Now accepting Telehealth appointments. Schedule a virtual visit.
Whether implantation failure is a real diagnosis remains to be determined. While scientific papers on the subject are published almost daily, nobody really knows what implantation failure really is and/or means. The definitions have varied: Some go by the number of failed IVF cycles, others by the number of good quality embryos transferred. Like with the pseudo-diagnosis of “unexplained infertility," it often seems that everybody has different definitions of what implantation failure is.
To go solely by the number of failed IVF cycles, quite obviously, makes little sense because three failed cycles at age 23 have very different meaning (and likely different causes) than the same number of failed cycles at age 43. The same applies to number of embryos transferred. Transferring consecutively five great-looking embryos of a 23-year-old and failing to achieve a pregnancy has different connotations than transferring five embryos, even all together, in a 45-year-old. How these two parameters at varying ages change has never been determined, and this is the principal reason why, strictly speaking, without proper prior diagnostic definition of this condition, there really cannot even be such a thing as “implantation failure.”
Yet, pretty much every time when they cannot figure out why patients do not conceive on repeated IVF attempts, fertility specialists all over the world use this diagnosis in abundance. Like in association with the so-called “unexplained infertility,” the apparent helplessness of the situation appears to mandate a solution, and one such solution is, of course, the invention of a hypothetical, yet undefinable, diagnosis, like implantation failure or unexplained infertility.
But giving the unknown a name does not solve the problem and, certainly, does not lead to appropriate treatments. Only clear definitions of a clinical conditions in their so-called clinical phenotype (i.e. clinical presentation) may do that. Indeed, the more diffuse the phenotype of a diagnosis, the more difficult will it be to learn what causes the problem.
This does not mean that implantation failure may not really exist. The implantation failure we are referring to, however, cannot be diffuse but must be highly specific. First, however, we need to understand how implantation works when everything goes as it should:
When an embryo enters the endometrial cavity, it does not immediately implant. Indeed, it spends ca. 48 hours within the biological microenvironment of the endometrial cavity, made up of thick mucus produced by endometrial glands. There must be a biological purpose for this 48-hour way station on the highway to implantation. The embryo during those two days very likely does a lot of different and important things we have absolutely no idea about. One can imagine that, once the embryo enters the uterus, it, for example, sends a message out that says to the mother’s uterus and her immune system, “hi, I’m here.” One can further assume that this message may be followed by a message that politely asks to be let in.
These are reasonable assumptions because the female uterus is in principle radically hostile to potential invaders. Were that not the case, women would be constantly infected by bacteria, viruses and even parasites. And, since we are already talking about parasites, this blastocyst-stage embryo that is now asking to be allowed to enter through a usually hermetically locked door, very obviously, must have a clue how to unlock this door without at the same time harming the well-being of the host. The reason is very clear: Like every other parasite (and as the name says), once the embryo is allowed to enter the mother’s organism, it becomes almost completely dependent on the host's well-being. For example, were the embryo dependent on weakening the maternal immune system in order to implant, implantation would endanger the mother’s life and, therefore, also that little parasite’s future as a parasite within the mother. Embryos, therefore, must have a way to unlock the gate without harming the effectiveness of the mother’s ability to continue refusing entry through the uterus to infectious agents and other parasites.
Increasing evidence now suggests that embryos do all of this in very similar ways to biological parasites (for example, helminths) by inducing the so-called tolerance pathways in the maternal immune system. Interestingly, this is also what many cancers are doing when they metastasize. Like parasites, in all kinds of tricky ways, they succeed in circumventing the mother’s immune system and, thereby, protect themselves from being attacked, as a normally functioning immune system usually does.
The reason why we are mentioning all of these seemingly unrelated facts in discussing implantation failure, is to demonstrate how complex the implantation process really is. And, by the way, we so far have addressed only one of the relevant players, the embryo. Then there is, of course, also the endometrium that, as recently has been suggested, “senses” the quality of embryos that are trying to implant and, literally, reaches out to normal good quality embryos and encases them in preparation for implantation, though does not do so for poor quality embryos. And, finally, as maybe the most important player, there is, of course, the previously mentioned maternal immune system, which must be viewed as the conductor of the orchestra because the immune system is in charge of determining which potential invaders into our bodies must be fought and rejected. To achieve successful implantation, all three of these players must interact in harmony.
Considering the very obvious complexities involved in this implantation process of embryos, it seems almost surprising that all goes well so frequently, and embryos are allowed to implant. One, however, also must consider the likelihood that failure may, actually, be more frequent than is generally appreciated. After all, even young couples at peak fertility, who likely produce an embryo every month, on average require between three and four months to conceive. From this observation alone, one can conclude that a large number of human embryos, likely even a majority, do not pass muster, and are not welcomed to implant.
Most of those are likely poor-quality embryos and we should be grateful they are not permitted to implant because this feature of the implantation process protects mankind from many abnormal pregnancies and potential births. It also appears reasonable to assume that not all failures to implant are caused by abnormal embryos. Clearly, there also must be malfunctions in the above-outlined normal implantation processes: The embryo may send no or the wrong signals upon entering the uterus; the endometrium and/or the maternal immune system may not properly register the embryo’s signals and/or respond in improper ways. For example, it has been demonstrated that hyperactive immune systems (i.e., immune systems affected by autoimmunity, inflammation, allergies) lose capacity to induce tolerance pathways.
