Abstract
Although biobanks are gaining importance as tools in the field of biomedical research, enabling investigators to access large numbers of catalogued samples and/or data, most have not reached their full potential. Numerous obstacles may prohibit the efficient sharing of, and access to their sample and data collections. In order to minimize or overcome these obstacles while meeting ethical criteria, the first step is to identify the challenges to sharing between biobanks and between biobanks and researchers, thus enabling targeted solutions to be implemented. To date, no one has specifically catalogued the full scope of the barriers to sample sharing currently identified in the literature, yet such a list is essential if these matters are to be addressed. We have reviewed the literature on biobanks in order to identify the issues mentioned as barriers to sample sharing with or without data. Our literature search identified 15 barriers, including logistical, ethical, and legal issues. In this article, we provide a description of these barriers, discuss key themes, and conclude that empirical research is required to determine the full extent of the challenges addressed in the literature.
Introduction
B
The general tone of the recent literature is that biobanking for biomedical research is now an established practice, and it should be exploited to its full potential.9–11 In 2008, the European Science Foundation appealed to the European biobank community to combine the wealth of biobank data and materials: “There is an urgent need for the coordination and harmonization of biobanking and biomolecular resource infrastructure”. 10 As a concrete sign of this desire to unite European efforts in biobanking, the European Union reserved 5 million Euros for the Biobanking and Biomolecular Resources Research Infrastructure (BBMRI) project. Harmonization of operating practices, networking, and increased funding are among the methods suggested for advancing the field, and are intended to allow biobanks to build up statistically significant sample collections, and bring them into closer contact with the researchers who use them.9,10,12–15 While these positive steps forward are crucial, it is also necessary to examine current barriers to sample distribution, which must be managed and/or removed before the full research potential of biobanks can be achieved. One of the primary goals of human biobanks is to make resources available to the researchers who need them. Barriers to the accessing or sharing of resources can therefore reduce the usefulness of biobanks, at the regional, national, and international levels. To date, no one has specifically catalogued the full scope of currently-identified barriers to sample sharing, yet such a list is essential if these matters are to be addressed.
We define a barrier to sharing as any factor that constitutes an obstacle in the sharing process, be it logistical, ethical, or theoretical. The term “barrier” is not used here in an absolute pejorative sense, or to suggest that every obstacle to sharing is completely undesirable; we recognize that some processes that can limit sample and data sharing, such as ethical review, are necessary and desirable. We consider sharing as the process of biobanks (or scientists with registries or laboratory collections) supplying samples with or without data to those requesting it. We do not address herein the particular problems of initial sample collection by the biobank, unless these specifically impact the subsequent sharing of samples with others. In order to provide a list of the barriers to sharing samples with or without data currently identified, we conducted a review of the literature. In identifying and discussing these issues, we aim to provide a unique and focused resource which will allow stakeholders to recognize and understand potential barriers to sharing in order to allow for a more concrete approach to devising solutions above and beyond the general calls for harmonization and standardization.
Methods
We conducted a literature search concerning the obstacles to sharing samples with or without data in biobanking. We searched three databases: PubMed, Web of Knowledge, and JStor. Key terms used in pairs and sequentially included: (biobank OR biobanks OR biorepository OR data bank) AND (sharing OR barriers OR challenges OR obstacles OR problems OR nomenclature OR terminology OR practical OR ethical OR consent OR governance OR legal OR data collection OR territoriality OR justice OR fair OR fairness). The articles retrieved included all those published up to June 26, 2012 (inclusive). The term biobank is relatively recent, appearing in PubMed for the first time in 1996, and used more frequently from 2000 onwards; 16 therefore, no lower date limit was set on the search. After eliminating all double entries, we read and evaluated the abstracts (or introductions for articles without abstracts), of all remaining articles to assess their relevance to our study; articles were included if they explicitly addressed barriers to sample sharing with or without data, in biobanking. Publications addressing only the sharing of data were excluded. We then used a “snowball” approach to obtain additional relevant articles from the reference lists of the initially selected articles. The full text of articles was then scrutinized to identify the obstacles to sample sharing with or without data, in biobanking. Since the goal of this article is to identify in as much detail as possible the barriers to sample sharing as discussed in the academic literature, we included only those issues which were addressed in the context of stopping, deterring, complicating, or hindering sharing. Themes or areas that were identified in general as benefiting from improvement in biobanking, of which the literature has many examples, were not interpreted as being barriers to sample sharing. For example, a call for increasing overall funding for biobanking activities in general would not have been identified as a barrier to sharing per se. Our aim in using this more restrictive interpretation was to avoid the reification of barriers by equating a call for optimization with the existence of an obstacle.
