Blastocyst formationðgood indicator of clinical results after ICSI with testicular spermatozoa

Human Reproduction Vol.18, No.5 pp. 1070±1076, 2003 DOI: 10.1093/humrep/deg221 Blastocyst formationðgood indicator of clinical results after ICSI with testicular spermatozoa I.Virant-Klun 1, T.TomazÏevicÏ, B.Zorn, L.BacÏer-Kermavner, J.MivsÏek and H.Meden-Vrtovec Reproductive Unit, Department of Obstetrics and Gynaecology, University Medical Centre Ljubljana, SÏlajmerjeva 3, 1000-Ljubljana, Slovenia 1 To whom correspondence should be addressed. E-mail: irma.virant@kclj.si BACKGROUND: The aim of this study was to evaluate the role of blastocyst culture in patients with azoospermia. METHODS: In 98 cycles were cultured for 2 days and in 128 cycles for 5 days to reach the blastocyst stage; a maximum of two of the most developed were transferred in each group. RESULTS: There was a negative correlation between a high (>20 IU/l) male serum FSH and embryo development, manifested as not reaching the morula stage on day 5 (r = 0.387; P < 0.05). After prolonged culture, 23% of reached the blastocyst stage. The pregnancy rates per transfer, and the abortion rates were approximately the same in the day 2 group and the day 5 group (20 versus 20% and 19 versus 18% respectively). After blastocyst transfer, a high clinical pregnancy rate (55%) and a low abortion rate (6%) were achieved, whereas the transfer of arrested provided a low pregnancy rate (2%) and a high abortion rate (100%). If only blastocysts had been transferred on day 5, the clinical pregnancy rate per started cycle would have been approximately the same in both groups (13 versus 16%). CONCLUSIONS: Blastocyst formation is a good indicator of clinical results after ICSI with testicular sperm. Key words: blastocysts/icsi/spermatozoa/testis Introduction A number of studies have reported the successful culture of human to the blastocyst stage in new sequential media without cell co-cultures (Gardner et al., 1998; Behr et al., 1999; MeÂneÂzo et al., 1999; Plachot et al., 2000; Langley et al., 2001; Karaki et al., 2002; Wilson et al., 2002). The transfer of blastocysts offers several advantages, the most important being synchronization of the embryo with the uterine endometrium, and selection of the best quality with a high implantation potential (Gardner et al., 1998; Plachot et al., 2000). Blastocyst transfer is highly recommended for the indications related to maternal, paternal and cytogenetic aspects on the embryonic side to avoid repeated failure of implantation (MeÂneÂzo et al., 1999). It has been proposed to use blastocyst culture to avoid the negative paternal effects on embryo development after ICSI (Sakkas, 1999). Abnormal sperm in terms of concentration, morphology and motility may manifest a negative effect on preimplantation embryo development to the blastocyst stage (Janny and MeÂneÂzo, 1994; Shoukir et al., 1998; Grif ths et al., 2000; Dumoulin et al., 2001; Miller and Smith, 2001). Additionally, the ICSI procedure in itself contributes to a reduced capacity of blastocyst formation in comparison with classical IVF (Grif ths et al., 2000; Dumoulin et al., 2001). The source and maturity of sperm, most likely to be indicative of the severity of spermatogenic disorders, affect the rate of blastocyst formation and blastocyst implantation. It has been reported that sperm from non-obstructive azoospermic subjects, when utilized for ICSI, result in that progress to the blastocyst stage at a lower rate than from obstructive azoospermic subjects and from ejaculated sperm (Balaban et al., 2001). Round spermatid injection is associated with a signi cantly lower embryo development rate and a signi cantly higher developmental arrest rate, when compared with the injection of testicular sperm (Balaban et al., 2000; Vicdan et al., 2001). It has been reported that pronuclear morphology is correlated with chromosomal complement, and impacted upon by the sperm source (Kahraman et al., 2002). Silber et al. (2002) found a higher rate of chaotic mosaicism in derived from TESE±ICSI than in those derived from ICSI with ejaculated sperm. The difference was highly signi cant. It has been found (Obasaju et al., 1999) that paternal factors, which are likely to derive from the centrosome, can contribute to embryonic numerical chromosomal abnormalities, which in turn may predispose to implantation failure. Sperm quality may adversely affect the chromosome constitution of that result from ICSI (Obasaju et al., 1999). These ndings indicate the importance of, and the need for, good embryo selection after TESE±ICSI. Little is known about the factors affecting embryo development to the blastocyst stage after ICSI with testicular sperm. The question of whether the use of blastocyst culture selects the 1070 ã European Society of Human Reproduction and Embryology

