniquement les chercheurs, les chercheurs du secteur public ou du secteur privé, etc.), comment ouvrir l’accès (directement ou par l’intermédiaire d’un chercheur interne), l’accès doit-il être libre ou payant (qui doit acquitter le droit d’accès et quel en sera le montant) et à quelles informations donner accès (toute la base de données, certaines parties seulement et lesquelles, uniquement certaines données et seulement sous forme anonymisée, etc.). Autre point à déterminer : quels motifs doivent justifier l’accès.

Bien qu’à l’heure actuelle il existe peu d’exemples de BRGH avortés ou sans lendemain (à part la base de données de Tonga qui devait être créée par Autogen Limited), il pourrait être utile d’étudier sans délai le cas de l’éventuelle suppression d’une BRGH, notamment du point de vue de sa gestion. Il faudra déterminer clairement les conséquences d’une telle suppression. Il conviendra de se demander, par exemple, si toutes les données et échantillons doivent être conservés ou détruits ; et si les sujets participants doivent être informés de la disparition de la base de données. Si la base de

RÉSUMÉ – 27


données est administrée par une entreprise privée, il faudra déterminer si des dispositions peuvent être prises pour que le gouvernement ait le droit de reprendre la base ou si un gouvernement peut se réserver au moins un droit de préemption. Il faudra pour cela examiner la législation applicable. Par exemple, de nombreux pays ont adopté une législation qui interdit la vente de tissus ou de matériels humains. Les conséquences de telles dispositions devront aussi être prises en compte.

Considérations relatives à la commercialisation

Les BRGH soulèvent de très nombreuses questions relatives à la commercialisation, notamment eu égard aux droits de propriété intellectuelle, aux modalités concrètes de commercialisation d’une base de données et au partage des bénéfices.

On entend généralement par propriété intellectuelle les droits conférés par une activité intellectuelle dans le domaine industriel, scientifique, littéraire et artistique. Les BRGH soulèvent toute une série de questions concernant les droits de propriété intellectuelle dans le cas de recherches ayant utilisé des données et matériels tirés d’une base de données. Dans ce cas, il faudra se demander qui est le détenteur de l’invention et qui a l’obligation de veiller à ce que les droits de propriété intellectuelle applicables soient protégés. Un autre aspect important est celui de l’accès à l’innovation dérivée des données et échantillons d’une base de données. Il pourrait être utile d’envisager une stratégie permettant à la fois de maintenir l’accès et d’obtenir un retour sur investissement. La base de données elle-même peut poser des questions de propriété intellectuelle, notamment les droits sur cette base, lorsqu’ils existent, la protection du droit d’auteur pour le logiciel et les autres droits permettant d’assurer le bon fonctionnement de la base de données.

En ce qui concerne la commercialisation, la première chose à prendre en considération est l’intérêt que présente la commercialisation de la base de données et si la commercialisation correspond aux attentes des sujets participants. Si l’exploitation commerciale de la base de données est approuvée, il conviendra de déterminer la procédure à suivre. Il importera notamment d’examiner si cette commercialisation doit se faire, ou non, sur la base de l’exclusivité. Si c’est le cas, il conviendra de veiller à assurer l’équité d’accès à la base de données et le respect de la législation sur la concurrence. Il sera essentiel de déterminer si la base de données peut ou non être vendue ou transférée moyennant contrepartie.

La question du partage des bénéfices est également complexe et présente de nombreuses facettes qui varient en fonction de la structure de la base de données. Par exemple, dans le cas d’une BRGH ayant le statut de partenariat public-privé ou d’entreprise privée, il conviendra d’établir si le gouvernement doit recevoir une compensation de l’entité privée et, dans ce cas, sous quelle forme. Par exemple, dans le cas de la Biobanque islandaise, le titulaire de la licence devrait, moyennant certains ajustements, payer au gouvernement i) un montant annuel fixe, destiné à financer la promotion des soins de santé et de la R-D ; et ii) 6 % des profits, plafonnés à 70 millions ISK par an. Si la compensation financière est l’option retenue, il sera important de déterminer comment seront utilisés les sommes dégagées. De plus, la compensation pourrait aussi prendre d’autres formes, notamment un soutien technique ou scientifique. Le partage des bénéfices pose aussi la question de savoir si les sujets participants

28 – RÉSUMÉ


pourront, ou non, bénéficier au niveau individuel de la base de données. Par exemple, les participants devraient-ils recevoir une partie des profits tirés d’une invention développée à partir des données, échantillons et informations contenues dans la base de données ? De même, il conviendra de déterminer si les donneurs pourront bénéficier d’autres avantages non monétaires, par exemple des produits développés grâce aux recherches menées à partir des données, échantillons et informations de la BRGH.

CHAPTER 1. INTRODUCTION – 29


Chapter 1

Introduction

1.1 Context

The development of biotechnology and bioinformatics affords the opportunity to store and analyse increasingly large amounts of genetic data. Genetic research involving the use of databases containing human genetic and genomic information, sometimes alone or in combination with other personal or medical information, has thus become increasingly important. More recently, new types of databases for genetic research are contemplated and being developed that are of a considerably different nature and larger magnitude. Many of these emerging databases focus on and include data, information and biological samples from populations. These population databases, also referred to as human genetic research databases (HGRDs), may contribute significantly to science’s understanding of the complex multi-factorial basis of disease (genetic and non-genetic components) and therefore to improvements in detection, prevention, diagnosis, treatment and cure. Such databases may also contribute significantly to the identification of genes associated with disease, an understanding of the frequency of genetic variants in particular populations, and to an improved understanding of the reasons for drug reactions (both positive and negative) and reactions to other environmental factors. Nevertheless, such databases also raise number of issues and concerns.

The potential contribution of such databases to knowledge of human health has resulted in heightened interest at the national, regional and international level in the governance and management of human genetic research databases. Over the past few years, there has been significant interest in fostering international co-operation in this field. In spite of this interest and the various concerns pertaining to such databases and in spite of international efforts, there is limited international guidance on the establishment, management and governance of human genetic research databases. Certain institutions, such as UNESCO and the Council of Europe, have developed instruments with respect to the use of genetic data.

In light of the limited guidance with respect to these issues, under the co-chairmanship and sponsorship of Japan and Canada, the OECD organised in February 2004 a workshop in Tokyo (“Workshop”) on the issues raised by human genetic research databases. The Workshop brought together experts from around the world with different backgrounds, training and experience.

The objective of the Workshop was to bring together OECD member countries and experts with a view to beginning the process of discussing the policy challenges associated with the broader question of the establishment, management and governance

30 – CHAPTER 1. INTRODUCTION


of human genetic research databases. It was intended that the Workshop would contribute to international consideration of the role such databases could play in translating scientific advances into innovation in health and provide a solid basis for evaluating future action in this area.

In the course of the Workshop, a significant number of complex issues raised by HGRDs were discussed. This Report aims to summarise and outline the discussions of these issues. In light of the breadth and complexity of the issues covered, this Report does not seek to exhaustively cover all aspects of each of these intricate issues or all of the relevant literature. Moreover, many of the databases discussed in this Report may be subject to national, regional or international legislation, regulation, policies, etc. While some of these national, regional or international legislation, regulation, policies, etc., are mentioned in this Report, they are examined only for illustrative purposes when discussing the relevant database. This Report does not aim to provide an exhaustive examination of all the national, regional or international legislation, regulation, policies, etc., that may be applicable to such databases.

The OECD is involved in this field to stimulate international consideration of the policy challenges associated with the establishment, management and governance of human genetic research databases and to determine approaches for addressing and managing some of these challenges.

1.2 Overview of issues

As mentioned above, human genetic research databases raise a number of issues and concerns. Different databases illustrate the numerous considerations raised by their establishment. Selection of the population to be covered by the database is fundamental. The targeted population may be that of an entire country, as in the case of the Estonian or Icelandic projects, may be a selection of a country’s population, as in the case of the UK Biobank or the Canadian CARTaGENE endeavours, or may be a selection based on specific criteria such as the TgRIAD project, aimed at individuals of African descent, or the GenomeEUtwin, aimed at analysing twins and other cohorts.

While some of the issues and concerns are not new, the increasing breadth and scope of such databases amplify them. Moreover, the combination of genetic data and personal information in these databases raises new issues relating to the use of such information, especially in a non-clinical or non-research context. For example, concerns arise with respect to the use of such information for law enforcement purposes or within the insurance or employment context. The question arises of what are the appropriate uses of these databases?

HGRDs are subject to national and regional laws and international instruments, including human rights and constitutional frameworks. Consequently, the challenge at the international level is to arrive at approaches that are consistent with national and/or regional legislation and international instruments while addressing the concerns raised by these databases.

The size of the database will have numerous implications, including for the collection and storage of data and biological samples, for the management of the database, for ensuring security and confidentiality. The UK Biobank aims to collect samples from approximately 500 000 volunteers, the Estonian Project aims to collect samples from approximately 100 000 individuals, and the CARTaGENE endeavour aims to include 50 000 individuals. The size of the population database will be linked to and affected by



the objectives of the endeavour. For example, the Estonian Genome Project aims not only to enable research on the genetic and non-genetic components of common diseases and diseases more prevalent in its population but also to create biological descriptions of a large and representative sample of the Estonian population (LD map donors).

The diverse nature of and structures adopted for the establishment of population databases will also have implications for their management, operation and governance. The diverse databases have or had adopted different structures. For example, the Icelandic Health Sector Database was intended to be established as a private/for-profit endeavour, which would have been undertaken by the Icelandic company deCODE Genetics. A public-private partnership (PPP) was initially adopted for the Estonian Genome Project, between the Estonian Genome Project Foundation and the company EGeen Limited, but was eventually terminated. A third approach is that of a public/not-for-profit undertaking, used for example by the UK Biobank or the GenomeEUtwin initiative. Moreover, different mechanisms may be adopted for establishing such databases. For example, the Canadian CARTaGENE initiative relies on a number of Statements of Principles and on existing legislation. Conversely, other initiatives, such as the Estonian Genome Project and the Icelandic Health Sector Database, are the product of specific, often detailed, legislation.

Irrespective of the nature and structure of the database, issues such as confidentiality/privacy, consent, public communication and engagement, access, commercialisation, intellectual property rights and benefit sharing need to be considered by all HGRDs.

The collection of a large number of data and information about a given individual raises numerous privacy and confidentiality issues, which are also tied to issues of security. Privacy, in the genetic research context may imply the right to know, but it may also mean the right not to know genetic information. Genetic information obtained in the research context may raise unusual privacy concerns due to its potential to generate information and knowledge beyond that which was originally sought, and it may create an obligation, in some situations, to provide that information back to the patients who contributed DNA samples.

Confidentiality, in the context of genetic research, relates to keeping genetic information that is collected in a research/clinical setting from third parties such as health insurers or employers. In the computer age, corollary concerns include the potential for violation of databases and the sale of genetic information for marketing purposes.

Other factors that influence the issue of privacy and confidentiality include the nature of the biological samples, data and information collected, the specificity or generality of the database, and the security systems established.

Informed consent is one of the most complex issues for human genetic research databases. Informed consent has become the pillar for protecting autonomy in research involving human subjects. Within the medical/scientific field, informed consent generally presumes the ability to indicate clearly to the participant the use and purpose of the particular research activity. While this is feasible for purpose-specific research, the very nature of HGRDs renders the provision of this type of information difficult. Therefore, the issue becomes what constitutes informed consent within the context of an HGRD given that the purposes for which the data, information and biological samples are collected and the uses for which they may be employed, usually, may be described only in a general manner. Many have queried whether the traditional model of informed

32 – CHAPTER 1. INTRODUCTION


consent is applicable in the context of HGRDs or whether a new model/paradigm should be developed. Additional consent questions raised by HGRDs include children’s consent and renewed consent. In situations where a determination has been made about the involvement of children in genetic studies and HGRDs, consideration of the consequences of their involvement and the manner in which to obtain consent is primordial. Issues that arise in the context of renewed consent include returning to the participant to obtain consent for new uses in the context of the HGRD.

In establishing a population database, public support is essential, given that participation is voluntary. This implies that the collection of biological samples, data and information and the inclusion in the HGRD of these depend on the consent of the donor. It may be difficult to estimate participation rates in projects where benefit is indirect, long-term and at the population level, especially in resource-poor communities or populations with different beliefs, cultures or languages. In order to build public trust, it will be important to bridge the distance between the research community and participants. However, the manner in which public support is elicited may vary considerably. In seeking to establish an HGRD, the methods employed to engage the public, the information that should be provided, and the manner for communicating it most effectively are important factors.

Given that the fundamental purpose of HGRDs is to foster research, access to the database raises crucial issues. Key questions for consideration include who should have access to the database (e.g. only researchers, public- and/or private-sector researchers), the manner in which access should be given (directly versus via an internal researcher), whether access should be free or for a fee (who should pay the fee and what should it be), and to what should access be given (e.g. to the whole database, to parts and which ones, only to certain data and then only in an anonymised manner). Another key issue is the purpose for which access should be granted.

With respect to commercialisation, one consideration is whether or not it is desirable to commercialise the database and/or whether commercialisation is in line with the participants’ expectations. If commercial exploitation of the database is undertaken, consideration should also be given to the manner in which it should be carried out. Should the database be commercialised on an exclusive or non-exclusive basis? If commercial exploitation is to be undertaken on an exclusive basis, how can fair access to the database and how can compliance with competition law be ensured? Should there be the possibility for the database to be sold or transferred in exchange for monetary consideration?

Another set of issues raised by HGRDs pertains to intellectual property rights. Intellectual property issues may arise with respect to the actual database, including database rights, where they exist, copyright protection for the software and other rights that ensure that would allow the database to operate effectively. Issues also arise with respect to those rights resulting from research employing data or biological samples accessed from such a database. In such circumstances, the question of who owns the invention and/or who is under the obligation to ensure that the relevant IPRs are protected arises. There is also the issue of access to a follow-on innovation developed using data and biological samples from the database. The policy adopted with respect to such issues may be influenced by the structure (e.g. private, public, or mixed) chosen for the establishment of the database.



The issue of benefit sharing is a complex one, many aspects of which vary with the structure of the database. For example, in the context of an HGRD established as a public-private partnership or as a private undertaking, it would be useful to consider whether or not the government should be entitled to some compensation from the private entity and if so, the form such compensation should take. If monetary compensation is the option favoured, for what purpose should these monies be employed? Compensation may equally take the form of technical or scientific support. Benefit sharing also raises the issue of whether or not participants should be entitled to individual benefits arising from the database. For example, would participants be entitled to share in the profits of a successful invention developed using data and information accessed from the database? Should participants have the right to access other, non-monetary benefits, for example, the products developed as a result of research involving data and samples from the HGRD?

These and other issues are examined in greater depth in the following chapters of this Report through the examination of different examples and approaches.

CHAPTER 2. HUMAN GENETIC RESEARCH DATABASES – 35


Chapter 2

Human Genetic Research Databases*

Human genetic research databases (HGRDs) can help to increase scientific understanding of the complex multi-factorial basis of common diseases (genetic and non-genetic components) and therefore to improve detection, prevention, diagnosis, treatment and cure. They can contribute significantly to the identification of genes associated with disease, of the frequency of genetic variants in particular populations, and of the reasons why different individuals respond differently to drugs or to other environmental factors. The development of biotechnology and bioinformatics affords the opportunity to store genetic data for analysis. For this reason, genetic research involving the use of databases containing human genetic and genomic information from many individuals, sometimes alone or in combination with other personal information, is increasingly important. The latter information may include information such as names, addresses or hospital identification numbers, medical information such as diagnoses and prescribed drugs, or other research information such as pharmacokinetic data.1 National,2 regional3 and international4 communities have shown heightened interest in and given greater attention to the collection,5 processing and storage of genetic data and to the management of genetic databases. The last few years has seen considerable activity in this field, including the development of international instruments6 and the launching of projects7 aimed at fostering international co-operation, framing research and harmonising ethical and legal guideposts for the management of human genetic databases.8

HGRDs have also been referred to as biobanks,9 population databases,10 gene banks,11 cohorts,12 and genome databases.13 A possible definition for HGRDs is “any collection of samples from which genetic samples can be derived and related data (e.g. genealogical, clinical, etc.) organised in a systematic way and used for purposes of research”.14 Existing HGRDs according to such a definition would be rather diverse in respect of the population included (affected or/and unaffected individuals, specific or general community), the nature and size of the biological samples and related data collected, the time of collection (clinical or research settings), the form of storage, the underlying scientific purpose (from screening programmes, association studies, and genetic epidemiology to pharmacogenetics or pharmacogenomics), and the structure of the database (public, private or mixed).15 Since there is no commonly accepted terminology and as the terms referred to above connote different meanings, this Report, for purposes of convenience, will employ the terms population databases or HGRDs interchangeably.

This chapter also draws on the background documents prepared for the workshop by Mildred Cho, Yoshinao

Katsumata, and Clementine Sallé. See Bibliography.

36 – CHAPTER 2. HUMAN GENETIC RESEARCH DATABASES


In this fast-evolving field, the number and types of HGRDs continue to grow. For example, the 14 April 2005 edition of Nature16 contained an article on a new population database initiative. In light of the continuous growth of population databases, this Report will not aim to comprehensively review all population databases. Rather, a few databases are reviewed and employed as examples throughout the Report. This selection was based not only on the available sources of information (in terms of quantity, quality and language) but also on the specific issues raised by these databases. HGRDs which have been abandoned17 or terminated are not considered.

2.1 Examples of human genetic research databases

This section examines the nature, objective, funding and structure, describes the population studied as well as the type of activity envisaged for a selection of HGRDs. In order to provide a breadth of examples of HGRDs, a number of diverse initiatives were selected for overview below and are employed throughout this Report.

2.1.1 The Personalized Medicine Research Project18 (“Marshfield Project”)

Launched in 2002 by the Marshfield Medical Research Foundation,19 the Personalized Medicine Research Project (PMRP) has as its objective the creating of a Personalised Medicine Research Database, an HGRD, to study the effect of genetic variation on diseases and medication and, ultimately, the provision of personalised medicine (health-care system improvement). The PMRP aims to collect genetic data from blood sample analysis, medical information contained in medical records and environmental, background and family information from a questionnaire from over 40,000 volunteers over 18 years of age, and from a geographical area20 identified by 24 postal codes21 within the state of Wisconsin. Currently, there are over 18,000 participants involved in the Marshfield project. The Centre expects the research project to last at least 20 years in order to allow some longitudinal studies.

Carried out over three phases and several years, the project will i) create the DNA foundation of the personalised medicine database and build the bioinformatics tools to store securely and analyse genotypic and phenotypic data, ii) create the phenotypic database, establish the scientific and administrative infrastructure to support genetic mapping of the DNA and the initial discovery projects, and genotype a sufficient portion of the genetic samples to support these discovery projects, and iii) expand the discovery projects, complete the genotyping of the genetic database and expand physician/health care provider education and community consultation.

The Project’s website22 specifies the confidentiality approach that will be applied to the information and data collected, the security features that will be employed to safeguard the information and data, provides a right to withdraw, and outlines the approach to any benefits that may accrue, both monetary and non-monetary ones.

2.1.2 CARTaGENE23 (“CARTaGENE”)

The CARTaGENE project, an HGRD of the Quebec population, was initiated in 1999 by a multidisciplinary team of the Quebec Network of Applied Genetic Medicine (RMGA).24 CARTaGENE’s aim is to create a database of genetic, physiological, medical and social and environmental data and a biobank for blood samples in order to draw a first map of the genetic diversity of the population of Quebec, Canada. CARTaGENE is under development as it is still awaiting ethics approval and approval from the



Commission d’accès à l’information (CAIQ) – the privacy commission of Quebec.25 When it is underway, CARTaGENE aims to randomly recruit 50 000 individuals from the ages of 25 to 69 (representing approximately 1% of the Quebec population) from the diverse regions of Quebec.26 As Figure 2.1 illustrates, it is intended that the participants will be recruited by an independent entity (Santé-Québec) on behalf of the CARTaGENE team and that they will be randomly selected from the Régie d’assurance maladie du Québec’s (RAMQ) large data bank containing the personal information of all Quebec residents who have a health insurance card.

Figure 2.1 CARTaGENE’s Governance and Organisational Structure27



This initiative aims to be semi-longitudinal and the draft consent form refers to a period of 50 years. Participants will need to “agree to allow information on their health events found in the registers of public organisations such as the RAMQ and the Hospitalisation Insurance to be disclosed to CARTaGENE on a regular basis”.28 It is stipulated, that at no time will CARTaGENE have access the medical records of participants. Each participant will be required to complete a detailed health questionnaire which will cover medical history and lifestyle, will need to provide a blood sample, and will submit to physiological measurements.29 The website also specifies the manner in which the security and confidentiality of the information will be ensured (i.e. coded information), a right for participants to withdraw from the study and its modalities, as well as a detailed consent form.30 The initiative will be overseen by an independent entity, the Institute for Population, Ethics and Governance (IPEG), which will ensure governance and ethics.

2.1.3 Genome Database of the Latvian Population31 (“Latvian Genome Project”)

The stated aim of the Latvian Genome Project is to create a unified national network of genetic information and data processing, to collect representative amount of genetic samples for genotyping of the Latvian population and to compare genomic data with the clinical information and the information available about specific pedigrees. It is anticipated that the direct outcome of the project for each individual will be seen in the possibility to consider his/her risks to develop certain diseases due to their genetic features and to eliminate these risks, particularly by application of the individual therapy based on genetic characteristics of the patient. It is intended that the project of establishing a genome database of the Latvian population will be carried out over a period of ten years and in three stages. Launched in January 2001 and funded by the Latvian Council of Science, the pilot phase, currently underway, has collected samples and information from over 2 700 participants.

While this project has numerous objects, generally it purports to study DNA polymorphisms in the Latvian population and their ethnogenesis in order to improve scientists’ knowledge of monogenetic and multi-factorial diseases and ultimately to generate personalised prevention, diagnosis and treatment tools. Scientists from various medical and educational institutions (the Latvian Medical Academy, the State Centre of Medical Genetics, the University of Latvia and the Biomedical Research and Study Centre) will be involved in the studies. Specialised clinical centres have also opened the doors to their diagnostic facilities.

Samples as well as medical, genealogical and lifestyle information will be collected from the inhabitants of Latvia and coded. All personal identifying information will be coded and the key will be kept by the State Genome Register. Whether or not the database will be connected or linked to other HGRDs or more traditional databases is currently unknown.

The basis for this wide scale genetic research in Latvia was established by the Human Genome Research Law, adopted in 2002.32 Section 2 states that the purpose of this Law is “to regulate the establishment and operation of a unified Genome Database of the Latvian population, the genetic research, to provide the voluntary nature and confidentiality of gene donation regarding the identity of gene donors, as well as to protect persons from misuse of genetic data and discrimination related to the genetic data”.



2.1.4 Icelandic Health Sector Database (“Icelandic HSD”)

In 1998, the Icelandic parliament, the Althingi, enacted an Act on a Health Sector Database33 (“HSD Act”) to enable the creation of a database containing medical, genetic and genealogical data for the Icelandic population. The database was to be designated as the Health Sector Database (HSD). The aim of this initiative was to increase knowledge in order to improve health and health services. The data within the database could be used to develop new or improved methods of achieving better health, prediction, diagnosis and treatment of disease, to seek the most economic ways of operating health services, and for reporting on the health sector.34 Pursuant to the HSD Act, in order for the HSD to be created a licence was to be granted by the government to a licensee. The Icelandic government granted such a licence to the for-profit company, deCODE Genetics35 (deCODE).

Pursuant to the HSD Act, the Licensee, deCODE, could incorporate into the database any data and information derived from medical records provided that it was obtained with the consent of the health institutions or self-employed health workers.36 If a patient did not want their data entered into or removed from the HSD, they had to specifically request that it be removed. Thus, the HSD Act created a presumption of consent to patients’ data being incorporated into the HSD. This approach created immediate controversy,37 culminating in a constitutional court case.

A case involving the Act on a Health Sector Database was brought before the Supreme Court of Iceland in 2003.38 At issue in this case was whether an adult child could request from the Icelandic Director General of Public Health that information pertaining to her deceased father not be included in the HSD.39 The Icelandic Supreme Court essentially decided this case on the basis of the Icelandic Constitution. The Court affirmed that Paragraph 1 of Article 71 of the Constitution applied to extensive information contained in medical records and that this provision guaranteed protection of privacy in this respect. Furthermore, it stipulated that to ensure this privacy the legislature must ensure, inter alia, that legislation does not result in any actual risk of information of this kind involving the private affairs of identified persons falling into the hands of parties who do not have any legitimate right of access to such information, irrespective of whether the parties in question are other individuals or governmental authorities. After a detailed examination of the provisions of the Act on a Health Sector Database, the Court concludes that it is impossible to maintain that these provisions will adequately ensure, in fulfilment of the requirements deriving from the Constitution, attainment of the objective of the Act of preventing health information in the database from being traceable to individuals. The Icelandic government intends to introduce legislation to amend the Act on a Health Sector Database, in light of the Supreme Court’s decision.

As a result of the controversy surrounding the establishment of the HSD, the Supreme Court decision, as well as possibly other considerations, such as economic factors, the HSD has never been established. Consideration of the Icelandic example throughout this Report will be based on the legislation, regulation as well as the constitutional court decision.

deCODE Genetics has developed its own proprietary databases containing genetic, phenotypic and genealogical data. The genetic and phenotypic data are gathered from individuals who must provide their written informed consent and according to traditional scientific research protocols. The genealogical data are assembled mainly from public domain sources.40 deCODE is using this data in order to identify the genetic causes of common diseases with a view to improving their treatment and diagnosis. deCODE’s



research protocols are reviewed by the National Bioethics Committee. Moreover, deCODE employs a third-party encryption system. This implies that genetic, phenotypic and genealogical data of an individual is anonymised by the Icelandic government’s Data Protection Authority (DPA) so that it is identifiable only with a PIN code created by the DPA, who also maintains the key.41 This enables the data to be analysed without any personally identifiable data being within the company’s research process. deCODE is employing this approach to carry out research in approximately 50 common diseases.42

2.1.5 Estonian Genome Project 43(“Estonian Project” or “EGP”)

The Estonian Genome Project Foundation (EGPF) is a non-profit organisation founded by the Government of Estonia in 2001. The EGPF carries out the Estonian Genome Project (EGP) with the goal to create a database of health, genealogy and genome data that would comprise a large part of the Estonian population.44 The Human Genes Research Act,45 which entered into force on 8 January 2001, regulates the conditions for the establishment and maintenance of the Gene Bank of the Estonian Genome Project as well as the ethical principles to follow in terms of recruitment, collection, protection of participants’ privacy and data security. The legal framework for the activities of the EGP consists of the Constitution of the Republic of Estonia, the Human Genes Research Act (HGRA), the Personal Data Protection Act, the Databases Act and the Council of Europe Convention on Human Rights and Biomedicine.46

The goals of the EGP are to create a population database founded on the collection of health care status descriptions and tissue samples, to carry out scientific research in order to find the genes that cause and influence common diseases, and to implement in public health the knowledge derived from genomics research. The aim is to collect data and tissue samples from 100 000 gene donors by the end of 2007.47 In order to contribute, participants are required to sign a consent form, complete a questionnaire and provide a blood sample. Personal information on the questionnaires delivered to the Gene Bank are separated and replaced with a 16-digit code in the coding centre. Health data that has been separated from personal information are stored in the database of the Gene Bank. DNA is extracted from the blood sample in the laboratory of the Gene Bank. Separated DNA is placed in the storage facility of the Gene Bank. On the basis of DNA preserved in the storage facility, it will be possible to prepare personal LD maps of gene donors in the future.

The Estonian Genome Project, which could be viewed as a medical care model, aims not only to enable research on the genetic and non-genetic components of common diseases, particularly in pharmacogenomics, but, as mentioned, also to create biological descriptions of a large and representative sample of the Estonian population (LD map of donors). This would enable the marketing of the project through software development, the education of the public in the fields of genetic and genomic research and biotechnology and ultimately the use of the knowledge to benefit public health.48

While the project was initially envisaged as a public-private partnership, through collaboration between the Estonian Genome Project Foundation and EGeen Limited,49 a for-profit company, this collaboration was terminated in December 2004.50 EGeen had benefited from an exclusive commercial licence and would have returned a percentage of its profits to the EGP. The termination of this collaboration will signify that the Estonian Genome Project Foundation will need to procure financing for the Gene Bank through other sources and that EGeen will loose its exclusive rights and will have access to the information, data and samples of the Gene Bank on the same basis as other companies.



2.1.6 The United Kingdom Biobank51 (“UK Biobank”)

The UK Biobank results from collaboration amongst the Wellcome Trust,52 the Medical Research Council,53 and the Department of Health54 (as co-founders and co-funders). Begun in June 1999, the elaboration of the project was preceded by an examination of its feasibility by an expert working group (issues addressed included the nature and organisation of the collection, scientific rationale, size and population to be studied) and through public consultations.55

The UK Biobank is a medical research study of the impact on health of lifestyle, environment and genetics. The purpose of the UK Biobank project is to provide a resource for research with the aim of improving the prevention, diagnosis and treatment of illness and promoting health throughout society for public benefit.56 While there were pilot projects carried out over the past couple of years, it is intended that the full project will get underway in 2006, when the Biobank will begin to gather information on the health and lifestyle of 500 000 volunteers aged between 40 and 69.57 This age group was selected for study as it involves people at risk of developing complex, multi-factorial diseases – including cancer, heart disease, stroke, diabetes, dementia – over the next few decades.

The project will follow their health for a number of years, collecting environmental and lifestyle data, information from medical and other records and biological samples (e.g. blood and urine). The Biobank will contain coded (reversibly anonymised) data. All identifying information will be kept separate.58

The UK Biobank is centrally managed by a charitable company, UK Biobank Limited, which coordinates the activities of six regional collaborating centres.59 An independent body, the Ethics and Governance Council, has been established and will act as an independent guardian of the project’s Ethics and Governance Framework.60 While the Ethics and Governance Framework is intended to remain a live document that will continue to evolve, it was handed over by the Council to the UK Biobank late 2005.61

2.1.7 Translational-Genomic Research in the African Diaspora62 (“TgRIAD”)

The Howard University’s National Human Genome Centre (NHGC), a private non-profit institution, announced its intention to develop a biobank covering the African Diaspora. Formerly designated as GRAD, the aim of the Translational-Genomic Research in the African Diaspora (TgRIAD) is “to understand the implications of genetic and environmental variation for differential disease distribution and variable drug response within the context of the historic and cultural experiences of African Americans and other populations of Africa and the Diaspora”.63 It is intended that this initiative will cover the populations of southern Africa, eastern Africa, western Africa, the Caribbean, the United States of America and possibly Brazil. It is intended that entire households, including children, will be the population of this database.64

The proposed infrastructure would allow researchers to study common diseases (for example, asthma, high blood pressure, breast cancer) prevalent among the African Diaspora and to improve their prevention (onset and progression), diagnosis and treatment. This project assumes that disease susceptibility may be explained by ethnic genetic factors and seeks to ensure the representation of the African-American community in genetic and genomic research and to improve its health. The intention is that for each participant biological samples, from which DNA will be extracted, will be taken and genealogical, medical and other phenotypic information be collected via a



questionnaire on their family and medical history. It is not intended that information contained in medical records will be included in the database.65 This initiative is established as a private initiative and not pursuant to specific legislation.

In 2003, a pilot project was launched with participants from Kenya, Nigeria and the United States. Currently, the pilot project has enrolled approximately 3000 participants.66 Howard University is revising the TgRIAD initiative as substantial funding will be required, and is not rapidly forthcoming.

2.1.8 GenomEUtwin67 (“GenomeEUtwin Project”)

The GenomEUtwin project is co-ordinated by the National Public Health Institute68 and the University of Helsinki69 and received funding under the European Commission’s “Quality of Life and Management of the Living Resources” of the 5th Framework Programme.70 This project aims to apply and develop new molecular and statistical strategies to analyse unique European twin and other population cohorts to define and characterise the genetic, environmental and life-style components in the background of health problems like obesity, migraine, stature, coronary heart disease, stroke and longevity (Figure 2.2).71

Figure 2.2 A schematic presentation of the strategy of GenomEUtwin72

The population cohorts for the GenomEUtwin project are derived from a multitude of sources including the Danish, Dutch, Finnish, Italian, Norwegian and Swedish twins cohorts and the MORGAM population cohort.73 In addition to analysing twins and other



cohorts to assess the influence of genetic and non-genetic factors, it aims to create synergies between scientists and research programmes in genetic epidemiology (e.g. collaboration, training of young researchers).74 The project will collect and aim to make available the epidemiological, phenotypic and genotypic information on the different population cohorts accessible to investigators worldwide.75

The project is well under way. Tens of thousands of DNA samples with the related informed consent for genetic studies of common diseases have already been collected and stored from the population-based twin cohorts.76 The study samples will be analysed in four intellectual core facilities using accumulated expertise by the various partners in genetics, epidemiology and biostatistics. The core facilities are DNA isolation and genotying (Helsinki, Uppsala), epidemiological expertise (Odense), database expertise (Stockholm) and biocomputing expertise (Leiden).

2.1.9 The International HapMap Project77 (“HapMap Project”)

The International HapMap Project was launched in October 2002 as a USD 100 million private-public endeavour.78 The International HapMap Project79 is a multi-country effort80 to identify and catalogue common genetic variants that occur in human beings. It describes what these variants are, where they occur in our DNA, and how they are distributed among people within populations and among populations in different parts of the world. The International HapMap Project is not using the information to establish connections between particular genetic variants and diseases. Rather, the Project is designed to provide information that other researchers can use to link genetic variants to the risk for specific illnesses, which will lead to new methods of preventing, diagnosing, and treating disease.

Although most of the common haplotypes in human chromosomes occur in all human populations, any given haplotype may be more common in one population and less common in another, and newer haplotypes may be found in just a single population. In order to efficiently choose the tag SNPs needed to identify haplotypes requires looking at haplotype frequencies in multiple populations. In recognition of this significant factor, the International HapMap Project is collecting DNA from diverse populations, including those with African, Asian, and European ancestry.81 The DNA samples for the HapMap Project come from a total of 270 people. The Yoruba people of Ibadan, Nigeria, provided 30 sets of samples from two parents and an adult child (each such set is called a trio). In Japan, 45 unrelated individuals from the Tokyo area provided samples. In China, 45 unrelated individuals from Beijing provided samples. Thirty US trios provided samples, which were collected in 1980 from US residents with northern and western European ancestry by the Centre d’Etude du Polymorphisme Humain (CEPH).82

No medical or phenotypic information was collected, but samples are discernable in terms of sex and population (limited identifiers). The individual DNA samples used in the Project are identified as coming from a male or a female, from one of the four populations participating in the study, and, in the case of the parent-child trios, from either one of the parents or the child. The samples are anonymous with regard to individual identity. Samples cannot be connected to individuals, and no personal information is linked to any sample. As an additional safeguard, more samples were collected from each population than were used, so no one knows whether any particular person’s DNA is included in the study. Since the samples include no personal identifiers, the risk to individual donors that privacy might be compromised is considered minimal. All data generated by the Project will be released into the public domain.83



A Community Advisory Group (CAG) was established at each site where new samples were collected. Each CAG will serve as liaison between the non-profit Coriell Institute for Medical Research,84 where the samples are being stored as cell lines, and the donor community. Through newsletters (translated into the languages of the donor community) and quarterly reports, the Coriell Institute will keep the CAGs informed about the general progress of the HapMap Project, how the HapMap is being used, and how the stored samples are being used.

The Populations, Ethical, Legal and Social Implications Group,85 composed of a group of geneticists and experts in the ethical, legal, and social implications of genetics research, proposed an informed consent form template, which was subsequently modified by the researchers for use in Nigeria, Japan, and China, so as to be culturally appropriate locally. The modified consent forms were also translated into Yoruba, Japanese, and Chinese, and approved by the relevant local ethics committees.

2.1.10 COGENE86 (“COGENE”)

With the objective to co-ordinate and further genomic research related to human health in Europe, the European Commission created, in November 2000, the “Forum of Genome Programme Managers”87 with representatives from 25 countries. Funded by the European Union under the Quality of Life Programme of the 5th Framework Programme (FP5), the associated “Co-ordination of Genome Research across Europe” scheme (COGENE) aimed to implement the Forum’s objectives through a web service that listed the European HGRDs and offered data, organised by country, on genome research in the European Union (infrastructures, major funding agencies, and genomic research projects). Furthermore COGENE organised workshops and established a platform for discussion, in order to inform the public and enable individuals from various fields and perspectives to address important issues on a number of topics (e.g. biobanks). Administered by the Academy of Finland, COGENE ended in 2004.

2.1.11 P3G – Public Population Project in Genomics88 (“P3G”)

Similar to the European COGENE initiative, the Public Population Project in Genomics (P3G) is an international endeavour in which a number of HGRDs (with their own independent governance structure, objectives, etc) have agreed to collaborate. P3G’s stated objective is to foster collaboration between researchers and projects in the field of population genomics. Its contribution is to develop research tools and networks for effective collaboration between biobanks and to facilitate knowledge transfer. Initiated in 2003,89 P3G is funded by its partners (up to 75% of the overall costs) and by Genome Canada.90 Thus far, P3G has formed three main international working groups: i) Social, environmental, biochemical and investment, ii) Knowledge, curation and IT, and iii) Ethics, governance and public engagement.91

2.2 What is a human genetic research database?

The linchpin question is the determination of which databases should be considered human genetic research databases. There exist numerous different types of databases, including nucleotide sequences, sequences variations, mutation sequences, gene expression, gene loci, protein structures, as well as model organisms and diseases (i.e. pathology databases). However, this key question is much broader.



