RESEARCH ALERT: When it comes to PGT-A, IVF practice in the U.S. is, unfortunately, not changing
By Norbert Gleicher, MD, Medical Director and Chief Scientist at The Center for Human Reproduction in New York City. He can be contacted though The Reproductive Times or directly at either ngleicher@thechr.com or ngleicher@rockefeller.edu.
After an electronic preprint in Fertility and Sterility, ASRM and SART finally published the paper version of their new Practice Committees’ Opinion on preimplantation genetic testing for aneuploidy (PGT-A) (1), and that is good as well as bad news: The good news is that this document for the first time clearly states that “the value of PGT-A as a routine screening test for all patients undergoing IVF has not been demonstrated.” In other words, after over 20 years of various forms of PGT-A and constantly growing utilization in U.S. IVF cycles, ASRM and SART finally came out with a document that clearly stated that such embryo testing in general populations really has no practical clinical purpose.
This statement clearly distinguishes this Committee Opinion from its 2018 predecessor document and, unquestionably, represents significant progress. ASRM and SART, indeed, deserve a round of applause because this one sentence alone defines routine utilization of PGT-A in IVF cycles – as currently practiced in a substantial number of IVF clinics – as no-longer an appropriate standard of care. Better late than never!
And now to the bad news: After with above-noted sentence in a few short words defining the purpose this document was produced for, it ultimately ended up lacking meaningful context. The reason is obvious: the document failed to follow up on its own main conclusion that PGT-A has not demonstrated any clinical value. And if something does not work, why would one continue using it? The question, therefore, arises: why was this very obvious follow-up conclusion omitted?
What makes this omission even more remarkable is that PGT-A utilization in U.S. IVF clinics is continuing to increase. Starting years ago when the procedure still was called preimplantation genetic screening (PGS) and was performed at cleavage stage through biopsy of one or two blastomeres of 6-to-8-cell embryos, its utilization has steadily increased, even though not only was there no evidence for any clinical utility for PGT-A, but increasing evidence was produced that it, indeed, did not offer any utility and in some patient sub-populations may actually harm pregnancy and live birth chances.
That PGT-A as a screening test in IVF has in over 20 years failed to establish even one single clinical utility, is by this ASRM/SART document, therefore, just left hanging out there for IVF providers, individually, to decide on how to react. To a degree, this is not only unfair to care providers, their clinics, and of course patients, but is also ethically questionable because, like other professional societies, ASRM and SART are not only expected to describe what should be done, but also what should not be routinely done.
What one should not do in medicine, more often than not, appears obvious: Anything proven not to work and, therefore, not producing a benefit that warrants its use, should no-longer be offered to patients. This means that, based on this new ASRM/SART document performing PGT-A routinely in every IVF cycle no-longer appears to represent standard-of-care. This conclusion, however, immediately raises a second question: under which circumstances is PGT-A then still indicated?
And this is where the ASRM/SART document gets into further troubles because not only does it not provide an answer, but the document appears to struggle trying to find indications for PGT-A. An example of this struggle is the following “word-salad” already in the abstract: “The value of PGT-A to lower the risk of clinical miscarriage is also unclear, although these studies have important limitations.”
In simple English this sentence likely is meant to communicate that a reduction in miscarriages – by proponents of PGT-A now often considered a new primary indication for PGT-A after live birth rates have proven unchanged – is also no longer defensible. So, what is then left?
That PGT-A shortens time to pregnancy may be true, though at best only to very minor degrees and only in good prognosis patients who produce large embryo numbers. The smaller the embryo number, the smaller this benefit, of course, will be. And assuming such a marginal benefit, is it even really worth the effort and additional costs?
In short, the ASRM/SART document1 represents important progress in defining the utility of PGT-A in association with IVF. The two Committees, interestingly, also responded in writing to a critical Letter-to-the-Editor which, among other issues, raised how troubling it was that, despite increasing reports about the ineffectiveness of PGT-A in IVF, utilization of the procedure has been increasing (2). In their response, the two Committees agreed that “it was troubling that the use of PGT-A in the U.S. is increasing, despite the lack of evidence to support this practice (3).”
