Main

Stored tissue sample collections for genetic research exist in different forms. Some of these collections provide a resource for potentially unlimited genetic research, and gather samples and data from specific populations. An example is the ‘UK biobank’.1 Other collections are stored for research on a specific disease. Collections that were originally gathered for different purposes, for example blood spot cards for newborn screening, could be reused for genetic research.2

Genetic research on biological material from minors and the associated medical records can yield valuable information on the development and genesis of early-onset genetic disorders and the early interaction of environmental and genetic factors. For example, Rasmussen3 describes the incorporation of DNA sample collections into the ‘National Birth Defects Prevention Study’ in the United States to identify the risk factors for birth defects. Studies such as the ‘Avon Longitudinal Study of Parents and Children’ in Bristol (children of the nineties) use genetic, phenotypic and environmental information of 14 000 babies from their conception onwards to study the interaction between these data.4

An extensive ethical literature exists on the collection, storage and use of biological samples for genetic research. The overwhelming majority of these documents discuss issues of privacy, confidentiality, commercialization and consent.5, 6, 7, 8, 9, 10, 11, 12, 13, 14 However, research on pediatric data raises specific ethical questions with regard to consent and privacy. For example, who should give consent to the inclusion of tissue and data from children? Is the general requirement that non-therapeutic research can only be done with children if it involves no more than minimal risk, applicable to biobank research? We shall review whether and how guidelines and policy documents discuss children in the context of storing biological samples and DNA for non-therapeutic research.

Methods

MedLine, Embase and Google Scholar were used as a primary source of information to identify relevant literature. Official websites of ethical committees, professional organizations and regulating bodies from the US and the European Union were equally searched.15, 16 More information was gathered on the basis of the bibliographical data in these documents. We focused on documents about genetic databases (so-called biobanks) and about stored biological samples that mentioned genetic research. Documents discussing archived human tissue without mentioning genetic research were discarded. General documents on genetic research were preserved, as long as they at least mention banking of data. Only documents in French, German, English or Dutch were preserved, that were no older than 1990. As the focus was on guidelines and recommendations, legally binding documents were not included.

Some confusion may exist regarding the term ‘children’. We have included documents that mention ‘children’ or ‘minors’ and have covered the lifespan from birth till the legal age of maturity. In the text, ‘children’ and ‘minors’ will be used synonymously.

Results

Guidelines and position papers

We have retrieved 29 different guidelines and position papers (Table 1, numbered references to this table) pertaining to the storage and use of biological tissue samples for genetic research. The documents covered the period 1994–2007, with a peak in number of guidelines in 2002 (7 out of 29 guidelines). The guidelines originated from 27 different organizations. Guidelines were issued by national bioethics committees (13), medical associations (6), genetic associations (5), other bioethical associations (1), UNESCO (1) and the Council of Europe (1). Five documents had an international scope, three had an European scope and 21 had a national scope: United States of America (4), United Kingdom (3), Germany (2), Australia (1), Iceland (1), The Netherlands (1), Finland (1), Israel (1), Singapore (1), Canada (1), France (1), Ireland (1), Greece (1) and Austria (1).

Table 1 Overview of sources used
Full size table

One international guideline (11) dealt with biomedical research in general, but as it contained a section on stored tissue samples, we included it as well. One document (4) dealt with clinical investigation, but was included because the document from the same organization (AMA) that deals with DNA databanks (16) explicitly refers to this document for vulnerable persons.

At the moment of writing of this paper, the OECD (Organization of Economic Co-operation and Development) is working on a document provisionally titled ‘Guidelines for Human Biobanks and Genetic Research Databases’. As this is still open for comments, we have decided not to include the draft version of the document in our discussion.

We first acknowledged whether the guidelines mentioned the use of tissue samples from minors. It turned out that 11 out of 29 documents did not mention such use. We then categorized the remaining 18 documents according to the themes they discussed with regard to this topic. We distilled four major themes: consent, principles of non-therapeutic research on vulnerable populations, ethics committee approval and difference between anonymous and identifiable samples (Table 2).

Table 2 Content analysis
Full size table

Consent

A major theme discussed is the issue of consent. First of all the guidelines discuss who should give consent to the donation, storage and use of the tissue samples from children. Some guidelines also discuss the issue of a child's assent (permission) or dissent (refusal). A third question is whether and when a minor should be recontacted to give full consent.

17 out of 18 documents state that a legal guardian should give consent. In this regard, a legal representative, guardian or legal proxy is named (1, 4, 6, 11, 14, 17, 19, 20, 22, 23, 26, 27, 28, 29), a parent (11, 23, 27, 28, 29), someone with parental responsibility (2, 9), a relative (18, 28) and someone designated to protect the interests of the subject (18). Four mention parents (9, 22, 27, 6). Only one guideline mentions that it must be both parents (6): ‘In general consent should be obtained from both parents’.

