Improving embryo biotech could radically reduce disease for future generations
Few societal conversations elicit such a unique outpouring of curiosity, grief, vitriol and laconic passivity than the global fertility rate, which has halved since World War II. Roughly half of countries today are below population replacement, leading researchers to tirelessly search for underlying causes. Our former Riskgaming podcast guest Leslie Schrock emphasized how to improve fertility for men and women, while John Burn-Murdoch last week argued that the decline is from fewer couples rather than couples having less children.
Conceiving is one of the most important decisions a couple can make, which is one reason the discussion around fertility has been so passionate. It’s a decision that will become even harder in the next few years though due to emerging biotechnologies that will dramatically improve the lives of future generations — but only for parents who choose to wait to conceive.
Before we jump to the future, let’s take stock of the present. Outside of pleasure and period trackers, technology hasn’t radically transformed traditional sex (Sex 1.0?), although books like Schrock’s Fertility Rules offer frameworks to improve fertility success through healthier lifestyle choices. Where fertility technology has rapidly progressed though is for couples who struggle to naturally conceive and need alternatives. In-vitro fertilization is among the most well-known, and roughly 2% of American births today come via IVF.
Cycles for IVF are very expensive (about $15,000-$20,000) and the rate for successful pregnancies can range from 32% for women under 35 to just 4% for women over 44 years old according to Britain’s health service. An immense amount of work is being done to optimize these numbers, including by Lux’s own Alife Health, which uses extensive datasets and specialized AI models to optimize fertility clinics.
As part of each IVF cycle, multiple embryos can become available for implantation, generally in the single digits. Today’s technologies are designed to select the embryo most likely to implant and reach term. For instance, many IVF clinics offer (for an additional fee) screening options like preimplantation genetic testing for aneuploidy, which in theory can select for healthier embryos. However, there are concerns that such testing has too high of a false positive rate and there are even class-action lawsuits that have been filed against some service providers. Even so, the combination of more optimization in the laboratory coupled with better screening options has dramatically raised IVF success rates since the technology’s infancy.
Improved success is allowing a new question to be asked: rather than just selecting one of, say, five embryos as the most likely to implant, could parents select an embryo based on the future health prognosis of their child? The cutting-edge option today is known as embryo screening using polygenic scores (or polygenic embryo screening). These products score an embryo for both complex diseases like high blood pressure or Alzheimer’s as well as human characteristics like height. Startups like Orchid (founded by the wife of a close friend) now offer whole-genome reports to couples, who can then select the embryo that has the best score.
This is an advance, but its utility is limited by the number of embryos available during a cycle of IVF. Since most couples will only have a handful, serious tradeoffs between embryos is inevitable. Perhaps one is more likely to develop epilepsy but also to have lower blood pressure. This fundamental constraint is why claims that parents are designing their children are ridiculously overblown.
Turning toward the future though, intentionally designing babies for better health isn’t too far off. Deena Shakir passed along this analysis published in Nature last week by a group of bioscientists exploring how a combination of CRISPR-Cas9 (aka gene editing) and the massive growth of genome-wide association data will combine to offer couples the option for heritable polygenic editing — or gene editing of embryos for more complex expressions.
What they find is striking. Through just a small number of genetic modifications, doctors might have the ability to mostly strike out entire diseases. That would not only benefit these children, but also future generations, since these edits would also be inheritable by grandchildren as well.
Exploring major conditions like Alzheimer’s, major depressive disorder and type 2 diabetes, the scientists simulated that the prevalence of these conditions could be reduced by orders of magnitude, although results varied. For instance, the prevalence of type 2 diabetes could be reduced from 10% to merely 0.2% if just 10 alleles in the genetic code were changed. On the other hand, major depressive disorder showed much less improvement (15% to 9% with 10 changes), since bioscientists are currently unaware of specific edits that would vastly improve the disease.
