Mapping the Personal Genomics Landscape
Last week saw the first annual Genomes, Environments, Traits (GET) Conference, in Cambridge, Massachusetts. Timed to coincide with DNA Day 2010, the conference marked one decade since the publication of the draft consensus human genome sequence. The GET Conference was billed as “the last chance in history to collect everyone with a personal genome sequence on the same stage to share their experiences and discuss the important ways in which personal genomes will affect all of our lives in the coming years.” Not quite everyone with a public personal genome sequence attended – Craig Venter, Desmond Tutu, Glenn Close were all unavailable – but a majority of the genomic pioneers were in attendance and the GET Conference was a one-of-a-kind event.
For those who missed the GET Conference, several high quality recaps are available. The most detailed is A Day Among Genomes, by Carl Zimmer of Discover’s blog The Loom. More targeted reflections on the conference and related events come from Emily Singer of Technology Review summarzing key trends highlighted by the genome pioneers (Singer also has a related piece on the difficulties of understanding human genomes), David Dobbs of Neuron Culture on genomes, cool conferences, and what the hell to tell people about behavioral genes, and Turna Ray of Pharmacogenomics Reporter on the recent Myriad Genetics decision, and its impact on the business of patenting genes. If you’d like even more detail, the Twitter community provided real-time play-by-play.
While there’s no need for a further summary, the GET Conference does provide an occasion to look at the evolving personal genomics landscape in a more holistic fashion.
Genomes GET Personal. Personal genomics refers to the generation and delivery of an individual’s genomic or genetic information. The data itself ranges from testing a single base (referred to as a single nucleotide polymorphism, or SNP) to attempting to sequence each of the approximately six billion bases that make up a human genome. Data generation occurs on a variety of platforms, but it takes more than data to make genomics personal. We must move beyond merely inexpensive genomics to truly personal genomics. That requires analysis of the data, linking it to the life of the individual donor, and, ultimately, using the data in some fashion.
For those of us who frequently read and think about such topics, it’s easy to develop a slightly myopic view of the significance of personal genomics. For example, as Carl Zimmer noted in his review of the GET Conference, it was a challenge to evaluate personal genomics critically “in front of an audience made up of genome scientists, people from the biotech sector, venture capital folks, and other assorted people who are, shall we say, already in the genomic tank.”
The reality is that, to date, personal genomics has been the province of a comparative few. Academic researchers, a fraction of healthcare patients supported by too few providers conversant in clinical genetics, and a handful of companies, entrepreneurs and early adopters striving to deliver genetic information to consumers outside of the clinical setting. But the rest of the world – including a majority of consumers, patients, healthcare providers and payors – is waiting in the wings.
With the cost of generating genomic data dropping, and their possible uses expanding, personal genomics is poised to enter the mainstream. When that happens, certainly by the end of this decade, and possibly far sooner, what will the personal genomics landscape look like? To put it another way, what are the channels or pathways through which ordinary individuals – those of us who are not geneticists or early adopters – will explore their own genomes?
Personal Genomics Pathways. The first step in answering that question is to sketch the personal genomics landscape as it exists today – to understand the pathways through which individuals are currently entering personal genomics.
The following sections outline four different categories of personal genomics: clinical, consumer, research and unintended. Delineating these categoris is not an easy task, and there are frequent examples of companies or technologies that reside in more than one of these four categories. Nevertheless, as the field continues to evolve, mapping the “big picture” can facilitate more precise dialogue, regulatory actions and commercial predictions.
Research. Genomic research is distinguished from the categories described below by its intended use (to improve our understanding of the genetic bases for complex human diseases and traits). But it is important to note that not all genomic research is personal genomic research. This is due to the fact that, in most research settings, genomic information flows in only one direction: from the individual to the researcher. Even aggregate research findings, let alone individualized data, are rarely returned to volunteer participants. Thus, despite the explosion over the past five years of genome-wide association studies (GWAS) and, more recently, the construction of large-scale genomic databases (including the UK Biobank and the Kaiser Permanente Research Program on Genes, Environment, & Health), the vast majority of genomic research does not qualify as personal genomic research.