Only four documents give further specification on the representative who gives consent. CIOMS (11) mentions that the parent or guardian who gives permission for a child to participate in research should be given the opportunity, to a reasonable extent, to observe the research as it proceeds, so as to be able to withdraw the child if the parent or guardian decides it is in the child's best interest to do so. The Bioethics Advisory Committee, Singapore (14) mentions that the extent and scope of a given legal proxy's lawful authority to give consent may well depend on the particular circumstances and on the putative proxy's legal relationship with the donor. The French Comité Consultatif National (18) stresses that although there is possibly no other solution than to gather consent from some trusthworthy person, one cannot ‘ignore that a collection based on samples from children necessarily carries knowledge about a whole lifetime’. Hence, they question the extent of such proxy consent. UNESCO (19) states that ‘the legal representative should have regard to the best interest of the person concerned’.

Should a minor's decision also be taken into account, next to the proxy consent by the parent(s) or legal guardian? 11 out of the 18 documents discuss the issue of assent (permission) and dissent (refusal). All 11 agree that, when appropriate, the child should be consulted and their agreement should be obtained (2, 6, 9, 11, 17, 20, 19, 22, 23, 27, and 28). Some documents state that a procedure should not be carried out if a child objects or seems to object to the procedures (2, 11, 17, 20, 22, 23, 27, and 28). In this respect, CIOMS (11) makes a distinction between the deliberate objection of an older child and ‘the behavior of an infant, who is likely to cry or withdraw in response to any stimulus’. Therefore, older children, in whom an apparent objection can be more easily evaluated, should be preferred over younger children. CIOMS also states ‘that a deliberate objection by a child to contribute to research should always be respected, even if the parents have given permission, unless the child needs treatment that is unavailable outside the context of research.’ The Australian National Health and Medical Research Council (6) mentions the concept of the mature minor. If the child is a mature minor, his or her consent should be sought in addition to that of the parents or legal guardian. CIOMS (11) mentions that assent alone is insufficient to permit participation in research; additional consent from a legal guardian must be sought.

Should children be recontacted to give their autonomous consent for research on their samples? Eight documents think so (1, 9, 23, 24, 6, 11 and 29). The criterion when this should happen varies however. Two guidelines mention ‘when they are old enough to understand/comprehend or when they are capable of discernment’ (1, 9, 23, and 24). CIOMS (11) mentions ‘If such research subjects, including children, become capable of giving independent informed consent during the research, their consent to continued participation should be obtained’. They do not specify whether they mean a legal age or a level of maturity. Two documents (6, 29) specifically state that reconsent must occur when the child is legally capable of doing so.

Principles of non-therapeutic research on minors

A second topic is the reference to general principles of non-therapeutic research on children. These principles include first the fact that such research should not entail more than minimal risk, second that it can only be done to benefit persons of the same age or condition, third that they should have some direct benefit to the participant, and fourth that they can only be carried out if the same research cannot be done on adults.

The first general principle with regard to non-therapeutic research on children that is quoted in the context of stored tissue samples is that of minimal risk. Eight documents mention this risk (2, 11, 17, 20, 21, 22, 23 and 27). MRC (9) does not explicitly mention the principle, but states that ‘Tests of known predictive value for adult onset diseases should not be done for research purposes on individually identifiable samples from children’, thus quoting a specific interpretation of risk. CIOMS (11) gives a definition of what minimal risk might entail: ‘the risk from research interventions that do not hold out the prospect of direct benefit for the individual subject should be no more likely and not greater than the risk attached to routine medical or psychological examination of such persons’. They make a distinction between the experiences of healthy children and those with a condition that would already require interventions as lumbar punctures or bone-marrow aspirations. Those would be ruled out for healthy children. Nationaler Ethikrat (22) mentions residual material from therapy or diagnosis: as research on such material does not require extra bodily interventions, the risk is considered minimal.

The next three principles are a reflection of the WMA's Declaration of Helsinki (1964),45 that states that ‘these groups should not be included in research unless the research is necessary to promote the health of the population represented and this research cannot instead be carried out on legally competent persons’.

The second principle states that research should be beneficial to minors with the same age or same condition. This principle is quoted by five documents (11, 17, 20, 22, 23). The WHO (20) states, however, that data should be coded to prevent identifiable links being made, and that ‘such permissions would only normally be granted to the direct clinical benefit of the child’. Nationaler Ethikrat (22) quotes a controversy surrounding the concept of ‘group benefits’. They state that ‘some hold that, if research is of minimal risk and done on residual material from therapy or diagnosis, non-therapeutic research can be done on subjects incapable of giving consent for the benefit of fellow-sufferers. Others hold that consent is a personal matter and must be left to those concerned, and as it is impossible to determine whether risks are minimal, research on persons not able to consent should be prohibited’. The Nationaler Ethikrat does not, however, give a definite solution to the controversy.

