"Recurrent implantation failure - It's time to get personal"
Recurrent implantation failure (RIF) is a controversial and poorly understood diagnostic dilemma. It can be defined as a failed embryo implantation, where b-human chorionic gonadotropin is in the negative range after a certain number of embryo transfers. The term ''implantation failure'' should be reserved for this specific scenario rather than for early failed biochemical pregnancies that may have a different underlying etiology. Although the concept of recurrence clearly implies repeated failed embryo implantation, the precise number of failed transfers to be classified as RIF is not well established. It is apparent that a cleavage-stage embryo produced by a 45-year-old woman would have a drastically different chance of implantation compared with a top-quality blastocyst from a 32-year-old woman. Therefore, a simple number of failed transfer attempts, without incorporating individual patient and embryo characteristics, is likely to grossly exaggerate the prevalence of RIF, especially in older adults and patients with poor prognosis. This overestimation of the true prevalence of RIF has serious consequences that hamper future research efforts and, most importantly, result in patients and clinicians utilizing unproven and mostly unnecessary tests and interventions, of which some may be detrimental to the overall chance of pregnancy.
The basic question that must be asked is ''after how many transfers is a recurrent failure to implant more likely due to some unrecognized underlying pathology, rather than chance alone?'' In the issue of the journal Fertility and Sterility, Ata et al. pose this crucial question and endeavor to answer it on the basis of an embryo's euploidy status. They postulate that the diagnosis of RIF should be made once recurrent failure of embryos to implant is unlikely to be due to embryo aneuploidy. They use the concept of ''cumulative implantation probability,'' which is similar to the one we proposed recently, termed ''theoretical cumulative implantation rate''. The basic concept behind their model is rather elegant in its simplicity: if 50% of embryos at a certain female age are euploid and a euploid embryo has a 50% chance of implantation, then it follows that an untested embryo produced at this age has a 25% chance of implantation. The next step in their model is to estimate how many embryos with this probability of implantation would be required (cumulative implantation probability) to achieve a certain threshold, beyond which RIF can be diagnosed.
The investigators provided not only the theoretical justification for their model but also a user-friendly calculator, which estimates the cumulative implantation probability on the basis of 3 inputs: age of the patient, euploidy rate at this age, and the probability of implantation per euploid embryo. The output of this calculator, based on the underlying assumptions of the model, determines whether a patient should be diagnosed with RIF based on the predeterminant cumulative implantation probability threshold value, which can also be entered, if desired. The result is a simple and extremely valuable tool which has enormous potential to rationalize the definition and diagnosis of RIF and invigorate the entire spectrum of research and possible treatment of this common condition.
There is, however, one major shortcoming of this model that must be addressed before its widespread use can be recommended. It is obvious that the model's reliance on the assumed euploidy rate is problematic. It is well-recognized that the euploidy rate is mostly determined by the female's age, but it is not the only factor. There are recognized variations from patient to patient, from cycle to cycle in the same patient, and even between laboratories. There are also well-articulated concerns regarding reliance on genetic testing of embryos and their presumed euploidy, where it appears that the indiscriminate use of preimplantation genetic testing for aneuploidy (PGT-A) may be detrimental to the cumulative pregnancy rate per cycle started. Furthermore, although euploidy may be the main factor that determines the implantation potential of an embryo, it is not the only one. Other embryonic characteristics, such as its grade and thaw survival, may also be important. Endometrial factors and hormonal milieu cannot be ignored as they also play a crucial role in implantation.
It is apparent that for implantation to be successful, all stars need to align: an embryo must be at least potentially euploid and structurally capable of implantation, the endometrium must be receptive to such an embryo, there must be no interference with the initial communication between an embryo and the endometrium so that the process of implantation can proceed unhindered, and the hormonal intrauterine environment must be able to nourish an embryo during this process. Therefore, authors propose that a more versatile and realistic model dispense with reliance on the presumed euploidy rate, and instead be based purely on the chance of implantation per embryo, irrespective and not dependent on its euploidy status.
Reliance on predeterminant euploidy rate in the calculation of the cumulative implantation probability introduces an additional level of uncertainty that is neither necessary nor required. Clearly, the implantation rates of qualitatively different untested embryos, as well as the tested ones, will vary greatly, and the estimation of the cumulative implantation probability should be based on local outcomes, rather than published rates that may overestimate the chance of a positive result.
There is one last point that needs to be briefly discussed: the threshold at which the diagnosis of RIF should be entertained. The cumulative implantation probability at which RIF is diagnosed could be set in a variety of ways, including 80%, which is similar to the diagnosis of infertility itself, or by incorporating some form of statistical measure, such as the standard deviation. Authors believe that this issue is of great importance and may require an expert panel discussion or a workshop to reach a robust and clinically meaningful consensus.
The management of imprecisely defined RIF poses a major clinical challenge. Finally, new avenues toward standardized and personalized definitions are being explored, which will enhance clinical decision making, patient counseling, and future research in this area.
