A PIECE OF MY MIND: Many Twin Pregnancies Have Mistakenly Been Defined as Adverse IVF Outcomes, Leading to Overuse of eSETs — In View of Real Cycle Outcomes and Patient Desires, Time to Reconsider!
Today’s posting offers an article by Norbert Gleicher, MD, the Medical Director and Chief Scientist at The Center for Human Reproduction in New York City. The article addresses the fact that for about 20 years, the worldwide infertility field has almost universally accepted that twin pregnancies represent an unacceptable adverse outcome of infertility treatments. This has led to what must be considered a radically illogical dogma of almost universal elective single embryo transfer (eSET) in IVF.
The CHR has been investigating this issue for many years and – for what we believe are very good reasons – has disagreed with this consensus in several publications and on many different grounds. Following the CHR’s longstanding practice of at times counter-arguing established dogma in infertility practice by “thinking differently” and, therefore, following several prior precedents, through which the CHR has contributed to worldwide improvements in misguided fertility practices, we have not been willing to accept the dogma of almost universal eSET as a settled subject. We, therefore, hope to once again reopen the discussion of this subject with this article.
The CHR’s Editorial Staff
A PIECE OF MY MIND: Many Twin Pregnancies Have Mistakenly Been Defined as Adverse IVF Outcomes, Leading to Overuse of eSETs — In View of Real Cycle Outcomes and Patient Desires, Time to Reconsider!
By Norbert Gleicher, MD, Medical Director and Chief Scientist, at The Center for Human Reproduction in New York City. He is also the Editor-in-Chief of the CHRVOICE and The Reproductive Times and can be contacted though either the CHR’s editorial office or directly at either of his two e-mails: ngleicher@thechr.com or ngleicher@rockefeller.edu.
Preamble
During the last two decades, professional societies have increasingly characterized twin pregnancies after assisted reproductive technologies (ART) as adverse treatment outcomes. This concept originated largely from Scandinavian investigators and subsequently became incorporated into international IVF guidelines on both sides of the Atlantic, promoting almost universal elective single-embryo transfer (eSET). The obviously declining twin pregnancy rates as a consequence have, moreover, been celebrated as evidence of “success” by ESHRE and ASRM, the two leading professional societies in the fertility field.
Although reducing preventable obstetrical and neonatal complications are, of course, appropriate outcome goals for IVF practice, the universal characterization of twins as treatment failures is based on several very questionable ethical assumptions and methodological approaches that clearly not only deserve but mandate reconsideration. Current dogma, moreover, ignores the likely most basic tenant of proper risk-benefit and cost-benefit considerations: While many are gleefully pointing out the alleged benefits of significant reductions in twin pregnancy rates, the discussion of the subject has completely ignored any discussion of adverse consequences of reducing twinning rates and those, of course, do exist.
This article, therefore, argues that the current eSET doctrine suffers from five major shortcomings: (i) It limits patient autonomy; (ii) It applies a uniform standard to highly heterogeneous infertility patients; (iii) It relies on inappropriate statistical comparisons between one twin pregnancy and one singleton pregnancy instead of comparing equal reproductive outcomes; (iv) It ignores the reality that infertility patients cannot be assured of achieving a second singleton pregnancy; and (v) it has completely ignored what must be considered adverse outcomes from reductions of at least some twin pregnancies.
