Introduction

The last few years have witnessed an important expansion of human DNA sampling and data collecting in order to exploit and study the genetic information collected. The strategic importance of this activity for genetic research and its applications is obvious. Human DNA, tissue or cell collections, Guthrie cards as well as databases which are attached to such biological resources are necessary for a wide range of purposes and these collections have been extensively exchanged for scientific purposes. However, the status of collections is not very well known and most laboratories that bank DNA have no written policies or agreements regarding this activity. Still, many DNA banks are concerned about how to obtain valid informed consent, safeguard the privacy of samples and data, and avoid potential misunderstandings with depositors.1,2,3,4 Regulations pertaining to the storage of human biological materials and genetic data are at their beginning stage in most European countries but the multiplicity of actors and of rules that regulate them (public versus private, hospitals, research centers, laboratories) make the situation increasingly difficult to comprehend. For instance, the rules that regulate access are still governed by practices that vary widely with the type of collection. The type of informed consent asked for at the time of constitution of the bank may be very different from collection to collection, and makes it even more difficult to understand the condition of their use. The rules for exchange and sharing of information and material are not clear. The notion of return of benefits to the communities, which recognizes their contribution, is fairly recent.

The purpose of this document is to formulate a professional and scientific view on the social, ethical, and legal issues that impact on data storage and human DNA banking practices for biomedical research in Europe. The use of human embryos in research as well as the use of human DNA for forensic purposes raise special ethical issues and are not addressed in this document. After the methods, the second section outlines the requirements for data storage and DNA banking in the public and private sectors in Europe. The third section addresses the issues relating to DNA banking, such as the consent requirements for the banking and further uses of DNA samples, their control and ownership, and the return of benefits derived from DNA exploitation to the community. The fourth section of the document presents a series of case studies involving the use of large DNA sample collections and personal medical information. Finally, the fifth section addresses the issues debated during an international workshop which aimed to compare the different approaches to DNA banking in Europe, and to come up with practical guidelines and recommendations.

Methods

The methods used for analyzing the professional and scientific views on the social, ethical and legal issues that impact on data storage and DNA banking for biomedical research was primarily the review of the existing professional guidelines, legal frameworks and other documents related to the data storage and DNA banking practices in public and private sectors in Europe. Then, with the help of the existing guidelines and a review of literature, the method was to examine questions which need debate, in particular the consent requirements for banking and further use of samples, the control of banked samples and quality issues, the ownership of banked samples and the return of benefits to the community. The following case studies were described: the ALSPAC study, deCODE genetics and the Act on a Health Sector Database, and Guthrie cards. These questions and case studies were debated during an international workshop organized by the European Society of Human Genetics Public and Professional Policy Committee in Paris, France, April 07–08, 2000.

The purpose of the workshop was to identify, from a professional viewpoint, the most important/pressing/burning ethical issues relating to the data storage and DNA banking practices in public and private sectors in Europe. The formal workshop presentations covered the following themes: setting the scene (banking practices, policies, return of benefits to the community), point of view of data and DNA banks managers, analysis of practical examples and elaboration of statements and recommendations. Small multidisciplinary groups were convened to take these discussions further. Their initial task was to explore the practices and needs in the countries represented and to consider the extent to which these needs were currently being met. Following the small group sessions, conclusions were fed back to the whole group where there were opportunities for further discussion.

A group of 50 experts from 12 European countries was invited. These experts were representatives of the seven following sectors:

  1. 1

    Medical Genetics;

  2. 2

    Human Genetics Societies;

  3. 3

    Ethical, Legal and Social Issues;

  4. 4

    Support Groups;

  5. 5

    Biotechnology/Pharmaceutics;

  6. 6

    Insurance/Employment;

  7. 7

    European Union Institutions.

A first background document was discussed during the workshop. A second document, including discussions of the workshop, was sent for comments to representatives of the human genetic societies and European experts in the field of DNA banking for biomedical research, as well as to all ESHG members. This document was also put on the ESHG website (www.eshg.org) for public consultation and discussion. The final document was approved by the ESHG board.

Data storage and DNA banking practices in public and private sectors in Europe

In 1988, The American Society of Human Genetics (ASHG) statement on DNA Banking and DNA Analysis defined a DNA bank as ‘a facility that stores DNA for future analysis’, whereas a DNA diagnostic laboratory was ‘a facility that analyzes DNA to provide information about the diagnosis of a disease state or susceptibility thereto, about the diagnosis of a carrier state, or for identification purposes’. In 1989, the British Clinical Genetics Society stated that the purpose of a DNA bank is ‘to provide for the future requirements of families affected by serious single gene disorders and who require DNA analysis for the purpose of: (1) confirmation of diagnosis at the molecular level; (2) presymptomatic diagnosis; and (3) carrier detection’. After 10 years, the American National Bioethics Advisory Commission5 defined a DNA bank as ‘a facility that stores extracted DNA, transformed cell lines, frozen blood or other tissue, or biological materials, for future DNA analysis’. The same Commission defined a DNA databank as ‘a repository of genetic information obtained from the analysis of DNA, sometimes referred to as ‘DNA profiles’. The genetic information is usually stored in computerized form with individual identifiers’.

Ethical principles for the prospective use of human genetic material and data have been introduced in several European countries. The British Clinical Genetics Society6 and the Danish Council of Ethics7 emphasized the quality assurance of banked tissue samples and consequently the origin of samples that may be banked. In Denmark, a Data Surveillance Authority, which has jurisdiction over biobanks, was created in 1996. The Health Council of the Netherlands8 issued the recommendation that human material cannot be stored ‘without a good reason’ (1994). In France, a 1996 ordinance pursuant to the 1994 Bioethics Laws mandated that ‘no person may take samples with a view to constituting a collection of human biological specimens, or use, to this same end, samples already taken or derivatives thereof if they have not notified the competent administrative authority of the proposed collection’. In Iceland, the Biobanks Act (2000) states that ‘the interest of science and of the community shall never be given priority over the interests of the donor of a biological sample’ (Article 1). The Estonian Human Genes Research Act (2000) also insists on ‘the voluntary nature of gene donation and the confidentiality of the identity of gene donors’. Recently, the British Human Genetics commission9 suggested ‘the establishment of independent oversight bodies for all large genetic databases’.

Conditions of data storage and DNA banking

Under what conditions may samples be banked? In its statement on informed consent for genetic research (1996), the ASHG distinguished between retrospective and prospective studies: ‘Retrospective research studies utilize previously obtained samples collected for a purpose that is different from that of the project under study; prospective research studies are those in which the collection of the new samples is part of the study design’. Whatever the study, the ASHG defined four types of identification that may be employed to store the samples as the following:

  • Anonymous: ‘Biological materials that were originally collected without identifiers and are impossible to link with their sources’.

  • Anonymized: ‘Biological materials that were originally identified, but have been irreversibly stripped of all identifiers and are impossible to link to their sources’.

  • Identifiable: ‘Biological materials that are unidentified for research purposes, but can be linked to their sources through the use of a code. Decoding can only be done by the investigator or another member of the research team’.

  • Identified: ‘Biological materials to which identifiers, such as a name, patient number, or clear pedigree location, are attached and available to the researchers’.

The extent to which a research sample can be linked with the identity of its source is a significant determinant in assessing the risks and potential benefits to human subjects.5 The American Commission defined the different types of research samples as follows:

  • Unidentified: ‘Sometimes termed ‘anonymous’, these samples are supplied by repositories to investigators from a collection of unidentified human biological specimens’.

  • Unlinked: ‘Sometimes termed ‘anonymized’, these samples lack identifiers or codes that can link a particular sample to an identified specimen or a particular human being. Typically, repositories send unlinked samples 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.’

  • Coded: ‘Sometimes termed ‘linked’ or ‘identifiable’, these samples are supplied by repositories to investigators from identified specimens with a code rather than with personally identifying information, such as a name or Social Security number’.

  • Identified: ‘These samples are supplied by repositories from identified specimens with a personal identifier (such as a name or patient number) that would allow the researcher to link the biological information derived from the research directly to the individual from whom the material was obtained’.

A great variability exists in the identification of the samples used, depending upon the source of the material and the purpose of the research. In some studies where individuals just serve as the sources of the samples, identifying them is not necessary. For other studies in which extensive information on diagnosis, family history, and demographics is crucial, the ability to identify the source of the sample is essential. But whatever the type of sample identification, with modern – DNA – identification techniques, it is possible even in anonymized collections to link a sample with an individual if one wishes to spend the effort and if the individual provides a fresh sample for matching.10

The consent requirements for the banking and further use of samples will be explored in the third section.

