A MILESTONE STUDY: Comparing cumulative live births between cleavage- and blastocyst-stage embryo transfers
By Norbert Gleicher, M.D., who is the Medical and Laboratory Director as well as Chief-Scientist at the Center for Human Reproduction (CHR). He can be reached through The Reproductive Times at hello@reproductivetimes.com, at the CHR at ngleicher@thechr.com, or as a Visiting Researcher at Rockefeller University at ngleicher@rockefeller.edu.
Briefing: There are few questions in IVF potentially as consequential as whether an IVF cycle should be managed with cleavage- or blastocyst-stage embryo transfer. Unsurprisingly—considering that in many IVF clinics blastocyst-stage embryo transfers have become routine in most or even all IVF cycles—a recent SART outcome data review demonstrated that, between 2014 and 2020 alone, utilization of blastocyst-stage transfers increased by 20.1%. Likely because blastocyst-stage transfers are increasingly perceived to lead to favorable rises in live birth rates, those numbers probably moreover significantly increased since then. On several grounds, the almost universally accepted current impression of improved live birth rates with blastocyst-stage transfers must, however, be questioned. One obvious reason is the fact that most outcome data regarding this issue reported on first embryo transfer cycles, an obviously biased assessment for many patients with often greatly varying blastocyst numbers. Dutch investigators have now published in the British Medical Journal (BMJ) a prospectively randomized multicenter trial in good-prognosis patients with at least 4 blastocysts for transfer in a cycle which demonstrated that cumulative pregnancy rates in this patient population were similar between cleavage- and blastocyst-stage transfers. This article discussed the conclusions that can be drawn from these findings.
Especially after publication of this study by Dutch investigators (another among several such “masterpieces” the collaborative effort of the IVF community in The Netherlands appears to produce every few months) (1), one must conclude that, likely, only very few papers in the history of IVF have caused as much damage to IVF practice as the original paper by Gardner et al., promoting the use of blastocyst-stage embryo transfer under the very enticing heading, “Culture and transfer of human blastocysts increases implantation rates and reduces the need for multiple embryo transfers (2).”
That this was unintended can be assumed; nevertheless, it must serve as a warning for how a single deeply flawed paper can not only affect, but dominate, medical practice for almost a quarter of a century. Not that there hasn’t been criticism of the very general concept that extended embryo culture to blastocyst-stage was “the thing to do” because it (universally) improved IVF outcomes. Indeed, there was considerable criticism based on biased patient selection in the original study, including from the author of this article (and colleagues) (3). Only very good prognosis patients were included; yet that went not only unnoticed, but the results of the study were presented—and promoted—as applicable to every IVF patient.
That a level of discomfort persisted with how blastocyst-stage transfers were promoted was witnessed by attempts at meta-analyses by Glujovsky et al (4,5). trying to compare cycle outcomes after cleavage- and blastocyst-stage transfers; but those were, of course, dependent on previously published studies and, if those were equally or even only similarly biased to the original Gardner et al. study, it should not surprise that they were unable to offer clear answers.
Here addressed recent Dutch study, therefore, is deserving of applause. Once more, this time under the leadership of two veterans in the field, Simon Cornelisse, MD and Sebastian Mastenbroek, PhD, the unique collaborative research effort The Netherlands’ academic IVF centers have been able to establish has again born plenty of fruit in answering a crucially important question for the IVF community: Does it really matter in good-prognosis patients whether their embryos are transferred on day-3 (cleavage-stage) or days 5 or 6 (blastocyst-stage) if the desired outcome is not live birth chance in first transfer cycle but cumulative pregnancy and live birth chances? And the answer the study provided was very clear: Cleavage- and blastocyst-stage transfers achieved similar cumulative live birth rates in women who after one retrieval had at least 4 blasts available for transfer.
And, importantly, the methodology of this study was solid: it involved 21 hospitals and clinics during August 2018 and December 2021. Though it is unclear why publication of this study took almost three years—speaking from personal experience and, obviously, speculating—it would not surprise if the message of the paper had run into resistance during the peer review process and, therefore, required repeated submissions to different journal, and that takes time. The infertility field, unfortunately, has a history of often refusing publications which argue against what is considered the common wisdom of the moment.
The study recruited 1202 women, with 599 randomized to cleavage-stage and 603 to blastocyst-stage transfer. Primary analysis took place based on intention-to-treat, with secondary analyses based on protocols. Primary outcome was cumulative live birth rate per oocyte retrieval, including all frozen embryo transfers of excessive embryos within one year after randomization if a first transfer had failed (it is important to remember that inclusion into the study required at least 4 transferrable embryos). Secondary outcomes included cumulative pregnancy rates, pregnancy loss, live birth after fresh transfer, number of embryo transfers needed, number of frozen embryos, and obstetric and perinatal outcomes.
