The Calculated Child

Polygenic embryo screening

“Eugene suffered under a different burden — the burden of perfection.”

In the dystopian world depicted by the movie Gattaca, Vincent masquerades as the genetically superior Eugene (meaning “well-born”) to join the astronaut program, which his “in-valid” status from birth previously disqualified him. Eugene was born with social and intellectual advantages based on genetic selection and modification as an embryo, but a minor failure in a sports competition was a crushing weight for one created “perfect.” He failed a suicide attempt that left him paralyzed. Afterward, he found fulfilment helping Vincent (now called Jerome) to achieve his dream and deceive the cruel system of discrimination.

As reproductive technology continues to absorb advances in genetics, I am musing about closing in on the path to a Gattaca-like society.

A Brief History of Embryo Screening

The story can begin in different places or from other breakthroughs, but I shall start in Cambridge, where preimplantation embryos were genetically screened for the first time. Bob Edwards and Richard Gardner identified female rabbit blastocysts by detecting the sex chromatin body. The sexed embryos were segregated into male and female groups for transfer to surrogate mothers, achieving a predicted 100% separation. But the striking result for its time seemed too arcane for an application before the launch of clinical IVF and the polymerase chain reaction (PCR) gave the precision needed for preimplantation genetic testing (PGT). Validation came in 1990 when Alan Handyside and Robert Winston announced the birth of twin girls selected at a preimplantation stage to avoid a sex-linked disease.

New technology is quickly adopted by clinical embryology. PGT avoided a growing list of single-gene disorders and was used to check chromosome copy number/ structure, respectively called PGT-M (M for monogenic disorders) and PGT-A (A for aneuploidy). The moral case for avoiding congenital diseases is undeniable, and while aneuploidy screening has mixed results and stands accused of wasting healthy embryos, it has helped to reduce multiple pregnancies by the transfer of only one or two embryos at a time.

Advances in culture technology enabled embryos to reach the blastocyst stage. More cells can be biopsied from the trophectoderm than just one or two blastomeres during cleavage (polar body testing is almost redundant). Meanwhile, great strides in genetics after the first full draft of The Human Genome Project in 2003 provided gene sequence data from the 20,000+ protein-coding genes for testing DNA in cancer and potentially thousands of heritable diseases. Some diseases are monogenic, like cystic fibrosis, sickle cell anemia, and Tay-Sachs disease. They are rare, congenital, frequently severe, and acquired by Mendelian inheritance (dominant/ recessive, autosomal/ sex-linked). Polygenic diseases are more common, non-Mendelian, and affected by environmental conditions, sex, and lifestyle. They tend to develop later, like type 2 diabetes, coronary artery disease, and Alzheimer’s disease. The distinction has immense practical consequences for testing.

In a quirk of scientific history, while Bob Edwards laid the foundations for PGT of single gene defects, his wife, Ruth, had completed her Edinburgh Ph.D. on quantitative genetics, which is close to but not quite identical to polygenic inheritance. Douglas Falconer, a pioneer in the field, supervised her project on the polygenic inheritance of body size in animals by linkage mapping from breeding experiments, the standard method before genome sequencing. The three of them would be astonished to know that PGT is now used to test embryos for susceptibility to polygenic diseases such as type 2 diabetes (which Douglas himself had).

Polygenic Futures

Without laboring the history of DNA analysis, a standard method used for embryos is Next Generation gene sequencing (“Next” is now a misnomer). DNA strands are fragmented before high-throughput, massively parallel sequencing, followed by complex bioinformatics to identify variations at millions of loci, or single nucleotide polymorphisms (SNPs).

The data from 22 autosomes and two sex chromosomes are interrogated using genome-wide association studies (GWAS) based on biobanks for linking genetic, environmental, and lifestyle factors to disease. A prime example of that database is the U.K. Biobank, containing over half a million entries from men and women of middle- to older age, mostly of Caucasian descent.

Anyone who has imported their raw DNA data from a testing service to Promethease is familiar with the depths of information. A long list of SNPs helps to build family trees in ancestry research. For instance, I have the “A’ allele rs1426654(A) common to light-skinned Europeans, and M269 for a male with Celtic ancestry. But the markers are far more valuable in research and medical care by linking predispositions to specific diseases. They also report associations with physical traits like height and eye color. I paused every time I found a SNP with the slightest risk score for disease, although no alarm was necessary. When data cause real concern, they need expert interpretation, and the same professionalism is needed when PGT is used for polygenic embryo screening, called PGT-P.
This is now a hot topic in reproductive medicine and has drawn criticism from medical authorities for making unproven claims and wild advertisements. For example, posters on New York subway station walls exhort young adults to Have Your Best Baby. Commercial services use sophisticated algorithms to assign polygenic risk scores for “disease risk reduction” in babies. There is a temptation to take the numbers at face value and overlook their tentative meaning. If one embryo has an estimated lifetime risk of, say, 8% for type 2 diabetes and another 13%, most people will choose the higher percentage—no matter how trivial the difference or complicated the deep interpretation.