Confidentiality issues

Studies that maintain identified or identifiable specimens must maintain subjects' confidentiality, that is ‘information from these samples should not be provided to anyone other than the subjects and persons designated by the subjects in writing’.11 However, the American Society recommends that investigators inform individuals that they cannot guarantee absolute confidentiality. The ASHG ethics committee recommended12 that researchers ‘consider a way of coding samples by a third, independent party, who would keep the codes inaccessible unless there are specific circumstances in which the code needs to be broken’. In other instances, the only way to protect the confidentiality of identified and identifiable samples is to code them 5,7,9,13,14 The Health Council of the Netherlands8 considers that wherever possible, anonymous samples be used; identifiable samples should be limited to studies that could not otherwise be performed. For the Swedish Medical Research Council,15 ‘if biobanks are to fulfill important purposes for future biomedical research, the identity of the samples must be preserved. It is essential for biological material to be coded, and for the code key to be stored separately. The code must be kept within public institutions. Strict rules for storage and the use for the code key must be established, preferably in consultation with the local personal data representative’. In the United Kingdom, the Human Genetics Commission9 considers that one-off consent should be sufficient if identifiers are encrypted and recommends that the Government support research into robust methods for encryption for use in situations where the encryption needs to be reversible.

Requests for DNA banking

Who may request that DNA samples be stored? In 1988, the ASHG stated that ‘a DNA bank or a DNA diagnostic laboratory should accept samples only in response to requests from health-care professionals and not in response to requests from individuals or families without the mediation of health-care professionals. (…) If an individual should bank DNA without a genetic evaluation, such an evaluation is desirable before the DNA is analyzed’. In order to avoid misunderstandings between the depositor and the DNA bank, it has been recommended that the depositor be informed about the following issues: (1) the services to be provided; (2) the duration of storage; (3) the disposition of the DNA at the end of the agreed-upon term of storage or upon the death of the depositor; (4) the conditions under which DNA can be used for purposes not requested by the depositor; (5) a discussion of risks associated with DNA banking, such as loss of samples; and 6) an agreed-upon method of maintaining contact between the depositor and the bank.15,16

Security of banking facilities

What are the security mechanisms for DNA banking? HUGO's 1996 Statement on the Principled Conduct of Genetic Research recognized privacy and the need to protect against unauthorized access by putting in place mechanisms to ensure the confidentiality of genetic information. It advocated that information and samples be coded, that procedures be put in place to control access, and that policies be developed for the transfer and conservation of samples. The Council of Europe's Recommendations on the Protection of Medical Data (1997) included new provisions applicable to the collection and automatic processing of medical data, including genetic data. Article 9.1 states that ‘appropriate technical and organizational measures shall be taken to protect personal data – processed in accordance with this recommendation – against accidental or illegal destruction, accidental loss, as well as against unauthorized access, alteration, communication or any other form of processing. Such measures shall ensure an appropriate level of security, taking account, on the one hand, of the technical state of the art and, on the other hand, of the sensitive nature of medical data and the evaluation of potential risks’. At the national level, most countries would subject genetic information to the security mechanisms in place for nominative and medical data generally (7,17; French Bioethics Law, 1994; 8; Swedish Act Concerning the Use of Gene Technology on Human Beings, 1991). Austria (1994), Estonia (2000) and Iceland (2000) have specifically adopted a law to ensure that appropriate measures are applied to the storage and use of genetic information.

Quality assurance for banking

As with the principle of quality assurance, it has been argued that the implementation of security mechanisms to ensure the confidentiality of genetic information and long-term conservation of genetic material should be a sine qua non condition of banking.18,19 These mechanisms should be in place before sampling is done. Very strict guidelines should be designed concerning the practical aspects of data storage and DNA banking, which imply coordination by experienced researchers, analysis by qualified personnel as well as modern storage facilities. Evaluation of the validity of the obtained results should also be included. The Swedish MRC15 recommends that ‘all biobanks containing both biological material as well as associated information about individuals must have an organization with explicit procedures for quality assurance, including systems for storage, coding, and registration’.

Regulation of banking facilities

What are the regulatory mechanisms for banking facilities? A first type of regulatory mechanism relates to the accountability and oversight of these facilities.20 Some policy statements recommend that DNA banks be covered by institutional regulations, such as Institutional Review Boards (British Clinical Genetics Society 1989, Health Council of the Netherlands 1994).9,15 Means of ensuring the oversight of banking facilities vary from accreditation of banking facilities21 to their voluntary certification.2,16 A second class of regulation relates to the quality assurance of the techniques used by the banks and the competence of technicians.20 Proposed requirements to ensure maintenance of high standards include the training and certification of personnel and the listing of bankers' responsibilities (Health Council of the Netherlands 1994).16 A third type of regulation relates to safety measures for banks, such as storage conditions or means of protection from hazards including unauthorized access, loss of samples, computer breakdowns, etc.9

Self-regulation is another approach to ensuring that DNA banks conduct their activities in a way which is ethically and legally sound.2 McEwen & Reilly2 argued that DNA banking could be responsibly managed if professionals developed and adhered to a code of conduct and developed and used written agreements that describe the rights and obligations of all parties with respect to storage. Because of the differences in orientation between academically based and commercial DNA banks, the authors recommended two different approaches: in the case of academically based banks, they recommend that governmental agencies, like the NIH Office of Protection from Research Risks in the USA, advise local institutional review boards to follow specified guidelines to ensure that DNA banks adhere to uniform standards. As for commercial DNA banks, voluntary agreement should be sought on a core set of rules for DNA banking to which all banks should adhere (a code of professional conduct) (ibid). A professional body, such as the American College of Medical Genetics in the USA, should play an oversight role to ensure compliance by the DNA banks that adopt the code of conduct. To a voluntary code of conduct, McEwen and Reilly2 added the use of a depositor's agreement that delineate the rights and obligations of both those who provide samples for storage and those who bank them. This may indicate one way in which academic and commercial DNA banks can conduct their activities in an ethically sound manner. More specifically, the Swedish Medical Research Council15 considered that ‘rules on research ethics review should apply in the same manner for biobanks maintained by industry. Such biobanks should only include coded material. A code key should be stored with a suitable public authority such as a university or county council’.

Use of DNA sample collections by academia and industry

There are a number of potential sources of DNA samples, including the military, teaching hospitals, new-born screening, pathology or research labs, commercial laboratories, pharmaceutical or biotechnology companies, patient associations, forensic services, and various blood, cell and tissue banks. DNA sample collections vary considerably in size, ranging from large collections formally designated as repositories to samples informally stored in a researcher's lab freezer. It is difficult to quantify the scale of existing sample collections or to describe the use of such collections. An inventory of stored tissue specimens in the United States revealed that a conservative estimate of the number of specimens held in the country was in excess of 282 million (National Bioethics Advisory Commission 1999).22 New samples are being collected in the US at the rate of 20 million a year and the NIH spent some $53 million in 1996 alone supporting extramural tissue repositories.22

DNA sample collections are used for various purposes, namely for clinical, research, and industrial uses.23,24 This tends to legitimate different institutional and legal frameworks (profit or nonprofit organizations, private or public). It may be necessary, however, to bring institutions operating in the same domain closer when it comes to defining the conditions of exchange of biological material or information. Different conditions and prices could create a distortion among various institutions maintaining the same material. Further information is also missing if the contractor wants to evaluate the structure and the regulations that govern the observed collection or database in the light of its objectives and social and economic context. Gathering information on the main issues related to the constitution and the maintenance of collections as well as on research, the clinical or the industrial purposes they are to be subjected to and the expected results appears crucial in order to provide an exact picture of the diversity, role and importance of banking institutions (either public or private). The type of information to be gathered can be described as follows:

  • How are collections constituted? What kinds of samples are collected and are they subject to transformation? Under which criteria is genetic material included in the collection and how is it documented? How is the information linked with the material managed?

  • How are collections maintained? Are there standards of conservation, regeneration, distribution and if so, how are they established? What are the safety provisions?

  • What is the scientific work done on these collections? Content of the scientific activity: evaluation, characterization, gene mapping, genetics of diseases, etc. Organization of research: modalities of cooperation, further uses of samples.

  • How are scientific results made available to the scientific community and what are the conditions of access to such results for private as well as public institutions? The analysis of the structure of the existing collections and databases and the way they operate appears to be the only way to provide regulators with the information and analyses that will help them make necessary choices.

