MS Patient, Ph.D.: Is 23andMe Doing It Right?
I was conflicted about whether direct-to-consumer DNA analysis services were a good idea, a stance I seemingly share with the FDA
As an MS patient, the last few years have been very exciting on the genetic frontier of the disease. Genome-wide association studies have found more than 300 MS genes, although it is not likely to be one individual gene that is the disease culprit (Wang et al., 2011; International Multiple Sclerosis Genetics Consortium, 2013). Researchers are a lot closer to understanding how different genes affect MS, and they might be able to determine which combination of genes makes a person more susceptible to the disease. Now is the time to consider the implications of anyone being able to access their genetic profile.
On the one hand, having personal access to the information encoded in our own DNA is very alluring. But researchers are only beginning to correlate the presence of a gene with a particular disease. Direct-to-consumer kits make this information easily accessible to anyone willing to spend the $100 to $200, but they vary in their ability to accurately predict risk for complex diseases (Kalf et al., 2014). So care must be taken when a patient infers diagnostic information based on genotype results.
A few months ago, my husband persuaded me to order a kit from 23andMe, have them analyze our DNA, and give us our health and ancestry report. It only took a month or so and then I accessed the results on their website, where a summary page shows a glimpse of the health risks, inherited conditions, traits, and drug responses for the single-nucleotide polymorphisms (SNPs) analyzed with the Illumina chip. From there I was able to navigate to a page with a list of all the health risks. Clicking on a particular disease name took me to a page with more details: the SNPs used to calculate the risk of a disease, my odds of having the disease, the percentage attributed to genetics, information about the disease, etc. I was happy to see that all this information is given along with a list of citations that the claims are based on.
My odds of having MS were reported as generally lower than those of the general population. This was somewhat ironic, since I have MS. However, it’s not unexpected that these results are not precise, given that 23andMe calculates these odds based on only two SNPs that have been attributed to MS: the higher-risk versions of the genes for IL-7Ra and HLA-DRB1. Even if all the known genes were incorporated into the odds ratio, there are still unknown factors involved in predicting who will have MS.
What happened next might surprise you.
I forgot about the report. My lifestyle didn’t change at all by receiving this report. I didn’t go running to my doctor to tell him that he misdiagnosed me because 23andMe said I actually have a lower risk for the disease. Or ask him about the many other diseases that the website said I was at a much higher risk for.
Nonetheless, I was worried that the average person using these kits might be alarmed by their results and take drastic measures that would be larger in scale to their actual risk. So in November, when I heard that the FDA issued 23andMe a letter saying that they needed to stop selling and marketing their kits until they had approval for their medical device, I was also on board for more regulation. The FDA claims that these kits are a medical device because they can be used to diagnose diseases and that “potential health consequences … could result from false positive or false negative assessments.”
However, most people who use these DNA analysis services don’t actually change their behaviors or get unnecessarily anxious over the results of these tests, either. It probably helps that for every disease there's an article providing a clinician's perspective—explaining the disease, the genetic loci used to identify them, and answering some common patient questions.
But what I find most interesting is the numbers of willing customers who participate in further research studies, filling out surveys that appear on the site and agreeing to share their SNP information. In one recent example, more than 53,000 customers filled out a survey about their allergies, which led to a genome-wide association study that identified a new genetic locus that was associated with allergies to cats (Hinds et al., 2013).
23andMe isn't (solely) a medical device but a powerful online community that allows customers to be active participants—contributing genetic material and epidemiological information, in return for access to their data. This differs from the normal research subjects that often never know what resulted from their contributions.
I am excited to see connections made between genes, the environment and MS in the next decade of research. Luckily, the International Multiple Sclerosis Genetics Consortium already compiles genetic material from MS patients and controls and has published many of the recent studies that have found MS genes. But researchers can learn from 23andMe to better involve patients in the research, which might help recruit more patients to the studies.
I am no longer advocating for more regulation. Especially since customers of these services seem capable of understanding what their results mean, provided that 23andMe keeps updating the results with the latest research findings and clinical interpretation. Maybe companies like 23andMe will make it so more people are comfortable providing researchers with their genetic material and in turn help them better understand how MS and other diseases get started and progress.