The analysis of genetic variations for various populations or communities (affected and non-affected individuals) via large-scale population databases or HGRDs is being undertaken in numerous jurisdictions. These databases, collections of biological samples and data from whole or subsets of populations or communities, constitute more a research infrastructure than hypothesis-driven protocols. The consideration is whether only these types of database should be considered HGRDs. In other words, is the element of containing information for “populations” or for large subsets of populations an essential criterion? A sub-element of this consideration is the issue of what constitutes a population. For example, the Estonian, Latvian and Icelandic projects aim to cover the populations within the jurisdiction of Estonia, Latvia and Iceland, respectively. Conversely, the GenomeEUtwin and HapMap initiatives aim to cover cohorts of twins from numerous jurisdictions and cohorts from selected jurisdictions, respectively. The criteria for defining what constitutes an HGRD could also take into consideration the aim of the database. For example, consideration could be given to whether the aim of allowing research on multiple diseases and conditions should be important. Another consideration is whether the characteristic shared amongst many databases that the overall benefit from the establishment and use of these databases be realised at the societal or population rather than individual level is significant. In ascertaining the essential aspects of an HGRD, another consideration is whether the private biobanks/tissue banks being amassed by many private sector entities should be included.

Another key consideration is the content of the database. This includes consideration not only of the type of biological samples or information to be collected and stored but also the source of that data and information. Genetic data have been broadly defined as “all data, of whatever type, concerning the hereditary characteristics of an individual or concerning the pattern of inheritance of such characteristics within a related group of individuals”. Genetic data do not necessarily include information derived from DNA or RNA specimens; they may also be inferred from family history, medical records or phenotype. Genetic data may be distinguished from personal genomic data which have been defined as “detailed personal data derived from analysis of DNA specimens”. Should HGRDs be considered to include databases only containing genomic data or should this designation include databases containing personal, medical and other data?

Consideration should also be given to whether HGRDs should designate those databases that contain personally identifiable information or only databases that contain data and information that cannot be associated with or result in the identification of individuals. Personal data have been defined as “any information relating to an identified or identifiable individual”. Genetic data may be collected from individuals through different approaches, permitting identification to varying degrees. This complex issue is related to those raised by anonymisation and linkability of data, which are considered separately.



Notes

1. R.E. Klein, J.T. Chang, M.K. Cho, K.L. Easton, R. Fergerson, M. Hewett, Z. Lin, Y. Liu, S. Liu, D.E. Oliver, D.L. Rubin, F. Shafa, J.M. Stuart and R.B. Altman (2001), “Integrating genotype and phenotype information: an overview of the PharmGKB project”, The Pharmacogenomics Journal, Vol. 1, pp. 167-170.

2. In its “Population-Based Large-Scale Collections of DNA Samples and Databases of Genetic Information” Report, the Israeli Bioethics Advisory Committee of the Israel Academy of Science concluded that Israel needed genetic database(s) of 10 000 to 100 000 samples from homogenous populations, linked to medical records and overseen by an independent public body (the Human Genetic Israeli Collection Authority, HUGIC). Consent for participation would be specific to genetic studies analysing the interaction between genes and environment. In their report, the Committee outlined the ethical rules and principles that population HGRDs would have to abide by, distinguishing for each issue, when applicable, between public and commercial collections (in conformity with the Israel Government, “Genetic Information Law”, 5761-2000”, 13 December 2000. The Bioethics Advisory Committee of The Israel Academy of Sciences and Humanities, Population-Based Large-Scale Collections of DNA Samples and Databases of Genetic Information, Jerusalem, December, 2002, http://stwww.weizmann.ac.il/bioethics/new-e/PDF/Finalized_Dna_Bank_Full.pdf (accessed 9 May 2006).

Similarly the Australian Law Reform Commission states that “[t]he inquiry has concluded that new regulation of human genetic research databases is necessary” (Part E. § 18.6). It further recommends the creation of a system of registration, the adoption of new rules to provide guidance for database management (not limited to research databases (e.g. law enforcement)) and participants’ protection (gene trustees). It is supervised by the National Health Research Council (NHMRC). Australian Law Reform Commission, Essentially Yours: The Protection of Human Genetic Information in Australia, Sydney, 29 May 2003, www.austlii.edu.au/au/other/alrc/publications/reports/96 (accessed 9 May 2006).

3. See e.g. European Society of Human Genetics (2001), “Data Storage and DNA Banking for Biomedical Research: Technical, Social and Ethical Issues, Recommendations of the European Society of Human Genetics”, Birmingham, November, www.nature.com/ejhg/journal/v11/n2s/index.html (accessed 9 May 2006); see also Council of Europe – Steering Committee on Bioethics (CDBI) (2002), Proposal for an instrument on the Use of Archived Human Biological Materials in Biomedical Research, Strasbourg, 17 October at www.coe.int/T/E/Legal_Affairs/Legal_co-operation/Bioethics/CDBI/07abstract_mtg24.asp (accessed 9 May 2006).

4. See e.g. UNESCO, International Declaration on Human Genetic Data, Geneva, October 16, 2003, http://portal.unesco.org/en/ev.php@URL_ID=16632&URL_DO=DO_TOPIC&URL_SECTION=201.html (accessed 9 May 2006).

5. The US National Bioethics Advisory Commission (NBAC), although referring to collections of human biological materials (thus encompassing collections for scientific as well as clinical or anthropological purposes) estimated in 1998 that “at least 282 millions



specimens (from more than 176 million individuals cases) are stored in the United States, and the collections are growing at a rate of over 20 million cases per year”. National Bioethics Advisory Commission (NBAC) (1999), Research Involving Human Biological Materials: Ethical Issues and Policy Guidance, Volume I, Report and Recommendations of the National Bioethics Advisory Commission, Rockville, Maryland, August, http://bioethics.georgetown.edu/nbac/hbm.pdf (accessed 9 May 2006).

6. See e.g. Human Genome Organisation, Statement on Human Genomic Databases, London, December 2002, www.hugo-international.org/PDFs/Statement%20on%20Human%20Genomic%20Databases%202002.pdf, (accessed 9 May 2006).

7. See e.g. the World Health Organisation (WHO) research project: “Human Genetic Databases: Towards a Global Ethical Framework” (a description of the project can be found at www.who.int/ethics/topics/hgdb/en/ (accessed 9 May 2006).

8. As mentioned by Nicole Palmour (2003) in “A Survey of the Variability of DNA Banks Worldwide”, “not only is DNA banking going on globally but in many cases there are no unified standards governing the practice”, in Bartha Maria Knoppers (ed.), Populations and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden, p. 123.

9. The United Kingdom uses the term. See www.ukbiobank.ac.uk (accessed 9 May 2006).

10. Bovenberg, J.A. (2003), “Background paper on Ownership and Commercialisation of Human Genetic Research Databases”.

11. Estonia uses the term for its HGRD, see www.geenivaramu.ee/index.php?show=article&lang=eng&id=457&pid=3&offset=30 (accessed 9 May 2006).

12. The COGENE project ended in 2004. Consequently, information will be accessible by contacting the Research Directorate of the European Commission. For some general background information on European Union’s Quality of Life Programme of the 5th Framework Programme (FP5), see cordis.europa.eu/life/ (accessed 11 September 2006).

13. This term is used in Latvia. See http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc/ (accessed 9 May 2006).

14. C. Sallé, “Existing Human Genetic Research Databases: Context (Consent Mechanisms and Communication Strategies”. p. 2.

15. For an in-depth study of HGRD, see e.g. Australian Law Reform Commission (2003), Essentially Yours: The Protection of Human Genetic Information in Australia, Sydney, 29 May, www.austlii.edu.au/au/other/alrc/publications/reports/96/ (accessed 9 May 2006). Part E. “Human Genetic Databases”, Chapter 18, “Human Genetic Research Databases”. See also J. Kaye. and P. Martin (1999), “The Use of Biological Sample Collections and Personal Medical Information in Human Genetic Research, The Wellcome Trust, London, especially pp. 9-13; National Bioethics Advisory Commission (NBAC), Research Involving Human Biological Materials: Ethical Issues and Policy Guidance, Volume I., Report and Recommendations of the National Bioethics Advisory Commission, Rockville, Maryland, August 1999, http://bioethics.georgetown.edu/nbac/hbm.pdf (accessed 9 May 2006). On population and human genetic research databases, see e.g. Bartha Maria Knoppers (ed.), Populations and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden,



2003; M. A. Austin, S. Harding and C. McElroy (2003), “Genebanks: A Comparison of Eight Proposed International Genetic Databases”, Community Genetics, Vol. 6, p. 37.

16. David Cyranoski and Rachael Williams (2005), “Health study sets sights on a million people - Huge Asian project to track genes, lifestyle and health”, Nature, Vol. 434, No. 7035, p. 812.

17. One such project was launched by AutoGen, an Australian biotechnology company in collaboration with Merck Lipha, a subsidiary of Merck Germany, upon the signing of an agreement with the Tongan Ministry of Health in November 2000. It contemplated collecting samples and data from inhabitants of the South Pacific country of Tonga and creating a database in order to “identify genes that cause common diseases using the unique population resources in the Kingdom of Tonga” and ultimately improve treatment through better designed drugs. (B. Burton (2002), “Opposition Stalls Genetic Profiling Plan for Tonga” Inter Press Service, 18 February. The Framingham Project, in Massachusetts, a database to be developed by a genomics company, Framingham Genomic Medicine (FGM) was abandoned in early 2001 when FGM was dissolved. For information on the Framingham Heart Study, see http://framingham.com/heart/ (accessed 10 May 2006); for information on the Framingham Project, see D.J. Craig (2001), “BU dissolves Framingham Heart Study Spin-off” B.U. Bridge, January 12, www.bu.edu./bridge/archive/2001/01-12/framingham.html (accessed 10 May 2006).

18. See www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pers_med_res_prj (accessed 8 August 2006).

19. The Marshfield Medical Research Foundation is the research and education department of the Marshfield Clinic, a not-for-profit organisation established in 1916 by six Marshfield physicians. With 40 centres in Wisconsin, the clinic, an integrated private multidisciplinary medical institution, benefits from extensive electronic clinical records. The Personalized Medicine Research Project is one of the numerous research projects currently being conducted at the Marshfield Medical Research Foundation. See www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pers_med_res_prj (accessed 11 May 2006).

20. The geographical area may be enlarged in the future.

21. See www.marshfieldclinic.org/chg/pages/Proxy.aspx?Content=MCRF-Centers-CHG-Core-Units-PMRP-Consent-form_4-26-06.1.pdf (accessed 8 August 2006).

22. See www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pmrp_faqs (accessed 8 August 2006).

23. See www.cartagene.qc.ca/ (accessed 8 August 2006).

24. The Network of Applied Genetic Medicine (RMGA) is a non-profit organisation established and sponsored by the Fonds de la Recherche en Santé du Québec (FRSQ) whose mission is the furtherance of genetic research through networking and knowledge exchange and transfer. See www.rmga.qc.ca/en/index.htm (accessed 10 May 2006).

25. See www.cartagene.qc.ca/diagramme.cfm (accessed 8 August 2006). Also from author’s personal communications with CARTaGENE representative (August 10, 2006).

26. See www.p3gconsortium.org/cartagene.cfm (accessed 24 August 2006).

27. See www.cartagene.qc.ca/diagramme.cfm (accessed 8 August 2006).






31. See http://bmc.biomed.lu.lv/gene/ (accessed 5 September 2006). See also �� and E. � �� – Unified Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc; (accessed 5 September 2006); www.privireal.org/content/dp/latvia.php (accessed 5 September 2006). A. Putnina, “Exploring the Articulation of Agency: Population Genome Project in Latvia”, www.ifz.tugraz.at/index_en.php/filemanager/download/122/putnina.pdf (accessed 11 May 2006).

32. On 13 June 2002, the Latvian Parliament enacted the Human Genome Research Law (unofficial translation), which sets the legal basis for the creation of a genome database and related genomic research in Latvia, and regulates consent mechanisms, participants’ confidentiality and protection against abuse and discrimination. Author’s communication with the Principal Investigators. See www.privireal.org/content/dp/latvia.php (accessed 9 �� !��"��#��$��“Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc (accessed 18 May 2006).).

33. Iceland, Minister of Health and Social Security: Act on a Health Sector Database No. 139/1998, entered into force 17 December 1998, http://eng.heilbrigdisraduneyti.is/laws-and-regulations/nr/659 (accessed 8 August 2006); Act on Biobanks No. 110/2000, entered into force 1 January 2001, http://eng.heilbrigdisraduneyti.is/laws-and-regulations/ (accessed 11 September 2006); see also the associated regulation: Regulations on the Keeping and Utilisation of Biological Samples in Biobanks No. 134/2001, entered into force 6 February 2001, also at http://eng.heilbrigdisraduneyti.is/laws-and-regulations/ (accessed 12 May 2006).

34. See Article 10 of Act on a Health Sector Database No. 139/1998.

35. See www.decode.com (accessed 11 September 2006).

36. See Article 7 of Act on a Health Sector Database No. 139/1998.

37. See, for example, R. Lewis, “Iceland’s Public Supports Database, But Scientists Object” Vol. 13, No. 15, 19 July 1999, The Scientist; Rose, H. (2001), The Commodification of Bioinformation: the Icelandic Health Sector Database, Wellcome Trust; Mannvernd, a community-based organisation which is extremely active in opposing the Health Sector Database project, www.mannvernd.is/english/ (accessed 11 September 2006); G.J. Annas (2000), “Rules for Research on Human Genetic Variation – Lessons from Iceland”, New England Journal of Medicine, Vol. 342, No. 24, p. 1830; J. Gulcher and K. Stefansson (1999), “An Icelandic Saga on a Centralised Healthcare Database and Democratic Decision-Making”, Nature Biotechnology, Vol. 17, No. 7, p. 620; J. Kaiser (2002), “Population Databases Boom, From Iceland to the U.S”., Science, Vol. 298, p. 1158; M.R. Anderlik and M.A. Rothstein (2001), “Privacy and Confidentiality of Genetic Information: What Rules for the New Science?”, Annual Review of Genomics and Human Genetics, Vol. 2, p. 401, http://arjournals.annualreviews.org/doi/full/10.1146/annurev.genom.2.1.401 (accessed



12 May 2006, requires subscription); B. Andersen and E. Arnasson (1999), “Iceland’s database is ethically questionable”, British Medical Journal, Vol. 324, p. 443.

38. Ragnhildur Guomundsdottir v. The State of Iceland (2003), Icelandic Supreme Court, No. 151/2003. Unofficial translation provided by Mannvernd, see www.mannvernd.is/english (accessed 12 September 2006).

39. Article 8 of Act on a Health Sector Database No. 139/1998 permits a patient to request, at any time, that information on him/her not be entered into the HSD. The patient’s request may also apply to all existing information or that which may be recorded in the future.

40. Author’s communication with deCODE Genetics representative (23 August 2006) and see www.decode.com/Population-Approach.php.

41. Author’s communication with deCODE Genetics representative (23 August 2006) and see www.decode.com/Population-Approach.php.

42. “The deCODE Population Approach”, www.decode.com/Population-Approach.php (accessed 12 May 2006).

43. See www.geenivaramu.ee/mp3/trykisENG.pdf (accessed 12 May 2006) and www.geenivaramu.ee/index.php?lang=eng&show=article&pid=3 (accessed 12 May 2006); A. Koik, “The Estonian Genome Project: A Hot Media Item”, www.opendemocracy.net/theme_9-genes/article_1335.jsp (accessed 12 May 2006); Mark Frary, “Estonian Genome Project Ahead of Schedule”, Reuters Health, 21 December 2002, www.genomics.ee/index.php?lang=eng&show=16&sub=40&nid=125&PHPSESSID=6ac84d8b5608915038d5635a215af9ff (accessed 12 May 2006); T. Tasmuth, “The Estonian Gene Bank Project – An Overt Business Plan”, Open Democracy, 29 May 2003, www.opendemocracy.net/debates/article-9-79-1250.jsp (accessed 12 May 2006); T. Sild and T. Mullari (2001), “Population-Based Genetic Research: Estonia’s Answer to the Legal Challenge”, European Journal of Health Law, Vol. 8, p. 363; A. Rannamaë (2003), “Estonian Genome Project – Large-Scale Health Status Description and DNA Collection” in Bartha Maria Knoppers (ed.), Populations and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden.

44. See www.geenivaramu.ee/index.php?lang=eng&sub=58 (accessed 9 August 2006).

45. Estonia, Human Genes Research Act, entered into force 8 January 2001, www.legaltext.ee/text/en/X50010.htm (accessed 12 May 2006).

46. See www.geenivaramu.ee/index.php?lang=eng&sub=58 (accessed 9 August 2006).

47. See www.geenivaramu.ee/index.php?lang=eng&sub=58 As of the end of 2005, there were over 10 000 gene donors. See www.geenivaramu.ee/mp3/flaier_EGPF2005.pdf (accessed 9 August 2006).

48. See www.geenivaramu.ee/mp3/trykisENG.pdf (accessed 12 May 2006).

49. See www.egeeninc.com/public/ (accessed 12 May 2006).

50. See “Genome Project Ends Cooperation with Current Financier” (27 December 2004), www.geenivaramu.ee/index.php?lang=eng&show=uudised&id=172 (accessed 7 September 2006).



51. See www.ukbiobank.ac.uk (accessed 12 May 2006). Two public interest groups have been intensely involved in the debates surrounding the establishment of the UK Biobank: GeneWatch, at www.genewatch.org/default.htm (accessed 12 May 2006); and Liberty, at www.liberty-human-rights.org.uk (accessed 7 September 2006). For more information on the UK Biobank Project, see e.g. F. Rawle (2003), “UK DNA Sample Collections for Research” in Bartha Maria Knoppers (ed.), Populations and Genetics: Legal, Socio-Ethical Perspectives, Martinus Nijhoff, Leiden, p. 3; D. Shickle, R. Hapgood, J. Carlisle, P. Shakcley, A. Morgan and C. McCabe (2003), “Public Attitudes to Participating in UK BioBank: A DNA Bank, Lifestyle and Morbidity Database on 500,000 Members of the UK Public Aged 45-69”, in Bartha Maria Knoppers (ed.), op cit., p. 323; Pallab Ghosh, “Will Biobank Pay Off?”, BBC News, 24 September 2003, http://newswww.bbc.net.uk/2/low/health/3134622.stm (accessed 12 May 2006); P. Hagan, “Tracking Genes and Disease”, The Scientist, 21 January 2002, www.biomedcentral.com/news/20020121/03/ (accessed 12 May 2006); P. Hagan, “UK Biobank Reveals Ethics Framework”, The Scientist, 24 September 2003, www.biomedcentral.com/news/20030924/03 (accessed 12 May 2006).

52. The Wellcome Trust, www.wellcome.ac.uk (accessed 12 May 2006).

53. See www.mrc.ac.uk/ (accessed 16 May 2006).

54. See www.doh.gov.uk/ (accessed 16 May 2006).

55. UK Biobank, www.ukbiobank.ac.uk/about/why.php and www.ukbiobank.ac.uk/ethics/consultations.php (accessed 16 May 2006).

56. UK Biobank Ethics and Governance Council, Annual Report 2004-2005.

57. See www.ukbiobank.ac.uk/about/overview.php (accessed 9 August 2006).

58. See www.ukbiobank.ac.uk/about/overview.php (accessed 7 September 2006).

59. See www.ukbiobank.ac.uk/about/organisation.php (accessed 7 September 2006).

60. See www.ukbiobank.ac.uk/ethics/efg.php (accessed 9 August 2006) and UK Biobank Ethics and Governance Council, Annual Report 2004-2005.

61. See www.ukbiobank.ac.uk/docs/EGF%20Version02%20May%202006.pdf (accessed 7 September 2006).

62. See www.genomecenter.howard.edu/TGRIAD.htm; www.genomecenter.howard.edu/tgriad_component.htm; www.genomecenter.howard.edu/tgriad_strategy.htm; (accessed 7 September 2006). For background information see www.genomecenter.howard.edu/intro.htm (accessed 16 May 2006); “College of Medicine and First Genetic Trust Form Biobank Data to Advance Study of Disease Risks Among People of African Descent” Capstone online, 2 June 2003, www.howard.edu/newsevents/Capstone/news2.htm (accessed 16 May 2006); A.K. Erickson (2003), “Ethnicity Puts Clinical Trials to the Test”, Nature Medicine Vol. 9, No. 8, p. 983; M. Szalavitz (2001), “Race and the Genome: The Howard University Human Genome Center”, www.genomecenter.howard.edu/article.htm (accessed 16 May 2006); M. Ritter (2003), “Research and Race”, The Washington Times, www.washtimes.com/national/20030803-123153-8383r.htm (accessed 16 May 2006).

63. See www.genomecenter.howard.edu/tgriad.htm.



64. Author’s discussion with Principal Investigator.



67. Information on the GenomEUtwin project was found at www.genomeutwin.org (accessed 16 May 2006).

68. See www.ktl.fi/ (accessed 16 May 2006).

69. See www.kaksostutkimus.helsinki.fi/ (accessed 16 May 2006).

70. See www.genomeutwin.org (accessed 7 September 2006).

71. See www.genomeutwin.org and www.p3gconsortium.org/genomeutwin.cfm (accessed 7 September 2006).

72. L. Peltonen (2003), “GenomEUtwin: A Strategy to Identify Genetic Influences on Health and Disease” Twin Research Vol. 6, No. 5, p.354.

73. See www.genomeutwin.org/desc.htm (accessed 7 September 2006) and L. Peltonen (2003), “GenomEUtwin: A Strategy to Identify Genetic Influences on Health and Disease” Twin Research Vol. 6, No. 5, p.354. Part of GenomEUtwin network, the MORGAM project brings together several European countries in an effort to better understand the influence of genetic risks in the development of cardiovascular diseases (follow-up of the WHO-MONICA cohort, see www.ktl.fi/monica/ [accessed 16 May 2006] and other studies). For more information on the MORGAM cohort see www.ktl.fi/morgam/ (accessed 16 May 2006).

74. See www.genomeutwin.org/desc.htm (accessed 16 May 2006).

75. L. Peltonen (2003), “GenomEUtwin: A Strategy to Identify Genetic Influences on Health and Disease” Twin Research Vol. 6, No. 5, p.354.

76. See www.genomeutwin.org/desc.htm (accessed 8 September 2006).

77. See www.hapmap.org (accessed 8 September 2006). See also www.sanger.ac.uk/HGP/Chr6/MHC/consortium.shtml (accessed 16 May, 2006); www.genome.gov/page.cfm?pageID=10001688#1 (accessed 16 May 2006); P. Recer (2002), “The International ‘HapMap’ Project”, CBS News, 29 October, at www.genome.gov/10001688 (accessed 16 May 2006); D. Bentley (2003), “The International HapMap Consortium”, Nature www.nature.com/nature/journal/v426/n6968/full/nature02168.html#The%20International%20HapMap%20Consortium (accessed 16 May 2006).

78. See www.genome.gov/10005336 (accessed 8 September 2006).

79. As Francis S. Collins, Director of the National Human Genome Research Institute, stated, “The HapMap will provide a powerful tool to help us take the next quantum leap toward understanding the fundamental contribution that genes make to common illnesses like cancer, diabetes and mental illness.” See http://genome.gov/10005336 (accessed 16 May 2006).



80. Public and private organisations in six countries are participating in the International HapMap Project. While the SNP Consortium (TSC) is co-ordinating private funding, the Wellcome Trust charitable resources and the US NIH provide approximately 40% of the overall budget for the HapMap Project. See www.genome.gov/10005339 (accessed 16 May 2006). Public funding is provided by the Japanese Ministry of Education, Sports, Science and Technology, the Chinese Ministry of Science and Technology, the Natural Science Foundation of China, the Chinese Academy of Science, Genome Canada and Genome Quebec. For funding allocation see www.sanger.ac.uk/HGP/Chr6/MHC/ (accessed 16 May 2006) For a table of all the research groups involved in the project, their funding, role, percentage of genome and chromosomes assigned, see www.hapmap.org/groups.html (accessed 8 September 2006).

81. See www.hapmap.org/hapmappopulations.html (accessed 8 September 2006).

82. See www.hapmap.org/hapmappopulations.html (accessed 8 September 2006). This geographic or ethnic variation was based on the assumption that a haplotype map built from samples from those geographic areas would apply to all populations. However, a parallel study has been undertaken to verify that postulate. Furthermore, it has been decided for confidentiality reasons that additional samples would be gathered and that researchers would blindly choose those studied. See www.sanger.ac.uk/HGP/Chr6/MHC/ (accessed 16 May 2006).

83. See www.hapmap.org/datareleasepolicy.html.en (accessed 8 September 2006). On 10 December, 2004 - The International HapMap Consortium announced that it was ending computer-based “click wrap” licence restrictions on data so as to facilitate access, see www.genome.gov/12514423 (accessed 8 September 2006).

84. See http://coriell.umdnj.edu/ (accessed 16 May 2006). The Coriell Institute has developed a policy for data storage from named populations; see Coriell Institute for Medical Research, “Policy for the Responsible Collection, Storage and Research Use of Samples from Named Populations for the NIGMS Human Genetic Cell Repository” at http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html (accessed 16 May 2006).

85. See www.genome.gov/10001685 (accessed 8 September 2006).


87. The creation of the forum was seen as primordial by the European Commission not only in terms of competitiveness but also in terms of overall health-care improvement. As asserted by the European Commission: “The development of new genome-based technologies, as well as new bioinformatics tools, is of primary importance for competitive genome research and for the development of new diagnostics and new therapeutic approaches. It is also important to create and maintain in Europe infrastructures to support genome research (for example, databases and animal model resources).” See European Commission (2000), “A New Initiative on Genome Research for Human Health”, Press Release, 15 November, at http://europa.eu.int/comm/research/press/2000/pr1511en.html (accessed 16 May 2006). The Forum dedicates an important part of its funding to the financing of integrated projects that aim to create genomic or proteomic databases.

88. Information on this project on the P3G website, see www.p3gconsortium.org (accessed 12 May 2006).



89. Based on the identification of common scientific objectives and a shared philosophy of open access, delegates from some leading population genomics projects met in London, in February 2003, with Dr. Bartha Maria Knoppers, the proposed leader, and representatives from Genome Canada and Quebec to discuss the idea of a P3G consortium. Subsequently, experts from diverse backgrounds (ethicists, lawyers, public health researchers, epidemiologists) refined the P3G objectives. See www.p3gconsortium.org/Pevents.cfm (accessed 11 September 2006).

90. Genome Canada is the primary funding and information resource relating to genomics and proteomics in Canada. Dedicated to developing and implementing a national strategy in genomics and proteomics research for the benefit of all Canadians, as of 2006, it has received CAD 600 million from the Government of Canada. Genome Canada has established five Genome Centres across the country (Atlantic, Québec, Ontario, Prairies and British Columbia) and has as a main objective to ensure that Canada becomes a world leader in genomics and proteomics research. See www.genomecanada.ca (accessed 11 September 2006).

91. See www.p3gconsortium.org/docs/blueprint_Draft2005.pdf (accessed 11 September 2006).

CHAPTER 3. ESTABLISHMENT OF AN HGRD – 55


Chapter 3

Establishment of an HGRD1

3.1 Nature and scope of a database

3.1.1 Ensuring representativeness of populations

When determining the scope of a database, it is essential to ensure its “representativeness” both in terms of the desired population and the diversity of populations.

It is critical for the scientific legitimacy of a project that the sample population is genetically representative of the population it is to serve. Ideally, all groups should be included in a given study. However, owing to financial and practical constraints, this is not always feasible. Thus, a careful and rigorous selection process is crucial.

Examples of the approaches adopted by several databases to achieve this representativeness include:

� The HapMap Project involves samples from individuals of northern and western European ancestry in the United States (CEPH samples, Utah), Yoruban (Ibadan/Nigeria), Japanese (Tokyo) and Han Chinese descent. This geographic or ethnic selection was based on the assumption that a haplotype map built from samples from these geographic areas would apply to all populations. A parallel study was undertaken to verify the postulate.1

� The Marshfield Project team noticed that attracting interest in becoming a gene donor can be quite cumbersome in rural farming areas. Although the non- or low participation rate is not worrying in terms of genetic representation, it is detrimental to some of the intended research, in terms of exposure to specific environmental or social factors (e.g. smokers and individuals who abuse alcohol tend to be unwilling to participate in the research).2

� CARTaGENE intends to do a random sampling of more or less 1.5% of the entire Quebec population (approximately 7.5 million).3

This chapter also draws on the texts prepared for the Tokyo Workshop by: Jasper Bovenberg, Ruth

Chadwick, Mildred Cho, Pall Hreinsson, Ryuichi Ida, Clementine Sallé, David Weisbrot, and Erich Wichmann. See Bibliography.

56 – CHAPTER 3. ESTABLISHMENT OF AN HGRD


3.1.2 Involvement of children and protected adults

While the aim of an HGRD is to be representative of the population, there are constraints to the achievement of such an objective. One constraint has been the involvement of children in genetic studies and as contributors to HGRDs. Some proponents advocate that genetic studies should not involve young children. However, this position could severely obstruct research on genetic diseases which occur early in life. Others maintain that children should be involved in genetic studies and should be permitted to contribute to HGRDs. However, many of these latter advocates have tied their support for the inclusion of children to the issue of consent (see Chapter 4 for discussion). Other considerations pertaining to the inclusion of children in genetic studies and in HGRDs include the necessary safeguards and information that can or ought to be disclosed to the child, the manner of communicating with the child, as well as the foreseeable future impact of participation on the child. Equivalent issues and challenges arise with respect to the inclusion in HGRDs of participants who are or become protected adults (see Chapter 4 for discussion). Similarly, the exclusion of protected adults from research, especially populational research, could delay or impair research into certain diseases and conditions, including those of a neurological nature or origin. In order to achieve some balance between these competing interests, the Latvian initiative has enacted, in its enabling legislation, provisions on the inclusion of persons recognised as not having the capacity to act. In this regards, the criterion for the inclusion of such a participant is that it must provide a direct benefit for the health of the participant and that the risks must be commensurate with the benefits.

3.1.3 Nature of database

A key issue for the establishment of an HGRD is the type of biological samples or information to be collected and stored in the database. As well, consideration of the manner in which the biological samples, data, information, and consent forms will be organised will be important, including for privacy and security purposes. For example, biobanks may consider establishing separate and independent databases, one containing the research materials and the other containing participant’s personally-identifying information (whether anonymised or coded). Various approaches that may be adopted or have been used by diverse databases are examined below.

Pursuant to the Act on a Health Sector Database (HSD Act),4 the Licensee would be allowed to collect health data,5 including genetic data,6 derived from medical records of individual patients for entry into the Health Sector Database (HSD) and to link these data to a genealogical and a genetics database.7 Prior to entry in the HSD, the data must be coded so as to ensure that only non-personally identifiable data is used in the HSD. The collecting of health data must be based on agreements to be concluded between the Licensee and the health-care institutions and physicians keeping the medical records. To date the HSD is not operational. deCODE is currently operating its own proprietary sets of genealogical, genetic and phenotypic data, in collaboration with specific research and health institutions.8

According to the Act establishing the Estonian Gene Bank9 it is defined as a database established and maintained by the chief processor consisting of tissue samples,10 descriptions of DNA, descriptions of state of health, genealogies, genetic data and data enabling the identification of gene donors. The creation of the gene bank involves two interrelated and parallel data collection processes: collection of health data and collection of genetic data (Figure 3.1). Consenting Estonians sign a Gene Donor Consent Form, fill



out a health questionnaire and provide a blood sample. After coding, the blood sample is sent to the laboratory of the Gene Bank for DNA extraction (Figure 3.2). Separated DNA is placed in the storage facility of the Gene Bank. Personally identified data on the questionnaires is separated and replaced by a 16-digit code in the coding centre. Health data that have been separated from personal data are stored in a database of the Gene Bank. A description of state of health is also coded and entered in the Gene Bank by the chief processor on the basis of data on the gene donor stored in medical institutions. The genealogy of the gene donor may be prepared in the Gene Bank on the basis of the results obtained from the participant’s questionnaire, information from other databases and from genetic research. The consent forms are stored in a separate archive of the Gene Bank.

Figure 3.1. Estonian Gene Bank’s proposed operation scheme11

Figure 3.2. Illustration of blood sample and information processing12



The UK Biobank is a multi-purpose data resource whose aim is to have approximately half a million participants aged between 40 and 69 years be involved. Each participant will be asked to contribute a blood and urine sample, have some baseline physical measurements taken (e.g. blood pressure, size, grip strength, lung function) and complete a confidential lifestyle (e.g. diet, exercise, smoking, alcohol) and other factors (e.g. mood, cognitive function, medical history) questionnaire to create a national database.13 The UK Biobank will follow participants over time in order to obtain an in-depth progression of their health, illness and incapacitation.14

3.1.4 Intended purposes and uses of the database

The intended purpose of the database has implications for its establishment. This should be determined at its inception in order to ensure that all relevant information, data and biological samples are collected. This determination is also important for ensuring that the appropriate level or type of consent is obtained and in order to promote and maintain public trust in the initiative. The obvious reason for the establishment of HGRDs will be to carry out research. However, the important issue is whether the specific nature of the intended research may be determined or determinable at the time the HGRD is established or at the time of the collection of biological samples and data. This determination, or the inability to make such a determination, will also have implications for the issue of consent (see Chapter 4).

The UK Biobank’s approach is to present information to participants as an “opportunity to contribute to a resource that may, in the long term, help enhance other people’s health”. Moreover, as it is impossible to anticipate future research uses, participants’ consent will be sought for medical research in general.15 While the consent will be general to all medical research, it is specific so as to exclude other non-medical research purposes. The UK Biobank has indicated that it will not close off any avenue of research. Proposals will have to receive independent scientific and ethics approval and will be reviewed by the UK Biobank to ensure that they are consistent with participants’ consent. The UK Biobank has indicated that it “will encourage and provide access to the Biobank resource and the results that flow from it [as] widely and openly as possible in order to maximise its use and value for research. This will include access for researchers from the academic, commercial, charity and public sectors, both in the UK and overseas”.16 For the intended Icelandic population database, the Interdisciplinary Ethics Committee would have had to approve all requests for research and searching of the database. Any decision of this Committee could have been appealed to the Minister (Health and Social Security) who, in such circumstances, could have sought an opinion from the Science Ethics Committee.

Another issue is whether the biological samples and data collected in the database may or should be allowed to be employed for other, non-research purposes. Other purposes for which the contents of a database/biobank may be employed include the delivery of clinical genetic services, law enforcement, insurance, legal actions and identification (e.g. military or civil). Changes in the demand for genetic information, whether or not it is collected for research purposes, will certainly raise new issues for policy makers, within and beyond the medical and biomedical research sphere. Moreover, the determination of these issues are significant in terms of the establishment of the database, the attraction of participants to contribute to the database, consent, security, enforcement, public trust and communication.



Various databases take quite different approaches to these questions. For example, the UK Biobank does not automatically prohibit access for purposes of law enforcement. However, such access will only be acceded to pursuant to a court order, and the UK Biobank may resist such action.17 For the Icelandic database, the legislation does not specifically give the police authorities access to the population database. Thus, if and when the database is established, this aspect will remain unclear. In Estonia, the Human Genes Research Act specifies the type of medical research that may be carried out. Moreover, this Act explicitly prohibits the use of the gene bank “to collect evidence for criminal or civil proceedings or for surveillance”.18

3.2 Funding of a database

HGRDs may be established under different structures: for-profit (i.e. private undertaking), not-for-profit (i.e. public undertaking) or a mixed model of public-private partnership. These types of undertakings are examined below using three model databases. The issue of whether a population database should be publicly or privately funded has been the subject of much debate. Arguments supporting the view that a database should be public include i) health data are a national heritage; ii) genetic data are essentially public; iii) health data recorded at the public’s expense should remain a national resource; iv) it is essential that this resource remain open for public research; and v) in order to avoid duplication.19 The arguments in favour of private funding of population databases include: i) health data are a commodity; ii) there is no inherent difference from traditional clinical trial data (data which are owned by the sponsor of the clinical trial, data embodied in a drug master file for registration purposes, and data exclusivity); iii) de novo collections require additional funds; and iv) the marketplace is always right.20 The Icelandic database was to have been established as a for-profit endeavour, but was never commenced. The Estonian database initially established as a private-public partnership, has been converted into a public undertaking. The UK Biobank was established as a not-for-profit undertaking.

3.2.1 Private sector funding

Iceland had chosen to have its population database developed by the private sector. The official rationale is provided in the Parliamentary Notes from the passage of the Act on the Health Sector Database.21 Recorded data on the health of the Icelandic people are seen as a national resource. High-quality health data have been collected for decades by health professionals, patients and scientists and have been paid for through public funds. Health data cannot be evaluated in monetary terms, because their value consists primarily in their potential for improving health. Owing to the nature of the data and their origin, the data cannot be subject to ownership in the usual sense by institutions, companies or individuals. However, a duty to utilise the data in the interests of the health sciences and the public health arises. This can best be accomplished through government authorisation for the creation and operation of a single, centralised database, which offers the following benefits: i) acquisition of new knowledge on health or disease; ii) improved quality and economy in the health system; iii) development of high-technology industry and employment in Iceland; and iv) potential for attracting business to Iceland. The creation and maintenance of the database were considered too costly to be financed by public funding. Therefore, the Act on the Health Sector Database22 authorises the issuance by the government of an exclusive licence to a single entity (the Licensee) to create and operate a centralised health sector database (the HSD). The exclusive Licensee would have to pay for the establishment, operation, maintenance, etc. of the database and related



activities (e.g., work of the Data Protection Authority). A publicly traded company, deCODE Genetics, received the exclusive licence. While the Licensee must fund the collection, it may commercialise certain aspects pertaining to the collection, but not the database directly. As mentioned, the HSD has not been established.