Some History
Beginning in the late 1990s and early 2000s, investigators from Scandinavian countries and Belgium increasingly promoted the concept that IVF success should no longer be measured simply by establishment of pregnancies or live births, but by the birth of only a single healthy infant. These views eventually resulted in widespread advocacy of eSET,1-3 and as recently as 2020, in the argument that “one (embryo) at a time should be the normal routine.”4 The most recent ESHRE Practice Guideline, published in 2024, noted that “no clinical or embryological factor per se justifies a recommendation of double embryo transfer (2ET) instead of eSET,” - a rather astonishing comment in an even more astonishing Practice Guideline.5
Though practically saying the same thing, the ASRM said it more sensibly by noting that “the most direct way to limit the risks of multiple gestation from ART is to transfer a single embryo.” The ASRM, in its last policy statement on the subject in 2022, at least noted the potentially improved pregnancy chances with 2-ETs (or transfer of more than one embryo in general), and, therefore, acknowledged that there is a “price” to pay for an (almost) universal eSET policy, - and that “price” is lowered pregnancy and live birth chances.6
And this conclusion then automatically clashes with extensive literature that suggests that successful conception within a desired timeframe clearly represents a frequently top-ranked desire of infertility patients,7 - though it, of course, does not universally outweigh everything. The relative weight of “success/quick conception” versus “risk and burden” varies by individual, age, prognosis, and life circumstances; the literature very clearly emphasizes heterogeneity rather than a single hierarchy.8
This is, of course, an incredibly important point that must be made, with especially age being likely the best example: The urgency of a 25-year-old to conceive will likely be smaller than the urgency of a 35-year-old or 45-year-old patient, with the latter representing roughly the median age of the CHR’s patient population over the last three years (compared to 36 years for all U.S. IVF clinics).
The Marginalization of Patient Autonomy
Modern medicine increasingly emphasizes shared decision making. Yet, in IVF, one finds a remarkable exception: Professional societies increasingly prioritize the physician – not the patient – in determining what constitutes an acceptable treatment outcome.
Are physicians entitled to demand eSET and refuse 2-ET, - and are medical societies entitled to do so?
Patients may reasonably value completing family building in one pregnancy, - especially if they are older, have already failed multiple IVF cycles, and/or are running out of money. Don’t they have a right to try to avoid repeating IVF cycles, minimizing repeated anesthesia, trying to avoid potential miscarriages, and/or repeated implantation failures (if such a thing really exists)? And shouldn’t they at least try to avoid – as already noted – financial consequences while trying to “beat” advancing maternal age?
For many patients, particularly women over age 40, twins represent success – not failure. So where is patient autonomy, when physicians practically take it away?
Uniform Standards in Highly Heterogenous Patient Populations
When all of medicine is pivoting toward the highly-individualized care of precision medicine, the infertility field is going the opposite way. And this is the case not only when it comes to eSET for (almost) all patients. It’s applied to almost all of infertility practice: We stimulate ovaries with gonadotropins produced in glycolization patterns that mimic gonadotropins extracted from postmenopausal urine of nuns because that had been the original source of gonadotropins. Why is nobody asking why we treat young women with nature’s gonadotropins for older, postmenopausal women?
Or why is one prominent NYC fertility clinic allegedly denying women above age 43 appointments, - if they don’t have a prior agreement to use third-party egg donations? Haven’t they noticed yet that there are women with “young” ovaries over age 43 and women with very old-behaving ovaries at age 33? Or how about all-blastocyst-stage culture for everybody, or all-freeze embryo banking cycles for everybody, or how about refusing patients IVF cycles unless they agree to PGT-A of their embryos?
Medicine increasingly emphasizes individualized care, - while embryo transfer policy has evolved toward exactly the opposite. Important variables in patient populations that are overlooked in the process are, of course, female age (the average 28-year-old woman differs fundamentally from the average 43-year-old), a patient’s prognosis (is it her 1st or 10th IVF cycle?), and her cause of infertility (is it tubal disease, male factor, diminished ovarian reserve, endometriosis, immune-mediated implantation failure, or something else?).
And then there are, of course, financial considerations at play. Repeated IVF cycles may be affordable in Scandinavia, where they in Sweden, for example, are usually covered by government insurance until age 40 (there are exceptions, of course). A dogmatic eSET policy under such circumstances may – at least at the margins – appear more logical but – when carefully considered – still remains totally non-sensical.
Moreover, in the United States, many patients still do not have insurance coverage for infertility services, especially IVF. Healthcare economics, therefore, directly affect acceptable risks for patients.