What is still missing from the ASRM and SART can be summarized in three paragraphs:
• PGT-A should no-longer be offered as a routine add-on to IVF cycles.
• PGT-A should only be offered for specific clinical circumstances where the benefits outweigh costs
and risks. Examples are parental inheritable chromosomal abnormalities (i.e., balanced
translocations), repeated miscarriages with abnormal karyotypes, medical need or social desire
for sex selection, anxiety, etc.
• Based on the right to self-determination, patients should be allowed to request PGT-A in
association with IVF, though they must receive informed (written) consent which clearly states
that PGT-A so-far has not demonstrated any outcome benefits in IVF and, therefore, except in
situations as outlined above, still must be considered an experimental procedure.
Related, a recent paper by investigators from several U.S. institutions analyzed 28,562 donor oocyte cycles reported to the SART registry and found that among patients who had at least one blastocyst available for transfer (i.e., in relatively good prognosis patients), utilization of PGT-A was actually associated with a decreased live birth rate after first transfer in fresh oocyte donor cycles (58.4% vs 66.6%: P<0.001), thereby reaffirming prior reports suggesting that in younger patients PGT-A in 1st cycles may, indeed, adversely affect IVF cycle outcomes (4).
This difference disappeared, however, in frozen-thawed oocyte donor cycles with PGT-A, where PGT-A cycles nominally (though after adjustments not significantly) produced higher live birth rates (48.3% vs. 40.5%, N.S.). That PGT-A is frequently used even in donor-recipient cycle, which already produce the highest pregnancy and live birth rates among all cycles is only further evidence for the inappropriate utilization of PGT-A in the U.S.
Also related, the currently very actively promoted hypothesis of non-invasive PGT-A (niPGT-A) is that the reason why PGT-A so far has failed to improve IVF outcomes is the increased manipulation of embryos a biopsy requires, adversely affecting an embryo’s ability to implant and to lead to pregnancy. First suggested by Richard F. Paulson, MD (5), this hypothesis and the hope for a simpler and less-invasive PGT-A procedure have led to an onslaught of reports using different methodologies in determining DNA in spent media from blastocyst-stage embryos. niPGT-A, therefore, basically, attempts to follow the same methodology successfully implemented in obstetrical practice with introduction of non-invasive prenatal testing (niPNT) of maternal blood. What in this context, however, must also be noted is that niPNT – like current standard PGT-A – demonstrates a significant false-positive rate.
A recent review article in Human Reproduction summarized the experience with niPGT-A to this point in time, concluding that “like traditional PGT-A, niPGT-A must attain the high standards of precision and reliability expected of any genetic testing platform used in clinical settings before it can be adopted into routine protocols of IVF (6).”
We, of course, fully agree with this sentiment, and also agree that the quick implementation of niPGT-A by several IVF clinics in the U.S. and overseas appears premature. But that, of course, should not surprise, considering that PGT-A has been sold to the public for over 20 years and just now was finally “discovered” by ASRM/SART not to improve IVF outcomes (1).
Unsurprisingly, considering this article’s three authors have been strong proponents of PGT-A, it avoided the most important argument against niPGT, namely that niPGT-A is still in pursuit of the same by now fully debunked PGT-A hypothesis which suggested that the ploidy of a blastocyst-stage embryo can be reliably diagnosed from a small DNA sample. In other words, PGT-A has failed for over 20 years not because of poor techniques and technologies (those have greatly improved and yet PGT-A still fails in its efforts) but because the biology of early-stage embryos has been widely misunderstood. This biology, however, remains the same, whether fetal DNA is obtained from a 5-6-cell trophectoderm biopsy or from spent media.