Any issue that was cited specifically in the context of hindering sharing was selected, and the full sentence (or sentences) placed in a table. The procedure was performed by two of the authors (FC and HCH) independently, and the results compared. Any discrepancy was discussed until both authors agreed on the full list of barriers identified. Once agreement was reached, the barriers were further defined and organized into groups based on commonalities. The final category labels given to each group of barriers were agreed upon by all authors. While these categories are an attempt to understand from where some of the barriers arise, we emphasize that they are indeed non-exclusive and do overlap with each other, such that some barriers have some characteristics from all categories.
Results
Twenty-seven articles published between 2003 and June 2012 authored by 23 distinct first authors were retrieved using the search strategy described above (Table 1). The nature of the articles ranges from empirical studies, to reviews, to commentaries and discussions, with the latter two types the most numerous.
SOPs, standard operating procedures.
Fifteen barriers to sample sharing with or without data were found through this search (Table 2). The identified barriers were organized into three macro-categories: Internal Issues, External Issues, and Ethical Issues. These macro-categories are designed to focus the review and group similar topics. They are not mutually exclusive or exhaustive, but are meant as a way to organize the barriers and reflect the way in which we have chosen to approach these issues.
The number of times the barrier was mentioned explicitly as a barrier to sample sharing with or without data in the articles analyzed. We recognize that some of these barriers have been mentioned in articles as a general problem in biobank operations, but these were not included.
SOPs, standard operating procedures.
Internal issues
Internal issues are those over which the biobank exerts most control, either because it established the systems (which cause problems), and/or because it has the wherewithal to change them. They are issues that arise from the way the bank runs, and crucially, over which professionals involved with that bank exert influence.
Standard operating procedures
As defined by Asslaber and Zatloukal, standard operating procedures (SOPs) “should define the whole process of sample acquisition, sample processing, and preservation, as well as storage and retrieval”. 12 Sample acquisition is a part of SOPs. However, since this review is focused on the sharing of samples with or without data, we chose to not focus on barriers to initial acquisition of samples unless they were mentioned specifically with respect to sharing. See “Sample Availability” for more information.
Given the large range of activities included in this category, it is not surprising that barriers in SOPs are mentioned frequently in the literature, with only legal issues being more frequently cited. The differences in procedures between biobanks are described as a challenge, and in some cases, the biobank's current practices may be suboptimal, and therefore a barrier, regardless of the differences with other banks.
Many of the problems with respect to SOPs are related to the quality of the stored samples and this is identified by Myles et al. 17 as the primary challenge in biospecimen exchange. Biobanks may be unable to assure the quality of their samples,17,18 or correctly link the patient's treatment history to a particular sample, or there may be a lack of clinical data. 17 The variety of methods by which biobanks store and process samples was mentioned as a barrier, as preparation methods in different laboratories and countries may not be uniform, leading to lack of compatability.13,19–21 Samples are thus not exploited to their fullest potential, as researchers are unlikely to continue to request samples that are not standardized. 11 Along the same line, and specifically with respect to international interoperability, Kiehntopf and Krawczak 20 address in more detail the following aspects: pre-analytical sample-handling, the standard pre-analytical code, sample quality management, technological issues (storage, retrieval, sample identification, and automated technical processes), sample-related IT, and data-management and shipping. 20 From an empirical study of European biobanks, Zika et al. also identified the complications related to shipping with dry ice as a barrier. 21 Budimir et al. sum up the issue by stating that biobanks with “different designs and different settings” can hinder sharing, and that international harmonization is therefore needed. 22
Sample availability
Sample availability, meaning the existence of a useable number of samples in a bank, is mentioned on three occasions as a hindrance to sharing. It is addressed here because it is mentioned in the specific context of sharing, despite the contradictory issue that if samples are unavailable in the first place, sharing is a moot point. Sample availability is brought up as a barrier to biobanks that are engaged in sharing, but could do more if more samples were obtained. The lack of “local” biobanks that could facilitate sample collection by their ease of accessibility to both researchers and donors is mentioned, in an interesting contrast to the general recommendations for more centralized banking. 23 An example at the micro-level is the sharing of prostate cancer samples, which is limited by the fact that such samples are in high demand, yet little tissue is, in fact, collected. 24 The general scarcity of certain kinds of samples is also mentioned by Myles et al. 17