Blastocysts and ICSI with testicular sperm best after ICSI with testicular sperm and leads to a better implantation rate still remains to be answered. The aim of this study was to nd the factors affecting the developmental potential of after ICSI with testicular sperm, and to evaluate the clinical role of blastocyst culture in patients with azoospermia in comparison with conventional cleaving stage. Materials and methods Experimental design This prospective, randomized study was performed at the Department of Obstetrics and Gynaecology, University Medical Centre Ljubljana, from January, 2000 to April, 2002. The study was randomized so that each day of TESE±ICSI cycles with cultured for 2 days (Group 1) was followed by the day of TESE±ICSI cycles with cultured for 5 days (Group 2). In this manner 226 cycles of ICSI with frozen±thawed testicular sperm were included in the study. In 98 cycles (Group 1) ICSI-derived were cultured conventionally for 2 days in the Universal IVF Medium (Medi-Cult, Denmark) and a maximum of two cleavagestage were transferred into the uterus on day 2. In 128 cycles (Group 2) were cultured for 5 days in the Medium 1 and Medium 2 of Blast Assist System (Medi-Cult) and a maximum of two most developed Ðblastocysts, delayed Ðmorulae, or arrested Ðwere transferred into the uterus on day 5. Supernumerary were cryopreserved. We compared the clinical results between the groups. All men gave informed consent to participate in the study. The study had been approved by the National Medical Ethics Committee. Studied population In this study 226 ICSI cycles with testicular sperm were performed in 154 infertile men (74 men in Group 1 and 80 in Group 2) suffering from azoospermia: 29 obstructive azoospermia (16 in Group 1 and 13 in Group 2) and 125 non-obstructive azoospermia (58 in Group 1 and 67 in Group 2). Data on the age and testicular volume were collected for all the patients. Histological diagnosis of spermatogenesis was made in 62 patients (28 in Group 1 and 34 in Group 2), evaluation of serum FSH concentration in 77 patients (48 in Group 1 and 29 in Group 2). Serum FSH concentration was de ned as increased at >10 IU/l and high at >20 IU/l. Testicular biopsy and tissue cryopreservation In all couples ICSI was performed using frozen±thawed testicular sperm. In our centre the diagnostic testicular biopsy has been followed by testicular tissue freezing to avoid possible negative effects of subsequent multiple biopsies since 1997. Moreover, with sperm cryopreservation we avoid ovarian stimulation and oocyte retrieval if no testicular sperm are retrieved. Testicular tissue was extracted by open biopsy or aspiration. Each testicular specimen was analysed for the presence of sperm. A part of the sample was immediately xed in the Bouin's solution for histological analysis of spermatogenesis using modi ed Johnsen's score (from score 1: Sertoli cells-only and thin epithelium, to score 10: normal spermatogenesis and thick epithelium) (Holstein et al., 1994). The remaining testicular specimen was frozen to be used in subsequent ICSI treatments. Testicular tissue was dissected, mixed with cryoprotectant solutionð10% glycerol in Flushing Medium (Medi-Cult), put into ampoules, and frozen automatically (MiniCool 40, Air Liquide, France) to ±150 C. Then it was plunged into liquid nitrogen (±196 C) and stored until thawing for ICSI. It was thawed at room temperature. IVF Female partners were stimulated using rst a desensitizing protocol of GnRH agonist (s.c. buserelin acetate 0.6 mg once daily; Suprefact; Hoechst AG, Germany), started on day 22 of the menstrual cycle. If after 14 days serum estradiol concentrations were <40 pg/ml and no ovarian cystic structures were observed on ultrasound (US) examination, stimulation was started with 225 IU/day of hmg (Pergonal; Serono, Switzerland) or FSH (Metrodin, HP 75 IU; Serono). hcg (Pregnyl; Organon, The Netherlands) 5000 IU was administered when a leading follicle measuring >18 mm in diameter was obtained. Transvaginal ultrasound-guided aspiration of ovarian follicles was performed 36 h after hcg administration using a single lumen needle. Oocyte and sperm preparation before ICSI Cumulus±corona cell complexes were isolated from the follicular uid and put into 5 ml Falcon tubes with 1 