Comparison of obstetrical and neonatal outcomes between fresh versus frozen-thawed testicular sperm derived from microTESE

Abstract

Purpose

To compare obstetrical and neonatal outcomes of embryo transfer cycles using fresh vs. frozen-thawed testicular sperm derived from microTESE in non-obstructive azoospermia (NOA) patients.

Design

The retrospective cohort study included a total of 48 couples diagnosed with NOA who underwent 93 ET cycles, both fresh and frozen-thawed embryos, and resulted in pregnancy. ET cycles were divided into two groups according to sperm type, fresh (46 cycles, 49.5%) or frozen (47 cycles, 50.5%) testicular sperm. The primary outcome was the birth weight of newborns correlated with gestational week (birth weight percentile).

Results

A comparison of patients’ basic characteristics and ET cycle parameters showed no significant clinical differences between the groups. A total of 172 embryos were transferred, 86 (50%) in each group. A higher rate of good-quality blastocysts was found in the fresh testicular group (83.3% vs. 50%, p = 0.046). A comparison of pregnancy outcomes showed no significant differences in clinical pregnancy, implantation, or live birth rates. A total of 53 cycles resulted in live birth, 26 (49%) and 27 (51%) in the fresh and frozen groups, respectively. No difference was found in pregnancy length, delivery mode, or obstetrical complications. A total of 61 newborns were included, 31 (51%) and 30 (49%) in fresh and frozen testicular groups, respectively. No significant differences were found in mean birth weight or birth weight percentile between the groups.

Conclusion

No significant differences were found in obstetrical outcomes when comparing ET cycles using fresh or frozen-thawed testicular sperm retrieved from microTESE. Moreover, there is no association between the sperm source and the birth weight of newborns.

Introduction

Azoospermia is defined as the absence of sperm cells in a routine sperm analysis. The frequency of azoospermia in the general male population is 1% and up to 10–15% among infertile men [1, 2]. Azoospermia is usually divided into obstructive azoospermia (OA) and non-obstructive azoospermia (NOA). NOA is often more severe, due to a disruption or complete absence of the sperm production process, and therefore requires more complex treatment methods. In the past, a man diagnosed with azoospermia was defined as “infertile,” and the therapeutic alternatives included sperm donation or adoption. Currently, the assumption is that there might be local production of cells in the testicles [3–5]. Therefore, several surgical methods have been developed that allow the extraction of sperm cells directly from the testicle. These surgical techniques along with the development of intracytoplasmic sperm injection (ICSI) allow azoospermic men to have genetic offspring.

The microsurgical testicular sperm extraction (microTESE) procedure was first described in 1999, using a sophisticated surgical microscope during the surgical process, where the surgeon can view the testicular tissue at a magnification of 20–25 times [6]. In microTESE, the specimen is carefully selected for thicker tubules, where there is a greater chance of finding sperm and also allows avoiding bleeding that may interfere with the quality of the testicular sample [7].

Sperm preservation and freezing are sometimes required and even mandatory following surgical procedure for sperm extraction since repeated attempts are not always possible or equally effective while repeated ICSI treatments may be necessary until the desired pregnancy is achieved. Studies that compared pregnancy outcomes of fresh versus frozen-thawed surgically retrieved testicular sperm showed no difference in the fertilization, implantation, clinical pregnancy, and live birth rates [8–11]. Although sperm retrieved by microTESE result in healthy offspring [12–14], there is a lack of high-quality prospective clinical trials that focus on evaluating the short- and long-term health of the newborns and children in general and specifically following using fresh versus frozen testicular sperm.

The main purpose is to study the association between embryo transfer cycles using fresh versus frozen-thawed microTESE-derived sperm and birth weight of newborns as well as other obstetrical outcomes.

Materials and methods

Study design and setting

This is a single-center, retrospective cohort study, conducted in a tertiary IVF clinic between January 1, 2010, and December 31, 2021.

