Virtual Azoospermia and Cryptozoospermia—Fresh/Frozen Testicular or Ejaculate Sperm for Better IVF Outcome?
Institute for the Study of Fertility, Lis Maternity Hospital, Tel Aviv Sourasky Medical Center, 6 Weizman Street, Tel Aviv, 64239 Israel (e-mail: Ronh@tasmc.health.gov.il).
ABSTRACT: Men diagnosed as having azoospermia occasionally have a few mature sperm cells in other ejaculates. Other men may have constant, yet very low quality and quantity of sperm cells in their ejaculates, resulting in poor intracytoplasmic sperm injection (ICSI) outcome. It has not been conclusively established which source of sperm cells is preferable for ICSI when both ejaculate and testicular (fresh or frozen) sperm cells are available. It is also unclear whether there is any advantage of fresh over frozen sperm if testicular sperm is to be used. We used ejaculate, testicular (fresh or frozen) sperm cells, or both for ICSI in 13 couples. Five of these couples initially underwent ICSI by testicular sperm extraction, because the males had total azoospermia, and in later cycles with ejaculate sperm cells. Ejaculate sperm cells were initially used for ICSI in the other 8 patients, and later with testicular sperm cells. The fertilization rate was significantly higher when fresh or frozen-thawed testicular sperm cells were used than when ejaculated sperm cells were used. Likewise, the quality of the embryos from testicular (fresh and frozen) sperm was higher than from ejaculated sperm (65.3% vs 53.2%, respectively, P < .05). The use of fresh testicular sperm yielded better implantation rates than both frozen testicular sperm and ejaculate. Therefore, fresh testicular sperm should be considered first for ICSI in patients with virtual azoospermia or cryptozoospermia because of their superior fertility.
Men diagnosed as having azoospermia, following several semen analyses with no evidence of sperm, occasionally may have a few mature sperm cells in other ejaculate. They are considered as having virtual azoospermia (Tournaye et al, 1995), which is probably caused by fluctuations in spermatogenesis in cases of nonobstructive azoospermia (Bendikson et al, 2008). After thorough examination or extended sperm preparation, these sperm cells can be used for intracytoplasmic sperm injection (ICSI; Ron-El et al, 1997; Swanton et al, 2007). There is, however, no way to guarantee the detection of sperm cells on the day of ovum pickup. When repeated ejaculated specimens yield no sperm cells and no backup of frozen sperm cells is available, it is possible to consider a rescue testicular sperm extraction (TESE) procedure. Some men may have constant severely low quality and decreased quantity of sperm cells in their ejaculates, which may include either solely nonmotile ejaculated sperm cells (Swanton et al, 2007) or isolated motile sperm cells found after a meticulous search only. This is defined in the World Health Organization (2010) manual as cryptozoospermia. When the outcome of fertilization is low or when no pregnancy is achieved after several trials with these ejaculated sperm cells, the use of testicular sperm cells may be a viable option. Another group of virtually azoospermic patients include those who underwent TESE after having been diagnosed with nonobstructive azoospermia, but mature sperm cells could no longer be found in the retrieved frozen-thawed testicular specimens on the day of ovum pickup, or those specimens yielded very poor fertilization and embryo development. For such cases of virtual azoospermia, fresh ejaculated semen could be a source of sperm cells for use in fertilization.
When both sources of spermatozoa are available (ie, ejaculate or testicular), it is unclear which is preferable for the ICSI procedure. Furthermore, it has also not been established whether fresh testicular sperm has any advantage over frozen. Few studies compared the fertility outcome of ejaculate with the outcome with fresh testicular sperm cells in the same patients. Weissman et al (2008) showed better results with fresh testicular sperm cells in 4 cases with multiple failed IVF/ICSI cycles using ejaculate sperm. Bendikson et al (2008) compared 16 cases in which either fresh testicular or ejaculated sperm were used, and their results showed a trend favoring testicular sperm. Our literature search failed to come up with any study in which the results of frozen-thawed testicular sperm were compared with those of fresh testicular sperm cells in patients with virtual azoospermia or cryptozoospermia. To the best of our knowledge, this is the first clinical study that aimed to compare the outcome of ICSI cycles in patients defined as having virtual azoospermia or cryptozoospermia, where multiple sources of sperm (ie, ejaculated or fresh or frozen-thawed testicular) were available and were used in different ICSI cycles of the same patients. We reasoned that the outcome might indicate which sperm source is more likely to lead to better IVF outcome in this unique group of patients.