ml of Universal IVF medium (Medi-Cult; the tubes were gassed in 5% O 2,5%CO 2 and 90% N 2. The removal of the cells of the cumulus and the corona radiata by hyaluronidase has been described extensively elsewhere (Palermo et al., 1992; Van Steirteghem et al., 1993). Oocytes were placed in 80 IU hyaluronidase (Type VIII; Sigma, USA) and the cells of the cumulus and the corona radiata were removed with a denudation pipette (Swemed Lab, Sweden) and washed with Flushing Medium (Medi-Cult). After testicular tissue preparation on PureSperm (Nidacon, Sweden) discontinuous concentration gradient and washing, 5 ml droplets of sperm suspension were prepared under liquid paraf n (Medi-Cult). ICSI procedure Metaphase-II oocytes were placed in 5 ml droplets of Universal IVF medium under liquid paraf n (Medi-Cult). ICSI was performed by the conventional method (Van Steirteghem et al., 1993) without using polyvinylpyrrolidone (PVP). A metaphase-ii oocyte was aspirated with a slight negative pressure by the holding pipette (Swemed Lab) and a single spermatozoon was injected into the ooplasm by the microinjection pipette (Swemed Lab) under an inverted microscope (Diaphot-TMD, Nikon, Japan) equipped with a hydraulic micromanipulator (Narishige, Japan). Before micro-injection, no special selection of sperm was performed. Embryo culture to the cleaving stage and transfer on day 2 In each patient, all were cultured in a group in 0.5 ml of Universal IVF Medium (Medi-Cult) in a double-well dish (Falcon); a maximum of two were transferred into the uterus on day 2 using the TDT embryo transfer set (Prodimed, Neuilly-en-Thelle, France). Supernumerary of good quality (<15 % fragmentation) were cryopreserved. Embryo culture to the blastocyst stage and transfer on day 5 Embryos were cultured to the blastocyst stage in sequential media, Medium 1 and Medium 2 of Blast Assist System (Medi-Cult). All of each patient were cultured in a group. On day 1, fertilized oocytes were transferred into 0.5 ml of fresh Medium 1 in a doublewell dish (Falcon). On day 2 they were transferred into fresh Medium 2. Then were transferred each day into a fresh Medium 2, and on day 5 a maximum of 2 multi-cell were transferred into the uterus. On day 5 we classi ed as most developed Ð blastocysts, delayed Ðmorulae, and arrested Ð lower than morula stage. We transferred two blastocysts into the 1071

I.Virant-Klun et al. uterus; if there were no blastocysts, we transferred the two most developed that did not reach the stage of blastocystð morulae, or arrested. Supernumerary blastocysts were cryopreserved. Luteal support and follow-up of pregnancy On the day of embryo transfer, the luteal phase support was started using oral dydrogesterone, 300 mg/day (Dabroston; BELUPA, Croatia). Biochemical pregnancy was con rmed by serum beta-hcg determination 15 days after embryo transfer. Clinical pregnancy was con rmed by US demonstration of at least one gestational sac at 6 to 7 weeks of pregnancy. Statistical analysis For statistical analysis the SPSS program (SPSS Inc., Chicago, Illinois, USA) was used. Differences between the groups were evaluated by Mann±Whitney U-test. Correlations between the clinical parameters (female age, female indications of infertility, male age, serum FSH, testicular volume, modi ed Johnsen's score), and the developmental potential of after ICSI with testicular sperm were evaluated by Spearman's correlation coef cient. Statistical signi cance was set at P < 0.05. Results General clinical results In the studied population of men, the mean serum FSH concentration was 12.1 6 10.3 IU/l (range: 1.9±38.3), modi ed Johnsen's score 6.8 6 2.9 (range: 1±9), and testicular volume 29.9 6 13.3 ml (range: 4.0±50.0). Serum FSH was elevated (>10 IU/l) in 47 of the 77 (61%) men, and high (>20 IU/l) in 22 (29%) men. The highest serum FSH concentration at which testicular sperm were retrieved was 38.3 IU/l. There was a strong negative correlation between the FSH level and modi ed Johnsen's score (r = ±0.649, P < 0.0001). With FSH serum concentrations 10 IU/l and 20 IU/l, the mean Johnsen's score was 5 and 2, respectively. Clinical results concerning ICSI with testicular sperm are presented in Table I. The group receiving on day 2 did not differ from the group receiving on day 5 in the male and female age, and proportion of female indications of infertility. After 226 ICSI cycles, were transferred in 189 (83%) cycles resulting in 38 clinical pregnancies, i.e. in a 20% pregnancy rate per embryo transfer. Seven pregnancies (18%) ended in a spontaneous abortion. Clinical results of the group receiving on day 2, and of the group receiving after prolonged culture on day 5, arrested included, were approximately the same. In both groups the twin pregnancy rates were approximately the same; there was no triplet pregnancy in either group. Clinical results after day 2 transfer versus day 5 transfer Blastocyst transfer provided a high clinical pregnancy rate per embryo transfer (55 %). After the transfer of blastocysts on day 5, pregnancy and implantation rates were signi cantly higher than after the transfer of on day 2; the abortion rate was signi cantly lower (Table I). 