ICSI outcomes in obstructive azoospermia: influence of the origin of surgically retrieved spermatozoa and the cause of obstruction

Transcription

1 Human Reproduction Vol.21, No.4 pp , 2006 Advance Access publication December 16, doi: /humrep/dei418 ICSI outcomes in obstructive azoospermia: influence of the origin of surgically retrieved and the cause of obstruction C.Buffat 1, C.Patrat 1,6, F.Merlet 1, J.Guibert 2, S.Epelboin 4, N.Thiounn 5, A.Vieillefond 3, A.Adda-Lievin 1, C.Lebon 1 and P.Jouannet 1 1 Laboratoire de Biologie de la Reproduction; CECOS, Hôpital Cochin Saint Vincent de Paul, Hôpitaux de Paris Université Paris V, 2 Service de Gynécologie Obstétrique and 3 Service d Anatomo-Pathologie, Hôpital Cochin, 4 Service de Gynécologie Obstétrique, Hôpital Saint Vincent de Paul, and 5 Service d Urologie, Hôpital Necker, AP-HP, Paris, France 6 To whom correspondence should be addressed. catherine.patrat@cch.ap-hop-paris.fr BACKGROUND: Spermatozoa can be retrieved from the testis and epididymis of men with obstructive azoospermia (OA) and used for ICSI. However, it is unknown whether the outcome of ICSI depends on the cause of obstruction or the origin of surgically retrieved. METHODS: A cohort of 171 men with OA and normal spermatogenesis were included in this retrospective study. They were divided into three groups according to the site and origin of obstruction: 83 men had congenital bilateral absence of vas deferens; 55 and 33 had acquired epididymal and deferent duct obstructions, respectively. The outcome of 368 ICSI cycles was determined and compared according to the origin of : epididymal (n = 253) or testicular (n = 115). RESULTS: Fertilization and clinical pregnancy rates did not differ between of different origin (58.9% versus 51.9% and 22.1% versus 24.3% with epididymal and testicular, respectively). However, the miscarriage rate was significantly higher for testicular (35.7% versus. 12.5% P < 0.05, 2 test). Findings were similar whatever the aetiology of the OA. CONCLUSION: This study suggests that the use of testicular, even those generated during normal spermatogenesis, alters embryonic development and that epididymal should be preferentially used, irrespective of the aetiology of OA. Key words: ICSI/microsurgical epididymal sperm aspiration/obstructive azoospermia/testicular sperm extraction Introduction Ten percent of male infertility cases involve azoospermia (Hendry, 1994), which is classified as either obstructive or non obstructive. Obstructive azoospermia is characterized by an occlusion or an absence of the reproductive tract preventing sperm cells from entering the seminal plasma before ejaculation. Non obstructive azoospermia is caused by absent or severely reduced sperm production by the testis, such that the ejaculate does not contain sperm (Jarow et al., 1989). When the technology was developed to obtain pregnancies by intracytoplasmic injection of surgically retrieved from the testis or the epididymis, new hope was given to azoospermic men that they would to be able to use their own gametes to become fathers (Craft et al., 1993; Silber et al., 1994). Spermatozoa can be obtained by microsurgical epididymal sperm aspiration (MESA) in cases of azoospermia due to a congenital bilateral absence of the vas deferens, failed vasoepididymostomy and otherwise irreparable obstruction (Tournaye et al., 1994). Spermatozoa can also be collected straight from the testis if the epididymis is inaccessible or if no motile are present in the epididymis (Devroey et al., 1994; Nagy et al., 1995). High fertilization and good pregnancy rates can be obtained for both obstructive and non-obstructive azoospermia (Dohle et al., 1998; Palermo et al., 1999), although the quality of collected from the epididymis may differ from that collected from the testis. High abortion rates were reported when using testicular (Ghazzawi et al., 1998; Pasqualotto et al., 2002), suggesting that the male gamete could affect the developmental competence of resulting embryos. However, in most studies, it was not possible to determine whether the miscarriage rate (MR) associated with the use of testicular was related to a defect of spermatogenesis or to sperm immaturity. In this study we compared the outcomes of ICSI for men with obstructive azoospermia and normal spermatogenesis according to the use of epididymal or testicular and the cause of obstruction The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 ICSI outcome according to the cause of obstructive azoospermia Materials and methods Patients This retrospective study included 171 men presenting obstructive azoospermia at one of the two infertility centers, at Cochin or Saint Vincent de Paul Hospitals, between January 1996 and December Most (91%) suffered from primary infertility of 1 to 18 years duration. The mean age of the men was 36.7 ± 6.4 years (range: years) at the time of semen collection. Azoospermia was confirmed on at least two occasions. All men had normal testicular volume and normal serum FSH concentration. Karyotype analysis was performed systematically and was normal in all the men. biopsy revealed normal spermatogenesis in all cases for which it was performed (n = 56). Therefore, it can be presumed that spermatogenesis was normal in all cases. The mean age of the female partners was 