For many years, physicians at a cancer clinic have been storing biological samples left over after being used for diagnosis in clinical testing. Prior to 2000, no consent for storage or research was obtained. In 2000, the clinic changed its policy and began to systematically request consent for the use and storage of leftover biological samples “for future cancer research.” From that point on, the clinic has been storing samples only when the patient consented. It discards samples when the patient does not consent. Many of the sample donors are still alive (some are still patients at the clinic), while others have died. The clinic now has over 4000 samples, with comprehensive clinical data. Two groups of geneticists would like to use the samples for research, one examining the genetic basis of certain cancers, and the other examining the genetic basis of ethnicity and drug response in a randomized, heterogeneous population study.
Our knowledge of genetics has largely transformed the manner in which biomedical research takes place. From the Human Genome Project and the International HapMap Consortium, we now know the sequence of the human genome, and we have created a haplotype map of the human genome, describing the common patterns of haplotype ancestry. While in the past genetic research tended to focus more on the identification of single genes that follow a Mendelian pattern of inheritance (i.e., presence of single gene being both necessary and sufficient to cause the disease in question), there has been a shift in interest to the search for genetic risk factors in common diseases (e.g., cardiovascular disease, cancer, diabetes), as well as to pharmacogenomics research (understanding the role of genetic variation in individual drug response) and to studies of normal genetic variation across entire populations (Knoppers and Sallée, 2005).
Researchers need access to systematic collections of tissue or fluid samples and related clinical data (bio-banks) to maximize their research progress in such areas, ultimately leading to better understanding of the role genes play in health, disease, and interactions with the environment (World Health Organization, 2003; Canadian Biotechnology Advisory Committee, 2004). Despite some controversy regarding their scientific utility, most believe that bio-banks are both important and extremely useful (Barbour, 2003; Finkelstein et al., 2004; Foster and Sharp, 2005). Bio-banks differ in a number of respects, including the number of samples collected, the types of sample collected (disease specific/general, prospective/archival, individual or family/populations), the degree of identifiability of the samples, the range of possible or permitted uses, the status of the institution(s) in charge of their constitution and management (public/private), and the sector in which the samples were collected (clinical, research, forensic, etc.).
In this chapter, we discuss a key ethical issue that, although important for all research, has particular considerations in the context of bio-banking for clinicians, namely consent. We cover the general situation plus that applicable to deceased individuals and approach the issue from a variety of national and international perspectives. Related topics such as return of research results and commercialization are beyond the scope of this chapter.
Since the Nuremberg trials following World War II, international ethics protocols have been designed to protect human subjects who take part in medical research. The Nuremberg Code, the first international instrument on the ethics of medical research, established that, “voluntary consent of the human subject is absolutely essential” (Nuremberg Code, 1947; World Medical Association, 1964; Council of Europe, 2005). Voluntary consent requires the free (i.e., uncoerced) decision of a legally capable individual who has “sufficient knowledge and comprehension of the elements of the subject matter involved as to enable him to make an understanding and enlightened decision.”
Informed consent is now the touchstone of ethical biomedical research and is codified in many policy documents, including those of the World Medical Association, the United Nations Educational, Scientific and Cultural Organization (UNESCO), the Council for International Organizations of Medical Sciences (CIOMS), the Human Genome Organisation (HUGO), the World Health Organization (WHO), and the Council of Europe. This principle is equally applicable in the context of genetic research (UNESCO, 2003, 2005; Canadian Institutes of Health Research, 2005; Council of Europe, 2005), as in the even more specific context of bio-banking (HUGO, 1998; WHO, 2003; Council of Europe Steering Committee on Bioethics, 2006). In each case, the underlying values that informed consent serves to protect are the same, and include dignity, autonomy and privacy. However, because of factual differences between the different types of research, how these values are weighted against the competing value of facilitating research that benefits humanity does change. For research on biological samples, as opposed to research on a human subject directly, there are minimal, if any, physical risks to the research subjects; the potential harms relate to human dignity (e.g., unconsented use of specimens that goes against religious or personal beliefs) and individual (e.g., insurance) or group-based discrimination (Rothstein, 2005). For populational bio-bank research, the paradigm shifts even further: traditional research on biological specimens generally involves a single researcher or group of researchers obtaining and using samples in defined ways to research a discrete area. In contrast, population bio-banks often involve entities that obtain the sample but are not directly engaged in research, who supply specimens to other researchers. The purpose of a population bio-bank is to develop a “resource” that can be used for many research protocols, often in numerous scientific areas and in ways that cannot be foreseen at the time of collection (Canadian Biotechnology Advisory Committee, 2004). The movement towards population-based genetic research (requiring infrastructures such as bio-banks) has led to a concurrent movement to rethink the “paramount position of the individual in ethics” (WHO, 2003; Knoppers and Chadwick, 2005).