1072 Table I. Clinical results after ICSI with testicular sperm according to the day of embryo transfer (day 2 versus day 5) Embryo transfer on day 2 (cleavage stage ) Embryo transfer on day 5 (blastocysts) Started cycles 98 128 Couples 74 80 Male age (years) (range) 34.05 6 6.85 (21±54) 34.74 6 7.09 (21±61) Female age (years) (range) 32.93 6 5.01 (20±43) 32.78 6 5.15 (20±43) Female indications of 22 (30) 29 (36) infertility (%) Oocytes 749 850 Oocytes per woman 7.6 6 5.3 6.6 6 4.9 Fertilized oocytes (%) 240 (32) 305 (36) Embryos 215 86 Transferred (%) 147 (68) 62 (72) Frozen (%) 56 (26) 24 (28) Excluded (%) 12 (6) 0 (0) Embryo transfers (%) 81 (83 ) a 31 (24 ) b Clinical pregnancies 16 17 Clinical pregnancy rate per 20 % c 55 % d embryo transfer Clinical pregnancy rate per 16 % 13 % started cycle Twin pregnancies (%) 1 (6) 2 (13) Triplet pregnancies (%) 0 (0) 0 (0) Implantation rate 12% e 31.7% f Spontaneous abortions (%) 3 (19) g 1 (6) h Extrauterine pregnancies (%) 0 (0) 0 (0) a,b/c,d/e,f/g,h Statistically signi cant differences as detected by Mann±Whitney U-test (P < 0.05). Clinical results after embryo transfer on day 5 After prolonged culture, at least one blastocyst was transferred in 31 of the 108 (24%) ICSI cycles. In the remaining transfer cycles, that did not reach the blastocyst stage were transferred: morulae in 30 (28%) cycles and arrested at the lower than the morula stage in 47 (44%) cycles (Table II). After the transfer of blastocysts clinical pregnancy and implantation rates were signi cantly higher than after the transfer of morulae and arrested (Table II). Blastocyst transfers resulted in a signi cantly higher twin-pregnancy rate, and in a signi cantly lower abortion rate than transfers with morulae and arrested (6 versus 50% and 6 versus 100%; P < 0.05). With arrested that did not reach the morula stage, the pregnancy rate was very low (2%). After 47 transfers only one pregnancy was achieved, which ended in a spontaneous abortion (Table II). If only blastocysts had been transferred on day 5, the clinical pregnancy rate per started cycle would have been approximately the same as in the day 2 transfer group: 13% (17 pregnancies in 128 cycles) versus 16% (16 pregnancies in 98 cycles) (Table I). Developmental stage of on day 5 After prolonged culture, 68 of the 290 (23%) reached the blastocyst stage and 54 (19%) reached the morula stage, whereas 35 (12%) arrested before the morula

Blastocysts and ICSI with testicular sperm Table II. Effect of the developmental potential of transferred on day 5 on clinical results Development potential of on day 5 Blastocysts Morulae Arrested All Number of cycles with embryo transfer 31 30 47 108 Patients 28 24 28 80 Female age (years) 32.6 6 4.9 34.4 6 9.5 33.5 6 4.7 32.8 6 5.1 Male serum FSH concentration (IU/l; range) 8.6 6 5.4 (2.0±15.9) 10.0 6 9.7 (2.0±33.2) 15.0 6 11.8 (2.5±38.3) 11.4 6 9.8 (2.0±38.3) Transferred 62 54 73 189 Transferred per woman 1.9 6 0.2 1.8 6 0.5 1.5 6 0.5 1.5 6 0.7 Clinical pregnancies 17 4 1 22 Clinical pregnancies per embryo 55% a 13% b 2% c 20% transfer Twin pregnancies (%) 2 (13) d 0 (0) e 0 (0) f 2 (9) Triplet pregnancies (%) 0 (0) 0 (0) 0 (0) 0 (0) Implantation rate 31.7% g 7.4% h 1.4% i 13% Spontaneous abortions (%) 1 (6) j 2 (50) k 1 (100) l 4 (18) Extrauterine pregnancies (%) 0 (0) 0 (0) 0 (0) 0 (0) a,b/a,c/d,e/d,f/g,h/g,i/j,k/j,l Statistically signi cant differences as detected by Mann±Whitney U-test (P < 0.05). Table III. Developmental stage of according to the male serum FSH concentration Serum FSH concentration All stage having more than 6 cells, and 134 (46%) arrested at the 2±6-cell stage. Clinical factors affecting developmental potential of There was no correlation between the female age (r = ±0.083, P = 0.390), female indications of infertility (r = 0.039, P = 0.687), and the number of blastocysts in the couple. In the male partner, there was no correlation between the number of blastocysts and age (r = 