Awareness
Related to the topic of availability, but not strictly the same notion, Zika et al. mention the lack of knowledge or awareness of existing sample collections by researchers seeking materials as a barrier to sharing. 21
Networks
The lack of efficient networks of biobanks is mentioned four times as a barrier. The need to elaborate and decide on the details of biobank management, ownership, and confidentiality, issues which are sometimes problems for sample sharing, limits the effectiveness of creating and or sustaining collaborative networks.20,21 Asslaber and Zatloukal emphasize the fragmented nature of biobank networks, especially in Europe, as a key factor preventing them from pooling their samples, and hence prohibiting researchers from accessing large sample and data sets. 12 The lack of functioning biobank networks is an obstacle to increasing the number of “well organized and accessible” collections. 11 Yuille et al. provide the example of the BBMRI, a network with a “distributed hub structure” whose aim is to “enable access by researchers to different sample types (with associated data) collected under different study designs,” as a way to create functioning biobank networks. 11
Fees
The financing of infrastructure and personnel is a big challenge to the existence of biobanks.11,25 However, the specific issue of cost or fees with respect to sharing samples was not mentioned often. The fees charged by biobanks (cost recovery rather than profit-generation) are mentioned as an obstacle on two occasions. Somewhat surprisingly, not mentioned are users being excluded due to prohibitively high costs, nor for-profit biobanks. Rather, the issue appears to be confusion about costs. Biobank fee structures, which are elaborated internally, differ between biobanks and are said to be confusing. 17 At a more basic level, however, the very practice of assigning a financial value to a sample is problematic. To do so responsibly, the level-of-effort, the potential usefulness of the sample, and the finances of those who might wish to access it, must all be considered. However, to date, concrete guidelines to help calculate how these factors can and should be quantified do not exist. 26
Governance
Biobank governance, widely discussed in the biobanking literature, is mentioned four times specifically in the context of problems in sample sharing. Governance in biobanking refers to the systems, procedures, and documents that regulate the banks' activities, and to the people who oversee this sphere; without innovation in this realm, the flow of research materials will be limited. 27 In certain institutions, biobanks sometimes place restrictions on access to their data in order to protect their mandate, 12 an issue which can also be considered in the context of territoriality. At the other end of the spectrum, attempts by those who govern biobanks to broaden their sharing policies can be met with resistance from researchers, who are either confused by the changes or seek to limit access to their samples for the territorial reasons identified below. 28 While we have included governance in our list of internal issues, certain aspects of the problem, such as legislation and guidelines, may be imposed from outside the bank itself. Governance may also be considered an external issue and/or an ethical matter, and as such, may share some characteristics with those issues mentioned below. With respect to guidelines, Elger and Caplan summarize the problem regarding international collaboration by saying that many national and biobank-specific “guidelines contain clearly divergent recommendations in important areas, which interfere with international collaboration. Not only do different systems exist for the collection of data and the processing of samples, but also the guidelines reflect fundamentally different ethical frameworks.” 16
External Issues
The constraints imposed on biobanks from external sources constitute some important barriers to sample and data sharing, and in contrast to internal issues, are less controlled by the biobank operators. Typically, the obstacles take the form of requirements imposed by non-biobank organizations that are designed to improve some aspect of biobanking, but which also have the unintended secondary effect of causing difficulties in sharing.
Commercialization
The potential to commercialize biobank samples represents a further barrier to sharing. While Hewitt et al. 18 merely mention that this is so, suggesting that the broader issue is the effect of commercialization on public trust, elsewhere it is noted that the policies which influence sharing are dominated by the market interests of the biotechnology industry. 29 In other words, biobanks may be prevented from making their samples available, at least to certain groups, or researchers may likewise be prohibited from storing their samples in banks with an open sharing policy, due to the conditions of funding received from private biotechnology companies.
