Eight years' experience in autologous oocytes vitrication for male factors: efforts to nd relevant clinical predictors of oocytes survivability

Objective The objective of this study was to provides a descriptive analysis of the clinical outcomes achieved in oocyte vitrication in cases of unavailable sperm on oocyte retrieval day, and to nd predictors of oocyte survival. Methods This retrospective cohort study used data from a university-aliated reproductive medicine center. There were 321 cycles carried part or all oocytes vitrication due to unavailability of sperm from March 2009 to October 2017. A descriptive analysis of the clinical outcomes including both fresh embryo transfers and cryopreserved embryos transfers was provided. The ability of an individual parameter to forecast oocyte survival per thawing cycle was assessed by a binary logistic regression analysis. The cumulative probability of live birth (CPLB) was estimated by using the K-M method according to the total number of oocytes consumed in consecutive procedures.


Introduction
Oocyte freezing is no longer considered an experimental method by the American Society for Reproductive Medicine [1]. Oocyte vitri cation is gradually becoming a useful adjunct to routine in vitro fertilization (IVF) in various clinical scenarios such as the unavailability of sperm at the time of egg retrieval [2,3], and for couples who do not wish to cryopreserve supernumerary embryos in cases of plenty of oocytes are retrieved [4]. Another applicable indication for oocyte vitri cation that has now become a reality is the establishment of donor oocyte banks [5,6]. Oocytes cryopreservation for deferring child bearing and fertility preservation in cancer patients is also applied in clinic [7][8][9][10].
Reports of donor oocyte vitri cation of have so far been encouraging. In a sibling cohort study of recipient cycles, similar embryo development has been shown from fresh versus vitri ed oocytes [11].
Several well controlled studies about donor oocyte have shown clinical outcomes with vitri ed oocytes are comparable to those of fresh oocytes [6, [12][13][14]. A large study of donation program reported by Cobo et al. has demonstrated comparable obstetric and perinatal outcomes from vitri ed versus fresh oocytes [15]. These results have con rmed the further application of oocyte vitri cation in assisted reproduction treatment for medical indications.
Although oocyte vitri cation has been demonstrated as a successful and stable technique in donor program, these information from donation program might provide overly optimistic evidence for oocyte vitri cation of medical indications in infertile patients. Different oocyte sources may exist with a different inherent quality that affect the vitri cation outcomes [7,16]. However, reports related to autologous oocytes vitri cation in infertile patients was few and inconsistent [10,17]. A study of sibling oocytes from 44 patients undergoing IVF showed reduced rates of fertilization and embryo development after oocyte vitri cation [17]. Another study included 128 autologous vitri ed/warmed oocytes cycles from IVF cycles demonstrated signi cantly higher implantation rates (43% vs. 35%) and clinical pregnancy (57% vs. 44%) with vitri ed-warmed compared to fresh oocytes [10].
This study aims to describe the outcomes we have achieved from 8 years' experience of oocyte vitri cation due to unavailable sperm on oocytes retrieval day. Analysis was carried out to nd relevant factors on the oocytes survivability. The reported relatively large data would add to limited information yet available on the clinical application of vitri ed autologous oocytes for medical indications.

Materials And Methods
The ethics committee at the Center for Reproductive Medicine, Shandong University approved this clinical application of oocyte vitri cation. Couples chose oocytes cryopreservation because of unavailability of sperm at the time of oocyte retrieval instead of using donor semen. The control group consisted of age and BMI-matched patients, who were undergoing ICSI treatment for male factor infertility ( Figure I) EG + 7.5% PROH (DMSO), and the VS constituted of 15% EG + 15% PROH (DMSO) + 0.5 mol/L sucrose, per the instructions. The Modi ed kit was prepared with M-199 (Gibco Invitrogen Corp., Grand Island, NY, USA) as the basal media. A 20% serum plasma substitute (SPS) (SAGE, Trumbull, CT, USA) was also added. The ES for the Modi ed kit comprised 7.5% EG + 3.75% DMSO + 3.75% PROH, and the VS comprised 15% EG + 7.5% DMSO + 7.5% PROH + 0.5 mol/L sucrose in a M-199 medium with 20% SPS. [18] Oocyte vitri cation was performed at the room temperature. The oocytes were equilibrated in ES for 5-10 min until they recovered their shape, and then they were placed into the VS for 1min. Finally, the vitri ed oocytes were placed on a CryoLoop (Hampton Research, Laguna Niguel, CA, USA) and immediately immersed in liquid nitrogen. No more than four oocytes were loaded onto each CryoLoop. Oocyte warming was performed at room temperature, except for the rst step. The CryoLoop with the vitri ed oocytes was taken out of the liquid nitrogen and immediately placed in 1.0 mol/L sucrose in a M-199 + 20% SPS solution at 37°C for 1.5-2.0 min. Next, the oocytes were placed in 0.5 mol/L sucrose in an M-199 + 20% SPS solution for 3 min at the room temperature, after which they were transferred into another M-199 solution with 0.25 mol/L sucrose for 3 min. Finally, they were washed in M-199 + 20% SPS for 5-10 min while the stage was warmed slowly. After warming, the surviving oocytes were cultured for 2 h in G-IVF (Vitrolife, Göteborg, Sweden) in an incubator with 37°C, 6% CO2 before being inseminated by ICSI.
[18] Embryo transfer was performed on Day 2 or Day 3 depending on embryo quality or quantity. No more than 3 embryos were transferred per transfer. The supernumerary embryos were cultured into blastocysts, and high-quality blastocysts were vitri ed.