0.167, P = 0.197), serum FSH (r = 0.249, P = 0.143), modi ed Johnsen's score (r = 0.098, P = 0.504) and testicular volume (r = 0.039, P = 0.687). The only correlation we observed was a positive correlation between the serum FSH concentration in the male partner and the number of arrested that did not reach the morula stage on day 5 (r = 0.387; P < 0.05). In men with high serum FSH concentrations (>20 IU/l) there was a signi cantly lower proportion of morulae and a higher proportion of arrested before the morula stage than in patients with normal serum FSH concentrations (Table III), but there was no statistically signi cant difference in the proportion of blastocysts between these two groups. In men with high serum FSH concentrations (>20 IU/l) a signi cantly lower proportion had at least one morula and a signi cantly higher proportion of men had only arrested Blastocysts Morulae Arrested Johnsen's score >20 IU/l (7 men/11 cycles) 27 5 (18%) 0 (0%) a 22 (81%) c 5.3 6 3.0 >10 IU/l (15 men/22 cycles) 58 13 (22%) 8 (14%) 37 (64%) 4.7 6 2.9 <10 IU/l (18 men/27 cycles) 45 11 (24%) 14 (31%) b 20 (44%) d 8.2 6 1.3 a,b/c,d Statistically signi cant differences as detected by Mann±Whitney U-test (P < 0.05). in their ICSI cycles, compared with the men with normal serum FSH concentrations (Table IV). There were no statistically signi cant differences in the developmental potential of between the men with obstructive and those with non-obstructive azoospermia (Table V). Discussion In this study we provide evidence that prolonged culture of after ICSI with testicular sperm does not decrease clinical results in infertile men with azoospermia. Blastocysts have a very good prognosis for pregnancy, whereas after the transfer of that do not reach the blastocyst stage, clinical results are inferior: low pregnancy and implantation rates, most pregnancies ending in a spontaneous abortion. In this study, 23% of developed to the blastocyst stage after ICSI with testicular sperm. From our own experience, a signi cantly lower proportion of develop to the blastocyst stage after ICSI with testicular sperm than after ICSI with ejaculate sperm (45%) or classical IVF (60%) thus con rming the negative paternal effects on embryo development. Some recent studies show that blastocysts may have chromosomal abnormalities. It has been shown that some chromosomally abnormal human develop to the 1073

I.Virant-Klun et al. Table IV. Men and developmental stage of after ICSI with testicular sperm according to the serum FSH concentration Serum FSH concentration All men Men with at least one blastocyst in all cycles (%) Men with at least one morula in all cycles (%) >20 IU/l 7 (11 cycles) 2 (29) 0 (0) a 5 (71) d >10 IU/l 15 (22 cycles) 4 (27) 4 (27) b 7 (46) <10 IU/l 18 (27 cycles) 7 (39) 6 (33) c 5 (28) e a,b/a,c/d,e Statistically signi cant differences as detected by Mann±Whitney U-test (P < 0.05). Men with arrested only in all cycles (%) Table V. Type of azoospermia and developmental potential of Azoospermia All Blastocysts (%) Morulae (%) Arrested (%) Serum FSH (IU/l; range) Johnsen's score Obstructive (13 men/25 cycles) Non-obstructive (67 men/82 cycles) blastocyst stage (Sandalinas et al., 2001). However, in our study a high pregnancy rate per embryo transfer (55%) and a very low spontaneous abortion rate (6%) by blastocysts were obtained. Clinical pregnancies (13%) were also achieved by the transfer of morulae, although half of these pregnancies ended in a spontaneous abortion. Morulae are delayed and a good percentage of these should reach the blastocyst stage on day 6. Our ndings t with the observation of a lower implantation rate of the reaching the blastocyst stage on day 6. After 47 transfers of arrested at a lower cellstage than the morula only one pregnancy was achieved, which ended in a spontaneous abortion. By using blastocyst culture the genetic risk is not avoided, but is much decreased. A signi cantly higher proportion of genetically abnormal is found among arrested ; aneuploidy mosaicism was found in almost 37% of arrested (MeÂneÂzo and Ben Khalifa, 1995). In yet another study the percentage of abnormal cells was 54% in arrested, and 17.1% in blastocysts (Veiga et al., 1999). Our study con rms the nding that after prolonged culture, arrested lead to a spontaneous abortion, therefore the transfer of arrested is not recommended. Moreover, if on day 5 only blastocysts had been transferred, the clinical pregnancy rate per started cycle would have been approximately the same as with the transfer on day 2. In this case, however, many women would have been saved from having a transfer and pregnancy that would end in an abortion. Also, prolonged culture of improves the selection of for cryopreservation thus avoiding the freezing of arrested. We did not observe any correlation between the female age, female indications of infertility, and the number of blastocysts in the couple. Most cycles were performed for the male factor of infertilityðazoospermia. 