Legal
Legal issues were the most frequently listed barriers to sharing from the pool of articles reviewed. For the purposes of this review, legal issues refer to national or international laws, and exclude nonlegally binding agreements and guidelines created between biobanks and other parties. Agreements and guidelines are not mentioned in the literature as problematic because of their legal status, and are therefore included in the internal issues section in subcategories such as governance. Despite the frequency with which legal barriers are mentioned, there is little variation in the main problems. Broadly, the divergence of regulations on the uses, storage, transfer. and nature of tissues and data is repeatedly mentioned as an obstacle to international collaboration.11,12,21,30–33 Laws and guidelines contain “clearly divergent” instructions, which can be either prohibitive or confusing. 20 Furthermore, tissue export or import is limited or banned by certain countries. 34 Even within the European Union, a disruptive amount of variance exists.11,32 Based on their survey of 126 biobanks, Zika et al. state that most of the problems in sample sharing were related to legislative barriers. 21 The bureaucratic effort and biobank-specific organization required to understand and comply with these regulations is a secondary barrier. 27
Intellectual property and patents
While intellectual property (IP) rights and patenting are also primarily legal issues, they merit separate attention because the barrier they pose is described distinctly from broader legal questions. IP and patents are mentioned six times as a challenge in sharing in biobanking. When a patent or claim of intellectual property on some work derived from banked samples is anticipated, access to those samples may be restricted, at least for a set time period. 35 This fact may further erode public trust in biobanking, as the public are not keen to see beneficial research restricted for ostensibly financial reasons. 30 Such restrictions, when they exist, may prohibit the optimal use of the samples by eliminating other research groups.17,18 However, as Cambon-Thomsen et al. point out, confusion about the very nature of IP and patents, even when no restrictions are in place, may be a hindrance to some researchers. 28
Nomenclature
Nomenclature concerns the terms used in the medical or scientific field, and not just those terms that are specific to biobanking. The profusion of terms relating to sample and data types, research methods, and databases is, in part, a natural result of evolving language habits over the years. However, this is of course highlighted when samples taken 10 years ago are compared with those obtained more recently. The problem is exacerbated on the international level, where even if all collaborations take place in English, the various translations and culturally different uses of words are still not likely to make for easy collaboration. Nomenclature may, therefore, be an important barrier to international sharing. 33 Pearson 13 also points out that medical histories which accompany samples are useless if those from different institutions use different words.
Publication credit
Publication credit and proper recognition of the time and effort devoted to creating a useful tool (like bioresources) can be a difficult issue in all spheres of academic research, not just biobanking. Shickle et al. point out that access to samples might be conditional on publication credit, and not simply “altruistically” available. 36 This is distinct from researchers not wanting to make samples available at all (addressed below, as territoriality); in this case, the obstacle is meeting conditions, rather than sheer unwillingness to share. 36 According to Cambon-Thomsen, 37 biobanks that are not properly credited or recognized for their work in establishing and maintaining their samples will not receive the professional recognition that they need to keep attracting researchers, and will risk becoming storage warehouses with no “clients,” or potentially losing funding, and their samples will not be shared with a wider pool of researchers.
Ethical issues
Ethical issues appear as a distinct category, although in some cases, they may be both internal (i.e., under the control of the biobank) and external, as they are to some extent imposed by sociocultural context and/or national standards. However, in all cases, they are examples of decisions which require some degree of moral reasoning to resolve and we therefore judge that they merit separate attention here.
Consent
The question of informed consent for biobank research is ubiquitously mentioned in the literature.16,38–41 This is not surprising considering the huge role the topic plays in the ethical sphere of biomedical sciences. Indeed, not obtaining the proper consent can prevent samples and associated information from ever being used at all in biobank research. 42 However, the explicit mention of informed consent as a hurdle to sample sharing is relatively rare in the articles included in our analysis. Consent, when mentioned strictly as an obstacle to sharing, appears in different contexts. Consent forms differ between biobanks, or between the institutes that initially collect the samples and then send them to biobanks.13,20,39 Therefore researchers may be unable to use the samples or data for certain projects. This is linked to the difficulty of obtaining consent for prospective research, as it is difficult to inform donors of as-yet unplanned projects. 39 It is stated that the current formulation and interpretation of most biosample consent forms hinders data sharing, as they were not conceived with transnational projects in mind. 43 Kiehntopf and Krawczack 20 specify that, with respect to the interoperability of consent to use samples and data, the prime issues of interest are the right to transfer to third parties and the scope of research allowed.