Endometrial preparation and Pregnancy assessment
All patients used hormone replacement therapy as endometrial preparation protocol, which had been described in previous study [19]. In short, 4-8mg oral oestradiol valerate was administered daily for at least 10 days starting on day 2-5 of the menstrual cycle. When the endometrial thickness reached ≥ 8 mm, oral dydrogesterone 20mg twice daily plus vaginal micronized progesterone 200 mg once daily was initiated on the day of oocyte warming. Clinical pregnancy was determined as the presence of a gestational sac identi ed by vaginal or abdominal ultrasound 4-5 weeks after embryo transfer (ET).
Gestational age, birth weight and congenital malformations outcomes were followed.

Statistical analysis
The main outcome measurements were survival rate and the cumulative live birth rate (including live birth from fresh ETs and subsequent cryo-ETs) per warming cycle. The secondary outcome measures included laboratory outcomes of vitri ed-warmed oocytes, implantation, clinical pregnancy rates, and the delivery rate per fresh embryo transfer and vitri ed embryo transfer, as well as gestational age, birth weight and congenital malformations outcomes.
The difference in means and prevalence among the groups were analyzed by Student's t-test for continuous data and Chi square for categorical data. A p value < 0.05 was considered statistically signi cant. A binary logistic regression model was performed to nd predictable parameter of oocyte survival per thawing cycle. The oocyte-to-baby rate was calculated by dividing the number of live births by the total number of oocytes consumed×100. The cumulative probability of live birth (CPLB) was estimated by the Kaplan-Meier method according to the total number of oocytes consumed in consecutive procedures, including oocytes from cancelled ETs and from fresh or cryo-ETs, until a live birth was achieved.