The authors describe the timeline for heritable polygenic editing as “feasible in the next three decades,” a prediction heavily dependent on how quickly we can build tools for very precise and large-scale gene editing within embryos (an area of immense interest and funding right now). If we accept their timeline though, children born today will have these options available to them if and when they are ready to conceive.
While the researchers show utmost concern to avoid over-hyping this technology (perhaps one reason why it took 18 months to get through Nature’s editorial process), there are also reasons to believe that this timeline could be much faster.
For instance, one of the largest problems for realizing heritable polygenic editing is building much more robust genome-wide association datasets. This requires not just more data, but also more diverse data. On the former, this is increasingly possible given the rapid decrease in cost for whole-genome sequencing, which has crashed from more than $100 million in 2001 to less than $1,000 today. That’s faster than even Moore’s Law would predict, although recent evidence is that further cost reductions are stalling a bit.
Then there’s the question of diversity. The hardest part of building these comprehensive datasets is that bioscientists need to find as many different combinations of genes as possible in order to predict what any edits might do. That’s hard: since genes are inherited and therefore correlated with parents, that also means that genes are relatively correlated among people from similar backgrounds. Instead, we need as much data from uncorrelated people as possible.
Lux has a solution here with Variant Bio, which had a front-page story in The New York Times last weekend in the context of its new $50 million development deal with Ozempic-parent Novo Nordisk. Variant, which was co-founded by our own Josh Wolfe with Stephane Castel and Kaja Wasik, is designed from the ground up to purposely seek out these diverse populations, which are often isolated from the rest of the world. From the article:
“What’s the value of sequencing the 500,001st British person?” Dr. Castel said. “There are only so many insights to find by studying the same group of people.”
He and Dr. Wasik were more enthusiastic about recent findings from underrepresented groups, such as the discovery of novel gene variants affecting metabolism that were first identified in Inuit populations in Greenland.
Variant’s main goal is developing novel therapeutics by characterizing unique genetic variants (hence the Novo partnership). In the future though, its database could also power technologies like heritable polygenic editing and directly improve the disease outcomes of future generations.
Beyond more depth and diversity in genetic datasets, the next major hurdle will be identifying the positive and negative effects of individual edits to an embryo’s DNA. Outside a narrow number of diseases that derive from the expression of a single gene, most genes affect the body in multiple complex ways, making it difficult to find “the perfect edit.” As the authors of the Nature paper wrote:
For genomes edited for a large increase or decrease in a trait or disease, the resulting zygote may be unviable. There are known examples of variants that are protective against one disease but are risk factors for other diseases. It is very difficult to prospectively predict the pleiotropic effect [a single gene influencing multiple expressions] of new combinations of variants on pre-natal development, which will be a significant source of uncertainty in the future use of [heritable polygenic editing].
That uncertainty is real, and its solution comes down to the future of artificial intelligence in biology. If AI accelerates rapidly and we have fairly accurate models of functional gene expression (essentially a computational testbed for genetic experiments), then it’s certainly reasonable to predict that viable heritable polygenic editing could be only a decade away rather than three.
The authors have an extensive ethics section that includes discussion around a relapse into eugenics and all of the terrors of that era. Given their predicted timeline though, they ignored a major ethical challenge for (younger) couples that I think is crucial to consider: conceive today, or wait and perhaps radically reduce the incidence of large classes of diseases for your future offspring? Given that this form of gene editing must be done at the embryonic stage, the decision is binary. Gene therapies might be developed as an alternative, although this is less likely given the complex genes that lead to these diseases in the first place.
There’s no easy answer. If you are over 35 and looking at the daunting success curve of IVF that is rapidly approaching, it makes sense to just move forward. If you are in your early 20s though, the question is much more difficult. At a minimum, we will likely have a much better sense of the timeline for these transformative technologies by 2030.