Patient selection

Inclusion criteria included fresh and frozen-thawed embryo transfer cycles that resulted in pregnancy and the sperm source was either fresh or frozen-thawed microTESE retrieved. Patients with negative pregnancy test were excluded from the analysis. The female partner’s age was between 18 and 45 years, and pregnancy was defined as βHCG blood test levels ≥ 25 mIU/ml. Exclusion criteria included patients with incomplete or missing information regarding pregnancy outcome, treatment failure (i.e., treatments resulted in the absence of sperm extracted in microTESE, oocytes, fertilizations, embryos available for transfer, or negative pregnancy test), male partner diagnosed with infertility other than NOA (i.e., OA), concomitant pre-implantation genetic testing (PGT) of any cause, and egg or sperm donation. The two study groups included pregnancies following the usage of fresh or frozen testicular sperm. Each group included fresh and frozen-thawed embryo transfer cycles.

Data collection

Data regarding the patients’ clinical characteristics, fertility and cycle parameters, pregnancy, and obstetrical information as well as neonatal sex and birth weight was collected retrospectively from the computerized and manual patients’ files at our IVF Unit. When there was missing data concerning obstetrical issues, it was completed via an annual report published by the “Gertner Institute for Epidemiology and Health Policy Research,” which holds information of the successful live birth of each IVF unit in Israel since 2016. If data regarding the main outcome was still missing, we contacted the patients and questioned about the missing information. Data were reported per cycle and not per patient; hence, couples with more than one successful treatment were included more than once.

Blastocyst quality was scored into good/fair/poor quality by the IVF lab team based on the morphological appearance and developmental parameters.

Testicular sperm cryopreservation

Following ICSI using testicular sperm, the remaining testicular tissue extract was cryopreserved using a standard freezing protocol. The sperm-containing extract was slowly mixed in a 1:1 ratio with Quinn’s Advantage™ Sperm Freezing Medium, a commercial HEPES-buffered salt solution containing human serum albumin, glycerol, sucrose, and gentamicin (Cooper Surgical, CT, USA). The mixture then is transferred to cryovials of 1 mL and incubated at room temperature for 10 min; afterwards, the cryovials are placed in a liquid nitrogen vapor-containing chamber at about 10 cm above the level of N2 (− 80 °C) for 8–10 min to allow initial slow freezing and stored finally in liquid nitrogen at − 196 °C [15].

Embryos morphological scoring and transfer

Fertilized embryos were incubated until the day of embryo transfer. During this time, a routine assessment of embryos’ quality and scoring was performed by the IVF lab team. Cleavage-stage embryos were evaluated by the number of blastomeres and morphological scoring according to Alpha and ESHRE consensus [16] and blastocyst grading performed according to the Gardner embryo grading system [17]. It should be noted that all the embryos in our study were frozen using the vitrification methods.

The number of embryos to transfer is determined according to the Israel Fertility Association (IFA) and Israeli Ministry of Health guidelines. Moreover, the developmental stage of embryo transfer is often determined by the women’s age, the number of embryos achieved, and their morphological quality.

Primary and secondary outcomes

The primary outcome of this study was the birth weight of newborns correlated with gestational week (birth weight percentile). Birth weight percentile was calculated according to Dolberg birth weight curves, which take into consideration sex, birth week, and number of newborns, based on the live-born population in Israel [18]. Birth weight percentile outcomes were calculated for all neonates and in sub-analysis for singleton and multiple pregnancies. Secondary measures included pregnancy outcomes: (1) chemical pregnancy, (2) clinical pregnancy (defined as at least one intrauterine implanted sac that is seen during the first ultrasound scan) and the number of implanted sacs, (3) spontaneous missed abortion (MA) (defined as spontaneous pregnancy loss in the 1st trimester or 2nd trimester till 21 + 6 weeks of pregnancy), (4) termination of pregnancy (TOP), (5) extrauterine pregnancy (EUP), (6) intrauterine fetal death (IUFD) defined as spontaneous pregnancy loss after 22 weeks of pregnancy, (7) ongoing pregnancy, and (8) live birth. Moreover, obstetrical data were collected including pregnancy complications: hypertensive diseases, gestational diabetes (GDM), intrauterine growth restriction (IUGR), placental factors (i.e., placenta previa, placental abruption or hematomas, velamentous cord insertion, abnormally invasive placenta), pre-term labor (PTL) defined as delivery before 37 weeks of gestation, pre-term premature rupture of membrane (PPROM), multiple pregnancy rate, mode of delivery, and birth weeks.