Results
Three hundred and twenty-one oocytes vitri ed-warmed cycles were carried out from March 2009 to October 2017 due to unavailable sperm on oocytes retrieval day. These oocytes had been vitri ed during 2007-2013 previously. The incidence rate was around 0.3%-0.5% in all IVF/ICSI cycles during the years.
The majority cases for all oocyte vitri cation were unavailable sperm from ejaculated sample or surgical sperm extraction (73.23%), followed by unable to provide ejaculated sample through masturbation (17.72%) and unexpected absence of male partner (9.05%) ( Table I). The median age of the female patients at oocyte vitri cation was 30.24 years (95% CI 29.82-30.89). The median preservation duration oocyte vitri cation was 6.52 weeks (95% CI 5.56-7.62). The overall oocytes survival rate was 83.13% (95% CI 81.81-86.35%). Data were also obtained from age and BMI matched controls undergoing fresh ICSI cycles for severe male factor with autologous oocytes. Similar fertilization rates shown between vitri ed and fresh groups, while high-quality embryo rate decreased signi cantly (33.33% vs. 53.75%, P < 0.0001) in vitri ed group. (Table II)   Table III shows information from two groups divided by the median survival rate (91.67%). The serum total cholesterol (TC)in the ≥ 91.67% survival rate group was higher. The blood glucose level was also higher in the ≥ 91.67% survival rate group. There were more cycles due to absolute male factors included in the ≥ 91.67% survival rate group. And the preservation time was longer in < 91.67% survival rate group. Table IV shows comparison between two oocytes vitri cation groups divided by different reasons for lack of sperm availability on oocyte retrieval day. The Relative Male Factor group (due to Unable to provide ejaculated sample through masturbation or unexpected absence of partner) presented a higher serum triglyceride (TG). The oocytes survival rate was higher in Absolute Male Factor group (due to unavailable or insu cient sperm from ejaculated sample or surgical sperm). Table V shows comparison of sibling oocytes between fresh and vitri ed groups in part oocytes vitri ed cycles. A total of 67 cycles had part MII oocytes vitri ed because of male factors. Forty-one cases inseminated with husband sperm in both fresh oocytes and vitri ed oocytes. No signi cant difference was found between the vitri cation and fresh groups of sibling oocytes in fertilization rate (66.93% vs. 59.77%), but few high-quality embryo developed in vitri ed group (27.68%% vs. 53.46%).
There were 81 vitri ed embryo transfer cycles, including 53 DF transfer (i.e. vitri ed oocyte and vitri ed embryo) cycles and 28 TF (i.e. vitri ed oocyte, frozen sperm and vitri ed embryo) transfer cycles, which yield 22 and 11 neonates respectively. The delivery rate per transfer, gestational age, birth weight and congenital malformations outcomes were similar among groups. (Supplementary Table IV) One hundred and forty-two babies have been born as a result of 262 fresh ETs and 81 subsequent cryo-ETs. The cumulative live birth rate per warming cycle was 41.40%. The oocyte-to-baby rate was 4.3%. At the end of present study, 110 blastocysts were still cryopreserved from these oocyte warming cycles included in this work. Suppose the delivery rates are maintained with this cohort, a rough estimation after their use could yield an outcome of 36 additional babies, which would enhance the oocyte-to-baby rate to 5.4%. The Kaplan-Meier analysis showed no signi cantly different CPLB between patients ≤ 35 versus > 35 years (Log-rank (Mantel-Cox); P = 0.231; Breslow (generalized-Wilcoxon); P = 0.458; and Tarone-Ware; P = 0.388). The CPLB improved when more oocytes were warmed and the curve for older patients reached the plateau earlier (with 15 oocytes) than those for young women (with 23 oocytes). ( Figure II) A Binary Logistic regression model was performed to nd predictable parameter of oocyte survival per thawing cycle. Several parameters were introduced into the initial model as predictors, including age, BMI, metabolic indicators (includin g TG, TC, HDL, etc.), basic hormones, infertility years, PCOS/non-PCOS, endometriosis/non-endometriosis, ovarian stimulation protocols, reason for lack of sperm availability, vitri cation kits, storage duration. As shown by the OR, the effect of the reason for lack of sperm availability was acknowledged, and the effect of serum TC on survival was put forward (Supplementary  Table II).

