Investigation and Clinical Approach of Recurrent Implantation Failure

Recurrent implantation failure (RIF) derives from the practice of IVF. The ability to identify implantation failure after the transfer of embryos raised the possibility that there exists a patho-physiological state leading to repeated implantation failure. RIF can be defined as the repeated lack of implantation after the transfer of embryos. Prior to IVF, treatment resulted either in a pregnancy or not. IVF added the ability to further compartmentalize the treatment process so that it became possible to know when an embryo was transferred and if an implantation occurred. Hence RIF became a possible clinically identifiable phenomenon. The relevance of defining such a problem lies in the ability to diagnose patho-physiological conditions, which might be amenable to treatment and thus increase the implantation rate thereby increasing a couple’s chance of conception.

Many practical procedural problems exist when trying to define implantation failure. One limitation when defining implantation and implantation failure is that at present, implantation can only be determined by a rising quantitative hCG level which occurs a number of days after implantation has actually occurred or failed. There remains a time-frame which cannot be further compartmentalized and thus limits the ability to precisely define the cause of implantation failure. Thus, for practical purposes, implantation is defined as a rise in the quantitative hGC level determined at some point after embryo transfer. The most accurate way to determine this would be to do daily serum HCG levels starting after embryo transfer. But again practicality dictates that the first level is drawn at some point after embryo transfer and for many programs this is usually 10–14 days after hCG is given.

The timing of when the first hCG determination is performed will affect the implantation rate. The earlier the first hCG determination is done, the lower the failed implantation rate will be since some pregnancy losses will occur prior to the hCG determination for those programs measuring it later in the cycle. Any definition of RIF requires a standardization of the time from embryo transfer to the time used to determine a positive hCG level. The definition also needs to establish what will be considered a pregnancy when using hCG. For example, will any value above the threshold value be considered evidence of implantation or will there need to be rising hCG levels? Finally, the day of embryo transfer needs to be established. For example, suppose there was a preventable cause of failed implantation that occurred at the morula stage. Programs transferring day three embryos would have a higher failed implantation rate than those programs transferring blastocyst embryos.

The accuracy of a clinical definition is to a large extent determined by the simplicity or, conversely the complexity, of the problem. A physiologic process that has a single factor is easy to define in its entirety. A pathological process that alters the physiology of the process in question may be caused by a single factor and, if measurable, the definition of the disease process is simple and accurate. The thesis of this manuscript has two prongs. First, implantation is extremely complex and highly redundant. Continuation of the species is evolutionarily a high priority. Therefore, the evolutionary process has preserved the ability to implant through accruing a number of individual processes, many of which are, in and of themselves, not necessary, but collectively contribute to a maximum chance for implantation. Second, because of the complexity and redundancy of implantation, assigning etiologies for implantation failure and RIF will be difficult.

There are some processes involved in implantation which are simple and absolutely required for implantation to occur. But there are many processes which are not necessary but contribute to implantation. Therefore, the problem with implantation failure as well as RIF is not in defining the problem, which is self-evident, but in assigning causality to the phenomenon. Some causes of IF and RIF will be a single factor. Their identification and their contribution will be easier to discern. However, implantation failure and RIF may be a result of a threshold phenomenon whereby there are a multitude of factors required for maximum implantation efficiency. A failure of any one of these factors might, or might not, reduce the overall implantation rate. But to reach a state whereby implantation failures, or implantation fails repeatedly, might require errors in a number of factors and might not be the same factors for each case of IF or RIF. The identification of this problem becomes complex and very difficult to approach experimentally.

Any deviation from or interruption of the normal structure or function of any part, organ, or system, or combination thereof, of the body that is manifested by a characteristic set of symptoms or signs, and whose etiology, pathology, and prognosis maybe known or unknown. Implantation failure and especially to recurrent implantation failure demonstrates just how difficult it becomes to define a disease. There are two ways of approaching a definition for a normal implantation rate: evaluate spontaneous pregnancy rates in fertile couples, or evaluate implantation rates when using donor oocytes. There are many factors intervening when couples have intercourse which could result in a less than perfect system so that more than 50% of the cycles do not result in a pregnancy. Therefore, using the natural pregnancy rate in maximally fertile women does not provide the best definition of a normal implantation rate.