Statistical analysis

Statistical Package for the Social Sciences (SPSS Inc. V28.0.1.1, 2021) was used for all statistical analyses. Association between categorical variables was assessed using the Pearson chi-square test and Fisher’s exact test, where appropriate. Continuous variables were compared using either Student’s t test or Mann–Whitney nonparametric test. The strength of the association between two continuous variables was estimated by calculating the Pearson correlation coefficient and the Spearman nonparametric correlation coefficient. A two-sided p value ≤ 0.05 was considered statistically significant. When a statistically significant association was found between two categorical variables in which one of the variables had more than two categories, post hoc comparisons were carried out. These multiple pair-wise comparisons were performed in order to find out which categories most contributed to the significant association and were assessed according to the Bonferroni method for the correction of the significance level. Logistic regression tests were used to adjust outcome measures of the study for confounding variables.

Ethical issues

The research project was approved by the ethics committee, IRB protocol number 0200–22-SZMC.

Results

During the study period, 125 couples underwent a microTESE procedure. Of them, 48 couples, who met the inclusion criteria, underwent 93 embryo transfer cycles, which were divided into two groups according to sperm source. fresh (46 cycles, 49.5%) or frozen (47 cycles, 50.5%) testicular sperm (Fig. 1).

Fig. 1.

Flow chart showing the design, inclusion, and exclusion criteria of cases in the study. mTESE, microsurgical dissection testicular sperm extraction; PGT, pre-implantation genetic testing; ET, embryo transfer; FET, frozen embryo transfer

Basic patient characteristics were compared between the two study groups and are presented in Table 1(a). No significant differences were noticed regarding mean female and male BMI, medical history, and infertility cause or type. Abnormal male karyotype rate was similar in the study groups as well as the prevalence of previous testicular surgical procedure. Mean female and male ages were significantly higher in the frozen group compared with the fresh group (33.0 ± 4.7 vs. 31.0 ± 4.9 and 37.4 ± 7.0 vs. 34.2 ± 6, p = 0.045; 0.020, respectively). Female smoking habits were no different between the groups, whereas male smoking rates were significantly higher in the fresh testicular sperm group (42.9% vs. 22.0%, p = 0.042). Regarding female partner obstetrical history, the frozen testicular sperm group showed significantly higher rates of women who previously conceived (41.3% vs. 63.8%, p = 0.030), while the CS delivery rate was higher among women who underwent embryo transfer cycles with fresh testicular sperm compared with frozen sperm (37.9% vs. 14.3%, respectively, p = 0.022). Other pregnancy and obstetrical categories did not show statistically significant differences.

Table 1.

The distribution of a couple of basic characteristics (a), embryo transfer cycle parameters (b), and embryo characteristics (c) based on the status of fresh versus frozen microTESE. Data in bold emphasis indicates a difference with statistical significance (p<0.05)

