The Future of Fertility Care Begins Before the Embryo

At the annual meeting of the European Society of Human Reproduction and Embryology in London this July, two currents ran through the sessions at the same time.

One was the field’s steady push toward smarter, gentler ways to read an embryo: noninvasive ploidy prediction from spent culture medium, image-based artificial intelligence (AI) tools that estimate chromosomal status, and algorithms trained to forecast a blastocyst’s fate from the oocyte itself. The other current was quieter, and in some ways, unsettling: growing evidence demonstrating how current methods of embryo screening do not fully dictate the outcome of an embryo transfer and certainly not of an entire pregnancy or long-term generational health.

Consider three findings all presented at this conference. In one, a large Australian cohort of more than 18,000 patients, authors reported a 68 percent cumulative live birth rate over three cycles, achieved largely with single-embryo transfer. However, doctors screened for abnormal chromosome numbers using preimplantation genetic testing for aneuploidy (PGT-A) in only about a quarter of the patients, prompting the authors to question whether routine genetic testing benefits everyone, a debate that has remained ongoing since the advent of the technology.¹

A separate Australian study found that more technically difficult biopsies were associated with less favorable genetic outcomes, a reminder that the test’s own mechanics can shape its answers.² Finally, a third Danish study linked low-grade inflammation to the loss of pregnancies with embryos carrying the correct chromosome numbers, indicating that chromosomally normal embryos were lost not because of their genome, but because of the environment that received them.³

That last finding is the whole argument in miniature. It is also one of the most common conversations in the clinic. A patient sits across from her physician after months of injections, a retrieval, and a biopsy report that reads, reassuringly, euploid: containing the complete chromosome set. The embryo is chromosomally normal, but once it is transferred, it does not implant, or it implants and is lost a few weeks later, which leaves the patient asking: The test said the embryo was normal, so what happened?

Situations like this highlight an important gap that PGT-A cannot fill. In patients with good prognoses, transferring a single euploid embryo leads to a live birth roughly 55–75 percent of the time, depending on the laboratory’s quality, the embryo’s developmental day, the embryo’s grade, and the presence of other relevant variables influencing implantation.^4,5 In other words, even a genetically normal embryo chosen by the best available test fails to produce a baby one in three times.

That residual gap of 30 percent or more is where the rest of reproductive biology comes into play. It is almost entirely invisible to any test performed on the embryo after fertilization because the information simply is not there to be found. Preconception medicine, which focuses on optimizing reproductive and general health for men and women prior to pregnancy, is one lever that can help improve egg, sperm, and embryo quality, as well as the ability to have a successful implantation and pregnancy. Proactive lifestyle-based approaches optimizing diet, stress management, and more can supplement laboratory techniques that help improve the success of IVF. But these broad recommendations are only a starting point; the next frontier is understanding which specific biological levers matter for each individual and intervening while they are still modifiable.

What Chromosomal Screening Can, and Can’t, Reveal

None of this means PGT-A is not worth doing. In the more precise genetics laboratories, it identifies embryos with the wrong number of chromosomes exceedingly well. Aneuploidy is the leading cause of implantation failure and early miscarriage, particularly in older patient populations, so prioritizing a euploid embryo can shorten the path to pregnancy and spare patients the grief of transferring an unviable embryo. In the United States, with a growing population of women 40+ years old utilizing IVF and political restrictions on managing non-viable but ongoing pregnancies, PGT-A can provide critical insight and the ability to avoid transferring abnormal embryos.

But the limits of PGT-A are real, and researchers continue to acknowledge them. The tissue tested in PGT-A, the trophectoderm, is a biopsy sample of a handful of cells from the outer layer of the embryo at a single moment in time. Mosaicism, where the results suggest a mixture of cells with both correct and incorrect chromosome numbers, is a gray area, since many embryos labeled “mosaic” can go on to implant and produce healthy children, while some “mosaic” results may reflect the biopsy and amplification process as much as the embryo itself, a point the biopsy-difficulty finding underscores.^6,7

In short, trophectoderm biopsy is invasive by design, and the resulting PGT-A report says nothing about an embryo’s developmental ability, metabolic health, capacity to implant, or ability to develop into a viable pregnancy. An embryo can be designated “correct,” yet the embryo transfer can still fail.

The clinical value of routine PGT-A has been debated in the peer-reviewed literature, and professional bodies have cautioned that newer genetic tools, polygenic testing among them, are being offered ahead of the evidence.^8,9 Even for physicians or clinics that are high utilizers of PGT-A, the conundrum over how to manage recurrent implantation failure remains.

