Introduction

There have been some significant technological advances in genetics research in recent years that have importantly enabled the availability of genetic material and information for research into diseases and their possible treatments. Today, through clinical and research activities, it is possible to obtain, access and manage large amounts of genetic information, data that are enabling progress in research into various diseases and their treatments [1].

In reference to the acquisition and management of personal information and data, one of the ethical aspects that has been a subject of interest in the scientific literature of recent years concerns the criteria that need to be followed when informing the person involved. The prior consent of the person is required before any information or data can be used for scientific purposes. However, given that such information could be useful on later occasions or in subsequent research studies, the question arises as to whether that person’s consent needs to be sought again each time that information is reused or whether it is acceptable to seek an initial consent that includes future reuse, or at least the possibility of future reuse.

When the personal information involved concerns genetic material, and more specifically that person’s genetic sequence, such questions acquire special relevance. When the DNA is sequenced, it could be argued that the explanation given to the person concerned and the handling and management of that information obtained should be the object of more restrictive or special ethical criteria [1]

In effect, this is associated to a fundamental question as to whether genetic information should receive a treatment different to that given to other types of information when it is used for research purposes beyond that for which it was initially obtained (whether clinical or scientific) [1,2,3].

The present paper, the result of part of a research project developed by the Borja Institute of Bioethics (Instituto Borja de Bioética – Universidad Ramon Llull, Barcelona, Spain), aims to undertake a critical review of extensive literature, analyse these questions and propose ethical criteria and guidance to be taken into account when informing persons about the reuse in research of sequenced genetic data and when managing the information and the consent that an individual must grant before such data can be reused correctly in scientific research. The analysis carried out focuses only on the ethical framework, and the analysis of the legal framework is not the purpose of this work, since it presents different nuances in different countries and their jurisdictions.

The debate about genetic information

The first question that needs to be answered is whether genetic information should in fact be subject to greater protection and vigilance than other types of information. Many authors consider that this is indeed the case, often referring to what is known as genetic exceptionalism [3,4,5,6].

McGuire et al. [5] in reference to clinical practice though translatable to research, defend this exceptionalism, arguing that genetic information requires greater protection given its uniqueness, immutability, and identifiability. However, as McGuire et al. [5] or Martani et al. [3] acknowledge, it is also true that a person’s clinical history contains data that make them highly identifiable and that some of their non-genetic conditions are also immutable. In addition, a person’s clinical history also contains non-genetic information that can be highly sensitive, such as a diagnosis of multiple sclerosis or a sexually transmittable disease.

Nonetheless, like Evans [4], Mascalzoni et al. [7] or Slokenberga [6], we believe that genetic information appears to have greater sensitivity because it is related more than other types of information to an individual’s personality or to intimate or fundamental aspects of that person. Our genetics determine in part who we are and how we behave. Genetic information can and should require special protection when it can be associated to, for example, personality traits, certain behaviours, or psychiatric disorders. Additionally, the prediction of genetic risk can involve information that could significantly impact an individual’s self-perception and indeed their future life [3]. Certain genetic characteristics can, for instance, be subject to stigma or discrimination, something that is very often a significant source of concern and worry for the individual involved [4, 8, 9].

In general, transitory data or data liable to change (e.g. a person’s level of triglycerides) entail very little risk for the privacy of an individual, but fixed or permanent information (e.g. if the person is an HIV “carrier”) entails greater risk, and genetic information tends to be of the second type [10]. If that information is, in addition, uncommon, the potential harm for the individual is even greater because the “rarer” or “more unique” a datum is the more likely it is that the individual can be easily identified [9].

As rightly pointed out by Evans [4], defending exceptionalism seems to be more plausible in the sphere of genomics research than in the clinical sphere. It is important to bear in mind that genetics not only allows the identification of an individual but can also permit a prediction of that person’s risk of contracting particular diseases. This predictive power can be put forward as an argument in favour of exceptionalism, although it is limited in that there are genetic variants which, in the case of many diseases with which they are associated, are not the only determining element or factor and do not necessarily mean that the disease in question will emerge.

A person that offers a DNA sample, even if they usually consider that the sharing of data is beneficial for scientific research, also feels that such a process carries with it risks related to issues of privacy [11,12,13]. According to Khan et al. [14], when a subject makes a risk evaluation in this regard, loss of privacy is ranked highest, followed by uncertainty as to secondary use of their data and the possibility of some form of medical insurance discrimination. Of course, medical insurance discrimination can be avoided through other criteria or other measures (both ethical and legal), but, at this point, what is important is ‘trust’, that is, the fact that the professionals inspire confidence in the research subjects, so that they feel that professionals are truly interested in protecting their rights. A guarantee of confidentiality and privacy is key to ensure trust between the subject and the professional and between researchers and institutions [11, 14,15,16,17], and the professionals must protect that right, as well as other rights, even if the research subject is unaware of them or does not know how to exercise them [10].