Materials and Methods
Selection of Patients
All patients with virtual azoospermia and cryptozoospermia who underwent ICSI cycles with either testicular or ejaculated sperm between October 1996 and May 2009 in our IVF unit were candidates for inclusion in this study. Thirteen couples met the study criteria after exclusion of cases with poor ovarian response and cycles with mixed sperm sources. The current study was approved by the local institutional Review Board Committee in accordance with the Helsinki Declaration of 1975.
The sperm source used in each patient and their order of use in the sequence of treatments is presented in Table 1. TESE was performed first in 5 cases (patients 1–5) of total azoospermia on the basis of several semen analyses using high-speed centrifugation and careful examination of the pellet. Testicular sperm cells were successfully retrieved and used for ICSI in all 5 cases. Fresh testicular sperm cells were used in the first ICSI procedure, and frozen-thawed testicular sperm cells were used in the following cycles in 3 cases (patients 1–3). The TESE procedure was performed electively in the other 2 cases (patients 4 and 5): retrieved sperm cells were first frozen and later thawed and used for ICSI. In these TESE cases, the search for sperm was long and tedious and yielded few and sometimes solely nonmotile sperm cells. When fertilizations, embryo development, or both outcomes were poor after several cycles, or when sperm cells could not be found in the thawed testicular specimens, the male partner was asked to provide an ejaculated semen sample. The ejaculate sperm cells that were found were then used for ICSI cycles.
Table 1. . In vitro fertilization outcome in correlation with sperm source
|1||TESE||28|| ||1|| ||0||0||0|
| || ||28–29|| || ||5||1||0||0|
| || ||30–31||4|| || ||0||0||0|
|2||TESE||35|| ||1|| ||0||0||0|
| || ||37||1|| || ||0||0||0|
| || ||38–41|| || ||3||0||0||0|
|3||TESE||27–28|| ||2|| ||1||1||1|
| || ||27–31|| || ||8||1||0||0|
| || ||31–32||3|| || ||1||1||1|
|4||TESE||27–28|| || ||4||0||0||0|
| || ||28||1|| || ||1||1||1|
|5||TESE||26|| || ||1||0||0||0|
| || ||26||1|| || ||0||0||0|
| || ||27|| ||1|| ||1||1||2|
|6||Ejaculate||34–35||3|| || ||0||0||0|
| || ||36|| ||1|| ||1||1||1|
| || ||37–38|| || ||4||0||0||0|
|7||Ejaculate||42–43||3|| || ||0||0||0|
| || ||43|| ||1|| ||0||0||0|
|8||Ejaculate||29–34||3|| || ||0||0||0|
| || ||31–32|| || ||2||0||0||0|
| || ||36|| ||1|| ||0||0||0|
|9||Ejaculate||36||3|| || ||0||0||0|
| || ||37|| ||1|| ||0||0||0|
| || ||37–44|| || ||11||2||2||2|
|10||Ejaculate||29–30||4|| || ||0||0||0|
| || ||31–36|| || ||6||2||2||3|
|11||Ejaculate||38||2|| || ||0||0||0|
| || ||39|| || ||2||1||1||1|
|12||Ejaculate||27–33||5|| || ||1||1||1|
| || ||38|| || ||1||0||0||0|
|13||Ejaculate||32||1|| || ||1||1||1|
| || ||35–36|| || ||3||2||1||1|
| ||Total|| ||34||9||50||16||13||15|
Ejaculate sperm cells were initially used for ICSI in 8 other cases (patients 6–13). Two of them (patients 12 and 13) had become completely azoospermic during the treatments and they underwent the TESE procedure, during which sperm cells were retrieved. ICSI trials continued using testicular sperm cells. TESE was performed only after several ICSI cycles using ejaculated sperm resulted in poor fertilizations and no pregnancy had been achieved in the other 6 cases (patients 6–11).
Three patients (3, 5, and 8) underwent a second TESE procedure. These procedures were performed after all frozen-thawed testicular sperm cells were used up in previous cycles. The TESE procedure was performed on the day of ovum pickup and fresh testicular sperm cells were used.
According to Israel health regulations, repeated IVF treatments are paid for by the government until 2 children are born, which explains why several cycles of IVF treatments were performed in some insistent couples despite repeated failures. Altogether, the study groups included a total of 34 ICSI cycles using ejaculated sperm, 9 cycles using fresh testicular sperm cells, and 50 cycles using frozen testicular sperm.
Karyotype analysis and microdeletions in the Y chromosome were examined in 9 patients. All patients examined demonstrated a normal 46XY karyotype and only 1 (patient 3) demonstrated an AZFc deletion.