3.2.2 Public-private partnership

Estonia had initially chosen to have its population database built through a public-private partnership. To that end, the Human Genes Research Act23 provides for the incorporation by the government of a non-profit foundation, the Estonian Genome Project Foundation (EGPF). As the chief processor of the Gene Bank, the EGPF has the right to organise the taking of tissue samples, to prepare descriptions of state of health and genealogies, to code, decode, store, destroy and issue descriptions of state of health and genealogies, to perform genetic research and to collect, store, destroy and issue genetic data. Its objectives are to: i) promote the development of genetic research; ii) collect information on the health of and genetic information concerning the Estonian population; and iii) use the results of genetic research to improve public health. The chief processor has the right to delegate the rights of processing, except for coding and decoding, to an authorised processor, while retaining ownership of the gene bank (“the Project”). In order to secure funding, the EGPF founded a public, limited liability company, EGeen (“EGeen”). In exchange for funding the Project and granting the EGPF a minority stake, EGeen was granted the status of “authorised processor” and exclusive commercial access to all data emerging from the Project for 25 years. EGeen International Corporation (EGI) was an organisation of US investors and was responsible for attracting investments that would be directed to EGPF through EGeen in order to prepare and carry out the Project. The activities of EGPF, EGeen and EGI were regulated by a system of agreements developed pursuant to Estonian and American law.

In December 2004, however, the contracts between the EGPF and EGeen were terminated.24 There are several implications flowing from such a decision. EGeen is no longer under an obligation to finance the Project. The EGPF has had to revise it financial plans and search for other finance avenues. EGeen has lost its exclusive right to the Project and any access or collaboration with the EGPF will be on a similar basis as other entities. EGPF will also have the opportunity to sign access agreements with other entities interested in carrying out research with the resources of the Project.

3.2.3 Public sector funding

The UK opted25 for a population database funded by the public/charity sector. Financial support for the UK Biobank is provided by the Department of Health, Medical Research Council, Northwest Regional Development Agency (NWDA), Scottish Executive, and the Wellcome Trust to UK Biobank Ltd., a charitable company, under a joint venture agreement. A board of directors, accountable to the members of the company (the Medical Research Council and The Wellcome Trust), act as company directors and as charity trustees.26 UK Biobank Ltd. serves as the legal custodian of the data and samples and operates through a co-ordinating centre centrally managed and hosted by the University of Manchester, which will have overall responsibility for delivering the project, including data management and quality assurance, computing and financial management, and formal custodianship of the data and biological samples. It will also co-ordinate the activities of several regional collaborating centres which will be responsible for participant recruitment and initial data and sample collection.



3.3 Legal structure

3.3.1 Legal context for databases

A number of international legal instruments provide some context for issues pertaining to human genetics. First, the Universal Declaration on the Human Genome and Human Rights27 declares that the human genome is the common heritage of humanity and that the human genome in its natural state shall not give rise to financial gains. This Declaration is not legally binding. It deals with the human genome in its natural state and not with the type of data assembled in the databases being considered in the context of HGRDs. Second, the Convention on Biological Diversity28 sets forth universal principles for the exploitation of genetic resources, but it is generally held not to apply to human genetic resources. Third, the Convention on Human Rights and Biomedicine29 adopted by the Council of Europe deals primarily with protection in the context of the application of biology and medicine. While there are a number of Recommendations by the Council of Europe, these are not legally binding.30 In brief, there is currently no international, comprehensive framework setting forth global consensus on the issues of ownership, commercialisation, exclusive licensing, access for researchers, benefit sharing and other issues, as these pertain to population databases.

Owing to the number of elements and instruments by which population databases may be evaluated and to which they may be subject, any global analysis will necessarily be selective. Population databases may be subject not only to the legislation and regulation enacted to govern their creation and operation, but also to more generally applicable laws in their respective jurisdiction as well as to regional and international commitments or instruments. In addition, precedents may not yet be available on the interpretation of national legislation specifically governing the creation and operation of the database. Another complicating factor is that such databases may also be partly governed by the terms of privately negotiated agreements, which may be amended over time.31 32 For example, the creation and operation of the Health Sector Database would be governed specifically by the HSD Act, the government regulations on a Health Sector Database, the Operating Licence between the government and the Licensee, general Icelandic law and the international commitments to which Iceland is a party, including various business regulations of the EEA agreement, such as rules of free competition and the EU Directive for the protection of databases.

Some issues have been addressed at the national and international level in policy statements issued by interest groups. At the international level, these include the European and American Society for Human Genetics, the National Centre for Human Genome Research (NCHGR), the World Medical Association, and the Human Genome Organisation (HUGO). Numerous national health councils have also issued policy statements. While these policies may play a significant role in shaping the debate and offering guidance, they do not provide a complete global framework.�

One key issue is the determination of which legal instruments, if any, are applicable, ought to be applicable or should be developed to be applicable to population databases and whether or not these instruments should be legally binding. Interested observers in the area (e.g. participants, scientists, private entrepreneurs, academics, policy makers, politicians, patients, financiers, and legal scholars) would most likely offer divergent perspectives as to the appropriate combination of legal instruments to which population databases should be subject. Although many of the population databases discussed



previously are not, at the date of publication, fully operational, they do illustrate different approaches to the key question enunciated above.

Some have advocated that a global, especially legally binding, framework should be developed and adopted for population databases. For example, the US National Research Council has suggested that for a worldwide collection of DNA, a “more sophisticated and complicated approach would be to form an international organisation to serve as a trustee and fund-holder for all the sampled populations”.33 One reason to adopt some form of global framework is that it could extend to or provide guidance for local databases. Many national-scale databases developed in some jurisdictions would mostly likely be governed by specific legislation and thus subject to democratic control or at least to certain checks and balances. For example, the Estonian Genome Project was established and the Icelandic HSD would be established pursuant to specific legislation. However, in other jurisdictions, a framework could provide some or additional guidance. For example, the CARTaGENE and the UK Biobank initiatives were not established pursuant to specific legislation. Also, it is unlikely that specific and detailed legislation would be adopted for smaller-scale local initiatives. These initiatives may find an international legal framework very useful. The global framework could also be employed by governmental and charitable funding agencies when setting conditions for financing such initiatives and by universities involved in setting up international or national databases. However, a considerable obstacle to the establishment of an international, legally binding framework is its feasibility. In other words, given the diversity of applicable national and regional legal systems, the different nature, structures and approaches adopted for each current and future HGRD, would an international, legally binding instrument provide guidance while preserving sufficient flexibility to permit the achievement of the diverse goals sought by each current and future HGRD?

The issue of whether an international instrument is possible will also depend on its scope and structure. One example is the UN framework set up to prospect, explore and exploit deep-seabed minerals, which, like the human genome, have been declared the common heritage of mankind. This framework, carried out by the International Seabed Authority, based on the principles of technology transfer, unlimited access for research and benefit sharing, is set forth in a UN Convention.34 The “Global Trust” of all sampled populations, proposed by the US National Research Council, could be framed in this way. While a few prospecting agreements have been concluded with pioneer investors, the seabed framework has not yet attracted any institution, either for-profit or not-for-profit, to exploit any of these resources under the terms of the Convention. Given this experience and given that human genetic research databases may be subject to national or even local sensitivities, global consensus for such an extensive framework, let alone practical implementation thereof, may face considerable difficulties and challenges.

There may be alternatives to internationally legally binding instruments or frameworks. The HGP and the SNP Consortium,35 for example, have demonstrated that international co-operation in the area of genomic research, rather than legally binding instruments, is possible. Both of these endeavours adopted policies on the release of raw genetic sequence data, mirroring positions taken at the political level (see Clinton-Blair statement36). Internationally non-legally binding statements of principles are also an alternative. The Statement on Human Genomic Databases adopted by the Human Genome Organisation (HUGO) is an example.37 The HUGO Statement recognises the potential global good arising from genetic research, the scientific and clinical uses of genomic databases as well as the potential for conflicts between the free-flow of information that is crucial to research advances and the legitimate rights to return from



research expenditure. Building upon its previous Statements covering principled conduct of genetic research, the HUGO Statement adopts detailed Principles. These include i) human genomic databases are global public goods; ii) individuals, families, communities, commercial entities, institutions and governments should foster the public good; iii) the free flow of data and the fair and equitable distribution of benefits from research using databases should be encouraged; iv) the choices and privacy of individuals, families and communities with respect to the use of their data should be respected; v) individuals, families and communities should be protected from discrimination and stigmatisation; and vi) researchers, institutions and commercial entities have a right to a fair return for intellectual and financial contributions to databases. Subsequently the principles are worked out into practical recommendations governing, inter alia, the above issues of ownership of data and commercialisation.

While an international instrument may provide considerable international guidance and structure, national approaches, both binding and non-binding, have also been effectively employed. Some jurisdictions have chosen to develop a national legally binding framework or legislation. For example, the Icelandic government has enacted the Act on the Health Sector Database, as well as related regulations as the instruments that would govern the creation and management of the Icelandic population database. These instruments contain detailed provisions covering a breadth of issues, including rights of patients, access by researchers, use of the database, and penalties. Similarly, the Estonian government also enacted framework legislation, the Human Genes Research Act, which also covers a breadth of issues including the rights of gene donors, storage of samples and data, data protection, prohibitions of discrimination, etc. In both cases, the population database would also be subject to other applicable national and international legal instruments. The enactment of a specialised framework statute may create clarity and transparency for interested parties (initiators, researchers, participants, funders, public, etc.). Another advantage is the possibility of creating mechanisms, especially for funding, governance, enforcement, and penalties, specific to population databases.

Other jurisdictions have chosen non-legally binding approaches. Examples of non-legally binding approaches include the enunciation of principles at the national level. For instance, the creation of CARTaGENE and the UK Biobank is not based on the enactment of legislation. The CARTaGENE initiative is governed, amongst others, by two Statements of Principles, the Statement of Principles Human Genome Research (“2000 Statement”) and the Statement of Principles on the Ethical Conduct of Human Genetic Research Involving Populations (“Population Statement”).38 The Population Statement is based on a framework of ten fundamental principles that give rise to specific recommendations and procedures for their implementation. The Principles include individuality, diversity, complexity, reciprocity, security, and universality, with their application as Recommendations including coverage of such issues as consultation, consent, confidentiality, communication of research results and commercialisation. For example, the Recommendation on Communication of Research Results entreats researchers to regularly share information out of respect for the participation of the population. Research results should be made public cautiously and, where appropriate, researchers, in collaboration with the population involved, should facilitate the development of a follow-up plan. Another example, pursuant to the Recommendation on Commercialisation, researchers do not own the genetic sample but may aspire to the acquisition of intellectual property rights over inventions derived from genetic data. Similarly, the UK Biobank has developed a detailed Ethics and Governance Framework,39 which will be administered by an independent Ethics and Governance



Council, as well as a Policy on Intellectual Property and Access.40 These two instruments will principally govern the UK Biobank, although it will remain subject to other applicable national legislation.

3.3.2 Human rights statutes

The judgement of 27 November 2003 of the Supreme Court of Iceland demonstrates that constitutional human rights statutes and applicable international human rights treaties may have a bearing on the operation, management and governance of human genetic research databases. As discussed previously, in the case of Ragnhildur Guomundsdottir v. The State of Iceland,41 the Appellant challenged the inclusion into the Health Sector Database of medical data pertaining to her deceased father. The Supreme Court highlights that Article 7 of the Act on a Health Sector Database provides for a private party, which is neither a health institution nor a self-employed health worker, to gain access to data in health records, without the data subject having given his or her express consent. Although this, in and by itself, need not be interpreted as a breach of Article 71 of the Icelandic Constitution, the legislature should, when developing a statute, do its utmost to ensure that the data cannot be traced back to the data subjects.42

The Supreme Court pointed out that the Act on a Health Sector Database stipulates repeatedly that data in the HSD should not be personally identifiable. The Court went on to say that the Act does not, however, contain the main principles for achieving this goal. Owing to the responsibilities imposed by the Icelandic Constitution, Article 71, Paragraph 1, on the legislature to protect citizens’ personal privacy, the Supreme Court held that the aforementioned main principles could not be substituted with provisions on various surveillance measures to be taken by governmental agencies, if these agencies had not been furnished with clear and lawful parameters on which to base their evaluations. Neither could the matter be referred to the Minister of Health to implement relevant provisions into the system’s operating licence, or to entrust some other governmental bodies with developing codes of practice, since such codes could be subject to a variety of changes, given the very vague boundaries set by the Act on a Health Sector Database.

Consequently, the Supreme Court concluded that the Act on a Health Sector Database did not meet the requirements of Article 71, Paragraph 1 of the Icelandic Constitution to provide adequate protection against the risk of the data being traced back to the relevant data subjects for the purpose of protecting their personal privacy. It was in part owing to this conclusion that the Court held that the plaintiff was entitled to not have data on her deceased father transferred to the HSD. This case does highlight, among other things, that stricter security requirements may be imposed in cases of database containing health data that have been transferred without the explicit consent of the participants, versus cases where such transfers are based on consent.

3.3.3 Jurisdictional aspects of population databases

A database may involve more than one jurisdiction in a number of ways. First, the population database may itself be established as a partnership or joint venture across multiple jurisdictions. Second, the population database may be established in one jurisdiction but collect data and samples from populations from another or numerous jurisdictions. Third, the population database may send data and biological samples to researchers situated in other jurisdictions. Any international dimension raises even more challenges. These include the application of diverse legal systems, of divergent, and



possibly conflicting legislation and regulation, as well as the diversity of cultures and values with respect to scientific and medical research.

For example, one of the key challenges that population databases which aim to include data, information and biological samples from more than one jurisdiction will need to address pertains to consent mechanisms. If a common or commonly-accepted consent mechanism is not agreed across the jurisdictions at the establishment of the database, it may become difficult to establish such mechanism later. Such a failure may impede the collection of samples and data across borders. Another issue is the determination of which data, information and biological samples may or will be shared across borders. A lack of clarity on this issue may subsequently impede the exchange of such data, information or samples across borders. For example, the HSD Act provides that the database will not be transported out of Iceland and any processing may only be carried out in Iceland. The UK Biobank’s IP and Access Policy states that it will provide access to researchers from the UK and overseas. The application of other existing legislation and regulation in the relevant jurisdiction may have implications and will also need to be taken into consideration.

3.4 Privacy and confidentiality

Genetic research increasingly uses databases that contain human genetic and genomic information from many individuals, sometimes alone but often in combination with other information, of a medical, personal, etc. nature. This other information may include identifying information, such as names, addresses, or hospital identification numbers, medical information such as diagnoses and prescribed drugs, or other research information such as pharmacokinetic data.43 These databases can contribute significantly to the identification of genes associated with disease, the frequency of genetic variants in particular populations, and to achieving a new level of understanding of why individuals respond differently to drugs or other environmental factors. These databases, however, also raise concerns about privacy and confidentiality of the individuals and groups that contribute the data and information.44 Many of these concerns are not new,45 but the increasing size and scope of the databases and new informatics capabilities exacerbate the potential for breaches of privacy and confidentiality.

Collecting large number of data points about each individual means, however, that even if frank identifiers such as names, addresses, social security numbers or patient identification numbers are removed from data sets, the remaining data may uniquely identify an individual. If the data in a dataset are linkable to other datasets that contain identifiers, a unique identification can be attached, for example, a name or an address. Since both health data sets and identified non-health-related datasets are available, either publicly or for purchase, the potential for linking such sets together, and with research data sets, is not remote. For example, Latanya Sweeney points out that the majority of states in the United States have legislative mandates to gather hospital-level data, and that many collect outpatient data as well. The National Association of Health Data Organisations in the United States has recommended that data fields include patient numbers, ZIP code, race, birth date, gender, visit dates, ICD-9 codes, procedure codes, physician ID number, physician ZIP code and total charges. States have provided this information to researchers and sold it to companies.46 Although these databases do not specifically include patients’ names or addresses, they contain many data fields with personal information that individuals could be uniquely identified through the database and potentially linked to a publicly available database that includes names.



In the United States, examples of publicly available databases with identifiers are records of the Department of Motor Vehicles and voting lists. In records that are limited to a particular locale, many may be uniquely identifiable even by a single data field. For instance, of the 54 805 individuals in the 1997 voting list for Cambridge, Massachusetts, birth date alone uniquely identifies the name and address of 12% of voters, birth date and gender identify 29%, and birth date and 5-digit ZIP code identify 69%.47 Furthermore, these voting lists could be readily linkable to outpatient databases that include ZIP codes and birth dates, thus linking health data to names and addresses. These outpatient databases could in turn be linkable to research data sets if they include fields such as race, birth date, gender, and information such as ICD-9 codes, procedures or diagnostic information. This type of linkability, and therefore possible identification, raises privacy and confidentiality concerns. 48

Legal scholars have coined the term “genetic privacy”.49 Allen describes four different types of genetic privacy, which are the right to limits on access.50 These are: i) informational privacy of personal information, ii) physical privacy of the body, iii) decisional privacy in making personal choices, and iv) proprietary privacy and ownership interests in bodily information or body parts. Here, only informational privacy is discussed.

Privacy has generally been considered to mean the right to be left alone. In the context of genetic research, it usually means the right not to know certain genetic information. Genetic information obtained in the research context raises unusual privacy concerns because it has the potential to generate information beyond what was originally sought, and also because it raises the possibility that researchers who obtain the information may be obligated, in some situations, to provide that information back to the patients who contributed samples and data. For example, many genes have clinical implications for more than one disease or condition. The Apo E gene is thought to have implications both for predisposition to Alzheimer’s disease and cardiovascular disease. When individuals provide samples for research about one condition, databases may reveal information, sooner or later, about other conditions. These conditions may be associated with the genotype originally studied, or a genotype at another locus.

Confidentiality has generally been considered to capture the notion of a professional keeping information private once it has been revealed by a client, and to be part of the fiduciary relationship between the professional and the client/patient. In the context of genetic research, the most significant concern pertains to keeping genetic information that is collected in a research setting from third parties, such as health insurers or employers. However, in the computer age, concerns also arise with respect to violations of databases or the sale of genetic information for non-research purposes.

In discussing the issues of privacy and confidentiality, the type and size of database has considerable implications. Database systems that are built for the purpose of finding answers to only one or a limited number of research areas will involve different considerations from population databases such as the Estonian or Latvian initiatives. In the case of specifically oriented databases, both the health data and genotypic data are obtained with the direct consent of the subjects in the research projects. Commonly, in specifically oriented databases, the data are collected by health-care professionals, most often the very doctors who already treat the patient in question for the disease or condition that is the subject of the research project. Generally, when data collection is complete, all personally identifiable information on the patient is encrypted or removed. The anonymised data are subsequently moved to a database system for further analysis.



Scientists participating in processing the data and blood samples would generally not have participated in their processing when the data and blood samples were in identifiable form. Moreover, access to such databases is commonly limited to individual scientists, who have been granted a licence to carry out the research project in question, and to their employees, according to rules set forth by the parties responsible for the research project. In many cases, these types of database systems are not connected to external computer networks. However, in the case of specifically-oriented databases, the number of participants is often lower than in HGRDs. Given the more limited number of participants and the possibility of linking data to identifiers as discussed above, the risk of uniquely identifying individuals may also increase.

In the case of the Icelandic Health Sector Database (HSD), the data would have been obtained not on the basis of patients’ consent, but from agreements of the entity holding the medical records, information, etc. Data on Icelanders would have been transferred to the HSD, except for the medical and health data or information of those individuals who specifically would have notified the Director General of Public Health that they did not wish their data to be entered into the system. The HSD has not been built. Nevertheless, the proposed design on which the system was to be based assumed that the HSD would be networked and that a large number of researchers would have access, based on agreements with the HSD’s licence holder. According to the Act on a Health Sector Database, the database system could be used for the development of “new or improved methods of achieving better health, prediction, diagnosis and treatment of disease, to seek the most economical ways of operating health services, and for making reports in the health sector”. Such a broad approach to accessibility, while beneficial for fostering research, may increase the risks of breach of privacy and confidentiality. The relatively small population of Iceland, the relatively homogenous nature of this population, the considerable amount of genealogical data available on this population, combined with the inclusion of rather detailed medical and health data in the HSD, provide an example of circumstances that could increase the risk of breach of privacy and confidentiality. Conversely, depending on their design and operation, large-scale population databases may raise less of a risk to privacy and confidentiality, than specifically-oriented databases. For example, population databases where identifying information is removed from the samples and data entered into the database and are archived in a stand-alone database, without any external links, may provide a higher level of protection for participants.

Another factor that impacts considerably on the issue of privacy and confidentiality is the nature of the biological samples and data collected. Evidently, identified samples pose the most direct challenges for privacy and confidentiality. Consequently, many researchers have chosen, or have been required, to use coded, unlinked or unidentified samples. Even when direct identifiers have been removed from coded samples, they may still be identifiable and continue to present privacy and confidentiality concerns. Unlinked and unidentified data appear to create fewer privacy or security risks than linked or identified data. However, these data may still be identifiable, depending on the ease with which they can be re-linked to datasets (especially publicly available data such as voter registry information) containing identifiers through the data that were not removed in the de-identification process (quasi-identifiers). The increasing availability of public data in digital form and the increasing size of genetic research databases suggest that re-linkage may be increasingly feasible, either on a record-specific or probabilistic basis. The technical details of how data may be linked back with identifiers are beyond



the scope of this study. Nevertheless, methods for maintaining privacy and confidentiality of unlinked and unidentified data remain an important consideration.

Researchers often find coded samples more valuable (critically so, in some types of research) than unlinked or unidentified samples because the link with identity provides a way to follow individuals in longitudinal studies. Thus, this type of linkage should not be completely prohibited. To maintain privacy and confidentiality of databases with coded material, technical and procedural computer security (including control and monitoring of access to, and transport of, data) is essential. While it has been advocated that anonymisation through double randomised coding (encryption) is a better method, anonymisation does not necessarily imply de-identification. (See Chapter 5)

3.5 Public engagement in the establishment of a population database

When establishing a population database, public support is essential given that participation is voluntary. This implies that the collection of biological samples and data and their inclusion in an HGRD depend on the consent of participants.51 It may be difficult to estimate participation rates in projects for which the benefit is indirect, long-term and at population level, especially in resource-poor communities or populations with different beliefs, cultures or values. In order to build public trust, the distance between the research community and participants must be bridged. However, the manner in which public support is generated may vary considerably. In seeking to establish an HGRD, consideration of the methods to be used to engage the public, the information to be provided, and the manner in which it may be effectively communicated are, amongst others, important considerations.

3.5.1 Explaining the purpose of the population database

Some critics of population databases have focused on the disproportionate expense of such projects relative to their anticipated and yet unknown long-term benefits, on the diversion of funds from under-funded health care systems,52 or from other research areas,53 on the futility or ineffectiveness54 of such large-scale studies in light of existing collections of DNA. Thus, “although many geneticists agree that these databases will yield a plethora of useful information, it is not clear that they will deliver on their most ambitious promises”.55

In light of such criticisms, initiators developing population HGRDs will have to focus on the overall beneficial aspects of their project, ensure scientific transparency, educate the public and the media, and explain their projects to stakeholders in order to obtain funding.56 For example, the Estonian Genome Project, underlining the long-term benefits of the project for the Estonian health-care system and the Estonian population’s health, specified that although the Republic of Estonia put money into the establishment of the gene bank, could initially defend its initiative as being mostly funded through private means.57 Population database initiatives that rely heavily on public support may face even more challenges.

In the absence of any potential direct individual benefit and the possibility for commercialisation of the results, individuals may be reluctant to participate either for fear of misuse of their samples or data or for lack of interest. Research teams will need to consider the most appropriate approaches for creating interest in and incentives for participation. Given the reluctance to remunerate donors beyond reasonable expenses incurred for participation or to share benefits directly (since there are no direct individual



benefits), project research teams may have to emphasise the long-term beneficial aspects of the studies they undertake. For example, the UK Biobank has chosen to emphasis the long term benefits to research and the population’s health. The need to reassure potential participants with regards to confidentiality and privacy issues (notably access) as well as the commercial aspect of their undertaking (e.g. if industry is involved) will also need to be addressed.

The Estonian and Latvian projects differ from other such projects in that participants might obtain direct personal benefit from participation. For example, the Act governing the Estonian initiative grants participants the right to personally access their data, but not their genealogies, stored in the Gene Bank.58 As a result volunteers not only donate their samples with the intention to improve (public) health and play a role in the advancement of science (long-term, indirect, and population-level benefit) but also with the expectation that it may one day benefit their own health. Indeed, both projects contemplate participants’ access to data in the gene bank, accompanied by genetic counselling.59 In such situations, it may be easier to provide an explanation of the purpose and possible benefits of the initiative.

However, the possibility for gene donors to have access to their data has been the subject of some criticism. It has been considered by some to constitute a strategic move to lure people into the project: “[t]he personal gene card/map is a small but important part of a myth created in order to persuade Estonians that their health is going to be vastly improved”.60

3.5.2 Approaches to engaging the public

The manner in which public engagement or support was sought varies considerably across the different projects. The establishment of the Latvian,61 Estonian62 and Icelandic63 initiatives (e.g. creation, functioning, supervisory authority) and the general rules they abide by (e.g. consent, confidentiality principles) were decided politically by the enactment of legislation.

In the case of the Icelandic database, the public and media attention garnered, at the national and international level, was negative, arising in part from the use of the presumed consent approach. In Estonia, the project and the legislation which enabled its existence received much attention from the media (newspaper articles and television programmes).64 Furthermore, early on the Estonian Project commissioned Emor,65 an Estonian marketing and consulting company, to conduct public polls in order to assess the population’s knowledge of and opinion on the project.66 The Latvian initiative has been scrutinised67 as being “articulated in exclusively biotechnological and biomedical terms which allows excluding the public”. The legislation was discussed by scientists and the media and amended accordingly.68

Other HGRD projects were initiated by researchers or scientific institutions and have tended to involve more the population or communities in their development by way of consultation or engagement mechanisms.

In accordance with the Policy for the Responsible Collection, Storage, and Research Use of Samples from Identified Populations of the NIH National Institute of General Medical Sciences (NIGMS) Human Genetic Cell Repository of the Coriell Institute,69 the HapMap team carried out a community engagement or consultation process in each of the communities selected for the study. The process was designed to inform investigators about the organisational structure and culture of the communities and assess the



suitability of the methods to be used (consent, sample collection) by communicating with and informing the public. It was carried out by the NIH community leaders in Nigeria and local investigators in Japan and China, along with members of the International HapMap Steering Committee. Professionals in genetics, ethics and social sciences worked under the supervision of local government and ethics committees. Funded by the NIH or national organisations, the process took the form of dialogues with representatives of the community, focus and working groups, individual interviews and public meetings, polls and surveys. While the communities did not enjoy veto rights, the HapMap had declared that strong opposition to the project would have led to the withdrawal of the community from the research. This stance is in accordance with the NIGMS Human Genetic Cell Repository policy which states that “if a significant portion of the members of a community has serious reservations about the advisability of collecting samples from the community, collection should probably not be carried out”.70 The community engagement process, which lasted longer and was more expensive than expected, had to undergo modifications and changes to adapt to each of the communities selected. However, it proved fruitful in terms of feedback and information collected and enabled potential participants to express their concerns and participate actively in the debate. It is intended that the conclusions of the process will be published.71

The UK Biobank project conducted an extensive consultation process in which the public and interested parties (researchers, health professionals, ethicists, lawyers, and industry) were informed of the development of the project and asked to take part in the elaboration of the protocol and framework. These consultations, workshops, studies, public meetings and individual interviews were either commissioned by the initial partners (Wellcome Trust, Medical Research Council and Department of Health) or funded by grant programmes. The public consultation process was instructive and helpful as it assisted the team in drafting protocols, informed the framing and design of strategies for communication and further consultation, and assessed the interest of the sample population in the Project.72 In October 2000, prior to the completion of the Project development phase, an initial study was commissioned to examine public perceptions with regard to tissue sample collection73 (issues were raised, interest in the project ascertained and the MRC guidelines74 refined). This study was followed by a formal consultation with primary health-care professionals on recruitment issues75 and an informal consultation76 with primary health-care professionals who would be involved in the development of the scientific protocol. The draft protocol being completed, the team focused in 2002 and 2003 on sketching out an adequate ethical and governance framework. It conducted a public consultation to assess public trust in the project. Four issues were addressed: individual feedback at the time of recruitment, re-contact of participants, withdrawal and access to the database.77 As well, supplemental information was acquired when a second consultation was undertaken with groups that had been underrepresented in the first consultation, general practitioners and research nurses. In the course of these consultations, the UK Biobank realised that most fears related to issues of confidentiality and commercialisation. The Biobank noted participants’ reservations with regard to commercial use of their samples and data as well as industry involvement. However, the Biobank also noted that when industry’s role was explained to participants, they acknowledged that industry could be invaluable partners in the development of drugs.78 The Ethics Consultation Workshop79 greatly helped the Biobank in the drafting of the Ethical and Governance Framework.80 Finally, the team also met with the pharmaceutical industry.81 It is intended that the UK Biobank will stay in communication with public representatives throughout the project.82



The Marshfield Project’s approach was to hold focus groups in order to obtain the opinion of the public on the approach adopted by the team for recruitment and information of the public. Over the course of these meetings, the idea of results being published on the web was vetoed and as a result abandoned.83

The CARTaGENE Project was developed on the basis of prior and ongoing communication and consultation with the public and it is intended that it will only be implemented if sufficiently supported by the public. As “research on a particular population may bring about consequences for the entire population, particularly with respect to the interpretation of results, … individual consent as well as group support need to be obtained”.84 The CARTaGENE communication’s strategy includes consideration of at least two major elements: i) consultation of and communication with the public (study of the Quebec population’s perception of the project and related issues, workshops, public consultation, integrated approach to communication during the project); and ii) the setting up of a communication strategy for research results which would prevent discrimination and stigmatisation, ensure proper understanding, release and publication of results, including scientific publication, discussion panels and participation in conferences.85 Two workshops were organised in which the scientific rationale, legal aspects, ethical and governance framework, and communication strategies of the project were presented to specialists in various fields, including social scientists, ethicists, lawyers, clinicians, researchers and representatives of the public. The project benefited from these two events by re-evaluating certain aspects based on the constructive feedback received and by modifying its intended approach (semi-longitudinal study/double-coded information).86

3.5.3 Approaches to ongoing communication strategies

Whether or not the teams undertaking the project have decided to directly engage or consult the public in the course of the project development process, consideration needs to be given to what information will be provided and the manner in which it will be communicated to the relevant population, especially participants. Project teams may wish to consider communicating information about the purpose of the project, the intended objectives, and the manner in which results will be communicated to the population, if at all. Project teams may also consider providing the public with general information on genetics and diseases. Failures in communication may jeopardise a project’s feasibility and credibility. A communications strategy includes the approach to communicating and engaging the population or community at the point of establishing the project as well as ongoing communication with the population or community throughout the life of the project.

In the phase of establishing an HGRD project, project teams may wish to consider an integrated approach to communication. Such an approach may consist of providing information through the creation of a website, the printing of leaflets, the setting up of a toll-free telephone number, establishing contacts with the media and organising public meetings. For example, the UK Biobank team will use the services and expertise of the Wellcome Trust and a team of experts to ensure the adequate provision of information to the public. It “will look into a variety of ways for communicating with, including listening to, representatives from the general public, research users and the scientific community”.87 The UK Biobank’s Ethics and Governance Framework acknowledges that regular communication will be important for encouraging continued participation. The UK Biobank will employ a variety of media, including websites, helplines, newsletters, and public meetings, to inform participants about the development and use of the



biobank. It will also aim to communicate with the general public, research users and the scientific community.

The Marshfield Project will publish a bi-monthly newsletter on its website, to keep participants informed about the project and to provide general education in genetics.88 The CARTaGENE team carried out public consultations whose aims were to undertake a study of social perceptions and to develop a plan of communication.89

The HapMap Consortium wished to extend community engagement beyond the time of collection. It established a Community Advisory Group (CAG) in each of the communities in which samples were collected. The CAG will ensure that participants are properly informed and will oversee further uses. Indeed, the CAG will be consulted by the Coriell Institute to ensure that proposed further uses of samples are appropriate.90

Members of P3G, whose communication strategies are similar, have decided to take advantage of experience in individual projects to assess these strategies and implement more successful public relations and communication initiatives. To build public and governmental awareness, increase credibility and ultimately ensure public commitment, it was contemplated designating an international spokesperson (or a group of international spokespersons) to be entrusted with public communication and responsibility for a harmonised message for the P3G projects. This approach could include involving international journalists, a website and links to the respective projects’ websites.91

3.5.4 Implications of communication strategies

The need for transparent, clear and unambiguous language, a well-conceived scientific rationale, and a communication strategy tailored to the needs of the participants, the media, advocacy and community-based groups, and the public are key components of building public trust.92

Projects and initiatives may need to be re-worked or re-considered in light of the feedback from the public consultations. Some authors have indicated that the HapMap Consortium had to rethink and modify its research protocols and communication strategies for the Japanese population in light of the public’s lack of interest and unwillingness not only to attend public forums, but also to participate in the study. The factors identified as explaining the public’s reaction were insufficient financial resources and the inadequacy of the methods initially developed for other countries, when transposed to a society with a different social structure, cultural and ethical environment, and health-care system.93 Similarly, the CARTaGENE initiative modified some of its parameters in reaction to the public consultation. There appeared to be greater preference for employing a double-coding approach to confidentiality rather than anonymisation of the samples and data. In addition, re-contacting replaced the monitoring of medical files.94

Large-scale HGRD projects have realised that the media and community-based or advocacy groups may influence views in regards to the endeavour. In Iceland, the Health Sector Database to be developed was the target of criticism focused mainly on ethical issues and the commercially driven philosophy of the project. Opposition from the national and international scientific, legal and ethical communities and the campaign mounted against the database by the Mannvernd group induced approximately 20 000 individuals (about 7% of the total population) to opt out of the Health Sector Database.95 The Tongan HGRD, which was to be developed by AutoGen Limited, was abandoned due to the strong opposition from church and human rights groups.



The Estonian Project team, when first establishing its HGRD, was criticised by the media for its method of soliciting the public.96 It was alleged that the information made available was composed more of propaganda than an objective assessment of the strengths and weaknesses of the project.97 The head of information of the Estonian Project addressed the criticisms (e.g. inclusion of children, secrecy, funding, public awareness and support for the project) through a number of initiatives. For example, the Estonian Project team was interviewed, permitted journalists and TV cameras to visit their laboratories, and established a website which provides the public with articles published by the national and international community in order to respond to these initial criticisms.98

An important consideration in the establishment of an HGRD is not only the communications strategy for providing information to the relevant population but also contemplation of the manner in which criticisms will be addressed. Criticisms may arise from the population, the media, the scientific community, community-based or advocacy groups, etc. In each case, the feasibility of the project may depend on the manner in which these criticisms are considered and addressed and the manner in which this information is communicated to the relevant entity.



Notes

1. See www.hapmap.org/hapmappopulations.html.en (accessed 8 September 2006).

2. Cathy McCarthy, Director of the Personalized Medicine Research Project, “The Personalized Medicine Research Project”, P3G and North-American Stakeholders Meeting, Montreal, 26 August 2003.

3. CARTaGENE, See www.cartagene.qc.ca (accessed 11 May 2006).

4. Iceland, Act on a Health Sector Database no. 139/1998. See ministryofhealth.is/laws-and-regulations/nr/659#allt (accessed 11 September 2006).

5. According to Article 2 of the HSD Act, the legislation does not apply to the medical record systems of individual health and research institutions, data collections made in connection with scientific research into individual diseases or groups of diseases, nor to records kept by health and social security authorities on users of the health service and operation of the health service. The legislation does not apply to the storage or handling of, or access to, biological samples. The legislation does not apply to the storage or handling of biological samples which are governed by the Act on Biobanks. Iceland, Act on Biobanks No. 110/2000.

6. Article 3 of the HSD Act defines “genetic data” as any data, of whatever type, concerning the heriditary characteristics of an individual or concerning the pattern of inheritance of such characteristics within a related group of individuals. It also refers to all data on the carrying of any genetic information (genes) in an individual or genetic line relating to any aspect of health or disease, whether present as identifiable characteristics or not.

7. See Articles 7 and 10 of the HSD Act.

8. Author’s communication with deCODE spokesperson, 23 August 2006.

9. Estonia, Human Genes Research Act¸ (200), RT I 2000, 104, 685, entered into force 8 January 2001. See Section 2. Translation obtained from www.geenivaramu.ee/index.php?show=main&lang=eng (accessed 13 September 2006).

10. It should be noted that tissue sample implies mostly the taking of a blood sample.

11. Report on Estonian Genome Project, 2002, p.6. Please note that this scheme is somewhat out of date since the exclusive relationship with EGeen has been terminated. Thus, the activities designated by the dark blue as EGeen can be carried out by any company.