The biotechnology and pharmaceutical industries are also developing DNA banking and it is highly relevant to understand their objectives in this respect.25 The majority of biotechnology companies are primarily concerned with generating and selling information about the relationship between specific genetic sequences and particular diseases, rather than developing drugs themselves. However, it has been reported that some firms are planning to develop diagnostic tests based on this data, a number are offering contract genotyping services, and others are looking to develop drugs in partnership with large pharmaceutical companies (ibid). As to pharmaceutical companies, many of them have major interests in the collection of biological samples. There has been a rapid growth in pharmacogenetics research in recent years, with companies increasingly taking samples from participants in clinical trials to support their work in three main ways: (1) to analyze the molecular basis of disease; (2) to understand disease stratification and (3) to analyze drug response, including dosage and adverse events. Pharmaceutical companies also obtain DNA samples from patients via agreements with academic researchers in order to have additional material for points 1 and 2. Relatively little seems to be known about the scale of these collections by large pharmaceutical companies, but industry reports suggest that this is now a routine activity (ibid). It should also be noted that research is international in scope, with companies working in many countries across different continents.

In other respects, private sector activities depend heavily on both publicly funded research and widespread public participation. It is therefore difficult to draw a line between public and private research, as researchers from both sectors are often involved in supporting the same project. Very close academic-industry links are a general feature of research in genetics. While this enables effective technology transfer, it also gives rise to concerns about academic conflicts of interest (ibid).

Issues

Consent requirements for the banking and further use of samples

For research use of human biological materials, two basic protections for individuals usually come into play: (1) informed consent is required, and (2) Institutional Review Board (IRB) or ethical committee oversight is required to ensure an acceptable balance between risks and benefits. However, there are variations that are pertinent to research using human biological materials.

Most countries explicitly mention that (1) consent should be written, and (2) specific protections should be provided for vulnerable populations. The majority mention the possibility of anonymous studies as an exception to an express consent. There is some debate as to whether the principle of consent can be met in every research situation. Consent requirements can depend on the study (prospective or retrospective) to be conducted and on the category (identified, identifiable, anonymized, anonymous) of samples to be banked.1,11,26

Nonanonymous samples

At the time that new collections are created it is difficult to foresee all the potential research applications that the collection may be used for. It can be argued that different consents should be required for the physical taking of the sample, its use in a specific research study, its use by third parties, its subsequent use for other research purposes, and its use in a commercial application. Or rather than providing choices, participants may simply be informed of institutional policies concerning, for instance, the length of time of storage, approaches to discarding, or conditions of access by others.19

The British Medical Research Council Working Group27 suggested that at the time of the collection of the sample, broad consent should be obtained for the future unforeseeable uses of the sample in research, without the need to recontact individuals. As for the British Royal College of Physicians,28 it regarded that the secondary use of human material samples does not require the express consent of the individual. The concerns of the College are that requiring consent may bring to a halt all research on existing, archived material. It has been reported that this would be similar to the approach that has been used in epidemiological research, where the practice has been that consent need not be obtained for nonintrusive research as long as approval has been obtained from an ethics board.25 Finally, it is also recommended that researchers provide subjects with choices concerning subsequent use of a sample.29 For instance, it would be reasonable to give the individual the option to refuse permission for any secondary use, permit any use but only if the sample is rendered anonymous, permit only research involving a certain disease or diseases, or permit any use without anonymization at the discretion of the investigator.

In the case of existing collections it may be impractical to gain consent for new research uses from the donors of the samples. Concerning already-collected, coded and identified samples, many bodies have agreed that they may be stored and used for other purposes than those originally intended if informed consent for banking and subsequent use has been obtained.4,7,8,14,15,30 The Ethics Committee of HUGO13,31 states that ‘research samples obtained with consent and stored may be used for other research if there is general notification of such a policy, the participant has not yet objected, and the sample to be used by the researcher has been coded or anonymized’. An exception to the principle of consent has been proposed for prospective statistical and epidemiological purposes13 The ASHG11 considers that ‘investigators should be required to recontact the subjects to obtain consent for new studies’. The National Bioethics Advisory Commission5 states that for research samples that are identified or coded, the possibility that the investigator will contact the source or the source's physician for additional information should be discussed during the consent process. In other respects, the Commission considers that research on existing samples that are identifiable does not require informed consent and may receive IRB review, provided that it does not exceed ‘minimal risk’ to the donor. Minimal risk is defined as any risk exceeding those encountered in daily living or during routine medical visits.

Anonymous samples

Regarding already-stored anonymous samples, several bodies have agreed that they may be stored and used for other purposes than those originally intended.5,7,8,11,26,32 In 1993, the Danish Council of Ethics recommended specific legislation to this effect.

Concerning anonymized samples, a number of statements consider anonymizing samples acceptable in order that they be banked and subsequently used in retrospective and prospective studies without obtaining an explicit consent. The advantage is that there is no possibility of breach of confidentiality, of duty to communicate results, of risks of stigmatization or of discrimination.5 Others consider that this approach is unacceptable since researchers have an opportunity to seek consent on whether the samples may be anonymized or not but do not use it.20 According to the NIH guidelines, when samples are obtained in a research setting, consent of the ‘source’ should be obtained for the stripping of identifiers and for prospective research.33 As for the ASHG,11 ‘investigators should consider the appropriateness of anonymizing samples, especially when there is available medical intervention for the disorder being tested’. The HUGO Ethics Committee31 mentions that while necessary demographic and clinical data may accompany the anonymized sample, careful consideration should be given before proceeding to strip samples of identifiers since other unknown, future uses may thereby be precluded as well as the subsequent validation of results.

Scope of informed consent

When consent is required, what is the scope of informed consent? In obtaining consent for the banking of identifiable or identified samples and subsequent use, policy statements recommend that several elements be disclosed, such as the purpose of the research, its limitations and outcomes, its risks and benefits, the types of information that could result from genetic research, communication of results, or means of maintaining confidentiality. 7,8,11,13,14,15,16,21,26,27,34,35 Individuals should be given options with respect to the types of research that can be carried out, the access to or sharing of stored samples, the duration of storage as well as the right to withdraw samples. ASHG16 considers that unless immortalized cell lines have been established, patient DNA is exhaustible and the patient's needs should take priority.3,36 Individuals should also be informed if the sample will be stored for later study as well as the possibility of storage failure. For the ASHG,16 ‘it is inappropriate to ask a subject to grant blanket consent for all future unspecified genetic research projects on any disease or in any area if the samples are identifiable in those subsequent studies’. Individuals whose samples are stored should also receive general information pertaining to the institutional policy with respect to banking (Health Council of the Netherlands 1994).16 Documents now focus on what constitutes appropriate disclosure during the consent process and on the documentation thereof in consent forms. Most assert that the potential subject must be told about the investigator's intent to retain samples and the scope of intended subsequent use. They have also asserted that the subject has the right to decline to permit secondary use of his or her sample or to limit it in general terms.29

The Health Council of the Netherlands8 and The Nuffield Council of Bioethics35 set out special rules for the storage and subsequent uses of samples procured from vulnerable persons. According to the Nuffield Council on Bioethics, when the patient is lucid his or her consent should be required and sought, and when the patient is in no position to give that consent the guiding principle must be that what is in his or her best interests should be determined by family members and/or physicians.

Concerning postmortem uses of samples, a policy of unrestricted access cannot be justified on the grounds that no ethical issues are at stake.37 In cases in which research using samples from deceased individuals involves identifiable private information about their living relatives, the research may pose risks for them.38 When they are still alive, individuals may want to establish policies to ensure that some of these outcomes do not occur. According to the National Bioethics Advisory Commission,5 if individuals restrict use of their samples when they are still alive, those restrictions should also apply after their deaths.

Generic consent

The ASHG10 considers that a blanket consent for all future unspecified genetic research projects is inappropriate if samples are identifiable. On the contrary, the American Society for Investigative Pathology39 as well as the Association of American Medical Colleges40 believe that specific requirements for informed consent will discourage patients from participating in biomedical research, and that researchers will be unable to pursue such studies without additional financial resources to support such an administrative burden. The Pathology Societies agreed that for research with identifiable specimens, the Institutional Review Board approval remains the norm but add that such research should be possible with a general consent procedure. The AAMC recommends that research on archival patient materials, whether linkable or not, should be permitted under a general informed consent mechanism. This mechanism should protect genetic information from unauthorized disclosure and misuse. WHO14 also proposed that existing stored samples ‘should not be subject to new rules for consent or recontact that may be established in the future’: (…) ‘a blanket informed consent that would allow use of a sample for genetic research in general, including future, as yet unspecified projects, appears to be the most efficient and economical approach, avoiding costly recontact before each new research project’.