This is partly due to a timing delay. The proliferation of individual-level genomic research data is a relatively new phenomenon, and research norms have been slow to adapt to a growing body of evidence suggesting that people are interested in learning the results of research carried out using their DNA, and that it is ethical for researchers to return such results. It also reflects some legal uncertainty, specifically whether research conducted (in the United States) in non-CLIA environments can be returned directly to participants without violating federal law. Driven by increasingly vocal calls from both research participants and researchers themselves – including several commentators in the GLR’s What ELSI is New? series – the government agencies that supply the bulk of the funding for genomic research are continuing to examine the issue of genomic data-sharing in the research context.
For the moment, the number of individuals participating in personal genomic research is on the rise. At the GET Conference, George Church provided an update on the Personal Genome Project, which is using unique informed consent protocols to build a research cohort of 100,000 individuals who will have the opportunity to actively participate in personal genomic research, and who will have direct access to their individualized genomic sequence information. The first ten participants (the “PGP-10”) have already made their data available online.
There have also been attempts to develop DTC genomic research initiatives and, while the yields so far have been modest, the model is an intriguing one that promises to involve increasing numbers of individuals in the research aspect of personal genomics.
Clinical. One of the key drivers of the personalized medicine movement is clinical personal genomics. It is defined by its application of genomic data (to clinical care) and its mode of delivering that data to the individual (through a licensed healthcare provider). Extremely wide-ranging, clinical personal genomics has the potential to integrate individualized genetic or genomic information into nearly every aspect of patient care.
Clinical personal genomics includes genetic testing for autosomal dominant genetic traits (e.g., Huntington’s disease), diagnostic testing to predict the likelihood of the development or recurrence of a disease with a known genetic component (e.g., breast cancer) and carrier testing for prospective parents concerned about passing on genetic traits (e.g., cystic fibrosis) to their children. (Arguably, reproductive personal genomics – including carrier testing and other reproductive technologies, such as prenatal testing and pre-implantation genetic diagnosis (PGD) deserve their own category but, since such services are typically offered under the supervision of healthcare providers, they are considered clinical personal genomics in this post.) Clinical personal genomics also involves testing for genetic variants that influence whether and how certain therapeutics will behave in an individual patient, often referred to as pharmacogenetics.
Providers of clinical personal genomics include numerous laboratories offering either FDA-approved genetic testing “kits” or laboratory developed tests (LDTs) (which are not currently regulated by the FDA) targeted at specific genes (as well as at other biomarkers). In addition to the companies that supply the tests or kits, clinical personal genomics also requires genetic counselors, clinical geneticists and other healthcare providers capable of helping patients to understand and act on their genomic data.
A pair of recent announcements by CVS Caremark (acquiring a majority stake in Generation Health) and Medco (acquiring DNA Direct), the country’s two largest pharmacy benefit managers (PBMs), suggest that personal genomics is primed to play an increasingly prominent role in the delivery of medical care. However, there is broad-based concern that there are insufficient numbers of trained healthcare professionals, especially genetic counselors and primary care providers with an adequate understanding of genetics, to handle the expected increase in patients seeking, or needing, clinical personal genomics services.
Consumer. Also referred to as direct-to-consumer (DTC) genomics, the distinguishing features of consumer personal genomics are its intended use (informational, educational or recreational, but not clinical) and its mode of delivery (directly to the consumer, without the requirement of a licensed intermediary).
Consumer genomic services run the gamut from genealogy (Ancestry.com), to paternity (Paternity Experts) to genetic matchmaking (Scientific Match), and everything in between. While some consumer personal genomics services are both popular and uncontroversial (ancestry testing) or are clearly niche products (MyRedHairGene.com), others have straddled the line between consumer and clinical personal genomics, creating confusion for consumers, healthcare professionals and regulators alike.
As an example, 23andMe tests more than half a million SNPs and reports back information relevant to more than 130 traits and conditions, many of which appear unambiguously aimed at influencing their customers’ clinical or medical decision-making. 23andMe also offers a popular genetic genealogy service, and has repeatedly expressed an interest in using its customers as the basis for a personal genomics research platform. What results is a single company with multiple overlapping products that could easily be viewed as a hybrid of the clinical, consumer and research personal genomics types.
What keeps companies like 23andMe in the consumer personal genomics category, at least for the time being, is an insistence on direct-to-consumer access. With a few important exceptions (e.g., New York and Germany), individuals worldwide can purchase and use these services without the involvement of a healthcare provider.
With the list of DTC providers growing rapidly, it can be difficult to keep track of everything that is out there. At present, the only publicly accessible registries of DTC providers are maintained by private entities, including AccessDNA, DNA Test Index, and the Genetics & Public Policy Center (pdf) at Johns Hopkins University.