And then there is one more issue to be discussed regarding niPGT-A, recently brought to our attention in two papers. Spanish colleagues in collaboration with Boston-based colleagues collaborated in a study – this time in Fertility and Sterility (7) – which extended routine embryo culture beyond 5 to 6 days in the hope of obtaining more and better spent media for their niPGT-A assay. Somewhat surprisingly, they, seemingly, were unconcerned about several recent reports on increasing epigenetic changes observed in embryos with prolonged cultures (a subject we will return to below).
The authors then reported allegedly high concordance between niPGT-A and PGT-A by embryo biopsy, as the authors “cautiously” concluded – and we are quoting – “possibly limiting the need for invasive PGT-A and making chromosomal testing of embryos available for an (even) wider range of patients.”
Especially considering above-noted recent ASRM/SART Practice Committee Opinion (1), how can a comment like that still make any sense? Indeed, we believe it doesn’t, and therefore are bringing this article here to hopefully wider attention.
But this is, of course, not all: The same group (in different order of authors) published another paper on the same subject, this time in Human Reproduction (8). In this fraction of what obviously was one study (in contrast to our basic science colleagues, we clinicians, unfortunately, often break up our studies into multiple publications), the authors looked for potential damage embryos may have encountered as a result of longer blastocyst culture. And, considering the superficiality of this search, unsurprisingly, there was none. This is how this finding was interpreted: “The implementation of an embryo culture protocol to accommodate iPGT-A had no impact on blastocyst viability or pregnancy outcomes.” And how about the epigenetics of the embryos? Not tested!
To summarize, in these two studies reported patients were – we presume by their physicians –determined to need a test called PGT-A. ASRM/SART, finally, after over 20 years of clinical (ab)use just declared this test useless (and in some cases, indeed, harmful to IVF cycle outcomes) (1). The investigators in these two (rather poorly controlled, but that is a separate issue) studies, moreover, chose to offer patients not the established PGT-A test, but an experimental niPGT-A test, in various formats in several prior studies uniformly found to be inferior in accuracy to the traditionally already inaccurate PGT-A test requiring an embryo biopsy. The investigators in addition exposed their embryos to extended embryo culture, without (to our knowledge) any preliminary outcome studies using animal models and/or to research-donated embryos, and without knowing the effects from these changes on embryo quality, implantation and pregnancy rates, and – as already noted above – potential life-long epigenetic changes. One is left wondering how this study made it through an IRB and how informed consents were worded?
It is likely that, as of 2024, a substantial majority of IVF cycles in the U.S. already include PGT-A at an average added cost to usual IVF cycle costs of at least $5,000. These added costs, moreover, come out of the patients’ pockets, even if they have full insurance coverage for IVF, because insurance companies (in this case smartly) do not cover PGT-A. Also smartly, IVF clinics and PGT-A laboratories roughly split the fees for PGT-A (that is unless clinics have their own in-house PGT-A laboratory).
IVF clinics, therefore, have become dependent on PGT-A fees which, in contrast to increasing insurance coverage for IVF cycles (generally greatly discounted by IVF clinics), is usually received in cash and undiscounted. One therefore would have to be naive to assume that the recent ASRM/SART document on PGT-A, declaring this test basically useless for most patients, will make even a small dent in PGT-A utilization. With rapidly growing introduction of niPGT-A, we, indeed, would not be surprised to see overall utilization of PGT-A even further expanding and producing even more inferior results compared to traditional PGT-A. But why should anybody care if PGT-A does not work in the first place!
References
1. Practice Committees of the ASR and SART. Fertil Steril 2024;122(3):421-434
2. Orvieto R. Fertil Steril 2024;122(3):555
3. Practice Committees of ASRM and SART. Fertil Steril 2024;122(3):556
4. Gingold et al., Fertil Steril 2024; DOI: 10.1016/j.fertnstert.2024.08.315. Ahead of print.
5. Paulson RF. Fertile Steril 2017;108(2):228-230
6. Volovsky et al., Hum Reprod 2024;39(9):1899-1908
7. Ardestani et al., Fertil Steril 2024; 122(3):465-473
8. Sakkas et al., Hum Reprod 2024;39(9):1952-1959