Personal Genomes: How will we use them?
Brendan Maher
Tuesday, 28 October 2008 20:05 UTC
As the number of human beings with their genomes fully sequenced ticks higher and direct-to-consumer gene profiling companies push the limits of what medical genetics can do, the once fantastical notion that any given human can walk into a doctor’s office with his or her genome on a hard drive looks more and more like a reality. Still the question remains to be answered: how do we use this wealth of information?
In a special focus issue of Nature issue, published 6 November, Barbara Prainsack and co-authors discuss the challenges presented by direct-to-consumer testing companies and how regulators should respond. Read it free for one month, here
In another commentary, Patrick Taylor offers his thoughts on how privacy protections for medical data that are based on autonomy endanger useful sharing of information without providing protections. Read it free for one month, here
What do you think are the right approaches to regulating consumer genomic testing and protecting privacy?
Updated 06 November 2008 12:25 UTC
-
Replies
-
Prainsack et al make an important contribution to the debate about regulation of the emerging field of consumer genetics. I agree with their view that transposing wholesale the governance frameworks from the world of clinical genetic testing is inappropriate to testing for susceptibility to common diseases. In the world of clinical genetics, governance issues have been driven primarily by the monogenic disease model of high-penetrance genes. To discover that one has the gene for Huntington’s Disease is to know that one will die of that condition if one lives long enough. A couple who discover that they are both carriers of CF genes face profound questions relating to reproductive decision-making. The familial implications of monogenic disease require careful management of the testing process.
Susceptibility testing for common conditions such as heart disease, where genetic factors reveal at most moderate risk elevation, bears little resemblance to the traditional world of clinical genetics. In most respects such tests are more akin to traditional risk predictors such as cholesterol. Hence, it is appropriate that we consider carefully whether the governance arrangements we have developed for clinical genetics are appropriate to these new tests. Do we, for instance, require detailed pre- and post-test counselling for a susceptibility test? Should the provision of such testing be restricted to a small number of laboratories, as is the case in some countries, such as Belgium?
In fact discussion of these issues is well-established, and the need for a nuanced approach to governance is reflected in international standards such as the OECD guidelines on quality assurance for molecular genetic testing. For instance, these guidelines emphasise the importance of counselling but acknowledge that the level of support required will depend on the nature of the test. At a recent meeting convened by the Human Genetics Commission to discuss development of a code of practice for consumer genetics, there was agreement amongst industry and the other stakeholders present that a pragmatic and nuanced approach is required and that such an approach required two things: a set of minimum common requirements which all tests and testing services would be required to meet (such as laboratory quality assurance systems and the provision of accurate and comprehensive information about tests to end users) and a means to identify those tests for which stricter regulation was required (there was no support for the idea that a test for Huntington’s Disease should be made available direct to consumer).
Given that the development of policy is, at least in some countries, already informed by an appreciation of the need for what we might term a risk-based approach to the regulation of genetic tests, we disagree with Prainsack et al’s view that policy action must wait on the delivery of new empirical research findings on how consumers are using the new consumer genetics services. Such research will be extremely valuable and is sorely needed, but we know enough now to be able to take sensible and measured action. To fail to do so would be to indulge in a form of unnecessary genetic exceptionalism; after all if susceptibility tests are akin to cholesterol tests, then why should they not be subject to the same regulatory constraints? To require companies providing such tests to adhere to the laws and regulations governing clinical laboratories and in vitro diagnostic medical devices is not unreasonable and would provide some minimal level of protection for consumers. That such protection is not already in place can not be explained by lack of prior discussion, because the debate about the need for enhanced oversight of genetic tests has been going on for well over a decade. Our first policy challenge is moving beyond recommendations to action to ensure this base level set of protections for consumers.