1074 47 7 (15%) 16 (34%) 24 (51%) 4.4 6 1.9 (2.5±10.3) 197 52 (26%) 30 (15%) 116 (59%) 15.7 6 10.7 (1.9±38.3) 8.8 6 0.4 5.9 6 3.1 The only correlation we found was a positive correlation between the serum FSH concentration in the male partner and the number of arrested that did not reach the morula stage on day 5. In men with high serum FSH concentrations (>20 IU/l) there was a signi cantly lower proportion of morulae and a higher proportion of arrested prior to reaching the morula stage. We also found a higher proportion of men with only arrested after ICSI with testicular sperm than in patients with normal serum FSH concentrations. Negative paternal effects were expressed at lower cell stages, whereas the proportion of blastocysts was approximately the same as in patients with normal serum FSH. We consider it very likely that an increased number of arrested in men with high FSH is related to the switch from maternal to embryonic genome at the 4±8-cell stage (Bolton et al., 1989). Serum FSH concentration is not a highly predictive factor for successful testicular sperm recovery in azoospermic patients (Tournaye et al., 1997). No strong predictors for successful testicular sperm recovery are available except histopathology (Tournaye et al., 1997). FSH concentrations correlate with spermatogenetic activity. However, the prediction of successful sperm retrieval is not absolutely reliable. Testicular sperm extraction can also be successful when FSH hormone concentrations are outside the threshold levels (Bohring et al., 2002). It is possible to obtain sperm from patients with markedly elevated serum FSH (38.7 IU/l) as reported by Gil-Salom (Gil-Salom et al., 1995). In our study the highest serum FSH concentration at which testicular sperm were retrieved was 38.3 IU/l. Besides, 61% of men included in our study had increased serum FSH levels (>10 IU/l), and 28% of men had high serum FSH concentration (>20 IU/l). Therefore, the negative effect of high FSH should not be overlooked.

FSH concentrations might re ect the testicular sperm quality. Egozcue showed that meiotic disorders in infertile men are frequent, and increase with high FSH concentrations (Egozcue et al., 2000). These patients produce more sperm with autosomal and sex chromosomal disomies and diploid sperm, originating either from meiotic mutations or from a compromised testicular environment. It has been stated that sperm concentration <1 3 10 6 /ml and/or serum FSH concentration >10 IU/l are the only predictors of male germ cell meiotic abnormalities (Vendrell et al., 1999). There is a link between gonadal failure (high serum FSH levels) and the occurrence of sperm chromosome aneuploidies (Levron et al., 2001). Testicular sperm have higher rates of sperm aneuploidy and diploidy than ejaculated sperm (Bernardini et al., 2000; Mateizel et al., 2002). Paternal factors negatively affecting the embryo development are related to anomalies of male germ cell DNA condensation (Francavilla et al., 2001) and fragmentation (Tesarik et al., 2001; Evenson et al., 2002), as well as to centrosome anomalies (Hewitson et al., 2002). The factors leading to azoospermia may affect testicular sperm morphology (Yavetz et al., 2001). In our study there was a strong negative correlation between FSH concentrations and Johnsen's score. The serum FSH concentrations of 10 IU/l provided a mean Johnsen's score of 5, and the FSH concentration of 20 IU/l, a Johnsen's score of 2. These elevated FSH concentrations mainly correspond to immature germ cells or absence of germ cells. In most patients with high serum FSH concentration (>20 IU/l) we extracted sperm from the sites of focal spermatogenesis; immature germ cells prevailed. FSH level >20 IU/l might represent the threshold level of testicular sperm maturity and explain the arrest of embryo development. We found no statistical difference in the developmental potential of in patients with obstructive versus nonobstructive azoospermia. Some men with non-obstructive azoospermia had a normal or only slightly elevated serum FSH concentration (i.e. azoospermia with maturation arrest), similar to that in men with