Interesting News From The Field of IVF
The IVF field’s lead journal in the U.S., Fertility and Sterility, is finally publishing some interesting papers again
Hard to believe that this could be a section heading, but it has been some time since we could say good things about an article F&S published. One reason, of course, has been that “expert opinions”—the least valuable publications in the evidence pyramid—and (mostly Chinese) meta-analyses, which, combined, in many journal issues consumed more print pages than data-driven original articles. Moreover, the writers of most “opinion” pieces were the reviewers of papers F&S had accepted for publication, thus establishing an often highly biased and self-reaffirming process akin to a hall of mirrors in how papers and commentaries made it into the journal.
And, as our readers by now probably have noticed, we are also not big fans of meta-analyses. Their popularity—and not only in F&S—has greatly increased because they require very little work. All one needs to do is sit at the desk and work on the computer, and even that has become so much easier with A.I. They, therefore, are often the often-favored products of “paper mills.” But even if not products of “paper mills” and A.I. abuses, they are often extremely biased because their conclusions, of course, largely depend on the selection of published studies for inclusion in the meta-analysis.
Upon careful examination, in many meta-analyses, one can, indeed, find significant selection biases, often supporting the authors’ predetermined opinions on the subject. In other words, consciously or subconsciously, we all like to confirm our preexisting biases, and meta-analyses allow for that with much more ease than a well-designed original study.
When we discussed F&S papers in recent years in these pages, it, therefore, was mostly to criticize them.
We, therefore, are delighted to have here the opportunity to bring to our readers’ attention several papers we found of interest in the good and bad. So, let’s start with one we liked and its accompanying commentary.
Some general principles regarding preimplantation genetic testing for aneuploidy (PGT-A) in IVF
That F&S would publish some objective PGT-A-related paper was probably our most pleasant surprise, considering the fact that papers critical of PGT-A in the past rarely made it through peer review. British colleagues have now published a paper in F&S that suggested that current PGT-A testing misclassifies roughly a third of embryos signed out as “mosaic” (both low and high level) because they, in reality, likely represent meiotic aneuploidies (i.e., aneuploidy in all cells). The authors, therefore, concluded that single-nucleotide polymorphism genotyping should be combined with standard next-generation sequencing (NGS) in PGT-A in order to be able to determine accurately whether a chromosomal abnormality is meiotic or mitotic in nature.1
In an accompanying Reflections commentary, Mili Thakur, MD, from Genome Ally in Grand Rapids, Michigan, and our September GrandRounds speaker here at the CHR (you can view her talk here), presented a short but very insightful summary, correctly concluding that “not all mosaic embryos are the same.”2
The introduction of a “mosaicism” diagnosis to PGT-A in 2016 by the PGT-A provider community only further confused the IVF field because it arose not out of a need, but because the field, under the guidance of the highly conflicted PGDIS (Preimplantation Genetic Diagnosis International Society) found itself exposed after, first the CHR,3 and shortly after that, Italian colleagues in October and November, respectively, reported normal pregnancies after transfers of what then were called “aneuploid” embryos (our Italian colleagues in their paper, indeed, described their transferred embryos already as “mosaic.”4
These reports were, of course, irrefutable evidence that at least some of the embryos which PGT-A (then still called preimplantation genetic diagnosis, PGD) laboratories signed out as “aneuploid” could still produce perfectly normal pregnancies. The PGD/PGT-A industry, in other words, faced a major credibility crisis and had to do something. And what the industry did was, what it had already done once before in a similar crisis, a few years earlier, instead of in general rethinking the concept of the procedure, the industry announced a “technically better” procedure that, finally, would fulfill all the promises for the procedure for IVF that the industry had claimed.
And this technically better procedure also received at this point the new name PGT-A and, rather than reporting embryos with only a binary result of euploid (chromosomal normal) and aneuploid (chromosomal abnormal), embryos now could also be “mosaic.”
And here comes the likely biggest joke in this whole sequence of charades, the definition of how the genetic testing industry—under the totally incompetent guidance of the PGDIS5—defined “mosaicism:” Though “mosaicism” has had a universally accepted definition for ages, the presence of more than one cell lineage in a complete organism derived from a single cell,” the PGD/PGT-A field decided to invent its own definition, “the presence of more than one cell lineage in an on average of 5-7 cell biopsy of the trophectoderm of a blastocyst-stage embryo.”