In general, consent requirements will vary with the degree of identifiability of the sample and the associated data. This makes sense: the weaker the link between sample and donor, the lower the chance of harm to the donor. It must be noted that there is considerable confusion in the terminology used to describe the identifiability of genetic samples (Knoppers and Saginur, 2005; US National Bioethics Advisory Committee, 1999). For the purpose of clarity, we use the term anonymized to refer to samples that were originally identified or coded but are then stripped of all possible identifiers. Coded is used to refer to samples that are identifiable only through breaking the unique (single coded) or the two unique (double coded) codes given the sample. Double coding involves a keyholder who can link the two codes when necessary for research or clinical purposes.
After over a decade of inconsistency and uncertainty, certain trends are beginning to emerge in this area. First is the increasing recognition of the validity of waiver of consent for secondary research on double-coded specimens and data, and second is the increasing acceptability of broad consent for population projects.
At the international level, HUGO, CIOMS, WHO and UNESCO all advocate similar approaches. Consent may be waived for secondary uses of samples if the samples are anonymized or if they are double coded provided certain conditions are met: there is a general notification of such a policy and the patient has not objected (HUGO, 1998), patient confidentiality is protected and research ethics board approval is obtained (CIOMS, 2002), no future identification is possible of the sample source (WHO, 2003), and the data has medical or scientific significance and research ethics committee approval has been obtained (UNESCO, 2003).
At the regional (European) level, the Council of Europe allows waiver of consent for research uses of coded samples if an independent evaluation finds that the research addresses an important scientific interest, the aims of the research could not reasonably be achieved using biological materials for which consent can be obtained, and the individual did not expressly oppose such research use (Council of Europe Steering Committee on Bioethics, 2006). The European Society of Human Genetics (2001) considers that the consent requirement can be waived when samples are anonymized (rec. 9), and, provided it is approved by an ethics committee, in situations where the collection can be considered as abandoned (rec. 14). For collections of coded information, while in principle re-consent of participants for new studies is necessary, ethics review committees can waive the requirement for such consent when re-contact is impracticable and the study poses minimal risks (rec. 12).
At the national level, different countries have taken different approaches to regulating consent requirements, yet the actual content of the norms are beginning to converge. The UK (Human Tissue Act, 2004; Parry, 2005), France (Loi du 6 août, 2004), the USA (US Department of Health and Human Services, 2003a), Canada (Tri-Council Policy Statement, 1998), and Germany (Nationaler Ethikrat, 2004), for example, all hold that re-consent is not required when samples are anonymized. With respect to coded samples, the UK allows research on samples that are not anonymized where reasonable efforts have been made to obtain the consent of the donor (Human Tissue Act, 2004). In France, the secondary use of samples removed during medical care is permissible provided the donors have been notified of the secondary use and have not objected to such use (Code de la santé public, 2004). Further, the obligation to inform individuals can be waived if it is impossible to find the person, or when an ethics committee is consulted by the research investigator, and concludes that such information is not necessary (Code de la santé public, 2004; L. 1123–1). In the USA, consent is not required for secondary research uses of double-coded samples, if there are assurances (either through private agreement, institutional review board policies, or other legal requirements) that the keyholder will not under any circumstances release the key to the investigators until the individuals are deceased, as once deceased this is no longer considered “human subject research” (US Office for Human Research Protections, 2004). For identifiable samples, an institutional review board may alter or waive the requirement of informed consent if the research involves no more than minimal risk, the waiver or alteration will not affect the individual’s rights and welfare, and the research could not be carried out without waiver or alteration of consent requirements (US Department of Health and Human Services, 2005).
In Canada, research ethics boards may waive some or all consent requirements if the research poses no more than minimal risk to the subject, the waiver is unlikely to adversely affect the rights and welfare of the subjects, and the research could not practicably be carried out without the waiver (Tri-Council Policy Statement, 1998). In the context of secondary use, when determining whether consent may be waived in a given circumstance, a research ethics board must consider the following factors: the necessity of the personal data, whether potential harm to individuals is minimized and potential benefits of the research outweigh potential harms, whether seeking consent is inappropriate (e.g., psychological harm, risk of threat to privacy, or contact with individuals not permitted under a previous data-sharing agreement, law or policy) or impracticable, what the individuals’ expectations are (no previous objections to the secondary use and expectations of a reasonable person), and what the views of relevant groups or communities are (Canadian Institutes of Health Research, 2005). Finally, in Germany, the balancing of patients’ rights and freedom of research was addressed in the 2004 opinion of the German National Ethics Council. It concluded that an ethics committee can waive consent requirements when samples and data are double coded provided researchers do not have access to the code (Nationaler Ethikrat, 2004). Further, consent can be waived for research on identified data and samples when donors’ interests are outweighed by the scientific importance of the research and the research cannot proceed otherwise or can proceed only at too high a cost, and disproportionate efforts (regulatory proposals 3 and 4).