Territoriality
“Territoriality” describes the phenomenon of unwillingness to share data or samples; this can also be described as wanting to keep samples and data exclusively for one's own research. In the literature, territoriality is generally attributed to individual researchers, rather than to biobank managers or operators. The axiom “publish or perish” is mentioned in light of the fact that researchers may well be keen to foster an environment of trust and altruism in their field, but are equally aware that the samples, especially rare ones, can be an entry into prestigious research teams. 36 More broadly, “elitism and competition” 17 among researchers is a threat to sharing, and access to data is therefore “vulnerable to researchers' incentives.” 29
Prioritization
Prioritization refers to the weight the biobank assigns to competing research projects (i.e., those which seek to use the same samples simultaneously). It is mentioned only once in the literature reviewed here. Fortin et al. 26 indicate that banks which seek to prioritize the work of local researchers, even with the laudable goals of supporting their community and keeping donors close to the work, will be hindering others, at the national and international level, from accessing their samples. 26 This is not a barrier to all forms of sharing, as the samples are still in use, but rather an obstacle to wider sharing.
Safe transfer/confidentiality
The ability to transfer useful data or information relating to samples while maintaining the confidentiality of the donors' information is difficult particularly in international collaboration. 33 Although data are usually transferred with the samples, the reverse is not always true, and data-sharing has become a topic in and of itself. In this respect, data-sharing may include the sharing of both clinical information, as well as results of analysis using samples and clinical data. Since this review is focused on the sharing of samples with or without data, we have not included articles dealing only with data sharing.
Discussion
Based on the articles included in this review, 15 obstacles to sharing samples with or without data in biobanking were identified. This is the first review that specifically identifies and discusses these obstacles to sharing. Our findings show that a broad variety of barriers have been discussed. The majority of issues receive one or two mentions; however, a small number are cited by numerous authors. Two identified barriers (legal issues and SOPs) are mentioned frequently, and IP and patents also emerged often.
Although the attention which the barriers receive appears to vary significantly, the conclusions to be drawn from this require caution. It cannot be presumed that the frequency directly corresponds to the actual severity or extent of these barriers in practice. Certain frequently-cited factors may seem overemphasized or relatively unproblematic to some readers and stakeholders. For instance, certain banks may experience no problems at all with storing or shipping samples, while others will feel that a shortage of samples is unlikely to hinder sharing. 9 While we do not suggest that any of the issues mentioned in the literature were invented or were purposefully ignored by the authors, it is important to note the following: i) the large majority of articles retrieved through our search do not aim to address barriers to sharing per se. Many articles simply mention barriers in the introduction and/or discussion and hence do not elaborate on these issues; ii) some aspects may have been repeated due to their perceived importance, whether or not this reflects the current state of affairs in biobanking; iii) only about a third of the articles under study actually addressed empirical data on barriers to sharing. In the absence of such data, it is impossible to assess how accurately the issues in the literature reflect current experiences. Our findings are a valuable initial approach, but require comparison with further empirical research.
However, our results do provide indications of what the authors feel are important hurdles to sharing. Therefore, we suggest that legal issues and SOPs, and to a lesser extent IP and patent considerations, are particularly problematic issues facing the biobanking world. An interesting aspect of this result is that each of these issues affects biosample sharing in different ways. National laws are an obstacle to sharing because they conflict with one another, but the effect of this is magnified because the laws, as they stand, are non-negotiable, and the very networking solutions designed to overcome this obstacle can be hampered by it. 12 Goebel et al. 44 point out that biobanks can circumvent some gaps in the transnational legal setting by drafting their own contracts to regulate cross-border sharing. However, as some authors emphasize, uniform regulation must be developed, something biobank stakeholders alone cannot achieve.31,45
SOPs, on the other hand, can, to some degree, be influenced from within the biobanking sphere. Our findings suggest that the problem in this case is not only one of divergence, but also one of quality. As Shaw and Patterson note, the rise in translational and genomic research has meant a greatly increased demand for samples that are not only well preserved, but are accompanied by comprehensive clinical data. 46 This places an extra burden on surgeons and pathologists who are often not the end-users of those samples, or even directly involved with the biobank itself. 46 Numerous suggestions for enhancing the sample extraction and preservation process already exist;47,48 a number of the articles that identified this barrier did so in the context of discussion on biobank networks.11,13 It may be the case that barriers which are in the process of being overcome are more frequently cited, by way of “introduction” to the proposed solutions.