The difference between these two definitions is, of course, of crucial importance because it makes one hell of a difference if one finds DNA from more than one cell lineage in the DNA of 5-7 cells or in the DNA of hundreds of cells of a complete blastocyst stage embryo. Moreover, the 5-7 cells biopsied off a blastocyst-stage embryo are taken from the outer layer of the embryo, the so-called trophectoderm, which, after implantation, becomes the placenta. The fetus, however, arises from a very different cell lineage of the blastocyst, the so-called inner-cell-mass, which is inaccessible to biopsy and, probably, would lead to birth defects if biopsied. And the placenta is now well recognized to maintain an island of chromosomal abnormal cells until birth.6 At the same time, the fetus, of course, cannot afford to remain host to chromosomal abnormal cells and, therefore, often self-corrects by elimination aneuploid cells from its own confinement.
In other words, what PGT-A laboratories sign out as “mosaic” embryos, with great likelihood are, indeed, mosaic embryos, but to consider percentages of mosaicism or low vs. high mosaicism, as many IVF clinics now use as a criterion as to which “mosaic: embryos to transfer (low mosaic) or not transfer (high mosaic), is pure nonsense and, biologically makes absolutely no sense because percentages of DNA in a 5-6 cell biopsy, of course, can never accurately reflect the complete embryo of several hundred cells.
Similarly, 5-7 trophectoderm cells, even if all euploid, do not preclude “aneuploid” cell islands elsewhere in the embryo, which then would mean that the whole embryo, of course, is “mosaic.” Even 100% aneuploid DNA in a 5-6 cell biopsy, moreover, does not mean that the whole embryo is not mostly euploid and was just at random biopsied in the midst of an aneuploid cell island. Again, such an embryo would, under the traditional definition of “mosaicism,” be exactly that.
The publication of truly absurd new guidelines for PGT-A in 2016 by the PGDIS,5 which the field and even ASRM/SART, paradoxically, took seriously even though they lacked even minimal evidence, therefore, just increased the confusion surrounding PGT-results to such a degree, that many IVF clinics came to insist to again receive only binary “euploid” and “aneuploid” reports from PGT-A laboratories. And, suffice to say, the 2016 promises of a technological superior PGT-A to the earlier PGD were, of course, not fulfilled, and—jumping a little ahead in PGT-A history—in September of last year (2024), even ASRM and SART finally concluded in a Guideline document that after over 20 years of clinical use, the test previously called PGD and now called PGT-A to this day has basically failed to improve any IVF cycle outcome.8
Yet to this day, ASRM and SART (and ESHRE) have not demonstrated enough guts to equally state that PGT-A for many infertility patients actually reduces pregnancy chances. And one, of course, does not even have to be a statistical genius to discover that, because everybody by now knows that PGT-A leads to non-use or even disposal of large numbers of embryos with perfectly fine normal pregnancy and live birth potential means that every time a laboratory’s PGT-A falsely leads to non-use or erroneous disposal, that patient has been harmed by PGT-A. Why our professional organizations, therefore, don’t come out and very clearly state that PGT-A should no longer be used in IVF is unclear to the CHR!
What does it mean when an embryo in PGT-A is reported as “chaotic?”
Colleagues from Mexico, in another paper in F&S, asked what the meaning was of an embryo after PGT-A reported as “chaotic.”9 And the reason for this question is once again that, based on how the PGT-A industry has been handling this issue, it is just as bizarre, or maybe even more so than even the handling of “mosaicism.”
This issue hit the headlines for the first time when Igenomix USA, in early 2023, sent a mailer to U.S. IVF clinics announcing that preliminary data of a study they had conducted in which they rebiopsied embryos that on first biopsy were found to be “chaotic.” In their laboratory, this meant they demonstrated 7 or more abnormal chromosomes, though the definition of “chaotic” varies between laboratories. On re-biopsy, ca. 40% of embryos were—surprising for them—found to be “euploid.”
Their explanation? None, or maybe a laboratory error?
The Mexican study, now published in F&S, reported on 93 “chaotic” embryos (here defined as 5 or more chromosomal abnormalities, including segmental abnormalities) that were investigated. Among those 69 (74.2%) survived thawing and were rebiopsied, 17 (24,6%) were reclassified as “euploid,” 26 (37.7%) were confirmed as “chaotic,”, 18 (26.1%) as “aneuploid” but “non-chaotic,” 5 (7.2%) as “mosaic” (under incorrect PGT-A definition), and 3 as “non-informative.” The authors then subdivided their chaotic embryos further, depending on different kinds of “chaotic” chromosomal findings, with, again, widely varying results, but the subgroups were so small that one cannot take their results really seriously.