A third principle that research on subjects who cannot consent should only be done if there is any direct benefit to participants, is quoted by four guidelines (11, 17, 20, 22). For example, the Nationaler Ethikrat (22) states that ‘No one disputes that those incapable of giving consent may be involved in research from which they themselves are likely to benefit therapeutically’. None of the documents give examples as to how this can be interpreted in the context of stored tissue samples, as probably no such examples exist.

A fourth principle of research on children is that it should only be done if it cannot be done on adults. This is mentioned by six documents (1, 2, 4, 11, 22, 23). For example, the Irish Council of Bioethics states the condition that ‘the research relates directly to a clinical condition from which the minor concerned suffers and/or is of such a nature that it can only be carried out on minors’. The other five documents have similar provisions.

Ethics committee approval

Seven guidelines mention the need for ethics committee approval for the inclusion of children in research (2, 11, 14, 20, 21, 29 and 27). The circumstances under which such approval is needed vary.

Four guidelines mention the need for ethics committee approval in general for research involving children (2, 20, 21, and 27). CIOMS (11) mentions the need for such approval when there is a slight or minor increase above minimal risk. The Bioethics Advisory Committee, Singapore (14) mentions that ‘the standard protocols and forms to be used in the taking of consent given on behalf of incompetent donors by their legal proxies should be reviewed and settled by the institution's ethics board or institutional review board, acting with the advice of the institution's legal advisors. The same ethics boards and legal advisors may also be consulted for a review of consent formalities in research projects for which it is anticipated that a significant proportion of the donors are or are likely to be legally incompetent’. NCI (19) mentions that the IRB should review reconsent issues when the child reaches the legal age to consent, at the time the board reviews the initial protocol.

Difference between anonymous and identifiable

Three documents make a distinction between the use of anonymous and linked biological samples from children (9, 20, 29). MRC (9) states that ‘tests of known predictive value should not be done for research purposes on individually identifiable samples from children’, without further explanation. WHO (20) states that ‘some research will require the linking of clinical and genetic data in order to proceed and that the main beneficiaries of this research may be future children rather than the child who provides the sample. In such cases data should be coded to prevent identifiable links being made with access to the key to the code being restricted and subject to separate permission on each occasion’. The NCI (29) mentions the need for reconsent when a child reaches the legal age to consent if identifiable samples are used.

Discussion

Of the 29 documents that mention stored tissue samples, 18 mention special provisions when these samples originate from children. These provisions include thoughts about consent, assent, reconsent, minimal risks, group or individual benefits, ethics committee revision and the difference between anonymous and identifiable stored samples. We have observed that these themes are not consistently mentioned by each document. Documents that discuss the same themes were mostly in agreement in their recommendations. The guidelines do not provide specific recommendations regarding ownership of tissues, return of results or recommendations when minors are concerned, probably because they do not consider these different from issues regarding tissue from adults.

Consent

Most policies agree that an informed consent must be obtained when biological samples and genetic material are collected for research. Moreover, 15 documents state that this should be written (4, 5, 6, 7, 8, 9, 10, 13, 17, 21, 25, 26, 27 and 28). When minors that are incompetent, be it through lack of maturity or because they do not have the legal age to consent, are concerned, this is problematic. 17 guidelines mention that a legal guardian or parent is the most suitable person to consent. However, the nature of genetic material might suggest that, if possible, both parents should consent, as their genetic material is equally present in the child. This is suggested by Bauman.46 However, in a society in which many children do not live with both (biological) parents, this may be not feasible.47

With regard to assent, guidelines remain vague regarding at which age a child should be asked for his or her assent and whether the criterion is one of maturity or age. Most guidelines use general terminology such as ‘his or her opinion should be taken into account depending on age and maturity’ (28). Only one document mentions the age of 12 or 13 years old, in its general section on biomedical research involving minors (11). This uncertainty is reflected in the literature on assent: there is no consensus at which age a child should be allowed to assent for non-therapeutic research, and whether this age should be fixed,48 or is dependent on social context and personal experience.49 Moreover, Alderson has shown that already very small children are able to show objections to certain practices. It is fruitful to look at the United Nations Convention on the Rights of the Child, that asserts that children have a right to say what they think should happen when adults make decisions that affect them (Article 12) and have the right to get and share information (Article 13). As this is a legally binding document for most countries, it makes sense that in longitudinal genetic research, small children are also given appropriate information and their opinions are taken into account.