12. Report on Estonian Genome Project, 2002, p. 8.

13. See www.ukbiobank.ac.uk/about/overview.php (accessed 13 September 2006).



14. UK Biobank, Ethics and Governance Framework, May 2006, see www.ukbiobank.ac.uk/docs/EGF%20Version02%20May%202006.pdf (accessed 13 September 2006).

15. UK Biobank, Policy on Intellectual Property and Access, January 2005, Section B.1, see www.ukbiobank.ac.uk/docs/UKBiobankIPandAccesspolicyfirstpublicdraft11.1.5final2.pdf (accessed 13 September 2006).

16. UK Biobank, Policy on Intellectual Property and Access, January 2005, Section A.3, see www.ukbiobank.ac.uk/docs/UKBiobankIPandAccesspolicyfirstpublicdraft11.1.5final2.pdf (accessed 13 September 2006).

17. UK Biobank, Ethics and Governance Framework, May 2006, see www.ukbiobank.ac.uk/docs/EGF%20Version02%20May%202006.pdf (accessed 13 September 2006).

18. Estonia, Human Genes Research Act (2000), RT I 2000, 104, 685, entered into force 8 January 2001. See Section 16.

19. Jasper A. Bovenberg (2003), “Ownership and Commercialisation of Human Genetic Research”, Background Workshop paper.

20. Jasper A. Bovenberg (2003), “Ownership and Commercialisation of Human Genetic Research”, Background Workshop paper.

21. Taken from Notes to the Bill submitted to Parliament at 123rd session, 1998-99 (the Notes).

22. Iceland, Act on a Health Sector Database, No. 139/1998.

23. Estonia, Human Genes Research Act (2000), RT I 2000, 104, 685, entered into force 8 January 2001. See Section 2. Translation obtained from www.geenivaramu.ee/index.php?show=main&lang=eng (accessed 13 September 2006).

24. EGP, Genome Project Ends Cooperation with Current Financier, 27 December 2004, see www.geenivaramu.ee/index.php?lang=eng&show=uudised&id=172&PHPSESSID=197f1b8a33cef08aa94f377cb65cf596 (accessed 13 September 2006).

25. The establishment of the UK Biobank is not governed by specific legislation. Instead, its Funders committed to developing a public Ethics and Governance Framework (EGF) to set standards for the project and establish scientific and ethical safeguards. This has been put in place, and will be revised periodically.

26. UK Biobank, See www.ukbiobank.ac.uk/about/directors.php (accessed 13 September 2006).

27. UNESCO, Universal Declaration on the Human Genome and Human Rights, Adopted at 29th Session of the General Conference, on 11 November 1997. See portal.unesco.org/en/ev.php-URL_ID=13177&URL_DO=DO_TOPIC&URL_SECTION=201.html (accessed 14 September 2006).

28. United Nations Environment Programme, Convention on Biological Diversity, Adopted 1992. See www.biodiv.org/convention/default.shtml (accessed 14 September 2006).



29. Council of Europe, Convention for the protection of Human Rights and dignity of the human being with regard to the application of biology and medicine: Convention on Human Rights and Biomedicine, Adopted 4 April 1997, CETS No.: 164. See conventions.coe.int/Treaty/Commun/QueVoulezVous.asp?NT=164&CL=ENG (accessed 14 September 2006).

30. For a recent overview, including national laws in the EU and the United States, see the report prepared by the European Society of Human Genetics, “Eurogapp Project, 1999-2000”.

31. The creation and operation of the Gene Bank is governed by the Human Genes Research Act, the Personal Data Protection Act and the Databases Act and EU legislation, such as rules of free competition and the EU Directive on the protection of databases. The activities of EGPF and EGI are further governed by a system of agreements developed pursuant to Estonian and US laws.

32. The establishment of the UK Biobank was not established by specific legislation, but is nevertheless subject to British law.

33. See David E Winickoff and Richard N Winickoff (2003), “The Charitable Trust as a Model for Genomic Biobanks”, New England Journal of Medicine, Vol. 349, pp. 1180-1184; see note 26.

34. United Nations, Convention of the Law and Sea, Adopted 1982, see www.un.org/Depts/los/convention_agreements/convention_overview_convention.htm (accessed 15 September 2006).

35. The SNP Consortium, see http://snp.cshl.org/ (accessed 15 September 2006).

36. United States, White House, Office of the Press Secretary, Joint Statement by President Clinton and Prime Minister Tony Blair of the U.K., Statement released March 14, 2000, see http://clinton4.nara.gov/WH/EOP/OSTP/html/00314.html (accessed 15 September 2006).

37. HUGO Ethics Committee, Statement on Human Genomic Databases, adopted December 2002, see www.hugo-international.org/Statement_on_Human_Genomic_Databases.htm (accessed 15 September 2006).

38. RMGA, Statement of Principles: Human Genome Research (VERSION 2000), see www.cartagene.qc.ca/docs/enonce.pdf (accessed 15 September 2006); RMGA, Statement of Principles on the Ethical Conduct of Human Genetic Research Involving Populations, see www.rmga.qc.ca/doc/ENONCE2002.ENG.pdf#search=%22Ethical%20Conduct%20of%20Human%20Genetic%20Research%20Populations%22 (accessed 15 September 2006).

39. UK Biobank, Ethics and Governance Framework, see www.ukbiobank.ac.uk/ethics/efg.php (accessed 15 September 2006).

40. UK Biobank, IP and Access Policy www.ukbiobank.ac.uk/docs/UKBiobankIPandAccesspolicyfirstpublicdraft11.1.5final2.pdf (accessed 15 September 2006).

41. Ragnhildur Guomundsdottir v. The State of Iceland (2003) Icelandic Supreme Court, No. 151/2003. Please note that this analysis is based on the unofficial translation provided by Mannvernd, see www.mannvernd.is/english (accessed 12 September 2006).



42. Ragnhildur Guomundsdottir v. The State of Iceland (2003) Icelandic Supreme Court, No. 151/2003. Please note that this is extracted from the website of the Icelandic Data Protection Authority, see www.personuvernd.is/tolvunefnd.nsf/pages/60CD0F820FBD71D700256E4D004B1108 (accessed 12 September 2006).

43. R.E. Klein, J.T. Chang, M.K. Cho, K.L. Easton, R. Fergerson, M. Hewett, Z. Lin, Y. Liu, S. Liu, D,E, Oliver, D.L. Rubin, F. Shafa, J.M. Stuart, and R.B. Altman (2001) “Integrating genotype and phenotype information: an overview of the PharmGKB project”, The Pharmacogenomics Journal 1:167-170.

44. L.T. Vaszar, M.K. Cho and R.A. Raffin (2003), “Privacy issues in personalised medicine”, Pharmacogenomics 4:107-112.

45. W.W. Lowrance (2002), “Learning from experience: Privacy and the secondary use of data in health research”, The Nuffield Trust, London.

46. L. Sweeney (1997), “Weaving technology and policy together to maintain confidentiality”, Journal of Law, Medicine and Ethics 25:98-110.

47. L. Sweeney (1998), “Datafly: a system for providing anonymity in medical data”, in T. Lin and S. Qian (eds.), Database Security XI, Chapman and Hall.

48. L. Sweeney (1997), “Weaving technology and policy together to maintain confidentiality”, Journal of Law, Medicine and Ethics 25:98-110.

49. W. Zimmerli (1990), “Who has the right to know the genetic constitution of a particular person?”, pp. 93-102 in R. Chadwick and G. Bock (eds.), Human Genetic Information: Science, Law and Ethics, John Wiley and Sons, Chichester.

G. Annas, L. Glantz and P. Roche (1995), Drafting the Genetic Privacy Act: Science policy and practical considerations”, Journal of Law, Medicine and Ethics 23:360-6.

50. A. Allen (1997), “”Genetic privacy: emerging concepts and values”, pp. 31-59 in M. Rothstein (ed.), Genetic Secrets: Protecting Privacy and Confidentiality in the Genetic Era, Yale University Press, New Haven, Connecticut.

51. Except where the population database is established based on an opt-out approach, such as that selected for the Icelandic Health Sector Database.

52. See e.g. T. Tasmuth (2003), “The Estonian Gene Bank Project – an Overt Business Plan” Open Democracy, 29 May, www.opendemocracy.net/debates/article.jsp?id=9&debateId=79&articleId=1250 (accessed 18 May 2006).

53. See e.g. P. Ghosh (2003), “Will Biobank Pay Off?” BBC News, 24 September 2003, see news.bbc.co.uk/1/hi/health/3134622.stm (accessed 16 September 2006).

54. See e.g. Kaiser J. (2002), “Population Databases Boom, From Iceland to the U.S.”, Science, Vol. 298, p. 1158, www.sciencemag.org/ (accessed 19 May 2006).

55. See e.g. Kaiser J. (2002), “Population Databases Boom, From Iceland to the U.S.”, Science, Vol. 298, p. 1158, www.sciencemag.org/ (accessed 19 May 2006).



56. Cathy McCarthy, Ph.D., Director of the Marshfield Personalized Medicine Research Foundation explained that while the US scientific community seems supportive of the project, the NIH finds it hard to fund an infrastructure rather than a hypothesis-driven protocol. Cathy McCarthy, “The Personalized Medicine Research Project”, North-American Stakeholders Meeting, Montreal, 26 August 2003.

57. See the critique of the project and the answer given, T. Tasmuth (2003), “The Estonian Gene Bank Project – an Overt Business Plan” Open Democracy, 29 May, www.opendemocracy.net/debates/article.jsp?id=9&debateId=79&articleId=1250 (accessed 18 May 2006); and A. Koik (2003), “The Estonian Genome Project: a Hot Media Item” Open Democracy, 7 July, www.opendemocracy.net/debates/article-9-79-1335.jsp (accessed 18 Nay 2006).

58. Estonia, Human Genes Research Act, Article 11.

59. See www.privireal.org/content/dp/latvia.php (accessed 9 May�� and J. Šteinbergs, “Data Protection in the Project “Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc (accessed 18 May 2006).

60. T. Tasmuth (2003), “The Estonian Gene Bank Project – an Overt Business Plan” Open Democracy, 29 May, www.opendemocracy.net/debates/article.jsp?id=9&debateId=79&articleId=1250, p. 3 (accessed 18 May 2006).

61. www.privireal.org/content/dp/latvia.php (accessed 9 May 2006); The Human Genome Research Act cited in V. � � �� “Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc and A Putnina, “Exploring the Articulation of Agency: Population Genome Project in Latvia”, www.ifz.tugraz.at/index_en.php/filemanager/download/122/putnina.pdf (accessed 11 May 2006).

62. Estonia, Human Genes Research Act, 13 December 2000, at www.legaltext.ee/text/en/X50010.htm (accessed 16 May 2006).

63. Iceland Minister of Health and Social Security, Act on Biobanks No. 110/2000, 13 May 2000, http://eng.heilbrigdisraduneyti.is/laws-and-regulations/ (accessed 12 May 2006); see also the associated regulation of the Ministry of Health and Social Security, Regulations on the Keeping and Utilisation of Biological Samples in Biobanks No. 134/2001, Reykjavik, 6 February 2001, http://eng.heilbrigdisraduneyti.is/laws-and-regulations/ (accessed 12 May 2006).

64. As noted by Mylène Deschênes and Geneviève Cardinal (2003), the “[m]edia first covered the topic in spring 1999” and “[s]ince that time, approximately three hundred articles, directly or indirectly connected to the projects, have been published by the different media”. See “Surveying the Population Biobankers” p. 44 in Bartha Maria Knoppers (ed.), Population and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden.

65. See www.emor.ee/eng/ (accessed 16 May 2006).



66. See A. Koik (2003), “The Estonian Genome Project: a Hot Media Item”, Open Democracy, 7 July, at www.opendemocracy.net/debates/article-9-79-1335.jsp (accessed 16 May 2006); see also www.geenivaramu.ee (accessed 16 May 2006).

67. A. Putnina, “Exploring the Articulation of Agency: Population Genome Project in Latvia”, www.ifz.tugraz.at/index_en.php/filemanager/download/122/putnina.pdf (accessed 11 May 2006), p. 9.

68. For criticisms, see more generally A. Putnina, “Exploring the Articulation of Agency: Population Genome Project in Latvia” www.ifz.tugraz.at/index_en.php/filemanager/download/122/putnina.pdf (accessed 11 May 2006).

69. NIGMS Human Genetic Cell Repository, “Policy for the Responsible Collection, Storage, and Research Use of Samples from Identified Populations”, 25 August 2004 (last revision), http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html (accessed 16 May 2006). See also National Institutes of Health, “Points to Consider When Planning a Genetic Study that Involves Members of Named Populations”, (last update 25 January 2005), www.nih.gov/sigs/bioethics/named_populations.html (accessed 16 May 2006).

70. NIGMS Human Genetic Cell Repository, “Policy for the Responsible Collection , Storage, and Research Use of Samples from Identified Populations”, 25 August 2004 (last revision) http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html (accessed 16 May 2006).

71. NIGMS Human Genetic Cell Repository, Policy for the Responsible Collection, Storage, and Research Use of Samples from Identified Populations, 25 August 2004 (last revision), http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html (accessed 16 May 2006). On the HapMap public engagement process, see e.g. “Background on Ethical and Sampling Issues Raised by the International HapMap Project”, Release Genetic Variation Mapping Launch, NIH News Advisory, October 2002, at http://genome.gov/10005337 (accessed 16 May 2006); E. Suda and D. Macer (2003), “Ethical Challenges of Conducting the HapMap Genetics Project in Japan”, in S.Y. Song, Y.M. Koo and D.R.J. Macer (eds.), Bioethics in Asia in the 21st Century, Eubios Ethics Institute.

72. See D. Shickle, R. Hapgood, J. Carlisle, P. Shakcley, A. Morgan and C. McCabe (2003), “Public Attitudes to Participating in UK BioBank: A DNA Bank, Lifestyle and Morbidity Database on 500,000 Members of the UK Public Aged 45-69”, p. 323 in Bartha Maria Knoppers (ed.), Populations and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden.

73. The Medical Research Council, the Wellcome Trust, Public Perceptions of the Collection of Human Samples, London, October 2000, at www.ukbiobank.ac.uk/docs/perceptions.pdf (accessed 16 May 2006).

74. The Medical Research Council (2001), “Human Tissue and Biological Samples for Use in Research”, London, April, www.mrc.ac.uk/pdf-tissue_guide_fin.pdf (accessed 16 May 2006).

75. The Medical Research Council, the Wellcome Trust (2001), “Consultation with Primary Health Care Professionals on Issues Relating to the Recruitment of Patients to a DNA Collection Study”, London, 31 January, www.ukbiobank.ac.uk/docs/GPreport.pdf (accessed 16 May 2006).



76. “Informal Consultation with Health Care Professionals”, Protocol Development Workshop, London, April 2001.

77. The Medical Research Council, the Wellcome Trust (2002), “The UK BioBank A Question of Trust: A Consultation Exploring and Addressing Questions of Public Trust”, London, March, at www.ukbiobank.ac.uk/docs/consultation.pdf (accessed 16 May 2006).

78. The Medical Research Council, the Wellcome Trust (2002), “The UK Biobank a Question of Trust: A Consultation Exploring and Addressing Questions of Public Trust”, London, March www.ukbiobank.ac.uk/docs/consultation.pdf (accessed 16 May 2006).

79. The Wellcome Trust, Medical Research Council, Department of Health, “The UK Biobank Ethics Consultation Workshop”, London, April 2002, www.ukbiobank.ac.uk/docs/egf-comment-version.doc (accessed 16 May 2006).

80. UK Biobank, Ethics and Governance Framework, www.ukbiobank.ac.uk/ethics/ethicsgov.php (accessed 18 May 2006).

81. Report from the meeting which took place in April 2003, see www.ukbiobank.ac.uk/ethics/consultations.php (accessed 18 May 2006).

82. See www.ukbiobank.ca.uk/ (accessed 18 May 2006).

83. The Personalized Medicine Research Project, www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pers_med_res_prj (accessed 11 May 2006); Cathy McCarthy, Director of the Personalized Medicine Research Project, “The Personalized Medicine Research Project”, P3G and North-American Stakeholders Meeting, Montreal, 26 August 2003.

84. M. Deschênes and G. Cardinal (2003), “CARTaGENE Project: Governance and Ethical Clearance”, Centre de recherche en droit public, Université de Montréal, May, www.cartagene.qc.ca/docs/enonce2003_1.pdf (accessed 11 May 2006), p. 6.

85. B. Godard and E. Racine (2003), “La Stratégie de Communication de CARTaGENE”, CARTaGENE, Atelier II, Montreal, 11 June, www.cartagene.qc.ca/atelierII/Questions.htm (accessed 11 May 2006).

86. The first semi-public workshop was held in Montreal in June 2001; the second took place in Montreal on 11 June 2003. The presentations as well as discussions of the second workshop are available at www.cartagene.qc.ca/atelier2.cfm (accessed 18 May 2006). See e.g. B. Godard and E. Racine, “La Stratégie de Communication de CARTaGENE”, at www.cartagene.qc.ca/atelierII/questions.pdf (accessed 18 May 2006 See also, Claude Laberge, “Update of the CARTaGENE Project”, Montreal, November 2003.

87. UK Biobank, Ethics and Governance Framework, www.ukbiobank.ac.uk/ethics/ethicsgov.php (accessed 18 May 2006), p. 12.

88. Personalized Medicine Research Project, Why Should I Participate?, www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pmrp_faqs (accessed 18 May 2006)

89. The texts of this conference have been compiled in Bartha Maria Knoppers (ed.) (2003), Populations and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden.



90. On the HapMap public engagement process, see “Background on Ethical and Sampling Issues Raised by the International HapMap Project”, Release Genetic Variation Mapping Launch, NIH News Advisory, October 2002, http://genome.gov/10005337 (accessed 18 May 2006); E. Suda and D. Macer (2003), “Ethical Challenges of Conducting the HapMap Genetics Project in Japan” in S.Y. Song, Y.M. Koo and D.R.J. Macer (eds.), Bioethics in Asia in the 21st Century, Eubios Ethics Institute.

91. See P3G Montreal Meeting, Workshop Minutes, “Working Groups 5: Public Engagement”, Montreal Meeting 2-4 July 2003; see also Montreal Meeting Executive Summary, see http://www.p3gconsortium.org/events/2003Montreal/ExecutiveSummary.pdf (accessed 15 September 2006).

92. We will not provide details on consent and public engagement mechanisms at this stage.

93. For a more detailed discussion on such ethical challenges in Japan, see E. Suda and D. Macer (2003), “Ethical Challenges of Conducting the HapMap Genetics Project in Japan” in S.Y. Song, Y.M. Koo and D.R.J. Macer (eds.), Bioethics in Asia in the 21st Century, Eubios Ethics Institute; see also D.R.J. Macer (2003), “Ethical Considerations in the HapMap Project: An insider’s personal view”.

94 See http://www.cartagene.qc.ca/focus.cfm (accessed 15 september 2006).

95. For insight on the debate see G.J. Annas (2000), “Rules for Research on Human Genetic Variation – Lessons from Iceland”, New England Journal of Medicine, Vol. 342, No. 24, p. 1830 ; J. Gulcher and K. Stefansson (1999), “An Icelandic Saga on a Centralized Healthcare Database and Democratic Decision-Making”, Nature Biotechnology, Vol. 17, No. 7, p. 620; J. Kaiser (2002), “Population Databases Boom, From Iceland to the U.S.”, Science, Vol. 298, p. 1158; M.R. Anderlik and M.A. Rothstein (2001), “Privacy and Confidentiality of Genetic Information: What Rules for the New Science”, Annual Review of Genomics and Human Genetics, Vol. 2, p. 401, at http://arjournals.annualreviews.org/doi/full/10.1146/annurev.genom.2.1.401 (accessed 18 May 2006); B. Andersen and E. Arnasson (1999), “Iceland’s database is ethically questionable”, British Medical Journal, Vol. 324, p. 443.

96. See e.g. F. Storms (1999), “Storms Brews over Gene Bank of Estonian Population”, Science Magazine, Vol. 286, No. 5443, p. 1262, www.sciencemag.org/cgi/content/full/286/5443/1262 (accessed 18 May 2006).

97. See T. Tasmuth (2003), “The Estonian Gene Bank Project – an Overt Business Plan” Open Democracy, 29 May, www.opendemocracy.net/debates/article.jsp?id=9&debateId=79&articleId=1250 (accessed 18 May 2006); see also A. Weber (2001) “Sold Nation”, Süddeutsche Zeitung, 23 November.

98. See e.g. A. Koik (2003), “The Estonian Genome Project: a Hot Media Item”, Open Democracy, 7 July, www.opendemocracy.net/debates/article-9-79-1335.jsp (accessed 18 May 2006).

CHAPTER 4. DATA AND SAMPLE COLLECTION AND MANAGEMENT – 83


Chapter 4

Data and Sample Collection and Management1

The heart of a population database is the data, information and biological samples collected, stored and made available for research uses. The collection, storage, use and disposal of data and biological samples raise a number of issues ranging from the nature of the information to its ownership to the question of consent.

4.1 Data and samples

4.1.1 Nature of genetic information

Various sources provide indications about an individual’s genetic inheritance. Some of these sources are genetic in nature (e.g. testing for the BRCA1 and BRCA2 genes) while others are not (e.g. a blood test for high cholesterol). The issue of what constitutes genetic information is more complex than this simple dichotomy implies. For the purposes of its report, the Australian Law Reform Commission defined “genetic information” as “information gained from DNA (or related) testing or tissue samples which may be the subject of testing”.1

With the advent of genetic testing, it has been argued by those who support the “genetic exceptionalism” approach that genetic information is exceptional or special and thus should be treated differently from other information. For these advocates, genetic information is unique and raises particular issues with respect to privacy and discrimination. On the other hand, the “genetic inclusivists” argue that it is no more special than other personal, health-related or medical information.

The proponents of the “genetic exceptionalism” perspective argue that genetic information is uniquely powerful and uniquely personal and thus merits unique privacy protection.2 They maintain that a person’s DNA can predict that person’s likely future for a variety of medical conditions. Moreover, they argue that genetic information can divulge personal information not only about the individual but also about the individual’s parents, siblings and children. Furthermore, as there exists a history of use of genetics to stigmatise and victimise, there is a legitimate concern that genetic information may once again be misused in a discriminatory manner.

The “genetic inclusivists” argue that genetic information is not fundamentally different from other types of health and medical information. Rather, they are concerned

This chapter also draws on the texts prepared for the Tokyo Workshop: Jasper Bovenberg, Mildred Cho,

Clementine Sallé and Erich Wichmann. See Bibliography.

84 – CHAPTER 4. DATA AND SAMPLE COLLECTION AND MANAGEMENT


that the motivation behind the consideration of genetic information as exceptional is based on the perception that genetic information is a mysterious, powerful and inexorable force that dominates and will control an individual’s future.3 The genetic inclusivists argue that genetic exceptionalism is flawed because: i) strict protection of autonomy, privacy and equal treatment of persons with genetic conditions may threaten the accomplishment of other public goods; and ii) there is no clear line of demarcation between genetic data and other health data. In other words, genetic information is not unique in its ability to predict an individual’s future health. Rather it provides a range of probabilities which need to be coupled with information on the individual’s lifestyle (e.g. smoker, heavy drinker, etc.) and other environmental factors (e.g. lives near polluted river, electrical power lines, etc.) in order to obtain a clearer view of the individual’s health status.

The workshop experts concluded that there was no consensus on granting genetic information a special status. However, it was considered important to note that existing laws, regulations and policies should ensure that they are applicable to genetic information in addition to other types of health-related information.

4.1.2 Type of data and samples

Issues pertaining to the type of data and sample will be examined under two broad categories. The first category will pertain to the specific types of data and samples to be collected and stored in the HGRD. The second category pertains to the manner in which they are collected and stored.

Almost all population databases intend to collect a blood sample from participants. From these samples, usually the intent is to extract DNA for subsequent analysis. Some databases, such as the UK Biobank, intend to collect urine samples in addition to blood samples from participants. Almost all of the population databases will take basic physical measurements of each participant (e.g. blood pressure, height, weight, etc.). The UK Biobank, for example, will also measure grip strength and lung function. The CARTaGENE initiative adds cholesterol and blood sugar levels.

Most of the population databases will ask participants to complete questionnaires. However, there is considerable variation amongst the questionnaires with respect to the level of detail and the breadth of areas covered. For example, the UK Biobank’s questionnaire will cover lifestyle (e.g. diet, exercise, smoking, alcohol, etc.) and other factors (e.g. such as mood, cognitive function, medical history). The Latvian questionnaire will also request information on the demographic and socioeconomic background of the participant. The questionnaires of the Estonian and Latvian initiatives are more detailed as they also contain questions pertaining to the participant’s genealogy.4

Some of the biobanks will also incorporate into their database information gathered from the medical records of each participant. However, most of the HGRDs are not explicit as to which data will be collected from medical records. The UK Biobank will require a link to full record of medical and other health relevant data, both past and ongoing. The UK Biobank will collect data from NHS record systems (especially GP, hospital and dental records, prescriptions) and other relevant record systems (such as disease registries or occupational health records). In order to accumulate essential data, the UK Biobank considers it essential to be able to track health events, the development of disease and the course of treatments. Interestingly, for the Estonian project, a “description of the state of health” is prepared for each participant based on the information gathered from the questionnaire and from medical records by the chief or an



authorised processor. Most of the biobanks also do not specify which data and information will be excluded. For example, one issue is whether information about past or present use of illicit substances will be included in the database. The CARTaGENE initiative explicitly indicates that it will not have access to any medical file held by a doctor or in hospital archives.

Table 4.1 Comparison of data collected by selected biobanks5

United Kingdom CARTaGENE Iceland Estonia

Measurements: Blood pressure, pulse rate, height, weight, body fat, grip strength and lung function. Source: Participant.

Measurements: Height, weight, waist, hips, body fat, blood pressure, oral acidity, cholesterol level and blood sugar, heart and blood vessel. Source: Participant.

Measurements: All measurement data included in past, present and future medical record of participant, as negotiated between health institutions holding the records and the operator of the HSD. Source: medical record.

Measurements: Height, blood pressure, heart rate and weight. Source: Participant.

Medical: Illness, treatment, hospitalisation, or death. Source: Medical and other records and re-contact of participant.

Medical: Health data (medical diagnostics by category, reimbursement for medication, hospitalisation and death) from the last 5 and the next 50 years (mandatory) Source: government hospitalisation database, public drug insurance program, cancer registry. No access to medical record.

Medical: Diagnosis, dates of admission/discharges excretion, medications, lab results

Medical: Diseases, medication, Source: Participant’s response to questionnaire

Health, environment, diet: Health, memory, work, family, medications, vitamins or supplements, surgical history, birth-weight, whether breastfed and place of birth. Source: Participant’s response to questionnaire.

Health, environment, diet: No specification. Source: Participant’s response to questionnaire.

Health, environment, diet Marital status, personal characteristics, occupation, education, allergies, nutrition, metabolism, self-preservation, perception, intellect.

Health, environment, diet: Employment, physical exercise, nutrition habits, smoking, alcohol, questions for women. Source: Participant’s response to questionnaire.

Updates: Re-contact, re-questioning and re-measurements of participants is optional, but foreseen. Long term access to medical and other records.

Updates: Optional: five annual updates of health data.

Updates: ‘Supplements’ to be added by chief or authorised processor on the basis of statements by the participant and data on the participant stored in medical institutions.

Genealogy: Not mentioned.

Genealogy: Optional.

Genealogy: Family pedigree. Source: public genealogy data.

Genealogy: Family medical history. Source: Participant’s response to questionnaire and other databases and genetic research.

The second category of issues pertains to the manner in which the data and samples

are collected and stored. The key question is the level of identification for the collection and storage of the biological samples and data. Various terms have been used to describe



the different levels of identification of data and biological samples collected within HGRDs.6 In order to facilitate discussion, the definitions employed by the US National Bioethics Advisory Council (US NBAC) in its Research involving Human Biological Materials: Ethical Issues and Policy Guidance: Report and Recommendations of the National Bioethics Advisory Council7 will be broadly followed. Other definitions exist, such as the definition of de-identified data used in the US Health Insurance Portability and Accountability Act (HIPAA).8

The US NBAC Report indicates:

Unidentified samples – sometimes also termed “anonymous”, “anonymisation” – are those from an unidentified collection of human biological specimens.

Unlinked samples – sometimes also termed “anonymised”– are those that lack identifiers or codes that can link samples to identified specimens or particular individuals. Typically, unlinked samples are sent from identified human biological specimens to investigators without identifiers or codes so that identifying particular individuals through the clinical or demographic information that is supplied with the sample or biological information derived from the research would be extremely difficult for the investigator, the repository or a third party. Unlinked samples also include samples that are already in an investigator’s possession and whose identifiers have been removed by a disinterested party.

Coded samples – sometimes also referred to as “reversibly anonymised”, “linked” or “identifiable” samples – are those from identified specimens. However, these samples do not include any identifying information, such as patients’ names or social security numbers. Rather, they are accompanied with codes. In such cases, although the repository (or its agent) retains the ability to link the research findings derived from a sample with the individual source by using the code, the investigator (or one reading a description of the research findings) would not be able to do so.

Identified samples – are those supplied from identified specimens with personal identifiers (such as names or patient numbers) which are sufficient to allow the researcher to link directly the biological information derived from the research with the individual from whom the sample was obtained.

Unidentified samples can be obtained only from collections of unidentified materials. Given the scarcity of truly anonymously-collected human biological materials, few research samples are truly unidentifiable.9

Coded (whether single or double-coded) or reversibly anonymised samples are the most commonly employed approaches to sample storage in research. Coded samples may be used when a researcher anticipates the need to obtain additional information about or from the source, to provide information to the source, or to obtain additional samples over time. For coded samples, the identification of the individual is not kept with the sample nor provided to the researcher. Instead, each sample receives a unique identifier, and the repository, for quality control or other purposes, maintains the possibility of linking the unique identifier and the identity of the individual. This link also provides the potential for a one-way flow of information from the biobank to the participant and, at times, for a reverse flow of information from the participant to the repository. Thus, the use of coded or reversibly anonymised samples may allow researchers to obtain follow-up data on treatment, recurrence and survival and may permit the communication of research findings to participants or to their physicians.



While the specific measures that will be employed by the various projects differ, most of them have opted for a policy of coded (single or double) or reversibly anonymised data and biological samples. In light of the longitudinal nature of these projects, the possibility to link back with identifying information will be essential to updating and verifying the data as well as re-contacting the participants. The CARTaGENE project provides the most detailed information about the complex, multi-layer approach that will be employed to protect the participant’s confidentiality.

Some projects, such as the HapMap, have opted for complete anonymisation. In the case of the HapMap, anonymisation is a sine qua non condition as it aims to create databases that make the data available in the public domain. In other circumstances, anonymisation, however, has been much criticised, as it may prevent further updating of the information contained in the HGRD, thereby eliminating the possibility of a longitudinal study and the right of withdrawal. In the case of the CARTaGENE project, the initial decision was to anonymise the samples and data. However, after public consultations the team re-considered its decision and determined that data would be double-coded and the study will be semi-longitudinal.10

Table 4.2 Comparison of data confidentiality approaches adopted by selected biobanks11

UK Biobank CARTaGENE Iceland Estonia

Coding: Identifying information is removed from study data and samples after collection. All study data are stored anonymously in the study database, with personal identifiers kept separately under strict control. The code will have no external meaning. Only those with access to the “key” to the code will be able to re-link the participants’ identifying information with the data and samples.

Coding: All data from the clinical visit, the questionnaires and the blood test will be coded differently. Each CARTaGENE partner holding information about participants will have its own code. No partner will hold the totality of the codes associated with a participant’s personal information.

Coding: Personal identification shall be coded one-way. One-way coding is defined as the transformation of words or numbers into an incomprehensible series of symbols which cannot be traced by means of a decoding key. Information may be transferred to the HSD through the Encryption Agency of the Data Protection Authority (DPA).

Coding: Each sample, description of DNA, description of state of health and genealogy receives a unique code consisting of at least sixteen random characters.

Key: Code can only be accessed by a limited number of UK Biobank staff solely for the purposes of linkage. UK Biobank staff signs confidentiality agreements and are trained in the appropriate handling of personal data.

Key: There is only one organisation with information identifying participants; however, this organisation has no access to any medical information.

Key: Not applicable for data (one-way coding). Biological samples shall be kept securely and labelled but stored without personal identification. The linking of biological samples with personal identification shall be in keeping with standards laid down by the DPA.

Key: The chief processor shall store the written consent together with the code indicated thereon in the database of the Gene Bank and it shall be the only possible key for decoding. Decoding can be performed only by the EGPF, in cases and for purposes stipulated by law.



4.2 Ownership of data and samples

HGRDs raise issues with respect to the ownership of the data and biological samples collected. Irrespective of whether the database is a private, public or mixed endeavour, the question of who should be able to claim property title to the data and biological samples is raised. Should the data and samples belong to the participant or to the HGRD? Equally important is the issue of remuneration for the provision of the biological samples and data. Should participants be remunerated? If so, what should be the criteria for such remuneration? This is a separate issue from benefit sharing, which is discussed in Chapter 6.

As discussed above, a population database may be established as a private endeavour, a public endeavour or a mixed undertaking. In reviewing the Icelandic, Estonian and the United Kingdom undertakings, which were initially established as very different endeavours, it is interesting to note the commonality of characteristics between these three different models. In none of these databases is the participant either entitled to a property claim with respect to their data or samples or remunerated. However, in all three cases, the participant retains some control over the data and sample via either an opting out procedure or the removal of the data/sample from the database. Finally, in all three cases the database is owned by the collecting entity. The policies applicable to each of the three examples are outlined below.

In Iceland, the HSD Act does not contain a provision on the ownership of the data to be entered into the HSD. According to the Parliamentary Notes, health data cannot be subject to ownership in the usual sense. However, the Licence provides that all data that enter the HSD are the common property of the Icelandic nation and in the care and under the responsibility of the Icelandic Government. Rather than proclaiming that the data are owned by the Licensee, the HSD Act provides that only those who have an operating licence can collect the data (i.e. create and operate the HSD). Strictly speaking, the Act does not limit the number of licences, but according to the Parliamentary Notes, the Licence is exclusive. The Licence does not confer unencumbered “ownership” of the HSD; it is contingent upon certain conditions and subject to monitoring by the government; it is not perpetual but temporary. The Licensee is not allowed to transfer any data to which they are granted access to other databases or merge or connect them with activities taking place elsewhere, without the consent of the authorities. Also, the HSD may not be transported out of Iceland and may only be processed in Iceland; the Licensee may not allow direct access to the HSD. A patient may at any time request that his/her data not be entered into or, apparently,12 removed from the HSD. Upon expiry of the Licence, the government may, in certain circumstances, claim title to the HSD. There is no remuneration for participants, but they may choose to opt-out.

The Estonian Genome Project has right of ownership of tissue samples, descriptions of state of health, other personal data and genealogy. This right is created from the moment the tissue sample or personal data are provided or the moment the state of health or genealogy is prepared. Tissue samples and uncoded information owned by the EGPF are not transferable. Pursuant to the informed consent form, the gene donor is aware that the right of ownership of the tissue sample, of the description of their state of health and of other personal data and genealogy shall be transferred to the EGPF. Gene donors may apply for the destruction of data which can be decoded and have the right to access their own data stored in the Gene Bank at no cost and to prohibit the supplementation, renewal and verification of descriptions of their state of health stored in the Gene Bank. Gene



donors do not have the right to access their genealogies. There is no remuneration for participants.

The UK Biobank Ltd. will be the legal owner of the database and the sample collection. Participants will not have property rights in the samples. UK Biobank Ltd. will serve as the “steward” of the resource (.e. database, collection, etc.), maintaining and building it for the public good in accordance with its purpose. According to its Ethics and Governance Framework, ownership confers a number of rights, such as the right to take legal action against unauthorised use or abuse, the right to sell or destroy the samples, etc. However, the Biobank has affirmed that it will not exercise all of these rights, such as the right to sell samples. Participants may withdraw at any time and are not remunerated.

Population databases are built as for-profit, not-for-profit and mixed databases. Another basic issue is whether privatisation should be allowed in the first place. Opponents to privatisation argue that, in view of their nature, human health and genetic data cannot be subject to appropriation and/or that these data are a pre-competitive, national resource, recorded at the public’s expense. Proponents of privatisation argue that the licences do not grant exclusive access to existing or future health and/or genetic data but only the right to create and operate a new database, at great expense and with uncertainty as to its economic and scientific viability. They ague that if Iceland and Estonia cannot or do not want to finance their respective databases out of public funds, no private party will finance the work without its being granted exclusivity. Since the collected information will have commercial value regardless of whether it has been assembled by a private or a public institution, privatisation should be allowed at some point in order to provide the incentives necessary to produce new diagnostics, drugs and/or therapies. Also, the “privatisation” of population databases need not be absolute. For example, the privatisation foreseen for an Icelandic HSD is limited, in the sense that the Licence is granted for a specified period, the government retains a certain interest in the HSD and its exploitation, and the exploitation thereof by the Licensee is subject to governmental regulation. On the other hand, the public character of the Estonian Gene Bank is limited in the sense that only legal ownership vests with the government whereas beneficiary ownership had vested with the company that financed the Gene Bank. While the UK Biobank model is completely public, even this model foresees the commercialisation of research results.

