The first live birth from a cryopreserved oocyte was reported by Chen [17] in 1986. Due to improved freezing and thawing techniques, the pregnancy rate with cryopreserved oocytes has improved considerably. Recently, high survival rates (≥90%) and pregnancy rates (40% to 55%) have been reported in egg donation programs [18]. A number of studies have reported no differences in the pregnancy rate or implantation rate of embryos obtained from cryopreserved mature oocytes compared with those from fresh oocytes, and further, no noticeable differences in chromosomal abnormalities have been reported [192021]. At the present time, mature oocyte cryopreservation is a preferred fertility preservation method if the individual is postpubertal and if chemotherapy or radiation treatment can be delayed to allow time for the full procedure.

Aspiration of immature oocytes followed by IVM is an option for patients who are unable to undergo ovarian stimulation, including prepubertal girls, women with hormone-sensitive cancers, and patients with polycystic ovarian syndrome (PCOS) who cannot undergo stimulation due to ovarian hyperstimulation syndrome. Aspiration of immature oocytes offers the advantage of allowing immediate cancer treatment, since COS is not required. In addition, it is thought that less damage is caused by cryopreservation of immature oocytes than mature oocytes, as immature oocytes lack the spindle and nuclear membrane that protect the chromatin. However, studies have reported that cryopreserved mature oocytes achieved better outcomes compared to IVM of cryopreserved immature oocytes [22]. For this reason, immature oocytes are often matured by IVM and then cryopreserved as mature oocytes [2223]. There have been only a few reports of successful birth using cryopreserved oocytes or embryos from immature oocytes that were matured in vitro [24]. So far, this technique has mainly been used in PCOS patients, and data on efficacy in cancer patients are not available.

In a conventional COS protocol for embryo or oocyte cryopreservation, cycles require up to 5 weeks, since ovarian stimulation must be started on day 2 to 3 of the menstrual cycle (Fig. 2A). In many cases, this time frame is appropriate, since many patients undergo tumor resection prior to chemo- or radiation therapy. That said, some patients need to start cancer treatment sooner and therefore random start COS protocols have been developed. In this procedure, ovarian stimulation is initiated irrespective of the phase of the cycle [252627]. This random start protocol is based on the recent concept that there are multiple waves of follicle recruitment within a single interovulatory period [28]. Moreover, this approach is possible because the purpose of the protocol is to retrieve mature eggs from the ovary and does not require the full maturation of the uterus. Several case reports and comparative studies have shown that the number of total and mature oocytes retrieved, oocyte maturity rates, and fertilization rates were not compromised in random-start COS cycles compared to conventional cycles [2930]. This protocol minimizes delays in treatment and enables more patients to pursue fertility preservation with egg or embryo cryopreservation prior to chemo- or radiation therapy.

For women who have hormone-sensitive cancers, such as breast cancer, mature oocyte retrieval can be attempted during a natural cycle to avoid exposure to supraphysiological levels of estrogen during COS. However, the number of oocytes collected is reported to be low. Recently, daily letrozole administration during COS has been used to prevent high levels of circulating estradiol in patients with estrogen-sensitive tumors. A commonly used protocol involves the use of 5 mg of letrozole from the 2nd to 3rd day of the cycle, followed by initiation of gonadotropins. The use of letrozole did not affect the number of oocytes retrieved, but a lower oocyte maturity rate in letrozole cycles was reported in some studies [3031]. This protocol is used only to obtain oocytes for cryopreservation, after which patients can then proceed to their planned cancer treatment. Some studies have examined the use of letrozole immediately prior to embryo transfer; in a study of 911 infants, there was no difference in the incidence of birth defects in children when letrozole versus clomiphene was used during ovulation stimulation [32]. Thus far, prospective data regarding the risk of breast cancer recurrence in patients undergoing COS with the letrozole protocol are reassuring when compared to the risk of recurrence in women with breast cancer who did not undergo COS [33].

In 2004, Donnez et al. [37] reported the first successful pregnancy and birth from transplantation of strips of cryopreserved ovarian cortex (Fig. 2C). To date, at least 60 live births from either natural conception or IVF have been achieved after re-implantation of cryopreserve ovarian tissue [38]. Women who receive an ovarian tissue transplant are generally poor responders to COS, since of the 500 to 1,000 primordial follicles that are present in the transplanted ovarian tissue, more than 50% are lost [37]. Restoration of ovarian activity is higher if primordial follicles are present in the transplanted tissue and the chance of live birth after replacement of ovarian tissue is roughly 20% [39]. Recently, the first case was reported of successful fertility restoration after transplantation of ovarian tissue cryopreserved before menarche [40]. According to studies published to date, it takes approximately 4 to 6 months after transplantation of ovarian cortex tissue for follicular development to resume, with a rise in estradiol [41]. This is consistent with the 4- to 6-month period required for a primordial follicle to progress to the large antral follicle stage with a fully mature oocyte. Follicle-stimulating hormone values during the follicular phase were relatively higher after transplantation, which can be explained by the small number of primordial follicles that survived in the transplanted tissues. Following transplantation, ovarian function continues for approximately 5 years on average [3839].