a. Couple characteristics	Fresh mTESE (n = 46)	Frozen mTESE (n = 47)	p value
Female age, mean ± SD	31.0 ± 4.9	33.0 ± 4.7	0.045
Male age, mean ± SD	34.2 ± 6	37.4 ± 7.0	0.020
Female BMI (Cm/m2), mean ± SD	24.5 ± 5.8	26.0 ± 4.8	0.279
Male BMI (Cm/m2), mean ± SD	24.8 ± 3.2	27.0 ± 5.8	0.100
Smoking (female), n (%)	2 (4.5)	6 (14)	0.157
Smoking (male), n (%)	18 (42.9)	9 (22)	0.042
Obstetric history, n (%)
Gravidity	19 (41.3)	30 (63.8)	0.030
Parity	16 (34.8)	23 (48.9)	0.167
Miscarriage	8 (26.7)	13 (30.2)	0.741
EUP	0 (0)	1 (2.4)	1
CS	11 (37.9)	6 (14.3)	0.022
Infertility cause, n (%)
Male factor only	40 (87)	35 (79.5)	0.346
Combined	6 (13)	9 (20.5)
Infertility type, n (%)
Primary	26 (56.5)	18 (38.3)	0.078
Secondary	20 (43.5)	29 (61.7)
Male karyotype, n (%)
Normal	35 (89.7)	40 (97.6)	0.195
Abnormal	4 (10.3)	1 (2.4)
Previous surgical procedure, n (%)	19 (41.3)	16 (64)	0.470
Sperm motility, n (%)	40 (87)	41 (89.1)	0.748
Testicular histopathology, n (%)
Tubular sclerosis	1 (2.6)	0	0.050
SCO	8 (21.1)	1 (2.9)
Maturation arrest	12 (31.6)	12 (34.3)
Hypospermatogenesis	17 (44.7)	22 (62.9)
b. Embryo transfer cycle parameters	Fresh mTESE (n = 46)	Frozen mTESE (n = 47)	p value
No. oocyte retrieved, mean ± SD	14.1 ± 6.3	12.8 ± 4.0	0.464
No. M2 oocytes, mean ± SD	11.9 ± 6.3	10.2 ± 3.8	0.318
No. 2PN fertilization, mean ± SD	6.3 ± 4.0	6.1 ± 3.4	0.924
Fertilization rate, mean ± SD	0.59 ± 0.24	0.62 ± 0.31	0.686
Oocyte type, n (%)
Fresh	45 (97.8)	45 (95.7)	1
Frozen	1 (2.2)	2 (4.3)
ET cycle type, n (%)
Fresh	20 (43.5)	26 (55.3)	0.253
Frozen	26 (56.5)	21 (44.7)
Fresh protocol type, n (%)
Antagonist	15 (75.0)	20 (76.9)	1
Agonist	5 (25.0)	6 (23.1)
Frozen protocol, type n (%)
Natural	9 (34.6)	11 (52.4)	0.220
HRT	17 (65.4)	10 (47.6)
No. of ET/BT, n (%)
One	10 (21.7)	13(27.7)	0.767
Two	32 (69.6)	29 (61.7)
Three	4 (8.7)	5 (10.6)
Motile sperm injected, n (%)	30 (83.3)	31 (93.9)	0.263
c. Embryo characteristics	Fresh mTESE (n = 86)	Frozen mTESE (n = 86)	p value
Embryo/Blastocyst, n (%)
Day 2 embryo	11 (12.8)	8 (9.3)	0.144
Cleavage state	63 (73.3)	54 (62.8)
Blastocyst day 5	8 (9.3)	18 (20.9)
Blastocyst day 6	4 (4.7)	6 (7.0)
Day 2 embryos, mean ± SD
Mean no. of blastomeres	4.27 ± 0.78	3.75 ± 0.70	0.154
Mean morphology score	1.75 ± 0.42	2.62 ± 0.95	0.039
Day 3 embryos, mean ± SD
Mean no. of blastomeres	8.08 ± 1.25	7.92 ± 1.93	0.600
Mean morphology score	1.87 ± 0.43	2.01 ± 0.50	0.121
Blastocyst quality, n (%)
Good	10 (83.3)	12 (50)	0.046
Fair	1 (8.3)	11 (45.8)
Low	1 (8.3)	1 (4.2)

TESE testicular sperm extraction, mTESE microsurgical dissection testicular sperm extraction, SCO Sertoli cell only, M2 oocyte metaphase II oocyte, ET embryo transfer, BT blastocyst transfer, HRT hormone replacement

Motile sperm cells were found at equal rates in both study groups. Testicular histopathology results were not significantly different between the two groups. Noteworthy, 40 out of 93 (43%) male patients were also tested for Y-chromosome microdeletions and none was found as a carrier.

