Article Prognosis of oocytes showing aggregation of smooth endoplasmic reticulum

RBMOnline - Vol 16 No 1. 2008 113-118 Reproductive BioMedicine Online; www.rbmonline.com/article/3026 on web 15 November 2007 Article Prognosis of oocytes showing aggregation of smooth endoplasmic reticulum Thomas Ebner PhD graduated with honours from the Paris Lodron University of Salzburg, Austria, in 1992. He then worked on his thesis in cancer research at the Institute for Pathology at the General Hospital, Salzburg until 1994. After a 2-year period as a researcher in the Division of Animal Physiology, Salzburg University he began his IVF career in Linz. His postdoctoral thesis on non-invasive selection at different stages of preimplantation development qualified him as a university lecturer in Salzburg. He has published more than 60 papers and his research interests are non-invasive selection processes in IVF, apoptosis, and culture media. Dr Thomas Ebner Thomas Ebner 1, Marianne Moser, Omar Shebl, Michael Sommerguber, Gernot Tews Landes- Frauen- und Kinderklinik, IVF-Unit, Krankenhausstr, 26 30, A-4020 Linz, Upper Austria, Austria 1 Correspondence: Tel: +43 732 6923 0; Fax: +43 732 6923 24604; e-mail: thomas.ebner@gespag.at Abstract Few cytoplasmic dysmorphisms of oocytes have been reported to negatively influence the further fate of the ova. One such anomaly, namely the central aggregation of the smooth endoplasmic reticulum (SER), has recently been associated with suboptimal outcome in a limited number of patients. In order to increase prognostic value, it was decided to prospectively screen all intracytoplasmic sperm injection patients within 1 year for eggs showing aggregations of SER. In addition, all deliveries (obstetric and neonatal data) were analysed. Occurrence of SER cluster was related to duration (P < 0.001) and dosage (P < 0.01) of the stimulation. Fertilization (58.9%) and blastulation rate (44.0%) were lower (P < 0.01) in affected ova compared with unaffected counterparts (77.4 and 87.8%, respectively). Pregnancies in women with affected gametes were accompanied by a higher incidence of obstetric problems (P < 0.01) leading to a non-significant trend towards earlier delivery and significantly reduced birthweight (P < 0.05). It is strongly recommended to avoid transfer of embryos/blastocysts derived from SER cluster-positive gametes. Patients have to be informed that even transfer of sibling oocytes without this anomaly involves a higher risk of detrimental outcome. Keywords: aggregation of smooth endoplasmic reticulum, blastocyst formation, neonatal outcome, obstetric outcome, oocyte quality Introduction There is a general tendency to disregard oocyte morphology in the decision-making process for the selection of an embryo or blastocyst for transfer (Ebner et al., 2006a). This is probably due to conflicting data having been published in literature. Although fertilization rate and embryo quality were assumed to be more or less comparable in different oocyte scores (Alikani et al., 1995; De Sutter et al., 1996), it was later suggested that, though not being distinguishable with a light microscope, oocyte quality will have an impact on rates of implantation and clinical pregnancy (Serhal et al., 1997; Loutradis et al., 1999). However, this topic is still under discussion (Balaban et al., 1998). In most studies, several oocyte dysmorphisms were pooled or not analysed at all; hence, the actual negative impact of certain cytoplasmic features on implantation behaviour remains unclear (Balaban and Urman, 2005; Borini et al., 2005). Considering the complex maturation process of the gamete, pooling oocyte features of different origin could indeed be imprecise. To facilitate analysis of morphological oocyte criteria on further outcomes, embryologists tend to classify dysmorphisms as extracytoplasmic and cytoplasmic anomalies (Mikkelsen and Lindenberg, 2001; Ebner et al., 2006a). Focusing on cytoplasmic dysmorphisms of oocytes collected after ovarian stimulation, two types of anomalies have been reported to negatively influence the further fate of the eggs. Firstly, Ebner and co-workers (2005) found a significantly decreased fertilization rate in the presence of vacuoles larger than 14 µm in diameter, data that is consistent with previous reports (De Sutter et al., 1996). In addition, a detrimental effect of vacuoles on blastocyst formation of affected metaphase II (MII) oocytes was reported (Ebner et al., 2005). Secondly, Otsuki and associates (2004) showed that presence of smooth endoplasmic reticulum (SER) aggregations within the ooplasm are associated with a lower chance of successful 113 2008 Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB3 8DB, UK

