Tải bản đầy đủ - 0 (trang)
5 New Regimes for Common Genetic Data Repositories: ClinVar and BRCA Share

5 New Regimes for Common Genetic Data Repositories: ClinVar and BRCA Share

Tải bản đầy đủ - 0trang

Premises for Clinical Genetics Data Governance: Grappling with Diverse Value Logics



245



genetic data, with Quest licensing the data and forming sub-license agreements

with participants. Content-wise, it builds upon the Universal Mutation Database

that has been built through data sharing between 16 French laboratories over a

decade, based on Inserm’s gene-data curation process. To participate in BRCA

Share, participants will commit to sharing past, present and future data in order

to get access to the database. Commercial labs will pay according to their size,

while research entities will get access at no cost. This financial arrangement allows

BRCA Share to emphasize data curation arrangements to attend to data quality and

to conduct functional studies on the effects of mutations, without depending on

research or public funding. This alternative model was not received positively by

all: a Nature editorial describes the BRCA Share initiative as a “walled garden” and

laments its lack of sharing with the ClinVar open database (Nature Editorial 2015).

Throughout the past two decades various alternative configurations for BRCA

data collection and sharing have been established. In each of those configurations

the distribution of rights and obligations for data depositing, accessing and curating

differs significantly. In the table that follows (Table 1) we summarise these key

aspects for each configuration.



3 Unique Data Characteristics and Diverse Actors’ Logics

3.1 Data Characteristics

The emergence of BRCA data repositories over the past two decades is in itself a

clear indication that access to such data is considered a valuable resource amongst

the various stakeholders involved. However, what constitutes the value of such

access can vary across stakeholder groups. For clinical purposes, the reuse of

previous assessments of clinical significance is core. For research purposes, the

fact that a variant has – or has not – been previously observed can be of value in

its own right. BRCA data sharing means that variant-specific information can be

looked-up by scientists when a specific variant is re-encountered during testing.

Scientists evaluate past variant assessments and use available data as resources to

facilitate and enhance their own assessment. Due to the nature of the BRCA genes

there are thousands of possible variants, and new rare variants are continuously

identified. Both BRCA1 and BRCA2 are relatively “large-sized” genes (BRCA1 has

24 exons, BRCA2 has 27 exons) that include unusually long exons 11 where high

rates of harmful variation occur. Examples of variations are: changes in sequence,

changes in amount, and changes in position (Human Genome Variation Society

2007). Looking-up variants in the common repositories is possible because they

are expressed according to one or more standard typologies in structured ways.

This makes possible the unequivocal sharing of data on such complex scientific

objects (although the relevant nomenclatures are still being extended and revised).

BRCA1 and BRCA2 are related to monogenic predisposition syndromes (e.g. for



246



P. Vassilakopoulou et al.



Table 1 Alternative configurations for BRCA data collection and sharing

Breast

Information

Core (BIC)



Myriad’s

database



Sharing

Clinical

Reports

(SCRP) and

Free the Data



ClinVar



BRCA share



Description

Repository for

data on BRCA

variant

occurrences and

their

assessments.



Depositing

Any genetic test

performer after a

registration

process

(individual

investigators,

research, clinical

and commercial

labs).

Repository for

One specific

data on BRCA branch of private

variant

genetic testing

occurrences and labs.

their

assessments.

Initiative for the Any genetic test

collection of

recipient

data on BRCA (clinicians and

variant

patients).

occurrences and

their

assessments.

Repository for

Any scientific

consolidated

group that holds

data on variant assessment data

assessments.

(from clinical

testing, research,

literature).

Repository for

Genetic test

data on BRCA performers

variant

(research and

occurrences and clinical, private

their

and public) that

assessments

commit to share

past, present and

future data.



Accessing and reuse Curating

Anyone after a

By members on a

registration process. voluntary basis.

Unrestricted reuse if

source is

acknowledged.



Within the company. Within the

company.



Unrestricted (data

are submitted to

ClinVar).



No curation.



Unrestricted.



No curation.



Only for

participants: paid

access for

commercial labs, no

cost for research

entities.



Professionally

curated –

curation is part of

the configuration.



breast cancer), meaning that a positive find in this gene can be conclusive without

further analysis of the genome. Practically, this means that the assessment of a

variant’s clinical significance should be valid for any person anywhere, i.e. variant

assessments are globally relevant.

To sum up, BRCA variant assessment data have a number of interesting

characteristics: their significance is high because they convey information related

to human life and disease susceptibility, they are products of highly specialised scientific knowledge production processes, they relate to specific genetic material that

shows high variability resulting to high numbers of rare occurrences, they convey

information which is globally relevant and they are portable and combinable due to



Premises for Clinical Genetics Data Governance: Grappling with Diverse Value Logics



247



standardisation. Practically, each new variant assessment is both a person-specific

diagnosis and a contribution to general body of genetic knowledge. This is one of

the reasons why BRCA data are objectified and sought after. By conceptualising the

information content of BRCA data is in this way, the moral obligation to make this

information available to anyone that needs to decide on a medical course of action

after genetic testing can be argued. This information is not simply an outcome of

routine lab activities but entails the exercise of scientific judgement, the employment

of specialised support systems and reflects the capabilities and resources available

in the lab where the data are generated. Labs that aim to maintain a competitive edge

(ensuring their subsistence) are reluctant to share their data and make them available

for reuse.