Statistical Outcome Risk Comparisons Between Twin and Singleton Pregnancies Have Been Fundamentally Flawed in Most Cases
This is perhaps the most strident argument against current eSET recommendations because this argument questions the basic hypothesis behind routine eSET practice, which is dependent on the claim that twin pregnancies create excessive maternal as well as neonatal newborn outcome risks, in comparison to singleton pregnancies. Assuming that such a significant risk does not really exist, not only could consideration of 2-ET no longer be attacked as harmful, but – in view of how important timely pregnancy success should be in every IVF treatment – 2-ET, indeed, would for many, if not most patients, have to be considered a very viable option and, indeed, maybe a preferable treatment over eSET.
And – considering the history of how this argument was introduced into the medical literature – there, indeed, are some significant doubts: At least initially, when our Scandinavian colleagues proposed the eSET policy, all studies compared outcome complications between one twin pregnancy with one singleton pregnancy. But these outcomes are biologically unequal, - comparing one twin pregnancy, which produces two infants, to one singleton pregnancy, which, of course, produces only one offspring. A statistically proper comparison should therefore be made with a comparison of one twin pregnancy with two consecutive singleton pregnancies (more on this below), a point we, here at the CHR, have been making since the eSET hypothesis was first introduced into the literature.9,10
After years of making this argument, Swedish colleagues conducted such a study demonstrating practically no longer any meaningful differences in pregnancy outcomes – maternal and/or neonatal.11 Key findings were: (i) When comparing one twin pregnancy to two singleton pregnancies in the same mothers (or matched cohorts), differences in many serious outcomes were minimal or absent after adjustment. (ii) Neonatal risks were largely limited to minor/short-term issues like respiratory complications and jaundice (higher in twins), with no significant differences in perinatal mortality, Apgar scores <7, first-year mortality, or congenital abnormalities. (iii) Maternal risks (e.g., preeclampsia, gestational diabetes) also showed limited or no differences in key areas. The study emphasized that twin pregnancies carried higher preterm/low birth weight risks, but the overall profile supported nuanced risk assessments.
And this does not even take into account that Luke et al. reported that twin-risks for most adverse outcomes are elevated in sub-fertile and IVF twins, compared to fertile twins.12 This paper supported discussions on the added risks of multiples in association with IVF, - while highlighting factors like fertility status. It therefore contributed to the broader context of comparing twin vs. singleton (or cumulative singleton) risks in infertility treatment, a finding first already reported in the BMJ by the Dutch investigator F.M. Helmerhorst in 2004.13
Then – and to this day – widely ignored by the infertility field, he pointed out that IVF twins demonstrated significantly worse outcomes after assisted conception than after spontaneous conception (e.g., higher risks of very preterm birth [RR 3.27], preterm birth [RR 2.04], very low birth weight [RR 3.00], low birth weight [RR 1.70], as well as higher Cesarean section rates, NICU admissions, and perinatal mortality [RR 1.68]).
In contrast, the risks of twin births were much closer and significantly lower in assisted vs. natural conceptions. Notably, perinatal mortality was about 40% lower after IVF (RR 0.58 in matched studies). There were minimal or no increases in very preterm birth, very low birth weight, or other major issues; only some outcomes like Cesarean section and NICU admission showed small increases.
While these outcome data on first impression may appear counterintuitive, they actually do make sense if one considers the fact that IVF pregnancies are diagnosed much earlier than spontaneous pregnancies and that, of course, means that IVF twins receive from early on much closer medical attention. His data thus already in 2004 pointed out that the most basic premise of the eSET hypothesis – significantly increased statistical risks from twin pregnancies in comparison to singleton pregnancies – is faulty.
But That Is Not Even All
All of this, moreover, does not even yet take into account the time and effort required to establish a second pregnancy and achieve a second delivery, - nor does it consider related costs for a second round of infertility treatments. But probably most importantly, a comparison between one twin and two singleton pregnancies does not take into account that – especially in an infertile patient population – even the best fertility clinic cannot guarantee a fertility patient a successful second pregnancy. Indeed, with infertile patient populations aging all around the world, the success of a second pregnancy attempt – because of pregnancy length, lactation period, and recovery time for the mother on average at least a two-year period – has a notably reduced chance, as after age 38-40, pregnancy chances quickly decline.
