LITERATURE for the INFERTILITY CLINIC
BRIEFING: The Reproductive Times in this section offers commentaries on a broad survey of recent articles in the English literature that the CHR considered of interest regarding the practice of clinical infertility, even if, at times, they are not immediately applicable to daily clinical practice. Articles are mostly chosen for two reasons: their potential translational value for immediate clinical practice or to help in determining where clinical practice might evolve.
IN VITRO FERTILIZATION (IVF)
Does human growth hormone affect IVF cycle outcomes? And what else we can learn about medical publishing from this paper
In comparison to most other medical journals in the fertility field, Human Reproduction is usually quite good. However, no journal is, of course, perfect, and the journal recently got a black eye for publishing a truly ridiculous paper on the routine use of human growth hormone (GH) during in vitro fertilization (IVF) cycles.
The title of the paper already said it all and also reflected a truly exaggerated degree of unearned self-assurance: “Empirical use of growth hormone (GH) in IVF is useless: the largest randomized controlled trial” (1). As it turned out, both statements were unsustainable: The study did not demonstrate that GH in IVF cycles was “useless.” In fact, the study left this issue unresolved. And while the paper did report a study of considerable size, it was not the largest ever randomized study of GH use in IVF because, as administered, there was no GH effect on ovaries in the index cycle. Let us explain:
The study attempted to investigate whether supplementation with GH in a general population of infertile women undergoing an antagonist IVF cycle would improve IVF cycle outcomes. The absurdity of this study design is, however, similar to a hypothetical study of aspirin in a prospectively randomized study of a general population (where a large majority likely have no headache) to see whether aspirin, indeed, does improve headache. Very obviously, such an aspirin study would make no sense!
A medical treatment can show effectiveness, of course, only if the target population is affected by the medical problem the treatment is supposed to address. In an IVF cycle, this means that GH supplementation makes sense only if insulin-like growth factor-1 (IGF-1) levels are abnormally low because GH acts on small growing follicles during folliculogenesis synergistically with FSH and androgens through IGF-1. And since abnormally low IGF-1 levels in infertile women are very rare, this study – with great likelihood – treated a large majority of women with GH who did not need it because their IGF-1 levels were within the normal range.
Who, therefore, can be surprised by the conclusion of the study that GH showed no effects?
We have, indeed, in these pages repeatedly pointed out that the persisting controversy regarding GH supplementation in IVF is likely at least to a large degree due to the fact that different studies with varying outcomes just reflected the different prevalence of abnormally low IGF-1 levels in study populations.
This article, however, also raises another important point: namely, that because IGF-1 effects primarily occur only on small growing follicles between secondary and small-antral follicles – which still need at least 6-8 weeks of further growth and development before reaching the gonadotropin-dependent state, when they become available in an IVF cycle – GH supplementation must begin in women with low IGF-1 at least 6-8 weeks before the IVF cycle begins if this treatment is expected to affect oocytes. And in this study, GH was only given during the IVF cycle, thus treating the follicle cohort 1-2 months following the index IVF cycle reported in the manuscript.
That this paper made it through an otherwise responsible peer-review process is, of course, of concern but must be viewed as a symptom of a general deterioration in peer review. Objectively speaking, there are many good reasons for this development that explain – but do not excuse – this indisputable conclusion. Yes, the volume of submissions at almost all publications has dramatically increased despite the establishment of huge numbers of new journals in parallel. But most editorial offices are still staffed by part-time editors who work as volunteers or – at best – for only minimal payments. More submissions also require more reviewers, and good reviewers are scarce. They are also overwhelmed by the incredible number of review requests they now receive, a service that is generally expected to be done for free.
In practical terms, all of this means that the number of submissions which editors don’t even send out for review but reject “in-house” has dramatically increased. But while editors may be well-qualified in some areas of their specialty, they may be less so in a large majority of subspecialty areas within their own medical specialty. Consequently, their judgment about what to reject even before peer review will be more driven by what may be of interest to readers and/or may be appealing fodder for the media, rather than by the quality of a paper. The paper discussed here would be a good example of this, considering the very strong – though obviously incorrect – statements made by the authors already in the title of their paper.
Editors, unfortunately, still also pretty automatically overvalue prospectively randomized studies, a phenomenon likely even more pronounced in reproductive medicine than in other specialties because of how difficult (and costly) such studies are in reproductive medicine and how rarely one, therefore, sees such clinical trials in the field. Once again, this article offers a good example, proclaiming already in its title not only that it is a prospectively randomized study, but the largest ever done on the subject. We, of course, can see the allure in the editors’ first – and, of course, so important – impression. There, of course, exists ample literature that demonstrates the very special dangers of poorly executed and/or interpreted clinical trials because their results are so overestimated (2).
Relatedly, another recent paper is also worth mentioning. In it, investigators from several countries attempted to determine how the IGF system participates in the midcycle surge in women. As noted above, IGF-1 signaling is well known to affect follicular function, especially during small growing follicle stages, but also to a lesser degree beyond. How it is involved in the ovulatory peak, which greatly affects granulosa cell function, is still largely unknown. What this study now suggested is that downregulation of IGF signaling at that time is mediated by increases in the expression of the inhibitory protein, STC1, which then is instrumental in the sudden cessation of GC proliferation and the onset of differentiation during the ovulatory peak (3).