The second policy challenge will be in deciding where such minimal requirements are not sufficient and what additional regulation is required. At what point, does the degree of heritable risk identified or the potential clinical outcome, become sufficient to warrant that the test be delivered by specialists? Take for instance a genetic test for autism, which a couple might use for reproductive decision-making, or a test for breast cancer, which, if sufficiently high risks were uncovered, might lead to consideration of prophylactic mastectomy. What about a test (such as has been commercialised by Celera) which uncovers not only a heightened risk of heart disease but also information about the individual’s pharmacogenetic response to statins?
Here the policy challenge is that whilst we have mechanisms for deciding which drugs should be available over-the-counter and which require a doctor’s prescription, we have no equivalent for diagnostic tests. The arguments we are having are not new, controversy has flared around DTC pregnancy tests and HIV tests in past decades: what is new is the scale and pace of change. The problem is particularly acute in this field for a variety of reasons, the chief being that it is very simple to go from a new research discovery straight to commercialisation. In general the SNPs are already on the chips. Couple this technological advantage with the new medium of internet delivery as a means to engage directly with the public then you have the perfect storm for consumer diagnostics.
It is in this regard, above all, that we are witnessing something truly novel in the diagnostics sector. The IVD industry has long had a consumer market, and a growing range of tests are available for purchase over the internet or in pharmacies, but never before have we had a situation where new biomarkers move from discovery to DTC provision with no intervening period of gradual adoption by the medical profession.
Prainsack et al are right to suggest that our models for change cannot rest solely on the governance frameworks for clinical genetics, but they are wrong to suggest that we do not have at least some of the tools to hand, or that we must wait before acting. We can and should act now. If the nuanced approach being developed by the Human Genetics Commission in the UK bears fruit then it is likely that many of the concerns about inappropriate governance models may be addressed. But models developed now will probably require refinement as the field develops, and it is here that the kind of research which the authors call for will prove invaluable. Prainsack et al are right to suggest that this industry is still ‘in the making’; policymakers and regulators must show a similar willingness to improvise. Sometimes they will get it right, other times they will get it wrong, but the fear of failure cannot be an excuse for inaction.
Stuart Hogarth
Research Associate
Department of Social Sciences
Loughborough University
s.j.hogarth@lboro.ac.uk -
I fully support Stuart Hogarth’s emphasis on the need for a nuanced approach to governing personal genomics (PG) testing, and I join him in applauding the UK Human Genetics Commission’s efforts in this regard. However, I disagree with Hogarth on the point that ‘the laws and regulations governing clinical laboratories and in vitro diagnostic medical devices’ would be the solution to the problem. Even if we agreed that in many respects PG tests can be considered ‘medical devices’ (as they aim at disease prevention and include the material/technological element of a ‘device’), this begs several further practical and ontological questions: For example, how should we deal with the fact that PG companies derive both health-related and non-health related information from genomic data – should medical device regulation apply to certain contexts of interpretation of these data and not to others? Furthermore, do we really want a situation of ‘technological essentialism’ in which any information aiming at disease prevention, however vague and however small its clinical utility, should automatically require regulation based on the mere fact that it entails a ‘device’ element?
The extension of certain elements of existing regulatory regimes of in vitro diagnostic medical devices might indeed turn out to be the right answer to some of the challenges posed by the PG market. However, in the process of deciding to what extent this is the case, we should also keep in mind that these existing regulatory regimes do not address some crucial novel issues posed by PG. PG testing services are different from ‘traditional’ genetic testing in the sense that they do not test for the presence or absence of a certain disease-causing mutation but instead create large datasets from which information on a (potentially) unlimited number and range of phenotypes can be derived. This highlights the need to reconsider our use of categories such as ‘health-related’ vs. ‘non health-related’, ‘clinically relevant’ vs. ‘non-clinically relevant’, and ‘medical’ vs. ‘non-medical’. Similarly, we might also want to redefine who, and by what mechanisms, should be making these decisions on our behalf.
Barbara Prainsack
King’s College London
Centre for Biomedicine & Society (CBAS)
London WC2R 2LS, UK
barbara.prainsack@kcl.ac.uk
-