3.2 Research, Commercial and Clinical Logics at Play

The various groups of actors in the field operate according to different interests

and agendas, and this shapes their stance in relation to data sharing. We will in the

following distinguish between three different logics; a research logic, a commercial

logic and a clinical logic. These logics differ, overlap, and engage with each other

in non-trivial ways.

A research logic has the primary aim to further knowledge production, and thus

openness, sharing and transparency are highly valued. The orientation towards data

openness in genetic research is exemplified by the Bermuda principles that were

established in 1996 within the Human Genome Project and stipulated that DNA

sequence data should be published (i.e. uploaded to public repositories) within

24 h of production. These principles introduced institutional measures for rapid,

pre-publication data sharing, not only post-publication that had been common

(Contreras 2011) and were adopted by funding and policy entities such as e.g.

the US National Institute of Health and the NHGRI (National Human Genome

Research Institute). In the research field, funders and publishers of research are in

a position of power from which they can enforce data sharing. Scientific journals

demand disclosure upon publication. Human Mutation was the first journal to adopt

a full data sharing requirement, in 2010, and the European Journal of Human

Genetics has gone a step further, hiring curators to check that each paper’s variant

descriptions have been accurately transmitted to a public database (Krol 2014). This

regulatory model which ensures data sharing through demands to deposit data in

public repositories before a paper is accepted for publishing brings with it some

risks. The primary motivation for data sharing may be to get the paper published,

and the need to carefully curate, maintain and update the data deposited may take a

secondary importance. This, together with lack of sustained and predictable funding,

may explain the quality concerns of such open databases.

Within a commercial logic proprietary access to data equals competitive advantage and profit potential, as the case of Myriad exemplifies. Furthermore, a

commercial company must have sufficient public trust and confidence. Myriad



248



P. Vassilakopoulou et al.



emphasizes the quality and reliability of the information contained in its own

variant databases to gain customers’ trust. For instance, in a recent study conducted

by Myriad, the quality of data is compared across five different BRCA related

databases, BIC, ClinVar, HGMD Pro, LOVD and UMD (Vail et al. 2015). The

researchers investigated a set of 1.327 variants from Myriad’s database comparing

the classifications for internal and cross-database consistency. The general tendency

was that different sources more often than not reach what the authors characterise

as conflicting conclusions. Based on these findings Vail et al. argue that lack of

governance mechanisms that can assure proper quality control and adherence to

current standards regarding methods, nomenclatures and documentation, render the

open databases a potentially dangerous source of information for clinical purposes

and that “healthcare providers should exercise caution when using these research

tools for clinical purposes.”

A clinical logic primarily seeks resolution of questions regarding an individual

patient’s situation. The work of a healthcare provider is characterised by a pragmatic

search for relevant input to decision making following the Hippocratic mind-set of

benefitting and never harming the patient. This includes seeking for evidence in the

genetic data repositories. Given the varying quality of the available information,

a clinician that approaches the market to order a genetic test for a patient is

thus susceptible to the ‘superior quality’ argument that a specialised company

like Myriad can offer. Moreover, within this clinical logic, the primary aim is

not to contribute to knowledge production. This goes for individual clinicians

that use commercial services as well as for clinical laboratories that do in-house

testing, e.g. clinical genetic departments in hospitals. In these laboratories there

may be researchers who publish scientific results, but there are no comprehensive

mechanisms to ensure that the knowledge accumulated via the routine work of

clinical testing of patients is shared. Usually scientists within clinical laboratories

do look-up the variants identified in the open genetic repositories (so they have a

strong interest in having access to such repositories) but their own data including

their assessments of variants’ significance are accumulated in local tools (e.g. inhouse databases). For non-researchers there are few, if any, incentives to deposit

these data in shared repositories. The ensuing “silo effect” of isolated repositories

is problematic when we deal with a body of knowledge that is still young and under

rapid and continuous development and growth claims (Rehm et al. 2015) as it can

have life-or-death consequences for patients.

We have singled-out the three more influential logics shaping the field (research

logic, commercial logic and clinical logic). Nevertheless, in the current context

of clinical utilisation of genetic data there are more actors that add complexity to

the picture. The perspectives of individual patients, public health policy makers,

regulatory agencies, and funding bodies (public and private insurance funds)

are also important. The multiplicity of actors with different resources, interests,

engagement modes and time-horizons combined with the exceptional connectivity

afforded by current information and communication technologies and the continuous advancements in gene sequencing technologies create a very dynamic and

complex landscape.



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

5 New Regimes for Common Genetic Data Repositories: ClinVar and BRCA Share

Tải bản đầy đủ ngay(0 tr)

×