32.8 ± 4.4 years (range: years). The participants gave informed consent. Azoospermic men were divided into three groups according to the cause of obstruction. Eighty-three men had a congenital bilateral absence of the vas deferens (CBAVD) diagnosed by a lack of vas deferens at clinical examination, associated with a low semen volume, a semen ph <7 and very low fructose, carnitine and α(1-4) glucosidase concentrations in the seminal plasma. Ultrasound investigation confirmed the diagnosis. Before ICSI was undertaken for these men, the patient and his partner were tested for cystic fibrosis transmembrane conductance regulator (CFTR) gene mutations to allow appropriate genetic counselling. Fifty-five patients had an acquired epididymal obstruction due to a history of epididymitis, a turgescent epididymal head and flat epididymal tail at clinical examination, associated with low concentrations of carnitine and α(1-4) glucosidase but normal fructose in the seminal plasma. In all cases, the epididymal obstruction was confirmed by ultrasound examination or surgical exploration. In 33 patients, the acquired vas deferens obstruction was related to a history of vasectomy (n = 3), or of lower abdominal, scrotal or inguinal surgery. In this last group, the clinical examination revealed normal epididymis; the semen volume was normal with low concentrations of carnitine and α(1-4) glucosidase in the seminal plasma. The time between vasectomy and surgical sperm retrieval was 10, 12 and 15 years (n = 3). In all cases, the diagnosis was also confirmed by ultrasound examination and/or surgical exploration. Surgical sperm retrieval All testicular or epididymal sperm retrievals were performed under general anaesthesia. were aspirated by MESA from the most proximal part of the caput of the epididymis (Tournaye et al., 1994). When MESA did not provide sufficient motile for ICSI or when the status of the epididymis did not allow sperm collection, testicular were extracted from an open biopsy (TESE; Silber et al., 1994). In other cases, an open testicular biopsy was performed because it was technically simpler and less time-consuming. Large numbers of testicular were always collected, reflecting a presumably ongoing spermatogenesis. One small piece of the testicular biopsy was fixed in Bouin s solution for histopathological examination and the remainder was sent to the laboratory for sperm processing. Sperm processing Following MESA, epididymal fluid was diluted in FERTICULT - HEPES (JCD, Lyon, France) and the were prepared immediately for ICSI or for cryopreservation by addition of an equal amount (v/v) of freezing medium (SPERM FREEZE ; JCD) to obtain a final concentration of 15% glycerol. To perform ICSI, fresh epididymal were selected by centrifugation (20 min, 300g) through a two-step PURESPERM (JCD) density gradient (90 45%). The pellet was then washed (10 min, 600g) and diluted in Upgraded B2 INRA medium (CCD, Suresnes, France). A 1μl aliquot of the final suspension was placed in the injection dish. Cryopreserved epididymal samples were processed in the same way as fresh after thawing at room temperature. Following TESE, testicular biopsies were placed into Nunc tubes containing 1 ml of FERTICULT-HEPES medium before being transferred to Petri dishes containing 2 ml of the same pre-incubated medium. They were then crushed with sterile needles to dilacerate the seminiferous tubules and kept at 37 C, under 5% CO 2 in air for at least 1 h to facilitate the dispersal of into the medium. Spermatozoa were then used immediately for ICSI or cryopreserved by addition of an equal amount of cryopreservation medium (SPERM FREEZE). When used immediately for ICSI, the medium of the mixture was layered on a 1ml phase (45%; PURESPERM) for 20 min at 300g and then washed in B2 INRA medium (600g, 10 min.). The pellet was divided into 5μl drops in the cover of a Petri dish. The final suspension was covered with mineral oil (Medicult, Limonest, France) and kept for 1 h at 37 C and under 5% CO 2 in air before examination for the presence of motile (Royere et al., 1996). Cryopreserved testicular samples were processed in the same way as fresh after thawing at room temperature the day of ICSI. Ovarian stimulation and oocyte pick-up All female patients were treated with either a long or a short GnRH analogue (GnRHa, Decapeptyl, Beaufour Ipsen Pharma, Paris, France) protocol, followed by ovarian stimulation with recombinant FSH. Oocytes were retrieved by transvaginal ultrasound-guided puncture of the ovarian follicles 36 h after HCG administration. ICSI, assessment of fertilization, embryo cleavage, quality of embryo at day 2 and clinical pregnancy ICSI was carried out on the stage of an inverted microscope (Nikon) at 400 magnification using the Hoffman Modulation Contrast system. It was performed on all morphologically intact MII oocytes (second metaphase) and always with immobilized motile (Palermo et al., 1996). Oocytes were examined h after microinjection. Normal fertilization was defined as the presence of two distinct pronuclei (PN) and a second polar body. Embryo cleavage was assessed 24 h later and embryos were scored according to the number of blastomeres and to the percentage of enucleate fragments (A: no fragments; B: 1 20% fragments; C: 21 50% fragments; D: >50% fragments). Up to