Review and Commentary on Recent Literature in Reproductive Medicine
Here the staff of THE REPRODUCTIVE TIMES offers commentaries on recently published articles, primarily chosen for educational merit—in the positive but also in the negative—for clinical purposes, and for their potential translational values to clinical medicine when addressing basic science issues.
September 4, 2024. Revised with permission from the September CHR VOICE.
SUBJECT: Fertility Treatments Using Assisted Reproductive Technologies (ARTs)
Starting with the embryology laboratory...
IVF CULTURE MEDIA: Two leading voices in the international infertility arena, Richard J. Paulson, MD, and Eli Y. Adashi, MD, recently published an important commentary in Reproductive Medicine Online, which the editors somewhat confusingly placed into a category called “Countercurrent.” The article does not really represent a countercurrent opinion, when it complains that current culture media manufacturers do not adequately inform the public about the constituents of their culture media. 99.9% of embryologists and IVF providers would probably agree that this is unacceptable and must stop.
Nobody, of course, prescribes medications to patients without knowing what they contain. And like patients are from medications, embryos through culture media in embryology laboratories are—at epigenetically crucial times—potentially exposed to long-term effects that can last a lifetime. The two authors pointed out that a small number of analytical studies found considerable differences among culture media, but were unable to trace purposes and problems of many of the ingredients and to compare according to them IVF outcomes. The field, indeed, currently even lacks any adequately powered studies comparing between culture media sold. Imagine the uproar if a new cancer drug claiming benefits over older drugs would make claims of superiority without scientific backup from a properly executed comparative study; yet this scenario is fully accepted in embryology practice!
TIME-LAPSE IMAGING: An excellent study by mostly British investigators offered a final verdict on the clinical effectiveness and safety of closed incubation systems with time-lapse imaging and the news was not what most had anticipated: this very expensive addition to many—if not most—embryology laboratories in the world failed to significantly increase the odds of live births compared to standard embryology without time-lapse.
This, of course, does not mean that closed incubation systems with time-lapse imaging do not have a purpose in IVF—they of course especially for research do—but this publication once more lays bare the irresponsible introduction of “newness” into the IVF field we have been witnessing now for almost two decades based on unproven hypotheses.
In this case the false claims were that time-lapse would standardize and improve IVF outcomes and reduce manpower demands in embryology; however, none of these promises were fulfilled. One must wonder how many funds spent on new time-lapse equipment could have been used elsewhere and/or how much lower IVF cycle costs now would be.
A.I. FOR OOCYTE ASSESSMENTS: An obviously related subject to time-lapse imaging is the current A.I. frenzy surrounding the embryology laboratory. With the concept of “embryo selection” in IVF cycles recently under increasing criticism and considered just another unproven hypothesis that has overtaken IVF practice, follows the conclusion that embryo selection with A.I. will, likely, lead to similar disappointments as just reported for time-lapse.
That A.I. may potentially, however, find different applications in the embryology lab, was recently suggested in several presentations at the annual ESHRE conference in Amsterdam which suggested that, instead of using A.I. to assess embryos, freshly retrieved oocytes may be a more valuable target for A.I. In one of the first oral presentations of the conference ([Abstract O-004], a Canadian start-up [Future Fertility, Toronto, Canada] presented a novel A.I. model to predict ploidy status of a developing blastocyst based on the mature MII oocyte the embryo came from. The model’s AUC (0.61) in predicting aneuploidy was clinically insufficient but can be expected to improve with more learning by the system. Moreover, since the researchers of this project appropriately recognized obvious selection biases in embryos used to build their model, they should be able to correct those.
An A.I. model from Mexico [Abstract O-003] attempted to determine whether it could enhance the prediction of mature MII oocytes prior to stripping from cumulus-oocyte complexes (COCs) and claimed that the program improved the judgment before denudation in Grade III and Grade IV COCs (representing more difficult to predict COCs before denudation) (P<0.001). The abstract unfortunately did not allow for a clear enough understanding of study protocol to judge its ultimate validity; but assuming results hold up, this may be a potentially clinically important future application of A.I. in embryology.
