Introduction

Established fertility-preservation methods for women recognised at this time include fertility-sparing surgery and surgical ovarian transposition, shielding to reduce radiation damage to reproductive organs, and embryo cryopreservation after ovarian stimulation with gonadotropins and in-vitro fertilization (IVF). Currently, all remaining options are still considered experimental.

Cryopreservation techniques play a central role in assisted reproduction. Since initial reports that sperm could survive after being cooled at low temperatures, methods of preserving spermatozoa developed rapidly after accidental discovery of the cryoprotective properties of glycerol on sperm cells. Research on cryopreservation of sperm could advance relatively fast owing to the small size of sperm cells and their high number available for experiments. Since the first reported live birth after insemination with spermatozoa that had been frozen and thawed, clinical treatments have developed fast. Further development of sperm cryopreservation techniques has made the introduction of fertility treatments possible with banked frozen and thawed sperm from donors, and the establishment of sperm banking as a method for preserving male fertility.

With the development of IVF, methods for cryopreservation of embryos were required, as IVF treatment often resulted in surplus embryos. The human embryo has a relatively high permeability to cryoprotectants, and it is notably more resistant to cryodamage than the oocyte. After the first reported human pregnancy after cryopreservation, and thaw of an eight-cell embryo in 1983, methods for cryopreservation of embryos developed successfully. It is estimated that more than half a million babies have been born worldwide after intrauterine replacement of frozen and thawed embryos.

Cryopreservation of oocytes

Freezing of oocytes for fertility preservation is an option for women who do not have a partner and for women who do not wish to preserve embryos because of religious or ethical concerns. Oocyte cryopreservation for fertility preservation may also be an option for adolescent girls in selected cases, as shown in a recent report of a 14-year old girl interested in undergoing hormonal stimulation and oocyte freezing instead of ovarian tissue cryopreservation, which is usually the preferred option for children.

A valid concern in carrying out ovarian stimulation for oocyte freezing in women with breast cancer and other hormone-sensitive tumours is the exposure to supraphysiologial oestradiol levels during gonadotropin stimulation, which has been a main obstacle in offering fertility preservation by oocyte or embryo cryopreservation to those women. Alternative and potentially safer stimulation protocols with tamoxifen, and aromatase inhibitors have been developed. Those stimulation treatments have been shown to be effective, with no detrimental effects on relapse or survival rates in these women.

Freezing unfertilised oocytes, with the aim of later thawing and fertilising them by IVF or intracytoplasmic sperm injection (ICSI) is today a promising option for preserving fertility. The road to its current success, however, was long and difficult. The human oocyte is the largest cell in the body, and its large size and large volume with a high content of water makes it extremely fragile to intracellular ice formation during the freezing and thawing processes. Oocytes, in contrast to embryos and zygotes, also present with lower permeability to cryoprotectants. Research on permeability coefficients of human oocytes collected from different women has also revealed significant differences among individuals regarding permeability to water and cryoprotectants. Similar findings have been reported in several species, including rodents and cattle. For that reason, finding appropriate cooling and warming rates to ensure oocyte survival after cryopreservation has not been an easy task, and has been a great challenge in assisted reproductive technologies.

Cryopreservation of oocytes

Freezing of oocytes for fertility preservation is an option for women who do not have a partner and for women who do not wish to preserve embryos because of religious or ethical concerns. Oocyte cryopreservation for fertility preservation may also be an option for adolescent girls in selected cases, as shown in a recent report of a 14-year old girl interested in undergoing hormonal stimulation and oocyte freezing instead of ovarian tissue cryopreservation, which is usually the preferred option for children.

A valid concern in carrying out ovarian stimulation for oocyte freezing in women with breast cancer and other hormone-sensitive tumours is the exposure to supraphysiologial oestradiol levels during gonadotropin stimulation, which has been a main obstacle in offering fertility preservation by oocyte or embryo cryopreservation to those women. Alternative and potentially safer stimulation protocols with tamoxifen, and aromatase inhibitors have been developed. Those stimulation treatments have been shown to be effective, with no detrimental effects on relapse or survival rates in these women.