All these factors and arguments acquire greater influence and importance if we consider that it is not only the individual subject that they may affect. Others may also be impacted, whether past, present, or future family members in general [3]. In this regard, it is important to consider the right of family members to be made aware of particular health-related data, and indeed their right to remain unaware of such data if they so prefer it [18].

In view of the above arguments, we consider that genetic information (in the clinical sphere and, above all, in the sphere of scientific research) should be subject to greater care and protection than other types of information.

Informed consent and the reuse of genetic data

Based on the most recent literature and scientific experience, there is no doubt about the advantages of the reuse (also called secondary use) of data in scientific research. Such use has many applications in all areas of healthcare: organisational, educational, commercial, forensic, qualitative, in public health or in disease control [3, 13]. Samples collected may come from research samples or samples collected for diagnostic purposes. In the first case, there will already be a consent to participate in a specific investigation and the question is what type of consent should be applied to extend these samples to secondary uses. In the case of samples from diagnostic purposes, a new consent must be considered for the field of research that contemplates secondary use.

In public health studies, for instance, the secondary use of data is beneficial because it facilitates long-term research in which the data can be repeatedly analysed, enriching, and completing results and prior conclusions [3]. An important advantage of the secondary use of data lies in its enabling large-scale studies in shorter time periods and with lower personnel costs. It also makes it easier to have large groups of study participants and long follow-up periods that allow more precise estimation of the incidence rate of the outcomes and long-term effects of interventions [13, 19].

From an ethical perspective, one of the most important problems with such reuse is the consent of the person offering the data and to what extent and exactly how the individual concerned should be informed about reuse of that data [20]. As Petrini [21] points out, many samples and biological data are obtained in situations in which the person has not given their consent or has not given it for future reuse. Conflicts arise in such situations in relation to questions of confidentiality, autonomy, research interests, the advancement of knowledge, and public trust in science and its institutions.

Different models of consent can be found in the literature when considering the handling and management of information in this context [15, 21,22,23,24,25,26].

  1. (1)

    Generic or blanket consent. The subject allows unlimited use of their samples for any type of research.

  2. (2)

    Broad consent. The subject is informed of the possible use of the information in future projects of certain types and for certain purposes. There is a requirement to inform the subject about the operations and management of the biobank for such research.

  3. (3)

    Specific consent. Consent of the subject is required in this model for each project and each use of their sample.

  4. (4)

    Dynamic consent. Using digital media (website or software application), consent is updated and granted, or not as new projects and uses of the information arise.

These are the most important models. There are others (like ‘tiered consent’ or ‘meta consent’), but they can be reduced to these four models [27]. Although the debate about the benefits and drawbacks of each of the above models continues [7, 23, 28], at the present time the literature is generally inclining towards broad consent [3, 15, 21, 29,30,31], with the basic arguments in its favour being as follows [3, 21, 24, 32, 33]:

  1. (1)

    It is difficult or impossible to obtain specific consent for the reuse of samples for each later study.

  2. (2)

    Even if it were possible or simple to do so, renewing contact with a subject to renew a specific consent (or to update a dynamic one) could be an annoyance and/or badly timed, potentially constituting an abuse or excessive intrusion with respect to the privacy of that subject.

  3. (3)

    Using only those samples for which a new specific consent has been obtained could introduce bias in the selection process and reduce the scientific validity of the study.

  4. (4)

    There is no important risk for the donor subject, or the risk is at least very low.

We think that these reasons (especially the second one) make broad consent more useful and advantageous than specific or dynamic consent. There have been positive experiences in the use of dynamic consent [34, 35], but these experiences do not yet represent the most usual and common experience in scientific research and require an intense level of communication between researchers and research participants. In fact, according to Petrini [21], the WHO and many European committees and institutions favour broad consent. For example, the Danish Board of Ethics clearly sets out that specific consent for each envisaged use of a sample is not necessary and that informing the subject of this possibility at the start of the process is sufficient. In addition, The Board states that the samples of a biobank can be transferred to other centres with different potential applications. Although restricting sample use to the centre where it was obtained reduces the ethical implications of informed consent, the possibility of its transfer between institutions favours and benefits research.

