Abstract
This chapter provides a critical reflection upon the sociotechnical expectations that characterized the emergence of 'personal genomics' as a form of direct-to-consumer genetic testing in 2007. Situated in a broader social science debate about 'transformative' accounts of genomics, as representing a significant break with the past, it queries whether 'personal genomics' has amounted to a reconfiguration of the sociotechnical relationships between healthcare organizations, professionals, and patients, in which both the production and use of healthcare information is fundamentally changed.
| Original language | English |
|---|---|
| Title of host publication | Genomics and Society |
| Subtitle of host publication | Ethical, Legal, Cultural and Socioeconomic Implications |
| Publisher | Elsevier |
| Pages | 1-20 |
| Number of pages | 20 |
| ISBN (Electronic) | 9780127999210 |
| ISBN (Print) | 9780124201958 |
| DOIs | |
| Publication status | Published - 2016 |
Bibliographical note
Funding Information:23andMe, deCODEMe, and Navigenics were not the beginning of the DTC genetic testing market. This can be traced back as far as 1994 when University Diagnostics Ltd (UDL) became the first company to offer carrier testing for cystic fibrosis (CF) on a commercial basis in the United Kingdom. The firm initially marketed its test through the Cystic Fibrosis Trust and chose not advertise it more widely. Its Managing Director, Paul Debenham, in giving evidence to a House of Commons Science and Technology Committee in 1995, suggested that CF diagnosis “will be commercially viable because [there is] a desire by the public to know their CF status” but also noted that uptake had been low professionals only and withdrew from the DTC market and soon relocated to the United States. [24] . 3 3 Paul Debenham is now a Director at LGC, the company which bought out UDL, and served as a member of the Human Genetics Commission before its disbanding. He noted to the House of Commons Science and Technology Committee that UDL operated on the basis of “proactive self-regulation” and observed that critical media coverage would ensure that should any company not conduct itself according to the highest standards, it would soon go out of business. However, within a few years, the company ceased to offer this service due to low demand. In 2001, another company called Sciona began selling genetic tests via high street stores most notably the Body Shop, providing consumers with advice about lifestyle and nutrition. In response to criticisms of its activities, Sciona switched to marketing its services through healthcare 4 4 According to letters published on the GeneWatch web site, one reason given by several retailers for either stopping or even starting to sell Sciona’s product in the first place was the lack of credible scientific evidence to support the company’s claims (see http://www.genewatch.org/sub-425647 , accessed 27 April 2011). However, at the time of UDL’s move into the DTC market, British parliamentarians had expectations about a rapid growth in commercial genetic testing services sold to consumers outside of the context of the NHS. This prompted them to call for the establishment of an entirely new regulatory system for the expected deluge of DTC tests. However, as the HGC noted in its 2003 report Genes Direct , there had not been a rapid growth in DTC testing in the late 1990s [25] . While the DTC market in health- and lifestyle-related genetic testing—at least in the British context—faltered, genetic genealogy or ancestry services through companies such as Oxford Ancestors, Family Tree DNA, and DNAPrint Genomics (now ceased trading) flourished. According to Wolinksy [26] , by 2006, these three firms had “attracted more than 300,000 customers.” 5 5 To this figure of 300,000 can be added that of the 160,000 individuals who signed up to the 5-year Genographic Project paying US$99 per test to both help fund the research and receive information about their ancestry in return. In contrast to UDL and Sciona, these firms marketed and sold their services on the web. In the United States, other firms also entered the DTC market in the early 2000s, providing lifestyle and nutrigenetic tests as well as tests of medical significance on the web. This included newly created firms such as DNA Direct and Mygenome and more established operators such as Kimball Genetics and Cygene. In an opinion piece in While the vast majority of companies were located in North America and a number of European countries, firms also emerged in South Korea, India, Japan, Singapore, and Australia. The 23andMe, Navigenics, and deCODEMe services were launched in November 2007 into this heterogeneous and volatile market. While each adopted varying business models, they offered something different to existing DTC firms that sold only individual tests. These new firms utilized the latest genomic sequencing technologies and data from ongoing GWAS to test several thousand different polymorphisms, providing risk estimates for the onset of over 100 disease conditions, information on drug metabolism, as well as on (ostensibly) nonmedically significant questions such as ancestry. Nature published in 2002, the founder of the DTC firm Mygenome Fred D Ledley called for a “consumer charter for genomics services,” declaring that “every individual has a fundamental right to information to obtain genetic information about himself or herself […] The consumer-focused model empowers consumers with direct and private access to genetic tests and services” [27] . By 2007, evidence indicated that there were at least 25 companies located in the United Kingdom, continental Europe, and the United States offering a variety of health-related testing services, ranging from nutrigenetics, carrier testing for single gene disorders, and susceptibility testing for various conditions [28] . A further survey of the industry in 2010 indicated that there were more than 60 firms marketing genetic testing services on the Internet. 