EXTRACELLULAR VESICLES: In another ESHRE abstract, Swiss colleagues demonstrated that extracellular vesicles, which currently are the target of research in every last corner of the human body, apparently also play a role within follicles in oocyte maturation. More specifically, it appears the vesicles from mature oocytes—after being taken up by immature oocytes—help in their maturation. [Abstract O-005]. If confirmed, this would be a potentially very significant observation because some investigators have for some time already suspected that the largest lead follicles affect the fates of their smaller counterparts. Indeed, it appears that once a lead follicle over-matures, it already adversely affects the quality of smaller follicles even before ovulation. That this effect may be mediated by extracellular vesicles would not surprise.
…and continuing with allegedly new treatments
AN UNNAMED DRUG TO IMPROVE IMPLANTATION? And to remain at the ESHRE event, Spanish and Polish colleagues reported on a new, still unnamed drug (OXO-001) which allegedly improves implantation of embryos through a direct effect on the endometrium (Abstract O-027). What the drug does, and its chemical description, so far remains secret. But based on a randomized, double blinded, placebo-controlled multi-center phase II clinical trial with three parallel arms conducted in 28 European centers between 9/2021 and 1/2023, this pilot study improved implantation rates by more than 10% (46.3% vs. 35.7%).
Considering that this study was performed in donor egg-recipient cycles of women under age 40, the pregnancy rate in controls (35.7%) appears low, while the reported pregnancy rate after use of the new “wonder-drug” (46.3%) may just qualify as within the expected rate. These preliminary data, therefore, should be consumed with caution. This mysterious new drug supposedly is not a hormone, and it will be interesting to hear more on it in the future.
PLATELET RICH PLASMA (PRP): Since its initial introduction into clinical medicine in sports medicine and several years later into the fertility field by Greek investigators, PRP has in recent years become increasingly popular in many IVF clinics. Under such names as “ovarian rejuvenation” (without any evidence that it really rejuvenates anything in ovaries), PRP is widely utilized without standardization of the procedure and without properly conducted studies of appropriate statistical power to judge the treatments’ effectiveness and safety.
After initially being injected into ovaries in attempts to improve egg numbers and egg quality, it later found additional application in uterine perfusions, either for so-called implantation failure (whatever this diagnosis may mean) or for chronically thin endometrium.
Though published data are still sparse, some literature is slowly becoming available, with two recent papers briefly addressed here: A first involved intraovarian administration of PRP and was a prospectively randomized study of young women (<age 38) with poor prior ovarian response and two or more prior IVF cycles with <3 retrieved oocytes. The study involved only 83 patient cycles (41 PRP+ and 42 PRP-) and, therefore, was not an adequate study size to ever reach reliable conclusions regarding pregnancy or even live birth rates in IVF cycles. The authors, therefore, used as primary outcome the number of MII oocytes, which for several reasons leads to question about the reported study outcome and interpretation.
The authors found no difference in MII oocyte numbers and concluded that intraovarian PRP does not improve IVF outcomes in women under age 38. Based on these results, they furthermore recommended against the use of PRP in this kind of infertility population.
This study, however, does not warrant the authors’ conclusions and here is why: Though allegedly powered to define significance at a difference of one MII oocyte, this does not represent a valid cut-off. To make the point more bluntly, whether a 32-year-old female produces three or four MII oocytes likely does not make much practical difference. More importantly, however, a recent report suggested that, as ovaries age, their ability to produce good-quality embryos declines for MII oocytes, while the ability of immature oocytes improves. MII oocytes, therefore, reflect only a part of the cumulative pregnancy chance of an IVF cycle and the researchers in this study should have used not only MII oocytes but the total oocyte number retrieved (including MI and GV oocytes) as their primary endpoint.
This does not necessarily mean that the conclusions the paper presented may not turn out to be correct, but in this paper presented data do not allow for the conclusions the authors reached: that intraovarian PRP does not work and, therefore, should not be used. Whether intraovarian PRP improves IVF chances and, if so, in which kind of infertility patients, still remains to be determined. Following widely accepted ethical norms, intraovarian PRP, therefore, should be presented to patients as a still experimental procedure.