In effect, broad consent favours what Then et al. [25] call biobank networking: the sharing by researchers of samples from different biobanks with its significant benefits for biomedical research [36,37,38,39]. The technology presently available to store and analyse data is much more powerful and refined than in the past. As argued by McKeown et al. [31], the use in medical research of data platforms and machine learning, as well as the reuse and exchange of data among researchers, are trends or resources that are nowadays unavoidable and indeed necessary. For this reason, the concept of consent needs to be adapted and extended so that new technologies can make better and more beneficial use of existing data and generate new and, on many occasions, unexpected findings, something that is impossible if the consent is not broad in nature. It should also be noted that it is becoming more and more difficult to predict all the possible uses of certain types of data in the future [2, 31, 33].

One of the most important problems in the ethical and legal debate around biobank networking is precisely the most appropriate type of consent and the protection of privacy [12, 39]. In light of the rapid advances in data storage and analysis technologies, Khan et al. [14] are of the firm conviction that the privacy of a subject cannot be fully guaranteed, especially in the long term. In this respect, Byrd et al. [9] point out that future identification risks cannot be predicted and so it should be mentioned on the consent forms that, although some data may not be identifiable at the present moment, this may not be the case in the future.

Petrini [21] argues that broad consent is useful, acceptable and trustworthy provided that (1) personal data are handled in a confidential manner, (2) the subject has the right to withdraw their consent, and (3) the research has been approved by an ethics committee. In effect, in the opinion of Then et al. [25], the storage, secondary use and long-term sharing of samples should also always be supervised by ethics and scientific committees.

According to Petrini [21], such requirements lay bare the drawbacks of specific informed consent:

  1. (1)

    An excess of information about the aims, methods and risks of each research study can result in a subject having unnecessary concerns.

  2. (2)

    When a subject is exposed to important risks, consent is fundamental. However, if the risks are low or of only slight importance and the obtaining of consent could reduce the scientific validity of the study, consent may be considered unnecessary.

  3. (3)

    The advance of knowledge can be impeded by an excessive emphasis -or of little practical use- on the individual rights of a subject.

Although it constitutes a question of secondary relevance for the development of a research study, it is useful to reflect on whether the consent document should include the communication of secondary or accidental findings [1, 40]. Current guidelines support the communication of such findings if they are clinically valid and significant and if they may potentially benefit the subject in question [40, 41]. According to Pereira [1], most current guidelines concur in that the possibility of encountering secondary findings should be included in consent forms, with the option of allowing the subject to decide whether they would or would not like to receive any such results. Regarding the possibility of secondary findings and recontact, we believe it would be important to incorporate this discussion in the informed consent, as well as being able to grade the possibility of recontacting, for example, “only” in the event of serious secondary findings. In this case, it would be essential that recontact be made with people specifically trained in handling genetic information.

Likewise, the possibility not only of potentially serious secondary findings appearing should be incorporated into the consent, but also how these may affect the family, discussing the possibility or need to communicate these findings to family members. Currently genetic information is considered personal, putting the emphasis on patient’s confidentiality, however, if there is a risk of serious harm to the family, a harm prevention strategy may be necessary [42]. Informing family members of genetic risks requires a delicate balance between the ethical duties of respecting the confidentiality of the person and preventing harm to family members at risk, given both the personal and family nature of genetic information [42, 43]. It may be necessary to incorporate this discussion into the consent.

Khan et al. [14] argue, after a review of 30 studies, that subjects generally approve the secondary use of their genetic samples in future research studies, although determining exactly what such future research will consist of cannot be predicted with any exactitude.

It is our belief that broad consent is the most practical, adaptable, and trustworthy in terms of the reuse of genetic information and in particular sequenced DNA. This consent should be supported by appropriate measures, including a guarantee of confidentiality and supervision on the part of ethics committees. Only in this way can the ethical duties towards the persons from whom such information has been obtained be respected.

The next question that needs to be tackled is related to what information exactly should the consent deal with. Clayton et al. [16] and Martani et al. [3] argue that, in the sphere of research, a subject is more concerned about the possibility of being identified through their data than by the nature of the data, whether genetic or otherwise. That possibility is a factor of major importance for a subject when considering their decision as to whether to participate in a research study or not [12]. The use of different consent models for genetic and non-genetic information would complicate the process of drawing clear and easy-to-understand consent forms, which would have an impact on the transparency of and trust in the research [44]. In addition, it would entail non-expert subjects having to understand the differences between genetic and non-genetic data -and their respective implications- in the informed consent. The clarity of the consent forms, which is key to promoting a subject’s support and participation in a study, would be reduced.

