Whole genome sequencing as a diagnostic tool: Participant-centered consent

+Model

ARTICLE IN PRESS

JEMEP-145; No. of Pages 9

Ethics, Medicine and Public Health (2016) xxx, xxx—xxx

Available online at

ScienceDirect www.sciencedirect.com

DOSSIER ‘‘ETHICS, MEDICINE AND GENETICS’’ /Practices and concepts

Whole genome sequencing as a diagnostic tool: Participant-centered consent Séquenc¸age du génome entier comme outil de diagnostic : consentement centré sur le participant N. Nijsingh (Independent scholar) Utrecht University, Kerkweg 15, 3513 VA Utrecht, Netherlands Received 3 April 2016; accepted 27 June 2016

KEYWORDS Whole genome sequencing; Informed consent; Patient-doctor relationship; Recontacting; Screening; Blurring boundaries; Research ethics

MOTS CLÉS Séquenc ¸age du génome entier ;

Summary New developments in genetics and genomics offer a constant challenge to ethical reflection. One such development is the application of whole genome sequencing (WGS) as a diagnostic tool. The use of WGS in clinical practice carries with it the promise of personalized medicine, which will likely enable more efficient care and tailored prevention. However, the complexity and sheer amount of information involved may have significant implications for the patient-doctor relationship. The roles and expectations within this relationship are affected by both the institutional dimension of the logistics of handling genomic data and the changing possibilities of returning unsolicited or new information. As the relationship between doctor and patient changes under the influence of genomic data, the question emerges how the patient’s rights may be safeguarded when the role and responsibility of the clinician shift. I argue that this question necessitates a rethinking of the authority of the patient in handling the flow of information in the clinical context. Specifically, I argue that the authority of the patient in personalized medicine should be shaped from a participant-centered perspective. © 2016 Elsevier Masson SAS. All rights reserved.

Résumé Les nouveaux développements en génétique et en génomique offrent un défi constant à la réflexion éthique. Une de ces nouveautés est l’application de séquenc ¸age du génome entier (Whole Genome Sequencing [WGS]) comme outil de diagnostic. La possibilité de

E-mail address: [email protected] http://dx.doi.org/10.1016/j.jemep.2016.06.001 2352-5525/© 2016 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

2

N. Nijsingh cartographier et d’analyser l’ensemble du génome d’un être humain est de plus en plus abordable pour diagnostiquer divers troubles. L’utilisation de WGS dans la pratique clinique porte donc avec elle la promesse d’une médecine personnalisée, qui pourra permettre des soins plus efficaces et une prévention plus adaptée. Je soutiendrai dans cet article que cette évolution nécessite de repenser les droits et les attentes de la relation médecin-patient. La complexité et la quantité d’informations en cause peuvent avoir des répercussions importantes sur la relation patient-médecin. Les rôles et les attentes au sein de cette relation sont affectés à la fois par la dimension institutionnelle de la logistique de traitement des données génomiques et l’évolution des possibilités de retour des informations non sollicitées ou nouvelles. Comme la relation entre médecin et patient change sous l’influence des données génomiques, la question se pose de savoir comment les droits du patient peuvent être sauvegardés dans ce changement du rôle et de la responsabilité du clinicien. Je soutiens que cette question nécessite une remise en question de l’autorité du patient dans le traitement de la circulation de l’information dans le contexte clinique. Plus précisément, je soutiens que l’autorité du patient en médecine personnalisée doit être formée à partir d’un point de vue centré sur le participant. Bien que la possibilité d’une « médecine personnalisée », qui promet des soins plus efficaces et une prévention sur mesure, a été qualifiée de révolutionnaire et a rec ¸u une grande attention des auteurs de diverses origines, les conséquences de cette révolution pour la relation médecin-patient ont surtout été abordées très timidement. Ce document vise à contribuer à la littérature existant sur le sujet en discutant le cadre de cette évolution de la relation et le rôle de l’autorité du patient en son sein. À cette fin, je vais d’abord décrire le WGS comme un outil de diagnostic et son rôle dans l’idée de la médecine personnalisée. Je suggère que les défis du WGS portent sur les rôles et les attentes de la relation médecin-patient. Puis je montre comment ces rôles et ces attentes peuvent être affectés en raison d’un flou entre les lignes du dépistage et du diagnostic ainsi qu’entre la recherche et la pratique clinique. En outre, je décris la possibilité qu’ont les patients de recontacter leur clinicien pour reconsidérer la situation en fonction des données génétiques. Puis j’examine comment ces changements dans la relation entre le patient et le médecin affectent l’autorité de prise de décision du patient. Je conclus que l’émergence de WGS appelle à une réévaluation de l’autorité du patient. Je donne quelques indications sur la fac ¸on dont cette réévaluation devrait être constituée, et termine en posant un certain nombre de questions à l’égard de ce thème. © 2016 Elsevier Masson SAS. Tous droits r´ eserv´ es.