obstructive azoospermia. Male serum FSH seems to be one of the most important factors for testicular sperm quality and embryo development. Blastocyst formation is a good indicator of clinical results after ICSI with testicular sperm in terms of high implantation rate and low abortion rate. High male serum FSH concentrations negatively affect the developmental potential of, these negative paternal effects being manifested before the morula stage. Acknowledgements We would like to thank the andrologist SasÏo DrobnicÏ and gynaecologists Martina RibicÏ-Pucelj, Andrej Vogler, Eda Bokal- VrtacÏnik, and Bojana Pinter for collaborating in the IVF programme, Ms Mojca Pirc for revising the manuscript, and Ms Bibi Fissbechus for all her support. References Balaban, B., Urman, B., Isiklar, A., Alatas, C., Aksoy, S., Mercan, R. and Nuhoglu, A. (2000) Progression to the blastocyst stage of derived from testicular round spermatids. Hum. Reprod., 15, 1377±1382. Blastocysts and ICSI with testicular sperm Balaban, B., Urman, B., Isiklar, A., Alatas, C., Mercan, R., Aksoy, S. and Nuhoglu, A. (2001) Blastocyst transfer following intracytoplasmic injection of ejaculated, epididymal or testicular spermatozoa. Hum. Reprod., 16, 125±129. Behr, B., Pool, T.B., Milki, A.A., Moore, D., Gebhardt, J. and Dasig, D. (1999) Preliminary clinical experience with human blastocyst development in vitro without co-culture. Hum. Reprod., 14, 454±457. Bernardini, L., Gianaroli, L., Fortini, D., Conte, N., Magli, C., Cavani, S., Gaggero, G., Tindiglia, C., Ragni, N. and Venturini, P.L. (2000) Frequency of hyper-, hypohaploidy and diploidy in ejaculate, epididymal and testicular germ cells of infertile patients. Hum. Reprod., 15, 2165±2172. Bohring, C., Schroeder-Printzen, I., Weidner, W. and Krause, W. (2002) Serum levels of inhibin B and follicle-stimulating hormone predict successful sperm retrieval in men with azoospermia who are undergoing testicular sperm extraction. Fertil. Steril., 78, 1195±1198. Bolton, V.N., Hawes, S.M., Taylor, C.T. and Parsons, J.H. (1989) Development of spare human preimplantation in vitro: An analysis of the correlations among gross morphology, cleavage rates, and development to the blastocyst. J. In vitro Fertil. Embryo Transfer, 6, 30±35. Dumoulin, J.M., Coonen, E., Bras, M., Bergers-Janssen, J.M., Ignoul- Vanvuchelen, R.C., van Wissen, L.C., Geraedts, J.P. and Evers, J.L. (2001) Embryo development and chromosomal anomalies after ICSI: effect of the injection procedure. Hum. Reprod., 16, 306±312. Egozcue, S., Blanco, J., Vendrell, J.M., Garcia, F., Veiga, A., Aran, B., Barri, P.N., Vidal, F. and Egozcue, J. (2000) Human male infertility: chromosome anomalies, meiotic disorders, abnormal spermatozoa and recurrent abortion. Hum. Reprod. Update, 6, 93±105. Evenson, D.P., Larson, K.L. and Jost L.K. (2002) Sperm chromatin structure assay: its clinical use for detection of sperm DNA fragmentation in male infertility and comparison with other techniques. J. Androl., 23, 25±43. Francavilla, S., Bianco, M.A., Cordeschi, G., D'Abrizio, P., De Stefano, C., Properzi, G. and Francavilla, F. (2001) Ultrastructural analysis of chromatin defects in testicular spermatids in azoospermic men submitted to TESE- ICSI. Hum. Reprod., 16, 1440±1448. Gardner, D.K., Schoolcraft, W.B., Wagley, L., Schlenker, T., Stevens, J. and Hesla, J. (1998) A prospective randomized trial of blastocyst culture and transfer in in-vitro fertilization. Hum. Reprod., 13, 3434±3440. Gil-Salom, M., Remohi, J., Minguez, Y., Rubio, C. and Pellicer, A. (1995) Pregnancy in an azoospermic patient with markedly elevated serum folliclestimulating hormone levels. Fertil. Steril., 64, 1218±1220. Grif ths, T.A., Murdoch, A.P. and Herbert, M. (2000) Embryonic development in vitro is compromised by the ICSI procedure. Hum. Reprod., 15, 1592±1596. Hewitson, L., Simerly, C.R. and Schatten G. (2002) Fate of sperm components during assisted reproduction: implications for infertility. Hum. Fertil. (Camb), 5, 110±116. Holstein, A.F., Schulze, W. and Breucker, H. (1994) Histopathology of human testicular and epididymal tissue. In Hergreave, T.B. (ed), Male Infertility. Springer Verlag, London, pp 105±148. Janny, L. and MeÂneÂzo, Y. (1994) Evidence for a strong paternal effect on human preimplantation embryo development and blastocyst formation. Mol. Reprod. Dev., 38, 36±42. Kahraman, S., Kumtepe, Y., Sertyel, S., Donmez, E., Benkhalifa, M., Findikli, N. and Vanderzwalmen, P. (2002) Pronuclear morphology scoring and chromosomal status of in severe male fertility. Hum. Reprod., 17, 