6 6 In summer 2010, an unpublished internship project in ESRC Cesagen surveyed the consumer genetic testing industry. Of the 69 services then available, it was found that about a third of services on offer were predictive, carrier, or diagnostic tests for disease conditions, tests for “genetic relatedness” (e.g., genetic paternity) and ancestry accounted for another third, while pharmacogenomic and nutrigenetic testing along with various behavioral or lifestyle tests constituted another third of all the services offered. Given the heterogeneity and volatility of this market, assessments tend to be out-of-date quite quickly. The genome-wide association approach had only just been consolidated as the primary way to investigate the role of genes and their variants in the development of common, complex diseases in 2005. Utilizing commercial genotyping technologies, this approach marked a departure from previous ways of finding disease genes because it interrogated “the entire human genome at levels of resolution previously unattainable, in thousands of unrelated individuals” [29] . Unlike candidate gene studies, it was also said to be “hypothesis-free,” although this is not entirely the case as the GWA approach was predicated on the understanding that common diseases can be explained with reference to “common variants” [single nucleotide polymorphisms—or SNPs], which have a frequency of 1–5% in the population [30] . Using a case–control design, most GWAS have aimed to identify variants shared by groups of people diagnosed with a particular disease condition and which distinguish them from others who are disease-free. 7 7 Case–control GWA studies are based on the assumption that individuals can be sorted with a high degree of confidence into two groups: those affected by disease and those who are disease-free. This relies on the process of clinical diagnosis at a time when disease categories are becoming unstable and multiplying and the absence of recognized symptoms in individuals does not mean that they are necessarily disease-free. One of the most notable groups to use this approach is the Wellcome Trust Case–Control Consortium which received UK£9 million in funding in 2005. Its first study published in June 2007 assessed 14,000 cases of seven common diseases (bipolar disorder, coronary artery disease, Crohn’s disease, hypertension, rheumatoid arthritis, type 1 and type 2 diabetes) and 3000 controls, which were drawn from the 1958 British Birth Cohort and from donors in the UK Blood Service [31] . By 2013, the US NHGRI (National Human Genome Research Institute) catalog of GWAS (which assay at least 100,000 SNPs) contained 1653 studies and shows that more than a 100 loci for 40 common diseases have been identified and replicated [32] . In 2008, a review of the field summarized that: “GWAS have progressed from visionary proposals made when neither the sequence of the human genome nor many variations in this sequence were known, to routine practice of screening 500,000–1,000,000 SNPs in thousands of individuals” [33] . Despite the fact that few large-scale GWAS had been published prior to their launch, the appearance of 23andMe, deCODEMe, and Navigenics in 2007 therefore represented a significant commercial and technical innovation. These firms transformed basic research into the role of gene variants in common, complex disease conducted on different population cohorts into a consumer service, offering users direct access to findings from the latest science that promised to elucidate their individual susceptibility to disease. To do so, firms genotyped customers’ DNA and compared the variants found in their genomes to the variants in the published literature and their statistical association with certain disease states (to give the customer their “relative risk” of a future disease occurring) and statistical data on the incidence of diseases in the population (to determine the “average population risk”) to calculate the average person’s lifetime risk of developing a disease based on certain sociodemographic characteristics. Customers then receive their absolute risk based on these two calculations. 8 8 Alongside innovations in research on genetic associations, the commercialization and standardization of gene chips, and the use of the web to deliver personal genomics services, we might also add the often overlooked sociotechnical innovation in saliva collection and DNA analysis. This is the subject of a forthcoming paper by the author. Therefore, personal genomics firms, while they were not the first to offer DTC genetic testing, were the first to commercialize information from GWAS. However, they were not the first to see that genetic risk information from GWAS could empower individuals to take health decisions about their health. Since the late 1990s, a number of scientists had discussed the idea of using associations found between genetic variants and disease risk to create individual “genomic profiles” that could in turn predict individuals’ chances of developing certain diseases or response to medications [34,35] . In 2003, Collins and colleagues at the US NHGRI published their vision of the future of genomic research. For Collins and his colleagues, developing GWA approaches to the study of genetic variants would indicate to individuals their probable risk of future disease before clinically measured factors such as high blood pressure were even recorded. They saw that this knowledge had the potential to form the basis of an “individualized preventive medicine.” They speculated that: The steps by which genetic risk information would lead to improved health are: [1] an individual obtains genome-based information about his/her own health risks; [2] the individual uses this information