Introduction The possibility to map and analyze the entire genome of an individual human being is becoming more and more affordable for the purpose of diagnosis of various disorders. I will argue in this paper that this development necessitates a rethinking of the rights and expectations in the patient-doctor relationship.1 Although the possibility of ‘‘personalized medicine’’, which promises more efficient care and tailored prevention, has been called revolutionary and has received ample attention of authors from various backgrounds [1—3], the consequences of this revolution for the patient-doctor relationship have mostly been addressed only in passing or else very tentatively [4,5]. This paper aims to add to the existing body of literature on the topic by discussing the framework for this changing relationship and the role of patient authority in it. To this purpose, I will first describe whole genome sequencing (WGS) as a diagnostic tool and its role in the idea of personalized medicine. WGS brings both great promise for society as 1 As this paper focuses on the clinical context of genome sequencing, the health care professional that a patient will in most cases relate to will be a geneticist. However, part of my argument is that numerous professionals will be involved in personalized medicine, which is why I opt for the generic term ‘‘doctor’’ or ‘‘physician’’ to indicate this role. I will use these terms loosely and interchangeably.

a whole, as well as substantial challenges. I suggest that these challenges relate to the roles and expectations in the patient-doctor relationship. I then proceed to show how these roles and expectations may be affected due to a blurring between the lines of screening and diagnosis and as well as between research and clinical practice. Also, I describe how the possibility of recontacting patients is of consequence for the clinician’s exercise of the duty of care. Then, I examine how these shifts in the relationship between patient and doctor affect the decision-making authority of the patient. I conclude that the emergence of WGS calls for a re-evaluation of patient authority. I give some pointers to how this re-evaluation should be constituted, and end by posing a number of questions with regard to this theme.

WGS as a diagnostic tool Whole genome sequencing has its roots in the Human Genome Project, which commenced in 1988 [6]. The purpose of the Human Genome Project was to gather more information on the human genome, with the eventual aim of mapping the entire genome, which consists of approximately 3.3 billion base pairs. This goal was achieved in 2003 [7]. Since that moment, developments have progressed

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

Whole genome sequencing as a diagnostic tool: Participant-centered consent in rapid succession, leading to a fall in sequencing costs, and an increase in sequencing speed [8]. In particular, the decreasing cost of WGS is likely to provide more opportunities for gathering diagnostic data. Therefore, even though there is still much uncertainty, it is expected that WGS will become increasingly relevant as a diagnostic tool in clinical practice. The mapping of the genome can help to provide information on the inter-relatedness of phenotype and genotype and contribute substantially to our understanding of diseases both mild and serious, rare and common [9]. The anticipation of these developments has given rise to the idea of ‘‘personalized medicine’’: an approach to health care where, based on a genetic profile, the needs of the specific individual are taken as the point of departure, thus allowing for safer and more effective treatment as well as tailored prevention.2 Personalized medicine provides promising perspectives for prevention, diagnosis and treatment of a potentially large number of diseases and has even been heralded as a revolution in modern medicine, an entirely new approach, which affects the structures and logistics of health care itself. As knowledge on the relation between phenotype and genotype expands, it is expected to become possible to prescribe drugs that are particularly effective or have fewer side effects for that individual, to take preventative action on the basis of risk analysis and to treat root causes of diseases for which there currently is only symptomatic treatment [10]. These developments would imply a significant departure from the trial-and-error paradigm of current medical practice [2,11]. To achieve the ambitions inherent to the idea of personalized medicine, it is necessary to adapt the medical infrastructure. Hamburg and Collins liken this to the infrastructure created by the US federal highway system: When the federal government created the national highway system, it did not tell people where to drive — it built the roads and set the standards for safety. Those investments supported a revolution in transportation, commerce, and personal mobility. We are now building a national highway system for personalized medicine, with substantial investments in infrastructure and standards. We look forward to doctors’ and patients’ navigating these roads to better outcomes and better health [12]. Although this is an exciting prospect, it urgently raises the question how we should envision this new infrastructure and how the medical-ethical frameworks that have crystallized over the years will hold up in this environment. Admittedly, much of the predictions are uncertain and the availability of more information may not necessarily lead to clinically actionable understanding. The relation between genotype and phenotype is notoriously complex, and putting new

2 Of course, medicine has always been aimed at persons — who else should one treat? Therefore, the term ‘‘personalized’’ is a bit of a misnomer. It seems that ‘‘customized’’ would be more appropriate to describe these developments. Nonetheless, since in this paper, I want to relate to the present debate, I will follow the commonly used term ‘‘ersonalized medicine’’.

3

insights to use will thus not be an easy task. Furthermore, it is difficult to fully gauge how these developments will influence clinical practice. Still, it seems probable that as knowledge of the genome expands, more and more clinically relevant facts will be available and their significance better understood.3 This means that regardless of how exactly these developments will unfold, it is likely that knowledge of the genome will play an increasingly large role in clinical practice. The increasing importance of knowledge of the genome has been extensively commented upon [13,14]. However, the implications of personalized medicine for the relationship between doctor and patient have largely been neglected.4 The normative framework that has emerged over time with respect to the patient-doctor relationship is centered around the observation that this relationship is typically asymmetrical because the doctor tends to be in a position of power, whereas the patient is vulnerable [17]. Thus, the doctor has specific duties of care, confidentiality and information, whereas the patient has the authority to give or withhold consent.5 This framework has been dominant throughout medical ethics.6 As I will demonstrate in the next section, WGS challenges this framework by changing the streams of information: large amounts of information become available, which can be stored and accessed for various purposes. I will argue that, as more and more information becomes available and more and more people get involved, the framework in which the respective roles presuppose a more or less personal relation between patient and doctor no longer captures all the relevant normative aspects. This raises a number of questions: who controls this information and how can the patient exercise authority over what happens to it? How do the changing roles and expectations affect the normative framework that has become prevalent in modern medical ethics? How can the rights of patients be protected in personalized medicine? In order to address these questions, I will focus on how WGS and the emergence of personalized medicine may change the relationship between physician and patient.