3193±3200. Karaki, R.Z., Samarraie, S.S., Younis, N.A., Lahloub, T.M. and Ibrahim, M.H. (2002) Blastocyst culture and transfer: a step toward improved in vitro fertilization outcome. Fertil. Steril., 77, 114±118. Langley, M.T., Marek, D.M., Gardner, D.K., Doody, K.M. and Doody, K.J. (2001) Extended embryo culture in human assisted reproduction treatments. Hum. Reprod., 16, 902±908. Levron, J., Aviram-Goldring, A., Madgar, I., Raviv, G., Barkai, G. and Dor, J. (2001) Sperm chromosome abnormalities in men with severe male factor infertility who are undergoing in vitro fertilization with intracytoplasmic sperm injection. Fertil. Steril., 76, 479±484. Mateizel, I., Verheyen, G., Van Assche, E., Tournaye, H., Liebaers, I. and Van Steirteghem, A. (2002) FISH analysis of chromosome X, Y and 18 abnormalities in testicular sperm from azoospermic patients. Hum. Reprod., 17, 2249±2257. MeÂneÂzo, Y.J. and Ben Khalifa, M. (1995) Cytogenetic and cryobiology of human cocultured : 3-year experience. J. Assist. Reprod. Genet., 12, 35±40. MeÂneÂzo, Y., Chouteau, J., Guyader-Joly, C. and Veiga, A. (1999) Sequential media: why and how? Contracept. Fertil. Sex., 27, 449±451. 1075

I.Virant-Klun et al. Miller, J.E. and Smith, T.T. (2001) The effect of intracytoplasmic sperm injection and semen parameters on blastocyst development in vitro. Hum. Reprod., 16, 918±924. Obasaju, M., Kadam, A., Sultan, K., Fateh, M. and MunneÂ, S. (1999) Sperm quality may adversely affect the chromosome constitution of that result from intracytoplasmic sperm injection. Fertil. Steril., 72, 1113±1115. Palermo, G., Joris, H., Devroey, P. and Van Steirteghem, A.C. (1992) Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet, 340, 17±18. Plachot, M., Belaisch-Allart, J., Mayenga, J.M., Chouraqui, A., Serkine, A.M. and Tesquier, L. (2000) Blastocyst stage transfer: the real bene ts compared with early embryo transfer. Hum. Reprod., 15 (Suppl 6), 24±30. Sakkas, D. (1999) The use of blastocyst culture to avoid inheritance of an abnormal paternal genome after ICSI. Hum. Reprod., 14, 4±5. Sandalinas, M., Sadowy, S., Alikani, M., Calderon, G., Cohen, J. and MunneÂ, S. (2001) Developmental ability of chromosomally abnormal human to develop to the blastocyst stage. Hum. Reprod., 16, 1954±1958. Shoukir, Y., Chardonnens, D., Campana, A, and Sakkas, D. (1998) Blastocyst development from supernumerary after intracytoplasmic sperm injection: a paternal in uence? Hum. Reprod., 13, 1632±1637. Silber, S.J., Sadowy, S., Lenahan, K., Kilani, Z., Gianaroli, L. and MunneÂ, S. (2002) Chromosomal abnormalities in derived from testicular sperm extraction. Hum. Reprod., 17 (Abstract Bk 1), O-104. Tesarik, J., Greco, E. and Mendoza, C. (2001) Assisted reproduction with invitro-cultured testicular spermatozoa in cases of severe germ cell apoptosis: a pilot study. Hum. Reprod., 16, 2640±2645. Tournaye, H., Verheyen, G., Nagy, P., Ubaldi, F., Goossens, A., Silber, S., Steirteghem, A.C. and Devroey, P. (1997) Are there any predictive factors for successful testicular sperm recovery in azoospermic patients? Hum. Reprod., 12, 80±86. Van Steirteghem, A.C., Nagy, Z., Joris, H., Liu, J., Staessen, C., Smitz. J., Wisanto, A. and Devroey, P. (1993) High fertilization and implantation rates after intracytoplasmic sperm injection. Hum. Reprod., 8, 1061±1066. Veiga, A., Gil, Y., Boada, M., Carrera, M., Vidal, F., Boiso, I., MeÂneÂzo, Y. and Barri P.N. (1999) Con rmation of diagnosis in preimplantation genetic diagnosis (PGD) through blastocyst culture: preliminary experience. Prenat. Diagn., 19, 1242±1247. Vendrell, J.M., Garcia, F., Veiga, A., Calderon, G., Egozcue, S., Egozcue, J. and Barri, P.N. (1999) Meiotic abnormalities and spermatogenic parameters in severe oligoasthenozoospermia. Hum. Reprod., 14, 375±378. Vicdan, K., Isik, A.Z. and Delilbasi, L. (2001) Development of blastocyststage after round spermatid injection in patients with complete spermiogenesis failure. J. Assist. Reprod. Genet., 12, 319±328. Wilson, M., Hartke, K., Kiehl, M., Rodgers, J., Barabec, C. and Lyles, R. (2002) Integration of blastocyst transfer for all patients. Fertil. Steril., 77, 693±696. Yavetz, H., Yogev, L., Kleiman, S., Botchan, A., Hauser, R., Lessing, J.B. and Gamzu, R. (2001) Morphology of testicular spermatozoa obtained by testicular sperm extraction in obstructive and nonobstructive azoospermic men and its relation to fertilization success in the in vitro fertilization-intracytoplasmic sperm injection system. J. Androl., 22, 376±381. Submitted on November 5, 2002; accepted on January 1, 2003 1076