The second article is a review of PRP use for endometrial perfusions. One cannot expect from a review article new revelations and that applies also to this paper. However, it offers a single source for what little is known about endometrial applications of PRP.
… IVF cycle outcomes…
FOLLICULAR VOLUME: Spanish colleagues published in Fertility and Sterility a superficially very daring paper, suggesting that follicular volume (i.e. taking current standard follicular diameter measurements from 2 into 3 dimensions) and concluding that a follicular volume of > 0.56 cm3 denoted oocyte maturity.
It is rare that a paper can be judged as completely misleading simply from its title; but this is one of these papers because suggesting that—independent of age and other parameters affecting oocyte maturity like functional ovarian reserve—to assume that one universal cut-off value for follicle volume can define oocyte maturity is not only complete nonsense but, furthermore, demonstrates a surprising lack of understanding of the physiology of ovarian aging. As is now increasingly well understood, with advancing female age oocytes reach maturity at progressively smaller follicle sizes (i.e., therefore, obviously, also at progressively smaller follicle volumes). This is also a main reason why menstrual cycle length (indeed, the follicular part) shortens with advancing female age. In short, this paper would be best ignored.
SEMEN EXPOSURE AFTER EGG RETRIEVAL? Acceptance of the fetal semi-allograft by the maternal immune system is crucially important for implantation and early post-implantation development of pregnancies. Reproductive immunologists, therefore, have argued for decades that exposure of the female immune system to paternal semen (i.e., paternal HLA antigens) prepares the maternal immune system for the paternal allogeneity of the implanting embryo. And from this argument arose the hypothesis that post-retrieval semen exposure through inseminations may benefit IVF outcomes. Now Swedish investigators, however, seem to have put an end to it.
In a double-blind placebo (saline)-controlled prospective trial of 792 couple cycles (393+ and 399-) with primary outcome live birth, the seminal plasma group recorded positive pregnancy in 35.4% of cycles, clinical pregnancy in 28.8% (-6.6%), and live birth rates in 26.6% (-8.8%) while in controls the corresponding numbers were 37.3%, 33.6% (-3.7%), and 29.8% (-7.5%). While the paper correctly noted no effects of paternal semen exposure on live birth rates, we purposely here calculated miscarriage rates because those also demonstrate no difference between both study groups and some reproductive immunologists have in the past also argued for inseminations as treatment for repeated pregnancy loss. This paper, therefore, very clearly made the point that post-retrieval inseminations are useless.
WHAT IS THE BEST FET PROTOCOL? Another still controversial issue in IVF practice has been the question: what is the best method to prepare a woman’s endometrium for frozen embryo transfer (FET)? A Vietnamese study of 1428 patient cycles (476 in each of 3 group) has addressed this question in a randomized open label study and found no difference between natural, modified natural, or artificial endometrial preparations. Considering how long this question has remained unresolved, this is unsurprising since any obvious difference would have been detected already long ago.
This likely leaves the field with natural cycles (whenever women have regular ovulatory cycles), augmented natural cycles (when their natural cycles prove insufficient in generating a good endometrium), and with artificial cycles for the rest of all patients.
EFFECTS OF AIR POLUTION ON IVF: Another ESHRE 2024 abstract deserves mention in regard to this issue (Abstract O-075) because Australian investigators within this context reported that increases in particulate matter (in the study defined as PM2.5 and PM10) prior to oocyte collection were found to be associated with significantly decreased live birth rates in subsequent frozen embryo transfers. More specifically, exposure to PM10 (10 micron in diameter and can include dust, pollen, and mold) particles decreased odds of live birth by 38%. Finer particles of PM2.5 size, usually produced by vehicle exhaust and industrial pollution, were also associated with lower odds of live birth, but to a lesser extent.
This retrospective study involved 3,657 FET IVF cycles between 1/2013 and 12/2021 which besides PM10 and PM2.5 data from the government also included SO2, CO, O3, and NO2 levels at 24 hours, 2 and 4 weeks, and 3 months prior to egg retrievals. Only increases in particulate matter affected IVF outcomes.