Bio-bank research is moving beyond the one study/one informed consent model to a format of obtaining general or broad consent to participate in the research activities of a bio-bank (Rothstein, 2005). Allowing for broad consent eliminates the need to determine whether a waiver of consent should be granted in a specific instance. HUGO, in its Statement on Human Genomic Databases (2002) held that “[i]nformed consent may include notification of uses (actual or future) … or, in some cases, blanket consent. The CIOMS (2002) implied an acceptance of broad consent by stating “the original consent process [must] anticipate, to the extent that this is feasible, any foreseeable plans for future use of the records or specimens for research.” The Council of Europe similarly allowed for broad consent, yet emphasized that the individual “may place restrictions on the use of his or her biological materials” (Council of Europe Steering Committee on Bioethics, 2006).
While not all laws or policy documents address the issue of broad consent, those that do tend to endorse it. For example, the Canadian Biotechnology Advisory Committee (2004) held that, for bio-bank research to be most beneficial, consideration should be given to establishing an “authorization model” of informed consent specifically for prospective population genetic research. This model would require consent for the initial collection of the biological sample. Authorization of subsequent research would be given (or denied) by the donor at the time of the initial sample collection. Individuals must be able to specify which uses of their biological material and associated data are permitted or excluded as well as the degree of subsequent decision-making authority they want to retain; and individuals must have the option to give general or “broad consent” to any and all future uses. The German National Ethics Council went further. It recognized the necessity for archived samples (obtained during diagnosis and treatment) to remain available for further use and held that a “form-based” broad consent should be obtained at the time of collection and would be sufficient and so it would not take the “option” approach (Nationaler Ethikrat, 2004).
Despite this general trend towards more permissive rules governing secondary use of biological samples and consent, there remain jurisdictions that require a specific consent. For example, (though being re-evaluated), Sweden holds that “[t]issue samples preserved in a bio-bank may not be used for other purposes than those indicated in information submitted previously for which consent has been granted. In the event of a new purpose, the person who previously granted consent must be informed about the new purpose and grant new consent” (Swedish Ministry of Health and Social Affairs, 2002). However, overall there has been a gradual understanding by national policy makers of the difference between the degrees of identifiability of samples and data and corresponding levels of research access for secondary uses. There is also a move away from requiring an explicit re-consent for all secondary uses provided other safeguards are in place (i.e., through double coding, anonymization, data steward, research ethics board approval, etc.).
It is widely held that the death of a person does not extinguish the interests of that individual. Family members, and others who have physical possession or access to an individual’s body, tissue or cells, have to respect certain obligations and rights following the death of the individual. The WHO (2003) does hold though that death affects the primacy of this interest and allows for the possibility, through appropriate ethical approval, of readjusting the balance of interests in light of death.
At the regional level, the European Society of Human Genetics (2001; rec. 13) holds that postmortem uses of samples are subject to the advance wishes of the donors. In the absence of any known wishes, use of those samples should be regulated, a policy of unfettered use not being ethically justified. The Council of Europe Committee of Ministers (2006) did not explicitly differentiate between research on archived human biological materials based on whether the sample source is living or not. It simply stated that postmortem uses have to meet satisfactory information and consent measures. Finally, the European Commission (2004) recommended allowing samples from the deceased to be used for research provided the sample is anonymized.
At the national level, there is variation with respect to whether, and if so under what circumstances, research can be performed on biological samples from deceased individuals. National positions cover the range from the theoretically unlimited power of officials to “deem” consent from the deceased to the restrictive position that essentially disallows research on identifiable samples from the deceased unless the deceased previously consented to that research use (Genetic Privacy and Nondiscrimination Act, US Department of Energy, 2003). Yet, these differences have few practical consequences. Equally notable is the number of documents that do not address the issue at all.