TESE Procedure, Cryopreservation, and Thawing Technique
TESE procedures were performed under general anesthesia. Three biopsies were taken from each testis—from the center and 2 poles—each weighing approximately 50 g, as described elsewhere (Hauser et al, 1998). Wet preparations were performed during the operative procedure and examined immediately. The retrieved testicular material was transferred to the laboratory and underwent mincing of the tissue, centrifugation of the sperm suspension and thorough search under high-power magnification with an inverted microscope as described previously (Ben-Yosef et al, 1999). The total number, motility, and progressive motility sperm variables were recorded whenever possible. Spermatozoa were assessed for any indication of movement (ie, “twitching” of head or tail) in the sperm medium droplet without polyvinyl pyrolidine. A sluggish local motility was observed and this served as a sign of viability in most cases.
The motile fraction of sperm cells from seminal fluid, as well as from TESE specimens, was extracted with the use of ISolate, a 2-layer gradient system, according to the manufacturer's instructions (Irvine Scientific, Santa Ana, California). If no motile sperm cells could be detected, sperm samples were exposed to 8.4 mg/mL pentoxifylline (Sigma; 1:10 vol/vol in modified humena tubal fluid [mHTF] medium). In most cases, sperm cells demonstrated at least sluggish motility within 10–120 minutes of incubation.
Excess testicular material was frozen for subsequent thawed sperm ICSI cycles. The minced testicular tissue was diluted by the addition of medium composed of HTF (Irvine Scientific) with 1% human serum albumin (Kamada, Kibbutz Beit-Kama, Israel) and an equal volume of freezing medium test yolk buffer (Irvine Scientific) added in a droplet fashion. After dilution, the mixture was equilibrated for 15 minutes at room temperature, then sealed in 1-mL tubes (Nunc, Roskilde, Denmark) and cooled in a semi-programmable freezer (Nicool, LM-10, Air Liquid, Paris, France). The tubes were cooled from room temperature to 26°C at a rate of 1.7°C/min, then to 2100°C at a rate of 5°C/min. The tubes were then soaked directly in liquid nitrogen (2196°C) for storage (Yavetz et al, 1991). Thawing of testicular tissue was performed at room temperature on the day of ICSI. To remove the cryopreservation medium, the sample was diluted with medium and centrifuged. The pellet was then resuspended in fresh medium and examined for the presence of motile spermatozoa in the same manner as the fresh testicular specimen. Most of the vital spermatozoa acquired motility within 1–2 hours of culture.
Ovarian Stimulation and ICSI Procedure
Controlled ovarian stimulation was achieved using gonadotrophin-releasing hormone analogue (Decapeptyl, Ferring, Kiel, Germany) and human menopausal gonadotrophin (Menogon, Kiel Ferring) or recombinant follicle-stimulating hormone (FSH; Gonal F, Serono Aubonne, Switzerland). Oocyte retrieval was scheduled 34–35 hours after the administration of human chorionic gonadotrophin (hCG; Pregnyl 10,000 IU, Organon, Cambridge, United Kingdom or Ovitrelle 250 μg, Serono). The cumulus-oocyte complexes were isolated into mHTF (Irvine Scientific). A single motile spermatozoon was aspirated from the separate sperm droplet into the injection pipette and then transferred to the 10% PVP droplet to separate it from attaching cells and debris. When only completely immotile sperm cells were obtained, even after an exhaustive search, spermatozoa with normal morphology were selected for the microinjection procedure. Both motile and immotile spermatozoa were immobilized before their aspiration into the injection pipette. The isolated spermatozoon was then injected into the metaphase II (MII) oocyte retrieved from the spouse.
ICSI was performed at the same time, on the day of oocyte retrieval. The oocytes were denuded of cumulus cells with hyaluronidase and a fine hand-drawn glass Pasteur pipette. The procedure itself was performed by means of a Nikon inverted microscope (Diaphot 300, Nikon, Tokyo, Japan) with Narishige micromanipulators.
Fertilization and Embryo Assessment
Fertilization was checked 18–20 hours after ICSI using an inverted microscope at ×400 magnification (TE 200, Nikon). Each embryo was incubated in a separate droplet of medium covered with oil to allow individual assessment and documentation at different stages during preimplantation development. Morphological assessments were always performed at several time points between days 1 and 3 of development. Day 2 and day 3 assessments included the number of blastomeres, the degree of fragmentation, and the extent of compaction. Specifically, the cleavage rate was defined as normal when embryos reached the 4–5-cell stage on the morning of day 2 or when they reached the 7–8-cell stage on the morning of day 3. The degree of fragmentation was scored as 0 = none, 1 ≤ 10%, 2 = 10%–20%, 3 = 21%–60%, and 4 ≥ 60%. Compaction was described as being either partial or full.