These are interesting results for several reasons: First, because one would expect current IVF laboratories to filter out these large-size particles. And if this were to be correct, then observed associations could not be considered direct lab-effects on gametes and embryos, but, more likely, health effects on fertility patients themselves. And, if this were, indeed, to be the case, how would these effects so quickly translate into effects on IVF results?
Could this mean that some Australian IVF laboratories do not have clean enough air? Some clarity may come from a still unpublished multicenter U.S. study which assessed IVF cycle results during 2023 when huge wildfires in Canada brought major air pollution to the East Coast.
UTERINE PERFUSION STRATEGIES: We above briefly discussed uterine perfusions with PRP. But endometrial perfusions are now performed for several reasons, using varying agents. This, besides PRP, includes granulocyte colony stimulating factor (G-CSF), white blood cells, human chorionic gonadotropins, etc. All of these treatments have remained controversial because reported outcome data are insufficient to judge their effectiveness.
A recent review article by Chinese colleagues, however, summarized this limited experience and, therefore, is a worthwhile read. It is, however, difficult to agree with the authors’ conclusions that all of these treatments represent promising strategies, and that PRP likely is the best. A more cautious tone would have been preferrable, suggesting that some of these strategies may turn out to be effective in selected patients and only under specific clinical conditions. Moreover, since none of these strategies has been established as effective, they should be considered experimental procedures, and their use should be limited to study protocols.
MOSAICISM IN PGT-A: It is impossible to discuss IVF cycle outcomes these days without at least one reference to preimplantation genetic testing for aneuploidy (PGT-A). The opportunity presented itself with publication of a review article in Genes which attempted to explain reported live births after transfers of by PGT-A as “mosaic” reported embryos. The three authors specifically wanted to contribute an explanation to the “disappearance” of aneuploid cells during early stages of embryo development.
They failed!
And they did so for several reasons: First, they failed to acknowledge the difference between how PGT-A laboratories define mosaicism and how the rest of the biological world defines mosaicism. PGT-A labs uniformly define mosaicism based on an on-average 5-6-cell trophectoderm biopsy of a blastocyst-stage embryo that demonstrates more than one euploid cell lineage. This, however, contradicts the worldwide accepted biological definition of mosaicism which refers not only to 5-6 cells of any tissue, but to a complete organism and all of its cells (i.e., in this case a complete blastocyst-stage embryo) that contains more than one euploid cell lineage.
Here is one crucial reason why this review article really misses the point: A second–in this case aneuploid–cell lineage in 5-6 random cells of a blastocyst-stage embryo, of course, has significantly different biological meaning than a second such lineage anywhere in a complete 250 to 350 cell blastocyst-stage embryo. The table below explains this in detail:
__________________________________________________________________________________
5-6 cell TEB PGT-A result CORRECT INTERPRETATION __________________________________________________________________________________
EUPLOID EUPLOID or more likely MOSAIC*/**
MOSAIC MOSAIC
ANEUPLOID ANEUPLOID or possibly MOSAIC*/**
_________________________________________________________________________________
*In order of likelihood; **Often reported as low or high
As will be apparent from the table, whatever the official PGT-A diagnosis is, it will in a majority of cases be incorrect because most embryos signed out as euploid will be actually mosaic. This is likely also correct for most embryos signed out as aneuploid, though it is reasonable to assume that the diagnosis of aneuploidy will be more often correct than a diagnosis of euploidy. That a diagnosis of euploidy involves a majority of really mosaic embryos is indirectly acknowledged by the PGT-A laboratory community in how euploidy is defined by most laboratories: Almost all of them define an embryo as euploid if it has less than 20% intermediate DNA counts (i.e., <20% of a second aneuploid cell lineage) which, of course, even at 5% or 10% defines an embryo as mosaic. Some PGT laboratories, moreover, now have extended an euploid diagnosis to as much as 50% aneuploid DNA.
