The DNA Exchange

On October 25^th, Time Magazine ran an article about genetic testing of children with the provocative title, “What Your Doctor Isn’t Telling You About Your DNA.”  The piece begins by describing a dilemma in the cytogenomics lab at Children’ s Hospital of Philadelphia: a mutation for early-onset dementia is picked up through what the article describes as ‘genome analysis’ (it was microarray, actually) of a sick baby.  The doctors at CHOP, absent any notion of the family’s  preferences, decide that it is not in their best interest to have this information forced upon them – a choice that has drawn the ire of a number of prominent voices in the blogosphere.  “Nice to know that two physicians in Philadelphia not only have medical degrees, but specialize in mind-reading”, says Razib Khan in a post for Discover Magazine.

 Bam!  That noise you hear is the sound of a thousand genetic counselors smacking their foreheads in unison.  Really?  Why are they struggling with this after the fact?  Where was the pre-test counseling?  This was a particularly surprising story for me coming out of CHOP, given that on the same day the Time article was published, University of Pennsylvania genetic counselor Barbara Bernhardt was in Boston at the NSGC meeting describing the NIH-funded work being done at CHOP to document the wishes of families whose kids go through WGS.  So how did this situation happen?  Well, it turns out that clinical microarray testing, genomic or otherwise, does not require informed consent and no guidelines exist for microarray when it comes to return of incidental findings.  As other research coming out of CHOP has shown, physicians who order microarray vary widely in how they handle both the discussions with family members and the return of incidental findings.  For the record, in this case, there was no genetic counselor involved. 

 Would genetic counseling have changed the outcome?  Maybe.  Perhaps the lab personnel at CHOP would be sleeping a little better at night.  But it is important to note that counseling in and of itself is NOT a panacea.  Making sure that families get counseling is only a first step in the process, and the second step – the harder step — is figuring out what those counselors ought to say.  Having spent the past many months chairing the NSGC task force on genetic testing of minors* (too many months; it doesn’t speak well for my stewardship, frankly!) I can tell you with some authority no one can offer you any simple answers.

 The starting point for many genetic counselors is the familiar NSGC position that stresses deferring genetic testing of minors when feasible, in order to preserve the child’s right to decide for him or herself as an adult.  However, even this conditional commitment is far from universal; in the Time piece, Misha Angrist of Duke’s Institute for Genome Sciences and Policy responds: “We think that premise is nonsense…Parents should be given access to this information that’s derived from their bodies and their children’s bodies. This information is for everyone. It’s scary because we have chosen to make it scary. We exacerbate it by treating it like the bogeyman.”  Now I have some sympathy with this position – quite a lot, actually – but I do have to say that for me the prospect of mid-life dementia is pretty scary.  Jus’ saying.

 What I don’t like is the idea of inserting myself as an arbiter of parents’ decisions on behalf of their children, because I don’t think anyone knows my children as well as I do, and I imagine others feel the same.  On the other hand, genetic counselors have had a lot of experience with patients who have the kneejerk reaction ‘I want to know everything,’ and then change their mind when prompted to think through specific scenarios.  So I appreciate what Razib Khan means when he says, “this sort of fiat paternalism on the part of the medical community is frankly going to make enemies of exactly the sort of engaged high-information patients who can be their allies in staving off public hysteria about vaccination and the like.”  But – but – our practice guidelines and our instructions to students have to be relevant to the low-information parent as well, the one less inclined to information gathering – perhaps because some of them are aware that they lack the means, whether that’s financial or personal, to turn warnings about the future into prevention.  Sure, it is very satisfying to turn lemons into lemonade – but if you can’t afford the sugar, it just leaves a sour taste in your mouth.

 And so it comes around again to careful informed consent – which is a problem in and of itself, because everything comes round to informed consent, and we can’t just keep making it longer and longer.  For one thing, it will be an obstacle to clinical use of next-gen testing, since we don’t have that many counselors and if we did, who would compensate them for a consent process that lasts multiple hours?  And what about the poor parents?  Rationally, we have no choice but to design a consent that you can get through in something like half an hour, after which one can expect to encounter only blank faces and autonomic head-bobs from glassy-eyed participants. 

