Population-Scale Sequencing and the Future of Genomic Medicine

Context and Objectives

Genomic medicine, as a field, has the potential to change the way we prevent, manage and treat disease. However, the implementation of genomic medicine in routine clinical care—through “personalised,” “precision” or “stratified” medicine—remains a future prospect due to a range of scientific and social challenges. Frequently cited issues relate to test reliability and validity, cost-effectiveness, health system and workforce readiness, and regulatory and ethical concerns (McCarthy et al. 2013; Deloitte 2015; Milani et al. 2015; Manolio et al. 2015).

Paving the way for a genomic medicine era is in part being facilitated by the evolution of population-scale sequencing initiatives. A diversity of such efforts has emerged in response to scientific and technological advances in genomics over the past 20 years. They have varied aims and objectives, differ in scope and in scale of activities, and in management and governance arrangements. In this paper, we reflect on their variety and evolution, on the learning that they offer, and on implications for future research, policy and practice. We draw insights from national initiatives and international collaborations, both disease-centred and more general in their orientation, but all building on large-scale population-sequencing data. We highlight the social implications of scientific and technological progress in this transformative field, and particularly issues that influence how we might manage risk and reward in genomic medicine.

Methods

We conducted a scoping review, searching Google and Google Scholar for evidence on diverse population-scale sequencing initiatives and key wider literature on this topic (e.g. journal articles and reviews, websites of major initiatives, initiative reports, press releases and news articles). We complemented our search strategy with a snowballing approach. We do not claim to have profiled all population-scale sequencing initiatives that exist; through the initiatives we have reviewed, we have tried to represent the diversity that characterises the field. The 30 initiatives profiled in this study are listed in Table 1.

Below we highlight key insights gained.

Key Findings

Diversity in Form and Purpose

Population-based sequencing initiatives have diverse goals, but we have witnessed a general movement towards more clinically oriented efforts with time. Across the initiatives we profiled, objectives spanned advancing the knowledge base on genetic variation within and across populations; enriching disease specific, clinically relevant insights relating to diagnosis, risk prediction or treatment; the development of new tools and methods for genetic studies; capacity-building (human resource capacity, management and governance, infrastructure); catalysing translation and coordination across initiatives; and other aims (national security, surveillance, commercial services).

The plethora of opportunities that have been created by advances in genomic science are accompanied by a diversity of funders and partnerships, although the funding landscape is still dominated by public and third-sector players. This is in some ways reflective of the state of the field and markets. The rationale for public investment or public-private partnership is partially based on the pre-competitive nature of the research conducted by many of the current initiatives, as well as on government efforts to place their countries at the forefront of genomic research and to help catalyse genomics-based life-science industries. Most projects take place within academic and research centres and in partnerships between them (nationally or internationally). Sequencing tasks are often outsourced to specialist firms (with BGI and Illumina dominating the provider market). Aside from this, only a small number of population-scale initiatives currently engage private-sector partners in the conduct of research.

Progress, Achievements and Impact

Over two decades of progress in this field have yielded numerous achievements and impacts—both direct and indirect (see Figure 1). For example:

• Catalogues of population-specific genetic variation have enabled further research—namely, association studies that compare genetic and phenotypic traits in order to advance knowledge of the genetic basis of disease.
• Biobank-based projects are providing comprehensive, longitudinal datasets that could allow for a greater understanding of the interactions between genes, lifestyle and environmental exposures on phenotypic expression.
• Although still in its early days, progress in the clinical interpretation of genetic variation is starting to inform more effective disease management strategies as well as offering diagnoses to patients who suffer from previously undiagnosed conditions (e.g. rare developmental disorders).
• New sequencing and analysis techniques are opening up opportunities for further progress at pace and at scale.
• Diverse initiatives worldwide have made lasting contributions to the physical and human resource capacity that will allow for future research and innovation.
• Some studies have advanced our collective knowledge on patterns of human migration, divergence and evolution, as well as estimated rates of mutation in modern humans.