4.3 Consent

Informed consent is one of the most complex issues for medical/scientific research, and especially human genetic research databases. Within the medical/scientific field, informed consent generally presumes the ability to indicate clearly to the participant the use and purpose of the particular research activity. While this is feasible for purpose-specific research, the very nature of HGRDs renders the provision of this type of information difficult. Given that in the context of an HGRD the purposes for which the data and biological samples are collected and the uses for which they may be employed will often only be known in a very general manner, what constitutes informed consent raises a number of challenges. Many authors have queried whether the traditional model of informed consent is applicable in the context of HGRDs or whether new models/paradigms should be developed. For example, some authors have advocated a “blanket” consent. Others have advocated a general consent for limited purposes, with the undertaking to return to the participant should the proposed uses go beyond those limited purposes.



4.3.1 Nature of consent

While some commentators view consent from the perspective of a written form, others view consent as a process. “At least since the Nuremberg Code was enunciated in 1947, it has been understood that the voluntary, competent, informed and understanding consent of the research subject is a necessary (but not sufficient) condition for ethical and lawful experimentation involving humans.”13 It has been advocated that while the consent form is important, the substance of the consent process is vital.

As the US National Bioethics Advisory Commission (NBAC) has highlighted, one of the issues with informed consent is that the focus is placed on the form rather than on the substantive process. The Commission noted that institutional review boards (IRBs) and investigators often tend to focus on the disclosures found in the consent form rather than on the ethical standard of informed consent and what is entailed in the process of obtaining informed consent.14

It has been advocated that the informed consent process should include consideration of a number of elements. While the diverse aspects of informed consent are complex, they are briefly summarised. One element is that of disclosure of the relevant risks and benefits of the procedure. A second element is comprehension by the participant of those relevant risks and benefits. These aspects entail more than the simple physical risk of a needle stick or a buccal swab. While such activities may present some physical risks, the greater risks that must be acknowledged as part of the consent process include those of psychosocial harms, such as potential stigmatisation or intra-familial conflict. Similarly, research may create or augment the risk of stigmatisation or discrimination at a group level and participants would need to be aware of this potential. For example, negative racial stereotyping may be one such risk. Another consideration is competence on the part of the participant to make a decision of whether to accept the treatment or to participate in the research. A fourth aspect is choice – an expressed decision to accept the treatment or participate in the experimentation. A fifth element is “voluntariness” of the choice to accept treatment or to participate in research.15 With regards to these latter elements, the issue of prospective authorisation raises a number of challenges, including what constitutes consent. As it would be difficult for participants to provide specific, informed consent when approved research protocols are not in place, consideration of the type of consent possible arises.

4.3.2 Consent for population databases

Traditionally, medical and scientific research is carried out on human subjects once specific informed consent has been obtained to participate in research.16 The challenges of obtaining informed consent, as outlined in the previous section, arise even more in the context of population databases. For example, one issue is whether data and biological samples collected for a specific project may be used only for that project (consent to one-time use), for other research queries on related topics (purpose-limited consent), or more broadly for any research purpose, such as for a biobank (blanket consent). In the context of HGRDs, this segregated approach to consent may also raise some administrative challenges that may make it unfeasible. Another aspect that may be important in the context of HGRDs is whether only a specific set of known researchers will have access to the data, or whether others, either known or unknown at the time of the consent, will be able to have access.17 Some have argued that the informed consent process, in the context of population databases should also specify the circumstances, if any, under which subjects may be re-contacted and genetic or other information may be reported back to



subjects, physicians and/or family members.18 Fundamentally, the guidance that is to be provided in the consent process should focus on how to provide appropriate information to prospective participants, how to promote prospective participants’ comprehension of such information, and how to ensure that participants continue to make informed and voluntary decisions throughout their involvement in the research.19

While many issues already mentioned arise in the context of all research, some issues are more specifically germane to population-based genetics studies as they may increase the potential risk. In addition to respecting the requirements set out by national, regional and international statutes, regulation and policy, the information provided to the prospective participant should generally contain a description of the project to be undertaken, its benefits and risks, its objectives, the permitted uses and access policies, as well as an explanation of the donor’s rights and the confidentiality and security mechanisms to be implemented. Thus, the information divulged notifies the subject of the investigation, and ensures formulation of his/her rights, prior to the providing of consent. Amongst others, examples of areas for which information should be provided to the prospective participant include: i) ensuring that the participant understands any community objections to the research; ii) the manner in which the consent will cover the intended research, especially for broad research protocols; iii) whether a right of withdrawal exists and its scope; iv) the possibility for the participant to submit additional information;20 v) the right to know or to not know about one’s own genetic data; vi) the possibility of updating information; vii) the possibility for the project to have access to other databases;21 viii) the right for the project team to contact relatives; and ix) measures to be undertaken to protect participants from stigmatisation and discrimination.

Given the complexity of obtaining “informed consent” in the context of population databases, it has been questioned whether the “traditional/ordinary” informed consent procedure is sufficient or whether another model of consent is appropriate/required. Some have argued that in light of the nature of HGRDs, the unforeseeable purposes for which they are to be used22 and their length (for some projects the samples and data are to be stored indefinitely), specific consent mechanisms may be inadequate and cumbersome. Since the participant providing biological samples and data to the database is aware of the overall general purpose and is aware that, by their nature, the data and samples will be used for diverse purposes, blanket consent may be viable. Otherwise the value of the database will be considerably reduced. In such a hypothesis, it has been argued that the core element in the informed consent process is to ensure that the participant understands the nature of database. Still others have advocated a new model/paradigm. For example, the establishment of an HGRD as a charitable trust has been advocated. This mechanism would govern not only the issue of consent but the entire relationship between investigator, participant, researcher and funding entity.23

On a practical level, most of the projects discussed here have adopted a policy of voluntary, written informed consent.24 For example, the Estonian Human Genes Research Act explicitly prohibits presumed or coerced consent.25 Section 9 of the Estonian Act prohibits influencing a person’s decision to participate, including by threatening the person with negative consequences, promising material benefits or providing subjective information. Moreover, most of the HGRDs discussed here have opted for broad consent, with forward-looking research mandates, and based on a process which provides a broad spectrum of information to participants.

Exceptionally, Iceland has opted to make presumed consent sufficient to utilise an individual’s samples and data in the HSD certain circumstances. By virtue of Article 7 of



the HSD Act, the Licensee would need to obtain the consent of the health institutions or self-employed health workers to obtain data derived from medical records for entry into the HSD. Pursuant to Article 8, a patient may request from the Director General of Public Health that their information not be entered in the HSD. As previously discussed, this approach has been criticised by diverse stakeholders and was challenged before the Icelandic Supreme Court.

The UK Biobank will invite by mail potential participants to attend a meeting at a local assessment centre during which staff will answer questions about the project and explain the consent process. The Biobank will provide information leaflets to potential participants which will cover a broad spectrum of information. The UK Biobank26 has chosen a broad, general approach to consent, whereby individuals consent to “participate in the UK Biobank” as explained in the consent form. The consent is premised on the participant understanding and accepting, amongst other things, the long term nature of the project, its risks and benefits, the need for a link between the data and samples provided by the participants and their full medical and other health relevant information, the fact that the UK Biobank is the legal owner of the database and the sample collection, that the participant will have no property rights in the samples, the type of safeguards that will be employed for security and confidentiality, the expectation that commercial entities will have access to the database, the intention to hold and to allow access to health information even after the participant’s death, the participant’s right to withdraw and the possibility that participants will be re-contacted. The re-contact would be to obtain consent for ethically and scientifically approved uses of samples and data not falling within the ambit of the original consent, to update the participant’s information or to invite participation in a new study involving the collection of new samples and information. The UK Biobank has abandoned its original plan of an “opt-out mechanism” whereby participants could choose research areas for which they refused to provide consent. Interestingly, the Latvian project has chosen to include in its Consent Form the opportunity for participants to indicate, if they so choose, the areas of genome investigation (e.g. cancer, diabetes, etc.). It is also worth noting that should a participant die or lose their mental capacity to consent or should it be impossible to re-contact the participant, the samples and data will remain in the UK Biobank’s database and may be used (possibly even against family wishes).

In Estonia, the donor consents for the EGP to carry out “scientific and applied gene and health research in order to find genes that influence the development of illnesses”. Furthermore, the consent form explicitly specifies that “[r]esearch carried out with the help of the Gene Bank shall not be limited to the present scientific level” and that the donor cannot provide conditional or partial consent.27 Such broad consent has also been chosen for the Marshfield Project, with ethics review boards determining whether or not new consent is required should data and samples be used for a different purpose.28 In the Estonian case, the participant is explicitly accepting, in the Consent Form, that their sample may have commercial value, that commercial entities may receive anonymous data and that the Estonian Genome Project Foundation possesses the right of ownership of the sample, description of the state of health, other personal and genealogical data.

In the CARTaGENE project the consent process includes a 2-hour appointment with a nurse who will explain the project and answer questions. In the course of this appointment, a pamphlet will be provided to the prospective participant and the topics discussed will include type of data and samples to be provided, type of information that will be collected and from which sources, risks, benefits, commercialisation, security and confidentiality, return of results and the right to withdraw. In the consent form the



participant accepts, among other things, the CARTaGENE team accessing in coded form health information from a number of government databases, the conservation of coded data and samples for 50 years and to their destruction at the end of this period, that the coded data and samples will be used for research approved by an ethics research committee in order to study the contribution of genetic diversity to health and disease, that the coded data and samples will be exchanged with other public population projects, that general results will be published and that they will not receive personal financial benefit from any commercialisation.29

Under the HapMap Consortium participants agree that their samples can be used for the “Haplotype Map Project and other research on genetic variation”.30 It is noteworthy that although participants cannot be re-contacted, as their samples and related data are anonymised, a Community Advisory Group established in each community will monitor subsequent use of samples. This Group could request that the community be withdrawn from the study and the samples destroyed. For any further use of samples for research, a Statement of Research Intent must also be approved by the Repository IRB, taking into consideration the interest of the community at stake and the investigator(s).31

Finally, mention should be made of an original approach to population database research suggested by the German Research Foundation.32 The Senate Commission on Genetic Research proposes that participants in large-scale population studies should be given the option of either specifically consenting to a specific research protocol or giving a “blanket” consent to the research team. The latter is understood not as informed consent but as “a free and independent decision taken in deliberate partial ignorance that is considered legitimate under clearly defined circumstances”(i.e. low-risk studies). This approach would allow the competent, informed individual to prospectively consent to further research without re-contact or re-consent. It presumes ethical oversight of all research.

4.3.3 Children’s and protected adults’ consent

The creation of population databases raises issues with respect to those members of a population who are not able to provide consent, such as children or protected adults. The issue of whether or not to allow children and/or protected adults to participate in genetic studies and contribute to HGRDs is complex.

One of the more challenging difficulties to including children in population databases is what constitutes “appropriate” consent for children. Often, researchers look to parents to provide consent in lieu of the child, especially for younger children. An issue that arises is the continued validity of the consent provided by the parent. In other words, it remains an open question whether the consent provided by a parent when a child is 3 years old remains valid when the child is 13 years old. An approach would be for the HGRD to determine that below a certain age (e.g. 10, 12, 14 years) the parents decide on the child’s behalf. However, once that age is attained, it would be mandatory for the research team to actively seek the child’s permission for further research, based on a number of elements.

Another issue with respect to research involving children is the question of who should have authority to consent on behalf of the child. Specifically, the consideration is whether consent is only valid if provided by both parents. If this approach is adopted, consideration would have to be given to situations in which obtaining the approval of both parents may be difficult. For example, such situations may arise where one of the child’s parents may be unknown, difficult to locate, or refuses to consent. Moreover,



difficulties for obtaining consent may also arise in situations where the decision-making authority over the child may be more complex. This type of complexity may arise, for example, if the child has been placed under the guardianship of a relative or has been placed in judicially-sanctioned foster care.

As for adults, but much more acute, is the issue of which information should be provided to children and the manner in which it is to be communicated. HGRD project teams will need to consider the best approach to communicate, in a simplified manner, technical information about genetics and the child’s situation, if relevant, the implications this may have for the child and to anticipate and dispel any concerns the child may have.

For protected adults, the consent issue may be even more complex. A situation where an adult consents to participate in a population database while having full capacity but becomes incapacitated over the course of that initiative, raises issues of responsibility and obligations for the project, the participant, and the incapacitated participant’s guardian or representative. In this regards, the UK Biobank policy is that if a participant loses mental capacity or dies, the Biobank will be guided by the most recent record of the participant’s consent. Family members will not be able to withdraw the consent of the incapacitated or deceased relative unless stipulated in the participant’s consent.

In order to address this quandary, the Latvian legislature, in enacting the Human Genome Research Law, included provisions on “Genome Research of a Person Recognised as not having Capacity to Act” which covers protected adults as well as children.33 Protected adults may only be included in the population database in exceptional circumstances, and if it provides direct benefit for the health of the participant and if the risk is commensurate with the gained benefit. Even in such a case, the written consent of the protected adult’s guardian is required. For children, the default premise is that the consent of the parents is required. However, legal authorities may have regard to the child’s wishes, based on the child’s age and maturity.

4.3.4 Renewed consent and re-contact

The issue of renewed consent raises both retrospective elements and prospective elements. The retrospective element may arise where it is being contemplated that an existing collection, possibly amassed without consent or without specific consent, will be employed for different purposes. The prospective element of renewed consent raises the question under what circumstances do participants need to re-consent. Both of these scenarios also raise issues pertaining to re-contact.

There are instances in which large epidemiological studies established in the past for non-genetic research are now being used for pursuing genetic questions.34 These data and biological samples are especially attractive when they are available for a period of ten to twenty years back, with some follow-up information or the possibility of mortality follow-up. These situations raise the question of whether or not renewed consent is required if samples and data have been collected in the past, especially for specific purposes. The old informed consent may often not fulfil today’s requirements. Typically, in such consent forms the research questions are restricted to specific diseases and often no consent was asked regarding the extraction and use of DNA. The issue therefore concerns the circumstances under which it is permissible to use these valuable resources for different purposes and what the implications are for participants’ consent.

One approach would be to re-contact the participants and asked for a new, timely and pertinent consent. However, this would be an expensive undertaking, with many drop-



outs due to death, relocation, or loss of interest. Technical loss due to follow-up is probably much more important than the fraction of former participants who would explicitly refuse the use of the old data and samples for new questions. On the other hand, for studies aiming to be representative of the population, such a selection bias would not be acceptable. Another approach would be to completely anonymise the data. This may solve the confidentiality issue but makes it impossible to perform follow-up research (e.g. to include the participants and/or their families).

A number of the population databases, including the UK Biobank, have incorporated the element of re-contact in their projects. Potential participants of the UK Biobank will be informed that the Biobank will re-contact them to collect new information, to seek consent for new proposed uses or to ask whether researchers may contact them to discuss involvement in new studies that may require new samples or data.

4.4 Right to withdraw

Having voluntarily entered a large-scale population HGRD, participants may at some point wish to withdraw from the study and have their samples, information and data destroyed. The issue arises of whether this is possible, and if so at what point(s) in time, in what manner and for which information, data and samples. Should participants have the right to withdraw the data and samples pertaining to them only, or for deceased or incapacitated relatives? In some cases, it may be possible for participants to withdraw their data and samples throughout the life of the HGRD. However, in other cases, depending on how the HGRD is established, the right to withdraw data and samples may only be available up to the point of anonymisation. Finally, various problems for withdrawing data and samples may arise. For example, if the participant’s data have been included in information provided to a third party, it may not be possible to destroy that information. What should be the scope of the right to withdraw?

The UK Biobank’s Ethics and Governance Framework states that participants will have the right to withdraw at any time without having to provide a justification and without any penalty. If participants become incapacitated or die, they would no longer be able to withdraw themselves. The Biobank will not enrol potential participants who express a view that they would want to exercise their right to withdraw should they lose their capacity or die as this would reduce the value of the resource for research. Nevertheless, participants may change their minds at any time after enrolment and their wishes will be honoured. If a participant loses their mental capacity or dies, the UK Biobank will be guided by the most recent record of the participant’s consent. Family members will not have the right to withdraw data and samples of incapacitated or deceased relatives, unless the participant’s consent expressed this wish.35 Moreover, as long as a participant has not actively withdrawn, the Biobank will continue to use the samples and data.

The UK Biobank contemplates three different options for withdrawal:

i) ”No further contact”: the UK Biobank would no longer contact the participant directly, but would still have their permission to use information, data and samples provided previously and to obtain further information from their health-relevant records.

ii) “No further access”: the UK Biobank would no longer contact the participant or obtain information from their health-relevant records in the future, but would still have their permission to use the information and samples provided previously.



iii) “No further use”: In addition to no longer contacting the participant or obtaining further information, UK Biobank would aim to destroy all of their information and samples collected previously (although the participant would be told that it may not be possible to trace all distributed sample remnants for destruction). Only their signed consent and withdrawal would be kept as a record of their wishes. Such a withdrawal would prevent information about them from contributing to further analyses, but it would not be feasible to remove their data from analyses that had already been done.

For the Latvian project, participants have the right to prohibit the supplementing, renewal and verification of their health description stored in the database. The participant also has the right to withdraw their consent at any time. In such a situation the tissue sample, the DNA description, the description of the state of health and the genealogical data will be destroyed.36

The Estonian Project has adopted a more assertive approach. According to Section 12 of the Human Genes Research Act, a participant has the right to apply for the destruction of data which enables decoding, the right to withdraw consent until the tissue sample or the description of their state of health is coded, as well as the right to prohibit the supplementing, renewal and verification of descriptions of their state of health. Thus, the legislation allows withdrawal up to the time when data and samples are coded. It furthermore permits participants to sue the foundation for compensation for damages and request the destruction of any information (samples, DNA and health description) stored upon unlawful breach of a participant’s right to confidentiality.37

Since the HapMap Consortium anonymises data, making individual withdrawal impossible (mentioned in the consent form), and since samples are labelled by population, only community disengagement is permissible. The Coriell repository is careful to respect communities’ wishes as shown by their extensive community engagement policy. Thus, it will act upon a request for withdrawal by destroying all the samples or returning them to the community. “However, the Repository staff will be unable to return samples that have already been sent out to investigators.”38

4.5 Results

In many epidemiological studies, information or results are provided to participants. However, given the breadth of an HGRD, the issue is whether or not this is feasible or desirable. First, the issue arises of whether or not results from users of the data and samples should be returned to the database. While providing results back to the database may enrich it, quality assurance of the results provided and included in the database is an important consideration. A second issue arises from the provision of results back to the participant. Again, given the breadth and purpose of HGRDs, the issue is whether it is feasible to contemplate a policy of providing results back to each participant and the value of doing so, especially outside a clinical context. If results are provided to the participant, what information should be given, under what circumstances, and in which conditions?

4.5.1 Results back to database

With population databases, one of the most useful and potentially promising aspects will be the long-term collection and updating of the information collected. The databases will become increasingly useful as more and more information is collected and stored and is employed to carry out research. One way to continually improve the wealth of



information contained in the database is to ensure that any information derived and results obtained from data and samples accessed from it are incorporated back into the database.

The UK Biobank project has developed a comprehensive approach to this issue of results back to the database. It has chosen a policy requiring research users to provide back the results from all tests made on participants or their samples. It has determined that, as a condition of access to protected samples, every user will be required to enter into an Access Agreement. This will specify the user and the specific purpose for which the data are to be used. It will include standard terms as to the ownership, exploitation and dissemination of results, including return of data to UK Biobank. In order to enrich as much as possible the value of the database, the UK Biobank has also developed a policy with respect to the dissemination of research results. All research findings, whether positive or negative, from all research using the resources of the UK Biobank will need to be placed in the public domain. The CARTaGENE initiative has also adopted a policy that general research results, negative or positive, should be published and publicly available in anonymous, general and statistical format.39

In order to accelerate the dissemination of information and data, the UK Biobank may publish the title, a lay summary and a scientific abstract of each piece of research for which access to the Biobank has been granted, shortly after such access is granted, prior to the identification and/or publication of any results. Suitable modifications may be permitted to scientific abstracts prior to publication to preserve future patent rights or protect commercially sensitive information. Moreover, users given access to protected material will be required to undertake to disseminate the results of their research as rapidly and widely as possible. Users will be encouraged to discuss their research with other scientists and the public and to share relevant data and samples as openly as possible. A limited and reasonable delay prior to dissemination will be permitted in order to enable a paper to be published, a patent to be filed or other competitive advantage to be pursued. UK Biobank may consider requests to delay release of results if circumstances merit (e.g. in the event of publication delays outside of the researcher’s control).40

In addition to dissemination, the users of protected material will be required to provide UK Biobank with a copy of all of the results of their research based on this material, including negative findings and supporting data, for incorporation into the Resource. Users who have had access to samples will be required to provide sufficient details of the assay techniques used so that other researchers can comprehend the results. The users will be required to copy the data to UK Biobank on terms that permit them to be used for research by users of the UK Biobank Resource without charge (other than standard UK Biobank access charges where applicable). This will be the case even if the data concerned are the subject matter of a patent. The users will be required to send data to UK Biobank within six months of deriving results suitable for publication or patenting, although UK Biobank will consider requests to delay release of results if circumstances merit it.41

4.5.2 Results back to participants

In most epidemiological studies, participants receive the results of their initial recruitment medical exam (measurements and general observation).42 However, there is also the issue of the results of genetic analysis, which are often not easy to interpret. First, with respect to genetic data, participants should have the right to chose to know or not to know. Second, data that are not clearly interpretable should not be given to participants or their physicians. Conundrums arise when there are relevant results from genetic testing



but no treatment is available. What approach should be adopted? Should the case be given to an ethics committee to decide what is appropriate on a case-by-case basis? Or would it be better, as a general rule, not to provide any genetic information to the participants? The different HGRD projects have adopted very different approaches to this issue. The ability to return results to participants is also linked to the manner in which the samples and data are collected and maintained, as discussed above. For example, if a database decides to adopt an approach of coded samples, it may be possible for the researcher both to obtain follow-up information or data and to communicate research results back to participants or their physicians.

If the genetic data remain linked or linkable to the identity of the individual, researchers may be ethically obliged to provide clinically-relevant information back to research subjects, especially if avoidable, serious harm may ensue if the information is withheld.43 Participants may or may not have agreed to be re-contacted, depending on the terms of the informed consent to the research project. Thus, privacy issues have implications for anonymisation and informed consent. For example, in the United States, informed consent to participate in research is linked to the identifiability of the participants because consent can be waived if the research involves no more than minimal risk to the subjects.44 However, risks due to inadvertent breaches of privacy or confidentiality in genetics research databases are in part related to the identifiability and security of the information collected.

Database information may also have implications for family members of the participant. In the United States, several court cases about clinical care (as opposed to research) have indicated a trend towards a professional standard according to which health-care practitioners should disclose genetic test information to family members when certain conditions are met. In the case of Safer v. Estate of Pack, the Superior Court of New Jersey ruled that a physician has a duty to warn those known to be at risk of avoidable harm from the existence of a genetically transmissible condition in the physician’s patient or identifiable third parties, if the physician knows of their existence. In this case, Donna Safer was diagnosed with multiple polyposis and cancerous blockage of the colon. Many years previously, Safer’s father had died of colon cancer at age 45, but the family was told that the children were at no increased risk of colon cancer.45 In a similar case, Pate v Threkel,46 the Florida Supreme Court ruled that the physician may have a duty to warn third parties of genetic risks, but that this duty would be satisfied by informing the patient of risks to the third parties. It is not clear, at the present time, how these cases would be applied to genetic information obtained in the research setting.

The Estonian Genome Project47 and the Latvian project48 differ from other HGRD projects in that participants enjoy an explicit right to know or not to know about their personal genetic data. They may request access to their decoded data and are entitled to genetic counselling. Furthermore, their physicians can, with the knowledge and consent of the participant (except in cases of emergency), access such data for treatment purposes.

Exceptionally, a few HGRDs have decided to provide some basic test results back to participants. For example, CARTaGENE will offer participants the possibility of obtaining blood sugar level, cholesterol level, blood pressure and anthropomorphic measurements with an appropriate explanation and to provide these results to the participant’s personal doctor.49

Conversely, most projects have decided to not provide participants with individual research results. The absence of personal feedback can be explained either by the anonymity of the samples and data kept in the HGRD (HapMap, P3G) or by the nature of



the projects (research infrastructure and not health-related programmes) (Personalised Medicine Research Project). The Marshfield Project objects to the inclusion of DNA analysis results in patients’ medical records for confidentiality reasons but also because these results are considered “often preliminary, inconclusive, and not necessarily valid for decisions concerning patient care and treatment”.50

The UK Biobank51 has demonstrated a particular reluctance to provide any results to participants. Insisting on the research nature of their endeavour and the contemplated absence of interpretation, counselling or support to be provided by researchers, they conclude that data communicated outside of a medical setting would not be constructive and might even be harmful to the participant. At the initial assessment visit, participants will be provided with a printed report of their baseline measurements (e.g. blood pressure, height, weight, estimated body fat). However, the Biobank clearly states in its policy that the staff conducting the enrolment will not have the same duty of care they would have in a clinical setting; rather the legal duty of care would be that applicable in the research context. Any results from routine investigations carried out before the samples are stored and from later research studies will not be provided to participants.

4.6 Education and training of data collectors and researchers

The training of health-care professionals and researchers will be important for the success of an HGRD. Health-care professionals responsible for the recruitment of participants and the collection of data (e.g. questionnaires, medical examination, drawing of blood, transfer of the information collected) may not be familiar with genomic research so that a training policy may be required. The creation of a training policy raises issues such as, should such training be mandatory and what should be the nature of such training.

The field of genomic research which relies on advanced technology and bioinformatics is in constant evolution. Consequently, ongoing education and training, exchanges among researchers, as well as the organisation of and attendance at workshops and scientific meetings are essential. Some of the population database projects have included in their framework an educational component for the public (examined in Chapter 3), researchers or experts and data collectors. However, much more detailed planning of the manner in which the data collectors and researchers will be educated and trained will need to be developed by the population databases.

4.6.1 Education and training of data collectors

The training of data collectors, whether nurses or general practitioners, needs to be considered within the scope of the population database project, especially because those responsible for recruitment of participants and collection of data may not be familiar with genomic research. The Estonian Genome Project, the CARTaGENE and the UK Biobank all contemplate such situations. The Estonian Genome Project specifies that “[a]ll data collectors [will] have passed the data collector training course organised by the Estonian Genome Project Foundation”.52 The UK team will establish a protocol explaining to general practitioners their role, the information they can divulge to the subjects, the measures to be undertaken for security purposes and guiding them in the appropriate handling of certain contentious situations.53



4.6.2 Education and training of researchers

The genomic and genetic research undertaken by the various population database projects constitutes a unique opportunity for researchers, ranging from students to senior scientists, to be trained in this changing field and to become acquainted with the latest technologies.

The partners of the P3G consortium share an interest in the “exchanges [among] experts and young researchers in many human and social scientific disciplines in addition to biology, public health and genetics”.54 GenomEUtwin, for instance, aims to establish a research network in order to further collaboration within genome research groups. This network will be responsible for the development of training programmes in molecular epidemiology for pre- and post-doctoral researchers (courses, inter-institute exchange, workshops, etc.) and advanced clinicians, scientists (specialised training to enhance ongoing research, workshops and meetings of participants).55 Similarly, the COGENE project, by supporting genomic research, had wished to “stimulate progress in functional genomics”,56 promote dialogue among researchers, and organise scientific meetings and training courses.57

In the TgRIAD Project being developed within the Howard University School of Medicine, the training of fellows, residents and medical and basic science students is foreseen as one of the key components. They will have the opportunity to participate in and develop expertise in medical technologies in the future.58 In Estonia, the EGP wants to improve Estonia’s international competitiveness through educational support and the establishment of infrastructure.59



Notes

1. Australian Law Reform Commission, Essentially Yours – The Protection of Human Genetic Information in Australia, Report 96, 2003, Chapter 3, see www.austlii.edu.au/au/other/alrc/publications/reports/96/ (accessed 15 September 2006).

2. G. Annas, L. Glantz and P. Roche (1995), “Drafting the Genetic Privacy Act: Science, Policy and Practical Considerations”, Journal of Law, Medicine and Ethics, 23:360-6.

3. T. Murray (1997), “Genetic Exceptionalism and ‘Future Diaries’: Is Genetic Information Different From Other Medical Information?” p. 60, in M. Rothstein (ed.), Genetic Secrets: Protecting Privacy and Confidentiality in the Genetic Era, Yale University Press, New Haven, Connecticut.

4. Author’s discussion with biobank representative, August 2006. See also www.p3gobservatory.org/documentDetail.do?methodToCall=executeGet&studyId=100011&documentId=1 (accessed 15 September 2006).

5. J. A. Bovenberg (2006), Establishment, Management and Governance of Human Genetic Research Databases, Study commissioned by OECD, pp. 27-29, and additional information.

6. Mildred Cho, “Confidentiality and Data Management”, Workshop Background Paper, 2003.

7. National Bioethics Advisory Commission (1999), Human Biological Materials: Ethical Issues and Policy Guidance, Vol. 1, Report and Recommendations of the National Bioethics Advisory Commission, Rockville, Maryland, August, see www.georgetown.edu/research/nrcbl/nbac/hbm.pdf (accessed 15 September 2006).

8. United States, Health Insurance Portability and Accountability Act, 45 C.F.R. 164.514(b).

9. Mildred Cho, “Confidentiality and Data Management”, Workshop Background Paper, 2003.

10. CARTaGENE, see www.cartagene.qc.ca/focus.cfm (accessed 15 September 2006).

11. Source : J. A. Bovenberg (2006), Establishment, Management and Governance of Human Genetic Research Databases, Study commissioned by OECD, pp. 37, and additional information.

12. deCODE press release, joint statement by deCODE and the Icelandic Medical Association, 27 August 2001.

13. See e.g. G. Annas, L. Glantz and P. Roche (1995), Drafting the Genetic Privacy Act: Science policy and practical considerations”, Journal of Law, Medicine and Ethics 23:360-6. G.J. Annas (2000), “Rules for Research on Human Genetic Variation – Lessons from Iceland”, New England Journal of Medicine, Vol. 342, No. 24, p. 1830.

14. National Bioethics Advisory Commission (1999), Human Biological Materials: Ethical Issues and Policy Guidance, Vol. 1, Report and Recommendations of the National Bioethics Advisory Commission, Rockville, Maryland, August, see www.georgetown.edu/research/nrcbl/nbac/hbm.pdf (accessed 15 September 2006).



15. Faden, R.R. and T.L. Beauchamp (1986), A History and Theory of Informed Consent, New York: Oxford University Press, cited in National Bioethics Advisory Commission (1999), Human Biological Materials: Ethical Issues and Policy Guidance, Vol. 1, Report and Recommendations of the National Bioethics Advisory Commission, Rockville, Maryland, August, see www.georgetown.edu/research/nrcbl/nbac/hbm.pdf (accessed 15 September 2006).

16. See e.g. World Medical Association Declaration of Helsinki, Washington, 2002, www.wma.net/e/policy/b3.htm (accessed 18 May 2006); Council for International Organisations of Medical Sciences (CIOMS) (2002), “International Ethical Guidelines for Biomedical Research Involving Human Subjects, Geneva, November, www.cioms.ch/frame_guidelines_nov_2002.htm (accessed 18 May 2006); The Nuremberg Code (1949), Trials of War Criminals before the Nuremberg Military Tribunals under Control Council Law No. 10, Vol. 2, pp. 181-182, US Government Printing Office, Washington, DC; Conseil de l’Europe (CE), Convention pour la protection des droits de l’homme et de la dignité de l’être humain à l’égard des applications de la biologie et de la médecine: Convention sur les droits de l’homme et la biomédecine, Oviedo, 4 April 1997, http://conventions.coe.int/treaty/FR/Treaties/Html/164.htm (accessed 18 May 2006).

17. H. Greely (1999), “Breaking the stalemate: a prospective regulatory framework for unforeseen research uses of human tissue samples and health information”, Wake Forest Law Review, 34:737-66.

18. J.F. Merz, P. Sankar, S.E Taube and V. Livolsi (1997), “Use of human tissues in research: clarifying clinician and researcher roles and information flows”, Journal of Investigative Medicine, Vol. 45, No. 5, p. 252-257.

19. G. Annas, L. Glantz and P. Roche (1995), Drafting the Genetic Privacy Act: Science policy and practical considerations”, Journal of Law, Medicine and Ethics 23:360-6. G.J. Annas (2000), “Rules for Research on Human Genetic Variation – Lessons from Iceland”, New England Journal of Medicine, Vol. 342, No. 24, p. 1830.

20. Estonia, Gene Donor Consent Form, www.geenivaramu.ee/index.php?lang=eng&sub=74 (accessed 18 May 2006).

21. See deCODE Genetics, “An Informed Consent for Participation in a Genetic Study of [name of disease]”, www.decode.com/files/file148517.pdf (accessed 18 May 2006). UK Biobank, Ethics and Governance Framework, see www.ukbiobank.ac.uk/ethics/ethicsgov.php (accessed 18 May 2006). C. Laberge et al. (2001) “Formal Application to Genome Quebec”, Newsletter – Map of Genetic Variation in the Quebec Population, October 15, Vol. 1, p. 3.

22. “In the case of biobanks, … the specific uses generally cannot be determined in advance. That is the essence of a biobank. Samples and other information are often aggregated, analysed, and offered to pharmaceutical companies and other researchers.” [footnote omitted] M.A. Rothstein (2002), “The Role of IRBs in Research Involving Commercial Biobanks”, Journal of Law, Medicine and Ethics, Vol. 30, No. 1, p. 105.

23. H. Greely (1999), “Breaking the stalemate: a prospective regulatory framework for unforeseen research uses of human tissue samples and health information”, Wake Forest Law Review, 34:737-66.

24. As an example, a participant in Latvia will be asked to sign two consent forms (the forms are also to be signed by the main processor), which will then be transferred to the State �� “Data Protection in the Project “Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc). (accessed 9 May 2006).



25. Estonia, Human Genes Research Act, Chapter 2, §9.


27. Estonia, Human Genes Research Act, Chapter 2 § 12 ; EGP “Gene Donor Consent Form”, www.geenivaramu.ee/index.php?lang=eng&sub=74 (accessed 18 May 2006).

28. See www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pers_med_res_prj (accessed 18 May 2006).

29. Draft General Information Pamphlet, August 2006, see www.cartagene.qc.ca/docs/DocIntro_En.pdf (accessed 15 September 2006). Draft Information and Consent Form, August 2006, www.cartagene.qc.ca/docs/InfoConsent_En.pdf (accessed 15 September 2006).

30. Recruitment and Informed Consent Forms of the Haplotype Map Project (HapMap) and Other Research on Genetic Variations, cited in S.Y. Song, Y.M. Koo and D.R.J. Macer (eds.) (2003), Bioethics in Asia in the 21st Century, Chapter 2.5, pp. 31-45, Eubios Ethics Institute; see also www.sanger.ac.uk/HGP/Chr6/MHC/ (accessed 16 May 2006); “Background on Ethical and Sampling Issues Raised by the International HapMap Project”, NHGRI News Release, October 2002, www.genome.gov/10005337 (accessed 19 May 2006).

31. NIGMS Human Genetic Cell Repository, Policy for the Responsible Collection , Storage, and Research Use of Samples from Identified Populations, 25 August 2004 (last revision) http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html (accessed 18 May 2006). See also “Background on Ethical and Sampling Issues Raised by the International HapMap Project” Release Genetic Variation Mapping Launch, NIH News Advisory, October 2002, http://genome.gov/10005337 (accessed 18 May 2006); E. Suda and D. Macer (2003), “Ethical Challenges of Conducting the HapMap Genetics Project in Japan” in S.Y. Song, Y.M. Koo and D.R.J. Macer (eds.), Bioethics in Asia in the 21st Century, Eubios Ethics Institute.

32. See Deutsche Forschungsgemeinschaft, Senate Commission on Genetic Research, Predictive Genetic Diagnosis – Scientific Background, Practical and Social Implementation, Bonn, 27 March 2003, www.dfg.de/aktuelles_presse/reden_stellungnahmen/2003/download/predictive_genetic_diagnosis.pdf (accessed 18 May 2006).

33. Latvia, Human Genome Research Law, (2002), Section 12.

34. H. Erich Wichmann, “Comment on Access to Mutation Databases for Research Purposes”, Background Paper, 2003.

35. UK Biobank, Ethics and Governance Framework, www.ukbiobank.ac.uk/ethics/ethicsgov.php (accessed 18 May 2006), B. Understandings and Consent, 6. Right to Withdraw and 7. Respect for Incapacitated Participant’s Wishes.

36. See e.g�� “Data Protection in the Project “Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc (accessed 18 May 2006). Latvia, Human Genome Research Law, (2002), Section 11.

37. Estonia, Human Genes Research Act, Chapter 2 §10 and Chapter 3 §21; EGP “Gene Donor Consent Form”, www.geenivaramu.ee/index.php?lang=eng&sub=74 (accessed 18 May 2006).



38. NIGMS Human Genetic Cell Repository, Policy for the Responsible Collection, Storage, and Research Use of Samples from Identified Populations, 25 August 2004 (last revision) http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html (accessed 18 May 2006).

39. CARTaGENE, Information and Consent Form, August 2006, www.cartagene.qc.ca/docs/InfoConsent_En.pdf (accessed 15 September 2006).