Recently, however, the NIH launched a new genetic testing registry (GTR) which has the potential to serve as a more comprehensive resource for tracking DTC genomics services. The GTR, which will include providers of both clinical and consumer personal genomics services, is not yet operational. Listing in the GTR is also voluntary so, even once it is in place, it is unlikely to serve as a comprehensive directory of all consumer personal genomics services. There are reports, however, that Representative Patrick Kennedy is attempting to revive the Genomics and Personalized Medicine Act in a form that would include a mandatory genetic testing registry.
Of all of the personal genomics categories listed here, consumer services is the one most likely to rapidly splinter into multiple categories. At the moment, there are few regulations that deal directly with DTC genomics companies and the services they provide. As the generation of genomic data becomes increasingly inexpensive and commonplace, the spectrum of consumer services will expand considerably. As was true of the development of personal genomic research norms, regulatory activity in this area has lagged commercial and scientific development. At some point, however, additional regulations will arrive, helping to further define this category. For instance, it is possible that the GTR will serve as a precursor to a more comprehensive system of regulation for genetic testing. Additional regulation, whatever its impetus, would likely produce further fragmentation within this category, with some companies sliding into defined regulatory boxes and others changing their offerings to avoid regulatory control (and expense).
Predicting precisely which consumer services will be offered and how, if at all, they will be regulated, is impossible. All we know is that personal genomics consumers ten years from now are certain to have many, many more options than they do today.
Unintended. This final category is a catch-all, characterized by a single shared feature: these individuals did not intentionally confront their personal genomic information. At the Genomics Law Report, we have discussed a variety of ways in which an individual might receive an unintended, and possibly unwanted, introduction to personal genomics. Paternity identification, surreptitious testing, genetic testing of a first-degree relative, forensic activity and the re-identification of previously de-identified genetic information all have the capacity to introduce unsuspecting individuals to their genetic information. It’s also possible that individuals who have agreed to share or to explore only certain aspects of their genetic information will be unexpectedly presented with personalized genetic information beyond the originally intended scope of their agreement. No doubt there are other means of unintended exposure as well.
While not every unintended exposure to personal genomic information will be undesirable, such occurrences should clearly be minimized. Although the GET Conference featured a self-selecting audience largely enamored of personal genomics, not every individual shares the desire to peer deeply, or at all, into his or her own genome. An introduction to personal genomics, no matter the context, should be expected, if not always desired (e.g., certain clinical testing), with ample opportunity afforded for pre-test education and, where necessary, informed consent.
Unfortunately, as the cost of generating individualized genomic data declines, more and more such data will be generated. The proliferation of personal genomic data, and the increasing array of valuable applications of such data, is likely to increase the incidence of unintended personal genomics exposures. A combination of public education and policy and legal reforms will be needed to minimize the number of such events and mitigate their impact when they invariably occur.
The Future of Personal Genomics. The categories described above are roughly drawn, and they may well be incomplete. There is no question that they are neither exclusive nor exhaustive. All we really know is this: to the extent that they accurately reflect the current personal genomics landscape, they will not do so for long.
Genomic researchers with novel questions will continue to require novel, and increasingly participatory, research models. Clinical practice will grow and is likely to become simultaneously more specialized (e.g., increasing availability of genetic diagnostic tests) and more generalized (e.g., incorporation of whole-genome sequences into medical records as a default). Consumer personal genomics will go wherever the entrepreneurial imagination can take it and regulatory bodies permit it, leading to splintering and further blurring between its boundaries with other categories.
The 2010 GET Conference closed with the personal genomics company Knome awarding a free exome sequence to the most original audience-supplied idea applying personal genomics. The winning proposal, submitted by Jonathan Eisen, would supplement understanding of our ancestors by sequencing current and ancestral microbiomes. A sampling of the submissions that didn’t win – including sequencing of millions of sperm from an individual to understand germ line variation, replacing newborn blood-spot testing with genomic sequencing, using real-time genetic testing to identify and prevent allergic reactions, constructing encryption keys from an individual’s genomic code and the development of new commercial models to expand access to and participation in personal genomics – provides a glimpse at the untapped applications for personal genomics.
Where will personal genomics head from here? I, for one, am already looking forward to the 2011 version of the GET Conference by which time, if recent history is any guide, this roadmap will already be out of date. And that, without question, is the most exciting thing about personal genomics as we close the book on the 2010 GET Conference.