The Ethics of DTC Genetic Testing

It's ten years since the first draft of the human genome was published. This was truly one of the greatest feats the human race has accomplished. A decade later, and a scant sixty years since the double helix of DNA was identified, it's now possible to send a swab from your cheek off to a laboratory and be given a list of alleles that show whether you're more or less prone to certain diseases and conditions. This nascent industry, DTC (or Direct-to-Consumer) Genetic Testing, has been in the news for the past couple of months, and for none of the right reasons.

A short while ago, Berkeley College decided that it'd be fun to invite all new students have some of their genes analysed. The world as one freaked. The problem, as stated in an article in The Chronicle of Higher Education, is that

the substantial intellectual risk is that they'll be institutionally introduced into misunderstanding the precision, interpretation, and historically problematic execution of such research, and the subtle, unexamined undercurrent of coercion in their participation. Until students have a firm comprehension of all those aspects, such projects shouldn't be planned.

 As if that wasn't enough, it happened shortly after the US Government's GAO released the results of an investigation into companies that perform these DTC tests. The report, delightfully titled "Direct-To-Consumer Genetic Tests: Misleading Test Results Are Further Complicated by Deceptive Marketing and Other Questionable Practices", wasn't exactly favourable towards the industry. The title alone is enough to make you think that every company involved is guilty of quackery or incompetence or both, and that's before you read how one donor was told by four different companies that

he was at below-average, average, and above-average risk for prostate cancer and hypertension

And that's just the 'incompetence' part. Except it wasn't incompetence. But more of that anon.

No one seriously doubts that the potential of DTC testing is enormous. It's still a bit in the future, but given the rate at which it's going, it's entirely probable that in another ten years, the field will have advanced to such a stage that it'll be capable of doing what it's claiming to do now. Unfortunately, what it can do now is couched in a great deal of uncertainty, and every finding that's made should be qualified with a major list of provisos.

Given the publicity the field has been getting, and the lies some of its members have been telling (that GAO report is truly appalling), it's no wonder that the most moderate thing people are calling for is regulation. There's a significant crowd that wants — for ethical reasons — to be able to deny people access to information on their own genetic code.

And that's the reason I started writing this in the first place.

I must, at this juncture, point out that I'm just this guy with a blog, and therefore profoundly clueless. What you're read above are simple facts; from here on the opinions will come, and they're worth exactly the same as any other opinion you'd read on the internet.

When Berkeley backtracked in August, the Centre for Genetics and Society wrote that it

applauded [the] cancellation of controversial plans by the University of California, Berkeley to offer individual genetic testing to incoming students as both an educational exercise and a scientific experiment. The University says its researchers will now use students’ genetic samples provided for aggregate analysis only, and will not provide individual results to students.

Hooray!

So what are these ethical issues? The Chronicle article above mentions three; let's call them 'informed consent', 'peer pressure' and 'margin of error'. We'll also add another, more pressing one; 'Data use'.

Informed Consent

The term 'informed consent' normally refers to participants in a drug trial, and it's to ensure that they know what they're letting themselves in for. The context is quite similar here, but there's no trial. Instead, DNA donors (and which male student doesn't spend his time trying to be a DNA donor?) need to understand the results when they get them.

One of the genes that was tested was for alcohol tolerance. The fear was that if one student found out he had a high threshold, he'd get completely rat-arsed every night on the assumption that it wouldn't affect him. Of course, an allele that confers a high tolerance only does so in relative terms; he might be able to handle one or two per cent more alcohol than someone of similar mass but with a different allele (I made those numbers up). What it won't do — and it doesn't take many sentences to explain it this way — is allow him to have another few pints of an evening with no ill effects.

The other two genes that were tested in the Berkeley students engendered similar fears, and could be similarly explained to the students.

Of course, in the wider context of allele identification, this becomes a much greater problem. No one really gives a shit of a bunch of students drink too much, but if the test is for a significantly increased risk of some form of cancer, then it's vital that the testee be told exactly what the results mean and, more importantly, what they don't mean.

So is denying the students their results the best option? No. Better communication is.

Peer Pressure

Was there a 'subtle social coercion' to take part, as the Chronicle suggests? Given that the uptake was apparently about twenty per cent, it seems unlikely. Nonetheless, it could well be a legitimate concern in other circumstances. However, the suggestion that the best way to deal with such subtle coercion is an outright ban is ridiculous. An awareness of the difficulty, coupled with a categorical statement on the voluntary nature of participation, will yield results for those who want them. In other words, better communication.