IVF provides a more definable system since it involves oocyte, sperm and, endometrium. For this more defined system, the relevant statistic is not normal implantation rate but rather, the maximum implantation rate. Any deviation from this maximum implantation rate would represent a compromise that is, by definition, a disease. Identifying the maximum implantation rate requires evaluating implantation rates obtained through the use of normal oocytes, sperm, and endometrium when the IVF procedure is done by the technically most adept IVF units. This condition is frequently met for women using donor oocytes for age-related ovarian failure. The implantation rate for the double embryo transfer patients was equally high at 56%. It would appear that the maximum implantation rate for IVF is between 40% and 60%. For IVF programs with substantially lower implantation rates, the problem may be technical related to either the clinical or laboratory components of the program. Until a program’s implantation rates meet industry standards, evaluating the patient for problems with implantation failure will be premature.

Defining the term “recurrent” is very challenging, analogous to defining infertility or recurrent pregnancy loss. A problem when using pregnancy rates based upon number of transfer cycles relates to the variability of the number of embryos transferred on any given cycle, the quality of the embryos transferred, and the day of embryo transfer. Defining recurrent implantation failure as three unsuccessful IVF attempts actually defines RIF in terms of failed IVF cycles and does not address the issue of implantation rates. Some patients may have a large number of poor quality embryos transferred for a given IVF cycle; whereas, others may have only one or two embryos transferred. Thus, defining RIF in terms of pregnancy rate does not allow an evaluation of the implantation rates. Also, the day of transfer greatly impacts the pregnancy rate since blastocyst embryos have a higher implantation rate than day three embryos.

Therefore, time to pregnancy should be shorter for cycles with blastocyst transfers and especially if many blastocysts are transferred on a given cycle. For example, transferring three blastocyst embryos should result in a pregnancy sooner than transferring only a single blastocyst embryo. The definition of RIF would be fewer transfer cycles for multiple optimal embryos transferred per cycle than for either poorer quality embryos, high quality embryos transferred on day three, or fewer blastocyst embryos per cycle. Age also has an impact on the pregnancy rate and is a known factor for implantation failure. Thus, defining recurrent differs depending upon the age of the population studied. Finally, the choice of the number “three” deserves some attention. For three to be the operational number of IVF cycles, there needs to be some correlation with that number and the incidence of implantation failure. For example, the cycle fecundity needs to decline after three cycles of IVF if RIF is defined by the number of cycles for transfer. Where there is no implantation failure, the cycle fecundity rate remains constant. The cycle fecundity rates of 36.2% for each of the first four cycles of donor oocyte embryo transfers.

However, cumulative pregnancy rates for IVF have demonstrated a decline in cycle fecundity with successive failed IVF cycles. A clinical pregnancy rate of 19.0% for the first IVF cycle, 17.4 for the second, 11.8 for the third, 10.8 for the fourth, and less than 6% for the next three successive cycles of IVF. Defining RIF after three cycles of IVF with this data is problematic given the fact that is seems the number should be either two or four cycles of IVF. Furthermore, notes an even more dramatic decline in pregnancy rates with successive IVF cycles where the embryos transferred were blastocyst embryos. The cycle fecundity rates were 36% for the first transfer, 19% for the second, and 9% for the third. Given this data, RIF may need to be defined as failed two or even one cycle of transfers with blastocyst embryos.

A second method of defining RIF uses the number of embryos transferred. All consecutive pregnancies from a large IVF program were analyzed to determine the total number of embryos that had been transferred to achieve each pregnancy. The number to achieve 80% of the pregnancies was 8 and for 95% of the pregnancies it was 12. Arbitrarily defining normal at the 95 percentile, RIF was defined as having failed to achieve a viable pregnancy if more than 12 embryos had been transferred. The patients participating in this study were all infertility patients and, thus, the pregnancy rates were not maximal as they might have been had donor oocyte recipient patients been used. The pregnancy rates and implantation rates were below what they are today, and these pregnancies were achieved using day three transfers, not blastocyst transfers. Therefore, the number used to define RIF would be lower today and depend upon whether day three or blastocyst transfers were being considered.

Any definition of RIF will be found wanting given the complexity and redundancy of the implantation process. However, a working definition of RIF can be based upon the number of embryos transferred. The definition will depend upon the stage of embryo development. The actual number will vary by center and can best be calculated by using pregnancies delivered in patients using donor oocytes. The definition of RIF used in our program requires the transfer of =8, 8-cell stage embryos or =5 blastocyst embryos. Evidence of implantation is accepted when there is a rise in hCG for hCG measurements made 12 and 14 days after the hCG injection that is given to trigger the time of retrieval.