Intellectual property rights and patenting, while cited fairly often as barriers, differ from general legal issues and SOPs in that these are, to a degree, self-imposed desirable limitations to widespread sharing. In this sense, they may be considered less as obstacles than as tools to support innovation, with some restriction on sharing as a secondary side-effect. While in some cases, they curb sample sharing, the benefits accruing to researchers and the public may offset this.35,49
In contrast to those barriers that are frequently mentioned, it is interesting to note that ethical issues receive relatively little attention as barriers to sharing; even the issue of informed consent is mentioned infrequently. This is in marked contrast to the plethora of articles on the ethical aspects of biobanking, particularly the ongoing debates on the optimal measures for obtaining consent.50–52 This may mean that ethical issues are fairly unproblematic in terms of sample sharing, but have a greater impact on sample acquisition (or that the focus has been on the latter); and/or that as with IP and patents, ethical issues are a desirable measure to limit the risks (to participant privacy and autonomy) associated with widespread sharing. While both of these explanations may be applicable, we feel that the necessity of some restriction on sharing is well recognized, particularly in the case of informed consent. Consideration regarding the justifiability of the barriers discussed in this review, and the extent to which they should be “kept” in place, will also be an important step in optimizing ethical and responsible sample sharing.
Limitations
This review of the literature has certain limitations. First, it must be emphasized that our decision to exclude papers that imply barriers may mean that certain issues were not addressed. However, as previously mentioned, our goal was to avoid accidentally including issues which were not in fact intended to be seen as barriers, either through our own perception bias or the ambiguity of the writing. Second, it must be noted that some of the studies referred to here are out of date by as much as 8 or 9 years.13–33 Hence, while we do not feel we have included any completely obsolete issues, some problems may be less pressing in light of recent developments in networking or sample processing. Third, the articles included in our search range from surveys to reviews to commentaries. Therefore, as previously noted, some problems are mentioned in a more hypothetical way than others. While it might be useful to analyze the rates at which problems are concretely mentioned as occurring versus merely being suspected of occurring, the lack of empirical studies and relatively small literature pool on this topic means that such an exercise would be unlikely to produce valid results. Finally, our inclusion criteria and our desire to use explicit statements mean that articles that did not mention sharing in the abstract or introduction were excluded and thus we may have missed articles that discuss this subject later on in other sections. Furthermore, our decision to exclude articles that focus solely on data sharing means that we have not included such important papers as that by Kaye et al. 53 or Knoppers et al. 54 which provide important insight into aspects of data sharing. However, since data sharing, especially within the realm of genomics, has created an entire field of study of its own, we feel it would be best to address this issue separately.
Conclusion
Biobanks are gaining importance as tools for researchers and clinicians. Regardless of their size and scope, they exist to enable some extent of access to biological samples and/or data. Facilitating this access is an essential step to maximizing the research potential of biobanks; to date, no single article has sought to codify the numerous obstacles to sample sharing mentioned in the literature. As a first step towards addressing the barriers to sample sharing with or without data, a list of some of the challenges biobanks and their stakeholders face has been compiled. What is now required is empirical research to determine whether biobank operators and stakeholders are in fact experiencing the issues identified, and whether other obstacles exist that have not been mentioned in the literature. Biobanks have not yet fulfilled their potential, and overcoming obstacles to their efficient operation is essential.55,56 The consolidation of this list of issues is the necessary first step to studying the existing problems, and ultimately finding solutions.
Footnotes
Disclosure Statement
The authors declare no conflicting financial interests.
HCH is funded by the European Commission FP7 Marie Curie initiative. FC is supported by funds from the Käthe-Zingg-Schwichtenberg Fund (Swiss Academy of Medical Sciences) and a research assistant post at the University of Basel. BE is funded by the University of Basel.