Freezing unfertilised oocytes, with the aim of later thawing and fertilising them by IVF or intracytoplasmic sperm injection (ICSI) is today a promising option for preserving fertility. The road to its current success, however, was long and difficult. The human oocyte is the largest cell in the body, and its large size and large volume with a high content of water makes it extremely fragile to intracellular ice formation during the freezing and thawing processes. Oocytes, in contrast to embryos and zygotes, also present with lower permeability to cryoprotectants. Research on permeability coefficients of human oocytes collected from different women has also revealed significant differences among individuals regarding permeability to water and cryoprotectants. Similar findings have been reported in several species, including rodents and cattle. For that reason, finding appropriate cooling and warming rates to ensure oocyte survival after cryopreservation has not been an easy task, and has been a great challenge in assisted reproductive technologies.

Chilling injury and disruption of the meiotic spindle

Concern has also been raised about the disruption of the meiotic spindle observed when oocytes are being cooled to temperatures near 0 °C, which has been described as a sign of chilling injury. At low temperatures, tubulin depolymerises, and the support for the structure of the spindle, which maintains the chromosomes aligned on the metaphase plate, disassembles. The meiotic spindle is needed for the correct chromosome segregation during the oocyte maturational processes.

Variations in oocytes’s sensitivity to chilling injury have also been reported among individuals in human oocytes as well as in several animal species, including non-human primates.

Nevertheless, it has been shown that the spindle reappears in the cytoplasm after warming up the oocytes at 37 °C, and chromosomal aberrations do not seem to be increased in embryos obtained from frozen thawed oocytes compared with fresh embryos.

Increasing fertilisation rates by intra-cytoplasmic sperm injection

Over many years, the techniques for cryopreservation of oocytes were based on slow freezing protocols originally developed for embryo freezing, and success rates of oocyte survival and fertilisation were low. Since the first pregnancy reported in 1986, only a few live births have been achieved. Cryopreservation of oocytes induces hardening of the zona pellucida, which could only be overcome after the development of ICSI. In 1997, microinjection of a sperm in vitro was first used on oocytes that had been frozen and thawed, which resulted in a live birth, and over 100 deliveries were reported in the following decade by using ICSI.

As a result of legal restrictions on freezing embryos, research on cryopreservation of oocytes has been particularly successful in Italy, where reports of improvement of freezing protocols have abounded. Methods of evaluating oocyte quality have been further developed, and increasing survival rates of up to 75.9%, fertilisation rates of 76.2% after ICSI, and pregnancy rates of 21.3% per embryo transfer have been reported. The miscarriage rates reported in pregnancies obtained after fertilisation of frozen and thawed oocytes have also been similar to those found in control participants when embryos were obtained from fresh oocytes.

Oocyte freezing by vitrification

The process of vitrification involves the use of high concentrations of cryoprotectants and rapid cooling to achieve a glass-like highly viscous solution without formation of ice crystals. Given the high inter-individual variation in oocyte sensitivity to chilling, it has been suggested that rapid cooling might prevent such injury. In animal studies, cattle oocytes presented with increased survival and fertilisation rates when cooling them at high rates.

The first pregnancy after fertilisation of thawed vitrified oocytes was achieved in 1999, and few groups reported using this technique in the following years. Significant improvements in vitrification investigational protocols were reported in 2003, when Japanese researchers cryopreserved oocytes by vitrification in specially constructed devices aimed at cooling at the fastest rates, with the objective of avoiding problems with chilling and cytoskeleton damages. With those methods, a high survival rate after thawing and clinical pregnancies was obtained. The method of vitrification in those studies reached an oocyte survival rate of 91% after thawing and a 50% blastocyst stage development.

The efficiency of vitrification of oocytes has made possible the current establishment of oocyte banking in donor programmes. Survival rates of oocytes thawed after vitrification have been reported to be as high as 96.9%, and the fertilisation rates are currently non-significantly different from those of fresh oocytes (76.3% v 82.2%, respectively). Pregnancy rates have been reported to be as high as 65.2%, and implantation rates 40.8% with vitrified and thawed oocytes. In the largest, prospective, randomised-controlled study, including 600 recipients of donated eggs conducted in Spain, no superiority of using fresh oocytes over vitrified egg-banked and thawed oocytes could be demonstrated.

Commercially available vitrification kits use either open system containers with a direct contact with liquid nitrogen or closed containers with no direct contact. Open systems offer the possibility of more rapid cooling; however, those systems might present with a risk of contamination and transmission of diseases. New approaches for preventing the potential risk of cross-contamination are being investigated, such as avoiding direct contact with liquid nitrogen by long-term vapour-phase nitrogen storage, and ultraviolet radiation applied to the liquid nitrogen, with the aim of sterilising it when using open vitrification systems.