3 However, since medical research tends to focus first on cases where there is a positive correlation between a particular genotype and a specific illness, the correlation between phenotype and genotype may weaken as more unclear or healthy cases are added to the data set. Thus, the variants of uncertain significance may actually increase as more data becomes available. I owe this point to Niklas Juth. 4 For example, Cambon-Thomsen et al. ask how biobanks may influence the relation between the clinician and the patient, but do not provide an answer to that question [15]. See also [16, p. 592]. 5 This is not to deny that the patient has duties too. However, the framework is aimed at the protection of the patient, who is in a particularly vulnerable position [18, p. 273]. 6 See Beauchamp and Childress [19]. Contrary to Beauchamp and Childress, I will mostly speak of ‘‘authority’’, rather than ‘‘autonomy’’ when discussing the moral underpinnings of informed consent. I do not object to phrasing the issue in terms of autonomy, but I think that the notion of authorization more precisely captures what is at stake here. For the present point, the difference is not really of importance, but see [20] for more discussion on this topic.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

4

N. Nijsingh

Shifting normative frameworks In this section, I discuss the different ways in which WGS may change the normative framework that structures the relationship between doctor and patient. As I will show, the introduction of WGS in diagnostic practice may blur the lines between diagnosis and screening and between research and clinical practice, as well as the temporal delineation of the treatment relation. WGS has the potential to generate enormous amounts of more or less relevant health related information. Note however that WGS does not directly provide information on the health status of the subject. WGS is merely the mapping of an individual’s genome. Whether and to what extent WGS reveals meaningful data depends on the way the raw data are analysed. The analysis of the data can be of the whole genome (whole genome analysis [WGA]), the whole exome (the coding part of the genome) or instead a targeted analysis of certain specified areas in the genome or exome, for example by focusing on de novo mutations in a person’s genome.7 An advantage of a targeted approach is thought to be that it is less likely to produce as a ‘‘byproduct’’ results which are unrelated to the initial clinical question. Even a targeted approach, however, may involve thousands of known places in the genome and thus produce much information that is not directly connected to the initial reason for testing. Unintended and implicit information as a by-product of diagnostic tests is by no means new to medical science. However, the scale and diversity of the information that may surface in the implementation of WGS is unprecedented and raises with urgency the question what to do with the data that is made accessible by means of WGS [9]. Whereas other diagnostic tools, such as genome wide array analysis focus on particular known places on the genome, albeit thousands of known places, WGS has the potential for the disclosure of an even vaster amount (and variety) of data. The implication is that incidences of ‘‘unsolicited information’’ are likely to occur regularly when WGS is introduced as a diagnostic tool. Not all information that can be derived from the analysis of a genome is necessarily beneficial to the patient. Some information may be burdensome, whereas other findings may be of little or unknown clinical use. Burdensome information for example may concern untreatable diseases. The message that one may become very ill in the future, but that there is nothing to be done, may do more harm than good in terms of psychological stress, medicalization and social stigmatization. Also, some genetic information may cause financial harms: it is not inconceivable that certain outcomes may, for example, negatively affect one’s chances of insurance or employment. Some people, however, would still prefer knowing over not knowing, even when the clinical benefits are uncertain. Other information may be (too) difficult to interpret or of unclear significance, and therefore of contested worth. A question arises, therefore, about how much and what information that surfaces in the process of analysing the genome sequence should be communicated

7

By comparing the data with her parents’ genome [21].

and under what conditions. Specifically, the issue of what to do with information that ‘‘unintentionally’’ surfaces in the process has been much discussed [22]. The terminology with regard to information ensuing from diagnostic tests, but unrelated to the initial clinical question, has been a source of some discomfort to the authors dealing with the issue. Often these data have been described as ‘‘incidental findings’’ [22] or ‘‘unsought’’ information [23]. However, from the perspective of the laboratory, it is a matter of statistical fact that certain results will occasionally occur. Since we know for certain that WGS has the potential of generating a lot of unrelated information, it seems inappropriate to label this as being an ‘‘incident’’. Similarly, since the purpose of genome analysis is to map and analyze the entire genome, and thus to search for any deviation, the term ‘‘unsought information’’ is confusing. For these reasons, the term ‘‘unsolicited information’’ has been suggested [20,24]. This term avoids some of the awkwardness invoked by the other terms. In focusing on the clinical domain, the term ‘‘unsolicited information’’ enables taking into account the respective roles of the patient and doctor. After all, it is in general the patient who takes the initiative to establish a treatment relationship by consulting a physician.8 The patient means to address a specific problem, in order to be relieved of specific complaints. The doctor then formulates a hypothesis, on the basis of which a test is performed, either by the doctor herself or by third parties, such as a laboratory. Any information that surfaces in the course of this process that is not related to resolving the initial complaints could be considered ‘‘unsolicited’’. This conceptualization enables us to pinpoint with some more precision what is at stake when we consider the issue of WGS as a diagnostic tool: it allows us to take into account the expectations and will of the patient. If information surfaces that is unrelated to these complaints, it affects the respective roles of both doctor and patient. The initial clinical test in effect turns out to be a form of screening [21,26,27]: a means to detect non-symptomatic diseases in subjects who have not taken the initiative with regard to this discovery. In recent years, there has been a shift towards a more proactive stance. Doctors are increasingly seen not only as reactive with regards to a complaint of the patient, but also in their role as ‘‘health promoters’’ [28]. We see here a shift of expectations in the patient-doctor interaction. We expect a doctor to take our interests at heart and warn us when danger is imminent. The normative assumption behind this expectation has been explicated in medical ethics and law in the form of the ‘‘duty of care’’. The doctor-patient relation is widely regarded as a special type of relation, and has even been called ‘‘sacred’’ [16, p. 592]. The special