References
1. Mourad et al., Hum Reprod 2024; doi: 10.1093/humrep/deae251. Online ahead of print. [WORST PAPER AWARD]
2. Kis D. Medthority. March 30, 2023. https://www.medthority.com/news/news-trends/clinical-trials/how-can-flaws-in-clinical-trial-design-conduct-and-reporting-impact-clinical-decision-making/
3. Bøtkjær et al., J Clin Endocrinol Metab 2024;110(1):e160-e167
More on fresh vs. frozen embryo transfers
One would think that the question of whether fresh or frozen-thawed embryo transfers in subsequent cycles is the better treatment option would finally have been resolved, but no – FS Reports decided to publish another study on the subject after accepting a retrospective single-center study involving 8,319 autologous first blastocysts (in the absence of PGT-A). Among those, 6,755 transfers were fresh, and 1,564 were FETs (frozen-thawed transfers) after freeze-all cycles (1).
Considering the retrospective nature of this study, one cannot be surprised by significant differences in patient populations between the two study groups (Table 1 of the study). Though the authors apparently made appropriate statistical adjustments, this finding alone raises serious questions about the validity of the study results, because one must assume that there were good clinical reasons why patients ended up in one group or the other.
Moreover, without explanation, the study retrospectively investigated IVF cycles only between January 1, 2015, and December 31, 2020. Considering how much IVF practice has changed since 2015 – and even since 2020 – the relevance of any results from such a long time ago for current practice must be seriously questioned. After all, the concept of freeze-all cycles was introduced around 2014 (2), and by 2020, it was already refuted by many (3).
Though the paper’s conclusion that there is no major difference between both approaches cannot be denied, this still does not allow for the conclusion the authors reached: that a final decision about which approach should be taken should be left to patients and physicians. While that, in medicine, should probably be the rule in all circumstances, assuming both options are really equal, why would anybody choose a more expensive option, which would, in addition, delay time to pregnancy? The correct medical recommendation from physician to patient – unless special circumstances exist – should therefore be for fresh transfers.
References
1. Pavlovic et al., F S Reports 2024; 5(4):2666-3341
2. Shapiro et al., Fertil Steril 2014;102:3-9
3. Ben Rafael Z. Hum Reprod 2020;35(10):2179-2184
AND, OF COURSE, MORE ON PGT-A
Even though the ASRM has by now already acknowledged that PGT-A does not improve outcomes in a general population of IVF cycles (and, indeed, in certain subgroups, may affect outcomes adversely), it is always amazing to read some of the things colleagues are still coming up with to claim outcome improvements for PGT-A and offer reasons to continue utilizing PGT-A in association with IVF. Two of the most prolific among the still-persisting alleged “experts” (both, of course, geneticists with apparently no clue about the physiology of reproduction) now, finally, got together and – somehow – convinced a BMJ-group genetic journal to publish an absolutely ridiculous Opinion article, allegedly based on “mining the SART database” (1).
Unsurprisingly, the article only ends up regurgitating by now widely rebutted data interpretations and comments that simply do not make sense. We understand that it is difficult to have been so wrong for so long, especially if – as is the case here – one of them was for the longest time considered the “Wunderkind” of PGT-A (as a geneticist, the other author was a relatively latecomer to PGT-A and really never caught up).
It is also telling that this article appeared in the genetics literature and not in infertility literature. And even there, it apparently didn’t make it into an established journal but ended up as the third article in a brand-new journal in obvious search of content.
One characteristic of what can now – after over 20 years of PGT-A utilization – be described as “the PGT-A disaster” is the ignorance of basic biological and physiological facts the genetics community has exhibited when it comes to chromosomal testing of preimplantation-stage embryos. But it even went beyond that, because the genetic testing community, to this day, uses an incorrect definition of “mosaicism” when sending out PGT-A report cards on embryos. Since the infertility field viewed them as “the experts,” the field has been relying on their judgments, with nobody in the fertility provider community having the authority to stop the disaster from continuing.
A recent Fertility and Sterility publication from Russia, in vivid contrast to the above-noted rather shameful paper, addressed mosaicism at preimplantation stages in human embryos with more sophistication. The paper came from Ilya Volodyaev, PhD, and collaborators, a physicist by training and a senior embryologist in practice, who actually contributed a wonderful article to the Reproductive Times. As in his article for the Reproductive Times, this article for Fertility and Sterility also offers some clarity between true biology and artifacts in chromosomal testing of embryos. Though in our opinion he still maintains too rosy a viewpoint of PGT-A, his argumentation – in contrast to the prior paper – is sharp and logical. A worthwhile read!
And here is a technically interesting news item: Researchers in a letter in Fertility and Sterility reported that blastocyst re-expansion at 1 hour after trophectoderm biopsy may be an embryo predictor of live birth (2). One can only hope that proponents of PGT-A don’t take this as another indication to recommend PGT-A to their patients!
Less interesting was another PGT-A paper in the same journal issue, in which investigators produced what we have come to call a “self-fulfilling prophecy paper” (i.e., a paper where the answer a study is supposed to find is too obvious to even require a study). Here, the investigators from California were apparently surprised to find out that aneuploidy rates in cycles where embryos underwent PGT-M and PGT-A had the same age-specific aneuploidy rates as women undergoing only PGT-A (3).
Really?! Did they really ever believe that PGT-M could affect aneuploidy rates in any potentially conceivable way? Too absurd!
Finally, another technically interesting paper from the U.S. Juno-Genetics Laboratory and others reported that certain PGT-A platforms may be able to detect inherited small copy number variants (CNVs) with very high specificity without prior knowledge of parental status, and these were either confirmed as benign, likely benign, or variants of uncertain relevance. Since the study involved couples with no known pathological CNVs, the authors correctly noted that the fact that no pathological or likely pathological CNVs were detected does not mean that these platforms may not also be able to detect those (4).
References
1. Munné S. Griffin DK. BMJ Connections Clin genet Genomics 2024;1:e000003 [WORST PAPER AWARD]
2. Elkhatib et al., Fertil Steril 2024;122(6):1147-1149
3. Martel et al., Fertil Steril 2024;122(6):993-1000
4. Itturiaga et al., Fertil Steril 2024;122(5):789-798