Addiction science, meet translational genomics


What are we to make of efforts to sell genetic predictors of addiction vulnerability directly to ‘consumers?’. Mathews and colleagues [1] provide a comprehensive analysis of the phenomenon of direct-to-consumer (DTC) testing. They argue for strong regulation of genetic tests offered via the internet. A useful table sets forth the companies that offer such testing, including high-profile Silicon Valley participants such as 23&Me, founded by the wife of Google's founder. Tests fall into two basic categories, those that predict addiction vulnerability and those that predict response to known pharmacotherapies. The authors itemize the growing list of concerns raised by critics of DTC testing for a range of diseases or traits, not limited to substance use: imperfect predictive power, lack of clinical utility, failure to provide widely accepted standards of informed consent, and uncertain privacy protections.

Their proposed policy solution, however, falls short. Regulation is not a panacea. Conventional regulatory efforts exercised within the context of national borders are difficult to enforce in the internet era, when blood or saliva samples can be shipped easily across national borders and results delivered via secure, password-protected portals. In the United States, multiple blue-ribbon committees convened by the federal government have recommended enhanced oversight of the introduction of new genetic tests, with few concrete results [2-4]. As Mathews et al. note, the US Food and Drug Administration has expressed concerns and indicated its intention to oversee DTC testing rigorously.

Whether such efforts will be successful is difficult to predict and regulation will be hard-fought. The personalized genomics train has left the station, fueled by ever cheaper technologies of genome sequencing. Many argue that as long as testing companies are required to make full disclosure of the limited efficacy of their products—leaving caution to the discretion of the purchaser—that the responsibility of government has been met. This view casts DTC genetics as an entertaining consumer product, a contemporary version of the ouija board or crystal ball. Some companies market tests for drug response, ancestry and disease risk simultaneously, focusing on the ‘fun’ aspects of testing, such as comparing one's findings to those of family members or creating novel online social networks [5]. They claim they are not offering medical services, and demonstrating either harm or benefit has been difficult. Most studies of individual response to receiving DTC information indicate that impact on health behaviors is small and psychological harms are minimal [6-9]. Thus, achieving regulation will be a struggle.

Would limiting genetic testing to the clinic solve the potential social and privacy harms Matthews and colleagues fear? Most probably not. Tests predicting addiction vulnerability would still convey considerable social stigma, regardless of their technical accuracy, and it is unclear whether those harms would be balanced by clinical benefit. What sort of prevention efforts could follow from demonstrating a 1.3 increase in relative risk? Pharmacogenomic tests—those that predict drug response—are thought to avoid the issues of stigma associated with predicting addiction vulnerability [10]. However, that is not always the case. The genetic markers used to predict drug response often overlap with those that predict vulnerability to a disease or trait, as demonstrated in a recent study using variants in the nicotinic receptor genes to target smoking cessation therapies [11]. There is also considerable overlap in genes studied for addiction and those associated with a range of mental illness, enhancing the potential for stigma and privacy harms.

More intriguing is tracing the very existence of such tests and the intense pressure for translation into the clinic. DTC tests are a symptom of a much more widespread problem in how new and rapidly evolving technologies of genomic analysis are translated prematurely into marketable products, a result of political and economic pressure to capitalize on basic research, moving ever more swiftly from ‘bench to bedside’. This train is moving fast. In the case of addiction, the basic science that undergirds such tests reflects critical work interrogating individual and family differences in rates of addiction. This research was stimulated by the mapping of the human genome 10 years ago and expanded when linked with the increasing prominence of conceptual models focused on addiction as a ‘brain disease’ [12]. Addiction research also overlaps with basic neuroscience studies of learning and memory, providing additional pressure for research. Many of the ethical issues itemized by Matthews et al. are a consequence of premature translation. DTC tests move the early results of research into the consumer's hands without intermediaries. When premarket approval is not required, unvetted test results are delivered, and many believe access to such information about themselves is a market-place right. The translation train also has failed to integrate fully approaches devoted to exploring environmental correlates of addiction or population-focused interventions such as smoking bans or taxes.

Unfortunately, regulation of DTC tests, although a laudable goal, will not repair the flawed dynamics of translational genomic medicine. The current focus on rapid translation inevitably, if inadvertently, leads to premature creation and use of novel genetic tests. DTC testing is a symptom of a wider problem. My argument should not be interpreted as hostility to the use of genomic tools to study addiction susceptibility or to create new tools targeted toward those most vulnerable. These tools are needed. Rather, this is an argument for a simple virtue: patience. The case of DTC genomics reveals a need to slow down, rather than speed up, the translational genomics train.

Declaration of interests

The paper cited in James et al. [9] (a clinical trial) was funded jointly by the Mayo Clinic and the Navigenics company. Funding was awarded as a research contract directly to the Mayo Clinic.


BK's research on the ethical issues in genetics of addiction is supported by R01 DA014577 from the National Institute of Drug Abuse, NIH.