The Process

· A few trophectoderm cells are removed from each blastocyst.

· The blastocysts are vitrified for temporary storage at a low temperature.

· The biopsied DNA is amplified and genotyped with Next Generation sequencing or an equivalent genome technology.

· Customized bioinformatics estimates the statistical probability of a range of diseases and specific traits using GWAS data.

· The “optimal” embryo is identified with the lowest polygenic risk score.

· The embryo is devitrified for transfer at an appropriate stage of the menstrual cycle.

The Stumbling Blocks

PGT-P is based on genuine science, not a fraudulent technology. Diving into the embryonic genome is an extra step that can seamlessly integrate with conventional fertility care. The potential market is huge as IVF continues to expand worldwide, even to people who are not infertile. In a competitive marketplace with few regulatory constraints, we can expect more U.S. clinics will adopt PGT-P, followed by services in the Middle East, Singapore, and China. Europe will stand aloof, though powerless to stop cross-border traffic by patients. It is illegal in the U.K., but a loophole gives patients the right to their full-length genomic data from PGT-A, which they can send to American companies for analysis. Then, back in their home clinic, let’s imagine their physician advising them to transfer Embryo X based on its karyotype and morphology, but the parents insist on Embryo Y because it has a more favorable polygenic score for diseases they worry about. Who will forbid them? It could put their physician in a dilemma, who wants to provide the best care but feels control of treatment slipping away, and even agonizes over ethics.

How should they advise patients who want binary answers, not statistical probabilities? Who can guide decisions that are complicated by pleiotropy when an allele for a lower risk of condition A has a higher risk of B? For a real-life example, there is an associated risk of autoimmune disease with rs231779, but if an embryo is deselected for that SNP, it loses protection from melanoma. And if an allele is chosen for being innocuous in the present circumstances, it might be hazardous when the child grows up to live in a different environment, never anticipated by the parents or physician.

Patients need to beware if they have few embryos for selection. They should be prepared for the possibility of disappointment if a disease they thought ducked turns up later in their child, or if the trait they desired is not manifest. Patients might get angry. Physicians might worry if they are liable for a “mistake” despite careful informed consent and/ or recruiting a genetic consultant. Although they know an “optimal” embryo is a fairy story, how many other people do?

The Hard Edge

Critics worry that polygenic embryo tests could become a slippery slope. It was a familiar rant in the genesis years of IVF. But there is a specter, even though it is not genetic enhancement (at least yet). It is the expected evolution of PGT-P from disease avoidance to selecting the physical and cognitive traits that society values—height, skin color, athletic or musical talent, and IQ—the most contentious.

This is hard because it is another edge for people who are already privileged through wealth and connections, tilting society toward ever greater stratification. And yet, the edge is still blunt. What is the value of a few points higher on the IQ scale or a couple of extra inches at adulthood? Will they lift grade scores for college entrance or help career advancement? Academic brilliance and Olympic achievement can’t be gained in this way because other factors swamp slightly favorable genes. And then, if the “smarter embryo” is implanted, how will parents react if their children become college dropouts or develop diabetes? Disappointment may have a non-genetic cause, but the psychological burden on the subject could be heavy, echoing Eugene’s sense of failure.

Sometimes it is better if knowledge is left unknown, and to embrace ignorance as a kind of wisdom. I doubt that perverse philosophy will persuade anyone, but it seems pertinent to ask.

In Conclusion

Polygenic embryo testing is “soft selection,” not the eugenics practiced in Gattaca. Eugene needed genetic modification to be “perfect,” but there is not much public appetite to go down that path, even germline gene therapy at present. The Chinese scientist who used gene editing before implanting human embryos received a stiff prison sentence.

While selection doesn’t touch the genome and is limited by the number of embryos available, the door to PGT-P has swung open for business. Its apologists point out there’s no better investment than in children; we already give our kids an edge if we can afford private education and provide social advantages. The script has the emotional heft of promoting health, albeit without guarantees, but what has begun is surely a prologue to applications and new controversies driven by genetic discoveries.

Regulations and oversight are slow to keep up with the pace of technology. Robust guidelines, genetic counselors, and research registries are needed, but at this anxious moment in history, lawmakers are juggling other priorities and distracted by emergencies. The American Administration is drawing down regulations everywhere. While scientific progress is always welcome, we must be wary lest advances we cannot yet imagine take us closer to a Gattaca society without noticing.

What do you think?

Thanks for helpful suggestions to Jie Deng, M.D., Ph.D., H.C.L.D, F.A.C.O.G., a physician at Think Fertility in Bellevue, WA, and to Lucinda Veeck Gosden for advice.

Image: Molecules to Metrics — a composite image showing a DNA double helix transitioning into a heatmap, conveying the translation from genotype to statistical score. (Dall-E)