Margin of Error

The trivial nature of the tests is such that we don't really care if ten students eat an extra cabbage leaf as a result of an erroneous identification. In the larger context, though, this has rather severe implications. The answer, inevitably, better communication, but this may not be enough in this case. Thanks to the fact that DNA evidence has entered the public consciousness as infallibly telling you want you need to know, even someone who's aware of the limitations of allele identification as a predictor may be erroneously concerned if they're given faulty data. The answer? More rigour. Not the wholesale censoring of information.

Data Use

This is the one that's of most concern; to me, anyway. Again, it's not really an issue in the case of Berkeley, but in the grander scheme is the one that has the potential for the greatest harm. Quite simply: are your genetic data yours? Can the lab that sequenced your polymers (fnarr fnarr) do what they like with the information once they get it? The answer should be clear. The case of the Havasupai is illustrative of the point. While few of us would have sympathy for a complaint that DNA evidence contradicted a foundation myth, the fact remains that the data were used in ways that the donors didn't anticipate, and as a result personal information entered the public domain. One could argue that the knowledge gained benefitted humanity, and thus trumped such considerations, but this is nonsense, and extremely short-term thinking. Enlightened self-interest should compel us all to oppose the view that a lab has the absolute right to treat our DNA as theirs to do with as they will. In the longer term, this means that we'll have to protect our privacy with laws that will almost inevitably be more onerous than are required, thus limiting legitimate research and setting the entire field back, possibly by decades.

The State of the Art

A lot of the problem stems from the fact that a lot of people are unaware of just how imprecise these tests are in terms their capacity for prediction. I wouldn't go as far as to call them useless, but all the DTC companies are understandably reticent on the impact environmental factors have on such things. And even when they do provide this information, it's very often couched in impenetrable language that assumes a level of understanding of the science that most likely doesn't exist. I like to think of myself as someone with a decent understanding of probability and a good layman's knowledge of the subject of DNA sequencing, but I'm utterly bemused by one page on DTC company 23andme's web page. Here they give the figures for heart attacks. It strains credulity to breaking point to suggest that for men, 57% of the risk comes from genetics, and that the environmental component — including smoking, blood pressure, activity — is a mere 43%.

The DTC industry is well aware of the limitations of the field and is doing a really shitty job of communicating those limitations. Take that quote from the GAO above. It's entirely possible for three companies to give three different conclusions, and for each of them to be fully able to justify those conclusions based on their own, quite reasonable assumptions. The fact is that humans have approximately 20,000 genes, and that number is far too small for each gene to have one and only one purpose. Even discounting environmental factors, one gene standing alone can only tell you a limited amount of information, and with our current understanding of the genome in toto, no one — no one — can with any certainty give a definitive probability of a particular disease.

Unfortunately, the problem is sufficiently complex that it's tricky to communicate that in an easy-to-digest form. But these companies doubtless have working for them some very bright nerds to do the sequencing and analysis. They should also hire a bunch of people whose sole task is to make the process — including its limitations and benefits — understandable no just to potential clients, but to the world at large.

It's Not All Bad News

Fortunately, Berkeley wasn't the only college to try this, Stanford gave it a go, too, with results that seem a lot more positive. Also, whether as a reaction to recent, ahem, controversies, a number of groups and governments are getting their arses in gear and putting a framework in place that'll help bring confidence back to the industry. None of them is perfect (the ones I've read, anyway), but they're a start. The DTC industry is practically brand new, and to hope that an ideal framework would be in place already would truly be a triumph of optimism. However, they're well worth a read and, as long as the proper regulation is in place (if these are made voluntary, the industry is doomed), there's plenty of reason to look forward to a bright future. The Council of Europe's protocol on genetic testing has a lot of good stuff in it (VII.16.4 notwithstanding), and the UK government's Common Framework of Principles is a good start. And if you're looking for some light reading, the European Society of Human Genetics recently issued a statement on the industry (pdf) that's worth your time.

--

There's a lot more I'd like to say on this subject, but I've gone on far too long already; the rest'll have to wait. This was originally intended to be a couple of paragraphs, and instead I've prattled so much that I reached tl;dr status about fifteen hundred words ago. If you've made it this far, you have my humble thanks for your perseverance, and if you've any thoughts, I hope you'll share them.