As already noted, the cumulative live birth rate did not differ (cleavage-stage 58.4%; blastocyst-stage 58.9%). After fresh embryo transfer (i.e., 1st transfer)—as one would expect—blastocyst-stage transfer had a slightly higher live birth rate (CI 1.26, 1.00-1.58) and a lower pregnancy loss rate (CI 9,68, 0.51- 0.89), and a lower number of required transfers required for a live birth (CI, 1.55 vs. 1.82, P<0.001). The incidence of moderate preterm birth (32-36 weeks) was, however, higher in the blastocyst group (CI, 1.87, 1.05-3.34).
These are fascinating results that go far beyond just the finding of no difference in cumulative live birth rate and include the following:
(i) These are—as in the original blastocyst-transfer study of Gardner et al (2).—findings that in principle only pertain to (very) good-prognosis patients because the requirement of 4 blastocysts for inclusion in the study at all ages clearly selects out only best-prognosis patients. This selection bias, moreover, increases with advancing patient age and declining functional ovarian reserve of a patient because both of these cycle characteristics are obviously directly linked to the chance of producing at least 4 blastocysts in an IVF cycle.
(ii) These findings, however, also suggest that, since in this best-prognosis patient group there was no difference in cumulative live birth rates, in poorer prognosis patients (i.e., in older women and younger women with low functional ovarian reserve), blastocyst-stage embryo transfer will do poorer. In other words, at a still to be determined age and /or lower than normal age-specific ovarian reserve (i.e., in women with premature ovarian aging, POA) affected women will actually do more poorly with blastocyst-stage transfer, an important point that should lead to significant changes in how, for example, ASRM and SART judge when blastocyst stage-transfers are indicated (6).
(iii) The same also implies to secondary outcomes. Though this study demonstrates mildly lower miscarriage rates and quicker time (fewer embryo transfers) to viable birth, these relatively minor advantages at some point will also disappear as women get older or advance in their POA. They, indeed, at some point may reverse to an advantage for cleavage-stage transfer. Moreover, increased prematurity risk with blastocyst-stage transfer may, indeed, further increase.
(iv) The principal conclusion of this study, therefore, is that blastocyst-stage embryo transfer is greatly overused in current IVF practice and its increasing utilization (7), therefore, should not only be stopped but reversed.
But this is not all: One also must consider unrelated consequences of excessive utilization of extended embryo culture on IVF. A good example is the concept of almost universal elective single embryo transfer (eSET), by Gardner et al (and later by many others) already in their initial paper touted as a major secondary benefit of extended embryo culture (2). That 2-embryo transfer (2ET) offers higher pregnancy and live birth rates than eSET is undisputed. What this study, therefore, strongly suggests is that the current worldwide policy which in most IVF cycles calls not only for blastocyst-stage embryo transfers but also for eSET must be reevaluated.
And then there is, of course, also to consider the still rapidly growing practice of preimplantation genetic testing for aneuploidy (PGT-A) which is automatically linked to blastocyst-stage culture because current PGT-A technologies require extended embryo culture, whether for trophectoderm biopsy or for use of so-called non-invasive PGT-A, which uses spent media from blastocyst-stage embryos to detect fetal DNA (and has even been suggested to extend embryo culture by one day).
The IVF field, therefore, as of this moment in time not only faces the fact that ASRM and SART only just recently reaffirmed that PGT-A, still has not demonstrated any significant beneficial utility for IVF cycle outcomes (8). As a consequence of here discussed BMJ publication, the IVF field finds itself, therefore, at an important crossroad, suddenly confronting the reality that several of the most basic practice patterns at most IVF clinics have been determined to serve no useful purpose or may for some patients even be harmful.
It will be interesting to see how the field will react to this news. Logical consequences will, of course, not only require significant clinical practice changes but will have also considerable economic repercussions. But patients should benefit because less unnecessary treatments will mean lower IVF costs; and if IVF needs one thing to become available for more patients, it is affordability.
References
1. Cornelis et al., BMJ 2024;386:e080133
2. Gardner et al., Feril Steril 1998;69(1);84-88.
3. Gleicher et al., J Assist Reprod Genet 2016;33(4):455-459
4. Glujovsky et al., Cochrane Database Syst Rev 2016;30(6):CD002118. doi: 10.1002/14651858.CD002118.pub5.
5. Glujovsky et al., Cochrane Database Syst Rev 2022;5(5):CD002118. doi: 10.1002/14651858.CD002118.pub6.
6. Practice Committees of ASRM & SART. Fertil Steril 2018;110(7):1246-1252
7. Andrew C, Doody K. Fertil Steril 2023;120(1, Suppl) E15-E16
8. Practice Committees of ASRM & SART. Fertile Steril 2024;122(3):421-434