40. United Kingdom Biobank, Policy on Intellectual Property and Access, Section 9.

41. United Kingdom Biobank, Policy on Intellectual Property and Access, Section 9.

42. H. Erich Wichmann, “Comment on Access to Mutation Databases for Research Purposes”, Background Paper, 2003.

43. Mildred Cho, “Confidentiality and Data Management”, Background Paper, 2003.

44. US Dept. of Health and Human Services. Federal Guidelines for Research Involving Human Subjects. Code of Federal Regulations 1991(45 CFR 46).

45. Safer v Estate of Pack. 677 A2d 1188. (1996); M. Severin (1999), “Genetic susceptibility for specific cancers”, Cancer, Vol. 86, pp. 2564-69.

46. Pate v Threlkel, 640 So2d 183. (1994).

47. Estonia, Human Genes Research Act § 12, 16 and 24, EGP “Gene Donor Consent Form”, www.geenivaramu.ee/index.php?lang=eng&sub=74 (accessed 18 May 2006).

48. �� “Data Protection in the Project “Genome Database of the Latvian Population”, http://bmc.biomed.lu.lv/gene/print/Latvian%20Genome%20Project-raksts%20Judith%20Sandor.doc (accessed 18 May 2006).


50. Personalized Medicine Research Project, Is There a Risk to My Confidentiality, www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pers_med_res_prj (accessed 11 May 2006).

51. UK Biobank, Ethics and Governance Framework, www.ukbiobank.ac.uk/ethics/ethicsgov.php, p. 11 (accessed 18 May 2006).

52. Estonian Genome Project, www.geenivaramu.ee/mp3/trykisENG.pdf, p. 7 (accessed 18 May 2006).


54. See www.p3gconsortium.org/org.cfm (accessed 15 September 2006).

55. GenomEUtwin, “Application for the EC Quality of Life Programme”, Area 8,5: Integrated Projects in Functional Genomics Relating to Human Health Genome-wide Analyses of European Twin Cohorts to Identify Genes behind Common Diseases”.

56. The European Commission “A New Initiative on “Genome Research for Human Health” Press Release, Brussels, 15 November 2000, http://europa.eu.int/comm/research/press/2000/pr1511en.html (accessed 18 May 2006).


58. See www.genomecenter.howard.edu/tgriad_component.htm (accessed 15 September 2006).

59. See www.geenivaramu.ee (accessed 18 May 2006).

CHAPTER 5. DATABASE MANAGEMENT AND GOVERNANCE – 105


Chapter 5

Database Management and Governance1

5.1 Management and governance of databases

The management of databases connotes the oversight of activities. The governance of databases involves numerous operational, technical and legal issues, including consideration of applicable legislation and regulation, the role of ethics and oversight committees, issues of powers, compliance and enforcement granted to an HGRD, the security features of the database, issues pertaining to access to the database and the demise of an HGRD.

5.1.1 Legislation and regulation of HGRDs

When considering the governance of HGRDs, legislation may be taken to be the codification of ethics and governance by the state. As such, it is interesting to note the different approaches that have been adopted internationally. Some projects, such as those in Iceland, Latvia and Estonia, are products of legislation, created by acts of parliament and regulated by statute. Other projects are created independently of parliament, such as the UK Biobank and the CARTaGENE initiatives, and are subject to existing legislation. While the creation of an HGRD as a scientific endeavour may permit more flexibility, its establishment through legislation may facilitate the application of enforcement procedures and measures.

When the project is a creation of statute, its structure, operation and regulation will be, to varying degrees, laid out therein. For example, the Icelandic HSD is subject to the Act on a Health Sector Database (Act Number 139/1998) and the Act on BioBanks (Act Number 110/2000) of the Althingi, the Icelandic Parliament, which include provisions in the following areas: Establishment and Operation of Biobanks; Collection, Handling and Access to Biological Samples; Access to the Database; Monitoring and Obligation to Supply information; and Penalties.

This approach is reflected in the Estonian Human Genes Research Act, which contains similar provisions:1

� The Rights of Gene Donors (confidentiality, voluntary nature of participation, consent and other rights).

This chapter also draws on the texts prepared for the Tokyo Workshop: Vilhjalmur Arnason, Jasper

Bovenberg, Ruth Chadwick, Mildred Cho, Richard Cotton, Ryuichi Ida, William Lowrance, Nicole Questiaux and David Weisbrot. See Bibliography.

106 – CHAPTER 5. DATABASE MANAGEMENT AND GOVERNANCE


� Processing of Gene Bank (collection and storage of tissues and samples, ownership, access to the tissues, samples and gene bank).

� Data Protection (coding, decoding, access).

� Prohibition of Discrimination (right, employment context, insurance context).

� Supervision and Resolution of Appeals.

� Criminal Code amendment to enact the sanctions provided within the Act.

In the case of UK Biobank, Marshfield, TgRIAD, GenomEUtwin and the CARTaGENE projects, these issues have for the most part been the responsibility of the project initiators. In the United Kingdom, an interim advisory group was appointed by the “funders” to draw up the Ethics and Governance Framework (EGF) for the Biobank. Similarly, in Canada, the organisation that first conceived the project, The Quebec Network for Applied Genetic Medicine (RMGA), has published two Statements of Principles, the Statement of Principles Human Genome Research (“2000 Statement”) and the Statement of Principles on the Ethical Conduct of Human Genetic Research Involving Populations (“Population Statement”).2 To some extent, these documents outline the applicable principles and the associated recommendations to be applied. However, in some areas, such as penalties, policies will need to be established. It should be recalled that in each case, the population database is subject to the applicable national and local legislation.

There are interesting questions relating to the legitimacy of self-regulation that do not necessarily arise within a legislative context and may be related to the issue of trust. Although all of the population database projects discussed have some form of ethics and/or management committees, do those that are more firmly grounded in legislation have a more rigorous framework of regulation and therefore greater powers of enforcement? Do those created independently of legislation benefit from an inherent flexibility? Another important issue is whether existing mechanisms and statutes, which came into force prior to the conception of population databases, are equipped to deal with highly sensitive issues associated with these projects. The organisational structure of a project will have some effect on regulatory issues. For example in the United Kingdom, the Board of Directors is responsible for the direction, management and control of UK Biobank Ltd., and is accountable under both company and charity laws and regulations.

Canada, which has no federal or provincial legislation enacted specifically for the creation and regulation of the CARTaGENE project, provides an interesting example of the application of existing legislation to HGRDs. In this case, the Privacy and Access Commission of Quebec (CAIQ) will need to grant approval for access to and use of the provincial Health Insurance Board’s database as the method of recruitment.3 This demonstrates how the choice of structure and recruitment may alter the role of law and regulation. In this case, express permission is required for the use of existing records.

One area in which all of the projects encounter legislation is data protection and data security. In Iceland, this is dealt with in the specific legislation. Similarly in Estonia, the Human Genes Research Act includes provisions relating to access to the genetic database information. Existing data protection laws, in the form of the Personal Data Protection Act and the Databases Act, also apply. In the United Kingdom, access to information held in databases is governed by the Data Protection Act 1998, given the absence of specific enabling legislation applicable to the UK Biobank. Academic



comment has suggested that “the present UK law relevant to genetic biobanks could really be reduced to a set of issues arising from a breach of confidence”.4

5.1.2 Role of ethics and oversight committees

A review of the HGRD projects reveals that they generally require some form of oversight committee, but that the composition and formation varies. In most of the initiatives, these bodies are empowered to ensure that the activities of the HGRD are in compliance with the ethical and regulatory framework and in the interest of the participants. The role, function and nature of the oversight committee should be given consideration in the governance of the HGRD. For example, should its composition be multidisciplinary and for how long should individuals be appointed? Which issues should be submitted to the oversight committee for consideration? The Ethics Committee of the Estonian Genomics Database, for example, ensures compliance with ethical guidelines, and anyone with a project seeking access to the database may address it.

In its Ethics and Governance Framework,5 the UK Biobank has elaborated a detailed structure for its governance framework. The Ethics and Governance Council was established by the Wellcome Trust and the Medical Research Council as an entity independent of these organisations and independent of the UK Biobank. The Ethics and Governance Framework work will be supplemented by ethics committees, who will review requests for research access.

The remit of the Council includes: acting as an independent guardian of the Ethics and Governance Framework and advising the Board on its revision; monitoring and reporting publicly on the conformity of the UK Biobank project with this Framework; and advising more generally on the interests of participants and the general public in relation to UK Biobank. In pursuing its remit, the Council will engage with, and render accounts to, a number of internal and external audiences. Internal dialogues will be with the Board of Directors, the CEO/Principal Investigator and the funders. External dialogues could be with participants, regulatory or government bodies, other interested parties, and the general public. The Council will not speak “on behalf of” UK Biobank, as this will be the responsibility of the Board; instead it will speak “about” UK Biobank.

In order to be able to fulfil its remit, the Ethics and Governance Council will need to be appropriately knowledgeable about UK Biobank’s continuing activities. However, the Council will be able to require from parties involved in the UK Biobank whatever information and discussion are necessary to fulfil its remit. Normally, the Council will communicate its reflections and criticism informally. If the Council is not satisfied with UK Biobank’s response, it could make a formal statement of concern (e.g. to the Board or funders) or, if necessary, make a public statement that certain actions should or should not be taken. In the extreme, members of the Council could resign in protest and announce this publicly. It is intended that the Ethics and Governance Council will work in an open and transparent fashion and report to participants and the public. This may be achieved in a variety of ways, such as through publishing reports of its reviews or discussions, occasionally meeting in public, or holding meetings with the public.

In Iceland, the creation and operation of the HSD is, amongst others, subject to the oversight of a Monitoring Committee, an Interdisciplinary Ethics Committee and the Data Protection Commission. The Monitoring Committee shall be composed of three members, for a term of four years, who will be responsible to supervise the creation and operation of the Health Sector Database. One member shall be a health sector worker with knowledge in the field of epidemiology, while another shall be knowledgeable in the



field of information and/or computer science. The third shall be a lawyer and serve as Chairman of the Committee. Alternate members shall be appointed in the same way.

The HSD Act6 gives the Monitoring Committee quite extensive powers including overseeing the development of agreements between the Licensee and health institutions, monitoring the day-to-day operations of the database and ensuring compliance with legislation, regulation and the Operating Licence, advising the Ministry of Health and Social Security on utilisation of the database, reporting to the Science Ethics Committee about the queries submitted to the database, and reporting on a yearly basis to the Minister on the operation of the HSD and on the work of the Committee. The Monitoring Committee, along with the Data Protection Commission, is responsible for monitoring that the conditions established in the Operating Licence are fulfilled.

The Interdisciplinary Ethics Committee7 shall also be composed of three members appointed by the Ministry of Health and Social Security, who have expert knowledge in the field of health sciences, research ethics and human rights. The Interdisciplinary Ethics Committee ensures that the processing of data in the HSD is in compliance with “recognised international rules on science and ethics and rules established on the basis of such international rules and current in Iceland at any time”.8 The Interdisciplinary Ethics Committee must approve all requests for research and searching of the database. Additionally, it is specified in the legislation that the HSD will be monitored by the existing oversight framework: “The Directorate of Health is responsible for monitoring of the Licensee’s observance of the provisions of legislation and regulations regarding health in general and the security of patients and the public”.9 The Data Protection Commission is to establish the “Technology, Security and Organisation Terms” that the Licensee is to respect, is to operate an Encryption Agency which is responsible for the encryption of personal identifiers and shall carry out the transfer of all data to the HSD. The Science Ethics Committee, which evaluates multinational and other scientific research, also has a general oversight role.

The oversight bodies for the Estonian Genome Project are also specified in the legislation, which requires the formation of an Ethics Committee and gives the Estonian data protection supervision authority powers of supervision over data collection, coding, decoding and processing. The legislation stipulates that members of the Ethics Committee must be recognised experts in their field, with the necessary expertise to perform the duties required, be Estonian citizens and resident in Estonia. The Ethics Committee will provide non-binding advice on the processing procedures of the Gene Bank pursuant to generally recognised ethical rules and international conventions; draw the attention of the Chief Processor to breaches of ethical norms; present an annual report to the Supervisory board; and answer any requests for information, advice or assessment.10 The Estonian project also benefits from a Scientific Advisory Board, whose aim is to counsel the Supervisory Board and the Management Board in questions dealing with scientific aspects. In case of necessity, the Scientific Advisory Board provides advice on the scientific validity of the scientific researches carried out on the basis of the Gene Bank data.11

While the CARTaGENE initiative is still awaiting certain approvals when these are received, the Institute for Populations, Ethics and Governance12 (IPEG) will be established as an independent structure under the governance of the Ministry of Health and Social Services. The Health and Welfare Commissioner will be responsible for its administration. The IPEG’s objectives will be to ensure that rules of ethics and of responsible governance are respected in large-scale studies of population genetics and in



the set-up and use of data and biobanks of human biological material. The nine members of the Board of Directors will be independent and have an interest or an expertise in ethics as well as in the governance of cohort studies using human biological samples and in the conservation of these samples for future use in research. IPEG will be authorised to invite international experts to advise on particular issues, should this be required. Also, international experts may be invited to hold a more permanent position within IPEG’s governance structure.

The objectives of the IPEG will be: i) to develop rules of ethics and responsible governance regarding the collection and use of public or private biobanks; ii) to ensure respect for the rules of ethics and governance, especially those regarding cohort studies where human biological material is collected; iii) to periodically publish reports about the rules of ethics and governance; iv) to organise and hold information or consultation sessions that are related to the mission of IPEG; v) to provide opinions and advice to all interested parties about respect of the rules of ethics and governance; vi) to ensure monitoring of approved cohort studies in which human biological samples are collected; and vii) to ensure, along with the authorised institutions, the responsible governance of orphaned biobanks so as to allow their optimal utilisation in the interest of society-at-large.

5.1.3 Management of HGRDs

With the exception of the UK Biobank, few details are provided by the other HGRDs on the manner in which the population database will be managed. It will be analysed as an illustration of the different components of a management structure.

The core UK Biobank resource will be managed by UK Biobank Limited, which is a charitable company limited by guarantee. UK Biobank’s Board of Directors, which is accountable to the Members of the company (Medical Research Council and Wellcome Trust), will act as charity trustees under UK charity law and company directors under UK company law, and exercise management oversight of the UK Biobank. The UK Biobank will serve as the legal custodian of the data and samples, and will operate through a Coordinating Centre hosted by the University of Manchester. The Coordinating Centre will allocate funds through a collaborative agreement to six Regional Collaborating Centres representing more than 20 UK universities, which will be involved through a Steering Committee in the scientific design and management of the collection. As discussed above, an independent Ethics and Governance Council will advise the Board and Funders, and publish public reports on the conformance of the UK Biobank with this Ethics and Governance Framework and with the interests of participants and the public.

The Board of Directors of UK Biobank are accountable to the Members of the Company (Medical Research Council and Wellcome Trust), and to the Charity Commission for England and Wales, for the performance of their duties as directors and charity trustees, including the duty to act in the interests of the UK Biobank. Up to five Board members, including the Chair, will be jointly appointed by the Members of the Company. A further five members may be individually nominated by the Department of Health, the Medical Research Council, The Wellcome Trust, the Scottish Executive and the University of Manchester (which hosts the coordinating Centre). The Board will include persons with relevant scientific knowledge or other relevant expertise, selected for their ability to serve as directors and charity trustees of UK Biobank. The Board will adopt this Ethics and Governance Framework and be responsible for making sure that all UK Biobank policies and activities conform to it. The Board will also be responsible



for matters of corporate governance, including the management of conflicts of interest within the UK Biobank. Potential conflicts of interest among members of the Board and the Chief Executive Officer/ Principal Investigator (CEO/PI) will be recorded. The Board retains overall responsibility for the direction, management and control of the UK Biobank, but it delegates day-to-day management to the CEO/PI.

The Steering Committee is a committee of the Board chaired by the CEO/PI. Membership includes the lead investigator from each Regional Collaborating Centre, with UK Biobank’s Executive Director and Chief Scientific Officer as observers. Amongst other things, the Steering Committee is responsible for advising the CEO/PI on the development of the scientific protocol, and on the direction and scientific objectives of the UK Biobank, in accordance with the Steering Committee’s terms of reference.

The Steering Committee, Board and funders will receive independent scientific guidance from an International Scientific Advisory Board. This Advisory Board will meet annually to review the UK Biobank’s progress and future plans in accordance with its terms of reference.

5.1.4 Powers, compliance and enforcement

An important set of issues for HGRDs involves the powers and ability to ensure compliance and enforcement of any decision. Such powers, or the lack thereof, have implications for ensuring that privacy and security policies are respected, and for ensuring that a private entity to which commercialisation rights are granted respects these rights and the resources and does not act in an abusive manner.

Although the UK Government had suggested in its White Paper on Genetics that an independent body overseeing research should have the power of veto, this has not been incorporated in the UK Biobank Ethics and Governance Framework (EGF).13 The process by which the Ethics and Governance Council may communicate its criticisms is set out in the EGF. The options range from informal communication, to a formal statement of concern to “in the extreme, members of the Council could resign in protest and announce this publicly”.14 Although this is not covered in the EGF, recourse to criminal or civil legal proceedings may be available in the event of a breach of privacy or security; this would depend on the nature and circumstances of the breach and the nature of the applicable legislation. The weakness of the enforcement procedure and the applicable penalties is one factor that favours the enactment of legislation specific for the establishment of a population database.

The HSD Act provides for a number of penalties15 in case of breach of its provisions and annex legislation and regulation. As the Icelandic database would be operated by licence, the power to revoke that licence is one means of effecting compliance. If any of the provisions of the operating licence or the Acts were violated, the Minister may issue a written warning and set a deadline by which action must be taken. Inaction or intentional and gross negligence may result in revocation of the licence. Any disputes may be taken to arbitration or settled in the Icelandic courts.16 Furthermore, the HSD Act stipulates that violation of the legislation may lead fines or imprisonment for up to three years, unless stricter penalties apply under other legislation. Article 17 of the HSD Act stipulates that compensation is payable to anyone who suffers financial loss as a direct result of their data being made known.17

In the Estonian initiative, the responsibility for meeting the project’s objectives rests with the chief processor, a non-profit foundation within the Ministry of Social Affairs. In



the Estonian case, noteworthy are the detailed penalties provisions within the Act, which also contains sections specifying amendments to be made to the existing Criminal Code so as to take account of the project and the rights of participants. The amended Criminal Code would impose fines and terms of imprisonment in case of breach. It also states that complaints concerning discrimination in employment or insurance due to genetic risks will be dealt with by the Labour Inspectorate or the Insurance Supervisory Authority, respectively.

5.2 Security of databases

The collection of large number of data points about each individual implies that even if frank identifiers such as names, addresses, social security numbers or patient identification numbers are removed from datasets, the remaining data may uniquely identify an individual. If the data in a dataset are linkable to other datasets that do contain identifiers, a unique identification, such as a name, may be established. Although population databases do not specifically include patient names or addresses, there are so many data fields with personal information that individuals may be uniquely identified through the database and potentially linked to a publicly available database that includes names.

Given the potential for misuse of data and samples collected in HGRDs, security is a primary consideration. This is both a legal and a technical question. Given the objective of the database, the issue of what are the best methods for ensuring security, ensuring that access occurs only in the permitted manner, and ensuring that access to the data and samples is not unduly restricted are key considerations. One method is through custody of a code registry. Other approaches to ensuring security and protecting privacy are to limit the amount or type of data released or to limit accessibility to researchers using the database. Combinations of legislation and technical solutions are being formulated to this end. A common method of enhancing data privacy is encryption, either one-way or public-key. This method can be used in conjunction with the other methods.

5.2.1 Custody of code registry

In situations where links are maintained between the data and participants’ personal identifying information, probably one of the most valuable systems for protecting the confidentiality of data is to provide the custody of code registry through which the data are kept retrievably anonymised.18 The custodian of registry specifically designated for that purpose should manage the registry, which would be kept confidential. The custodian must be a person who is under the duty of non-divulgation of confidential information, such as medical doctor or pharmacologist, depending on the jurisdiction. The divulgation of personal information should be subject to legal sanctions. In using computers for anonymisation and for holding the registry, it has been advocated that stand-alone computers for handling the personal identifiers and other personal information including health data be employed so as to reduce the risk that network computers may be violated.19

5.2.2 Limited data release

Another approach to protecting privacy is to limit the amount or type of data released or accessible to researchers using the database. This would involve a combination of legislation and technical solutions. In the United States, for example, privacy regulations authorised under the Health Insurance Portability and Accountability Act (HIPAA) has



defined a “limited data set”, which stipulates a set of specific types of identifiers (such as addresses and medical record numbers) that must be removed from data.

Another solution is based on the concept of bin size. The US Social Security Administration (SSA) and other government agencies use the bin size concept to ensure anonymity of data in public-use files. Although the SSA files contain no geographic codes, they do contain other data fields that potentially uniquely identify an individual. Thus, the SSA uses the general rule that any combination of characteristics in a dataset should not yield a subset of fewer than five individuals. The “bin” is the subset of data that meet a set of combined characteristics, and in this case, the SSA has set a minimum bin size of five. For health data, a larger minimum bin size is probably more appropriate, given the sensitivity of the information and possible linkability of the data to other datasets.

If useful and available combinations of data yield subsets below the desired minimum bin size, there are several options for releasing data to researchers, based on the concepts of data generalisation, data suppression and data perturbation.20 Data suppression eliminates non-essential data fields until the bin size requirement is met. For example, if gender is not an important variable for a particular research question, that field could be eliminated for that particular query. A more nuanced variation on this approach is data generalisation, in which data in some data fields are de-specified such that the bin size requirement can be met. An example of this would be to change birth dates to birth years. This technique can be used only for hierarchical data that have multiple levels of specification.21 Examples of hierarchical data are drug or disease classifications. For instance, rofecoxib (Vioxx) can be described hierarchically as “rofecoxib”, “Cox-2 inhibitor”, “non-steroidal anti-inflammatory” or “anti-inflammatory”. Using data of lower specificity would increase the bin size but may not significantly decrease the value of the data to the researchers. Have been developed software systems for medical record-type databases for which anonymity levels can be set by the database owner and thus control the specificity of the information that database users can obtain.22 Sweeney and colleagues have also developed software to evaluate the identifiability of DNA sequences23 and the linkability of DNA sequences to other data.24

Another variation of this approach would be to actually modify (perturb) data. For example, this may be carried out sometimes by changing “male” to “female”, by randomly eliminating a proportion of data in a dataset to disrupt the potential linkage to individuals. However, these methods would result in data of lower integrity and thus of lower value to researchers.

5.2.3 Limited data access

Similarly, privacy and confidentiality may be protected by limiting or monitoring access to the data. A simplified way of doing this is to allow only researchers with approved passwords to access a database. A more sophisticated version of this strategy is to use rule-based control of access to data, instead of, or in addition to, human intervention. In this method, different users are allowed to access different information according to their roles. Such a rule-based system has been developed by the Trusted Interoperation of Healthcare Information (TIHI) project at Stanford University.25 Users are tracked by the system and their queries and the results of their queries are filtered to be sure that they are “appropriate”. For example, cardiovascular researchers might not be given the right to access patients’ HIV status. This system is advantageous because it not



only provides a mechanism for controlling access, but also creates an auditable log of users, queries and responses.

Another method is to allow only a very limited set of analysts to query the primary data directly. In this scenario, “outside” researchers would be allowed to query the data indirectly, through these analysts, and would only receive summarised answers to their queries (e.g. means, P-values, etc.). Various initiatives have adopted this approach, including the Estonian project, CARTaGENE, the Icelandic HSD and the VA Cooperative Studies Program. In these cases, researchers submit queries that are peer reviewed and processed only when approved by one or more review boards.

These methods have the advantage of allowing the database managers to know, to some extent, who is using the database and for what purpose. The disadvantage is that researchers using the database have less flexibility in making queries, and the process may be slower.

5.2.4 Data encryption

Encryption, either one-way or public-key, is a common method of enhancing data privacy. This method may be used in conjunction with the methods described above. Because encryption can be fairly easily implemented, it is assumed that data transferred to and from research databases will be encrypted in some way. However, data that are encrypted can also be decrypted, so encryption should not be relied upon as the only means of privacy protection.

5.3 Access to population databases

5.3.1 Principles underlying access to HGRDs

Given that the main purpose of HGRDs is to foster research, issues of database access are fundamental. Key issues for consideration include who should have access (e.g. only researchers, public- and/or private-sector researchers), the manner in which access should be given (e.g. directly versus via an internal researcher), whether it should be free or for a fee (e.g. who should pay the fee and what it should be) and what access should be given (e.g. to the whole database, to parts and which ones, to only certain data and then only in an anonymised manner). Another key issue is the purposes for which access should be granted. This is a broad issue that has implications for informed consent issues, discussed previously.

The issue of which primary interests need to be protected in the context of human genetic research databases raises the question: Whose interests are primarily at stake? In responding, consideration must be given to three main parties: participants; researchers; and the public, the community and humanity.

The main challenge in the discussion of access to genetic research databases is to strike an appropriate balance between the freedom of researchers and the interests of participants and the public. Good normative guidance can be found in Article 5 of the Helsinki Declaration:26 “In medical research on human subjects, considerations related to the well-being of the human subject should take precedence over the interests of science and society.” The notion of well-being may be too narrow, since the vital interests of the participants cannot be fully covered under that heading. It is of crucial importance to be clear about what these interests are and how they are best protected without thwarting the other interests at stake.



In responding to the issue of primary interests, the ethical interests of each of the three parties need to be considered separately. Here is a schematic suggestion:27

� Participants: welfare (reduction of risk; precaution); respect for persons/dignity (privacy, choices are observed); justice (not being subjected to discrimination; protection of the vulnerable).

� Researchers: scientific freedom (free flow of data); justice (fair return).

� The public, the community and humanity: welfare (health-care benefits); respect (communal consultation); justice (not being subjected to discrimination and stigmatisation).

The next consideration is how these interests may be affected, and how they may conflict, in the context of access to the databases for research purposes. From the perspective of participants, their main interests seem to lie in being protected from access by someone who, owing to their inherent interests, could violate their welfare, respect or claim to justice. Insurers and employers, or other institutional third parties, which have an obvious motivation to misuse the information, would be examples of parties who should be denied access to the database. But the reasons for restricting access should certainly not be limited to “negative reasons”, i.e. the risk of outright misuse. Access should also be justified with “positive reasons”, i.e. by scientific accreditation and competence. This is implied in statements such as access to the databases “should be allowed only to registered, bona fide researchers who have published in the field”. If this “positive reason” is too strong (e.g. requirement for previous publication), however, it could unnecessarily restrict scientific freedom which needs to be balanced with the interests of participants in database research. Of course, this relates to the linkability of the data to participants: the more easily linkable, the more stringent access should be.

This argument is based on an “objective assessment” of possible risks of harm to participants. The assessment is best conducted by professional review boards with policies characterised by transparency and accountability, which are the basis of trustworthiness. But limits to access are not only justified by “objective assessment” of risk or harm (welfare and justice considerations) but also by respect for the choices offered in the consent process (dignity). Regulatory institutions worthy of trust must see to it that access is in accordance with what participants “subjectively” consented to. The issues of access and consent cannot be completely separated since they meet in the assessment of rightful use and access of the data.

It appears broadly accepted that even “private” population databases should allow access for academic research and should not stifle research on other, existing or future databases. However, in practice, the demarcation between scientific and commercial access is blurred for various reasons. First, in human genetic research, there is no clear division between pure academic science, on the one hand, and commercial application of scientific inventions, on the other. Second, academic scientists are increasingly pressured by their funders (government or charity) to commercialise their research results, which requires secrecy rather than free public access. Third, even if scientists do not commercialise their research results, they may be forced to be secretive about the human health and genetic databases they build, since “their” database might be a unique selling point when they have to compete for national or international grants.

A detailed scientific and technical analysis of the different types of databases is, however, most important from an ethical perspective because the way in which moral interests are best protected will always be context-dependent. That task requires detailed



knowledge about the type of data, their linkability, the use and users of the data, and in general the purpose of access. All these facts matter not only for legal regulatory discourse but also for the ongoing critical ethical assessment of the appropriateness of access.

5.3.2 Purpose and conditions of access to HGRDs

Most HGRDs have developed a policy that they will allow access to researchers, whether domestic or foreign, from academia or the commercial sector. The UK Biobank’s policy is that all researcher and all avenues of research will not, a priori, be limited. In the Estonian initiative, researchers will have access at the discretion of the chief processor. For the Icelandic HSD, researchers will have access at the discretion of the database operator. For most of the HGRDs, access to its resources will be subject to an access agreement. In fact, the UK Biobank stipulates that the licence fee may be set at a higher rate for entities expected to derive a financial benefit from the use of the database. In all cases, the research protocol must have received approval from ethical and scientific bodies. The HGRDs will not allow direct access to the database, but will provide researchers with data and information in either anonymised or coded format. Most of the HGRDs have not explicitly set out whether they will allow access to the biological samples. Exceptionally, the UK Biobank has indicated that the use of biological samples will be carefully coordinated and controlled as they are limited and depletable. While most of the HGRDs will allow the sending of data and samples outside of their jurisdictions, the Icelandic HSD Act and the Estonian Human Genes Research Act require that the database and tissues be stored within the jurisdiction. Most of the HGRDs will not permit access to the participants directly. However, the UK Biobank has foreseen in its re-contact policy the possibility of the Biobank re-contacting participants to discuss the possibility of their involvement in new studies requiring new information and samples.

In Iceland, the HSD Act is not explicit with respect to the nature of the research request submitted and appears to leave the exact terms of such access to be negotiated by the parties. The main formal and substantive aspects which these agreements are required to incorporate28 only mention scientific work and “the demarcation of the centralised Database and external databases and the rights of medical institutions and employees to data”. They do not contain exact terms relating to researchers’ access to the HSD.29 The Licence requires that the health data be processed so that they can be used in scientific research. The Parliamentary Notes refer to the requirement that the computerised databases at health institutions, which will be created by the processing of the health data, be usable in scientific research and that the data in the HSD be usable, both in the Licensee’s own research and by other scientists. Pursuant to the Licence, the Monitoring Committee shall supervise the making of these agreements and thereby protect the interests of scientists, among others. The Licence does not specify on what terms and conditions scientists would be allowed to access the HSD. In brief, the general principle of access for public research to the HSD is not clearly established in the Act, with the exact terms of such access and the issue of ownership of the results of such research left to be negotiated among the Licensee, the health institution and physicians.

The Interdisciplinary Ethics Committee must approve all requests for research and searching of the database, although decisions may be appealed to the Minister (Health and Social Security) who will seek an opinion from the Science Ethics Committee.30 The legislation does not give police authorities specific access to the HSD, but it has been noted that the courts maintain jurisdiction over the database, and thus might subpoena



data or material. As noted, various bodies have responsibilities relating to access and security, but certain procedures are also specified in the legislation. For example, one key requirement is that “Processing in the Health Sector Database shall not begin until an assessment has been performed by an independent expert on the security of information systems.”31

Under the Estonian Human Genes Research Act, the Gene Bank may be used for scientific research, research into and treatment of illnesses of gene donors, public health research and statistical purposes. The chief processor (the EGPF) has the right to perform genetic research. It may also issue, to a gene researcher, the link between coded tissue samples, coded descriptions of DNA and coded descriptions of state of health based on blood relationships, but only in coded form. An authorised processor or gene researcher shall unconditionally deliver descriptions of DNA or parts thereof to the chief processor. The chief processor may grant an authorised processor or gene researcher the right to use a description of DNA or a part thereof with or without a charge. The chief processor shall grant gene researchers who are legal persons in public law or state agencies the right to use descriptions of DNA or parts thereof without charge. The Act also stipulates that results of genetic research and intellectual property rights related thereto shall be provided by law. The Act explicitly prohibits the use of the gene bank “to collect evidence for criminal or civil proceedings or for surveillance” but has little to say in relation to more general questions of access.

Uniquely among the databases studied, the Human Genes Research Act gives participants in the Estonian Genome Project full rights of access to their genetic data. This raises questions regarding the mechanisms for ensuring the appropriate distribution of such information. (In Iceland participants also have a right to access their data, but this was a pre-existing right, not specifically related to the HSD.)

The UK Biobank Ltd. will retain full control of all access to and use of the database and sample collection. Access will be provided for research purposes only. Exclusive access to the fully developed resource will not be granted to any party. The UK Biobank Ltd. will develop an overall policy and detailed terms of access, addressing fairness and transparency of decision-making, the handling of conflicts of interest and the prioritisation of use of samples. Access to data and/or samples will be granted under licence. Licences will be for specific uses under the terms and conditions of standard access agreements. Fees will be charged for a licence, with the possibility that charges may be higher for organisations that might be expected to derive financial benefit from use of the resource.

Research findings will be required to be incorporated back into the resource. Research users will be required to place both positive and negative findings (i.e. those failing to show association) in the public domain. The UK Biobank Ltd. will explore strategies for disseminating negative findings, e.g. by establishing an accessible archive. This policy will apply to all users, commercial and non-commercial. Researchers will be permitted to keep results confidential for limited periods, for example while they prepare papers for publication or file patent applications. Any population-based information having general medical usefulness should be published to allow it to be evaluated and applied in health care.

Requests for research access to the UK Biobank will be reviewed by a National Health Service (NHS) Multi-centre Research Ethics Committee, and must meet UK Research Governance Framework guidelines. These external organisations are accepted as the authority on this matter, as the EGF of the UK Biobank does not close off any



avenue of research, stating that “UK Biobank will not proscribe any research uses at the outset”.32 The UK Biobank does not automatically prohibit access for purposes of law enforcement, although this will only be granted under a court order, and such action may be resisted by UK Biobank. The details of security mechanisms for the UK Biobank are not spelled out in the EGF, and there is no specific mention of any procedure for assessing such security mechanisms.

5.4 Demise of database

While there are currently few examples of HGRDs that have failed or been terminated (such as the Tongan database established by Autogen Limited), consideration of the possible demise of an HGRD should be undertaken at the time of its establishment. Consideration should be given to clearly determining the consequences of its demise. For example, should all data and samples be destroyed? Should participants be notified of the demise of the database? If the database is operated by a private undertaking, should provision be made for a government to retain the right to have the database handed over to them or should they at least have a right of first refusal? Such considerations will be influenced by the applicable legislation. For example, many countries have enacted legislation prohibiting the sale of human tissue or material. The consequence of the application of such statutes would also need to be taken into account when developing such a policy. Moreover, the manner of disposal (destruction versus sale) needs to be addressed.

In Iceland, Article 5 of the HSD Act and Article 9 of the Operating Licence addresses the matter, stating that rights to the database return to the issuer (Ministry of Health and Social Security), and that the Monitoring Committee will operate the database until a decision is taken on its future. The HSD Act contemplates the duration of the exclusive Licence to not exceed 12 years. A separate agreement has been made relating to the transfer of rights from the Licensee to the Issuer.33 The Act clearly provides that the Licence and the HSD can neither be transferred nor subjected to attachment for any debt. Neither the Licence nor the HSD can be used as collateral for financial liabilities.

In Estonia, the samples and the data entered into the Gene Bank belong to the chief processor, i.e. the EGPF. This is a non-profit foundation created by the Republic of Estonia. The Act does not provide for what happens to the Gene Bank if the Foundation becomes bankrupt. The Act does provide that the tissue samples and any un-coded information under the ownership of the chief processor are not transferable. Consequently they are not subject to seizure or attachment and cannot be used as collateral for financial liabilities.

UK Biobank Ltd. will be the legal owner of the database and the sample collection. Such ownership contains rights, including the right to sell samples. UK Biobank Ltd. does not intend to exercise this latter right. Its stewardship of the resource implies that it will judiciously protect the UK Biobank, which protection extends to the careful management of any transfer of parts or all of the UK Biobank. A detailed strategy34 is being developed for handling contingencies in the event that UK Biobank Ltd. has to close or make other substantial transitions in the holdings or control of the resource. It emphasises the importance of protecting the rights of participants. The objective of this strategy will be to ensure that the protection and respect for the rights of the participants provided by the EGF continue to be maintained and that the Ethics and Governance Council is consulted on the proposed terms before any changes or transfers are made. It



addresses the possibility of partial or full transfer or sale of the resource, whether elective or as a result of insolvent liquidation.

While the CARTaGENE35 initiative has not spelt out a policy for its demise, it is foreseen that the project will have a limited lifespan of 50 years. In this regards, CARTaGENE has determined that at the end of the 50 years all of the samples, data and results will be destroyed. Participants are informed of this policy early on during the consent process. The samples, data and results will not be destroyed in the situation where the Minister of Health and Social Services orders CARTaGENE otherwise.



Notes

1. Estonia, Human Genes Research Act, RT I 2000, 104, 685, entered into force 8 January 2001 www.geenivaramu.ee/index.php?lang=eng&sub=18&eetika=1 (accessed 15 September 2006).

2. RMGA, Statement of Principles: Human Genome Research (VERSION 2000), see www.cartagene.qc.ca/docs/enonce.pdf (accessed 15 September 2006); RMGA, Statement of Principles on the Ethical Conduct of Human Genetic Research Involving Populations, see www.rmga.qc.ca/doc/ENONCE2002.ENG.pdf#search=%22Ethical%20Conduct%20of%20Human%20Genetic%20Research%20Populations%22 (accessed 15 September 2006).

3. The Commission d’accès à l’information is responsible for the enforcement of two acts: the Act respecting Access to Documents Held by Public Bodies and the Protection of Personal Information and the Act respecting the Protection of Personal Information in the Private Sector.