8 Exceptions are of course children and otherwise incompetent persons. I do not mean to suggest here that these exceptions are in any way marginal or rare; in many cases involving unsolicited information incompetence is a factor. To complicate matters even more, others, such as aunts, uncles or siblings may have interests too, and these interests persist after the patient has ceased to be [25]. The emphasis on informed consent has a tendency to downplay the fact that there are different parties involved who, perhaps, have conflicting interests or preferences.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

Whole genome sequencing as a diagnostic tool: Participant-centered consent nature of this relationship is reflected by the notion of a duty of care. This duty of care is thought to encompass, for example, the doctor informing patients on risks and benefits with regard to his health. The duty of care is thus one of the pillars of medical ethics and crucial to understanding the normative particularities of the patient-doctor relationship [18,29]. The question, however, emerges whether this duty of care also extends to diseases that may not occur for decades or to the carrier status for diseases that are only potentially relevant when one wishes to procreate. Also, findings of uncertain or unclear relevance, for which it is not at all certain that there is something meaningful to communicate, will inevitably appear. It seems that somewhere in the stream of information from laboratory to patient, filters need to be built in: not all findings should be disclosed to the patient, even if that were possible. Note that this situation is not merely that of a clinician withholding information; the filter could (in fact often is) also be applied on the level of the laboratory [4,24]. Given that unsolicited information is bound to occur on a regular basis, this problem calls for some sort of procedure to decide which information is worthwhile. We see here that the duty of care thus takes shape within a broader context than merely within the confines of the treatment room. The same development can be observed when we consider the way in which information is handled over time. Once the information on a person’s genome becomes available, the questions of whether and how it should be stored emerge [21,22]. There may be different purposes for storing information gathered in the process of WGS, which in practice are likely to overlap. When WGS is necessary in the context of clinical care, the data produced may need to be stored because it allows patients to access their own test results. Also, information may be retained for reasons of efficiency: when a new test is called for, or when the patient’s situation changes, it is cheaper to have the data available, instead of performing a new test. Simply discarding the resulting genome sequence after testing seems a waste of resources. Third, from the perspective of research, the data can also be considered as a valuable resource. The promises of personalized medicine rely on the possibility of comparing large amounts of genomic data. Therefore, it has been proposed to use the sequencing data, which has been gathered in the context of diagnosis for research [26,30]. Indeed, the ambitions of personalized medicine can only be achieved by comparing the sequencing data of a large number of subjects and mapping disease progression and different responses to medication and environmental variables to different genomic patterns. Large-scale research projects, such as ‘‘The Cancer Genome Atlas’’ and ‘The Personal Genome Project’’ which have emerged as successors to the Human Genome Project aim to do just that: to gather and analyze large amounts of data in order to determine the relation between genotype and specific phenotypic manifestations [31]. Angrist and Jamal have coined the term ‘‘clinical research hybrid frameworks’’ to capture the way in which research and diagnostic practice may — or rather, according to Angrist and Jamal, should — intertwine [32]. In this framework, large-scale databases (Angrist and Jamal fantasize about ‘‘One database to rule them all’’ [32, p. 3])

5

should be interactively linked to clinical practice, thus both enhancing the as yet insufficient scientific knowledge and benefiting the participants when actionable clinical data emerge, which leads to a ‘‘de-compartmentalising’’ of research and clinical practice. Evidently, such a proposal for the blurring of lines between research and diagnostics has further implications for the patient-doctor relationship [33,34]. As the segregation between clinical practice and research is one of the basic tenets of medical ethics, the blurring of the lines between the two raises a large number of ethical questions.9 However, here I am mainly concerned with the way in which this blurring affects the normative framework that structures the relationship between doctor and patient. I will return to this matter shortly, but first want to point out another way in which this relationship changes due to the application of WGS in clinical practice. The segregation between research and clinical practice is one aspect of the clinical research hybrid framework that merits consideration when we examine the changing relationship between physician and patient. Another complication concerns the way in which subjects may benefit from continued participation in such a framework. As science develops, the analysis of a stored genome sequence may deliver new results as time passes. This raises the question of ‘‘recontacting’’: the re-establishing of contact with patients on the basis of new information derived from a stored genome sequence. This recontacting can be done either on the initiative of the doctor, or that of the patient [37]. A typical scenario involves unsolicited information and the patient being unaware of a purported genetic threat of some sort. Other situations in which the possibility of recontact may arise are when there are improved tests to establish a diagnosis or when new information becomes available that somehow changes a patient’s prognosis or recurrence risk [18]. Various authors question whether and under what conditions it is appropriate to recontact patients whose genome was sequenced for purposes of diagnosis or research [24,29,38]. In some cases, the patient may not want to be informed or may be very difficult to trace down, perhaps due to a change in address. Although the question of what conditions justify or require an initiative to seek contact has been topic of some debate,10 there is another issue worth noting that is more often overlooked. As Ellen Wright Clayton notes [22] in the context of genetics, ‘‘often the work is done by investigators and at institutions with which the person has no prior contact’’. As a consequence, the borders between the responsibilities of the different professionals may also become difficult to draw: apart from the doctor who has face to face contact with the patient, there is a multitude of persons involved in the diagnostic process, each with different informational responsibilities. Whereas it seems false to hold that each of these individuals is in a treatment relationship with the patient, they are each part of an institutional structure on which the

9 Specifically, the ‘‘therapeutic misconception’’, where a subject mistakenly presumes to have a therapeutic benefit from participating in research, has traditionally been a source of worry [35,36]. 10 The issue of recontacting within a research context has been increasingly debated upon in recent years [39,40].

