Thoughtful blog post over at Nature recently by Erika Check, on a $25M set of 4 studies that will sequence the exomes of 1500 neonates, whether ill or not. Called the Genomic Sequencing and Newborn Screening Disorders program, it is essentially a pilot study for universal newborn genome sequencing. One could see such a study coming down the pike. But if this is a direction in which medicine is heading, we should be moving like a wary cat, not like a bounding puppy.
The dominant rhetoric for whole-genome screening sketches a benevolent world of preventive care and healthier lifestyles. “One can imagine a day when every newborn will have their genome sequenced at birth,” said Alan Guttmacher, director of NICHD, which co-sponsors the program with the genome Institute. In his genotopian vision, a baby’s sequence “would become a part of the electronic health record that could be used throughout the rest of the child’s life both to think about better prevention but also to be more alert to early clinical manifestations of a disease.”
But deeper in her article, Check responsibly quotes a skeptic, Stephen Kingsmore of Children’s Mercy Hospital and Clinics in Kansas City, who estimates that the program is likely to find 20 false positives for every true positive. In other words, only around 5% of what will loosely be called “disease genes” will in fact lead to disease. One of the reasons for that low rate of true positives is that many of the disease alleles we can screen for concern diseases of old people: Alzheimer’s, various cancers, and so on. Life experience plays a large and still imperfectly understood role in such diseases. Sure, we can test at birth or even before for the SNPs we know correlate with those diseases, but, Check asks, what does that really tell us?
In Guttmacher’s sunny scenario about early prevention, the parents and later the child could be regularly reminded of this individual’s elevated risk. This itself has not only direct health risks but potentially a significant inadvertent impact on the patient’s social life. Everything from the child’s temperament (is she anxious by nature?) to family situation (ill siblings? Alcoholic parent? Suicide?) to many other factors could profoundly modulate how this genetic knowledge would affect the child. Social context matters.
But such an individualized, lifelong health-maintenance program is unlikely ever to be accessible beyond medicine’s most elite customers. Personalized medicine has been around since the ancient Greeks, and, logically enough, it’s expensive. Only the rich have ever been able to afford truly individualized care. “Personalized medicine” seems to have almost as many meanings as people who use the term, but if what you mean by personalized medicine is a physician who knows you as an individual and tracks your healthcare over a significant part of your lifetime, you’re talking about elite medicine.
Medicine for the middle and lower classes tends to be much more anonymous and impersonal. Throughout medical history, the headcount–if they can afford a doctor at all–get more routinized, generalized care. Even many in that fortunate segment of the population today who have health insurance attend clinics where they do not see the same doctor every time. In any given visit, their doctor is likely to know them only by their chart. No one asks, “Has your family situation settled down yet? Are you sleeping better? How’s your new exercise program going?” What you get is a 15-minute appointment, a quick diagnosis, and, usually, a prescription. Genomic technology is unlikely to change this situation. If anything, it will enhance it.
For the hoi polloi, then, personalized medicine will likely mean personalized pharmacology. Some of those most excited about personalized medicine are biotech and pharma companies and their investors, because some of the most promising results from genomic medicine have been new drugs and tests. Should neonatal genome screening become part of routine medical care, middle and lower-class parents would likely be given a report of their child’s genome, the associated disease risks, and a recommended prophylactic drug regimen. Given an elevated risk of high cholesterol or other heart disease, for example, you might be put on statins at an early age. A SNP associated with bipolar disease or schizophrenia might prompt preventive anti-depressants or anti-psychotics. And so forth.
Such a program would be driven first by the principles of conservative medical practice. Medicine plays it safe. If there’s a risk, we minimize it. If you go to the ER with a bad gash, you’ll be put on a course of antibiotics, not because you have an infection but to prevent one. Second, it would be driven by economics. Drug companies obviously want to sell drugs. So they will use direct-to-consumer marketing and whatever other tools they have to do so. That’s their right, and in a comparatively unregulated market, arguably their duty.
But now recall Kingmore’s figure of 20 false positives for every true positive. This may sound high, but again, medical practice is conservative: we’d rather warn you of a disease you won’t get than fail to notify you of a disease you will get. False positives, in other words, are preferable to false negatives. Add to that the scanty state of our knowledge of gene-environment interactions. We are rapidly accumulating mountains of data on associations between SNPs and diseases, but we still know little about how to interpret the risks. We needn’t invoke any paranoid conspiracy theory: that kind of data is devilishly hard to acquire. Science is the art of the soluble.
If Kingmore is even in the ballpark, then, the more neonatal genome screening reaches into the population, the more unnecessary drugs people will be taking. Unnecessary medication of course can have negative effects, especially over the long term. Indeed, the long-term and developmental effects of many medications–especially psychiatric medications–are unknown.
The Genomic Sequencing and Newborn Screening Disorders program is purely an investigative study. Parents in this study won’t even be given their children’s genome reports. But the study is obviously designed to investigate the impact of widespread neonatal whole-genome screening. Currently, all 50 states administer genetic screening for phenylketonuria and other common diseases. The historian Diane Paul has written a superb history of PKU screening. It’s not hard to imagine a similar scenario playing out, with one state leading the way with a bold new program of universal newborn exome screening and, in a decade or two, all other states following its lead.
“Personalized medicine” is a term that’s used increasingly loosely. It covers a multitude of both sins and virtues, from old-fashioned preventive regimens to corporate profiteering. From here, widespread neonatal genome screening looks like an idea that will benefit shareholders more than patients.