Embryo transfer was performed on day 2 or day 3 of development, and serum βhCG was measured 12 days later. Implantation was confirmed when an intrauterine gestational sac was documented by sonography 4 weeks after embryo transfer.
The number of fertilized oocytes was compared among the 3 groups (ejaculate, TESE fresh, and TESE frozen) by the chisquare test or Fisher's exact test (as applicable) for all parameters (ie, MII oocytes injected with motile sperm, fertilization with motile sperm, fertilization with nonmotile sperm, high-quality embryos, implantation, normal cleavage, and pregnancies). Pairwise comparisons were performed when the results of the overall test were significant. The False Discovery Rate method for adjustment of significance level was used. To correlate between FSH levels and testicular size, Spearman Rates Correlation test was used. Continuous variables are expressed as x̄ ± standard deviation. SAS for Windows version 9.1.3 was used for all statistical analyses.
Baseline characteristics of the study groups are presented in Table 2. The female partner's age did not differ between the ejaculate (32.7 ± 4.5) and both the fresh and thawed TESE groups (33.3 ± 5.7 years and 34.4 ± 5.2 years, respectively, P not significant [NS]). The day 3 FSH levels of the female partners ranged from 4.4 to 11.8 mIU/mL (average 6.82 ± 2.13 mIU/mL). Estradiol and progesterone levels on the day of hCG administration, the mean number of retrieved oocytes per cycle and the average number of mature MII oocytes were also similar.
Table 2. . Baseline characteristics of the study groups
|Cycles, n||34||59|| |
|Age of female partner, y||32.7 ± 4.5||34.1 + 5.1||NS|
|Serum estradiol on day of hCG injection, pg/mL||1585 ± 682||1722 + 441||NS|
|Serum progesterone on day of hCG injection, ng/mL||1.3 ± 0.7||1.6 + 0.8||NS|
|Retrieved oocytes, n||11.2 ± 5.7||11.0 + 6.4||NS|
|Metaphase II oocytes, n||9.0 ± 5.1||9.1 + 5.8||NS|
The age of the male partners ranged from 28 to 52 years (average 36.7 ± 8.3 years), and their FSH levels ranged from 8 to 45 (average 19.1 ± 7.6 mIU/mL). No correlation was found between testicular size and FSH levels (correlations between FSH and left or right testes were 20.114 and 0.217, respectively, NS, Spearman rank correlation test). All types of testicular histology were found (normal, 2; maturation arrest, 5; Sertoli cell only, 3; and mixed atrophy, 3). Normal or abnormal FSH levels were found in all types of testicular histology. In 2 cases with normal FSH levels and relatively normal testicular histology, finding sperm cells was easier than in other cases. Yet, in other cases with normal FSH levels, finding sperm cells in the testicular minced specimens was long and tedious.
Significantly fewer oocytes were injected with motile sperm cells in the frozen TESE group compared with the fresh ejaculate and fresh TESE groups because of the lack of motile sperm cells after thawing of the testicular specimens (Table 3). Nevertheless, the fertilization rates in both the fresh and the frozen-thawed testicular sperm cells were significantly higher than when ejaculated sperm cells were used. This difference reached a level of significance in cycles where motile sperm cells were available in the fresh TESE, the frozen TESE, and the ejaculate groups (50.0%, 46.7%, and 38.2%, respectively, P < .05), as well as in cycles in which only nonmotile sperm cells were available (37.5%, 28.7%, and 13.6%, respectively, P < .05; Table 3). Likewise, the quality of the embryos that developed after microinjection of testicular (fresh and frozen) sperm was higher compared with the ejaculate group (58.3% and 66.7% vs 53.2%, respectively, P = .065; Table 3), reaching a level of significance when the 2 TESE groups were combined (65.3% high-quality embryos) compared with the ejaculate group (53.2% high-quality embryos; P < .05).