three embryos were transferred into the uterine cavity 48 or 72 h after the oocyte puncture. Supernumerary embryos with <20% fragmentation were frozen and cryopreserved (Mandelbaum et al., 1988). Pregnancy was detected by rising serum HCG concentration in two consecutive assays at least 12 days after embryo transfer. Clinical pregnancy was defined by the presence of at least one gestational sac observed by ultrasound scanning after 7 weeks. Miscarriage was defined as all spontaneous interruption of clinical pregnancy, not related to extra uterine implantation nor medical abortion. Statistical analysis Numerical variables were compared using analysis of variance (ANOVA). Categorical variables were compared using the χ 2 test. All statistical tests were assessed at the 5% level of significance. 1019

3 C.Buffat et al. Table I. ICSI outcomes according to the origin of from 171 obstructive azoospermic patients A No. of oocyte retrievals Age of women (years) (mean ± SD) 33.0 ± ± 4.2 Age of men (years) (mean ± SD) 38.2 ± ± 5.7 No. of oocytes retrieved (mean ± SD) 1164 (10.1 ± 6.2) 2623 (10.4 ± 6) No. of metaphase II (MII) oocytes injected (mean ± SD) 874 (7.6 ± 4.8) 1889 (7.5 ± 4.4) No. of 2 pronuclei (PN) oocytes (%) a 454 (51.9%) 1112 (58.9%) No. of cleaved embryos (%) a 483 (55.2%) 1192 (63.1%) No. of embryos with < 20% fragmentation (%) b 208 (43%) 501 (42%) a Percentage of the total number of MII oocytes injected; b Percentage of the total number of embryos. B p d No. of embryo transfers No. of transferred embryos (mean ± SD/transfer) 232 (2.1 ± 0.7) 526 (2.2 ± 0.8) No. of embryos implanted (%) a 33 (14.2%) 68 (12.9%) No. of clinical pregnancies (%) b 28 (24.3%) 56 (22.1%) No. of miscarriages (%) c 10 (35.7%) 7 (12.5%) < 0.05 No. of extra-uterine pregnancies 1 1 No. of medical abortion 0 2 No. of deliveries (% per cycle) 17 (14.8%) 46 (18.2%) No. of newborns No. of newborns with malformation 1 3 a Ratio between the number of gestational sacs and the number of transferred embryos; b Ratio between the number of clinical pregnancies and the number of oocyte retrievals; c Ratio between the number of spontaneous pregnancy losses before 20 weeks of gestation and number of the clinical pregnancies; d χ 2 test. Results ICSI outcome with testicular or epididymal A total of 115 ICSI cycles were performed with testicular (84 with frozen and 31 with fresh ) and 253 ICSI cycles with epididymal (200 with frozen and 53 with fresh ) (Table IA). A majority of patients (n = 115) had ICSI attempts with only epididymal and 40 with only testicular sperm. Spermatozoa from both origins were used for 16 patients on two different cycles. Mean female and male ages and mean number of oocytes retrieved were similar in the two groups. The fertilization rates, embryo cleavage rates and embryo quality did not differ according to the site of sperm retrieval, either testicular or epididymal (Table IA). No difference in the implantation rate or in the clinical pregnancy and delivery rates were observed. However the MR was significantly higher for retrieved from the testis (35.7%) than for that from the epididymis (12.5%, P < 0.05, Table IB, Fig. 1). When examining ICSI results of only the first rank of attempts, the MR after ICSI with testicular sperm was also higher compared with the miscarriage rate after ICSI with epididymal (40.0 versus 11.8%, P < 0.05; χ 2 test). Similar significantly differences were obtained for the other ranks. Although mean male and female age, and mean rank of ICSI attempts were slightly higher for those pregnancies that ended with a miscarriage, the difference to pregnancies without miscarriage was not significant (Table II; ANOVA). A significantly higher MR was also observed after ICSI using testicular in the subgroup of women under 35 years old (Fig. 1; P < 0.05; χ 2 test). The trend was similar for women who were 35 years old, but the difference not statically 1020 significant because the small size of this last group. The differences observed in MR were not influenced by the long or short GnRHa protocol used before ovarian stimulation. In case of short protocol, the MR was 44.4% when using testicular and 20% when using epididymal (n = 9 and 15 pregnancies, respectively; P = 0.05; χ 2 test). In case of long protocol, the MR were 31.6% when using testicular (n = 19 pregnancies) and 9.7% when using epididymal (n = 41 pregnancies; P < 0.05; χ 2 test). There was no significant difference in the MR coming after ICSI with fresh and cryopreserved sperm (17.3% and 28.5%, respectively; P > 0.05). Four children were born with malformations (two with pylori stenosis, one with periventricular leukomalacia and one with pulmonary artery valvular stenosis). Outcome of transfers with thawed embryos according to the site of retrieval One hundred and ninety-one transfers of thawed embryos, which had been frozen during the same ICSI attempts, were Table II. Comparison of variables according to the outcome of pregnancy Without miscarriage With miscarriage Age of women (years) (mean ± SD) 31.6 ( 0.6 (22 39) 33.3 ( 1.1 (26 39) Age of men (years) (mean ± SD) 36.0 ( 0.7 (27 51) 38.2 ( 1.6 (31 47) Rank of ICSI attempt (mean ± SD) 1.8 ( ( 0.4 ICSI with cryopreserved (%) a a Ratio between the number of attempts performed with cryopreserved and the number or attempts with fresh.