Cryopreservation of immature oocytes

Because immature germinal vesicle stage oocytes have not yet formed spindle, and because they have a higher membrane permeability, it has been suggested that germinal vesicle stage oocytes might be more resistant to chilling injury than mature metaphase II oocytes. Reported survival rates of immature oocytes cryopreserved with slow freezing protocols after thawing vary from 37% to 63%. Spindle abnormalities, however, have also been reported after in-vitro maturation, suggesting impaired developmental competence when oocytes are frozen at germinal vesicle stage.

Few reports have been published of live births after immature oocyte cryopreservation with subsequent in-vitro maturation and IVF.

Rates of fertilisation and cleavage in cryopreserved and in vitro-matured oocytes are still significantly lower compared with non-frozen control participants. Additionally, in-vitro maturation programmes have only developed in a few centres worldwide. For those reasons, current clinical treatments aim to obtain mature metaphase II oocytes for cryopreservation.

In a recent study, in-vitro maturation of immature collected oocytes after controlled ovarian hyperstimulation significantly increased the number of mature oocytes yielded and embryos obtained for cryopreservation in women with breast cancer (45% increase in mature oocyte yield with 86% fertilisation rate of in-vitro matured oocytes). This suggests that it might be a complementary strategy for the preservation of fertility.

Perinatal outcome after oocyte cryopreservation

Over 1000 pregnancies and infants born after fertilisation of frozen-thawn oocytes have been reported worldwide, and these figures are constantly increasing. Studies indicate that pregnancies and infants conceived after oocyte cryopreservation do not present with increased risk of adverse obstetric outcomes or congenital anomalies. Nevertheless, most of the oocytes have been frozen by slow freezing methods. And the use of vitrification is relatively recent in clinical programmes. Vitrification techniques, however, had previously been introduced for the cryopreservation of blastocysts in clinical IVF programmes. A Swedish report of neonatal outcomes of children born after transfer of vitrified blastocysts has presented with reassuring results, as no apparent negative effects in neonatal outcomes have been found when comparing children born after transfer of vitrified blastocysts to those born after slow-freezing or after the transfer of fresh blastocysts.

Although overall pregnancy rates are improving, freezing of oocytes is still considered experimental, and it is recommended that the method should be carried out as part of research projects under institutional review boards.

Cryopreservation and transplantation of ovarian tissue

Cryopreservation of ovarian tissue combined with orthotopic transplantation into an irradiated ovary restored fertility in rodents 50 years ago. It took more than 30 years, however, to improve the freezing protocols and transplantation procedures in larger animals and, in particular, in species that could be considered applicable to humans. In fact, the only available cryoprotectant during since the 1930s was glycerol. Although it was successful for freezing sperm, it was highly ineffective for cryopreserving oocytes and ovarian tissue. In the 1970s, more effective cryoprotectants emerged, such as propanediol, ethylene glycol and dimethylsulphoxide (DMSO).

The exhaustive studies of ovarian cryopreservation and transplantation conducted in animals, and particularly those in the lamb model, finally led to significant improvements in cryopreservation and transplantation techniques, and the establishment of a model for applying these investigational procedures to humans. In this model, the ovarian tissue was cryopreserved after improving current slow freezing protocols using DMSO as a cryoprotectant. The ovarian tissue was frozen in slices and the thawed slices were later transplanted orthotopically. In the lamb, recovery of cyclic ovarian activity was demonstrated, and pregnancies and live born lambs achieved, as well as long-term ovarian function. Since then, ovarian tissue has been cryopreserved from several species, with the aim of improving the methods, optimising existing freezing protocols, and developing new techniques, such as vitrification. The first live birth in non-human primates after ovarian tissue transplantation was reported in 2004.

Human ovarian cryopreservation

A method for cryopreservation of human ovarian tissue was first reported in 1996. Cryopreservation of ovarian tissue for fertility preservation is indicated for adult women when ovarian stimulation for oocyte or embryo freezing is not feasible or not desired. For prepubertal girls, the cryopreservation of ovarian tissue is their only option to spare their eggs.

As most eggs are within primordial follicles in the ovarian cortex, a significant number of eggs may be preserved by freezing pieces of ovarian cortex, even if they are small.