to develop an individualized prevention or treatment plan; [3] the individual implements that plan; [4] this leads to improved health; and [5] healthcare costs are reduced. Scrutiny of these assumptions is needed, both to test them and to determine how each step could best be accomplished in different clinical settings. [36] Therefore, the NHGRI vision was based on two propositions: the first was that GWA approaches would produce new knowledge about disease risk which could be applied in a clinical context at the level of the individual patient; and that people would wish to gain this information and act on it accordingly. This vision conjured up the image of a rational person who acts in a thoroughly rationalistic way, seeking out information about their future health and amending their behavior in the present to minimize or avoid the risk of potential disease. As they admit, however, whether people would actually behave in this imagined way required further scrutiny. Yet, some biotechnology entrepreneurs did imagine that some people might behave in this way and would avail themselves of genetic risk information. Fred Ledley, for example, in founding his company Mygenome Inc., envisioned that: Within this decade, the identification of discrete genetic factors involved in healthy development and disease will become routine, laying the foundation for truly personalized medicine in which individuals are empowered not only with self-knowledge of their genetic risk, but also with the ability to take informed actions to prevent disease and preserve health. [27] For Ledley, “truly personalized medicine” was not just a matter of pharmaceutical research; it was about individuals being empowered with self-knowledge of their genome to act in certain rational ways to reduce their risk. While Collins and colleagues imagined that “individualized preventive medicine” would happen within a clinical context, Ledley feared that traditional healthcare systems would only slow the uptake of genomic technologies and looked instead to a consumerist model to ensure their rapid application. By receiving information about their genetic risks directly without the need to go through a doctor, consumers would gain personal control over this form of self-knowledge. The leaders of 23andMe, deCODEMe, and Navigenics came to echo and reinforce these expectations when they launched their services 5 years after Ledley’s call for a “consumer charter for genetics” was published. Whether the leaders of these firms were aware of or had been influenced by Ledley’s intervention is an intriguing question. However, in one further twist, it is of note that as 23andMe, deCODEMe, and Navigenics were preparing to launch, geneticists working on GWAS had already begun to explore the question of whether the associations generated from these studies could be used to predict for individuals their likely future experience of disease The possibility of categorizing “individuals into groups with regard to risk of specific common diseases” was an open question [34] . Prior to the first GWAS being published, genetic epidemiologists argued that individual risk is “essentially impossible to predict” [37] from population-based studies and that risk factors identified in populations have little “discriminatory power at the individual level” [36] . This was a problem manifest in relation to clinical risk factors such as cholesterol or blood pressure. However, would the same also apply to genetic risk factors found by GWAS? In a historical review of the field, the geneticist Leonid Kruglyak noted that given studies had only identified variants with small effects and which even when taken together could only explain a small part of what was understood to be the heritable component of disease, “we also have to ask how we can piece together individual risk from so many small genetic contributions.” [33] . With reference to cardiovascular disease, Humphries et al. [38] did not see that individual risk estimates using GWAS data offered any advantage over conventional risk factors such as blood pressure, age, family history, or blood lipid levels. Other geneticists, however, argued that the knowledge produced by GWAS could be harnessed so as to provide risk estimates to individuals even when the variants involved were of small effect [39] . In sum, then, despite the prospect of “individualized preventive medicine” having been one of the major justifications for this research and its funding, scientists were divided on whether it really would deliver on this prospect. Away from the scientific and technical dimensions to personal genomics, what was new and specific to personal genomics to the way that it delivered genetic risk information to users? Sociologists have recognized since the turn of the century that the Internet has had a transformative effect on how people access and engage with medical and health information [40,41] . The sociologists Sarah Nettleton and Roger Burrows suggest that, with the emergence and expansion of the World Wide Web from the late 1990s onward, a new form of medicine which they call “e-scaped medicine,” has taken shape [40] . The notion of “e-scaped medicine” represents a purposeful play on words, to indicate not only that medical information is now ubiquitous online but that, as a result, it has escaped “and is thus no longer something that can be accessed and, more importantly perhaps, produced and regulated by medical experts” [40] . From this perspective, the emergence of online of DTC genetic testing services and personal genomics are very much a product of this wider transformation in how health and medical information is produced and circulated through online networks. However, the web has not remained the same and in the last 10 years what is commonly known as Web 2.0 has established infrastructures for the creation and sharing of