4 ICSI outcome according to the cause of obstructive azoospermia Miscarraige rate (%) n = 28 * * n = 56 n = 20 n = 41 n = 8 All population < 35 years _ 35 years Figure 1. Miscarriage rate (MR) after ICSI according to the sperm origin ( testicular epididymal ) and age of women in obstructive azoospermia. n = number of clinical pregnancies; *P < performed. The mean number of transferred embryos was similar in the two groups. The implantation and clinical pregnancy rates were significantly lower for oocytes injected with testicular (Table III). Moreover, the MR was also significantly higher for oocytes injected with testicular (Table III). Two and 19 deliveries went to term in the testicular and epididymal groups, respectively, resulting in the birth of three and 21 babies, respectively, with no obvious congenital malformations. When cumulating pregnancies obtained after transfer of fresh or frozen-thawed embryos, the MR was 39.4% when using testicular compared with 13.9% when using epididymal. ICSI outcome according to the cause of azoospermia and the site of retrieval The ICSI outcome was evaluated according to the site of sperm retrieval for each aetiological group (Tables IVA and IVB). Mean female and male ages were similar in all groups (Tables IVA and IVB). In the CBAVD group, the 2 PN fertilization, > n = 15 cleaved embryo (Table IVA) and delivery rates were significantly higher for epididymal whereas the MR was significantly higher for retrieved from testis (Tables IVB). For cases of azoospermia related to acquired epididymal or deferent duct obstructions, no difference was observed in 2 PN fertilization or embryo cleavage rates, or in embryo quality, according to the sperm origin (Table IVA). However, the MR was again significantly higher with testicular than with epididymal (Table IVB). The influence of the site of obstruction was also evaluated by analysing ICSI results with epididymal in each aetiological group. The percentage of good quality embryos and clinical pregnancy rate were significantly lower for ICSI with epididymal from cases of acquired epididymal obstruction than from that from the two other types of obstructive azoospermia (Tables IVA and IVB). Discussion The use of surgical sperm retrieval and ICSI has greatly improved treatment of azoospermic men, but the consequences of using immature sperm retrieved directly from the testis with normal or defective function are not fully known. We studied ICSI outcomes according to the site of sperm retrieval (testis or epididymis) and the aetiology of the obstructive azoospermia by analysing 368 consecutive ICSI cycles involving 171 obstructive azoospermic men with normal spermatogenesis. This is the largest series since a previous published report of a series of ICSI cycles in case of obstructive azoospermia (reviewed by Nicopoullos et al., 2004b). A number of methods have been proposed to surgically recover for ICSI. These methods include MESA, PESA (percutaneous epididymal sperm aspiration; Shrivastav et al., 1994), TESE or TESA (testicular sperm aspiration; Gorgy et al., 1998). The aim of this study was not to compare these various methods. The standard MESA procedure is a reliable technique that allows the retrieval of the maximum number of, which may be cryopreserved for several ICSI if needed. In our experience, up to five ICSI attempts could be performed and 43 mature oocytes could be injected after only one MESA. It allows the surgeon to make a complete scrotal exploration and, whenever indicated, Table III. Outcome of transfers of thawed embryos according to the source of P Age of women (years) (mean ± SD) 32.2 ± ± 4.2 Age of men (years) (mean ± SD) 37.4 ± ± 5.6 No. of embryo transfers No. of transferred embryos (mean ± SD/transfer) 101 (1.8 ± 0.7) 256 (1.8 ± 0.7) No. of thawed embryos implanted (%) a 6 (5.9%) 27 (10.5%) <0.05 e No. of clinical pregnancies (%) b 5 (8.4%) 23 (16.3%) <0.05 e No. of miscarriages (%) c 3 (60%) 4 (17.4%) <0.05 f No. of deliveries (%) d 2 (3.6%) 19 (13.9%) No. of newborns 3 21 a Ratio between the number of gestational sacs and the number of transferred embryos; b Ratio between the number of clinical pregnancies and the number of oocyte retrievals; c Ratio between the number of spontaneous pregnancy losses before 20 weeks of gestation and the number of clinical pregnancies; d Ratio between the number of deliveries and the number of oocyte retrievals. e χ 2 test; f χ 2 test for small samples. 1021

