30 Years of IVF: What Does the Future Hold?

‘Although IVF has provided the opportunity for couples to have children in circumstances where previous conception was highly unlikely or impossible, it has created … complications related to multiple births.’

The publication of the first successful human birth after IVF-embryo transfer (IVF-ET) in 1978 revolutionized the infertility industry [1]. The original technique involved oocyte retrieval in a natural ovulatory cycle. The tools for retrieval and the culture medium for embryos were rather primitive compared with those used today, hence, initial pregnancy rates were low. In addition, despite careful monitoring, there were often times when the one egg could be released without being ‘caught’. This combination of factors resulted in low pregnancy rates [2,3].

The pervading thoughts at this time were that in a given cohort of antral follicles, some of the oocytes would be genetically normal whilst some would have meiosis errors, and it would be merely fortuitous if the follicle containing a genetically normal egg became dominant. Thus, stimulating multiple follicles with clomiphene citrate alone or with clomiphene and human menopausal gonadotropin, became the usual method for preparing oocyte retrieval [4].

An even greater recruitment of follicles could be achieved by the use of gonadotropin stimulation instead of clomiphene. This would not only allow more fresh embryos to be transfered, but also produce extra embryos that could be cryopreserved for future use [5]. Thus, one could try to maximize the success from one expensive oocyte harvest (cost of medications, monitoring, oocyte retrieval and anesthesia) by performing a much less expensive frozen-embryo transfer in case of failure to achieve a pregnancy with the fresh-embryo transfer [6].

Unfortunately, in approximately 20% of the cycles, the high serum estradiol levels generated by the expensive gonadotropins caused premature luteinization and, thus, cancellation of the oocyte retrieval. This led to the introduction of the additional use of gonadotropin-releasing hormone (GnRH) agonists, either short term from the early follicular phase or longer protocols starting in mid-luteal phase, in order to suppress the spontaneous rise in luteinizing hormone (LH) [7,8].

Protocols using GnRH agonists, such as, buserelin or leuprolide acetate, dominated during the 1990s. GnRH antagonists, for example, cetrorelix or ganirelix, used to inhibit the LH surge, were introduced into controlled ovarian hyperstimulation protocols during the year 2000 [9]. They have the advantage of not requiring two menstrual cycles to perform IVF, as with GnRH agonists, and require a shorter duration of use, initiated in the late follicular phase.

Since IVF-ET is very expensive, couples want to maximize their success in the first IVF-ET treatment. Success rates of IVF centers are published in national registries and, therefore, commercial competition amongst IVF centers would influence IVF centers to try to maximize their success following the first fresh-embryo transfer. Furthermore, although some IVF centers had similar success using frozen-embryo transfer to that of fresh-transfer cycles, many other IVF centers did not appear to do well with frozen-embryo transfer [10]. This led to patients and physicians both willing to transfer three or more embryos, which ultimately led to a high multiple birth rate.

‘Successful pregnancies have even been recorded in women during apparent menopause who were treated in a similar manner in order to achieve follicular maturation followed by IVF-ET.’

Conclusions were made in the late 1980s, which have been somewhat confirmed in recent years, that a diminished egg reserve, as evidenced by an elevated serum follicle-stimulating hormone (FSH) on day 3 or a diminished antral follicle count, would not only result in the retrieval of fewer eggs, but would also reduce the quality of these eggs [11–13]. One recent study concluded that if a woman's FSH on day 3 was greater than or equal to 15 milli international units (mIU)/ml, live deliveries at any age were extremely unlikely [13]. Thus, if an increase in FSH on day 3 was found (usually ≥12 mIU/ml), the couple would be advised of very low success rates with diminished egg reserve, and the advice would be to not proceed with IVF treatment using their own eggs, in order to theoretically save the couple money. The couple would be advised to try the donor oocyte program, which began in the late 1980s [14,15]. In fact, some women with normal baseline FSH on day 3 would be advised to consider donor oocytes earlier if the serum FSH on day 3 was normal, but the FSH increased above 10 mIU/ml on day 10 [16]. Counting the number of antral follicles became a popular method. Additional research funds were spent in order to find even more ways to detect a woman with a diminished egg reserve, for example, measuring inhibin B and the antimullerian hormone.

‘Pregnancy rates per blastocyst transfer can be misleading since they do not take into account the fact that some attempts for transferring blastocysts result in no embryos left to transfer…’

In my opinion, these aforementioned studies regarding ovarian egg reserve did not advance the field of IVF-ET any more than they set it back. So many couples have been denied IVF services or pushed into donor-egg programs mistakenly with the false belief that pregnancies with their own eggs were not possible. Some women had insurance coverage with their own eggs, yet they had to spend a great deal of money for donor eggs that may not have been necessary. It has been shown that women aged 39 years old or younger, whose ovarian egg reserve was so low that only one embryo was formed, had a clinical pregnancy rate per single-embryo transfer of approximately 40% as long as the embryo at day 3 had six-to-eight blastomeres (which occurred 65% of the time) [17]. Even women aged 40–42 years with marked diminished egg reserve had a live delivered pregnancy rate of 21.7% per transfer [18].