A comparison of embryo transfer cycle parameters between the study groups is shown in Table 1(b). There were no significant differences in the number of oocytes retrieved, M2 oocytes, fertilization rates, fresh vs. frozen embryo transfer cycle’s rate, number of embryo/blastocysts transferred per cycle, protocol types, and presence of motile sperm injected during ICSI.

The basic embryo characteristics comparison is shown in Table 1(c). A total of 172 embryos were included in the study, 86 (50%) in the fresh and 86 (50%) in the frozen testicular sperm group. There were no significant differences in the rates of embryo transfer at day 2/3 or blastocyst transfer at day 5/6 in each group, mean number of blastomeres of day 2 and 3 embryos, and mean morphology score of day 3 embryos. However, the mean morphology score of day 2 embryos was significantly lower in the frozen vs. fresh testicular sperm (1.75 ± 0.42 vs. 2.62 ± 0.95, respectively, p = 0.039). Moreover, significant differences were found in blastocyst quality scores between the groups (p = 0.04). The post hoc test was carried out and implies that the main parameters that impact the difference are the good (83.3% vs. 50% for the fresh sperm group) and fair (8.3% vs. 45.8% for the frozen sperm group) scoring; hence, in total, the fresh testicular group has better quality blastocyst rates.

Pregnancy outcomes comparison is presented in Table 2(a). A total of 93 embryo transfer cycles resulted in a positive pregnancy test and were included for the analysis, 46 (49.5%) in the fresh vs. 47 (50.5%) in the frozen testicular sperm group. No significant difference was found in chemical or clinical pregnancy rates, as well as implantation, miscarriage, and EUP rates. No cases of second-trimester abortions or TOP were reported. Importantly, no difference was found in the live birth rate.

Table 2.

The distribution of pregnancy (a) and obstetrical (b) outcomes based on the status of fresh versus frozen microTESE

a. Pregnancy outcome, n (%)	Fresh mTESE (n = 46)	Frozen mTESE (n = 47)	p value
Chemical pregnancy	6 (13.3)	9 (19.1)	0.450
Clinical pregnancy	40 (87)	38 (80.9)	0.423
No. of sac implantation
One	30 (76.9)	30 (78.9)	0.256
Two	9 (23.1)	5 (13.2)
Three	0	2 (5.3)
1st trimester MA	10 (25.6)	10 (26.3)	0.946
2nd trimester MA	0	0	NA
TOP	0	0	NA
EUP	0	1 (2.6)	0.494
Live birth	26 (72.2)	27 (75.0)	0.789
b. Obstetrical outcome	Fresh mTESE (n = 26)	Frozen mTESE (n = 27)	p value
Mean delivery pregnancy week, mean ± SD	38.6 ± 1.7	38.6 ± 2.1	0.889
Delivery mode, n (%)
NVD	13 (50)	11 (44)	0.401
CS	12 (46.2)	10 (40)
VE	1 (3.8)	4 (16)
HTN/PET, n (%)	3 (12.5)	2 (9.5)	1
GDM, n (%)	3 (13)	3 (14.3)	1
IUGR, n (%)	2 (8.7)	0	0.489
IUFD, n (%)	0	0	NA
PPROM, n (%)	0	0	NA
PTL, n (%)	2 (8.3)	4 (16.7)	0.666
Placental factors, n (%)	2 (9.1)	1 (5)	1
Multiple pregnancy, n (%)	6 (25)	6 (22.2)	0.816
Multiple pregnancy complications, n (%)	1 (20)	4 (57.1)	0.293

MA missed abortion, TOP termination of pregnancy, EUP extrauterine pregnancy, NVD normal vaginal delivery, CS cesarean section, VE vacuum extraction, HTN hypertension, PET pre-eclampsia, GDM gestational diabetes mellitus, IUGR intrauterine growth restriction, IUFD intrauterine fetal death, PROM preterm premature rupture of membranes, PTL preterm labor