One successful twin pregnancy may, therefore, for many women, represent the only realistic opportunity to achieve a desired family size and, therefore, by itself, fundamentally alter the ethical balance in the discussion of eSET versus 2-ET.
This is also a fundamental reason why the argument by Finnish colleagues that eSET should also be the rule in older patients between ages 40 and 44 was so bizarre when it appeared unchallenged in 2013 in Human Reproduction,14 – yet receiving much more attention in the infertility field than, for example, Helmerhorst’s above-noted groundbreaking 2004 study.
How the Scandinavian Perspective About eSET Fooled the World
The original Scandinavian recommendations for eSET, defining twin births as basically adverse – and under all circumstances to be avoided - IVF outcomes, emerged from countries characterized by government-funded IVF, excellent neonatal intensive care, strong public health orientation, at that point at least still relatively homogeneous populations, and minimal financial burden for patients. These assumptions do not generalize internationally and, therefore, - even assuming that they, at least to a degree, support an eSET concept – do not apply to big parts of the world, including the U.S.
But aside from the fact that even these explanations do not warrant current eSET policies ethically, they, of course, appear almost bizarre in their undisputed acceptance in the U.S., where patients often pay tens of thousands of dollars out of pocket for IVF treatments. The calculus in the U.S. and in many other countries, therefore, differs substantially from the Scandinavian calculus and we wonder – considering the relative silence of our Scandinavian colleagues regarding this subject in most recent years – whether in view of increasing heterogeneity in their populations from immigration and considering increasing financial pressures on their health care systems – there has not been at least silent recognition of how much damage the current eSET policy they brought to the world really has caused and is still causing. One, therefore, would wish that they spoke out on the subject. As the original initiators of this practice, they, indeed, have a certain responsibility to speak out and, of course, have the standing to do so!
So, What is the Solution?
As in so many other areas in medicine, what we are witnessing regarding the here-discussed issue once more demonstrates that any rule that is applied “to everybody,” - i.e., rigid protocols like “eSET for everybody,” is non-sensical. Unfortunately, the infertility field has been going into this direction for much too long regarding so many different practices, from routine blastocyst culture for everybody, to routine all-freeze cycles, to preimplantation genetic testing (PGT-A) for all.
And all of this is happening in the fertility field at a time when almost everybody else in medicine is trying to implement “precision medicine,”- highly individualized medical practice specifically geared at the circumstances with which a patient presents. And these circumstances, of course, also include risk-benefit and cost-benefit assessments for every medical intervention.
Whether patients undergo eSETs or 2-ETs, therefore, must be individualized.
Rather than universal eSET, embryo transfer policy should incorporate maternal age, patient prognosis, number of previous failed IVF cycles, functional ovarian reserve, patient preferences, financial considerations, medical co-morbidities, desired family size, and obstetrical history, - in other words, an approach consistent with precision medicine.
The goal of infertility treatment is not merely to optimize obstetrical statistics, but to help patients build the families they desire. Accordingly, embryo transfer decisions must be individualized, rather than dictated by a universal preference for singleton gestations.
REFERENCES
Bergh C. Hum Reprod 2005;20(2):323-327
De Neubourg D, Gerris J. Hum Reprod 2006;21(4)843-846
Nygren et al., Acta Obstet Gynecol Scand 2007;86:774-782
Bergh et al., Fertil Steril 2020;114:673-679
ESHRE Guideline Group on the Number of Embryos to Transfer. Hum Reprod 2024;39(4):647-657
ASRM. A Committee Opinion. Fertil Steril 1022;117:498-511
Assaysh-Öberg et al., Sex Reprod Health 2023;37:100879
Carson SA, Kallen N. JAMA 2021;326(1):65-76
Gleicher N, Barad D. Fertil Steril 2009;91(6)”2426-2441
Gleicher N. Contemp Ob/Gyn September 1, 2013
Sazonova et al., Fertil Steril 2013;99(3):731-737
Luke et al., Am J Obstet Gynecol 2017;217(3):330.e1-330.e15
Helmerhurst FM. Brit Med J. 2004;328
Niinimäki et al., Hum Reprod 2013;28(2):331-335