Table 3. . Fertilization and embryo development
|Cycles, n||34||9||50|| |
|% metaphase II oocytes injected with motile sperm||281/305 (92.1%)||75/82 (91.2%)||321/457 (70.2%)a||<0.001|
|Fertilization rate with motile sperm (motile/injected)||107/281 (38.2%)a||37/74 (50%)||150/321 (46.7%)||<0.05|
|Fertilization rate with nonmotile sperm (nonmotile/injected)||3/22 (13.6%)a||3/8 (37.5%)||39/136 (28.7%)||<0.05|
|Normal cleavage rateb (%)||61/112 (54.5%)||19/35 (54.3%)||112/179 (62.6%)||NS|
|High-quality embryosc (%)||59/111 (53.2%)||21/36 (58.3%)||122/183 (66.7%)||=0.065|
|Average no. of embryos transferred per cycle (±SD)||2.3 ± 1.7||3.0 ± 2.3||2.8 ± 1.4||NS|
Normal cleavage rate and the average number of embryos transferred per cycle were similar among the 3 groups (Table 3). Thirteen clinical pregnancies were achieved. Four clinical pregnancies resulted in 4 singleton deliveries in the ejaculate group. The fresh TESE group had 3 clinical pregnancies, of which 2 were singleton pregnancies and 1 was a triplet pregnancy that turned into a twin pregnancy when 1 fetus vanished. The use of frozen TESE resulted in 3 chemical pregnancies and another 6 clinical pregnancies, of which 1 resulted in a twin delivery and 5 singleton deliveries (Table 4). The implantation rate was significantly higher in the fresh TESE group (18.5%) than in the frozen TESE or ejaculate groups (5.7% and 5.1%, respectively, P < .05; Table 4).
Table 4. . Pregnancy rates
|Cycles, n||34||9||50|| |
|Pregnancy rate (βhCG/ET)||4/28 (14.3%)||3/7 (42.9%)||9/47 (19.1%)||NS|
|Ongoing pregnancy (rate/ET)||4/28 (14.3%)||3/7 (42.9%)||6/47 (12.8%)||NS|
|Take-home baby (rate/ET)||4/28 (14.3%)||3/7 (42.9%)||6/47 (12.8%)||NS|
|Implantation rate||4/79 (5.1%)||5/27a (18.5%)||8/140b (5.7%)||0.041c|
In the present study, in only 70.2% of the oocytes, motile sperm cells were available for ICSI in the frozen TESE group compared with 91.2% for the fresh TESE group and 92.1% for the ejaculate group (P < .001). This reflects a loss of motility in the thawed testicular sperm specimens. Because a significantly higher fertilization rate was achieved in both TESE groups compared with the ejaculate group when nonmotile sperm cells were used, however, nonmotile sperm cells in the thawed specimens can be considered as still having maintained their fertility potential for ICSI, unlike ejaculated sperm cells, in which nonmotility is a reflection of reduction of its fertility potential. The fact that TESE sperm produced higher quality embryos in this group of virtually azoospermic and cryptozoospermic patients might also indicate that a higher rate of nonmotility of testicular sperm cells does not reduce the overall fertility outcome. In a previous study, we had shown that sperm motility is an important parameter contributing to fertility outcome when either fresh or frozen testicular sperm cells were used (Ben-Yosef et al, 1999). In the present study, however, the fertility results were still better for the TESE groups, even though more motile sperm cells were available in the ejaculate in patients with virtual azoospermia and cryptozoospermia, which could suggest that the source of sperm cells, namely the testis, plays a more crucial role in fertility outcome than their motility in this unique study group.
The finding of mature sperm cells in the ejaculate of a man formerly diagnosed as being azoospermic is not unique and probably dependent on the techniques used to determine azoospermia. Using an extended sperm preparation, Ron-El et al (1997) found mature sperm cells in the ejaculates of 17 out of 49 (35%) patients previously diagnosed as azoospermic. Jaffe et al (1998) described a technique using semen centrifuging (sperm pelleting) in 70 men with nonobstructive azoospermia based on routine semen analysis. Sperm cells were identified in 16 of these 70 (22.8%) patients, 15 of whom had motile sperm cells.
A similar technique, which is used in our laboratory, combines a thorough search of the pellet after high centrifugation with the examination of Papanicolaoustained slides of the ejaculate specimens (Hauser et al, 2005).
A unique evolution of virtual azoospermia was described in a case report of a pregnancy and subsequent birth of healthy twins after ICSI of ejaculated motile spermatozoa 6 hours after unsuccessful TESE where testicular sperm cells could not be extracted in a patient with infertility due to nonobstructive azoospermia (Zadori et al, 2003). Fertility outcome when ejaculate sperm is used for ICSI in cases with severe male infertility is controversial. Nagy et al (1995) showed that high fertilization and pregnancy rates were obtained by ICSI even in the most extreme cases when cryptozoospermia or total asthenozoospermia or total teratozoospermia was diagnosed in the initial semen sample. Contrarily, Strassburger et al (2000) claimed that an extremely low sperm count has a negative effect on the outcome of ICSI.