Obstetrical outcomes are presented in Table 2(b). A total of 53 cycles ended with live birth, 26 (49%) in the fresh testicular sperm group and 27 (51%) in the frozen testicular sperm group. No differences were found in pregnancy length, mode of delivery, and other obstetrical complications (i.e., hypertensive disorders, gestational diabetes, preterm labor, IUGR). No cases of IUFD or PPROM were reported. No difference was found between the groups regarding placental complications. One case of placental hematoma and one case of velamentous insertion of cord were reported in the fresh sperm group while one placenta previa case was reported in the frozen group. Multiple pregnancy rates were similar between the groups, 6 (25%) in the fresh group vs. 6 (22.2%) in the frozen testicular group.

Sub-analysis of obstetrical outcomes shows that out of 53 cycles that resulted in a live birth, 45 (84.9%) were singleton and 8 multiple gestations (15.0%). A total of 65.1% and 34.8% of singleton pregnancies had vaginal and cesarian delivery, respectively, in contrast to 12.5% and 87.5% in the multiple pregnancy group. Hypertensive disease complicated 6.66% of singleton pregnancies vs. 25% of multiple pregnancies. No cases of gestational diabetes or IUGR were found among multiple pregnancies. However, singleton pregnancies had 13.3% gestational diabetes and 4.4% IUGR diagnosis. Preterm labor was more prevalent in multiple pregnancies, where 50% delivered earlier than 37 weeks compared to 4.4% of singleton pregnancies. Placental complications were found in 4.4% of singleton pregnancies compared to 12.5% in multiple pregnancies.

Table 3 presents the comparison of birth weight, birth weight percentile, and sex in each group in general and in sub-analysis of singleton vs. multiple pregnancies. A total of 61 newborns were included, 44 singleton pregnancies, 14 twins, and one triplet. In the fresh sperm group, there were 31 (51%) newborns and 30 (49%) in the frozen testicular group. No significant difference was found in mean birth weight or birth weight percentile between the groups (Table 3(a)). Sub-analysis for singleton and multiple pregnancies was made for better evaluation of the primary outcomes. A total of 21 (47.7%) singleton newborns and 10 (59%) twin newborns were in the fresh testicular group whereas 23 (52.3%) singleton newborns and 7 (41%) twins were in the frozen testicular group. No significant difference was found in birth weight or birth weight percentile in any subgroup (Table 3(b) and (c)).

Table 3.

Birth weight and sex distributed by the status of fresh versus frozen microTESE (a) and sub-analysis to singleton (b) versus multiple pregnancies (c)

a. Neonatal outcome—all	Fresh mTESE (n = 31)	Frozen mTESE (n = 30)	p value
Birth weight (gr), mean ± SD	2918.55 ± 666.7	2981.17 ± 671.0	0.716
Birth weight percentile (%), mean ± SD	49.94 ± 29.9	53.03 ± 26.9	0.673
Sex, n (%)
Female	14 (45.2)	12 (40)	0.684
Male	17 (54.8)	18 (60)
b. Neonatal outcome—singleton	Fresh mTESE (n = 21)	Frozen mTESE (n = 23)	p value
Birth weight (gr), mean ± SD	3177 ± 657.6	3236 ± 448.8	0.728
Birth weight percentile (%), mean ± SD	50.86 ± 34.0	55.48 ± 25.2	0.615
Sex, n (%)
Female	8 (38.1)	6 (26.1)	0.393
Male	13 (61.9)	17 (73.9)
c. Neonatal outcome—multiple	Fresh mTESE (n = 10)	Frozen mTESE (n = 7)	p value
Birth weight (gr), mean ± SD	2374 ± 185.7	2141.43 ± 603.7	0.696
Birth weight percentile (%), mean ± SD	48 ± 19.9	45 ± 32.5	0.845
Sex, n (%)
Female	6 (60)	6 (85.7)	0.338
Male	4 (40)	1 (14.3)

No tendency for any sex was found in any comparison made between the groups in the total pregnancies, as well as in separation into singleton or multiple pregnancies (Table 3).