 So, after laying before all of you the Herculean task of imagining an all-inclusive and yet remarkably concise process of pre-test counseling, I guess the least I can do is offer something concrete.  In that spirit, five guidelines for the use of DNA sequencing in minors:

 1. Optimally, pre-test counseling should be a cooperative affair between lab and clinical counselors.  In general, I would suggest that the lab offer a framework for how to approach informed consent, since they have the most experience with testing.  However, the process of informed consent and specific decisions about how to proceed should come from the clinical side, since they have a relationship with the patient.  This is a psycho-social thing, but mostly it is a medical thing, since you cannot stress enough the importance of taking phenotype and family history into account in deciding how to use genetic information.  In the case of the baby at CHOP, would it make a difference if the parents were thinking of having another child?  What if one of the parents was showing signs of dementia? (note: people tend to think about information flowing downwards through the family tree, but it goes back up too, like xylem and phloem!). 

 2. Best practices protect a family’s right to know – AND a family’s right not to know.  I know it can be hard for hard-wired information gatherers to accept, but some people feel just as strongly about not knowing as they do about knowing.  I don’t care if you respect it in your heart so long as you incorporate it into your informed consent procedures.  And to be fair, preventive medicine may not seem like the holy grail to people who are struggling to afford healthy food and dental cleanings.  So try not to judge.

 3. For all that we talk about preserving the child’s right to decide (a worthy goal, all other things being equal), the most important reason to avoid giving out predictive information is that fact that it might be wrong!  Standards for what qualifies as clinically valid information had better be REALLY HIGH, because, so far, our track record as astrologers of the interior stars is not so hot.  Now, what people tend to assume is that it will get better but keep in mind that it will also get worse – when your information is built on tests that were done on people with a certain phenotype or a family history and then you move on to getting genetic information on people without that context, inevitably you are going to find the exceptions – the ones who would never have come to medical attention because they stayed healthy.  It happened with sickle cell anemia and cystic fibrosis, and it will happen with everything else.  Did you know James Watson has Cockayne syndrome?  Yup.

 Back in the day, my embryology teacher had a policy of taking off only some points if you missed an answer – and more points if you wrote down something incorrect.  When you are wrong, she pointed out, your best guess has dangerous implications in the clinical setting.  Words to the wise.

 4. Watch out for cascading liabilities.  Doing WGS on a child cannot make a clinician liable for every genetically-primed event that occurs in their lifetime.  I think it is important to draw a very clear distinction between the right to contact – the right to follow up with a patient when you have significant information that was not anticipated or not available at the time of testing – and the duty to re-contact, which places that burden on the lab, researcher or clinician – turning it into a legally-binding obligation.  The endless, boundless expectation of re-contact is a nightmare for genetics and every informed consent must make these limits clear in advance to all participants.

 5. And my sincere nod to anti-paternalism: parents know their children best.  They need a chance to think these issues through with someone familiar with how these choices may play out (Even high-information parents.  Especially high information parents).  But remember, they need to be counseled – not converted. 

 *NOTE: While I am currently serving as chair of the NSGC Task Force on Genetic Testing of Minors, the opinions expressed here (as always) are mine, and mine alone.

Let’s STOP talking about the $1000 genome!  Please.

Unless you’ve been living under a rock since 2001, you’ve been hearing about the $1000 genome for years.  The inevitable, the holy grail, the game-changer – the ultimate goal post as proclaimed by an entire chorus of Chicken Little genomicists crying, ‘The cost of sequencing is falling!  The cost of sequencing is falling!’  And it’s true! Not only has the cost of sequencing dropped faster than Facebook stock on IPO day, but the product has improved at the same time, in speed, accuracy and coverage.  Will it be transformational?  It already is. 