Two complications with individual informed consent must be noted: (1) Sometimes informed consent is not possible; in those circumstances, a carefully circumscribed use of community consent or ‘permission’ may be allowed if the potential benefits are sufficiently great and particular ethical care is taken (CIOMS 1991). (2) Informed consent by and for vulnerable subjects raises serious ethical problems; if it is absolutely necessary to sample such individuals, it should be done only after an explicit review and approval by an Institutional Review Board (ibid).

Consent at the population level: group consent

For populations, if a such population is to be the subject of research, then there is a case for saying that consent may be required at a group level.41 This requirement has been recognized by the Tri-Council of Canada42 and by the Human Genome Diversity Project,43 which supports the principle that population consent, as well as individual consent, should be sought for genetic research. Researchers should obtain the informed consent of the population, through its culturally appropriate authorities, before they begin sampling. Some may argue that the refusal by a population may violate the rights of an individual who wants to participate.

What form should group consent take? For some populations, that will be individual signed consent forms as well as a group consent form or agreement signed by the population's authorities.44 In other populations, written documents may not be appropriate. The precise form of the consent must take these differences into account. Nevertheless, there is a possible conflict between this form of consent and the law of some sponsoring countries. Many countries require signed consent forms from all individuals participating in funded research and affected researchers may have to abandon plans to work with that community.

Use of samples banked prior to collection of informed consent

A few clinical laboratories have formal policies for the use of stored specimens which has been obtained in the past without consent or without specific consent to their use in research and from which genetic inferences can be made.45 It is also not known how many clinical laboratories have a fully informed understanding of the way consent was obtained for specimens they are testing as part of their involvement in scientific or clinical research.46

A variety of scientific groups, including the British Medical Research Council Working Group27 the American Society of Human Genetics,11 the American College of Medical Genetics,26 the National Bioethics Advisory Commission,5 HUGO,31 and WHO,14 issued position papers that address the use of archived specimens. Although everyone recognizes that the consent process under which many samples were acquired does not meet today's standards, they recognize that there are many valuable archived specimens that could be used and that it would be prohibitively expensive to try to obtain a new consent from each donor for reuse.29 For ASHG,11 making samples anonymous will eliminate the need to recontact to obtain informed consent. This will also reduce the chance of introducing bias due to inability to recontact some, or the possible refusal of others to participate. Or if the samples are identifiable, new consent would have to be sought. In NBAC's judgment (1999), where the research uses identified or coded samples from previously collected specimens, such uses usually are not justified without the source's consent; the use of unidentified or unlinked samples for research could be justified in some cases if other appropriate protections were in place, despite the lack of consent. The HUGO Ethics Committee31 considers that routine samples obtained during medical care and stored before the notification of using such samples for research and with the patient's consent may be used for research if the sample has been anonymized prior to use. The British Medical Research Council Working Group's position27 is that for old collections, samples should be regarded as abandoned and therefore are able to be used for new research purposes as long as ethics committee approval is obtained.

Use of samples from archived pathology materials

Most researchers using human biological samples have relied on specimens from archived pathology materials. Some studies require samples with specific biological, clinical, or demographic characteristics. The creation of such collections can be of great value to researchers.

As noted in the previous section, some specimens are gathered during clinical procedures for which no informed consent was obtained. However, even when informed consent was given for the medical procedures that produced the specimens, the individuals may not have consented to possible future research uses of the material or may not be aware that their specimens might be used for various research purposes by a number of investigators. Even when individuals are asked to provide samples for possible use in future research and even though an approved research protocol does not yet exist, the question is whether individuals can give specific, informed consent today to the use of their materials at some time in the future. For the American Pathology Societies39 and the Association of American Medical Colleges40 research on archival patient materials should be permitted under a general informed consent mechanism. In a report on ‘research involving persons with mental disorders that may affect decision-making capacity’, the National Bioethics Advisory Commission47 considered that, within limits and with appropriate protections, individuals, while competent, could give prospective authorization to a particular class of research if its risks, potential direct and indirect benefits, and other pertinent conditions were explained. Perhaps more than for other samples, it seems appropriate to ask a subject in the setting of a clinical biopsy or a surgery to grant a blanket consent for future unspecified genetic research projects on the disease this person suffers from. Indeed, it is by no means possible at any given moment to predict exactly which further studies will be undertaken. As stated by the HUGO Ethics Committee31 while necessary demographic and clinical data may accompany the anonymized sample, careful consideration should be given before proceeding to strip samples of identifiers since other unknown, future uses may thereby be precluded as well as may the eventual validation of results. As to old collections, according to the British Medical Research Council Working Group's position27 samples could be regarded as abandoned and therefore be used for new research purposes as long as ethics committee approval is obtained.

The British recommendation may be particularly relevant for pathology samples. In fact, good clinical practice in pathology requires long-term storage of tissue samples for different reasons. First, initial results may need validation by additional techniques that were not yet available at the time the biopsy was taken. Alternatively, important progress made in the understanding of certain pathologies may demand re-evaluation of the material in view of current insights. Finally, the prognostic impact of certain evolutions in the patient's disease process often strongly depends on the pathologic findings at diagnosis that frequently cannot be adequately judged by means of the pathology report only. Electronically stored pictures will not solve this problem.

Control of banked samples and quality issues

Access to samples and sharing

Worldwide research endeavors have raised the issue of access to and sharing of banked samples. While protecting confidentiality, the free circulation and the availability of genetic information and samples for research has been promoted by many instances.13,14,48,49 At the international level, HUGO (1996) and UNESCO (1997) mandated that DNA samples should be openly available to the scientific community: ‘collaboration between individuals, populations, and researchers and between programs in the free flow, access, and exchange of information is essential not only to scientific progress but also for the present or future benefits of all participants. Cooperation and coordination between industrialized and developing countries should be facilitated’. WHO14 adopted the same position: ‘qualified researchers should have access to samples, provided that strict confidentiality is observed or that identifying characteristics are removed’. Concerning access by relatives, HUGO31 stated that ‘Where there is a high risk of having or transmitting a serious disorder and prevention or treatment is available, immediate relatives should have access to stored DNA for the purpose of learning their own status. These exceptional circumstances should be made generally known at both the institutional level and in the research relationship’. WHO14 went further and recommended that ‘control of DNA should be familial, not individual. All blood relatives should have access to stored DNA for purposes of learning their own genetic status, but not for learning the donor's status’.

At the regional level, in 1991, the Council of Europe proposed that ‘standard research contracts have a clause to the effect that the sharing of all knowledge and distribution of materials will be obligatory. (…) Data and materials as much as possible should be free of charge or available at a nominal cost or to cover distribution costs’. In 1997, the same instance adopted the Recommendation on the Protection of Medical Data which provided that ‘Medical data can be used by health professionals for their own medical research as long as the data subject has been informed of this possibility and has not objected’.

At the national level, access to medical records or to samples for genetic research is normally restricted to qualified investigators and subject to institutional oversight, be it legislative or via ethics committees.19,50 Generally, consent of the patient is also required.11,15,21,51,52 Ethics committees or legislative provisions may also grant access to records or samples without consent for research purposes, mainly if the data are anonymized (French Law No. 94-548 1994).7,32

In most international and national legislation and protocols, individuals are considered to have an absolute right to give or to withhold information about their genetic status, and equally an absolute right to prevent their stored genetic data being transmitted to a third party for whatever purpose. However, one of those purposes might be further research in which individuals' genetic data might assist in securing health benefits for a large number of other people. Some people wonder if in such a case should such individuals be able to exercise a right to withhold their genetic information, particularly if it is encrypted or anonymized, when it might play a part in establishing links and patterns with genetic defects in members of their families or some wider social grouping and thus contribute significantly to their well being.53 It has been argued that the principle of inalienable individual rights that lies at the heart of much of the legislation about data protection may not be in the best interests of the population's health (ibid).

Duration of storage

Several policy statements recommend that samples be stored for limited periods to be specified at the time of collection or only for the time needed to complete the purpose pursued in banking the samples.7,8,16,26 A prolongation of storage must be justified by a good reason, such as research or an absolute need,7,8 or when the original consent to collection and storage did not prohibit the use of anonymized samples for research.7 WHO14 recommends that DNA be stored as long as it can be of benefit to living or future relatives, while the British Clinical Genetics Society6 recommended an open-ended duration for the storage of DNA.

Concerning the destruction of stored samples, a few guidelines state that individuals who withdraw from a study may request destruction of their sample.26,33,54 For the HUGO Ethics Committee31 in the absence of need for access by immediate relatives, stored samples may be destroyed at the specific request of the person. The Committee recognizes that the destruction of samples is not possible for samples already provided to other researchers or if already entered into a research protocol or used for diagnostic purposes. Some authors point out that given certain techniques used in genetic research, such as the immortalization of cell lines, doubt may be raised as to the realism of offering individuals the possibility of having their samples destroyed upon withdrawal from a study.1,20 By their nature, anonymized samples cannot be withdrawn by their donors.