5 C.Buffat et al. Table IVA. ICSI outcomes according to the type of obstructive azoospermia and the origin of used Groups CBAVD Acquired epididymal obstruction Acquired vas deferens obstruction No. of oocyte retrievals Age of women a (mean ± SD) 33.4 ± ± ± ± ± ± 3.8 Age of men a (mean ± SD) 36.1 ± ± ± ± ± ± 5.0 No. of oocytes retrieved (mean ± SD) 468 (10.6 ± 7.1) 1397 (10.5 ± 5.7) 488 (9.2 ± 6.0) 764 (10.0 ± 6.2) 208 (11.5 ± 4.0) 462 (10.5 ± 6.4) No. of MII b oocytes injected (mean ± SD) 343 (7.8 ± 5.1) 995 (7.5 ± 3.9) 386 (7.3 ± 5.0) 568 (7.4 ± 4.8) 145 (8.0 ± 3.6) 326 (7.4 ± 5.1) No. of 2 PN c oocytes (%) d 169 (49.3%) f 612 (61.5%) f,h 202 (52.3%) 330 (58.1%) 83 (57.2%) 170 (52.1%) h No. of embryos (%) d 182 (53.0%) g 665 (66.8%) g,i,j 217 (56.2%) 345 (60.7%) i 84 (57.9%) 182 (55.8%) j No. of embryos with <20% fragmentation (%) e 82 (45.0%) 293 (44.0%) k 89 (41.0%) 125 (36.2%) k,l 37 (44.0%) 83 (45.6%) l CBAVD = congenital bilateral absence of vas deferens; a years; b metaphase II; c pronuclei; d of the total injected MII oocytes; e of the total embryos; g,i,k,l p < 0.05; f,h p < 0.01; j p < Table IVB. Groups CABVD Acquired epididymal obstruction Acquired deferent duct obstruction a Defined as the ratio between the number of gestational sacs and the number of transferred embryos; b defined as the ratio between the number of clinical pregnancies and the number of oocyte retrievals; c defined as the ratio between the number of spontaneous pregnancy losses before 20 weeks of gestation and number of the clinical pregnancies; d defined as the ratio between the number of deliveries and the number of oocyte retrievals; e,f,g,h,i p < 0.05; j p < a vasoepididymostomy may be performed concomitantly. Furthermore, if motile epididymal can not be collected, testicular extraction of sperm can be carried out during the same surgical procedure. However, MESA is more invasive and time-consuming and may cause more discomfort for the patient. In our study, 2 PN fertilization, cleavage, embryo quality, implantation, clinical pregnancy and delivery rates were similar for retrieved from the testis and the epididymis. However, we found that the use of testicular for ICSI was associated with a rate of pregnancy loss that was significantly higher than that for epididymal (35.7% versus 12.5%, P < 0.05). Similar results were found after transfer of both fresh and frozen-thawed embryos, whatever the aetiology of obstructive azoospermia. Given the length of the studied period and even though the ICSI practitioners were already trained when the study started, data were re-analysed omitting the first and the second year of the study. We found similar observations leading to similar conclusions since 38.4% and 13.9% of miscarriage rates were observed when using testicular (n = 101) and epididymal (n = 159) sperm, respectively, for ICSI (P < 0.05; χ 2 test). Thus, an hypothesized bias due to a learning period No. of embryo transfers No. of transferred embryos (mean ± SD/transfer) 84 (2.0 ± 0.7) 278 (2.2 ± 0.7) 112 (2.2 ± 0.7) 159 (2.2 ± 0.8) 36 (2.0 ± 0.5) 89 (2.3 ± 0.9) No. of embryos implanted (%) a 8 (9.5%) 38 (13.6%) 17 (15.1%) 18 (11.3%) 8 (22.2%) 18 (20.2%) No. of clinical pregnancies (%) b 7 (15.9%) 33 (24.8%) i 14 (26.4%) e 9 (11.8%) e,i,j 7 (38.8%) 14 (31.8%) j No. of miscarriages (%) c 2 (28.6%) f 6 (18.1%) f 6 (42.8%) g 1 (11.1%) g 2 (28.5%) h 0 h No. of extra-uterine pregnancies No. of medical abortion No. of deliveries (%) d 4 (9.0%) 24 (18%) 8 (15.1%) 8 (10.5%) j 5 (27.7%) 14 (31.8%) j No. of newborns No. of newborns with malformation during the study had no influence on the results. Similarly, to avoid possible bias related to the rank of treatment cycle and the fact that a proportion of treatments were in the same men twice or more, we also analysed only the results of the first ICSI attempt for each couple; the findings were similar (40% versus 12.5% of miscarriage after ICSI with testicular and epididymal, respectively, P < 0.05, n = 171). The difference in MR could not be explained by the age of women since it was also observed when analysing the subgroup of women who were <35 years old at the time of ICSI. It could also not be explained by the treatment used for ovarian stimulation. The proportion of miscarriage when using testicular or epididymal is very different in the current literature. Comparison with published data is difficult since the series are most often small and are reporting mixed cases of obstructive and non-obstructive azoospermia. Furthermore, the origin of the is not always mentioned. In our study, the 12.5% MR observed in pregnancies obtained with epididymal in case of obstructive azoospermia is within the previously reported range varying from 4.6% (Friedler et al; 2002) to 25% (Balaban et al., 2001). Griffith et al. (2004), analysing ICSI data using epididymal or testicular 1022