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

6

N. Nijsingh

patient is dependent. It seems not unreasonable to ascribe to this institutional structure a duty of care, similar to that of the physician. This raises the question who has what informational responsibilities; under what conditions should for example a laboratory employee warn a doctor if significant information emerges? To answer such questions, we need to consider the logistic, institutional dimension of the circumstances in which communication of medical data is appropriate or required [36]. In the next section, I consider how this dimension affects the requirements of consent.

Informed consent In the previous section, I have described how the introduction of WGS in clinical practice may affect the roles and expectations in the patient-doctor relationship. This introduction raises a number of issues that concern the normative framework, which structures the patient-doctor relationship. The patient-doctor relationship in personalized medicine is distributed over various persons and extends beyond the length of a particular treatment relationship. In this section, I explore how these shifts affect a basic moral claim of the patient in the medical-ethical framework: the claim to informed consent. Informed consent establishes control by enabling a subject to exercise discretion over the duties of others within a specific context: typically, by consenting the subject waives her relevant rights, thus allowing the other to perform certain specified acts [41]. This, of course, presupposes that the relevant rights and duties are clearly defined. However, I have argued that the roles and expectations in the doctorpatient relationship tend to shift when WGS is introduced to the clinical practice. In this section, I examine the role of consent in these dynamics: how can we continue to guarantee a level of control of the patient when traditional normative frameworks are subject to the blurring of boundaries? It has been argued that the possibilities of WGS exceed the competence of most patients, thus creating a gap between the ideal of autonomous control and actual practice [22]. The consent involved in the context of WGS does not only include choice of various types of information, but must also include issues concerning storage and access. The complexity of the different possibilities, clinical relevance, severity and character of the different afflictions as well as issues of insurance, privacy, storage and recontacting are likely be too much for even educated patients to grasp. In fact, even many professionals might struggle to make sense of all possible outcomes and their implications. In the context of recontacting, specifying what type of information should be communicated is even more difficult, since it might be completely unknown at the moment of the initial test what type of information can surface at a later date. Furthermore, information that is deemed irrelevant today might become important later. A person may, for example, not consider reproduction at a certain moment in time, but later on change his preferences, or in the future, discover new treatment options for diseases that are currently not treatable. Different authors have explored various forms of consent with the intent to do justice to the apparent

mismatch between patient comprehension and the requirement of informed consent [42]. In general, these investigations have demonstrated the importance of context: the specific relation between professional and patient determines what form of consent is appropriate. For example, in a research context, the extent of a duty to disclose results is often thought to depend at least partly on the question whether the researcher and the research participant have a pre-established relationship [20,27, 35, p. 44]. The problem with WGS in clinical practice, however, is that the patient-doctor relationship in which the complexities of the issues are to be discussed, is subject to shifts in roles and responsibilities. This raises the question as to what expectations are justified beyond the classic treatment relation. As the expectations shift, the relation between physician and patient becomes more complicated, and this necessitates redefining and managing the duties, rights and expectations. When personalized medicine becomes a reality, patients will have more complex relationships with a variety of people and institutions that can be captured within the traditional patient-doctor relationship. Furthermore, as we have seen, it is impossible to be sufficiently specific in the initial consent, if only for the reason that at any given point in time, it is unknown what information will become available later. For these reasons, some authors have emphasized that the institutional reliability, rather than consent, should be prioritized [43]. Such a prioritization would inspire trust and a mutually beneficial relationship between institution and participant [5,44]. For example, Allen and McNamara speak of consent in this context as a ‘‘declaration of institutional trust’’ [45]. However, it is easy to see that the notion of institutional trust is insufficient to replace the authority of a patient to determine what happens to her. As we have seen, the point of patient authorization within the normative medical-ethical framework is to protect persons who are in a vulnerable position, by acknowledging their control over the duties of others. The patient in the context of personalized medicine is as vulnerable as ever [16,pp. 589—91],11 which means that the control that the patient exercises in a consent cannot be unilaterally taken away from her and replaced by a notion of ‘‘trust’’, however important trust is to the functioning of institutions. The argument that participants indicate that they trust researchers to deal with the information with integrity [45] does not mean that a declaration to that effect constitutes a valid authorization.12 Trust, by definition, is at odds with the control necessary to effectively protect patients from infringements of their rights. We will thus have to look for new ways of shaping solid authorization in the changing medical landscape. How can we maintain patient authority in the changing clinical practice?