That the authors of above cited paper completely omit all of these facts in their review is unusual because not only should the correct definition of mosaicism be the basis for every diagnosis of an embryo by PGT-A, but should also be obvious considering that single cell studies of human embryos now have irrefutably demonstrated that, at blastocyst-stage, at least ca. 80% of embryos have aneuploid cells (i.e., are mosaic). That these authors, therefore, even in the title of their article raise the possibility that PGT-A may “overestimate” mosaicism is truly absurd.
CHEMICAL PREGNANCIES: What causes chemical pregnancies, which represent ca. half of all pregnancy losses, is still unknown. Spanish colleagues have published an “interesting” paper in Human Reproduction after trying to determine whether biochemical pregnancy loss was caused by the embryo or by the endometrium. They did this by testing in autologous IVF cycles embryo ploidy (by PGT-A and, therefore, with all of above outlined inaccuracies) and embryo transfer after performance of an endometrial receptivity assay (ERA, of course, another highly controversial test) or with the use of donor oocytes, and with embryos transferred in a frozen-thawed embryo transfer cycle (FET).
What they found was that the risk of a chemical pregnancy did not differ between autologous and donor eggs and with use of PGT-A and /or ERA and hypothesized that chemical pregnancies “may be related to a mechanism not associated with the embryos’ chromosomes or transcriptome of the endometrium.”
A more likely correct interpretation, however, is that neither PGT-A nor the ERA really do what they are claimed to do by so many interested parties, i.e., these two add-ons to IVF may simply not work!
That, however, does not necessarily mean that their conclusion that neither embryos nor endometrium contribute to chemical pregnancies is necessarily incorrect: The literature, indeed, suggests that chemical pregnancies may in many cases have an immune etiology.
…and social/planned oocyte freezing
Two recent papers from Israeli colleagues deserve mentioning under this subtitle. In a first, the authors conducted a systematic literature review on the subject and involved 3,847 records but included only ten studies performed between 1999 and 2020 of 8,750 retrieval cycles, a remarkably small number reflective of how little is really know about the efficacy of this clinical concept.
Their mean age at time of cryopreservation was 37.2 years, a reflection of the previously reported fact that most women fail to pursue egg-freezing at best ages. The return rate was only 11.1%, an especially low number considering the relatively advanced ages of these women at time of cryopreservation. Among those who returned, the mean age at cryopreservation was, indeed, even higher than for the complete group (38.1 years). The average number of cryopreserved oocytes, moreover, was only 12.6, likely barely enough for a reasonable live birth chance for one child. Oocyte survival was recorded in 78.5% of eggs, and the live birth rate per patient was only 28%. Among women who cryopreserved above age 40, the live birth rate was only 19.0%, while women under age 35, demonstrated a live birth rate of 52.0%.
These data are not very reassuring, especially as one can assume that reported outcomes than average community clinics can expected to produce. The message appears clear: The concept of social egg-freezing to extend reproductive capacity into more advanced years works reasonably well if pursued at relatively young ages. However, the field so far has not succeeded in educating the public appropriately.
For practical reasons, the responsibility for educating younger women primarily falls on our colleagues in general Obstetrics and Gynecology. It is the responsibility of the infertility field, however, to educate these colleagues about the concept and execution of social egg-freezing.
The second paper of special interest was a study which–in an attempt to maximize outcomes of retrievals in women undergoing fertility preservation–the authors concentrated on the timing of oocyte retrieval.
They matched 140 women undergoing fertility preservation cycles with 140 women undergoing standard IVF for male factor infertility. As one would expect, fertility preservation patients were stimulated more aggressively with higher gonadotropin dosages and, consequently, produced more oocytes.
In this group of patients, the investigators, however, noted a negative correlation between number of large follicles and number of retrieved oocytes. Though there was no association between number of large follicles and oocyte maturity for the whole study group, age stratification revealed a negative correlation, and the authors concluded correctly that these findings challenge traditional beliefs regarding oocyte retrievals.
They, indeed, support the two earlier discussed notions that as women age their follicle phases (and with it their complete menstrual cycles) shorten and that lead follicles (per size) at some point may exert adverse effects on smaller follicles. Both of these observations combined, of course, strongly support a concept of highly individualized timing of egg-retrievals, as now advocated by The Center for Human Reproduction in NYC for almost ten years.