Responsibility of the diagnosis of stored samples

Most researchers believe that findings from research should not be communicated to subjects unless they are confirmed and reliable and constitute clinically significant or scientifically relevant information. Revealing unconfirmed findings may place subjects at risk of harm. However, it has been argued that the principle of autonomy dictates that subjects have a right to know what has been learned about them and that interim results should be shared with subjects.55 In 1996, the ASHG recommended that ‘all genetic research studies involving identified or identifiable samples in which disclosure of results is planned have medical geneticists or genetic counselors involved to ensure that the results are communicated to the subjects accurately and appropriately. In other respects, it is the obligation of the subjects to keep the investigator informed of how they may be contacted’. In 1999, the National Bioethics Advisory Commission considered that disclosure of research results to subjects should occur only when all of the following apply: ‘(1) the findings are scientifically valid and confirmed; (2) the findings have significant implications for the subject's health concerns; and (3) a course of action to ameliorate or treat these concerns is readily available’. (…) ‘When research results are disclosed to a subject, appropriate medical advice or referral should be provided’.

Ownership of banked samples

Many of the agreements about ownership of banked samples and access to biological material and information are determined by multiparty contracts and are not regulated by legislation.25,56 The general practice is that information belongs to the researcher or team that creates it and the individual who may have been a subject of the research has no legal entitlements to that research. The claim that subjects should own their sample even while it is entered into research is the minority view.29 In 1988, the ASHG statement established that ‘banked DNA is the property of the depositor unless otherwise stipulated’. On the other hand, the British MRC Working Group on Collections of Human Tissue and Biological Samples for Use in Research4,27 considered that the funding body retains ownership of the collection while the researcher is the custodian of the collection. The custodian has the responsibility of control over the access to the collection and ensuring that standards of confidentiality are maintained. Funding bodies need to determine the purpose of the collection and if it is available to both commercial ventures and academic researchers. It has been suggested that potential subjects should decide whether they are willing to participate only after they have been informed about who will own the sample (as it is for tissue) and whether or not there is a plan for subsequent use of sample. Subjects should have the right to decide to donate their tissue to the research team. As with other aspects of life, competent individuals have the right to make gifts and these gifts are owned by those who receive them.29

The issue of ownership arises primarily because of the possibility that DNA samples could have some commercial value. This issue is most controversial in connection with the patentability of ‘inventions’ derived from the scientific analysis of human material. This patentability remains contested in various nations and within particular groups. At the international level, HUGO has maintained a consistent position on the unpatentability of human DNA sequences.57,58 At the regional level, in 1994, the Council of Europe's recommendation on the Protection and Patentability of Material of Human Origin stated that ‘human beings are subjects – not objects – of law’. After 3 years, in its Convention on Human Rights and Biomedicine, the Council of Europe insisted that ‘the human body and its parts shall not, as such, give rise to financial gain’. Finally, in 1998, the European Commission Directive concerning the Legal Protection of Biotechnology Inventions intended to complement international intellectual property provisions. By distinguishing between inventions and discoveries (the latter being unpatentable), the Directive responds to concerns that the human body or parts thereof could be patentable. According to the Article 3, ‘an element of the human body in its natural environment or even a sequence or partial sequence is unpatentable’, while Article 9 states that ‘even a patentable invention – according to the European patent law – can be excluded if contrary to public order and morality’.

At the national level, the status of human genetic material is not so clear: some countries have not taken any position on the issue of status, or among the few that do, the position is often through an indirect reference. Whatever the country, any inventions even if patentable are still subject to the public order and morality filter of European patent law.19 Finally, while a property position may allow for actual or potential financial return, the personality approach avoids individual returns but not the possibility of commercialization by the researcher, through traditional intellectual property rules. Thus, irrespective of the qualification, ultimately, patenting is still possible but the locus of the financial benefits is different (ibid). There are important questions which remain unanswered about the social acceptability of the private ownership of gene patents, and the impact this might have on scientific research, innovation and the costs of new medical technologies.

Return of benefits to the community

In the case of research using donated samples, it is generally assumed that the individual donor is giving his biological material to further the collective good of the community, rather than his personal profit or the private profit of a company.41,59 Simultaneously, while some policies aim to encourage the commercial exploitation of publicly funded research, there is also a recognition that this must be matched by a suitable social return, either in the form of technology transfer, local training, joint ventures, provision of health care or information infrastructures, reimbursement of costs, or the possible use of a percentage of royalties for humanitarian purposes.13,25,60 As we noted it, some consider that research teams, whether they be situated in academe or industry, should disclose that they may someday benefit economically from the research and that the subjects will not.2,29 Potential subjects will then be free to decide whether they are willing to participate regardless of this stricture. According to the Directive 98/44/EC of the European Parliament and of the Council of 6 July 1998 on the legal protection of biotechnological inventions, if an invention is based on biological material of human origin or if it uses such material, and where a patent application is filed, the person from whose body the material is taken must have had an opportunity of expressing free and informed consent thereto, in accordance with national law (Recital 26).

The remuneration of participants in genetic research is widely proscribed, irrespective of the fact that its use may ultimately lead to financial rewards for the researcher or industry. Although commodification (reification) and commercialization of the human body and its elements is not considered ethically sound,13,49,61 there is an emerging concern for a more equitable approach that provides some return of benefits to the community.13,14,60,62 Many new products, including vaccines and drugs for common diseases, are now based on genetic research. Much government or nonprofit research will eventually be commercialized and companies involved in human health may have special moral obligations.

The North American Regional Committee of the Human Genome Diversity Project63 has proposed guidelines for providing benefits to sampled populations. Three basic principles should govern researchers in this connection: (1) Honesty, that is any benefits that are promised must be both deliverable and delivered. If the sampling is a part of research aimed at studying a disease, the community must be told so, but it should also be warned that the research may not lead to any useful knowledge and that even useful knowledge may not lead to treatments. (2) Legality, that is national and local laws in the areas where research is being conducted must be consulted to see what kinds of benefits may legally be given to whom. The researchers' national laws, or the conditions of any financial support for their specific project, may also affect what benefits can be legally given. (3) Appropriateness: the benefits must be appropriate in their nature, in their scale, and in their distribution within the community. The range of possible benefits that researchers might confer on a community could include products, supplies, training, or services. Money is one kind of benefit that demands particular mention: paying a community for participating may raise special concerns about legality and coercion. Transfer of scientific technology is another kind of benefit that may be particularly useful in some contexts. The scale of the benefit is also a crucial concern: an enormous benefit may make the process of informed consent meaningless by making it effectively impossible for the community to say no. Finally, researchers are obligated to do no harm through a careless distribution of benefits within the community.

The North American Regional Committee of the Human Genome Diversity Project63 has identified two issues that might be viewed as providing benefits – medical services and financial interests in the samples and their use – but that might in fact raise special problems. The provision of medical services in connection with collecting samples for the HGD project may seem an almost uniquely good benefit, welcomed by almost everyone. Nevertheless, providing the medical services must be culturally appropriate for the population and the kind of medical service must be both useful and feasible. Researchers must be also prepared to provide services to the entire community, not just those who provide samples. To limit medical benefits to the latter contravenes that principle and risks coercing individuals to participate. It is recommended that the research group considering medical services investigate that possibility during their early contacts with the group, long before sampling begins.

Concerning the second issue – financial interests in the samples and their use – the HGD project requires that all financial benefits issued from patenting and commercial use of samples and of the information derived from them should be returned to the community. The manner in which this resolution should be implemented is that anyone seeking access to the Project's samples or data would have to agree to be bound by contract to a set of rules concerning the rights of the sampled populations.

The HUGO Ethics Committee60 presented a statement on benefit sharing. The Committee made six recommendations: (1) all humanity share in, and have access to, the benefits of genetic research; (2) benefits should not be limited to those individuals who participate in such research; (3) there should be prior discussion with groups or communities on the issue of benefit sharing – such prior discussion should include consideration of affordability and accessibility of eventual therapy, and preventive and diagnostic products of research; (4) even in the absence of profits, immediate health benefits as determined by community needs could be provided – immediate benefits include medical care, technology transfer, or contribution to the local community infrastructure (eg, schools, libraries, sports, clean water); (5) at a minimum, all research participants should receive information about general research outcomes and an indication of appreciation; the ethical advisability of provision of information to individuals about their results should be determined separately for each specific project; and that (6) profit-making entities should dedicate a percentage (eg 1–3%) of their annual net profit to health-care infrastructure or for vaccines, tests, drugs, and treatments, or, to local, national and international humanitarian efforts.