11 Boniolo et al. for example, introduce the term ‘‘trusted consent’’, which has an oxymoronic ring to it [46]. 12 The term ‘‘participant centered consent’’ derives from Katherine Hunt’s term ‘‘patient-centered consent’’ [55]. Although I am indebted to her work, my use of the term is slightly different than hers. Although patient centered, her conception is more strongly conceptualized from within the patient—doctor relationship, although she does also reserve a place for digital patient portals.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

Whole genome sequencing as a diagnostic tool: Participant-centered consent Discussions of consent in the context of WGS tend to focus on the question of specificity: how broad or narrow should the consent be [48—50]? The issue of specificity concerns the consent at a particular time, within the context of the transaction between the doctor and patient. The problem with consent that is too specific is that it is likely to be overly exacting, burdensome and costly as well as giving a false sense of control. Specifically in the case of genetic information, there is a real danger of information overload [51,52]. The problem with consent that is not specific enough is that it becomes hard to see how it can be considered (informed) consent at all [53, p. 30]. My suggestion here is that this frame tends to regard the patient-doctor relation as the default background against which this transaction takes place. As we have seen, the introduction of WGS in clinical practice complicates this relatively clear picture in which the doctor is the knowledgeable provider of information on whom the patient depends for the provision of care. However, if personalized medicine — more specifically, the research-clinical hybrids — become a reality, this image is no longer accurate. Here, a continuous relationship is established between various agents and the proper form of consent needs to reflect this. A number of authors have taken aim at traditional conceptions of consent. For example, Rotimi and Marshall have argued that developments in genomic medicine call for an ‘‘innovative consent’’ that steps beyond the borders of traditional consent in allowing for an interactive approach [47]. However, even though Rotimi mentions the possibility of new media, the understanding of consent in this context remains within the confines of the patient-doctor relationship. Therefore, we need to take patient authorization one step further and take seriously the idea of a ‘‘circular process of communication’’ [54], where the interaction is not based on a single consent, but rather based on a more continuous exercise of authority: a participantcentered approach to consent.13 As Bunnik et al. remark [56] the new media, and in particular the Internet, is particularly well suited to layered ways of offering information and facilitating consent. Participant-centered consent can be imagined as follows: all sequencing information is stored in a single location and related to the patient’s medical history. The patient is enabled to log in on a personalized account, which allows access to various parties (patients, clinicians, researchers), and from this digital platform is able to view clinical files, change permissions, or perhaps withdraw altogether. Default settings are applied when no choice is made. Of course, in order for participant-centered consent to provide a valuable expansion of patient authority, much depends on the way this default is set and what the patient needs to do in order to adjust these. In that sense, it is by no means an infallible alternative, or immune to abuse. Also, a participant-centered consent cannot be a substitute for patient-doctor interaction: the medical-ethical framework, insofar as it applies, remains unaffected by the addition of new forms of authorization. Still, an interactive platform, such as this allows control to the patient, while at the same

13

Think also of issues pertaining to discrimination [43].

7

providing a public service in enabling access for researchers to large-scale databases, in direct relation the personalized information. Allowing the patient to set the permissions is technically not problematic. The question is one of authority: who gets to control the information that is stored? Whether or not a patient is enabled to access the data and attach permissions to it is in that sense a matter of political choice. Database management involves just this: deciding who has access to what. It involves rules on how to create and access information. Given that both narrow and broad forms of consent fail in allowing the patient to dispose over her rights, it follows that a participant-centered approach that allows the patient control over the stored data is called for. A participant-centered approach will have consequences for the way in which we understand consent and the expectations that different persons may have towards each other. In the next section, I mention some of the questions that need to be discussed when we consider this model of patient authorization.

Personalized medicine and the participant perspective As the streams of data in clinical practice become bigger and more complex, the issues of regulation become similar to (or, rather, an extension of) issues of big data ethics [57,58]. The complications that I discussed in section on shifting normative frameworks involve the increasing amount of information that is available and that somehow must be managed in a way that respects the patients, preferable while also helping to realize the ambitious goals of personalized medicine. As I have demonstrated, this raises the question as to how the patient’s rights may be safeguarded when the role and responsibilities of the doctor shift: to what extent and how can the patient exercise control over what happens with information derived from WGS? I have suggested that a participant-centered view of consent may do justice to the changing relationships in this context. I have claimed that the authority of the patient in handling the flow of information in the clinical context needs to reflect these changes. The question however remains how this should be concretely accommodated in the new structures. Recall the quote by Hamburg and Collins in the beginning of this article, who compared the emergence of personalized medicine with the creation of the highway infrastructure. Although Hamburg and Collins emphasize the individual’s discretion to make use of an infrastructure as she sees fit, the other side of the coin is that the infrastructures determine the choices that an individual has. This structure within which individuals operate limit the possibilities of choice. Thus, obviously, decisions made on an institutional level have implications for the possibilities on an individual level.14

14 That this is not a trivial affair is illustrated by the massive information leaks that various (medical) institutions have been faced with: http://www.spaink.net/dutch-data-breaches/.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

8

N. Nijsingh

Firstly, participant-centered consent presumes a structure in which the authority of the participant can be exercised: a digital platform that can be accessed by different parties. This also presupposes that information cannot be swapped between different biobanks, but should rather be accessed from within such a structure. From this structure, we should be able to expect the same virtues that have traditionally shaped the patient-doctor interaction: confidentiality, privacy, a standard of care, and communication. Consequently, the default settings should be set with extreme care, since these determine much of the way in which an individual relates to the institutional framework. A patient should be able to confidently rely on the fact that her information will not be abused when she does not undertake any action. Also, of course, the user interface should be easy to understand and adjustments in restrictions should be transparent and as straightforward as possible. Secondly, the processing of vast amounts of information requires specialists, other than merely medical specialists; think of for example programmers and data analysts. Not only do the possibilities of WGS quickly exhaust the capacities of patients, doctors too are likely to be at a loss when confronted with spreadsheets that contain billions of variables. Therefore, it is of utmost importance that databases containing sequencing data and the interaction of these databases with both researchers and clinical practice is developed with an eye to the ethical and technological complexities involved. This simply means that there is a task for governments, hospitals and research institutions to invest in big data expertise, that is sufficiently safe and user friendly. Lastly, the introduction of participant-centered consent should be accompanied by a thorough re-examination of the relevant ethical and legal frameworks: to what extent can institutional trust be compared to interpersonal trust? How should authority be conceived in the context of personalized medicine? Who can be held responsible if the system fails? What claims can we derive from a notion of a duty of care that is implemented on an institutional level? These questions, and more, are in dire need of further examination. Whatever the answers, the precise shape of consent in personalized medicine and the form that the patient-doctor relationship takes, WGS in clinical practice will require a rethinking of the medical-ethical and research ethical frameworks that apply.