In the UK, the Human Tissue Act 2004 s.4(a)–(e) provides for powers to dispense with the need for consent by providing a mechanism for “deeming” there to be appropriate consent for a research use. In theory, therefore, there are wide powers to perform research on biological material from the deceased. In France, while in principle biomedical research on a deceased individual can take place only if the individual had expressed consent to such research while alive, or if family members testify to the existence of such wishes (Code de la santé publique 2004; art. L. 1121–14), there is a key exception to this principle: the postmortem collection of cells, tissues, and human body products is allowed for therapeutic or scientific purposes if there is no prior opposition (art. L.1241–6). In the USA, research on tissue samples of deceased persons is not considered human subject research, and, consequently, consent is not required (US Department of Health and Human Services, 2003b). Therefore, legally, research is permissible on tissue samples from deceased individuals. In Canada, with the exception of Quebec (Sallée and Knoppers, 2006) in the case of deceased donors, “free and informed consent shall be expressed in a prior directive or through the exercise of free and informed consent by an authorized third party” (10.1(c)). Germany imposes identical conditions on collection and subsequent use in research whether the individual is alive or deceased. Next of kin can provide consent, as long as this is not inconsistent with the deceased’s wishes (express or presumed), and there are broad provisions allowing for waiver (Nationaler Ethikrat, 2004).
Yet, all of these countries allow for waiver of consent when samples are anonymized. Since no further downloading of data from medical records is possible after the death of the sample source, it is, in fact, permissible to perform research on anonymized samples, as long as this is not inconsistent with the wishes of the deceased.
Bio-banking is not of trivial scope or significance. As of 1998 in the USA alone, over 282 million archived and identifiable pathological specimens from more than 176 million individuals were being stored. At least 20 million are added each year (US National Bioethics Advisory Committee, 1999).
Yet, the scientific community identifies the limited availability of carefully collected and controlled human tissue samples annotated with essential clinical data as a major obstacle to progress in post-genomics research (US National Institutes of Health, 2006). Thus, whether or not broad consent has been obtained, and whether or not consent can be waived for a given research use, can have a profound impact on whether a particular project is possible (Kaiser, 2006).
Allowing for broad consent options is consistent with empirical evidence of the views of individuals. Most patients do want their tissues to be used for research. Indeed, a recent review (Wendler, 2006) of studies involving in total more than 33 000 persons, specifically examining secondary research uses of leftover samples from medical care, found that the vast majority of individuals (83–99%) were willing to donate their samples for research; those unsure about donation were often concerned about spreading their disease. Further, most people (79–95%) were willing to provide a sample for research in general (i.e., broad consent for research), as long as ethics committees would determine which research their samples would be used for and would ensure that those uses pose no more than minimal risk to the sample donor (Wendler, 2006).
Before contributing clinical samples for research or starting a new collection, a policy for future uses should be set. The policy should outline the consent process and the necessary information to be given to patients. One fact that must be emphasized in a clinical setting is that patients’ decisions regarding future research uses of their samples will in no way affect their care (US National Institutes of Health, 2006). As much as possible, future uses of biological samples should be anticipated and consent obtained. Not mentioning this possibility at the time of collection during care constitutes a lack of transparency (Cambon-Thomsen, 2004). According to “tick in the box” consent forms, allowing individuals to specify which uses of their biological material and associated data are permitted or excluded as well as the degree of subsequent decision-making authority they want to retain “strikes a reasonable balance that is supportive of individual autonomy and of genetic research” (Canadian Biotechnology Advisory Committee, 2004). This approach may not be practical however. Moreover, it would not be feasible for patients wishing to contribute their tissues or medical data to longitudinal studies in population genetics as the same data must be collected from a large number of individuals over time.
The geneticist should contact the primary physician to ascertain the conditions under which the samples were obtained. Did the general consent signed upon admission or a more specific one prior to surgery or biopsy describe the policy of further uses of removed tissue? If so, was it for quality assurance programs or specifically for research? If the latter, was it limited in any way? The absence of a specific consent for research may be covered by a notification of the research policy in the general consent form signed at admission. In any event, the researcher will have to obtain the approval of a research ethics committee for the protocol. The committee may require a more specific consent from those patients that are still living. Tort laws vary on the access requirements for tissues of deceased persons. The same holds for access to the medical records, if needed for such research. The fact that the research is in the same field – cancer – may facilitate such research use but genetics raises issues of its own.
A randomized study of genetic variation to determine drug response would involve using only certain samples as representative of the general population. In all likelihood, such a study would not be foreseen in the admission consent. Access to the medical record for data on drugs prescribed would probably form part of the protocol. Therefore, while the study would not use any patient identifiers, the data collection stage would need to retrieve information from the randomly selected charts and then remove all identifiers before analysis. Usually, such studies only publish aggregate data, which serve as a resource for later, more specific disease protocols. Again, a research ethics board and local laws would determine whether a specific consent would be required. Anonymization may, however, obviate any consent and a waiver may be granted or even foreseen by law. The fact that ethnicity and relation drug response are under study, while no doubt helpful to the communities concerned, could also have untoward results such as insurance difficulties or no drugs at all being available to certain subpopulations.