Case-studies

The ALSPAC study

The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC) is a research initiative of the University of Bristol in the United Kingdom designed to monitor and analyze different courses of health and development in a geographical cohort of children.64 It is part of an international research program with counterpart studies in a number of European centers. The main aim is to understand the ways in which the physical and social environment interact, over time, with the genetic inheritance to affect the child's health, behavior and development. The study is designed to link together information from a variety of sources including hands-on examination of the children, questionnaires completed by parents, health records, assays of biological samples and specific measurements of the environment in the home, and to use these unique data to test hypotheses on the causes and prevention of childhood ailments and disorders. Prospective data from early pregnancy and from maternal as well as child DNA permit transgenerational studies. Since 1991, ALSPAC has followed 14,000 families originating in the Avon area (www.alspac.bristol.ac.uk).

The DNA resource available for elucidating genetic interactions with environmental exposures is unique. A biological sample is collected from all mothers and children for whom permission is received. DNA is obtained from the mother from maternal blood samples already collected, and from the child from the umbilical cord, from cord blood and from further blood samples or buccal samples taken later in childhood. Where maternal blood is not available, buccal samples are taken. It is also planned to ask partners for buccal samples (ibid).

ALSPAC has its own independent Law and Ethics Committee since 1989 and genetic studies are guided by its Genetics Advisory Committee. The ALSPAC Ethics Committee feels that, in general, collection of biological samples during pregnancy or at delivery can be carried out without maternal consent, but that analysis of the samples should only be carried out if the mother has given consent. The mothers received the study brochure in pregnancy; it included information on the research that was envisaged at that stage. ALSPAC staff discussed the issues with each mother during pregnancy and most mothers signed the consent form. Where written consent had not been obtained prior to the mother giving birth, attempts were made to contact her by letter or at the clinic visit. In cases where samples are collected but a consent form has not been filed, the samples have not been used for research purposes (ibid). In 1999, consent for DNA analysis had been obtained for over 10 000 children.65

The core activity of the ALSPAC study lies in the maintenance of data collection and ensuring its availability for collaborators. The costs of maintaining existing samples and data accumulation anticipates that all the tests will be sponsored. Funding has been obtained from a variety of sources, such as government bodies, charitable research organizations, individual sponsorship, and industrial companies. ALSPAC has received some industrial funding, but has not yet secured a collaboration with a biotechnology or pharmaceutical company to exploit its collection for genetics research.25 In order to address the problem of a short fall in core funding, ALSPAC has requested that all applications for collaborators include a contribution towards the core costs of the study. This includes the funding of particular core personnel, the costs of particular items of data on a fee per item basis, or a lump sum (www.alspac.bristol.ac.uk).

deCODE genetics and the Act on a health sector database

In order to investigate if some diseases thought of as being acquired have inherited forms, researchers have started to hunt for small groups or populations. In Iceland, genealogical and medical records that exist for most of the population greatly extend the scope of this approach.

In 1998, the Icelandic parliament, Althing adopted a law called the Act on a Health Sector Database, making it legal for the Minister of Health to grant a license to a private company to construct an electronic database containing selected and coded data from the country's health records. The database aims to increase knowledge in order to improve health and health services and can, by prudent cross-referencing with other databases, contribute to the discovery of genetic and environmental risk factors in common diseases and statistical data on disease and treatment. The extent to which the new genetics will affect the delivery of health care remains unclear and Iceland provides a unique opportunity for testing this (Icelandic Ministry of Health 1998). The health economics aim is that since the data in the records have been paid for out of public funds they are not owned by individuals or institutions and should be used for the public benefit (ibid).

The Act on a Health Sector Database grants deCODE genetics exclusive rights for 12 years to use the data for purposes of financial profit, under strict conditions laid down in the Act, the Regulation and the License and within provisions of the Icelandic Competition Act and the provisions of the EEA Agreement. The Licensee shall in the creation and operation of the health sector database refrain from abusing his position as Licensee in his business with parties purchasing his services and special business terms shall be based on general and transparent business terms. deCODE has entered into a nonexclusive arrangement with Hoffmann–La Roche for the purpose of researching the genetic origins of 12 common diseases.66

The proposal for the IHD has been highly controversial both in Iceland and internationally. The Icelandic Medical Association and the World Medical Association opposed the Act.67 Similarly, the Icelandic Psychiatric Association, the Association for Ethics in Science and Medicine (Mannvernd), and the National Bioethics Committee opposed the Bill mainly because there was no provision for informed consent. The Icelandic Parliament attempted to address the criticisms of the proposal by redrafting the Health Sector Database Act, and the regulation that governs the establishment and running of the database was issued in January 2000.

Some of the most important concerns about the creation of the Icelandic database have centered on issues of privacy and data protection. The dangers of invasion of privacy inherent in modern technology and science are judged to be considerable and this has led to attempts to ensure protection through various international policy documents and legislation.49,68,69,70 For instance, the recommendation from the European council30 is explicit about the issue of informed consent for clinical trials on human subjects and on gathering and storing identified health-care information. However, the Icelandic legislation does not require explicit consent of the individuals whose information is to be submitted to the database, since this information will not be identified or identifiable with reasonable effort. Even if the data in the Icelandic health sector database are seen as anonymous, those who object can opt out of it, but it will generally be assumed that those who do not opt out can be included in the patient records database.

There are also fears that data cannot be kept confidential, even if the Icelandic Act includes measures ‘to ensure protection of confidentiality in connecting information from the Health Sector Database with the databases of genealogical and genetic information’ (Article 10) and even if ‘employees of deCODE must sign an oath of confidentiality’ (Article 11).71 Also, Article 14 states that ‘providing information on individuals from the Health Sector Database is prohibited and only statistical information involving groups of individuals may be provided’. deCODE has developed a third-party encryption system supervised by the Data Protection Commission, but many people argue that in a country such as Iceland, where there are only 270 000 people, it will be possible to establish which data belong to which individuals, particularly in cases of rarer conditions. Moreover, since the database is a dynamic system, with data being added to it all the time from hospital records, newly added data may be identifiable.72

Another set of concerns surrounds who has access to sample collections and databases of genetic information. The Icelandic Act has provision to regulate access and the use of the database, but deCODE's exclusive license to build the database and its exclusive rights to commercial exploitation of the database for 12 years as well as its (nonexclusive) arrangement with Hoffmann–La Roche have given rise to some of the major criticisms. Some scientists in Iceland believe that the health sector database will diminish their research opportunities.73 The Icelandic Government's answers are that the database will increase the research opportunities for scientists in Iceland in relation to funding, access to patients and to patients' records.74

Guthrie Cards

There are a variety of potential sources of samples that may be stored and used for subsequent purposes. The subsequent use of stored Guthrie cards is receiving increased attention as a source of DNA for research and testing purposes.33,75 Guthrie cards are the name given to the way in which the blood taken from newborn babies for genetic screening programs has been traditionally stored. Recognition of the epidemiological utility of dried blood spots on Guthrie cards for HIV seroprevalence surveys and a growing interest for DNA analysis has intensified consideration of issues regarding retention, storage, and use of residual dried blood spot samples. These samples potentially provide a genetic material ‘bank’ for all newborns nationwide.76 In 1994, an American study showed that although most state newborn-screening laboratories retain Guthrie cards, if at all, for only a short time, a growing number plan to keep them for an extended period and, in several cases, indefinitely. A large number of laboratories would also consider sharing anonymous cards for research purposes.75

However, newborn screening programs vary widely in approaches and policies concerning residual dried blood spot samples collected for newborn screening. Most of the time, newborns' blood spots are obtained in screening programs for treatable diseases. In general, newborn screening programs do not require a specific written consent for the testing for treatable disorders, but if research is proposed, parental consent is necessary. In 1993, the Danish Council of Ethics recommended that ‘stored newborn blood spots may be used subsequently for research without consent only if samples are anonymous’ even if parental authorization remains the norm for research using identifiable newborn samples, the Danish Council of Ethics recognizes that in some cases, using identifiable samples, it may be inappropriate to approach the parents. Since 1993 the Danish biobank, where Guthrie cards are stored, has been regulated by specific legislation, and thus assumes a unique position among biological specimen banks. Its purposes are: (1) diagnosis and treatment of diseases screened for, including repeat testing, quality assurance and group statistics; (2) other diagnostic uses during infancy; and (3) research projects. The stored samples have been used successfully to diagnose a range of genetic diseases using biochemical and molecular genetic assays.77 Storage of neonatal screening samples is thus beneficial not only to the individual testees, but also to future generations of newborns.