Acknowledgements I am grateful to Maartje Schermer and the members of the OZSW study group Ethics of care and health for the opportunity to present an earlier draft of this paper. In particular, I would like to thank Yrrah Stol whose comments convinced me to rearrange the entire argument. Also, this paper greatly benefited from the comments of Niklas Juth and Dennis Cooley. I am deeply indebted to Anne van Bergen, for many destructive comments and constructive sparring sessions.

Disclosure of interest The author declares that he has no competing interest.

References [1] Burke W, Psaty BM. Personalized medicine in the era of genomics. JAMA 2007;298(14):1682—4. [2] Ginsburg GS, McCarthy JJ. Personalized medicine: revolutionizing drug discovery and patient care. Trends Biotechnol 2001;19(12):491—6. [3] Offit K. Personalized medicine: new genomics, old lessons. Hum Genet 2011;130(1):3—14. [4] McGuire AL, Cho MK, McGuire SE, Caulfield T. The future of personal genomics. Science 2007;317(5845):1687. [5] Lunshof JE, Chadwick R, Church GM. Hippocrates revisited? Old ideals and new realities. Genomic Med 2008;2(1—2):1—3. [6] Collins FS, Morgan M, Patrinos A. The Human Genome Project: lessons from large-scale biology. Science 2003;300(5617):286—90. [7] von Bubnoff A. Next-generation sequencing: the race is on. Cell 2008;132(5):721—3. [8] Mardis ER. Anticipating the $1000 genome. Genome Biol 2006;7(7):112. [9] Berg JS, Khoury MJ, Evans JP. Deploying whole genome sequencing in clinical practice and public health: meeting the challenge one bin at a time. Genet Med 2011;13(6):499—504. [10] Downing GJ. Policy perspectives on the emerging pathways of personalized medicine. Dial Clin Neurosci 2009;11(4):377—87. [11] Scutari M. Personalised medicine: taking a new look at the patient. In: Foundations of Biomedical Knowledge Representation. Springer International Publishing; 2015. p. 139—41. [12] Hamburg MA, Collins FS. The path to personalized medicine. New Engl J Med 2010;363(4):301—4. [13] Knoppers BM, Chadwick R. Human genetic research: emerging trends in ethics. Focus 2006. [14] Knoppers BM, Ma’n HZ, Sénécal K. Return of genetic testing results in the era of whole-genome sequencing. Nat Rev Genet 2015;16(9):553—9. [15] Cambon-Thomsen A, Rial-Sebbag E, Knoppers BM. Trends in ethical and legal frameworks for the use of human biobanks. Eur Respir J 2007;30(2):373—82. [16] Fedder RS. To know or not to know: legal perspectives on genetic privacy and disclosure of an individual’s genetic profile. J Legal Med 2000;21(4):557—92. [17] Snyder L, Leffler C. Ethics manual. Ann Intern Med 2005;142(7):560—82. [18] Hunter AG, Sharpe N, Mullen M, Meschino WS. Ethical, legal, and practical concerns about recontacting patients to inform them of new information: the case in medical genetics. Am J Med Genet 2001;103(4):265—76. [19] Beauchamp TL, Childress JF. Principles of biomedical ethics. USA: Oxford University Press; 2001. [20] Nijsingh N. Consent to epistemic interventions: a contribution to the debate on the right (not) to know. Med Health Care Philos 2015;24:1—8. [21] van El CG, Cornel MC, Borry P, Hastings RJ, Fellmann F, Hodgson SV, et al. Whole-genome sequencing in health care. Eur J Hum Genet 2013;21:S1—5. [22] Clayton EW. Incidental findings in genetics research using archived DNA. J Law Med Ethics 2008;36(2):286—91. [23] Clarke AJ. Managing the ethical challenges of nextgeneration sequencing in genomic medicine. Br Med Bull 2014;111(1):17—30. [24] Hastings R, de Wert G, Fowler B, Krawczak M, Vermeulen E, Bakker E, et al. The changing landscape of genetic testing and its impact on clinical and laboratory services and research in Europe. Eur J Hum Genet 2012;20(9):911—6. [25] Chan B, Facio FM, Eidem H, Hull SC, Biesecker LG, Berkman BE. Genomic inheritances: disclosing individual research results from whole-exome sequencing to deceased participants’ relatives. Am J Bioethics 2012;12(10):1—8.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001