Ovarian biopsies may be harvested by minimal invasive surgery, such as laparoscopy. Ideally, this procedure should be carried out before starting cytotoxic treatment. Nevertheless, it may still be worth carrying out ovarian cortex harvesting after cytotoxic treatments in girls and young women previously unable to undergo the procedure, because girls and young women normally have high follicle counts in their ovaries. Consensus is lacking about how much ovarian tissue should be harvested for cryopreservation, and some programmes for fertility preservation propose unilateral oophorectomy as standard procedure.

Several cryoprotectants for slow freezing of human ovarian tissue have been investigated, such as glycerol, ethylene glycol, DMSO and propanediol. Follicle survival has been evaluated after thawing pieces of frozen ovarian cortex with those cryoprotectants. The highest follicle survival rates have been obtained with ethylene glycol, whereas glycerol has been associated with the poorest results.

Studies investigating the most favourable cooling rates and dehydration times have also been conducted. After evaluating the tissue by electron microscopic morphology, slow programmed freezing with a relatively long dehydration time became a standard cryopreservation method for human ovarian tissue from the end of the 1990s.

Vitrification of ovarian tissue

A concern of slow-programmed freezing has been the relatively poor survival of the ovarian stroma, which has been demonstrated by transmission-electron microscopy, a method that accurately evaluates cryoinjury of membranes, mitochondria and other organelles not seen in detail with any other methods. The introduction of vitrification techniques for cryopreservation of ovarian tissue has been shown to improve the viability of all compartments of the tissue, with a survival rate of follicles similar to that after slow freezing, much better integrity of ovarian stroma and undamaged morphology of blood vessels.

Vitrification of ovarian tissue used in combination with orthotopic autotransplantation has been successful in rodents and sheep. Human studies comparing slow-freezing protocols with vitrification of ovarian tissue have produced conflicting results, which may be explained by differences in the protocols and the medium used.

As suggested by transmission-electron microscopy, vitrification could be more effective than slow-programmed freezing when cryopreserving ovarian tissue. In one study, thawing of vitrified ovarian tissue resulted in improved oocyte survival compared with slow freezing (89% v 42%, respectively). The methods for vitrification have further developed, achieving a clinical grade closed cGMP-compatible system, in which the ovarian tissue sample will not be in any direct contact with liquid nitrogen during vitrification or storage, thus meeting the requirements of international tissue directives.

Whole-organ freezing and transplantation

Whole-ovary freezing has also been investigated, with the aim of achieving vascular anastomoses and a functioning organ after transplantation. The method has been shown to restore fertility in rats and sheep, but a high rate of follicle loss is still a concern. Directional freezing combined with microvascular anastomosis have improved outcomes, and long lasting ovarian function has thus been obtained in sheep.

In humans, research of whole-ovary freezing and transplantation is still at the initial stages. Transplantation of the whole fresh ovary has been reported between monozygotic twins when premature ovarian failure has occurred in the recipient twin.

Use of experimental human ovarian transplantation in the xeno-transplantation model

As the investigation of transplantation of human ovarian tissue in humans is not feasible, the xeno-transplantation of human ovarian tissue into immune-incompetent severe combined immunodeficiency (SCID) mice was initially proposed. The SCID mouse presents with T- and B-cell immune deficiency owing to a genetic mutation, and it has been used for the investigation of various xenographs, as they revascularise and survive without being rejected. The experimental model of human ovarian transplant on SCID mice has also allowed the investigation of transplant survival, xenograph’s hormone production and the induction of follicle development experimentally by gonadotropin stimulation. In the xeno-transplant model, the thawed tissue transplanted under the kidney capsule of the SCID mouse has also allowed the efficacy of various freezing protocols and cryoprotectants to be compared and optimal size of cortical pieces for freezing to be investigated. Xenograph models have, in addition, been used to investigate the histological and molecular mechanism of the effect of chemotherapy agents in ovarian follicle loss after administration of chemotherapy agents in similar doses to those used for clinical treatments.

Studies in the xenograph model aimed at improving ovarian transplant survival

It has been shown that the survival of an ovarian transplant is greatly dependent on the revascularisation process, as the ischaemic phase that follows immediately after transplanting is associated with an initial massive follicle loss. In the SCID mouse model, experiments allow the investigation of pharmacologically active substances in xenographs. Antiapoptotic substances, such as sphingosine-1-phosphate, have previously been investigated in rodents, resulting in inhibition of oocyte apoptosis induced by cytotoxic treatments. Recently, in the xenograph model, spingosine-1-phosphate has been shown to increase vascular density after transplantation in the human ovarian graft, accelerated neo-angiogenesis and reduced necrosis and tissue hypoxia, resulting in significant improvement of oocyte survival compared with vehicle-treated controls.