Other bodies have proposed guidelines for the retention, storage, and use of Guthrie cards. In 1995, the WHO recommended that ‘blood spots collected in screening newborns for treatable disorders could be used to provide epidemiological information about genetic predisposition to late onset disorders. Care must be taken to ensure that such testing remains anonymous and that results cannot be traced to individuals or families’. However, retention of some demographic data may be of considerable use to epidemiologists. It has been argued that in countries that have adopted positions on genetic epidemiology research, those requirements could apply to anonymized research on newborn samples.19 In 1996, according to the American Council of Regional Networks for Genetic Services, a federally funded national consortium of representatives from 10 regional genetics networks, newborn screening programs should promulgate policies and rules for retention and use of residual newborn screening dried blood spot samples, based on scientifically valid information. Banking of newborn samples as sources of genetic material should be considered in light of potential benefit or harm to society.76

Actually, most state newborn screening programs have few or no procedures for retaining, storing, retrieving, and using Guthrie card samples. Few scientifically sound procedural systems for use of dried blood spot samples remaining after newborn screening currently exist. In light of the growing interest in novel uses of stored Guthrie cards, it may be time to do so. The ethical concerns about issues related to privacy may ultimately dictate policies related to retaining or destroying residual dried blood spot samples (ibid). To date, no regulations exist that provide anonymity or guarantee privacy of newborn blood samples on Guthrie cards.

Discussion

Setting the scene

It came out during the workshop organized by the ESHG (see Methods) that DNA banking for medical and research purposes is indispensable. It facilitates the constitution of large collections, sharing of samples, multiple testing on the same samples, and repeating testing over the years. However, banking organization is complex, requires multiple actors, and concerns are expressed in various countries. Points of discussion focussed on ethical issues vs law requirements as well as on primary use vs secondary use of DNA samples, or for clinical vs research uses, or for new collections vs existing collections, or with identifiable individuals vs anonymous banking. Among ethical issues, informed consent, protection of confidentiality, sharing and commercial uses, guidelines within the institutions and the professional organizations, as well as control process issues were addressed. Ethical principles concerning the protection of individuals participating in research projects state that consent must be informed and confidentiality and private life must be protected. Also, human body elements may not be commercialized and genetic testing must be limited to medical or research use with the agreement of an ethics committee. Nevertheless, consent forms often do not completely fulfill ethical requirements. It has been reported that some issues are poorly addressed, such as further use of samples, feedback of results, sharing, non profit or for profit uses, consent for using specimens of died individuals. In the same way, there is no consensus about how to manage and organize collections of human material for genetic studies. Many instances promote the free circulation and the availability of genetic information and samples for research. But it has been noted that research is a highly competitive activity, which is predominantly due to the academic pressure to publish as productivity contributes greatly to obtaining and keeping research grants. Considering the impressive costs associated with contemporary research techniques, the latter have become the principal mans of survival of any research group. In addition, financial gain and lack of understanding of the research process might impair collaboration between research institutes. Consent for genetic studies has specific features which should also be taken into account at an international level.

DNA banking practices raise economic issues, such as the funding of DNA banking for research, the role of private firms, and the use of DNA banking for health care. Although the running costs are difficult to estimate because few bank managers know which financial, material, and human resources are dedicated to the banking activities, these costs can be substantial, and financing a bank may be difficult for researchers. In order to maintain the collections, some academic laboratories collaborate with private firms; nevertheless, this collaboration raises the question of the autonomy of academic laboratories vis-à-vis the private sector and in extenso the question of the mediation by private firms. The transfer from research to routine laboratories is facilitated thanks to the mediation of the industry which transforms research results into commercial products. But this ideal scheme of transfer is not always possible for all DNA activities because of the uncertain profitability which leads to short-term investments, to the transfer of costs to public laboratories, and to funding by patient's associations. Although the recognition of DNA banking as part of medical practice appears essential, questions remain concerning the financing solutions for DNA banking. Would a national or European centralization of DNA banks be a solution in order to realize economies of scale and to cut costs?

DNA banking practices also raise legal issues, such as problems of human dignity, ownership, commodification and threat to privacy. The DNA qualification has an impact on the legal issues related to DNA: the physical aspect (biological material) raises questions about attitudes towards the human body and body parts; the informational aspect (code) is linked to prevailing attitudes towards medical information generally. The commercial interests in DNA banks have risen along with the development in molecular genetic techniques allowing replication of the DNA material. Consequently, it appears necessary (1) to control the flow of banked DNA and DNA data, (2) to develop policies to regulate DNA banking more closely, and (3) to insure that DNA banking can perform its function without impinging on the rights and interests of individuals who have their DNA sample or DNA data in a bank.

Because of the recognized value of banking genetic material, many policy statements have begun to address DNA banking in the 1990s. These policy statements define the scope of informed consent extensively to ensure that individuals who provide samples are given the opportunity to make informed choices with respect to the possible uses of their stored samples. But there is some debate as to whether the principle of consent can be met in every research situation. At the time that new collections are created it is difficult to foresee all future research applications. In the case of existing collections, it may be impractical to gain consent for new research uses from the donors of the samples. For instance, specimens taken at autopsy and stored in paraffin blocks are basically DNA banking and are occasionally valuable sources for family investigation or research. Who should give informed consent for using specimens of died individuals? In other respects, worldwide research endeavors have raised the issue of access to and sharing of banked samples. While protecting confidentiality, the free circulation and the availability of genetic information and samples for research has been promoted by many regulatory bodies. Many policy statements have also begun to identify security and regulatory mechanisms for DNA banking. However, there remains a diversity of positions on the banking and further use of DNA samples regarding coding and anonymization of samples, retrospective and secondary uses, and duration of storage.

In order to prevent the abuse of individuals' rights and to protect human integrity, return of benefits to the community were also analyzed. Principles concerning samples donated for research state that they are given for the collective good of the community without possible legal claim on the intellectual property. In compensation, a suitable social return is proposed either in the form of provision of health care or information infrastructure or the collective use of a percentage of royalties for humanitarian purposes. Such is the case, for instance, for the Icelandic pharmaceutical company, deCODE genetics, which has agreed to compensate participants by providing them with free access to any tests or pharmaceuticals developed from the studies, during the life of the patents. Independence of the decision maker and of the body to receive the royalties is crucial to avoid coercion. Potential donors must be informed of the potential financial benefits for the research team in order to be free to decide whether they are willing to participate. However, the actual or future benefits should not serve as an inducement to participation. Benefits should be provided to all members of the community regardless of their participation in the research.

The concept of benefit sharing is not limited to the possible therapeutic benefit of participating in research. In the past, many researchers sought no specific reward for biomedical research. More recently, due to increasing private investment, researchers and institutions often demand a share of monetary benefits deriving from their research. In other respects, some people consider that since the vast majority of the advances in science represent mere theoretical insight, usually procured by analyzing large numbers of samples and unfortunately not generally linked to immediate financial gain, a certain contribution to scientific progress should be considered an acceptable and suitable social return as well.

Issues to be raised

At present, there is no official census of genetic banks; the number, the kind of biological material stored, the services available, as well as the financial supports are not known. Many biological sample collections are derived from neonatal screening, research projects, and diagnostic laboratories, but the majority is not dedicated to the acquisition, characterization, identification, preservation, and distribution of DNA samples. In Italy, genetic banks supported by the Telethon are editing guidelines for quality control, in particular for material preservation and shipping and respect of privacy.

In other respects, it is reported that pathology departments have constituted the largest human biological collections and this has great potential in terms of genome investigation. DNA obtained from archived tissues, blood films, slides may be useful to establish the molecular diagnosis of a deceased patient. This information may be of value for relatives even after 10–20 years. Consequently, regulations concerning pathologic archives should be considered. However, it has been argued that apart from the interesting genetic information a tissue biopsy provides, often with regards to peculiar pathologic conditions, it also supplies invaluable immunomorphological information. In addition, although the collected material in principle allows genetic analyses of nearly any kind, it is especially suited to gain insight in the genetic mechanisms underlying particular diseases. As such, after an accurate diagnosis has been established, the biopsy is not so much of interest to the individual or his relatives, but rather serves the investigation of disease mechanisms.