Univariate analysis for several confounding variables such as patients’, cycles’, and embryos’ characteristics showed no significant predictors for fetus birth weight percentile; hence, the multivariable logistic regression model was redundant.

Discussion

Previous studies warn from the possible adverse effect of cryopreservation on sperm, which may be due to structural damage, increased sperm DNA fragmentation, and damage to mitochondrial function [19]. Therefore, some researchers have expressed concern about the pregnancy outcomes of ICSI with frozen testicular sperm and recommend the use of fresh oocytes with fresh testicular sperm as the first choice [20]. In our study, we analyzed embryo transfer cycles following a microTESE procedure that ended with a positive pregnancy test in order to compare obstetrical and neonatal outcomes between fresh and frozen-thawed testicular sperm in men diagnosed with NOA. Cycles that did not result in pregnancy were excluded from the analysis. We found no difference in pregnancy outcomes between fresh and frozen testicular groups.

In our study, we defined the birth weight percentile as the primary outcome. There were no significant differences in this outcome between the fresh and frozen testicular groups, including total number of deliveries/neonates and also after sub-analysis of singleton and multiple gestations (Table 3). A univariate analysis also did not show any significant predictors for fetus birth weight percentile from the parameters that were analyzed in our study. In our study, we also compared obstetrical outcomes and pregnancy complications and found no differences between the two groups (Table 2). Literature research found no study comparing obstetrical or neonatal outcomes between fresh and frozen testicular sperm following microTESE. Similar studies compared these outcomes in other types of surgical procedures or had a comparison between different sperm sources, among others—TESE/microTESE. Wu et al. [21] compared pregnancy and neonatal outcomes in ICSI cycles performed with frozen vs. fresh testicular sperm following testicular sperm aspiration (TESA) in men diagnosed with OA. No difference was found in pregnancy outcome and the average newborn birth weight was similar in both groups. However, a higher incidence of low birth weight (LBW) was found in the frozen sperm group (20.91% vs. 8.49%, p < 0.05), but sub-analyzing twin and singleton pregnancies in the two groups separately indicated similar LBW rates. Yi-hong et al. followed up children following ICSI treatment with TESA or percutaneous epididymal sperm aspiration (PESA) vs. ejaculated sperm [22] and found no difference in gestational age and rates of stillbirths, perinatal mortality, infant mortality, prematurity, and malformations. A large national cohort study in Denmark summarized the neonatal outcomes and congenital malformation between four groups: ICSI with epididymal or testicular sperm, ICSI with ejaculated sperm, conventional IVF, and natural conception [23]. The study showed higher rates of LBW and some congenital malformations after ICSI with epididymal or testicular sperm in comparison to some of the other groups. Finally, there is a comparison of pregnancy and neonatal outcomes between three groups: men with NOA who underwent microTESE, men with OA who underwent TESA, and pregnancies with sperm donor [24]. The authors concluded that the use of testicular sperm from male patients with NOA negatively affected fertilization and pregnancy outcomes. NOA patients achieved better fertilization and pregnancy outcomes by using donor sperm, but the chance of becoming a biological father was lost. Once a live birth was achieved, there was no difference in neonatal outcomes.

In the current study, the mean age of the male partner in the frozen testicular sperm group was significantly higher than in the fresh testicular sperm (Table 1(a)). This difference could be explained by our institute policy; accordingly, priority is given to fresh embryo transfer derived from fresh testicular sperm and fresh oocytes, later frozen embryos if available, and lastly, the frozen testicular sperm are used. Similarly, this policy may also explain the difference in the female partner age (Table 1(a)) that was older in the frozen testicular sperm group.