According to the World Medical Association (WMA) [21, 45], in the framework of biobank data storage donor subjects should always be informed about the following:

  • the purpose of the biobank

  • their fundamental rights and guarantees

  • the risks and liabilities associated to data acquirement

  • data storage and use

  • the procedures for the delivery of results, including accidental findings

  • biobank access rules

  • how privacy is protected

  • that, in the event the data and material are not identifiable, it may not be possible for the person to know what is done with their data or material, and that they will not have the option of withdrawing their consent

  • when applicable, questions related to commercial use, including the distribution of profits and intellectual property

  • the transfer of samples, and therefore their custody, to other institutions or countries

However, as Then et al. [25] point out, in the case of broad consent, however detailed it may be, it is not possible to inform the subject about exactly what the sample will be used for or how it will be used in the future, as these are questions that cannot be predicted. We believe that this is indeed the case and that not taking this into consideration could make the idea of broad consent inviable. Nonetheless, it should also be noted that, if necessary, it is certainly possible -and indeed is an obligation that can be required- to inform and explain to the subject up to a certain moment in time what use has been made of their genetic information.

The protection of information, privacy and integrity

Protecting the genetic information and privacy of an individual not only depends on how that information is managed in the informed consent but also on the following two factors: firstly, the place and manner in which the information is stored, and secondly the nature of that information.

As for the place and manner of storage, it is important to note that the most used storage place of genetic information at the present time is not a biobank but rather large local repositories, located in technical assistance centres (servers, hard discs…). Management of these repositories is the responsibility of the centres in which they are housed. In such cases, the control systems related to access to this information tend a priori to be less precise or formal in comparison to those of a biobank [46].

Consequently, there is less protection in terms of information reuse, or the rules and procedures in this respect are less strict than those applied to a biobank. In order to have access to the reuse of information stored in a biobank, such access must be framed within a research project, and that project must have been endorsed by a scientific committee and assessed by a research ethics committee [37, 47]. Such requirements are not obligatory in the case of access to repositories managed by clinical or scientific institutions.

Given the above, it is our conviction that genetic information should be ‘biobanked’. By using this expression, we do not mean that such information should be stored in biobanks but rather in local repositories which have safety standards similar to those of biobanks. Such repositories could be ‘federated’ as part of a repository network, in such a way that, with due security and authorisation measure, researchers from a particular centre could access the anonymized information stored in other repositories of that network (which is in essence the spirit of the previously commented biobank networking). We believe that this ‘biobanking’ process has two major advantages:

  1. (1)

    The genetic data -which as previously commented must be subject to special care- would have high standards of protection.

  2. (2)

    Thanks to the establishment of ‘federated’ repository networks, researchers could have easier access to the reuse of a much larger amount of data.

As for the nature of the information, we consider that this factor needs to be taken into consideration when evaluating the convenience of its reuse. The reuse of certain genetic information may be injurious to the rights of the individual [48], and we believe that, in such cases, reuse should be limited and that this circumstance should be reflected in the informed consent. There are cases in which, regardless of the original use and purpose of the data, reuse of that information could have negative and harmful consequences for the individual.

In such cases, we consider that reuse should largely be limited or even prohibited, with the individual, when signing the consent form, being made aware of the possibility that such situations could potentially arise and being offered the guarantee that their data would not be used should they in fact arise. For instance, in the case of genetic information from persons with very rare diseases: in these cases, reuse of the information could easily lead to their identification.

Conclusions

An ever-growing amount of genetic information is being stored and managed for the purposes of scientific research. To enable such research to progress, the information that is stored needs to be reused beyond the primary use and purpose for which it was obtained. However, this reuse must comply with certain ethical criteria to ensure the protection of the dignity, integrity and autonomy of the individual from whom that genetic information was obtained.

Considering the review conducted in the present study, we consider that these criteria should be as follows:

  1. (1)

    Genetic information requires special care and protection (genetic exceptionalism). Because of its nature and its implications, it should be treated differently to other types of information. ‘Biobanking’ the repositories in which such information is stored would improve the protection required.

  2. (2)

    Broad consent is the most practical, adaptable and, at the same time, trustworthy type of consent for the reuse of genetic information

  3. (3)

    There must be a guarantee that any reuse will be limited or even prohibited in situations which could potentially result in consequences that are injurious to the rights of the individual.