+Model JEMEP-145; No. of Pages 9

ARTICLE IN PRESS

Whole genome sequencing as a diagnostic tool: Participant-centered consent [26] Wright CF, Middleton A, Burton H, Cunningham F, Humphries SE, Hurst J, et al. Policy challenges of clinical genome sequencing. BMJ 2013;347:f6845. [27] Cornelis C, Bolt I, Van Summeren M. Contextualizing genetic testing and sequencing results for patients and parents: the need for empirical-ethical research. Am J Bioethics 2014;14(3):10—2. [28] Stewart M. Patient-centered medicine: transforming the clinical method. Radcliffe Publishing; 2003. [29] Fitzpatrick JL, Hahn C, Costa T, Huggins MJ. The duty to recontact: attitudes of genetics service providers. Am J Hum Genet 1999;64(3):852—60. [30] Vaszar LT, Cho MK, Raffin TA. Privacy issues in personalized medicine. Pharmacogenomics 2003;4(2):107—12. [31] Gurwitz D, Lunshof JE, Altman RB. A call for the creation of personalized medicine databases. Nat Rev Drug Discov 2006;5(1):23—6. [32] Angrist M, Jamal L. Living laboratory: whole-genome sequencing as a learning healthcare enterprise. Clin Genet 2015;87(4):311—8. [33] Beauchamp TL, Saghai Y. The historical foundations of the research-practice distinction in bioethics. Theor Med Bioethics 2012;33(1):45—56. [34] Nijsingh N. Blurring boundaries. Am J Bioethics 2012;12(10):26—7. [35] Bredenoord AL, Kroes HY, Cuppen E, Parker M, van Delden JJ. Disclosure of individual genetic data to research participants: the debate reconsidered. Trends Genet 2011;27(2):41—7. [36] Cho MK. Understanding incidental findings in the context of genetics and genomics. J Law Med Ethics 2008;36(2):280—5. [37] Griffin CA, Axilbund JE, Codori AM, Deise G, May B, Pendergrass C, et al. Patient preferences regarding recontact by cancer genetics clinicians. Fam Cancer 2007;6(3):265—73. [38] Hampel H. Recontacting patients who have tested negative for BRCA1 and BRCA2 mutations: how, who and why? J Genet Counsel 2009;18(6):527—9. [39] Pullman D, Hodgkinson K. Genetic knowledge and moral responsibility: ambiguity at the interface of genetic research and clinical practice. Clin Genet 2006;69(3):199—203. [40] Holtzman NA. ACMG recommendations on incidental findings are flawed scientifically and ethically. Genet Med 2013;15(9):750—1. [41] Beyleveld D, Brownsword R. Consent in the Law. Bloomsbury Publishing; 2007. [42] Foster MW, Sharp RR. Ethical issues in medical-sequencing research: implications of genotype—phenotype studies for individuals and populations. Hum Mol Genet 2006;15(Suppl. 1): R45—9.

9

[43] Altman RB, Benowitz N, Gurwitz D, Lunshof J, Relling M, Lamba J, et al. Genetic nondiscrimination legislation: a critical prerequisite for pharmacogenomics data sharing. Pharmacogenomics 2007:519. [44] Wolf SM, Annas GJ, Elias S. Respecting patient autonomy in clinical genomics: new recommendations on incidental findings go astray. Science 2013;340(6136):1049. [45] Allen J, McNamara B. Reconsidering the value of consent in biobank research. Bioethics 2011;25(3):155—66. [46] Boniolo G, Di Fiore PP, Pece S. Trusted consent and research biobanks: towards a ‘new alliance’ between researchers and donors. Bioethics 2012;26(2):93—100. [47] Rotimi CN, Marshall PA. Tailoring the process of informed consent in genetic and genomic research. Genome Med 2010;2(3):20. [48] Bullock E, Widdows H. Reconsidering consent and biobanking. In: Biobanks and Tissue Research. Netherlands: Springer; 2011. p. 111—25. [49] Grady C, Eckstein L, Berkman B, Brock D, Cook-Deegan R, Fullerton SM, et al. Broad consent for research with biological samples: Workshop conclusions. Am J Bioethics 2015;15(9):34—42. [50] Petrini C. Broad consent, exceptions to consent and the question of using biological samples for research purposes different from the initial collection purpose. Soc Sci Med 2010;70(2):217—20. [51] Juth N, Munthe C. The Ethics of Screening in Health Care and Medicine. Netherlands: Springer; 2012. [52] Nijsingh N. Informed consent and the expansion of newborn screening. In: Dawson A, Verweij M, editors. Ethics, prevention, and public health. Oxford: UP; 2007. p. 198—213. [53] Berkman BE, Hull SC. The ‘‘right not to know’’ in the genomic era: time to break from tradition? Am J Bioethics 2014;14(3):28—31. [54] Mascalzoni D, Hicks A, Pramstaller P, Wjst M. Informed consent in the genomics era. PLoS Med 2008;5(9):e192. [55] Hunt K [Dissertation] Individualizing the informed consent process for whole genome sequencing: a patient directed approach. Phoenix, Arizona: Arizona State University; 2013. [56] Bunnik EM, Janssens AC, Schermer MH. A tiered-layered-staged model for informed consent in personal genome testing. Eur J Hum Genet 2013;21(6):596—601. [57] Issa AM. 10 years of personalizing medicine: how the incorporation of genomic information is changing practice and policy. Pers Med 2015;12(1):1—3. [58] Stein LD. The case for cloud computing in genome informatics. Genome Biol 2010;11(5):207.

Please cite this article in press as: Nijsingh N. Whole genome sequencing as a diagnostic tool: Participant-centered consent. Ethics, Medicine and Public Health (2016), http://dx.doi.org/10.1016/j.jemep.2016.06.001