Regulating genetic testing, understanding probabilities, and knowing the future

Should we regulate people’s access to their own genome data?.

An image from the NIH that I guess has something to do with DNA and people...Credit: Jane Ades , NHGRI

More on the issue of regulating personal genetic testing, by Ars Technica‘s John Timmer.

The headline is a bit misleading, since the issue is more about whether the “Direct-to-Consumer” (DTC) genetic testing industry should be regulated like any other medical tests or not. Definitely worth reading in full from the link above, and the comments section also has a lot of further clarifications on the matter.

First though, to step back a bit: “if you could know the exact moment of your death in advance, would you choose to find out?”

Personally, I wouldn’t. I’d rather live in blissful ignorance, rather than have some sort of set deadline to wait for. What does this have to do with genetic testing? Well, there’s a lot of information it can provide, and I’d want to exclude any information about incurable diseases from my results (those that I couldn’t prevent no matter what).

I know this is an ethical issue even with genetic testing for parents, who want to find out what the chances are that any kids they have might be born with some incurable genetic disorder…. Of course, as parents you can choose to have or not have kids (not a happy choice, but still, it is a choice.)

But if you were faced with the potential for some disease with no current cure and no means of prevention, you couldn’t do anything about that knowledge, except live on in apprehension, waiting for some day to succumb. And the sad part would be that again, like everything else, it’s a probabilistic thing, not an absolute guarantee. But even though there’s a chance you wouldn’t get the disease, you’d still live with the possibility hanging over your head.

There’s a caveat to my question of whether I’d like to find out my moment of death. I would, if I could do something to delay it. And that could extend to my genetic information as well. For example, if I found out that I had a higher risk of heart disease, I could make even more of an effort to eat healthier, exercise more, etc. And knowing my increased risk could help me extend my lifespan.

In this case, that knowledge could prove incredibly powerful. And that’s why I think it’s really important that we not only get access to the information in our individual genomes, but make sure that we both understand and interpret it appropriately. And that’s where I think the current DTC testing falls short, and could benefit from some regulations and industry standards.

John adds in the comments section,

“If you just wanted to pay Complete Genomics (for example) the $11k or so it costs to sequence your genome and just have them give you back the results with no interpretation, the FDA wouldn’t give a fig. They also don’t care about 23andMe telling you that you have the gene for dry earwax or curly hair or that you can smell asparagus in your urine.

What they do care about are companies making medical claims. If you want to sequence someone’s DNA, or analyze their sequence data, and then tell that person that they have disease risks, that is very clearly medicine, and that is regulated by law, and has been since 1906.

The problem is that the DTC companies have wanted to have their cakes and eat them. They claim their tests are not medical, and then offer BRCA or APOE status, as well as reporting on predispositions for chronic multigenic diseases. Except for the multigenic diseases, you get 3 different answers from 3 different companies, and despite their claims several years ago that they were working together on industry standards, none exist. And that’s where FDA comes in.”

As he says, the FDA wouldn’t care if the companies only gave you non-medical information, but since you do get information about your risks for things like different diseases, genetic disorders, and cancer, it seems to me that they should be regulated like any other medical testing services.

I should point out, as the article does, that the current technology for most of these companies doesn’t actually sequence your genome (you CAN do that if you want to, but that’s a lot more expensive, although the dream is to have a $1000 genome in the near future). What they do use are SNP arrays (single nucleotide polymorphism arrays), which can generally tell you which variant of certain genes you have.

As you might expect, the error rates of SNP arrays are higher than a full-genome sequence when it comes to figuring out what gene variants you have. This means that in some cases the analysis may indicate you have a particular gene variant, when you actually don’t.

That doesn’t seem that serious, but if it just happens to be a gene that generally correlates with a 50% increase in some incurable cancer (for example), you want to be sure. (you’d have to actually sequence that region to find out for sure whether you had that gene variant or not.) Hopefully this example also underlines the power of the information you can get from these tests…

Of course, having a higher error rate is not really the problem with the current tests, and not the reason why I think these tests should be regulated. No, I think the really difficulty is with interpreting these tests.

Humans don’t deal well with probabilities. And genetic information is all about probabilities.

The fact is, it’s complicated. It takes a little while and a little effort to really understand what a 50% increased risk of getting a disease means for an individual. And that’s not to mention the error rate, which means you may not even have the gene on which that increased risk is based on.

There’s a reason there are trained genetic counselors out there, and that’s to explain what the results of genetic tests really mean. Of course, genetic testing/counseling currently mainly helps you figure out about the risks that your unborn children have, but there’s no reason a similar approach wouldn’t work with personal DNA testing. So you can get your genetic information, and attempt to interpret it, but for the vast majority of people you’d have to go to a qualified geneticist/doctor to figure out what it actually means for you. I’d rather that, than people deciding they’re as good as dead because they may have a variant of a gene that in some sub-populations increases the risk of some incurable disease by a certain percentage.

Another commenter points out how blood tests are currently regulated as medical tests, and it seems like that approach would work here as well.

Unfortunately, your regular doctor is going to be just as clueless dealing with this information as anyone else, hence the need to train geneticists, for a future in which this kind of genetic testing becomes more and more important and common.

The industry is just getting started, and I think the reason for a lot of these companies getting setup with SNP arrays now, is to establish a brand and to get people familiar with the idea of personal genetic testing (as well as to figure out the regulatory issues now), for when whole genome sequencing becomes cheap enough to replace the SNP arrays. I don’t think anyone knows what the industry will look like 10 years from now, and how common DTC genetic testing will become, but it’s certainly a powerful tool, and could have huge implications for personalized medicine. But there’s a long way to go – humans, and our genomes, are complex.

And another good comment from the Ars Technica article, has this argument for why you should get on board with DTC genetic testing right now:

Understanding genetic information is hard, but the more time you spend immersing yourself in the world of SNPs, odds ratios and relative risks the easier it gets. People who buy a genome scan now, and spend a little time digging around in their results, will have a head-start in terms of genetic literacy by the time whole genome sequences become available.

Alternatively, you could just sit back and wait 2-4 years for your genome, and then hope your doctor is one of the tiny fraction of medical practitioners who will actually have the faintest clue what to do with the information it contains. Good luck with that.

The companies are certainly trying their best to help popularize genetic testing and educate the public about it. I know 23andMe, for example, which has the cheapest tests and was Time’s invention of the year a few years back, has a pretty good website and also a blog aimed at educating the public (I actually met one of their science writers/bloggers a few years back, she was a former Stanford biosciences PhD student as well).

The company got some press initially for being founded by the wife of Google co-founder Sergey Brin, but the company has certainly been at the forefront of DTC genetic testing, and their blog is pretty good at explaining many of the issues involved.

Here’s a recent post on What DTC Genetic Testing Can—and Can’t—Tell You. Interestingly, the comments includes a response from 23andMe saying that the company is working with Informed DNA, an independent genetic counseling service, to make them available to customers who have questions. While that’s good, I think the original comment being responded to illustrates why leaving that up to the individuals may be insufficient:

I am a 23andme client. I am troubled by many of the 23andme community members who seem to believe that they are more expert at interpreting genetics data for medical conditions than the pharmacological and medical doctors and researchers who’d spent lifetimes to hone their skills and to gain practical experience. They don’t seem to know what they don’t know, but they believe they know. These anosognosic tendencies may prevent them from seeking consultation and thus do themselves harm.

And ultimately, isn’t the goal of a lot of this genetic testing to prevent harm?

Leave a Reply