The early references to the ‘$1000 genome’ were purely aspirational, tossed out in stark contrast to cost of the newly-completed 2.7 billion dollar genome generated by the Human Genome Project.  But the emergence of the $1000 genome as a meme began in earnest in 2005, when the Craig Venter Foundation offered $500,000 to whoever got there first, an incentive that has since evolved into the $10,000,000 Archon X Prize (100 human genomes/30 days/98% accuracy), to be contested in September, 2013.

But all that money aside (seriously, that’s a lot of zeroes!), the phrase, the ‘$1000 genome’ represents much more than a measure of technological prowess.  As George Church describes it, “The ‘$1,000 genome’ has become shorthand for the promise of DNA-sequencing capability made so affordable that individuals might think the once-in-a-lifetime expenditure to have a full personal genome sequence read to a disk for doctors to reference is worthwhile.”  Which is exactly why, as the technical parameters of the challenge have grown clearer and more explicit, the bandying about of the term has grown more and more misleading.

HERE is what the ‘$1000 genome’ DOES NOT MEAN: that getting your DNA sequenced will cost $1000.  This may be self-evident to all the genomics experts competing to win the Archon X prize, but it is anything but obvious to everyone else.   The $1000 figure covers only renewables – those things like reagents and chips that are consumed in the process of sequencing.  It does not include the cost of the sequencer or the cost of the tech who runs the sequencer.  It does not cover overhead or profits.  And most of all, it does not cover the costs associated with interpretation, without which a DNA sequence is merely an endless stream of A’s, C’s, T’s and G’s. 

Sequencing as a “once-in-a-lifetime expenditure”?  More caveats!  Integrated sequencing and interpretive processes make it difficult to re-examine old data, and with both our knowledge base and our sequence quality improving by leaps and bounds, lab experts have assured me that re-sequencing would make more sense than working with data that is even a few years old.  So the whole sequencing-at-birth idea?  Not so much – at least, not yet.

Finally, while the magic of sequencing may lie in the technology that makes it possible, the value of sequencing lies in our ability to translate that technological virtuosity into improved health.  A number of exciting early reports demonstrate the potential health benefits; unfortunately, most of them fail to acknowledge all the ways in which these early adopters do not represent the general public.  A highly-publicized article in Cell in spring 2012 described the experience of Michael Snyder, a molecular geneticist from Stanford who experimented on himself using genetic sequencing followed by a serially repeated, battery of tests designed to monitor his health and biochemistry.  When his genetic sequence showed a variant associated with an elevated risk for type II diabetes, Dr. Snyder added a close monitoring of his blood glucose and other markers for diabetes — a testing regimen unprecedented for someone without risk factors for the disease.  Lo and behold, following an attack of respiratory virus, Dr. Snyder’s blood glucose levels rose to a level consistent with type II diabetes!  The doctor improved his diet and increased his level of exercise, and six months later his blood glucose levels were normal.

Was this, as suggested, a miracle of preventative medicine?  It’s a little hard to know from a sample of one.  Because aggressive monitoring is not done for individuals with no signs or symptoms of diabetes, we don’t know much about the likelihood of transient high blood glucose.  But one thing is indisputable: like the Personal Genome Project participants and other high profile subjects of whole genome sequencing, Dr. Snyder had available to him levels of expertise and medical care that are not in any way typical.  For much of America, paying for routine medical care is a challenge, and paying for acute or chronic medical care the most likely cause of personal bankruptcy.  And even people with money to spare don’t usually get a sit-down with George Church to discuss their most disturbing sequence variants.

Whoever wins the Archon X Prize next year: I salute you.  The $1000 genome is an enormous technical achievement and you deserve every penny!  But let’s not confuse people about what it means.  Let’s not confuse the $1000 genome with the $10,000 interpretation or the $25,000 follow-up.  The meme that represents the future of genetics should not be a bait-and-switch.