Our study demonstrated a higher blastocyst quality rate and higher morphology score in day 2 embryos in the fresh testicular sperm than the frozen testicular sperm group (Table 1(c)). These results are in contrast to the results published in two previously mentioned meta-analyses. Yu et al. published a systematic review and meta-analysis of 17 studies and 1261 ICSI cycles and presented 601 (72.6%) vs. 496 (69.8%) good quality embryos in the fresh and frozen testicular sperm, respectively, with overall summary risk estimates showed no statistical difference between groups [9]. However, it included studies in which the sperm was extracted either with microTESE or TESE. Similarly, in the Madhat et al. meta-analysis, 10 out of the 26 included articles reported the percentage of good quality embryos and none had shown a significant difference between fresh and frozen groups [11]. However, this analysis included studies that compared patients with OA as well as NOA, and the surgical method used for sperm extraction was not exclusive to microTESE. Two studies compared fresh and frozen-thawed testicular sperm extracted from microTESE of NOA patient and demonstrated a difference in clinical pregnancy outcome between the groups while no difference was found in the good embryo quality between the groups [25, 26]. Accordingly, it seems that in the case of embryo and blastocyst quality, our results are not in concordance with the majority of data in the literature. However, in our study, this difference did not influence pregnancy, obstetrical, or neonatal outcomes between the groups; hence, the clinical meaning of this difference is unknown.

Women in the frozen testicular group had a significantly higher mean rate of previous pregnancies (Table 1(a)), which can be attributed to the previously mentioned policy that frozen sperm is last to be used, leading to higher chances of successful or unsuccessful pregnancies in preceding fertility treatments. This could lead to a potential bias since the same sperm source can lead to multiple embryos with similar pregnancy outcomes.

Sperm preservation and freezing are sometimes required and even mandatory following surgical procedure for sperm extraction since repeated attempts are not always possible or equally effective while repeated ICSI treatments may be necessary until the desired pregnancy is achieved. Moreover, repeated procedures may lead to testis damage including injury to blood supply, formation of hematomas, fibrosis, and testicular atrophy. Furthermore, there is an approach that advocates freezing all the sperm in the first place in order to avoid the logistical difficulties in synchronizing the partner’s oocyte retrieval and compromising the incomplete maturity of the oocytes collected in the same cycle as well as cycle cancellations in the case of absence of good sperm quality and quantity appropriate for ICSI. We found no clinical effects of freezing and thawing testicular sperm; hence, clinicians can reassure their patients regarding the use of frozen-thawed testicular sperm when needed.

We are aware of the limitations of the current study being retrospective rather than prospective randomized. Moreover, our study is based on a small number of patients, pregnancies, and deliveries that were analyzed in the study and were underpowered for the secondary outcomes studied. Nevertheless, no previous studies had researched the obstetrical and neonatal outcomes of microTESE procedure in general and specifically in the separation of fresh and frozen testicular sperm. More studies are needed for evaluating the long-term effects of microTESE on children with longer follow-up time.

Another possible limitation of this study is that the frozen sperm group had less SCO and more hypospermatogenesis cases (p = 0.05). This result is consistent with previous data; accordingly, men with SCO pattern in testicular biopsy have poorer prognosis and lower sperm retrieval rate compared with men with hypospermatogenesis [27–30]. Furthermore, the cryopreservation of sperm per se is an indication of the availability of surplus sperm, which is a favorable prognostic sign. Nevertheless, these differences are likely to affect mainly the chance of achieving pregnancy rather than obstetrical and neonatal outcomes once pregnancy is achieved.

In conclusion, no significant differences were found in obstetrical outcomes when comparing embryo transfer cycles using fresh or frozen-thawed testicular sperm retrieved from microTESE. Moreover, there is no association between the sperm source and the birth weight of newborns. However, in light of the small study population and lack of similar studies in the literature, further clinical prospective research is needed to support this conclusion.

Data Availability

All data generated or analyzed during this study are included in this published article. Shaare Zedek IVF and Fertility Unit does not allow public disclosure of patient data used in this study. In case of additional questions, please contact the authors or our IVF and Fertility Unit.

Declarations

We confirm that all authors have met the conditions of authorship, have approved the manuscript, and agree with its submission to the Journal of Assisted Reproduction and Genetics. Moreover, none of the co-authors has any conflict of interest.