Human Gene Editing: CRISPRs, Designer Babies, Eugenics, and Healing the Sick

The Story

I will call him Ryan (not his real name), and I can see him as though it was yesterday, not some 20 years ago. Three days old, a bit jaundiced, floppy, somnolent, showing up in my emergency room on night shift accompanied by first-time parents who were appropriately worried that something might be seriously wrong.  On first impression, he was adorable- as cute a child as you will ever see with a beautiful head of light brown hair. On second impression, the pediatrician in me saw something else- he looked elfin- cute, yes- but not right.  While the first thought in a floppy, sick baby is infection, a little alarm in the back of my head was saying “metabolic abnormality”- a genetic defect in biochemistry. 

Being a male infant, and having blood work that did not suggest accumulation of amino acids, I made my best guess at an x-linked recessive genetic disorder presenting in this age group (males have only one X chromosome, so one bad copy is all it takes) and guessed he had a defect in ammonia metabolism (a urea cycle defect, ornithine transcarbamylase- or OTC- deficiency). I was unable to get definitive testing - that would require a children’s hospital- but sent the child to the intensive care nursery on antibiotics, with a diagnosis of OTC deficiency, rule out sepsis.  I got two calls from the director of pediatrics the next day. The first was to ask me what kind of a jerk thinks they can diagnose OTC deficiency in the emergency room.  The second was to tell me I was, unfortunately, correct.  I still can’t tell you exactly what I saw that night that tipped me off- but having seen other children with this condition, I guess it just fit the picture. 

Ryan died at the Children’s Hospital a few days later- diagnosed too late to save him from the ravages of his abnormal biochemistry. Few of the first cases of OTC deficiency in any given family survive- it is only when the siblings come along and are rapidly tested and treated from birth that we can offer any treatment, and that is a lifetime struggle with little margin for error.

The challenge and the Hope

This is the nightmare of every pediatrician- the child you simply cannot help- and genetic disorders loom large in that category. Thus I was particularly touched by a recent piece in Business Insider, describing the plight of young parents who lost a child to a disorder that is, at least in part, determined by genetics.  The condition described is probably affected by multiple genes and other factors, and may in fact not be among the genetic disorders we can most simply address.  But there are many genetic disorders that are determined by a single genetic defect- from Tay-Sachs disease and PKU to cystic fibrosis - and that are either lethal or profoundly disabling.  To date, attempts to repair human genes have been limited in scope or fraught with hazards resulting from the use of viral vectors to carry new genetic content.  Now we seem to have in hand a better tool- but will we be allowed to use it?

Rather than inserting a new gene altogether, we now have technology that allows us to edit genetic errors. There are several versions in play (TALENs, Zinc Finger Nucleases), but the most promising to date appears to be CRISPR. A comprehensive discussion in The New Yorker (Nov. 16) covers this development as well as it can be covered, and I will not repeat the details of the technology here.  Rather, I want to focus on some aspects of the ensuing discussion.

Eugenics versus Gene Repair

Eugenics was a popular concept once, even in the U.S., until the horrors of Nazi Germany gave the whole concept of genetic improvement an enduring and well-deserved bad name. The concept, eliminating genetic defects by eliminating or sterilizing those who carry them, is nowhere near as promising as it may sound.  Dominant mutations with profound effects usually prevent reproduction anyway, and thus almost all cases represent new mutations.  The majority of genetic defects are recessive- we all carry them and only get in trouble when we have two bad copies (or one bad copy on a single X chromosome).   The majority of copies of these defective genes exist in asymptomatic individuals- the bad copy being masked by the good one.  Thus, eliminating the few afflicted individuals does little to reduce the frequency of recessive defective genes.    We had no tools to identify single, masked copies of bad genes and if we could (as we now can), it turns out we all carry them.  In short, eugenics as it was conceptualized in the early part of the preceding century simply will not work, even if one chooses to disregard the fundamental moral and ethical defects of an approach which systematically abuses the less genetically fortunate or supposedly inferior.

Today, we can do better- we can identify and fix the problem at its source by editing defective genes. Further, we may even be able to do this early enough in the embryonic development (or completely enough later) that we correct the defect in germ cells, preventing transmission to the next generation.

This has, unfortunately, almost immediately kicked off a conversation about whether we should allow designer babies and prompted a rehash of some of the horrors of early eugenics. Neither, I would argue, is terribly relevant to the issue of treating genetic disease.

The Specter of Designer Babies

I have conceded that early eugenics was a failure on both moral and scientific grounds- a bad idea that would not work. The designer baby concept is not about correcting an isolated and devastating defect, but rather it is about giving parents a choice of what they want in a child.  This is, for a myriad of reasons, a profoundly bad idea.  Clearly, as we have seen with gender selection in China, such decisions can badly distort populations, leading to major social issues.  More generally, most traits are likely to be highly complex and driven by multiple genes and gene interactions.  Trivial items like eye color may be subject to design- but complex traits like social intelligence, concert pianist, and astrophysicist will not be easy to select among potential embryos or to design via genetic editing.  These, however, are only technical concerns. 

My real angst about designer babies is about what we may think we want- gender bias being perhaps the best extant example. Assuming that we manage to address the gender issue, we are still left with the question of what, exactly, we want to design.  The real strength, real beauty, and real creativity of the human race are not to be found in our similarities, but in our differences.  My fear is that our desires, driven by whatever social “norm” is all the rage (males, multimillionaires, super-soldiers, take your pick) will be far too narrow to meet our very real needs.  Will we create poets, novelists, or fashion designers?  Surely we will fail to design what we do not know we need- who wanted an aerospace engineer or quantum physicist in 1825?  Herein lies the real hazard- who decides (individually or collectively) what we need and how can we assure that the gifts of human diversity remain in our possession?  As a culture, we risk becoming at best a faded remnant of our former selves and at worst reliving the horrors of the holocaust.

The Plea

My plea, and that of many others, is that we not throw out the proverbial (and in this case genetically deficient) baby with the bathwater. We have at hand an opportunity to relieve human suffering on a vast scale- devastating genetic disorders being the most obvious- but many other things far less devastating (lipid disorders, cancer risk factors) may prove subject to intervention.  It is easy to fall into the “slippery slope” argument- that a technology taken to its limits will produce unacceptable results- forgetting that we can and do put limits around technology.  We will need a thoughtful approach that allows for individual choice and ethical consideration least we relive the failures of early eugenics.  I will not pretend to know, for instance, whether conditions such as dwarfism should or should not be “treated” or whether the disease model is even the right concept to apply to what is on the one hand due to an identifiable genetic sequence and on the other hand simply an extreme of the variability we wish to maintain.  The decision is, in my opinion, far simpler in the case of a lethal or profoundly disabling genetic error, less so than in the case of a less disabling or treatable condition. We can do much good and it would seem foolish to me, as a pediatrician, not to do it.