There is a technique that has been described and successfully used for over 20 years, in which ovulation can be restored in women seemingly experiencing overt menopause by restoring the sensitivity of gonadotropin-resistant follicles. This is achieved by lowering chronically elevated serum FSH levels, which is theoretically responsible for the downregulation of critical FSH receptors [19,20]. Successful pregnancies have even been recorded in women during apparent menopause who were treated in a similar manner in order to achieve follicular maturation followed by IVF-ET [21,22]. In addition, there is a case report regarding a 40-year-old woman with hypergonadotropic amenorrhea with estrogen deficiency, and a markedly elevated serum FSH of 124 mIU/ml who conceived and delivered a healthy baby with her own eggs, despite having failed four previous transfers of embryos derived from donor eggs [23]. One theoretical explanation for the disparity of conclusions from different IVF centers is that these women should not be treated with high-dose gonadotropin protocols but only minimal stimulation protocols [24]. This is possibly explained by the suggestion that the high serum FSH downregulates a key FSH receptor whose normal interaction results in a protein that is important for implantation [17].

Although IVF has provided the opportunity for couples to have children in circumstances where previous conception was highly unlikely or impossible, it has created some problems, in particular, complications related to multiple births [25–31]. Thus, strategies were considered to improve the efficiency of the IVF process that would allow very high pregnancy rates from the transfer of only a single embryo. With further knowledge of nutrient requirements at different stages, sequential embryo-culture media were made allowing a reasonable percentage of embryos to develop to the blastocyst stage. Theoretically, the technique would select only the heartiest embryos (the weaker ones would fail to cleave from day 3–5), would result in less uterine contractility and the embryos would be entering the uterus at the normal time [32,33].

Unfortunately, there is little evidence that blastocyst transfer rather than the transfer of a six-to-eight-cell embryo on day 3 or even four cells on day 2 provides any significant improvement in pregnancy rates, especially if one considers the pregnancy rate from a given oocyte ‘harvest’ [34]. Pregnancy rates per blastocyst transfer can be misleading since they do not take into account the fact that some attempts for transferring blastocysts result in no embryos left to transfer, whereas by day 3 there may be embryos present, which, if transferred, could have resulted in live pregnancies. Generally, there is also a bias for performing this procedure on younger women with better egg reserves. Interestingly, blastocyst transfer did not result in the desired delivery of singletons because cost and competition have resulted in few IVF centers willing to transfer only a single blastocyst [35].

Unfortunately, performing preimplantation genetic diagnosis markedly increases the cost of IVF without providing the ideal answer for markedly improving the chance of a single embryo to result in a live delivery. Its best benefit is for the detection of inherited single genetic disorders.

Expensive research continues to find ways to identify the best embryo. Some of these techniques involve evaluating the spent culture media to either identify immune factors that may be important for implantation, for example, soluble leukocyte antigen G, or other metabolic factors, by profiling the culture media with Raman spectroscopy [36,37].

‘Emphasis on single-embryo transfers will lower the costs of caring for women…’

Although there are now more countries and insurers willing to provide financial coverage for some limited number of IVF attempts than in the past, the majority of couples do not have financial coverage. Many who need the service simply cannot afford it. Additional procedures performed to help identify the best embryo markedly increase the cost of an already expensive procedure, and if they only provide slight improved pregnancy rates, the increase in cost is not acceptable. The demonstration of implantation rates following the transfer of six-to-eight-cell embryos of 38–42% despite diminished egg reserves suggests that maybe we have been going about this in the wrong direction. By starting gonadotropins later in the follicular phase and reducing the dosage, we may enable nature to select the best egg. Certainly it is possible that there are properties that allow the woman to select the best single egg of the cohort each cycle. Thus, perhaps the use of protocols using minimal gonadotropin stimulation could prove to be the most beneficial technique [1,24].

Minimal gonadotropin stimulation protocols not only reduce the cost of expensive medications, but should markedly reduce the laboratory costs owing to the use of fewer embryos [24]. In addition, the risk of ovarian hyperstimulation syndrome is markedly reduced. Emphasis on single-embryo transfers will lower the costs of caring for women and children produced from a multiple birth, and these reductions in costs could influence more financial coverage from insurers for IVF procedures.

Other cost-saving measures would be the elimination of the overuse of intracytoplasmic sperm injection, which, when not needed, could even reduce implantation rates [38]. Furthermore, emphasis should be placed on improving a given IVF center's ability to properly freeze extra embryos. My own bias is that the the programable freezer used in the popular technique first described by LaSalle et al. is not beneficial and successful techniques that avoid this instrument should be used instead [39,40].

I predict that there will be governmental or insurer-derived policies emphasizing only single-embryo transfer of embryos produced on day 3 with six-to-eight blastomeres or blastocysts, and only allowing two embryos that are produced on day 3 if there are only four or five blastomeres present on day 3 [17].