What should be the scope of parental consent in time and in content? Literature supports that it is reasonable to recontact minors when they are able to do so and give them the opportunity to withdraw their data, which is there only by proxy consent.50 However, none of the guidelines studied is explicit about when this should happen: when they are able to understand the impacts of research or when they reach a certain age? An overview of legal regulations regarding the position of minors in a health care setting in the EU member states has shown that age and circumstances under which minors are allowed to take health care decisions vary in different countries.51 In some countries, the age of medical majority is the same as the legal age of majority (Cyprus, Greece and Slovakia). Others distinguish between the medical and legal majority (Portugal, Slovenia, Lithuania and Spain). The Czech Republic and Estonia consider medical majority on a case-by-case basis, dependent on age and maturity.

Another aspect of the scope of consent is the question of content of the consent. Are parents allowed to consent to any future genetic research on their children's DNA? This is called broad consent.52 Or can they only give specific consent to research on specific genes and diseases? Most biobanks have as an aim the accommodation of future research, the nature of which is undetermined at the moment of storage. Hansson53 has argued that broad consent at the time of storage is sufficient to keep the data available for research for a long period of time. In the case of children, however, it is not the donor herself who has consented and it may seem fair to restrict proxy consent only to specific research protocols or research on certain genes or diseases.

Principles of research on children

Most would consider biobank research in general as minimal risk.9 There is probably very little harm involved in the taking of a sample of blood of an adult to be included in a biobank. The harm is mostly situated in the area of possible stigmatization of certain groups through genetic research and the risk of loss of privacy.9 However, for children, procedures such as having blood taken and visiting a hospital or research center may well be frightening. The amount of anxiety that such procedures cause is probably dependent on the personality and experience of each individual child, and the question whether a child should be enrolled or not in non-therapeutic research is hence one that has to be made on a case-by-case basis.

The issue of privacy, which is often quoted with regard to biobanks in general, is a difficult one when children are concerned. It can be argued that the privacy of an infant is not that important and that his or her medical information is more or less public property. Most parents discuss a variety of health related issues about their baby with different people; they would probably refrain from doing this if an older child was concerned. Indeed, the need for privacy grows as the child grows older, and what a child considers private is in many cases different from what an adult would be sensitive about.54 Some consider genetic data as a ‘future diary’, a term framed by Annas:55 we would not find it acceptable to publish a diary of a ten year old. Moreover, the DNA of the child is the same as the DNA of the adult it is to become. The combination of the static characteristics of genetic information together with the evolution of the child towards maturity makes the question of privacy and minimal risk in biobanks that store pediatric samples a difficult one to answer. The topic of privacy and children was not explored further by any of the guidelines under consideration.

A next requirement is that of who should benefit from non-therapeutic research on tissue from children. Although some people acknowledge that there could be direct benefits to the individual from being enrolled in biobank research, such as benefits from regular health checkups,56 this is not the primary aim of the research. Therefore, it is difficult to envisage what direct benefit would mean in this respect. On the criterion of group benefits, Holm50 has argued that these are too restrictive and not well founded as criterion, and that parents should be allowed to consent to participation of their children regardless of these benefits. However, in the context of proxy consent and broad versus specific consent, it may seem reasonable that proxy consent is not given for research on the entire genome of a child, but restricted to research on specific genes or conditions, which would satisfy the requirement of group benefits as well.

Ethics committee review and anonymization

No consistent overview of the issues that an ethics committee should take into account when considering research on children could be found. Moreover, it is remarkable that only a subset of guidelines (7 out of 29) mentions this specifically with regard to research on minors, as it could be deemed good practice.

The Declaration of Helsinki45 considers research on identifiable human materials as human subject research, but can this be generalized to the use of anonymized samples? If research on anonymous samples is not considered human subject research, maybe the provisions that are put on research on tissue samples from children do not apply. In the context of risk, some authors in ethical literature have argued that the risk of privacy breaches is removed when anonymous samples are used.57, 58 However, others such as Lysaught,59 point out to the possibility that genetic research may lead to group stigma, and hence indirectly may affect the individual. Moreover, subjects may feel strongly about the type of research that is done on their tissue, regardless of the level of identifiability, as mentioned by Trouet.60

The storage and use of biological samples from children for research poses some specific questions that are not covered by ethical reflections on the use of samples from human beings in general, such as who should give consent, whether the opinions of the minors matter, to which risks children can be exposed and who should benefit from such research. The majority of the 29 guidelines does reflect this distinction and mentions children in a separate section. However, a systematic reflection is missing and should be subject to further study. Such study could also shed some light on existing general discussions with regard to the ethics of biobanks, most notably the issue of informed consent.