6 ICSI outcome according to the cause of obstructive azoospermia sperm from obstructive azoospermic sperm, did not find any difference in the MR according to the origin of. However, the series is small (28 treatment cycles and 12 pregnancies with epididymal sperm and 14 treatment cycles and six pregnancies with testicular sperm, respectively) and, consequently, it is difficult to draw any firm conclusion. The 35.7% MR observed in pregnancies obtained with testicular is also in the previously reported range which varies from 12.5% (Palermo et al., 1999) to 40% (Nicopoullos et al., 2004) whatever the nature of azoospermia. In a larger series of 180 clinical pregnancies obtained after ICSI with testicular for obstructive azoospermia, a 14% clinical MR was reported (Vernaeve et al., 2003b); however, this study made comparison with epididymal. The outcome of ICSI using non-ejaculated sperm may be influenced by various factors related to the female partner, sperm quality and the source of surgically retrieved sperm. The female partner s age or ovarian reserve defined as the number of oocytes available for ICSI may significantly contribute to the pregnancy outcome and delivery after ICSI (Silber et al., 1997; Friedler et al., 2002). Furthermore, high maternal age is an important risk factor for early pregnancy loss (Ziebe et al., 2001). In our study, there were no significant differences in maternal age, female infertility factor rates and mean number of retrieved or micro-injected oocytes for ICSI with testicular and epididymal sperm. Could the use of frozen-thawed sperm influence ICSI and pregnancy outcome? We did not find any differences in ICSI outcomes related to the use of fresh or frozen-thawed, whether epididymal or testicular. This result is in accordance with most published studies, which report that sperm cryopreservation does not affect fertilization, embryo cleavage, pregnancy or abortion rates compared with fresh for either epididymal (Nagy et al., 1995; Silber et al., 1997; Tournaye et al., 1999; Habermann et al., 2000; Cayan et al., 2001; Friedler et al., 2002; Nicopoullos et al., 2004a) or testicular (Fischer et al., 1996; Ben-Yosef et al., 1999; Palermo et al., 1999; Friedler et al., 2002; Nicopoullos et al., 2004a). Could the source of influence the ICSI outcome? Published results concerning fertilization, pregnancy and miscarriage rates after ICSI with testicular are discordant. Some studies have reported similar fertilization and pregnancy rates for ICSI with ejaculated and epididymal (Ghazzawi et al., 1998; Monzo et al., 2001; Windt et al., 2002; Friedler et al., 2002). Other studies report impaired fertilization rates with epididymal but similar pregnancy rates (Palermo et al., 1999; De Croo et al., 2000), and others similar fertilization rates but impaired pregnancy rates (Osmanagaoglu et al., 2003; Vernaeve et al., 2003a). Miscarriage rates after ICSI with testicular have been reported to be higher (Ghazzawi et al., 1998; Pasqualotto et al., 2002) or similar (Palermo et al., 1999; Monzo et al., 2001; Friedler et al., 2002; Nicopoullos et al., 2004a,b) to ICSI with ejaculated or epididymal sperm. It is difficult to interpret these results and their causes because, apart from two studies (Palermo et al., 1999; Nicopoullos et al., 2004a), there was no systematic distinction between obstructive and non-obstructive azoospermia. This prevents consideration of the respective influence of spermatogenesis defects and of immaturity in miscarriage. Indeed, produced by defective spermatogenesis may be associated with a higher chromosomal aneuploidy rate (Mateizel et al., 2002) or other genetic alterations and may be responsible for embryos being less able to develop to the blastocyst stage (Balaban et al., 2001) and to implant (Ubaldi et al., 1999). Our results show that testicular from normal spermatogenesis allow similar fertilization, early embryo preimplantation development and pregnancy rates as epididymal but that they are responsible for a significantly higher MR, as previously suggested by Pasqualotto et al. (2002). Failure of embryo development after ICSI could be related to DNA damage (Morris et al., 2002), but it has been previously shown that DNA fragmentation as measured by Tunel assay and Comet assay was higher in epididymal than in testicular of men with obstructive azoospermia (Steele et al., 1999; Ramos et al., 2002). Therefore, other sperm defects might be responsible for the higher MR when using testicular. We also found that the cause of obstruction may influence ICSI outcome when using epididymal. In this study, the embryo quality and pregnancy rates were lower after ICSI with epididymal in cases of acquired epididymal obstruction compared with CABVD or acquired deferent duct obstruction groups. However, no difference was found in the MR