Amazingly, the authors concluded from this study—and we are quoting: “Rebiopsy can potentially identify embryos with favorable reproductive potential, particularly those initially classified as “mosaic chaotic.”
This conclusion is, of course, ridiculous because how would the authors know that their second biopsy is more “accurate” than their first one was? Peer review, where are you?
Overall, this study, of course, confirms only one thing: PGT-A is a scam that not only does not offer any outcome advantages but significantly harms IVF outcomes. Not to be forgotten, of course, rebiopsy harms embryos.10
Does PGT-A shorten the time to live birth?
This is what our colleagues from Boston IVF claimed in a recent F&S paper,11 even though ASRM/SART in their September Practice Committee Opinion quite clearly stated that this was not the case.8 But, reflecting the most recent relentless efforts by proponents of PGT-A to find a clinical utility for PGT-A after all, the authors hooked up to the recently in several papers propagated notion that, while PGT-A does not improve IVF outcomes in young patients (who usually have good embryo numbers and, therefore, at least theoretically, might benefit from a good embryo selection method), the test does allegedly improve outcomes in older women (who usually are happy having enough embryos for transfer.
Not only does this notion, therefore, not make any sense whatsoever, but proponents of this nonsense forget that three of the first four papers in the world literature questioning the utility of what then was called PGD, of course led by Mastenbroek et al.’s since 2007 famous prospectively randomized study, demonstrating actually reduced clinical pregnancy rates in especially older patients (which, of course, makes perfect sense!).12-14
The sad thing is that our Boston colleagues – like all other authors who recently have made this argument - should know better because it really does not take a genius to understand that embryo selection, assuming it does have a function in IVF at all, which can be questioned (but that is a subject for another day), will have such a function only if there are enough embryos to select from, and that, of course, is the case more often in younger than older patients.
So, where is the error in the paper by Eliner et al. and in the papers of all the other authors claiming PGT-A benefits in older rather than younger patients using PGT-A? Once again, it does not take a genius to guess what the answer must be: patient selection bias!
In younger patients, in many clinics (and studies), almost all patients’ IVF cycles include PGT-A. In older patients, however, even the most fervent proponents of PGT-A will often abstain, unless these older patients have unusually good functional ovarian reserve. In other words, older patients are highly selected, and younger patients are not. That older patients demonstrate outcome advantages in certain studies, therefore, has nothing to do with PGT-A and everything with who gets PGT-A in their IVF cycles.
And if you don’t believe us, there will be an opportunity to see the evidence in a poster at ASRM 2025 later in October in San Antonio,15 and a full-length paper has been submitted. The CHR’s David Barad, MD, simply couldn’t stand it any longer reading all of these papers claiming outcome benefits for older women from PGT-A and decided to formally study the issue based on national U.S. IVF registry data, and it wasn’t at all difficult to prove where the error was.
Why has nobody before made this point?
Poor study design because every IVF practitioner, of course, understands the importance of patient age in IVF outcome; yet none of the studies claiming outcome benefits in older women ever adjusted results for patient age. It’s really not that difficult!
References
1. Popa et al., Fertil Steril 2025;124(2):307-318
2. Thakur M. Fertil Steril 2025;124(2):246-247
3. Gleicher et al., Fertil Steril 2015;104(3):e59
4. Greco et al., N Engl J Med. 2015; 373(21):2089-2090
5. Gleicher et al., Reprod Biol Endocrinol 2020;18(1):57
6. Coorens, et al., Nature 2021;592(7852):80-85
7. Yang et al., Nat cell Biol 2021;23(4):314-321
8. Practice Committees of ASRM & SART. Feril Steril 2024;122(3):421-434
9. Calull et al., Fertil Steril 2025;124(2):319-326
10. Nohales et al., J Assist Reprod Genet. 2023;40(8):1905-1913
11. Eliner et al., Fertil Steril 2025;124920;281-288
12. Platteau et al., Fertil Steril 2005;84(2):319-324
13. Mastenbroek et al., 2007;357:9-17
14. Gleicher et al., Fertil Steril 89(4):780-788
15. Barad et al., Fertil Steril. ASRM Congress Suppl 2025. In press; Abstract