4. Elsagen, UK Law report, Jane Kaye and Peter Petkoff, University of Oxford, September 2003.

5. United Kingdom, Ethics and Governance Framework, May 2006, See www.ukbiobank.ac.uk/docs/EGF%20Version02%20May%202006.pdf (accessed 15 September 2006)

6. Iceland, Government regulation on a Health Sector Database, No. 32/2000, Chapter V, Articles 15-24.

7. Iceland, Government regulation on a Health Sector Database, No. 32/2000, Chapter VI, Articles 25 - 29.

8. Iceland, Government regulation on a Health Sector Database, No. 32/2000, Chapter VI, Article 26.

9. Operating Licence for the Creation and Operation of a Health Sector Database, Article 3.3, January 2000. (“Operating Licence”).

10. See http://www.geenivaramu.ee/index.php?lang=eng&sub=72 (accessed 18 May 2006)

11. See www.geenivaramu.ee/index.php?lang=eng&sub=78 (accessed 15 September 2006).

12 See www.cartagene.qc.ca/ipeg.cfm (accessed 18 September 2006).

13. Our Inheritance, Our Future: Realising the potential of genetics in the NHS (A Genetics White Paper Presented to Parliament by the Secretary of State for Health By Command of Her Majesty), June 2003, see www.conversations.canterbury.ac.nz/resources/genetics/uk-%20white-



paper.pdf#search=%22Our%20Inheritance%2C%20Our%20Future%2C%20%22 (accessed 20 September 2006).


15. Iceland, Act on a Health Sector Database, no 139/1998, 17 December 1998, Section VI, Article 13- 17.

16. Operating Licence, Articles 11.6 and 13.1, January 2000.

17. Iceland, Act on a Health Sector Database, no 139/1998, 17 December 1998, Section VI, Article 13- 17.

18. Cho, Mildred (2003), “Confidentiality and Data Management”, Background Paper for the Tokyo Workshop.

19. Subrule concerning Measures for personal information protection. Ethics Guidelines for human genome/gene analysis research, 21 March 2001, Japan.

20. L. Sweeney (2002), “k-anonymity: a model for protecting privacy”, International Journal on Uncertainty, Fuzziness and Knowledge-based Systems, 10 (5), 557-570.

21. Zhen Lin, Teri Klein and Russ Altman (2001), Methodology on scrubbing data in PharmGKB, paper presented at the Pacific Symposium on Biocomputing, Kohala Coast, Hawaii.

22. L. Sweeney (1997), “Weaving technology and policy together to maintain confidentiality”, Journal of Law, Medicine and Ethics, 25:98-110; L. Sweeney (1998), “Datafly: a system for providing anonymity in medical data”, in T. Lin and S. Qian (eds.), Database Security XI, Chapman and Hall, London.

23. B. Malin and L. Sweeney (2000), “Determining the identifiability of DNA database entries”, Proceedings/AMIA Annual Symposium, 537-541.

24. B. Malin and L. Sweeney (2001), “Re-identification of DNA through an automated linkage process”, Proceedings/AMIA Annual Symposium, pp. 423-427.

25. G. Wiederhold, M. Bilello, V. Sarathy and X. Qian (1996), “Protecting collaboration”, pp. 561-569 in Proceedings of the National Information Systems Security 1996 Conference.

26. World Medical Association, Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects, Adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964, and last amended WMA General Assembly, Tokyo 2004, see www.wma.net/e/policy/b3.htm (accessed 20 September 2006).

27. Vilhjálmur Árnason (2003), “Comment on Access to Mutation Databases for Research Purposes”, Tokyo Workshop Background Paper.

28. Annex C to the Operating Licence, January 2000.

29. Iceland, The Regulation on Scientific Research in the Health Sector (No. 552/1999) does not address the issue of access for research on the HSD.



30. Iceland, Government Regulation on a Health Sector Database, No. 32/2000, Article 27.

31. Iceland, Government Regulation on a Health Sector Database, No. 32/2000, Article 5.


33. Operating Licence, January 2000.

34. United Kingdom, Ethics and Governance Framework, Part D Transfer of Assets or Closure, May 2006, See www.ukbiobank.ac.uk/docs/EGF%20Version02%20May%202006.pdf (accessed 15 September 2006).


CHAPTER 6. COMMERCIALISATION CONSIDERATIONS – 123


Chapter 6

Commercialisation Considerations1

HGRDs raise various commercialisation issues, including those pertaining to intellectual property rights, commercialisation of the database and benefit-sharing. However, interestingly, most of the population database initiatives do not have detailed policies with respect to intellectual property, commercialisation or benefit-sharing.

Intellectual property broadly refers to the legal rights that result from intellectual activity in the industrial, scientific, literary and artistic fields. One set of issues raised by HGRDs pertains to the intellectual property rights that may arise as a result of research employing data or samples accessed from a database. Questions arise with respect to the ownership of the intellectual property rights pertaining to the invention, who is obligated to ensure the protection of the relevant intellectual property rights, and access to the innovation developed using data and samples accessed from the population database. What should such a policy require to ensure follow-on access while permitting a return on investment? Intellectual property issues also arise with respect to the database itself, including database rights, where they exist, copyright protection for the software and other rights to ensure that the database can operate effectively.

With respect to commercialisation, the first consideration is whether or not it is desirable to commercialise the database and/or whether commercialisation is in line with participants’ expectations. If the population database is to be commercially exploited, consideration should be given to the appropriate manner. Should the basis be exclusive or non-exclusive? If exclusive, how can fair access to the database and compliance with competition law be ensured? Should it be possible to sell or transfer the database?

The issue of benefit sharing is complex, with aspects that vary depending on the structure of the database. For example, in the context of an HGRD established as a public-private partnership or as a private undertaking, consideration should be given to whether the government should be entitled to compensation from the private entity and the form of such compensation. If monetary compensation is the option favoured, for what purpose should these amounts be employed? The compensation may also take the form of technical or scientific support. Benefit-sharing also raises the issue of whether or not participants should be entitled to individual benefits arising from the database. For example, would participants be entitled to share in the profits of a successful invention developed using data and samples contained in the database? Should participants have the

This chapter also draws on texts prepared for the Tokyo Workshop: Jasper Bovenberg, Hiroshi Gushima and

Jaanus Pikani. See Bibliography.

124 – CHAPTER 6. COMMERCIALISATION CONSIDERATIONS


right to access other, non-monetary benefits, such as the products developed as a result of research involving data and samples from the HGRD?

6.1 Intellectual property

6.1.1 Intellectual property generally

Intellectual property broadly refers to the legal rights resulting from intellectual activity in the industrial, scientific, literary and artistic fields. The diverse types of intellectual property rights arise from international instruments, regional and national legislation and regulation as well as judicial or administrative decisions. Generally, jurisdictions will enact intellectual property regimes for two main reasons. One is to give statutory expression to the moral and economic rights of creators in their creations and the rights of the public for access to those creations. The second is to promote, as a deliberate act of government policy, creativity and the dissemination and application of its results and to encourage fair trading which would contribute to economic and social development.

Generally speaking, intellectual property law aims at safeguarding creators and other producers of intellectual goods and services by granting them certain time-limited rights to control the use made of those productions. Those rights do not apply to the physical object in which the creation may be embodied but instead to the underlying intellectual creation. In the life sciences, including human genetics, the intellectual property rights most commonly arising include patents, trademarks, database protection and undisclosed information (also referred to as trade secrets or proprietary information).

6.1.2 Intellectual property and population databases

Only a few of the population databases have developed any policy with respect to the intellectual property rights arising from the database and the intellectual property rights that may arise from the use of the data and samples in the database. Intellectual property rights may arise in many areas. For example, they may arise with respect to the database and its contents, to the software designed for the database, to inventions developed from research based on the data, information and samples accessed from the database and even to follow-on innovations. While it is important that population databases contemplate and develop a policy on intellectual property rights, the nature of that policy may vary widely and will reflect the biobank’s nature, purpose, structure and objectives. For example, even in a not-for-profit structure, not all intellectual property rights will automatically have to be attributed to the biobank. Equally, in a biobank established as a private/for-profit endeavour, it may be that some intellectual property rights would more appropriately be attributed to a public authority. An examination of the IP policies of the various biobanks reveals this diversity of attribution intellectual property rights.

For the Icelandic population database, neither the statute nor the regulations contain an explicit provision to the effect that the HSD and/or any intellectual property rights (IPRs) arising from any research conducted on the HSD belong to the Licensee. Nevertheless, the HSD Act allows the Licensee to operate the HSD for financial profit and thus such rights would reasonably vest with the Licensee. The Licence contains an extensive section on intellectual property rights. All of these provisions aim to secure that the government, upon expiry or termination of the Licence, can avail itself of the HSD and any rights, including intellectual property rights, necessary to operate it. The underlying premise would appear to be that these rights originally vest in the Licensee.



In Estonia, since rights of commercial access to the Gene Bank were originally transferred to EGeen1, the transfer agreement covered this issue and these rights would most likely have vested with either EGeen or the party that conducted the research. The terms of the agreement may not be disclosed without permission of the parties thereto. The Act simply provides that an authorised processor or gene researcher shall unconditionally deliver descriptions of DNA or parts thereof to the chief processor. The chief processor may grant an authorised processor or gene researcher the right to use a description of DNA or a part thereof for a charge or without charge. The chief processor shall grant gene researchers who are legal persons in public law or state agencies of the Republic of Estonia the right to use descriptions of DNA or parts thereof without charge. Results of genetic research are not specifically addressed and any intellectual property rights related thereto are provided by law. The Foundation must obtain the intellectual property rights to academic research results. Moreover, patents are jointly owned by the Foundation and the Licensee.

In January 2005, the UK Biobank published a new draft of its Intellectual Property and Access Policy.2 The Policy establishes that intellectual property accruing from the creation and development of the Biobank resource, such as a database rights, will vest in UK Biobank. Materials created by third parties outside the UK Biobank project, such as health records, will be licensed by such third parties to UK Biobank for use in the project. The intellectual property in the infrastructure (e.g. in the project information technology systems) will vest in or be managed by UK Biobank according to the needs of the project.

UK Biobank will be responsible for managing intellectual property in the database and its infrastructure and ensuring it has all the rights necessary to further the objectives of the UK Biobank. To this end, it will identify new entitlement to intellectual property rights and ensure they are protected. This will facilitate the protection, curation, control and use of the Biobank in accordance with the Ethics and Governance Framework, the Intellectual Property and Access Policy and other legal and ethical requirements. Third parties involved in its creation (e.g. the regional collaborating centres) will be required to assign any new intellectual property to the UK Biobank. Some of these IPRs (e.g. the intellectual property in new technologies developed to establish the information technology infrastructure) may have commercial value. Where these vest in the UK Biobank, they will be managed and exploited in accordance with UK Biobank’s duties as a charity, for the long-term benefit of UK Biobank and the public.

In standard arrangements for access, intellectual property arising out of research using the database will vest in the investigator creating it, his or her institution or, in appropriate cases, their assignee(s). The standard agreement for access will not assert rights to this intellectual property for UK Biobank. It will require users to have suitable arrangements to ensure that intellectual property arising from their use of the Biobank vests in them, their employing institution or, in appropriate cases, their assignees and is properly identified, managed and exploited. Where the research involves a number of users or institutions, suitable arrangements to manage any intellectual property rights arising from the collaboration should be made in advance. The UK Biobank will reserve the power but not the duty under the Access Agreement to “step in” to protect or exploit intellectual property in suitable cases if, after a period of time, the lead institution chooses not to do so. It is intended that six months’ notice of the intention to “step in” will be given before proceeding.

The Access Agreement will require users to provide the UK Biobank with details at the point of first publication and grant of patent applications based on the results of



research carried out under the Access Agreement. Arrangements for intellectual property in studies that involve the collection of additional information or biological samples or otherwise expand the scope of the Resource will be developed and reflected in the IP and Access policy in due course.

It is not expected that the UK Biobank will generally claim a share of royalties generated from the results of research conducted using its resources. However, in certain circumstances, collaboration could lead to arrangements for revenue sharing. Any royalty income will be used by the UK Biobank to implement its purpose for public benefit.

The Icelandic and Estonian legislation does not provide for specific patent protection or other intellectual property to be granted to the licensees of the databases. The ethical framework of the UK Biobank acknowledges that use of the Biobank may lead to patentable inventions. However, in either scenario the issue is not whether the results of genetic research should be patentable, since this is dealt with by the patent law (domestic and international) and the policies of patent offices. Rather the issues to be resolved with respect to the population databases include: i) who should own the intellectual property rights, the government or the licensee or should they be held jointly? Or should the government have a right of first refusal? ii) Should the licensee be under an obligation to use or to let others use the intellectual property? iii) Can intellectual property relating to a national population database be owned by a foreign entity?

Another, more specific, intellectual property issue is that of rights accruing in databases. In some jurisdictions, there is no statutory protection specific to databases, so they would be subject to existing intellectual property protection such as copyright. Conversely, other jurisdictions have enacted statutes applicable directly to the protection of databases. The European Union Directive on the Legal Protection of Databases3 deals directly with the ownership of databases by giving database builders specific protection – “database rights” – which recognise the work and costs entailed in compiling, verifying and presenting data. “Database rights” allow the holder to prevent extraction or reutilisation of substantial parts of the database for a period of 15 years following completion of the database. In fact, the rights may continue in perpetuity since they can be rolled over following any substantial modification to the database that requires a substantial investment. The research exemption provided for in this EU Directive is rather narrow and not mandatory. Although the EU database rights were not designed with population health or genetic databases in mind, licensees of such databases could rely on them for enforcing their proprietary claims within the EU. The EU database rights might be disputed before the courts, outside the EU, since a similar treaty is being negotiated at the World Intellectual Property Organisation (WIPO) and the reciprocity provision in the EU Directive might prompt other jurisdictions to adopt similar protection.

6.2 Commercialisation

In Iceland, many commercialisation issues are governed by the statute, regulations and operating licence. It is acknowledged that the Licensee is authorised to use the data in the HSD for deriving financial profit. This appears to include the issuance of licences to third parties. However, neither the Licensee nor third parties are permitted to transport the HSD outside of Iceland. The Licensee may not grant direct access to data in the HSD. Any third party research on or queries to the HSD may only be processed by using query layers. The exploitation of the Licence is subject to national and European Economic Association (EEA) competition rules. This implies that the Licensee must refrain from abusing its position as an exclusive Licensee. For example, the Licensee should refrain



from exacting unreasonable fees and from discriminating among business partners. Licensing conditions must be general and transparent and any research, individual queries or query classes processed on the HSD must pass scientific and ethical review.

In Estonia, the Act grants the chief processor, inter alia, the right to “issue” descriptions of state of health, genealogies and genetic data and to grant an authorised processor or gene researcher the right to use a description of DNA or a part thereof, with or without charge. The chief processor also has the right to assign all the processing rights to an authorised processor. The Project intends to grant non-exclusive licences to and partner with private and public sector entities.4

Commercial companies and other research endeavours that stand to profit will be allowed access to the UK Biobank if their proposal falls within the purpose of the UK Biobank and passes scientific and ethical review. It is acknowledged that distinctions between “commercial” and “non-commercial” users may be blurred. Academic researchers may create inventions (even if this is not the original intent), the rights to which are passed on to for-profit firms. In principle there is no objection to use of the resource by pharmaceutical and other firms. However, it has been proposed that such firms may be asked to pay higher fees for access than other users.

CARTaGENE informs potential participants in its information and consent form that their data and biological samples may lead to the commercialization of a test or product. However, it also explicitly states that the participant will not derive any personal financial advantage from this commercialization.

6.3 Benefit sharing

The term “benefit sharing” connotes various notions in different contexts. In some circumstances, it is employed to mean a payment back to the source of the information or sample. In other circumstances, it may be employed more broadly to include the sharing of technical information or the provision of services. Consideration would have to be given to the type of benefit-sharing approach that a database would adopt and to the resources available for carrying out such an approach. In the context of HGRDs, the issue appears to be which, if any, and how benefits should be shared. Should certain forms of benefit sharing be banned as possibly creating conflicts of interest? Should any benefits be available at the individual level?

In Iceland, the HSD Act does not contain a specific provision on benefit sharing. Nevertheless, the Act does stipulate that the Licensee will pay the costs for the issuance of the Licence, of the various supervising committees and for the collection of the health data, including a system of computerised medical records in clinics and hospitals. The Act also provides that the government and the Licensee may agree on further payments to the government. According to the Licence Agreement, the Licensee, subject to certain adjustments, was to pay to the government i) an annual fixed fee, earmarked for the promotion of health care and R&D; and ii) 6% of its profits, capped at ISK 70 million a year. Upon expiry of the Licence, the government may elect not to extend it. The Licensee must then deliver the HSD to the government and ensure that the government is granted use of all software and rights necessary for its operation. This could result in an indirect benefit to the Icelandic people. In the event that the government continues operation in the service of the public health-care system, for non-business purposes, the Licensee would not be entitled to remuneration for software and IPRs, provided that the government pays certain service fees. In the event that the government continues



operating the database for business purposes or recommends operation of the HSD for business purposes within five years, the Licensee would be entitled to remuneration for software and intellectual property rights, based on the future use by the government or its assignee of the software and/or rights for business purposes, to be agreed upon by the government and the Licensee.

In Estonia, the participant does not have the right to receive a share of the income of the investors, funds or the EGPF. The Act provides that a participant is not entitled to request a fee for providing a tissue sample, preparation and study of a description of his or her state of health or genealogy, or use of the research results. However, participants have the right to access their data stored in the Gene Bank at no charge, but not their genealogies, and they have the right to genetic counselling if they access their data. The Act does not provide however, that such counselling should be provided at no charge, although it may be part of the national health-care insurance package. The informed consent form requires the participant to declare that he/she is aware of the fact that: i) her/his tissue sample may have some commercial value and that commercial entities may receive anonymous data about gene donors; ii) the right of ownership of the tissue sample, of the description of state of health and of other personal data and genealogy shall be transferred to the EGPF; and iii) the EGPF is controlled by the government but may be financed by commercial entities. Additional benefits may be agreed upon in the commercial access agreement between the EGPF and a commercial entity, the contents of which may not be disclosed without the consent of the parties.

The purpose of the UK Biobank is to learn from participants’ collective health experience over time, in order to generate and disseminate new knowledge for the benefit of public health in the United Kingdom and elsewhere. Knowledge derived from studies on the UK Biobank will be published in the world’s scientific and medical literature, communicated to the UK Biobank’s participants, the National Health Service (NHS), and others, as appropriate, accumulated and made available by the UK Biobank as a resource for further research. Such knowledge may also be applied to the development or improvement of health-care technologies or techniques. The idea of rewarding participants or their communities if any profits accrue from use of the UK Biobank resource was discussed but it was decided that, given the scale of the UK Biobank, the fact that participation is voluntary and meant for the good of others, and the expectation that in the long run many people will benefit in many different ways, any financial income should simply be reinvested in the resource.

The CARTaGENE initiative indicates to participants during its consent process that while their data and samples may someday lead to commercialisation of products or tests, that they will not derive any personal financial advantage from this commercialisation. Nevertheless, Recommendation 7 of its Statement of Principles on the Ethical Conduct of Human Genetic Research involving Populations5 raises the issue of benefit sharing. This Recommendation proposes that eventual sharing of any benefits with the population should be discussed at the outset and that it may take different forms, including access to medical care, future treatment or drugs developed; a contribution of a portion of the benefits to humanitarian organisations; support for local needs, technological infrastructures or health services to the population. This Recommendation also stipulates that the benefit cannot be limited to the participants but should be provided to the whole population and that benefit sharing should not constitute undue inducement to participate. Moreover, it entreats freedom of research through the promotion of public access to the gene bank. The Recommendation on Contribution to the Welfare of the Population provides that population genetic research should endeavour to promote health and prevent



disease, especially for the population studied; partnerships with local research teams should be established and copies of samples must remain in the jurisdiction of origin. Pursuant to the Recommendation on Contribution to the Welfare of Humanity, the universality of the human genome mandates the sharing of knowledge at an international level.

As the analysis across the various initiatives reveals, HGRDs are only beginning to consider and to develop policies with respect to the commercialisation considerations. However, the challenges surrounding these issues remain mostly unexplored and suggest future areas for policy research.



Notes

1. The Estonian Gene Bank and EGeen terminated their exclusive agreement in December 2004. See Estonia, “Genome Project Ends Cooperation with Current Financier”, 27.12.2004, www.geenivaramu.ee/index.php?lang=eng&show=uudised&id=172&PHPSESSID=197f1b8a33cef08aa94f377cb65cf596 (accessed 15 September 2006).

2. UK Biobank, Policy On Intellectual Property And Access, January 2005, see www.ukbiobank.ac.uk/docs/UKBiobankIPandAccesspolicyfirstpublicdraft11.1.5final2.pdf (accessed 20 September 2006).

3. EU, Directive on the Legal Protection of Databases, (96/9/EC).

4. Estonia, Genome Project Ends Cooperation with Current Financier, 27.12.2004, see www.geenivaramu.ee/index.php?lang=eng&show=uudised&id=172&PHPSESSID=197f1b8a33cef08aa94f377cb65cf596 (accessed 20 September 2006)

5. RMGA, Statement of Principles on the Ethical Conduct of Human Genetic Research Involving Populations, www.rmga.qc.ca/doc/ENONCE2002.ENG.pdf#search=%22Ethical%20Conduct%20of%20Human%20Genetic%20Research%20Populations%22 (accessed 15 September 2006).

CHAPTER 7. CONCLUSIONS – 131


Chapter 7

Conclusions

The objective of the workshop was to gain an understanding of current practices in OECD member countries for acquiring and maintaining samples or specimens from which the data/information is derived as well as the data per se. The discussion aimed to develop a common understanding of current practices and the identification of the challenges raised by the establishment, governance and management of HGRDs. It drew on the presentations of numerous experts from academia, the private sector, and the public sector with experience in and knowledge of human genetic research databases.

The major conclusions of the workshop were that:

� Human genetic research databases are considered an invaluable tool for research into the genetic basis for disease.

� While there are divergent schools of thought, there remains no expert consensus on whether genetic information should be treated as distinct from other personal, medical or health-related information.

� Public, and in particular participants’, trust in the development, management and governance of HGRDs remains an essential element of an enabling environment for health research and innovation in this field.

� Public engagement in the development of such databases is essential for ensuring their viability as well as community support for and participation in such undertakings. The workshop considered a number of practical approaches to assessing and ensuring public engagement and trust.

� The issues of linkability or anonymisation of data were seen as closely related to those of privacy and confidentiality. Moreover, given the difficulty of ensuring complete anonymity and unlinkabiltiy of data, participants should be informed at the moment of consenting to participate in the HGRD.

� Clear procedures must be in place for informing patients about how their data may be used in HGRDs. Experts questioned whether current approaches to informed consent were sufficient to ensure patient privacy and achieve an appropriate balance with research access. Whether or not such a balance is achieved in public policy will affect how successful genetic science is as a driver for innovative products and processes and delivery of better health.

132 – CHAPTER 7. CONCLUSIONS


� The commercialisation dimension, while different across the diverse HGRDs, raises tensions between free access to data and the desire for commercial benefits. Nevertheless, more than one model is available for establishing an HGRD.

� While the breadth of population genetic databases being established implies that a set of mandatory rules applicable to all such databases may not be appropriate at this stage, the workshop concluded that identification of best practices for the management and governance of such databases was a valuable undertaking for the OECD.

7.1 Policy themes arising from the workshop

Various policy themes emerged at the workshop. They are discussed in the Rapporteur’s report1 and are briefly summarised below.

7.1.1 Is genetic information special?

An issue that permeated the discussion concerning the development of policies on genetic samples and genetic information is whether genetic information is truly unique and worthy of special protection. Is genetic information really different from other highly sensitive personal information? Are special privacy and consent policies needed? Participants at the OECD workshop did not reach consensus on these points. For example, it was noted that unlike some forms of personal information, genetic data is easily reduced to a standardised state that is amenable to analysis, sharing and storage. As a result, confidentiality concerns may be heightened. However, it was also suggested that other forms of health information, such as data collected in traditional clinical trials, are also highly sensitive and have not been the subject of the same degree of ethical scrutiny.

The differing views regarding the status of human genetic material are reflected in the academic literature.2 Some entities treat human genetic information as a unique form of personal information. An example is the 2003 UNESCO Declaration on Human Genetic Data. Article 4 states that human genetic data have a special status because: “i) they can be predictive of genetic predispositions concerning individuals; ii) they may have a significant impact on the family, including offspring, extending over generations, and in some instances on the whole group to which the person concerned belongs; iii) they may contain information the significance of which is not necessarily known at the time of the collection of the biological samples; and iv) they may have cultural significance for persons or groups.” As a result, “[d]ue consideration should be given, and where appropriate special protection should be afforded, to human genetic data and to the biological samples”. However, it was also noted that the appropriate characterisation of human genetic material should remain a live debate. Policies that are developed on the premise that genetic information requires special protection may legitimise inappropriate notions of genetic essentialism, albeit inadvertently.

7.1.2 Public perceptions

A closely related issue concerns the nature of the risks associated with HGRDs. A distinction appears to exist between how the research community and the public and policy makers view the issues of risk. Many in the research community characterise personal risks as minimal, in part because HGRDs often involve analysis of genes that are likely to have little or no clinical value to individual patients. In addition, it was noted that, when asked, there is generally a very high level of public participation in an HGRD,



even in countries like Iceland where there was a good deal of social controversy. This suggests, for some, that the public feels a degree of comfort with HGRDs and that they may perceive the risks as tolerable.

Despite these levels of public participation, HGRDs continue to be characterised as individually and socially controversial ventures. There is evidence that the public, rightly or not, views genetic information as highly sensitive.3 In addition, there is at least some evidence that the public believes that research databanks and research on health information involve more than minimal risk.4 Since HGRDs often involve research on specific populations or sub-populations, they may generate issues of stigmatisation or discrimination that are unique to the community studied. As a result of these and other concerns, several national and international policy-making entities have suggested that specific precautions need to be taken in relation to HGRDs. Additional research on the sources and nature of public perception of risk and benefit in this context was identified as useful.

In addition, the impact of public representations of genetic research was considered a key issue. Over the past 15 years, genetic research has been profiled by the research community as a critically important area of investigation. It has been suggested that genetic research will have profound benefits for health care and that our genes are centrally important to both disease development and behavioural characteristics.5 At the same time, there has been an unprecedented amount of research into the legal, ethical and social issues associated with genetic research. This has led to a great deal of high-profile speculation about potential social harm. What impact have these public representations had on the perceived risks and benefits of HGRD and genetic research generally?

7.1.3 Public trust

One of the most consistent themes emerging from the OECD workshop was an appreciation of the central importance of gaining and respecting the public’s trust. This seems especially important given that many of the large HGRDs involve a significant amount of public funding and that participants offer a genetic sample with little or no possibility of personal benefit. Participation in an HGRD can be viewed as an altruistic act done with a presumption that the research may benefit society.

A number of elements were seen as central to the maintenance of public trust. First, the public should be meaningfully engaged at both the design stage and throughout the life of the project. This public engagement should not be viewed as a marketing strategy or way to “sell” the HGRD initiative. Rather, it should be viewed as a means of creating an ongoing and informed dialogue about the research project’s benefits and risks, including a balancing of the possible scientific and health outcomes against privacy concerns. Second, there should be a high level of transparency in relation to the relevant research objectives, commercialisation strategies and the privacy and consent policies. Finally, the explicit, primary goal of large public HGRDs should be to increase knowledge and improve global health and well-being.

While maintaining the public’s trust was viewed as worthy end in its own right, experience has shown that an erosion of public confidence can also have a profound impact on the research environment and, even, the acceptance of novel technologies.



7.1.4 Human rights norms and existing legal frameworks

The policies surrounding HGRDs will need to be sensitive to existing national constitutional and human rights norms. Indeed, as exemplified by a verdict of the Supreme Court of Iceland, national constitutional law – which often flows from well-established human rights principles – can have an impact on the content of HGRD policy. In the Icelandic case, there was a challenge to the presumed consent model adopted for the development of the HGRD. Specifically, it was argued that the Act on a Health Sector Database violated Article 71 of the Icelandic Constitution.6 Though the court found that the presumed consent was not, in itself, unconstitutional, it concluded that the legislation must contain provisions for ensuring that an appropriate level of confidentiality would be maintained.

While not all jurisdictions have similar constitutional provisions,7 the basic tenor of Iceland’s constitution can also be found in international instruments – such as Article 12 of the Universal Declaration of Human Rights,8 which protects against “arbitrary interference with [one’s] privacy, family, home or correspondence”. At a minimum, such provisions should remind policy makers to carefully balance privacy interests against research needs and the public good. This balance must be consistent with human rights norms and regional constitutional law.

Existing legal frameworks relevant to consent and privacy law may also create challenges for the implementation of other HGRD policies. For example, many countries have a number of national and regional privacy laws and policies that must be satisfied before a large-scale HGRD can operate. These may include national research ethics policies and guidelines, professional standards, privacy legislation, human tissue laws, and, in some jurisdictions, judicially-created consent laws.9 To complicate matters further, countries such as Canada and Australia have a great deal of variation among regional jurisdictions (i.e. provinces and states) in the relevant consent and privacy laws. As such, the development of a harmonised national genetic databank policy may prove difficult.

7.1.5 International harmonisation

Variations in regional approaches also exist at the international level. And, as international co-operation on large-scale data banks becomes more common, differences between relevant national laws and research policies will become more problematic. There was much discussion about the possible need for the harmonisation of research ethics review processes, relevant terminology (e.g. anonymous, linked and linkable), and, where possible, privacy and consent rules.

Though international harmonisation may be a laudable, long-term goal, it was seen as premature. Differences in relevant domestic laws, specific cultural views and approaches to funding and organisational structures make true harmonisation unattainable, at the current time. However, increased collaboration seems inevitable and desirable. As a starting point, there should be a move towards the development of broad, agreed upon, principles.10

7.1.6 Protection of identifiable information

The protection of identifiable health information lies at the heart of many of the social concerns associated with HGRDs. A number of strategies were suggested that would help to minimise the disclosure of identifying information, thus lessening privacy concerns



(e.g. limited data release of identifiable information, “data scrubbing”, limiting data access to aggregate information, and developing schemes, such as encryption, to increase data security).11 While such strategies should be implemented, where appropriate, there is also the recognition that the more stringent the privacy protection, the more costly and potentially inefficient the research undertaking. As a result, there must be a careful and explicit weighing of the costs and impact of privacy strategies. This should be done prior to the implementation of the HGRD and should inform both the individual consent and community engagement process.

Despite reservations by the research community about the adverse impact of burdensome privacy policies on the research process, there was agreement that the maintenance of public trust demanded thorough and thoughtful privacy policies. It was also agreed, at least in relation to large public HGRDs, that release of identifiable health information should be restricted to the least amount necessary to achieve the research objective. The onus should be on researchers to justify the use of non-aggregate personal information. Such an approach is consistent with emerging international and national privacy norms.12

7.1.7 Linkability

There was much discussion about the implications of linking genetic data to other sensitive personal information. As noted above, it is the ability to link to other sources of data that makes HGRDs a tremendously powerful research tool and yet heightens privacy and confidentiality issues. The more data that are cross referenced – such as geographic location, date of birth, disease population – the easier it is to identify individual participants. The linkability of the data also has implications for the ability of participants to withdraw their consent and the obligation of researchers to re-contact patients with clinically significant information discovered in the course of the research. Once the data are in a truly aggregate and untraceable form, withdrawal becomes impossible. All the issues associated with linkability should form part of the initial consent process. For example, the type of information researchers will have access to, the privacy issues associated with linkable data, and the nature and scope of the right to withdraw should be explained.

7.1.8 Revisiting basic principles

One fundamental dilemma: Can one obtain truly informed consent when the future uses of genetic information are unknown at the time of collection? This has been a core issue in the development of national and international policy in this area. For example, the UK Human Genetics Commission noted that “the difficulties involved in tracing and securing re-consent for different forms of medical research may make obtaining fresh consent impractical and would seriously limit the usefulness of large-scale population databases”.13 The use of a “blanket consent” or “broad consent” is probably the most commonly suggested strategy for dealing with this policy issue. However, in many jurisdictions, the application of existing consent law and research ethics policies makes the legitimacy of such an approach uncertain, a point noted in the recent World Health Organisation (WHO) report which in fact concluded that “[b]lanket consent for future research is only permissible in circumstances where anonymity of future data can be guaranteed”.14

Rather than revisit this issue, workshop participants felt the consent dilemma served as an example of the need to reconsider many of the relevant and basic ethical and legal



norms. For instance, is the traditional autonomy-driven view of consent appropriate in the context of HGRDs?15 Are there other, novel approaches that could be adopted that would satisfy broad social goals and address concerns about HGRDs?16 There are cultural variations and differences in legal traditions that would undoubtedly cause societies to take different approaches and reach different conclusions. HGRDs stand as an example of the ways in which scientific and technological advances challenge existing legal and ethical frameworks and presumptions.

7.1.9 Commercialisation policy

It was recognised that there is a tension – or, at least, a perceived tension – between free public access and commercial exploitation. It was also recognised that there is some evidence that the public may feel less comfortable with a commercially funded and controlled HGRD. However, industry involvement in HGRDs is not inherently unethical or problematic and, indeed, is likely to be required if the benefits of genetic research are to be produced and disseminated. What is needed, then, are policies that allow for responsible and ethically-sound commercialisation while, at the same time, guarding against harm, whether real or perceived.

The issues of ownership and commercialisation of HGRDs have been handled differently by the projects examined. Other commercialisation themes that garnered a degree of support include the following:

� Intellectual property rights should be allowable for inventions resulting from research on the HGRD (with large public databases, intellectual property terms should be negotiated through the public entity).17

� Commercial use should not be allowed to adversely and inappropriately inhibit access.

� Commercial use should not inhibit the publication of both positive and negative results (indeed, publication should be a requirement).

� All HGRDs should have a well-developed and explicit benefit-sharing policy, and the concept of “benefit” should be broadly defined.

� Clear policies should be developed for the possible termination of the HGRD.

Finally, it was noted that all of the above issues should be canvassed and publicly debated prior to the development and implementation of a large-scale HGRD. This is particularly so given the high degree of sensitivity that surrounds many of these endeavours.

7.2 Future areas of work

While numerous issues are examined in this volume, others would benefit from additional research and consideration. They were not part of the agenda for the Tokyo Workshop. Below is a brief list of some of these issues:

� Pharmacogenetics – The potential that pharmacogenetics holds for personalised medicine, with reduced or eliminated adverse drug reaction, is a challenge. However, with respect to population genetics database, pharmacogenetics raises numerous issues, including with respect to the manner of collection and storage of samples for clinical trials.



� Use of HGRDs for developing populations – As science’s knowledge and understanding of population genetics and genetic variations among populations and sub-populations increases, the utility of HGRDs for the pharmaceutical and biotechnology industries will increase. Greater focus on and study of populations in developing countries will increasingly be of utility for the life sciences as this may, for example, highlight genetic components that are particular to certain sub-populations.

� New model of consent – One of the biggest challenges currently facing HGRDs is the manner in which informed consent is to be obtained from participants. Most HGRDs are opting for a cross between the detailed informed consent currently used in the scientific and medical communities and blanket consent. However, some have raised the consideration that the current model for informed consent may not be appropriate or applicable for HGRDs. Work on development of a new model of consent would be beneficial for many current and prospective HGRDs.

� International harmonisation – There is considerable variation at the international level. As international co-operation on large-scale databanks becomes more common, the difference between national laws and research policies will become more problematic. There is a need for harmonisation in approaches, especially for research ethics review processes, terminology, and where possible, privacy and consent rules.



Notes

1. The Rapporteur for the OECD Workshop was Professor Tim Caulfield, University of Alberta, Canada.

2. In part, this debate is related to the concern of “genetic essentialism”. In the context of health care this would include an inappropriate emphasis on genetics over other relevant determinants of health. See, for example, D. Wertz (1997), “Society and the Not-So-New Genetics”, Journal of Contemporary Health Law and Policy, 13: 307: “[The] essentialistic view pervades popular culture. ‘I am my genes’ is a phrase constantly used by questioners at public forums, despite the efforts of panellists to try to explain that ‘you’ are not the same as ‘your genes’ ”. See also Dorothy Nelkin (2001), “Molecular Metaphors: The Gene in Popular Discourse” Nature Review of Genetics, 2(7): 555.

3. The conflict between the evidence regarding the public’s perception of risk and the high levels of public participation was noted at the workshop.

4. See, for example, J. Merz (1997), “Psychosocial Risks of Storing and Using Human Tissue in Research”, Risk: Health, Safety and Environment, 8: 235; M. Robling et al. (2004), “Public Attitudes Toward the Use of Primary Patient Data in Medical Research Without Consent: A Qualitative Study”, Journal of Medical Ethics, 30: 104.

5. See, for example, Nuffield Council on Bioethics (2002), “Genetics and Human Behaviour: the ethical context”, London: Nuffield Council on Bioethics; L.B. Andrews (1997), “Past as Prologue: Sobering Thoughts on Genetic Enthusiasm”, Seton Hall Law Review, Vol. 27, p. 893; T. Caulfield (2000), “Underwhelmed: Hyperbole, Regulatory Policy and the Genetic Revolution”, McGill Law Journal, Vol. 45, pp. 437-460.