Theoretically, whole ovary transplantation with vascular anastomosis is the most ideal method for minimizing duration of ischemia from transplantation to angiogenesis because it allows immediate blood circulation after the transplantation. However, adequate dispersion of cryoprotectant throughout the ovary is difficult in large mammals or humans due to the size of the tissue. Vascular damage from formation of ice crystals in the blood vessels can also occur. Recently, Martinez-Madrid et al. [42] proposed a method involving cryopreservation of the whole ovary together with blood vessels, which showed high post-thawing survival rate (75.1%) of follicles, blood vessels, and stromal structures. Bedaiwy et al. [43] also achieved a high follicle survival rate after freezing and thawing of the whole ovary, using a method that involved dispersing cryoprotectant into the ovary parenchyma by injection through the ovarian artery. Further studies on whole-ovary cryopreservation and transplantation are necessary for this technique to be applied in clinical practice.

Oocytes collected from ovarian tissue prior to ovarian tissue cryopreservation could be cryopreserved or undergo IVM to produce mature oocytes for cryopreservation. Oocytes can be collected from antral follicles within the ovarian tissue or from the dissection medium. There is only one report of a live birth resulting from cryopreserved embryos that were created using IVM oocytes retrieved from ovarian tissue after oophorectomy [44].

Researchers are continuing to focus on finding better ways to protect oocytes from the toxic effects of cancer treatments. One approach is to encapsulate chemotherapeutic reagents inside nanoparticles that specifically target cancer cells [104105], thereby decreasing the plasma levels and toxicity of these drugs into non-target tissues like ovary. This strategy can be applied to other chemotherapeutics to more specifically target cancer cells while also protecting the gonads.

Other studies aim to develop fertoprotective agents that can kill cancer cells but also protect oocytes against chemo- and radiation treatment (Fig. 2B). Several studies have explored the development of such fertoprotective reagents, including the apoptosis inhibitors imatinib, AS101, T3, S1P, G-CSF, and tamoxifen. It has been suggested that imatinib mesylate (Gleevec, Novartis, Basel, Switzerland) protects oocytes through binding to c-Abl [106107], a protein tyrosine kinase. It was also proposed that c-Abl is upregulated by DNA damage induced by radiation or cisplatin and stimulates apoptotic pathways. TAp63 lies downstream of c-Abl and is a main factor involved in primordial oocyte death [108], indicating that c-Abl may be a target for blocking apoptotic pathways in oocytes using imatinib mesylate. Although c-Abl is upregulated by DNA damage [109], further studies are needed to understand the relevance of c-Abl in oocyte death or for the direct regulatory relationship between c-Abl and TAp63.

AS101 has been identified as an immunomodulator that protects oocytes in large growing follicles co-treated with Cy, through reduction of apoptosis and down-regulation of the PI3K/PTEN/AKT pathway, resulting in a significantly higher numbers of pups from mice treated with Cy and AS101 than with Cy alone [110]. This finding suggests that AS101 can block the abnormal activation and growth of follicles induced by Cy treatment. More importantly, clinical trials have shown that AS101 is not harmful to humans. However, further research is needed to determine how Cy activates primordial follicles without inducing apoptosis. 3,5,3’-Triiodothyronine (thyroid hormone, T3) has anti-apoptotic effects in various cell lines and protects granulosa cells against paclitaxel-induced apoptosis [111]. However, there is no clear evidence of a fertoprotective effect of T3 against chemotherapeutics, requiring further investigation. Sphingoshine-1-phosphate (S1P) is an inhibitor of sphingomyelin hydrolysis-induced apoptosis in cells. S1P (FTY720) pretreatment has protective effects against both radiation treatment and chemotherapy agents including dacarbazine (DIC or DTIC) [112]. This effect was confirmed in human ovarian tissue with S1P pretreatment followed by Cy and doxorubicin (DXR) treatment [113114]. These promising results were replicated in macaques, which had live births after co-treatment of S1P with radiation exposure [115]. However, the mechanism by which S1P protects primate ovaries from cytotoxic drug-induced damage in vivo still remains to be determined. Moreover, S1P must be administered directly into the ovary, making clinical translation a challenge. Granulocyte colony-stimulating factor (G-CSF) has shown to reduce the loss of primordial follicles induced by Cy, busulfan, and cisplatin through protection of the ovarian blood vessels [116]. Despite optimistic results showing a protective effect against chemotherapy, the mechanism by which G-CSF protects blood vessels is still not known. Finally, tamoxifen also showed a protective effect against follicular loss caused by Cy, 7,12-dimethylbenzanthracene, and DXR in rat ovaries [117] and against radiation in oocytes [118]. In summary, these data suggest several feasible fertoprotective agents with the potential to protect oocytes against chemo- or radiation therapy may be available in the near future.