Transplantation of cryopreserved ovarian tissue for recovering ovarian function and fertility

The successful development of cryopreservation techniques for human ovarian tissue stimulated the initiation of ovarian tissue freezing as an option to preserve female fertility from the late 1990s.

In the year 2000, the first case of auto-transplantation of cryopreserved and thawed ovarian tissue by laparoscopy in a woman was reported. She had undergone oophorectomy at age 28 years owing to intractable dysfunctional bleeding, and her ovarian tissue was frozen by using a slow-freezing protocol with propanediol as cryoprotectant. Because of serious menopausal symptoms, the woman requested autotransplantation of the ovarian tissue, and the procedure was carried out 8 months after cryostorage. The histological analysis of one piece of cortex revealed normal stromal cellularity and presence of follicles with one or two layers of granulosa cells. Six pieces of cortex were cultured in vitro for 6 days in the presence of gonadotropins, and increased amounts of oestradiol, progesterone and testosterone could be measured. After investigating the quality of the frozen and thawed cortical tissue, 80 pieces of cortex measuring 2 by 2 mm to 5 by 10 mm were thawed using stepwise cryoprotectant dilution, as previously described.

The ovarian cortical tissue was grafted by laparoscopy. After thawing, the cortical pieces were strung with a 6-0 delayed absorbable suture to two absorbable cellulose membrane frames and loaded retrogradely into a 13-mm trocar ( Fig. 1 ). The graft was thereafter placed in a pocket created in the pelvic sidewall (ovarian fossa) posterior to the broad ligament ( Fig. 2 ) flattened and sutured ( Fig. 3 ) and covered by peritoneum ( Fig. 4 ). The grafts could be seen immediately after the surgery, and blood flow to the transplant was confirmed by Doppler ultrasonography.

Retrograde loading of the graft that was reconstructed by stringing the ovarian tissue between two strips of Surgicel ^® . Published with permission.

Placement of the leading suture in the pelvic pocket and through the lower peritoneal edge. Published with permission.

Placement of the base suture through the upper peritoneal edge. By pulling on this suture, the graft is flattened against the vascular pelvic wall. Published with permission.

Closure of peritoneum with interrupted sutures. Note the placement of two grafts side by side. Published with permission.

Hormone stimulation with daily administration of gonadotropins 15 weeks after transplantation resulted in follicle development and ovulation, which was corroborated by increases of oestradiol and subsequently progesterone. A corpus luteum in the transplant confirmed ovulation by ultrasonography. The endometrium also presented follicular and luteal phase changes, and the woman menstruated 2 weeks after ovulation.

Since 2004, several centres worldwide have reported cases of orthotopic transplantation of cryopreserved ovarian tissue in cancer survivors, resulting in live births. A recent review of the first 10 cases of successful orthotopic transplantation, which has resulted in the birth of 13 children, indicates that age at cryopreservation may be an important prognostic factor, as all except one woman in that report were younger than 30 years, and six of the women were younger than 25 years. As noted by the American Society of Clinical Oncology recommendations on fertility preservation, the benefit of ovarian cryopreservation for women older than 40 years of age is uncertain because of their age-reduced ovarian reserve of primordial follicles.

In reports of successful ovarian transplantation, restoration of ovarian function has been shown by the rise of oestrogen levels and decrease of gonadotropin levels between 3.5 and 6.5 months after ovarian grafting. Ovarian transplants in those cases have been shown to become functional after grafting in a peritoneal window close to the ovarian hilus and on the ovarian medulla, as proved by follicle development. In all cases, slow-programmed freezing was used to cryopreserve the ovarian tissue, and both large strips of 8–10 mm × 5 mm or small pieces of 2 × 2 mm of tissue were shown to restore ovarian endocrine function effectively. More than 50% of the women in that report conceived naturally, which argues in favour of orthotopic transplantation; only a few women required IVF to become pregnant.

A recent report of preparation of thin ovarian fragments for transplantation (less than 350 m thickness) led to a successful pregnancy after failure of ovarian strip transplantation on two previous occasions.