Discussion
Given that oocyte cryopreservation techniques have changed from slow freezing to vitri cation according to the safety and e cacy for the past decade reports [20], oocyte vitri cation is gradually applied in assisted reproduction treatment in various clinical scenarios. Especially, oocyte vitri cation is becoming an indispensably alternative technique for the couples without enough available sperms at the time of egg retrieval [16].
Our study represents the ndings of the largest data set from a single center in China of vitri ed autologous oocytes, which were from the couples lack available sperms at the time of egg retrieval. This report comprises 321 oocytes vitri cation-warming cycles. A total of 142 healthy babies born from fresh and frozen embryo transfer and the cumulus live birth rate per warming cycle was 44.24%, and 118 cycles (36.76%) had successfully taken baby home.
Different oocyte sources, like cancer patients, fertility preservation women, oocytes donors or infertile patients, may exist with different inherent quality that affect the vitri cation outcomes [3,16,19,21].This study would add more information, which is not optimistic, to the oocyte vitri cation outcomes from infertile patients. Inconsistent with the reports about oocyte recipient cycles [11,15], the high-quality embryo rate of vitri ed oocytes group decreased signi cantly in comparation to fresh control group in present study (Table II). And fewer high-quality embryos developed in vitri ed group in the sibling oocytes comparation from part-vitri ed oocytes cycles (Table V). Similar results had also demonstrated by other studies from egg-sharing program or autologous oocytes vitri cation cycles in infertility women undergoing IVF [22,23]. Another supporting result in present study was from comparison between groups of different reasons for lack of sperm availability. The survival rate was signi cant higher in the absolute male factor group, and the result could be explained that women in this group were relative "fertile" (Table  IV). All these outcomes provided evidences that the oocytes from infertile women are more vulnerable to vitri cation injury and might not survival through the procedures.
In order to obtain more referential information for clinical work, we tried to nd some useful predictors for the success of oocyte vitri cation. Age was rstly taken into consideration. However, in present study, no signi cant difference was discovered between the two age groups (≤ 35 years vs. >35 years) in survival rate, fertilization rate and high quality embryo rate (Supplemental Table I). These results were con rmed by the K-M analysis of CPLB according to different age groups. No signi cantly difference was observed between two age groups ( Figure I). This result was inconsistent with previous studies [6, 7, 11, 15], most probably due to the small sample size and characteristic of the older patients involved in present work. Only 50 (15.68%) patients more than 35 years old were included, because most advanced age couples are more inclined to choose donor sperm in case of unavailable sperm on oocyte retrieval day. The average number of retrieved oocytes in this older age group was 11.46 (95% CI 9.70-13.22), which indicate a better ovarian reserve of these patients than their peers and further explain the unsigni cant difference between two age groups. However, we observed the curve of older patients' curve reached the plateau earlier than young women, which was similar with other studies [7,11,15].
The average survival rate in present study was 83.13% (95% CI 81.81-86.35%), comparable to the published data range from 68.6-96.8% [10,[23][24][25][26] in the literatures. We compared two groups divided by the median survival rate (91.67%). Statistical differences were found in the serum TC, blood glucose, the proportion of different reasons for lack of sperm availability and preservation time. The results were part consistent to multiple logistic regression analysis. As shown by the OR, the effect of the reason for lack of sperm availability was acknowledged. Another parameter entered the model was serum TC, which had never been analyzed in human oocytes vitri cation studies before. A higher serum TC level was found favorable to the oocyte survivability after vitri cation. Cholesterol is known to be the major non-polar lipid of mammalian cell membranes [27]. Modulation of plasma membrane cholesterol to increase postcryopreservation survival is currently a new topic in mammals oocytes vitri cation [2,28,29]. Additionally, several studies had reported that the serum levels of some lipid biomarkers were associated with IVF outcomes [30,31]. Large prospective studies and mechanism researches are needed to clarify if serum sterol lipids levels or other lipidome biomarkers are relevant to the oocyte survivability after vitri cation.
Oocyte vitri cation e ciency could be de ned as the way to a live birth with the lowest number of vitri ed oocytes. Although we have obtained a cumulus live birth rate per warming cycle as 44.24%, the oocyte-tobaby rate was only 4.3% in present study. About one third couples (36.76%) had successfully taken babies home. Other studies about oocytes vitri cation for medical indications had reported quite different outcomes with oocyte-to-baby rate. Kara et al. reported the live-birth rate per mature oocyte was 3.0% in oocytes cryopreservation group (< 35 years old) [22]. Doyle et al. estimated live birth per warmed oocyte as 6.5% (including predicted live birth from remaining cryopreserved blastocysts) [25]. The data herein would provide more information for clinicians to provide suggestions when patients face with the situation of unavailable sperm on oocytes retrieval day.
The outcomes of live delivery, including gestational age, birth weight, and live birth congenital defect were compared to fresh control group, and no signi cant difference was discovered. The limited data we could achieve here showed double vitri cation (oocyte and embryo vitri cation) or triple-cryopreservation (oocyte/embryo vitri cation and sperm cryopreservation) had no adverse effect on perinatal outcomes.
The main drawbacks of our study were that few patients > 35 years were included, and the couples in present study were mostly with severe male factors, which might in uence the subsequent embryo quality and pregnancy outcomes. Another drawback was the long time period this retrospective study had included, which might add some variabilities that could play into the data presented. However, we have a relative stable lab team with trained technicians for oocytes vitri cation. Furthermore, we had included stimulation protocols and vitri cation kits parameters that changed through time into regression model as potential confounders.
In conclusion, Oocyte vitri cation is proved to be an indispensable and effective alternative when lack of available sperm on oocyte retrieval day. Present study provided evidences that the oocytes from infertile population were more likely suffer to vitri cation injury. Clinicians need to take this into account when giving suggestions to patients for similar situations. Further studies will be necessary to clarify the correlation between serum sterol lipids levels and human oocyte survivability after vitri cation. Table 1 Reasons for lack of sperm availability on the day of oocyte retrieval

Groups
Cycles Reasons for oocytes vitri cation at fresh retrieval

Declarations
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This study was approved by the Institutional Review Board of Center for Reproductive Medicine A liated to Shandong University. Written informed consent was waived due to the retrospective nature, and patients' data were used anonymously.

Consent for publication
Not applicable. The cumulative probability of live birth according to age (≤35 year vs. >35 year) and number of oocytes consumed

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.