Few studies have compared the fertility outcome of ejaculate with testicular sperm cells. Weissman et al (2008) reported a series of 4 couples with long-standing male factor infertility and multiple failed IVF/ICSI cycles with poor embryo quality and repeated implantation failure using motile ejaculatory sperm cells. The use of fresh testicular sperm cells resulted in pregnancies in all cases. Bendikson et al (2008) compared the fertility outcome of 16 couples with cryptozoospermia who underwent ICSI cycles using either ejaculate or fresh testicular sperm cells. There was no significant difference in the fertilization rate, although the rate of clinical pregnancies showed a trend in favor of the group in which testicular sperm cells were used (47.4% vs 20.8%). A comparison between the first testicular cycle and the ejaculated cycle closest in time to the cycle with testicular spermatozoa in a subgroup of cases revealed a higher rate of normal fertilization with testicular spermatozoa (60.9% vs 48.5%, P < .05; Bendikson et al, 2008).
In the Hourvitz et al (1998) comparative study of ICSI results, 2 groups of subjects were inseminated by either fresh testicular or ejaculated sperm. Although the oocyte fertilization rate was significantly lower in the testicular sperm group (42% vs 55.5%, P < .005), the number of cycles with embryo transfer, the mean cleavage rate per cycle, embryo quality, and clinical pregnancy rates (22.5% in the TESE group and 20% in the ejaculate group) were not significantly different between groups. These results are different from ours, but each couple in their study was treated by a single sperm source, either testicular or ejaculate, unlike the several options in our study. Our results are in agreement with those of Bendikson et al (2008), who compared the outcome of ICSI in 16 couples using fresh testicular or ejaculate sperm cells in cases of cryptozoospermia.
The present study involves patients that were treated over a period of 13 years. Yet, we believe that this study group can be analyzed as a single group, since throughout the whole period of the study we used ICSI for all TESE patients in our IVF unit. This technique has been well established in our lab since 1995, and the IVF outcome did not change significantly during this period, as evident in previously published studies (Hauser et al, 1998, 2005; Ben-Yosef et al, 1999).
Another issue that might be influenced by the relatively long study period is female age, which has a substantial effect on fertility outcome. Some patients in our study were treated over a period of several years, as indicated in Table 1. However, those cycles that resulted in pregnancies were sometimes achieved in the same patient at an older age (see patients 9–11 and 13) independently of the sperm source used. Furthermore, in 2 cases only, treatments stretched beyond the age of 40. In one of them (patient 2), no pregnancy was achieved even in earlier treatments, and in the other (patient 9), it was achieved in later treatments only. These results demonstrate that oocyte aging was not the critical parameter responsible for the success of the treatment, in these cases.
A clear limitation of the present study is the low number of cases included because of the relative rarity of virtual azoospermia and cryptorchidism in the overall population. However, the results of the present study may assist physicians in choosing the most efficient treatment for such cases.
In the present study, no correlation was found between testicular size and FSH levels. Normal FSH levels were found in the cases with normal spermatogenesis, but also in patients with pathological testicular histology. This indicates that testicular histology could not be predicted before TESE and agrees with our previous results (Hauser et al, 1995).
We had previously shown no difference in fertility outcome when fresh and frozen-thawed testicular spermatozoa were used in a group of nonobstructive azoospermic patients in whom isolated or only nonmotile sperm cells were retrieved (Hauser et al, 2005). Those patients differ from our current cohort because they all had azoospermia and no sperm could be found in their ejaculates. This explains why fresh testicular sperm cells were not used in all cases, and why TESE procedures were also performed electively. In these cases, we used only thawed testicular sperm cells. In the present study, the use of fresh testicular sperm cells in cases of virtual azoospermia and cryptozoospermia yielded better implantation rates compared with frozen testicular sperm cells and with ejaculated sperm cells as well. On the basis of our results, in spite of the ease of obtaining motile sperm samples and of laboratory manipulations in ICSI procedures whenever ejaculated sperm is available, fresh testicular sperm should still be considered first for ICSI in patients with virtual azoospermia and cryptozoospermia because of their superior fertility potential. Offering such a treatment method that involves an operating procedure should be considered to provide the best success possible and to achieve a pregnancy in countries in which cost considerations of IVF procedures play a role.