DNA banking is progressively part of medical practice. A new field, called ‘Community Genetics’ is emerging, where DNA banking is expected to play an important role and for which the help of clinicians is requested. Although this activity may be justified, it may be problematic because some aspects risk to be in conflict within the role of the clinician, which is to contribute both to the health of individuals or families and the people as a whole. DNA banking may be in the interest of the patient when the diagnostic measure was preliminary or unsuccessful or when the patient may wish to declare this family's interest as his own. DNA banking may be also in the interest of the patient's family when a sample of an index case needs to be stored for future diagnostic purposes. But DNA banking may be also in conflict with the interests of the patient and of his family in case of fear of unwanted information or of breach of confidentiality or of undesired research. Conflicts between patients and families may be important when a patient may wish his sample not to be stored or his existing sample to be discarded even though his sample may be of value to his relatives. It has been argued that if storage of a patient's sample is in the vital interest of a relative, it may be ethically defendable to keep the sample. Finally, at the community level, DNA banking may be in the interest of the people for instance for monitoring susceptibility alleles contributing to complex diseases, in order to take preventive action. But DNA banking may be also in conflict with the interest of the people because of discrimination's risks. To contribute to research into the genetic diseases may be an important role of the clinician provided that the desired degree of intimacy between clinician and patient is safeguarded.

DNA banking is also expected to play an important role in pharmacogenetics. For monogenic diseases current linkage methods are now efficient in identifying mutant genes, depending mostly on the total amount of family structures and DNA samples available. For susceptibility genes, identifications of confirmed polymorphisms associated with the disease have been much more challenging. For this purpose, scientists have constructed a high-density single-nucleotide polymorphism (SNP) map for disease susceptibility gene searches through large linkage regions. The construction of a whole genome high-density SNP map focuses the next stage of susceptibility disease gene research on the availability of well-constructed, accurately phenotyped patient populations. Glaxo Wellcome is now generating patient collections from multiple diseases with large unmet medical need. Pharmacogenetics does not only refer to susceptibility gene identification, but also to ‘right medicine for right patient’. Genetic profiling can be used to recognize patients who will respond positively to a particular medicine or to identify patients who will have an adverse event by taking a particular medicine. It is expected that genetic profiling will be performed at a reasonable cost by using a standardized genetic map.

Analysis of practical examples

The functioning of two large-scale DNA banks was presented and discussed, ALSPAC and deCODE genetics. The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC) aims to be a general reference population for genetic and environmental epidemiology. A general description is available on website http://www.ich.bristol.ac.uk. ALSPAC has his own independent Law and Ethics Committee since 1989 and genetic studies are guided by its Genetics Advisory Committee. ALSPAC enrolled 14,000 pregnancies during 1991 and 1992 representing 85% of the eligible population. Samples for DNA extraction have been taken from the child (cord and ALSPAC clinic blood at 7 years old) mother (pregnancy blood) and fathers (mouthwash samples). The ownership of samples is transferred to the University of Bristol. Consent for confidential DNA analysis has been obtained on over 10 000 children. Samples are stored in an anonymous storage facility. Any proposal to de-anonymize the samples would require the cooperation of the Director and the consent of the Ethics Committee. This latter has not so far approved any such proposal and is very unlikely to do so except in the most unusual circumstances. The identity of those who provided the sample is contained in a wholly separate storage facility. Access to both stores is very limited. Individual genetic data may only be fed back if they are of clinical relevance and if a beneficial therapeutic intervention exists. This will be considered on a case by case basis by the ALPAC Ethics Committee. In other respects, the Ethics Committee would not willingly agree to use the material for purposes which were not clearly stated in the original consent form. In practice, the study is in a great deal of contact with the vast majority of the members of the sample and it is not therefore excessively difficult to make further applications for the consent.

Some scientists in England have taken the position that in principle DNA samples should be treated as in the public domain and available to all scientists free of charge. In the case of ALSPAC, although the arguments in favor of worldwide access to knowledge free of charge are acceptable, these arguments are not applicable in the present context. Mothers who entered the study received specific assurances as to the use which will be made of information and samples that they have provided. It does not seem acceptable to disregard these assurances. It is reported that mothers would probably not have cooperated in the way they have if assurances of this kind had not been made. It is an integral part of the ALSPAC study that good relationships should be maximized with mothers and children during the whole of the very lengthy period during which the study will run.

deCODE genetics is another large scale DNA bank. A general description is available on website http://www.decode.is. The Act on a Health Sector Database makes legal for the Minister of Health to grant a license to deCODE genetics. The terms and conditions for this license are described in a document, which, like the regulation, is available on the Health Ministry website (http://brunnur.stjr.is/interpro/htr/htr.nsf/pages/forsid-ensk). It came out during the workshop that 58% of the population were in favor of the Act against 19.4% who were against it. The Health Sector Database that is built by Decode Genetics will contain medical data generated within the Icelandic health system and could be linked to genealogy, environmental, and genetic data that is obtained through individual informed consent. Even if a centralized health-care information has advantages, such a database raises issues on community consent, protection of privacy and freedom of science. No genetic data would be added to the database without informed consent by the individual. On the contrary, the medical data would be brought in without informed consent but individuals may opt out of the database at anytime. The ‘opt out’ clause has been criticized because patients who halt participation are not allowed to withdraw their data after entry into the database. As justification, it has been argued that the Act results from an informed democratic decision, but according a Gallup survey, only 13% of the population considered themselves to have a good grasp of the bill. It is argued that while the decision was democratic, if defined as a majority vote in parliament, such a decision should not supersede individual consent when it comes to participation in human investigations.

Conclusion

Data storage and DNA banking is receiving increasing attention as a result of the explosion of genetic research. Like in Iceland, large population-based studies have been set up or are planned at a national level in several countries including Estonia, Singapore, Tonga and UK. Consent, confidentiality, and coding are the key principles for DNA banking found in most statements. With experience in banking, with the advent of the creation of cell lines, with the need for international collaborative research teams to share DNA, and with a heightened sensitivity to patient consent to secondary uses, consent to banking was becoming more specific and personalized. Despite a consensus on requiring consent for subsequent use of identifiable or identified samples and an emerging consensus on subsequent use without consent of anonymous samples, appropriate consent requirements when anonymizing samples for retrospective and prospective studies remain controversial. Ambiguities due to the type of banked sample also prevail with respect to controlling and authorizing access to and sharing of samples with relatives or other researchers. For identifiable or identified samples, should conditions under which samples may be shared be clarified or should individuals always retain control on their samples? The feasibility of an individual to request that samples be destroyed is questionable and it may be preferable to ensure protection of the subject through other means. International standardization of ethical requirements and policies with regard to the use of DNA samples and information has been recommended.31 A such standardization would facilitate a greater protection of individuals as well as future international cooperation in biomedical research.

In fact, biomedical research has always transcended borders. Now the demand to transfer DNA samples and data across national borders is increasing. But such transfer will not occur fluently until a number of issues are resolved. There is a pressing need for an organization to take an international leadership and coordinating role. Several organizations, either the OECD, or the European Commission, or the WHO, or professional and trade organizations could be involved. The mandated organization could examine transborder samples and data flows. It could be concerned with the facilitation of samples and data flows as well as with data protection, ownership and legal accountability, and access to networking. It might also encourage collaborative efforts between the public and private sectors on these issues. As it has been recommended by the National Bioethics Advisory Commission,5 health-care organizations and professional societies should continue and expand their efforts to train researchers about the ethical issues and regulations regarding research on human biological materials and to develop exemplary practices for resolving such issues.

Proper protection of individuals relies on good regulatory mechanisms but also on transparency. Research that will ultimately benefit human health is crucially dependent on commercial involvement. Therefore access by the commercial sector – as well as academic researchers - to DNA samples might be facilitated. However, it is essential that research participants are made aware that their sample or products derived from it may be use by the commercial sector, and that they will not be entitled to a share of any profits that might ensue.4,15 As to researchers, when obtaining samples of human material from abroad, they must be satisfied that samples have been ethically obtained. Researchers should obtain from the clinicians providing samples assurances that they were obtained with proper consent in accordance not only with these guidelines but also with guidelines applicable in the country of origin. Furthermore, for the Medical Research Council, it is not appropriate for any one particular company to be given exclusive rights of access to collections of samples made with the benefit of public funds.

Recognizing that DNA samples are essential for biomedical research, it is also essential to respect the interests of those who participate as research subjects in regard to deciding control over and subsequent use of sample. Where human genetic material is considered as part of the ‘person’, providing it for research purposes is a consequence that is naturally assumed not to be associated with any financial benefit. In contrast, consideration of human genetic material as ‘property’ allows an individual to negotiate eventual commercial benefits, at least in the absence of legislation.62 With the exception of the United States (The Genetic Privacy Act 1994),26 most international, regional, and national bodies prohibit payment for the procurement of human genetic material or a possible payment or interest in eventual profits. They are contemplating instead the possibility of benefit sharing from downstream profits arising from genetic information. The acceptance of the notion of benefit sharing may certainly be seen as a recognition of the contribution of participating communities and populations.