Technologies like heritable polygenic editing won’t solve any of the social factors that have caused global fertility declines, but they can ameliorate some of the fears of parents who are seeking the best lives for their children. Dramatically improving the healthiness of future generations could ultimately be one of the greatest gifts one generation has ever offered to another. Now to fix the planet.
Podcast: Which companies will suffer with globalization’s reversal?
The Washington Consensus of the past few decades that called for open markets, free trade and reduced regulation will officially die on Monday as President Donald Trump re-takes the presidency indoors with a radically different economic program. Free trade is out and tariffs are in; globalization is dead and national sovereignty is the rule of the day. Such a change has massive implications for companies all around the world, many of whom have designed their corporate strategies for a global world. Who is affected, particularly when it comes to U.S.-China trade in the years ahead?
That’s where Isaac Stone Fish comes in. He’s the CEO and founder of Strategy Risks, a data and research company that helps companies and regulators understand and reduce alternative forms of risk. He’s particularly noted for his China expertise, and his firm publishes the SR250 ranking, which highlights the largest American companies with the deepest ties with China, encompassing everything from financing and supply-chain interlinkages to public communications.
Fish joins me and our Riskgaming director of programming Laurence Pevsner to talk about Trump’s imminent arrival, why Ford is the most China-entwined company in the U.S., how China overtook the U.S. in electric vehicles, why American defense contractors are surprisingly engaged in China trade, why we might already be at war with China, how CEOs are managing these new strategic risks and finally, what the biotech and social media industries must do going forward in a more fractured world.
🔊 Listen to “Which companies will suffer with globalization’s reversal?”
Lux Recommends
- Our scientist-in-residence Sam Arbesman enjoyed Brice Ménard and Nikita Shtarkman’s interactive learning tool “biocubes,” which allows users to explore the entire biological world with properly-scaled cubes representing the total mass of life on Earth (example above).
- Laurence enjoyed Scott Alexander’s “Bureaucracy Isn't Measured In Bureaucrats” on Astral Codex Ten. “Why does it take six years to respond to a petition? My guess is because they knew they would get sued and so they have some sort of million-step process that addresses every single thing you can sue over, so that they can prove to the court that their process addresses all possible complaints and they followed it to the letter. If you cut their bureaucrats in half, that doesn’t mean there will be fewer steps in the process. It means they’ll keep wanting not to get sued, the process will stay the same, and everything will take twice as long.”
- Sam enjoyed Hannah Ritchie’s analysis of the famous bet between Julian Simon and Paul Ehrlich on how prices for five resources — chromium, copper, nickel, tin, and tungsten — would change a decade later. Her analysis shows either could have won depending on when the bet started. “The problem is that Simon and Ehrlich agreed to a bet about short-term changes. Their decade-long wager was vulnerable to short-term fluctuations, which often had nothing to do with actual changes in physical supply and demand. Instead, they were influenced by geopolitical or temporary economic forces. That means either could have won depending on random year-to-year fluctuations that didn’t reflect their worldview.”
- Discussions of geoengineering as a solution to climate change linger on the edges of intellectual thought, but it is increasingly heading toward the center of the discussion. So I thought Erin Sikorsky and Tom Ellison’s discussion in Lawfare on the security implications of geoengineering to be fascinating. “Profit-motivated tinkering with Earth’s atmosphere risks undermining legitimate research into geoengineering and threatens to feed geopolitical tensions and misinformation, a worldwide example of privatized gain and socialized risk.”
- Finally, Sam enjoyed Brian Potter’s latest essay in Construction Physics on “Why Skyscrapers Became Glass Boxes.” “Similarly, the fact that a glass and metal curtain wall was so light meant that less structural framing was required to support it. Beams and columns could be made thinner, foundation sizes reduced. Not only did this reduce the cost of the building itself, but it reduced the cost of constructing it: all else being equal, a lighter structure is easier and cheaper to erect.”
That’s it, folks. Have questions, comments, or ideas? This newsletter is sent from my email, so you can just click reply.