If in response to signals from the embryo the maternal immune system cannot induce tolerance pathways, the immune system will still perceive the implanting embryo as “foreign” (which it is because it is 50% genetically paternal) and will, therefore, as it is expected to do, attack in order to protect the woman from dangerous invaders. The consequence will be more resistance to implantation, early pregnancy loss (i.e., chemical pregnancies) and increased numbers of miscarriages in women with hyperactive immune systems.
What does all of this mean clinically for women who suffer from difficulties with implantation, early pregnancy loss and repeat miscarriages? Unfortunately, not very much because, though we believe to understand many of the basic principles of the implantation process, the crux, as always, lies in the details, and most details affecting the implantation process of human embryos are still a big black box. For example, we still cannot differentiate between causes why embryos do not implant (and/or miscarry). In other words, we never know for certain, why an IVF cycle failed: was it because embryos were abnormal, because the endometrium did not behave appropriately in welcoming the embryo or was it because the immune system sent out missiles that killed off implanting embryos or the early pregnancies?
Consequently, when colleagues publish studies of women with alleged implantation failure, what does this really mean but a basket of possible conditions which, each, on their own or in combination, can prevent the establishment of successful pregnancy? Even laypeople will understand that appropriate treatments will depend on the underlying problem. Indeed, if the reason why an IVF cycle fails are poor quality embryos, who would even want these embryos to implant? The solution, in such cases, therefore, is improved ability to identify “bad” embryo prior to embryo transfer. That has been attempted but has, largely failed so far, with preimplantation genetic testing for aneuploidy (PGT-A) and embryoscope with time-lapse imaging being only the latest examples.
What reflects proper responses of endometrium to an implanting embryos is also still unknown. Laboratories at Cambridge University, in the UK, and here in NYC at Rockefeller University (the Brivanlou Laboratory, with which CHR closely collaborates) have in recent years been successful in establishing in vitro human embryo implantation models. They are, however, limited by current international restrictions that allow human embryos to be cultured only for up to a limited number of days post-fertilization. Remarkably, these studies, however, demonstrated that embryos apparently can develop perfectly normal up to approximately day 14 without any maternal contributions.
This observation, alone, demonstrates the amazingly pre-programmed, completely independent developmental abilities of human embryos. To some degrees, this should not come as a surprise since ectopically implanting human embryos, quite obviously, can implant even in absence of endometrium (as in vitro they have been shown able to implant on basically a plastic membrane). In vivo, ectopic pregnancies, however, still have to deal with the maternal immune system, which so-far has not been added to reported in vitro implantation studies of human embryos.
These finding call into question the long-held concept of a hormonal implantation window, which assumes that the endometrium produces under hormonal influences a very narrowly restricted time window for implantation. If there, indeed, exists such a hormonal time window, it does not appear dependent on endometrial tissue! One can also argue that a hypothetical in vitro implantation model that also includes components of the maternal immune system, capable of producing an allogeneic immune response, might demonstrate significantly more restricted implantation potential than currently established in vitro systems that lack immune responses.
Successful implantation in humans, thus, in many ways almost appears to be an accident, more often destined to fail than to succeed. How can one, therefore, be surprised by the fact that IVF much more often fails than succeeds (except, maybe, in young donor egg cycles). That implantation chances are so strongly associated with female age (i.e., the age of the eggs in a woman’s ovaries), speaks to the overwhelming importance of oocytes for successful implantation and pregnancy, and the relative unimportance of everything else. But being relatively unimportant does not mean not having any relevance. Unquestionably, endometrium and the maternal immune system actively participate in the implantation process but how they exactly do that, has remained largely unknown.
We, therefore, have to acknowledge that when the talk turns to an alleged implantation problem or so-called “implantation failure,” nobody really knows what it is all about. Consequently, treatments proposed and applied must be viewed with caution and be acknowledged by patients as well as physicians as largely experimental. Like all physiological processes, implantation, unquestionably, can malfunction, and likely in many different ways. But we currently do not even know what percentage of failed IVF cycles is caused by malfunction of the implantation process rather than by poor embryo quality. Common wisdom in the field holds that true implantation failure is a rare occurrence; whether that means one percent or 15 percent, nobody really knows. And in absence of specific diagnostic tests for implantation failure, to find successful treatments will be exceedingly difficult since effectiveness of treatments can only be assessed in “clean” patient populations, uncontaminated by other conditions.
This is a part of the September 2018 CHR VOICE.
Norbert Gleicher, MD, leads CHR’s clinical and research efforts as Medical Director and Chief Scientist. A world-renowned specialist in reproductive endocrinology, Dr. Gleicher has published hundreds of peer-reviewed papers and lectured globally while keeping an active clinical career focused on ovarian aging, immunological issues and other difficult cases of infertility.
Follow on LinkedIn
Watch his videos on YouTube
DISCOVER YOUR TREATMENT OPTIONS