content where users are then expected to actively create, upload, and share content, often at no immediate cost to them. Personal genomics companies—23andMe and deCODEMe in particular—went further than existing online DTC services by bringing together predictive genetic testing and the practices of interactive, “social” media of Web 2.0. Both 23andMe and deCODEMe included features on their web site to allow consumers to share information with others such as friends or family and provide opportunities for social networking. For Lee and Crowley [42] and Levina [43] , it is the social networking aspect of how these web-based services are organized that mark them out from what has gone before. Levina [43] argues that personal genomics is emblematic of new forms of “network subjectivity” that encourages and even requires the constant sharing of information about oneself with others in the network so that the network grows. Despite a number of significant differences from existing DTC services in terms of the information provided and how it was delivered, I would contend that regulatory and policy bodies did not see a fundamental break between personal genomics and what was on offer already to consumers. Both British and US policy advisors and regulators have tended to see that DTC genetic testing services should be subject to regulation and to some kind of premarket review. Back in 1995, when the British House of Commons Science and Technology Committee learned about the DTC selling of CF genetic tests, it warned that: “there is a very real danger that unscrupulous companies may prey upon the public’s fear of disease and genetic disorders and offer inappropriate tests, without adequate counseling and even without adequate laboratory facilities” [24] . In the United States, too, a NIH-DOE (Department of Energy) Task Force on Genetic Testing in 1997 and the US Health Secretary’s Advisory Committee on Genetic Testing (SACGT) in 2000 both came out against the direct advertising and selling of predictive genetic tests to the public [44,45] . Moreover, individual US states such as New York and California had also taken action against DTC companies even before the emergence of personal genomics (for more detailed discussion, see Ref. [46] ). However, a further way in which personal genomics is different to earlier forms of DTC genetic testing is not in terms of the consumer service offered but the way that some firms, 23andMe especially, has used Web 2.0 techniques to establish “research communities” dedicated to various health-related conditions such as Parkinson’s disease. 23andMe stressed that the firm’s services would not only enable individuals to “see what genetics research means for them” but also to share information with others with the idea that “ultimately, they will become part of a community that works together to advance the overall understanding of the human genome” model of doing research. The company and its scientific collaborators went on to publish academic papers to demonstrate that what they called “web-based, participant-driven” research was a viable and valid model for undertaking research on gene association research linked to health and other phenotypic characteristics [47] . The emphasis of 23andMe has been more on the potential of social networking and forms of online community-building that would take concrete form with the launch of its 23andMe’ features in 2009 [48] . 9 9 In July 2009, the company launched the “Do-it-yourself Revolution in Disease Research” which called upon actual and potential consumers to “pledge allegiance” to a particular disease (out of a predetermined list of 10 conditions) on which they would help to support research, and to provide relevant personal data in connection to these diseases. Unlike deCODE Genetics and Navigenics, its business model combined both a consumer service and the establishment of a database comprising genetic, phenotypic, and lifestyle information from customers to provide a platform for further research [49] . 23andMe therefore made use of similar techniques to Facebook with its launch of what it calls “Research Revolution,” a consumer-led research model whereby individuals voted for studies to be undertaken into specific conditions and were offered a hugely discounted price on their tests if they agreed for their data to be used in these studies. The ultimate aim of 23andMe is to create a new [50] . The firm has also worked with other research organizations on further projects on the role of genetic variants in conditions such as Parkinson’s disease and allergies. In 2014, it was the recipient of an NIH grant valued at US . The firm has also worked with other research organizations on further projects on the role of genetic variants in conditions such as Parkinson’s disease and allergies. In 2014, it was the recipient of an NIH grant valued at US$1.3 million to run a 2-year long project to further develop its web-based database and “research engine for genetic discovery” .3 million to run a 2-year long project to further develop its web-based database and “research engine for genetic discovery” [51] . In sum, then, I argue that there are various aspects to personal genomics that are new and specific to personal genomics in relation to what we have seen in the context of DTC genetic testing more broadly: these include the use of GWAS data to predict individual risk of future disease, the use of Web 2.0 techniques to facilitate and encourage sharing of personal genomic data, and the creation of web-based genomic database for research. Given that, how then should we assess the scale of the sociotechnical transformations brought about personal genomics?
Publisher Copyright:
© 2016 Elsevier Inc. All rights reserved..
Keywords
- 23andMe
- Direct-to-consumer genetic testing
- Genetic risk information
- Personal genomics
- Science and technology studies