between the three groups. The epididymis is generally damaged because of infection or inflammation in case of acquired epididymal obstruction. Our results and those of others (Yamamoto et al., 1996; Nicopoullos et al., 2004b; Pasqualotto et al., 2005) suggest that there may be worse than normal results for ICSI with epididymal when the cause of obstructive azoospermia is infective or inflammatory. This tendency for poor ICSI results with epididymal in cases of obstructive azoospermia of infective origin requires further study and raises several questions, including what is the best source of to use in these cases. In summary, our results show that the origin of surgically retrieved has an effect on the success of assisted reproduction, even when spermatogenesis is presumably normal. The significantly higher MR after ICSI with testicular than with epididymal in cases of obstructive azoospermia suggests that the immaturity of testicular may affect embryo development. More studies are required to determine the mechanisms of paternal influence on embryo development and long-term follow-up studies are needed to prove the safety of testicular sperm retrieval in assisted reproductive techniques. Acknowledgements We thank Rachel Lévy for fruitful and critical discussions. 1023

7 C.Buffat et al. References Balaban B, Urman B, Isiklar A, Alatas C, Mercan R, Asksoy S and Nuhoglu A (2001) Blastocyst transfer following intracytoplasmic injection of ejaculated, epididymal or testicular. Hum Reprod 16, Ben-Yosef D, Yogev L, Hauser R, Yavetz H, Azem F, Yovel I, Lessing JB and Amit A (1999) sperm retrieval and cryopreservation prior to initiating ovarian stimulation as the first line approach in patients with non-obstructive azoospermia. Hum Reprod 14, Cayan S, Lee D, Conaghan J, Givens CA, Ryan IP, Schriock ED and Turek PJ (2001) A comparison of ICSI outcomes with fresh and cryopreserved epididymal from the same couples. Hum Reprod 16, Craft I, Tsirigotis M and Nicholson N (1993) Fertilizing ability of testicular (letter) Lancet 342,864. De Croo I, Van der Elst J, Everaert K, De Sutter P and Dhont M (2000) Fertilization, pregnancy and embryo implantation rates after ICSI in cases of obstructive and non-obstructive azoospermia. Hum Reprod 15, Devroey P, Liu J, Nagy Z, Tournaye H, Silber SJ and Van Steirteghem AC (1994) Normal fertilization of human oocytes after testicular sperm extraction and intracytoplasmic sperm injection. Fertil Steril 62, Dohle GR, Ramos L, Pieters MH, Braat DD and Weber RF (1998) Surgical sperm retrieval and intracytoplasmic sperm injection as treatment of obstructive azoospermia. Hum Reprod 13, Fischer R, Baukloh V, Naether OG, Schulze W, Salzbrunn A and Benson DM (1996) Pregnancy after intracytoplasmic sperm injection of extracted from frozen-thawed testicular biopsy. Hum Reprod 11, Friedler S, Raziel A, Strassburger D, Schachter M, Soffer Y and Ron-El R (2002) Factors influencing the outcome of ICSI in patients with obstructive and non-obstructive azoospermia: a comparative study. Hum Reprod 17, Jarow JP, Espeland MA and Lipshultz LI (1989) Evaluation of the azoospermic patient. J Urol 142, Ghazzawi IM, Saeeaf MG, Taher MR and Khalifa FA (1998) Comparison of fertilizing capability of from ejaculates, epididymal aspirates and testicular biopsies using intracytoplasmic sperm injection. Hum Reprod 13, Gorgy A, Meniru GI, Naumann N, Beski S, Bates S and Craft I (1998) The efficacy of local anaesthesia for percutaneous epididymal sperm aspiration and testicular sperm aspiration. Hum Reprod 13, Griffiths M, Kennedy CR, Rai J, Wilson L, Blacklock AR and Keay SD (2004) Should cryopreserved epididymal or testicular sperm be recovered from obstructive azoospermic men for ICSI? BJOG 111, Habermann H, Seo R, Cieslak J, Niederberger C, Prins GS and Ross L (2000) In vitro fertilization outcomes after intracytoplasmic sperm injection with fresh or frozen-thawed testicular. Fertil Steril 73, Hendry WF (1994) Vasectomy and vasectomy reversal. Br J Urol 73, Mandelbaum J, Junca AM, Plachot M, Alnot MO, Salat-Baroux J, Alvarez S, Tibi C, Cohen J, Debache C and Tesquier L (1988) Cryopreservation of human embryos and oocytes. Hum Reprod 3, Mateizel I, Verheyen G, Van Assche E, Tournaye H, Libaers I and Van Steirteghem A (2002) FISH analysis of chromosome X, Y and 18 abnormalities in testicular sperm from azoospermic patients. Hum Reprod 17, Monzo A, Kondylis F, Lynch D, Mayer J, Jones E, Nehchiri F, Morshedi M, Schuffner A, Muasher S, Gibbons W and Oehninger S (2001) Outcome of intracytoplasmic sperm injection in azoospermic patients: stressing the liaison between urologist and reproductive medicine specialist. Urology 58, Morris ID, Ilott S, Dixon L and Brison