AI is a Helpful Tool, But It Doesn’t Resolve Everything

The AI approaches showcased in London are genuinely promising, and they deserve serious attention. Noninvasive ploidy prediction and image-based models could make embryo assessment cheaper, gentler, and more scalable.

But caution is warranted about what AI approaches change and what they do not. Rigorous reviews have found that image-based AI outperforms morphology alone but remains inferior to invasive PGT-A for detecting aneuploidy, and they have also found that noninvasive PGT-A still contends with maternal DNA contamination that limits its agreement with the biopsy standard.^10,11 These are solvable engineering problems, and the progress is real.

Still, the smartest algorithm and the biopsy share one aspect: A better lens on the embryo is still only a lens on the embryo. AI does not add information that does not exist. If the factors determining a successful pregnancy were fully encoded in the embryo being imaged, better imaging would eventually solve the issue.

Notably, the most forward-looking AI work points back toward predicting outcomes from the oocyte. It’s an implicit admission that what matters most is decided before the embryo exists.

Screening the Embryo Can’t Explain the Conditions That Made It

An embryo is the output of a process that was largely complete before fertilization. Its quality is affected by the oocyte that carried it: cytoplasmic maturity, mitochondrial competence, and the metabolic and hormonal milieu of the months in which that egg developed. It is shaped by the sperm, whose DNA integrity and epigenetic state influence not just fertilization, but the trajectory of the embryo that follows.

Whether any embryo turns into a pregnancy depends on the environment it is transferred into, such as endometrial receptivity, the immune and inflammatory landscape, or the metabolic health of the person carrying it. The euploid losses tied to inflammation tell exactly this story.²

Crucially, none of these inputs are fixed. Eggs and sperm mature over months, and across those windows their quality is measurably shaped by modifiable factors such as nutrition, metabolic and inflammatory status, body composition, and oxidative stress. Sperm DNA fragmentation is associated with impaired reproductive outcomes and is higher in the setting of several modifiable health and lifestyle factors, including obesity-related metabolic dysfunction and smoking.¹² Factors ranging from the mother’s existing micronutrient status to metabolic health during the preconception period dictate fertilized eggs and early embryo development.¹³

A chromosome count cannot see nor move these levers, and none of this information is yet recoverable after the fact from a spent-medium sample or time-lapse imaging. The field has spent two decades building ever-finer instruments to grade embryos, but it has spent comparatively little effort on the biology that produces good embryos, sperm, and receptive uteruses in the first place. We have been, essentially, inspecting the harvest while ignoring the soil.

Closing the Loop

This is the gap that precision preconception medicine is designed to fill. A new category is emerging that proactively optimizes egg, sperm, and metabolic health before conception using genetics, biomarkers, and precision nutrition. Unlike traditional genetic screening, which primarily identifies inherited disease risk, this approach analyzes functional genetic variants that influence nutrient metabolism and reproductive biology, translating those insights into personalized, actionable interventions. Rather than grading embryos after the fact, precision preconception medicine characterizes the upstream biology. It studies the genetic, metabolic, hormonal, and inflammatory signals that shape egg and sperm quality as well as the receptivity of the uterine environment in the months before a cycle begins. Much of that biology can be modified during that window, and it is increasingly measurable.

The catch is that the right intervention is rarely the same for two people, which is why the blanket “eat well and exercise” guidance underdelivers so often. People differ in how they absorb and use micronutrients, how much oxidative-stress and inflammatory burden they carry, how they metabolize the inputs that shape gamete quality, and which specific lifestyle modifications are most likely sustainable for their individual life circumstances.

A better approach would be to read an individual’s genetic and biomarker signals to identify which specific nutritional, metabolic, and lifestyle levers may actually move egg and sperm quality for that person and their partner in the months before a cycle. This could be paired with a user-friendly interface that nudges lifestyle change in a way that aligns with that specific person or couple. This way, precision is applied upstream while the biology is still modifiable.

Framed this way, upstream optimization and advanced embryo selection become two halves of one loop. Proactive preconception care helps feed healthier gametes and a more receptive environment into the clinic, while advanced embryo selection helps clinicians make the best choice from a better starting population.

When someone asks why a normal embryo did not become a baby, the honest answer is sometimes that the cause is unknown. But that should be a smaller share of the truth every year. The science presented in London points the way, but getting there will require looking at everything that happened before the petri dish came under the microscope.