|November 27, 2013|
This month, the Food and Drug Administration issued a warning letter to direct-to-consumer (DTC) genomics company 23andMe, Inc. asking it to suspend services pending the analytical and clinical validation of its testing. Under the FD&C Act, the FDA has deemed the company’s $99 DTC service to be a device “intended for use in the diagnosis of disease or other conditions or in the cure, mitigation, treatment, or prevention of disease”. Many believe this moment of reckoning will eventually benefit both the company and the public. However, the FDA is not well qualified to address all the ethical issues related to DTC genome testing, and there may be subtle costs of what some have described as FDA overreach.
Despite concerns about the medical (mis)implications of 23andMe data, in the United States, physicians hold a secure monopoly on diagnosing and treating diseases. In the majority of cases, the data they need do not come from the kind of testing 23andMe sells. That some fear a physician would treat a patient wrongly based on results from a $99 DTC test points to larger, systemic problems attributable to lacking evidence and deficient provider education – ones that cannot be solved by FDA regulation. On the other hand, the indefinite storage of human genome data by a profit-making entity raises many ethical questions about data stewardship that remain peripheral to most discussions about the regulation of DTC genomics. It is also noteworthy that 23andMe has failed to communicate with the public transparently about its more controversial patent filings despite the company’s claims that users should retain control of their own data.
As a primer for those who are indifferent to genome wonkery: Whatever method you use, establishing the clinical validity of genomic testing is complicated, in part because the desired outcomes of genomic testing are perennially contested. Unlike many other labs marketing genomic tests without oversight, 23andMe does not examine its customers’ genomes down to every “letter” (nucleotide). This is probably because the process for doing so, called next-generation or genomic sequencing, uses gargantuan storage capacity, arcane analytical methods, and reference data of variable quality. Despite these issues, for conditions with key involvement from one or a few genes, genomic sequencing holds promise if and when we eventually get it right. In a move to quell uncertainty looming over the U.S. sequencing industry, last week the FDA issued its first approval of a next-generation platform.
Genomic sequencing differs significantly from the testing 23andMe has been offering for the past six years (though the company has more ambitious plans). For its core service, 23andMe has been testing each user for several million genetic markers scattered across the 23 chromosome pairs (the “bookshelves” housing most of the human genetic library). The company has then searched population studies for evidence of statistical associations between these markers and various human traits. Based on the quality of that evidence, 23andMe has assigned a simple confidence rating to each trait it reports on and has updated its findings for users over time– two things that other labs are not doing but arguably should be.
Those who would have the FDA regulate DTC genomics are motivated by concerns over the analytical and clinical validity of this approach, which varies by customer ethnicity and for each trait. In the three-year period since the FDA first openly considered regulating genetic testing, 23andMe has amassed data from some 400,000 users, raising a related set of questions about the company’s business model and long-term intentions. This is because large databases of genomic information are now globally recognized as valued commodities, and broader issues of property rights in genomics remain controversial. Circumspection is clearly warranted, as many traits 23andMe reports on are health related, some consumers have inflated notions of what DTC genomics can offer, and clearer standards could help the company provide a more honest and transparent service. It does not necessarily follow that the most important function of DTC genome testing is, or should be, to accurately predict a person’s present and future traits.
In medical genomics today, there is an urgent need to develop new ways of educating and communicating about genomic information. Even when there is weak evidence for association between a genetic marker and trait of interest, there may be residual value in a service that helps customers understand that most individual genetic markers have poor predictive power. Some people learn this in medical encounters when stakes are high; others never learn it at all. Should frightening doctor visits be the only times we ever interact with our own genomic information?
Nuanced communication is a core skill of genetic counselors, but not everyone needs our services. Whatever mistakes 23andMe has made, responsible and user-friendly DTC genomics could still be a powerful antidote to dangerous, reductionist notions about genomics. Along with concerns about validity and long-term data governance, these notions drive much of the anxiety roused by DTC genomics in the first place. This is why we must keep the big picture in sight, and not let a tiff between the FDA and 23andMe hinder creative new ways of communicating and teaching about human genomics.
Leila Jamal, ScM, CGC is a genetic counselor and a PhD student in Bioethics and Health Policy at the Johns Hopkins Berman Institute of Bioethics/Johns Hopkins Bloomberg School of Public Health.