DR (2002) The spectrum of DNA damage in human sperm assessed by single gel electrophoresis (Comet assay) and its relationship to fertilization and embryo development. Hum Reprod 17, Nagy Z, Liu J, Cecile J, Silber S, Devroey P and Van Steirteghem A (1995) Using ejaculated, fresh, and frozen-thawed epididymal and testicular gives rise to comparable results after intracytoplasmic sperm injection. Fertil Steril 63, Nicopoullos JDM, Gilling-Smith C, Almeida PA and Ramsay JWA (2004a) The results of 154 cycles using surgically retrieved sperm from azoospermic men. Hum Reprod 19, Nicopoullos JDM, Gilling-Smith C and Ramsay JWA (2004b) Does the cause of obstructive azoospermia affect the outcome intracytoplasmic sperm injection: a meta-analysis. BJU Int 93, Osmanagaoglu K, Vernaeve V, Kolibianakis, E, Tournaye H, Camus M, Van Steirteghem A and Devroey P (2003) Cumulative delivery rates after ICSI treatment cycles with freshly retrieved sperm: a seven-year follow-up study. Hum Reprod 18, Palermo GD, Schlegel PN, Colombero LT, Zaninovic N, Moy F and Rosenwaks Z (1996) Aggressive sperm immobilization prior to intracytoplasmic sperm injection with immature improves fertilization and pregnancy rates. Hum Reprod 11, Palermo GD, Schlegel PN, Hariprashad JJ, Ergün B, Mielnik A, Zaninovic N, Veeck LL and Rosenwaks Z (1999) Fertilization and pregnancy outcome with intracytoplasmic injection for azoospermic men. Hum Reprod 14, Pasqualotto FF, Rossi-Ferragut LM, Rocha CC, Iaconelli A and Borges E (2002) Outcome of in vitro fertilization and intracytoplasmic injection of epididymal and testicular sperm obtained from patients with obstructive and non obstructive azoospermia. J Urol 167, Pasqualotto FF, Rossi-Ferragut LM, Rocha CC, Guilherme P, Valdemar O, Iaconelli A and Borges E (2005) Etiology-specific outcomes of intracytoplasmic sperm injection in azoospermic patients. Fertil Steril 83, Ramos L, Kleingeld P, Meuleman EJ, van Kooy R, Kremer J, Braat DD and Wetzels AM (2002) Assessment of DNA fragmentation of that were surgically retrieved from men with obstructive azoospermia. Fertil Steril 77, Royere D, Barthelemy C, Hamamah S, Lansac J (1996) Cryopreservation of : a 1996 review. Hum Reprod Update, 2, Shrivastav P, Nadkarni P, Wensvoort S and Craft I (1994) Percutaneous epididymal sperm aspiration for obstructive azoospermia. Hum Reprod 9, Silber SJ, Nagy ZP, Liu J, Godoy H, Devroey P and Van Steirteghem AC (1994) Conventional in-vitro fertilization versus intracytoplasmic sperm injection for patients requiring microsurgical sperm aspiration. Hum Reprod 9, Silber SJ, Nagy Z, Devroey P, Camus M and Van Steirteghem AC (1997) The effect of female age and ovarian reserve on pregnancy rate in male infertility: treatment of azoospermia with sperm retrieval and intracytoplasmic injection. Hum Reprod 12, Steele EK, McClure N, Maxwell RJ and Lewis SEM (1999) A comparison of DNA damage in testicular and proximal epididymal in obstructive azoospermia. Mol Hum Reprod 5, Tournaye H, Devroey P, Liu J, Nagy Z, Lissens W and Van Steirteghem A (1994) Microsurgical epididymal sperm aspiration and intracytoplasmic sperm injection: a new effective approach to infertility as a result of congenital bilateral absence of the vas deferens. Fertil Steril 61, Tournaye H, Merdad T, Silber S, Joris H, Verheyen G, Devroey P and Van Steirteghem A (1999) No differences in outcome after intracytoplasmic sperm injection with fresh or with frozen-thawed epididymal. Hum Reprod 14, Ubaldi F, Nagy ZP, Rienzi L, Tesarik J, Anniballo R, Franco G, Menchini- Fabris F and Greco E (1999) Reproductive capacity of from men with testicular failure. Hum Reprod 14, Vernaeve V, Tournaye H, Osmonagaoglu K, Verheyen G, Van Steirteghem A and Devroey P (2003a) Intracytoplasmic sperm injection with testicular is less successful than in men with obstructive azoospermia. Fertil Steril 79, Vernaeve V, Bonduelle M, Tournaye H, Camus M, Van Steirteghem A and Devroey P (2003b) Pregnancy outcome and neonatal data of children in obstructive and non obstructive azoospermia. Hum Reprod 18, Windt ML, Coetzee K, Kruger TF, Menkveld R and van der Merwe JP (2002) Intracytoplasmic sperm injection with testicular in men with azoospermia. J Assist Reprod Genet 19, Yamamoto M, Hibi H, Asada Y, Suganuma N and Tomoda Y (1996) sperm retrieval by epididymal micropuncture combined with intracytoplasmic sperm injection: difference between acquired and congenital irreparable obstructive azoospermia. Urol Int 58, Ziebe S, Loft A, Peterson JH, Anderson AG, Lindenberg S, Peterson K and Anderson AN (2001) Embryo quality and developmental potential is compromised by age. Acta Obstet Gynecol Scand 80, Submitted on October 13, 2005; accepted on October 26,