But these initiatives have also had a lasting influence on the institutions that govern science more widely. Examples include:

• Transforming informed consent and research ethics practices and experimenting with new models of feeding data back to research participants. Population-sequencing efforts, in particular those that are conducted as part of broader biobanking initiatives, have been at the centre of debates which have advanced group consent and community engagement processes in biomedical research, championed broad consent versus single-use consent principles, and introduced novel models for feeding back research findings. Some initiatives have opted for a full-feedback policy that would include incidental findings; others provide feedback on findings directly related to the core research aim only; and some have opted out of feedback provision (due to the absence of prospects for treatment, absence of informed consent on feedback provision issues, or lack of certainty on the finding implications).
• Catalysing open access research practices and guidelines and driving debate over what is patentable matter. Most initiatives have an open-access policy, for non-commercial, and in some cases commercial, uses of anonymised datasets. Some have “delayed data release provisions” to enable researchers to publish their findings prior to making the data more widely available, and to enable a degree of competitive advantage. The principle of open data-sharing runs into more demanding challenges when the boundaries of what is considered pre-competitive and competitive research become more blurred, and when preserving data anonymity becomes higher risk. There is broad consensus that primary sequence data should not be patentable, but some initiatives allow researchers to claim intellectual property rights on downstream discoveries, and initiatives with a more applied long-term drug discovery ambition place more emphasis on patenting.
• Championing innovation in data management. Within the wider social context in which these initiatives operate, fears of security and privacy breaches have spurred experimentation with data protection strategies bridging technological interventions (code-based systems including two-tiered or multi-level access systems to different types of data), regulatory policy and legal levers (e.g. vetting of researchers applying for access to pseudonymised but potentially identifiable information, access review bodies, requirements for physical presence at data storage premises), and community norms and behaviours (advisory groups, education on risk-minimising behaviours for research participants, emergency response plans). At the other end of the spectrum, one initiative considers it impossible (and unethical) to promise participants confidentiality and anonymity, considering the possibility of individual re-identification as high and at conflict with any consent given on a data confidentiality basis. In this model, participants engage with research only if they can demonstrate an in-depth understanding of the scientific and technological context and of the risks of engagement, and if they volunteer to disclose extensive genomic, phenotypic, clinical and lifestyle data with no confidentiality or privacy clause. Population-sequencing initiatives are at the forefront of debate on how to manage the inevitable tension between data security, privacy and ambitions for facilitating access to a wider range of actors who could help advance the translation of insights into genomic medicine practice.

Regarding the ultimate aim of many of these initiatives—that of translating research findings into new drugs and diagnostics and integrating them into genomic medicine services—larger-scale impacts are yet to accrue despite some promising examples of clinical change and transformation of practice (e.g. developments of drug candidates helped by Iceland's pioneering study; the potential for more comprehensive tools for the management of high-burden diseases in Qatar; integrating whole-exome sequencing for patients with rare, undiagnosed disorders in clinical practice in Estonia). Many of the most clinically oriented projects have also put enabling mechanisms in place in order to accelerate translational research (by building partnerships with industry and clinical services and by laying a conducive national landscape through training programmes for clinicians and investments in databases, registries and interoperable IT infrastructure).

A Future Research and Policy Agenda

It remains to be seen whether the various enabling mechanisms and achievements to date will deliver on their promise to bring genomics into routine clinical care. The regulatory, ethical, legal, scientific and socio-economic challenges to overcome remain substantial, but the implications of doing so are profound. Drawing on our analysis of the findings presented in this study, and our wider experience in science policy, we propose five key areas of action for researchers and policymakers to consider. These are likely to be important for building on current achievements and supporting future efforts:

1. Scope for scaling-up international, interdisciplinary and cross-sector collaboration to enable clinically relevant sense-making from large amounts of distributed genotypic and phenotypic data. This will require data-sharing and collaborative clinical interpretation that crosses disease, disciplinary and professional boundaries, and will call for new ways of designing studies and collaborating. Innovative means of study design will be needed to ensure sample representativeness and optimal collection for personalised medicine innovations.
2. Engaging in new research on the implications of genomic interventions in a clinical setting. This includes research on the health-economics of genomic medicine, capacity-building needs and implementation options, and implications for patient-clinician communications and liability management.
3. Examining the likelihood of further changes in industry R&D models and in rationales for public and private investment in a genomic medicine era. In relation to this, there is a need to address public acceptability issues that accompany the joint pursuit of health and commercial interests. This includes considering the implications of personalised medicine on blurred boundaries between pre-competitive and competitive research, new risk-reward calculations, and potentially changing market segments (e.g. will industry focus on targeted treatments for specific patient profiles across a disease life-cycle, or on specific disease states across multiple patient segments, as discussed in Chataway et al. 2012, 736).
4. Consolidation of learning on the research ethics framework, based on the experiences of prior and current initiatives, and consideration of the legal framework needed for genomic medicine in practice. Although research ethics challenges associated with data security and informed consent remain, and are subject to much debate, the legal framework for dealing with genomic medicine is far more nascent than the research ethics one. The recent landmark case of a woman who is suing a doctor for failing to disclose a family history of hereditary brain disease is illustrative of the legal framework challenges ahead.
5. Wider evaluation and learning will also be central to accountable and effective progress in genomic medicine, for patient benefit. This includes ex-post evaluation of completed efforts and evaluation in real-time of ongoing initiatives, to ensure both formative and summative learning and accountability for the investments made in this transformative field.

List of Profiled Initiatives

The 30 initiatives covered in this study are listed below, sorted according to the scope of their sample collection (either national or international):

• International
  • The Human Genome Diversity Project (HGDP)
  • The International HapMap Consortium
  • The Global Network of Personal Genome Projects (PGP)
  • The 1000 Genomes Project
  • The Human Heredity and Health in Africa (H3Africa) Initiative
  • The African Genome Variation Project
  • The Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium
  • The International Cancer Genome Consortium (ICGC)
• National
  • deCODE genetics (Iceland)
  • The Estonian Biobank/Estonian Genome Centre, University of Tartu (EGCUT)
  • The Singapore Genome Variation Project
  • Genome of the Netherlands (GoNL)
  • GenomeDenmark
  • The Faroe Genome Project (FarGen)
  • Cymru DNA Wales
  • The National Centre for Indigenous Genomics (NCIG) (Australia)
  • Kuwait legislation introducing mandatory DNA testing (no project name)
  • The Precision Medicine Initiative Cohort Program (U.S.)
  • SardiNIA
  • China Kadoorie Biobank (CKB)
  • UK Biobank
  • The Slim Initiative in Genomic Medicine for the Americas (SIGMA) (Mexico)
  • UK10K
  • The Deciphering Developmental Disorders (DDD) Study (UK)
  • Genomics England (The 100,000 Genomes Project) A Weill Cornell Medical Study—Exome Sequencing
  • Identifies Potential Risk Variants for Mendelian
  • Disorders at High Prevalence in Qatar The Saudi National Genome Program
  • The Belgium Medical Genomics Initiative (BeMGI)
  • The Initiative on Rare and Undiagnosed Diseases (Japan)
  • The National Centre for Excellence in Research in Parkinson's Disease (NCER-PD) (Luxembourg)

References

Chataway, J., C. Fry, S. Marjanovic & O. Yaqub. 2012. “Public-private collaborations and partnerships in stratified medicine: making sense of new interactions.” New Biotechnology 29(6): 732–40. doi:10.1016/j.nbt.2012.03.006.

Deloitte. 2015. “Genomics in the UK: An industry study for the Office of Life Sciences.” London: Office for Life Sciences. As of 2 December 2015:
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/464088/BIS-15-543-genomics-in-the-UK.pdf

Manolio, T.A., M. Abramowicz, F. Al-Mulla et al. 2015. “Global implementation of genomic medicine: We are not alone.” Science Translational Medicine 7(290): 1–8. doi:10.1126/scitranslmed.aab0194.

McCarthy, J.J., H.L. McLeod & G.S. Ginsburg. 2013. “Genomic Medicine: A Decade of Successes, Challenges, and Opportunities.” Science Translational Medicine 5(189): 189sr4. doi:10.1126/scitranslmed.3005785.

Milani, L., L. Leitsalu & A. Metspalu. 2015. “An epidemiological perspective of personalized medicine: the Estonian experience.” Journal of Internal Medicine 277(2): 188–200. doi:10.1111/ joim.12320

The research described in this article was conducted by RAND Europe.