6. Article 71 reads as follows: “(1) Everyone shall enjoy freedom from interference with privacy, home and family life ... (3) Notwithstanding the provision of the first paragraph above, freedom from interference with privacy, home and family life May be otherwise limited by statutory provisions if this is urgently necessary for the protection of the rights of others.” As reported by Prof. Pall Hreinsson, “Human Genetic Research Databases”, OECD Workshop on Human Genetic Research Databases: Issues of Privacy and Security, Tokyo, 26-27 February, 2004.

7. See, however, B. von Tigerstrom, P. Nugent and V. Cosco (2000), “Alberta’s Health Information Act and the Charter: A Discussion Paper”, Health Law Review, Vol. 9, p. 3.

8. United Nations, Universal Declaration of Human Rights, adopted and proclaimed by General Assembly resolution 217 A (III) of 10 December 1948, see www.un.org/Overview/rights.html (accessed 20 September 2006).

9. Australian Law Reform Commission (2003), Essentially Yours: The Protection of Human Genetic Information in Australia, Report 96, Australian Government, ALRC, March.



10. A point that was raised during the governance session was the need to carefully balance pluralism with basic principles that have relevance for all.

11. See, for example, J. Wylie and G. Mineau (2003), “Biomedical Databases: Protecting Privacy and Promoting Research”, Trends in Biotechnology, Vol. 21, No. 3, p. 113; B. Huberman and T. Hogg (2003), “Protecting Privacy while Revealing Data”, Nature Biotechnology, Vol. 20, p. 332.

12. See, for example, sections 57 and 58 of Alberta’s Health Information Act. Section 57 requires health information custodians to collect, use and disclose health information in the most anonymous form possible. Section 58 requires custodians to disclose only the amount of information necessary to carry out the intended purpose.

13. Human Genetics Commission (2002). “Inside Information: Balancing interests in the use of personal genetic data”, Human Genetics Commission, London, May, www.hgc.gov.uk/Client/document.asp?DocId=19&CAtegoryId=8 (accessed 18 May 2006).

14. World Health Organisation, Genetic Databases – Assessing the Benefits and the Impact of Human & Patient Rights, Recommendation 9, p. 14.

15. A. Chadwick and K. Berg (2001), “Solidarity and Equity: New Ethical Frameworks for Genetic Databases”, Nat. Review of Genetics, Vol. 2, p. 318.

16. Other basic principles that were touched on at the OECD Workshop included the role of “public good” in the assessment of the ethical appropriateness of a given research protocol. It was noted that public good is often explicitly excluded as a factor in the ethics review process. Should this always be the case?

17. Though these issues surrounding patenting and human genetic material were raised several times at the workshop, this was not meant to be a focal point of the discussion. However, the OECD has other relevant work. See, for example, OECD (2002), Genetic Inventions, Intellectual Property Rights and Licensing: Evidence and Policies and OECD (2006) Guidelines for the Licensing of Genetic Inventions.

GLOSSARY – 141


Glossary

The following definitions are provided for ease of reference:

Biobank – also referred to as “HGRD” or “population database” – a collection of biological material and the associated data and information stored in an organised system, for a population or a large subset of a population.

Biological sample – also referred to as “sample” – a biological specimen including, for example, blood, tissue, urine, etc. taken from a participant.

Coded samples – sometimes also referred to as “reversibly anonymised”, “linkable” or “identifiable” samples – are those from identified sources, but the samples do not include any identifying information, such as patients’ names or social security numbers. Rather, they are accompanied with codes. In such cases, although the repository (or its agent) retains the ability to link the research findings derived from a sample with the individual source by using the code, the investigator or researcher (or one reading a description of the research findings) would not be able to do so.

Consent – also referred to as “informed consent” – A process by which information concerning the donation process is presented to the donor or the donor’s representative with an opportunity for them to ask questions, after which approval is documented.

Human genetic research database(s) (HGRDs) – also referred to as “biobank” or “population database” – a collection of biological material and the associated data and information stored in an organised system, for a population or a large subset of a population.

Identified samples – are specimens supplied from identified sources with personal identifiers (such as names or patient numbers) which are sufficient to allow the researcher to link directly the biological information and data with the individual from whom the material was obtained.

Intellectual Property Rights (IPRs) - refers to creations of the mind: inventions, literary and artistic works, and symbols, names, images, and designs used in commerce. Intellectual property is divided into two categories: (1) Industrial property, which includes inventions (patents), trademarks, industrial designs, and geographic indications of source; and (2) Copyright, which includes literary and artistic works such as novels, poems and plays, films, musical works, artistic works such as drawings, paintings, photographs and sculptures, and architectural designs. Rights related to copyright include those of performing artists in their performances, producers of phonograms in their recordings, and those of broadcasters in their radio and television programs.

Participant – also referred to as “donor” or “gene donor” - refers to the research participant or their representative. They are the individual who provides the data, information and biological sample.

142 – GLOSSARY


Population database – also referred to as “HGRD” or “biobank” – a collection of biological material and the associated data and information stored in an organised system, for a population or a large subset of a population.

Sample – also referred to as “biological sample” – a biological specimen including, for example, blood, tissue, urine, etc. taken from a participant.

Unidentified samples – sometimes also termed “anonymous”– are those from an unidentified collection of human biological specimens.

Unlinked samples – sometimes also termed “anonymised”– are those that lack identifiers or codes that can link samples to identified specimens or particular individuals. Typically, unlinked samples are sent from identified human biological specimens to investigators without identifiers or codes so that identifying particular individuals through the clinical or demographic information that is supplied with the sample or biological information derived from the research would be extremely difficult for the investigator, the repository or a third party. Unlinked samples also include samples that are already in an investigator’s possession and whose identifiers have been removed by a disinterested party.

AGENDA – 143


Agenda

Human Genetic Research Databases: Issues of Privacy and Security

26 & 27 February 2004 Tokyo, Japan

Workshop Focus

The Workshop will focus on genetic databases that contain human genetic and genomic information collected for research purposes. In this context a database is a collection of data that are arranged in a systematic way so as to be searchable.

According to the Hugo Ethics Committee, genomic data can include inter alia, nucleic acid and protein sequence variants (including neutral polymorphisms, susceptibility alleles to various phenotypes, pathogenic mutations), and polymorphic haplotypes. The work associated with a database includes collecting, annotating, curating, storing, validating and preparing specific sets for transmission.1

Objectives

The Workshop will explore the privacy and security issues raised by the development of human genetic research databases in order to:

� Promote mutual understanding of current practices and learn from the experience of others.

� Recognise areas where there are differences in approach, gaps, or need for further work.

� Identify general underlying principles.

� Seek to advise on the development of acceptable standards and approaches.

� Provide advice to the OECD’s Working Party on Biotechnology about policy implications.

Outcome

An OECD report will issue based on the pre-circulated Workshop background material, discussion leader papers, written responses from discussants, and the Rapporteur’s report. This Report will be for the attention of government agencies interested in understanding how the privacy and security of Human Genetic Research Databases are presently being assured and what recommendations ensue for their policy deliberations.

1. HUGO Ethics Committee, Statement on Human Genomic Databases, December 2002.

http://www.gene.ucl.ac.uk/hugo/HEC_Dec02.html

144 – AGENDA


DAY 1 — THURSDAY, 26 FEBRUARY 2004

Welcoming Comments

Takefumi FUKUMIZU, Director-General for Manufacturing Industries Policy Manufacturing Industries Bureau Ministry of Economy, Trade and Industry, Japan

Takayuki MATSUO, Director, Science Technology and Industry, OECD, France

Session 1a: Workshop Objectives

Yoshinao KATSUMATA, Professor, Nagoya University Graduate School of Medicine, Japan

Session 1b: Existing HGRDs – The Context

Bartha Maria KNOPPERS, Professor, Faculty of Law, Université de Montréal, Canada

Issues: HGRD objectives populations, typology, status; linkages to other databases; sponsors, funding; public consultation and confidence; adequate consent mechanisms; remuneration of sample/information donors; common trends and divergences among HGRDs.

Session 2: Confidentiality and Data Management

Chair David RIMOIN, Professor, Dept. of Pediatrics, Cedars-Sinai Medical Center, United States

Discussion leader Mildred CHO, Associate Director, Stanford Centre for Biomedical Ethics, United States

Commentators Ryuichi IDA, Professor, Graduate School of Law, Kyoto University, Japan

Páll HREINSSON, Professor, University of Iceland and Chair, Icelandic Privacy and Data Protection Authority, Iceland

Issues: Definitions of anonymisation; levels of anonymisation, coding and linkages with other personal information; “appropriate” security measures and software mechanisms; quality assurance for confidentiality; confidentiality issues related to sharing of data; feedback of information to research subjects; confidentiality and changes in the status or purpose of the HGRD (e.g. termination); international issues/options.

AGENDA – 145


Session 3: Access to Databases for Research Purposes

Chair Alan GUTTMACHER, Deputy Director for the National Human Genome Research Institute, NIH, United States

Discussion leader Richard COTTON, Director, Genomic Disorders Research Centre, Australia

Commentators Vilhjálmur ARNASON, Chair of the Centre for Ethics, University of Iceland, Iceland

H.-Erich WICHMANN, Professor, and Director, GSF – Institute of Epidemiology, Germany

Yusuke NAKAMURA, Director, Human Genome Centre Laboratory of Molecular Medicine, Institute of Medical Science, University of Tokyo, Japan

Issues: Criteria for access for private and public researchers (scientific merit, ethics of proposed use, assurances of confidentiality, fees); control of data and research results; quality assurance, validation and accreditation of research results; feedback of research results to the database; international issues/options.

DAY 2 — FRIDAY, 27 FEBRUARY 2004

Session 4: Ownership and Commercialisation of Database

Chair Kevin CHEESEMAN, Director, Development Pharmacogenetics, AstraZeneca R&D, United Kingdom

Discussion leader Jasper BOVENBERG, University of Leiden, The Netherlands

Commentators Hiroshi GUSHIMA, Biofrontier Partners, Japan

Jaanus PIKANI, Chairman of the Board, Estonian Genome Foundation, Estonia

Issues: For profit, not-for-profit, and mixed databases; ownership of data and research results; protection of intellectual property/copyright; licensing of commercial opportunities; business models; benefit sharing; issues surrounding the change in purpose or sale/transfer of database; international issues/options.

146 – AGENDA


Session 5: HGRD Governance

Chair William LOWRANCE, Chair of the Interim Advisory Group on Ethics and Governance of the UK Biobank project, and Consultant in Health Policy and Ethics in Geneva, Switzerland.

Discussion leader Ruth CHADWICK, Professor, University of Lancashire, United Kingdom

Commentators David WEISBROT, President, Australia Law Reform Commission, Australia

Nicole QUESTIAUX, Vice Présidente du CCNE, France

Issues: Governance mechanisms; role of legislation, regulation, research ethics boards, central authorities; typology of governance structures; roles and mandates; composition of members; funding; powers, compliance, enforcement, sanctions; reporting requirements; assessment of security mechanisms; monitoring and accountability; international issues/options.

Session 6: The Way Forward

Rapporteur Rapporteur Recap of Key Issues

Tim CAULFIELD, Professor, University of Alberta, Canada

All Chairs Panel on Emerging Themes and Policy Needs

Next steps for the OECD

OECD Secretariat Closing Remarks

Iain GILLESPIE, Head, Biotechnology Division, OECD

LIST OF PARTICIPANTS – 147


Human Genetic Research Databases: Issues of Privacy and Security

26 & 27 February 2004 Tokyo, Japan

List of Participants

Mr. Vilhjálmur ARNASON University of Iceland Reykjavik, Iceland Mr. Andrea BOGGIO University of Geneva/World Health Organization Geneva, Switzerland Ms. Tonje BORCH Norwegian Health Directorate Oslo, Norway Mr. Jasper BOVENBERG Universiteit Leiden Leiden, Netherlands Mr. Timothy CAULFIELD University of Alberta Edmonton, Alberta, Canada Ms. Ruth CHADWICK Lancaster University Lancaster, United Kingdom Mr. Kevin CHEESEMAN AstraZeneca R&D London, United Kingdom Mr. Osamu CHISAKI Japan BioIndustry Association Tokyo, Japan Ms. Mildred CHO Stanford University Palo Alto, CA, United States Mr. Richard COTTON Genomic Disorders Research Centre Victoria, Australia Mr. Yukihiro EGUCHI Mitsui Knowledge Industry Co., Ltd Tokyo, Japan

148 – LIST OF PARTICIPANTS


Mr. Motoi ETO Ministry of Education, Culture, Sports, Science and TechnoTokyo, Japan Mr. Pierrick FILLON-ASHIDA Delegation of the European Commission in Japan Tokyo, Japan Mr. Takefumi FUKUMIZU Ministry of Economy, Trade and Industry Tokyo, Japan Mr. SEIGO FUKUMOTO Ministry of Economy, Trade and Industry Tokyo, Japan Mr. Iain M. GILLESPIE OECD Paris, France Mr. Hiroshi GUSHIMA Bio Frontier Partners Co. Ltd. Tokyo, Japan Mr. Alan GUTTMACHER National Human Genome Research Institute Bethesda, MD, United States Ms. Lea HARTY Pfizer Pharmaceuticals Inc. Groton, CT, United States Mr. Alexander HASLBERGER Federal Ministry for Health and Women’s Issues Vienna, Austria Mr. Tadashi HIRAKAWA Japan Bioindustry Association Tokyo, Japan Mr. Toshiyuki HIRAKAWA Japan Biological Informatics Consortium Tokyo, Japan Mr. Pall HREINSSON University of Iceland Reykjavik, Iceland Ms. Yoshiko ICHIHARA Ministry of Education, Culture, Sports, Science and Technology Tokyo, Japan Mr. Ryuichi IDA Kyoto University Kyoto, Japan Ms. Sigrun JOHANNESDOTTIR PERSONUVERND Reykjavik, Iceland



Mr. Richard JOHNSON Arnold & Porter Washington, DC, United States Mr. Hiroji KATO Ministry of Economy, Trade and Industry Tokyo, Japan Mr. Yoshinao KATSUMATA Nagoya University Graduate School of Medicine Nagoya, Japan Ms. Bartha Maria KNOPPERS Faculty of Law, Université de Montréal Montreal, Canada Ms. Lise LAVOIE Consultant Ottawa, Ontario, Canada Mr. William W. LOWRANCE Consultant in Health Policy Geneva, Switzerland Mr. Tohru MASUI National Institute of Health Sciences Tokyo, Japan Mr. Daisuke MATSUDA Ministry of Economy, Trade and Industry Tokyo, Japan Mr. Takayuki MATSUO OECD Paris, France Ms. Kaori MUTO Shinshu University Nagano, Japan Mr. Yusuke NAKAMURA Human Genome Center, University of Tokyo Tokyo, Japan Mr. Jaanus PIKANI Estonian Genome Foundation Tartu, Estonia

Ms. Nicole QUESTIAUX CCNE Neuilly-sur-Seine, France Mr. David RIMOIN Cedars-Sinai Medical Center Los Angeles, CA, United States Mr. Stéphane ROY Ambassade de France à Tokyo Tokyo, Japan

150 – LIST OF PARTICIPANTS


Mr. Takanori SAKAMOTO Ministry of Economy, Trade and Industry Tokyo, Japan Mr. Maurizio SALVI European Commission Brussels, Belgium Mr. Youichi SATOU Ministry of Health, Labour and Welfare Tokyo, Japan Ms. Gabriele SATZINGER Federal Ministry for Health and Women’s Issues Vienna, Austria Mr. Yoji SHIMADA Ministry of Education, Culture, Sports, Science and Technology Tokyo, Japan Mr. Takeshi SHIMANO Ministry of Economy, Trade and Industry Tokyo, Japan Mr. Nigel SKIPPER Health Canada Ottawa, Canada Mr. Seizo SUMIDA Japan BioIndustry Association Tokyo, Japan Ms. Joanne SUMNER The Wellcome Trust London, United Kingdom Ms. Yuka SUZUKI Ministry of Education, Culture, Sports, Science and Technology Tokyo, Japan Mr. Toshihito TAKASHIBA Attorney at Law Tokyo, Japan Mr. Kenji TAKEZAWA OECD Paris, France Mr. Keiji TAKITA Ministry of Economy, Trade and Industry Tokyo, Japan Ms. Akiko TAMAKOSHI Nagoya University Graduate School of Medicine Nagoya, Japan



Mr. Dai TOYAMA Ministry of Education, Culture, Sports Science and Technology (MEXT) Tokyo, Japan Mr. Kazuhiko UCHIDA University of Tsukuba Ibaraki, Japan Ms. Kaori UEMURA Japan BioIndustry Association Tokyo, Japan Ms. Helen WALLACE GeneWatch UK Buxton, Derbyshire, United Kingdom Ms. Caroline WEBER Health Canada Ottawa, Canada Mr. David WEISBROT Australian Law Reform Commission Sydney, Australia Mr. Erich WICHMANN GSF – Institute of Epidemiology Neuherberg, Germany Ms. Lorna WILSON OECD Paris, France Mr. Yoshiyasu YABUSAKI Japan Bioindustry Association Tokyo, Japan Mr. Hiroshi YOSHIKURA Ministry of Health, Labour and Welfare Tokyo, Japan

BIBLIOGRAPHY – 153


Bibliography

Legislation and Instruments

Alberta, Health Information Act, R.S.A. 2000, c. H-5, entered into force 25 April 2001.

CARTaGENE, Information and Consent Form, August 2006, www.cartagene.qc.ca/docs/InfoConsent_En.pdf.

Council for International Organisations of Medical Sciences (CIOMS), International Ethical Guidelines for Biomedical Research Involving Human Subjects, November 2002.

Council of Europe, Convention for the protection of Human Rights and Dignity of the Human Being with regard to the application of biology and medicine: Convention on Human Rights and Biomedicine, CETS No. 164, entered into force 1 December 1999.

Estonia, Human Genes Research Act, RT I 2000, 104, 685, entered into force 8 January 2001.

European Union, Directive on the Legal Protection of Databases (96/9/EC).

HUGO Ethics Committee, Statement on Human Genomic Databases, adopted December 2002.

Iceland, Act on Biobanks No. 110/2000, entered into force 1 January 2001.

Iceland, Act on a Health Sector Database, No 139/1998, entered into force 17 December 1998.

Iceland, Government Regulation on a Health Sector Database, No. 32/2000.

Iceland, Regulations on the Keeping and Utilisation of Biological Samples in Biobank, No. 134/2001.

Iceland, Regulation on Scientific Research in the Health Sector, No. 552/1999.

Latvia, Human Genome Research Law, 2002, entered into force 1 January 2003.

Nuremberg Code (1949), “From Trials of War Criminals before the Nuremberg Military Tribunals under Control, Council Law No. 10”, Vol. 2, US Government Printing Office, Washington, DC, pp. 181-182.

RMGA, Statement of Principles: Human Genome Research (VERSION 2000).

RMGA, Statement of Principles on the Ethical Conduct of Human Genetic Research Involving Populations, 2003.

UK Biobank, Ethics and Governance Framework, May 2006.

UK Biobank, Policy on Intellectual Property and Access, Draft 1 January 2005.

154 – BIBLIOGRAPHY


United Nations, Convention of the Law and Sea, adopted 10 December 1982, entered into force 16 November 1994.

United Nations, Universal Declaration of Human Rights, General Assembly Resolution 217 A (III) of 10 December 1948.

UNESCO, International Declaration on Human Genetic Data, adopted 16 October 2003.

United Nations Environment Programme, Convention on Biological Diversity, entered into force 29 December 1993.

United States, Health Insurance Portability and Accountability Act, 1996 (45 C.F.R. 164.514(b)).

United States, Department of Health and Human Services. Federal Guidelines for Research Involving Human Subjects, Code of Federal Regulations 1991(45 CFR 46).

World Medical Association, Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects, Adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964; last amended by the WMA General Assembly, Tokyo, 2004.

Case law

Pate v Threlkel, (1994) 640 So2d 183.

Ragnhildur Guomundsdottir v. The State of Iceland, Icelandic Supreme Court, No. 151/2003.

Safer v. Estate of Pack, (1996) 677 A2d 1188.

Databases

CARTaGENE, www.cartagene.qc.ca.

deCODE Genetics, http://www.decode.com/

Estonian Genome Project, http://www.geenivaramu.ee/index.php?show=main&lang=eng

Genome Database of the Latvian Population, www.bmc.biomed.lu.lv/gene/

GenomeEUtwin, http://www.genomeutwin.org/

Icelandic Health Sector Database, http://eng.heilbrigdisraduneyti.is/laws-and-regulations/

Marshfield Clinic research Foundation, Personalized Medicine Research Project, www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pers_med_res_prj

Public Population Projects in Genomics (P3G), www.p3gconsortium.org/.

Translation-Genomic Research in the African Diaspora (TgRIAD), www.genomecenter.howard.edu/TGRIAD.htm

UK Biobank, www.ukbiobank.ac.uk/



Monographs and Articles

Allen, A. (1997), “”Genetic Privacy: Emerging Concepts and Values”, in M. Rothstein (ed.), Genetic Secrets: Protecting Privacy and Confidentiality in the Genetic Era, Yale University Press,

Andrews, L.B. (1997), “Past as Prologue: Sobering Thoughts on Genetic Enthusiasm”, Seton Hall Law Review, Vol. 27.

Annas, G., L. Glantz and P. Roche (1995), “Drafting the Genetic Privacy Act: Science, Policy and Practical Considerations”, Journal of Law, Medicine and Ethics, Vol. 23(4):360-366.

Árnason, Vilhjálmur (2003), “Comment on Access to Databases for Research Purposes”, Background Paper for the Tokyo Workshop.

Australian Law Reform Commission (2003), Essentially Yours: The Protection of Human Genetic Information in Australia, Report 96, Australian Government, ALRC, March.

Bovenberg, J.A. (2006), “Establishment, Management and Governance of Human Genetic Research Databases”, study commissioned by the OECD.

Bovenberg, Jasper A. (2003), “Ownership and Commercialisation of Human Genetic Research”, Background Paper for the Tokyo Workshop.

CARTaGENE, Atelier II, Montreal, 11 June, www.cartagene.qc.ca/atelierII/Questions.htm.

Caulfield, T. (2000), “Underwhelmed: Hyperbole, Regulatory Policy and the Genetic Revolution”, McGill Law Journal, Vol. 45.

Chadwick, A. and K. Berg (2001), “Solidarity and Equity: New Ethical Frameworks for Genetic Databases”, Nat. Review of Genetics, Vol. 2.

Chadwick, Ruth, Minakshi Bhardwaj, Anthony Mark Cutter and Sarah Wilson (2003), “HGRD Governance”, Background Paper for the Tokyo Workshop.

Cho, Mildred (2003), “Confidentiality and Data Management”, Background Paper for the Tokyo Workshop.

Cotton, Richard G.H. and Ourania Horaitis (2003), “Access to Mutation Databases for Research”, Background Paper for the Tokyo Workshop.

Deschênes, M. and G. Cardinal (2003), “CARTaGENE Project: Governance and Ethical Clearance”, Centre de recherche en droit public, Université de Montréal.

Deschênes, Mylène and Geneviève Cardinal (2003), “Surveying the Population Biobankers”, in Bartha Maria Knoppers (ed.), Population and Genetics: Legal and Socio-Ethical Perspectives, Martinus Nijhoff, Leiden.

Deutsche Forschungsgemeinschaft, Senate Commission on Genetic Research, Predictive Genetic Diagnosis – Scientific Background, Practical and Social Implementation, Bonn, 27 March 2003, www.dfg.de/aktuelles_presse/reden_stellungnahmen/2003/download/predictive_genetic_diagnosis.pdf.



Estonia, Gene Donor Consent Form, www.geenivaramu.ee/index.php?lang=eng&sub=74.

Estonia, Genome Project Ends Cooperation with Current Financier, 27.12.2004, www.geenivaramu.ee/index.php?lang=eng&show=uudised&id=172&PHPSESSID=197f1b8a33cef08aa94f377cb65cf596.

Estonia Genome Project (EGP) (2004), “Genome Project Ends Cooperation with Current Financier”, 27 December 2004, www.geenivaramu.ee/index.php?lang=eng&show=uudised&id=172&PHPSESSID=197f1b8a33cef08aa94f377cb65cf596.

European Commission (2000), “A New Initiative on “Genome Research for Human Health” Press Release, Brussels, 15 November, http://europa.eu.int/comm/research/press/2000/pr1511en.html.

European Society of Human Genetics (2002), “Eurogapp Project, 1999-2000”, Vienna.

Faden, R.R. and T.L. Beauchamp (1986), A History and Theory of Informed Consent, Oxford University Press, New York.

Godard, B. and E. Racine (2003), “La stratégie de communication de CARTaGENE”, CARTaGENE, Atelier II, Montreal, 11 June, www.cartagene.qc.ca/atelierII/Questions.htm.

Greely, H. (1999), “Breaking the Stalemate: A Prospective Regulatory Framework for Unforeseen Research Uses of Human Tissue Samples and Health Information”, Wake Forest Law Review, 34, pp. 737-66.

Gushima, Hiroshi (2003), “Comment on Ownership and Commercialisation of Database”, Background Paper for the Tokyo Workshop.

Hreinsson, Páll (2003), “Comment on Confidentiality and Data Management”, Background Paper for the Tokyo Workshop.

Huberman, B. and T. Hogg (2002), “Protecting Privacy while Revealing Data”, Nature Biotechnology, Vol. 20, No. 4.

Human Genetics Commission (2002). “Inside Information: Balancing Interests in the Use of Personal Genetic Data”, Human Genetics Commission, London, May, www.hgc.gov.uk/Client/document.asp?DocId=19&CAtegoryId=8.

Katsumata, Yoshinao (2003), “Workshop Objectives”, Background Paper for the Tokyo Workshop.

Ida, Ryuichi (2003), “Comment on Confidentiality and Data Management”, Background Paper for the Tokyo Workshop.

Laberge, C. et al. (2001), “Formal Application to Genome Quebec”, Newsletter – Map of Genetic Variation in the Quebec Population, October 15, Vol. 1.

Laberge, Claude (2003), “Update of the CARTaGENE Project”, Montreal, http://www.cartagene.qc.ca/historique.cfm.

Lin, Zhen, Teri Klein and Russ Altman (2001), “Methodology on Scrubbing Data in PharmGKB”, paper presented at the Pacific Symposium on Biocomputing, Kohala Coast, Hawaii.



Lowrance, W.W. (2002), “Learning from Experience: Privacy and the Secondary Use of Data in Health Research”, The Nuffield Trust, London.

Malin, B. and L. Sweeney (2000), “Determining the Identifiability of DNA Database Entries”, Proceedings, Journal of the American Medical Informatics Association. pp. 537-541.

Malin, B. and L. Sweeney (2001), “Re-identification of DNA through an Automated Linkage Process”, Proceedings/AMIA Annual Symposium, Hanley & Belfus, Inc., Washington, DC, pp. 423-427.

Marshfield Clinic Research Foundation, “Personalized Medicine Research Project – Frequently Asked Questions”, www.marshfieldclinic.org/chg/pages/default.aspx?page=chg_pmrp_faqs.

McCarthy, Cathy (2003), “The Personalized Medicine Research Project”, presented at the North-American Stakeholders Meeting, Montreal, 26 August.

Medical Research Council (2001), “Human Tissue and Biological Samples for Use in Research”, London, April, www.mrc.ac.uk/pdf-tissue_guide_fin.pdf.

Medical Research Council and The Wellcome Trust (2000), “Public Perceptions of the Collection of Human Samples”, London, October, at www.ukbiobank.ac.uk/docs/perceptions.pdf.

Merz, J. (1997), “Psychosocial Risks of Storing and Using Human Tissue in Research”, Risk: Health, Safety and Environment, Vol. 8, pp 235-248.

Merz, J.F., P. Sankar, S.E Taube and V. Livolsi (1997), “Use of Human Tissues in Research: Clarifying Clinician and Researcher Roles and Information Flows”, Journal of Investigative Medicine, Vol. 45, No. 5, pp. 252-257.

Murray, T. (1997), “Genetic Exceptionalism and ‘Future Diaries’: Is Genetic Information Different From Other Medical Information?” in M. Rothstein (ed.), Genetic Secrets: Protecting Privacy and Confidentiality in the Genetic Era, Yale University Press, New Haven, Connecticut.

Nakamura, Yusuke (2003), “Comment on Access to Databases for Research Purposes”, Background Paper for the Tokyo Workshop.

National Bioethics Advisory Commission (1999), Human Biological Materials: Ethical Issues and Policy Guidance, Vol. 1, Report and Recommendations of the National Bioethics Advisory Commission, Rockville, Maryland, August, www.georgetown.edu/research/nrcbl/nbac/hbm.pdf.

National Institutes of Health (NIH) (2005), “Points to Consider When Planning a Genetic Study that Involves Members of Named Populations” 25 January, www.nih.gov/sigs/bioethics/named_populations.html.

NIH News Advisory (2002), “Background on Ethical and Sampling Issues Raised by the International HapMap Project”, Release Genetic Variation Mapping Launch, October, http://genome.gov/10005337.

Nelkin, Dorothy (2001), “Molecular Metaphors: The Gene in Popular Discourse” Nature Review of Genetics, July, Vol. 2, No. 7, pp. 555-559.

NHGRI News Release (2002), “Background on Ethical and Sampling Issues Raised by the International HapMap Project”, October 2002, www.genome.gov/10005337.



NIGMS Human Genetic Cell Repository (2004), “Policy for the Responsible Collection, Storage, and Research Use of Samples from Identified Populations”, 25 August, http://locus.umdnj.edu/nigms/comm/submit/collpolicy.html.

Nuffield Council on Bioethics (2002), “Genetics and Human Behaviour: The Ethical Context”, Nuffield Council on Bioethics, London.

OECD (2002), Genetic Inventions, Intellectual Property Rights and Licensing: Evidence and Policies, OECD, Paris.

P3G Montreal Meeting, Montreal Meeting, Executive Summary (2003), www.p3gconsortium.org/events/2003Montreal/ExecutiveSummary.pdf.

P3G Montreal Meeting, Workshop Minutes (2003), “Working Groups 5: Public Engagement”, Montreal Meeting 2-4 July. www.p3gconsortium.org/events/2003Montreal/ExecutiveSummary.pdf.

Pikani, Jaanus (2003), “Comment on Ownership and Commercialisation of Human Genetic Research”, Background Paper for the Tokyo Workshop.

Questiaux, Nicole (2003), “Comment on HGRD Governance”, Background Paper for the Tokyo Workshop.

Robling, M. et al. (2004), “Public Attitudes Toward the Use of Primary Patient Data in Medical Research Without Consent: A Qualitative Study”, Journal of Medical Ethics, Vol. 30, pp.104-109.

Rothstein, M.A. (2002), “The Role of IRBs in Research Involving Commercial Biobanks”, Journal of Law, Medicine and Ethics, Vol. 30, No. 1, pp. 105-108.

Sallée, Clementine (2003), “Existing Human Genetic Research Databases: Context (Consent Mechanisms and Communication Strategies)”, Background Paper for the Tokyo Workshop.

Severin, M. (1999), “Genetic Susceptibility for Specific Cancers”, Cancer, Vol. 86, pp. 2564-69.

Storms, F. (1999), “Storms Brews over Gene Bank of Estonian Population”, Science Magazine, Vol. 286, No. 5443, p. 1262, www.sciencemag.org/cgi/content/full/286/5443/1262.

Suda, E. and D. Macer (2003), “Ethical Challenges of Conducting the HapMap Genetics Project in Japan”, in S.Y. Song, Y.M. Koo and D.R.J. Macer (eds.), Bioethics in Asia in the 21st Century, Eubios Ethics Institute, Christchurch.

Sweeney, L. (1997), “Weaving Technology and Policy Together to Maintain Confidentiality”, Journal of Law, Medicine and Ethics, Vol. 25:98-110.

Sweeney, L. (1998), “Datafly: A System for Providing Anonymity in Medical Data”, in T. Lin and S. Qian (eds.), Database Security XI, Chapman and Hall, London.

Sweeney, L. (2002), “K-Anonymity: A Model for Protecting Privacy”, International Journal on Uncertainty, Fuzziness and Knowledge-based Systems, 10 (5), 557-570.

von Tigerstrom, B., P. Nugent and V. Cosco (2000), “Alberta’s Health Information Act and the Charter: A Discussion Paper”, Health Law Review, Vol. 9.

United Kingdom, Secretary of Health (2003), Our Inheritance, Our Future: Realising the Potential of Genetics in the NHS (A Genetics White Paper Presented to Parliament by



the Secretary of State for Health By Command of Her Majesty), June, www.conversations.canterbury.ac.nz/resources/genetics/uk-%20white-paper.pdf#search=%22Our%20Inheritance%2C%20Our%20Future%2C%20%22.

United States, White House, Office of the Press Secretary, Joint Statement by President Clinton and Prime Minister Tony Blair of the U.K. (2000), 14 March, http://clinton4.nara.gov/WH/EOP/OSTP/html/00314.html.

Vaszar, L.T., M.K. Cho and R.A. Raffin (2003), “Privacy Issues in Personalised Medicine”, Pharmacogenomics 4:107-112.

Weber, A. (2001), “Sold Nation”, Süddeutsche Zeitung, 23 November.

Weisbrot, David (2003), “Comment on HGRD Governance”, Background Paper for the Tokyo Workshop.

Wertz, D. (1997), “Society and the Not-So-New Genetics”, Journal of Contemporary Health Law and Policy, Vol. 13(2), pp. 299-345.

Wichmann, H.-Erich (2003), “Comment on Access to Databases for Research Purposes”, Background Paper for the Tokyo Workshop.

Wiederhold, G., M. Bilello, V. Sarathy and X. Qian (1996), “Protecting collaboration”, in Proceedings of the National Information Systems Security 1996 Conference, pp. 561-569.

Winickoff, David E. and Richard N. Winickoff (2003), “The Charitable Trust as a Model for Genomic Biobanks”, New England Journal of Medicine, Vol. 349, pp.1180-1184.

World Health Organisation, Genetic Databases – Assessing the Benefits and the Impact of Human & Patient Rights, World Health Organization, Geneva.

Wylie, J. and G. Mineau (2003), “Biomedical Databases: Protecting Privacy and Promoting Research”, Trends in Biotechnology, Vol. 21, No. 3, pp. 113-116.

Zimmerli, W. (1990), “Who Has the Right to Know the Genetic Constitution of a Particular Person?”, in R. Chadwick and G. Bock (eds.), Human Genetic Information: Science, Law and Ethics, John Wiley and Sons, Chichester.

OECD PUBLICATIONS, 2, rue André-Pascal, 75775 PARIS CEDEX 16

PRINTED IN FRANCE

(93 2006 09 1 P) ISBN 92-64-02852-8 – No. 55289 2006


Cre

atio

n a

nd

Go

ve

rna

nc

e o

f Hu

ma

n G

en

etic

Re

se

arc

h D

ata

ba

se

s



BIOTECHNOLOGY INNOVATION SCIENCE HEALTH BIOTECHNOLOGY INNOVATION SCIENCE HEALTH BIOTECHNOLOGY INNOVATION

INNOVATION HEALTH SCIENCE BIOTECHNOLOGY HEALTH INNOVATION SCIENCE BIOTECHNOLOGY HEALTH









HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH

HEALTH SCIENCE BIOTECHNOLOGY INNOVATION HEALTH BIOTECHNOLOGY













SCIENCE BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY

BIOTECHNOLOGY INNOVATION HEALTH SCIENCE BIOTECHNOLOGY INNOVATION

INNOVATION HEALTH SCIENCE BIOTECHNOLOGY

HEALTH SCIENCE BIOTECHNOLOGY INNOVATION

SCIENCE

OECDPUBLISHING

Scientists have known for years that complex diseases, including cancer, heart disease, stroke, and diabetes, arise from a complex combination of lifestyle, environmental, genetic and random factors. Large-scale study of populations may contribute significantly to science’s understanding of the complex multi-factorial basis of diseases and to improvements in prevention, detection, diagnosis, treatment and cure. As a result of developments in biotechnology and bioinformatics, the opportunity to store and analyse increasingly large amounts of genetic data have rendered possible the creation of large-scale population databases. Genetic research, involving the use of such databases containing human genetic and genomic data, information, and biological samples, is thus becoming increasingly feasible.

More recently, the databases being developed include data, information and biological samples from whole populations. Large-scale population databases which contain a significantly broader range of information about individuals also raise a number of issues and concerns. While some of these are not new, the increasing breadth and scope of such databases amplifies them. Moreover, the combination of a broader set of genetic data and personal information in these databases raises new issues about the use of such information, especially in a non-clinical or non-research context. In addition, as such databases will increasingly be international in scope and cover populations from numerous jurisdictions, new sets of questions will arise.

The OECD organised a workshop in order to begin the process of considering, at the international level, policy challenges associated with the establishment, management and governance of human genetic research databases. This report provides an overview of the complex issues that were discussed at that workshop and which need to be considered or addressed, in recognition of the significant contribution that human genetic research databases could play in translating scientific advances into innovation in health.


www.oecd.orgISBN 92-64-02852-8 93 2006 09 1 P-:HSTCQE=UW]ZWY:

The full text of this book is available on line via these links:http://www.sourceoecd.org/scienceIT/9264028528http://www.sourceoecd.org/governance/9264028528

Those with access to all OECD books on line should use this link:http://www.sourceoecd.org/9264028528

SourceOECD is the OECD’s online library of books, periodicals and statistical databases. For more information about this award-winning service and free trials ask your librarian, or write to us at [email protected].

OECDPUBLISHING

Date post:	14-Dec-2016
Category:	Documents
Upload:	lekhue
View:	212 times
Download:	0 times