114 pregnancy, irrespective of the fact that embryos transferred might have been derived from SER cluster-negative gametes of the same pool. Documentation of a case of an imprinting defect possibly related to the investigated dysmorphism (Otsuki et al., 2004) has been even more worrying. Based on these alarming data, it was decided to prospectively screen all intracytoplasmic sperm injection (ICSI) patients who attended the study clinic within 1 year in order to identify and trace eggs showing one or more aggregations of SER. By doing so, the sample number of dysmorphism-positive ova, and in parallel ultimately also the predictive value of the study results, would be significantly increased compared to that reported in literature (Otsuki et al., 2004). In addition, it was planned to follow all deliveries (obstetric and neonatal data) in patients with or without at least one SER cluster-positive gamete in order to either support the previous data (Otsuki et al., 2004) or to give the all clear. Materials and methods Between March 2005 and March 2006 a total of 651 cycles were performed at the study IVF unit. Out of these, 487 were treated by ICSI; thus, some 75% of all cycles were involved in this prospective study on the actual influence of SER clusters on further outcome. Since this was an observational study based on routine in-vitro culture, permission of the Institutional Review Board was not sought. The vast majority of ICSI couples (n = 287) suffered from male factor infertility (58.9%), approximately 10% (n = 51) from female infertility, 3.3% (n = 16) from unexplained infertility and the remaining 27.3% (n = 133) showed male as well as female indications. Approximately three-quarters of women (n = 373) were stimulated using an antagonist protocol. Therefore, human menopausal gonadotrophin (Menogon ; Ferring, Kiel, Germany) was started on day 2 of the cycle. In addition, a gonadotrophin-releasing hormone antagonist (Orgalutran ; Organon, Vienna, Austria) was administered after 5 6 days of stimulation, depending on the presence of a 12 13 mm follicle in the ultrasound scan. A total of 114 patients had their ovaries stimulated according to a long protocol. In these cases, down-regulation of the pituitary was achieved with the gonadotrophin-releasing hormone agonist buserelin (Suprecur ; Aventis Pharma, Vienna, Austria) and stimulation was initiated with human menopausal gonadotrophin (Menogon). In all patients ovulation was induced with 5000 10,000 IU human chorionic gonadotrophin (HCG, Pregnyl ; Organon, Vienna, Austria) provided that the lead follicle reached an appropriate diameter and serum oestradiol appeared adequate, according to the stimulation protocol used. Oocyte collection was carried out transvaginally under ultrasound guidance 36 h after HCG administration. Immediately after oocyte retrieval, collected cumulus oocyte complexes were kept in culture for 2 3 h using Universal IVF Medium (MediCult; Copenhagen, Denmark). After this resting period, oocytes were denuded in 80 IU/ml hyaluronidase (MediCult) using hand-drawn glass pipettes for mechanical treatment. Once most of the cumulus cells were removed, oocyte morphology could be checked adequately. Since some of the cytoplasmic inclusions (incorporations, refractile bodies) were found not to influence fertilization and further cleavage, those gametes were considered for ICSI, while oocytes showing larger vacuoles were not taken into consideration for ICSI. If an MII oocyte showed an aggregation of SER (Figure 1a) after multilevel morphological assessment, the dimension of the dysmorphism was measured accurately using an imaging and archival software (Octax Eyeware ; MTG, Altdorf, Germany). In detail, if possible the SER aggregation was brought into focus prior to taking the picture. The two-dimensional extension of the dysmorphism was then calculated from the stored images after the actual scale had been adjusted (according to the objective used). The mean ± SEM of the measured aggregations was 30.9 ± 13.0 µm and the range varied from 11.8 to 79.0 µm. In the presence of more than one aggregation (n = 8) in the same oocyte (Figure 1b), diameters were pooled for statistical analysis (25.9 73.8 µm). ICSI was performed according to the routine technique (Ebner et al., 2001) using the 3 o clock position for injection. Briefly, micromanipulation was performed on an inverted microscope ( 200 magnification, Olympus IX50, Vienna, Austria) with Hoffman Modulation Contrast (Modulation Optics Inc., Greenvale, NY, USA), an electronically controlled heat stage and hydraulic micromanipulators (Luigs and Neumann, Ratingen, Germany). Special care was taken not to penetrate SER clusters. In the morning of day 1 (17 20 h after injection), regular fertilization (two pronuclei and polar bodies) was checked. On the following days (days 2 and 3), embryos were scored in terms of number and size of blastomeres, as well as for the presence of fragments, signs of compaction (day 4) and blastocyst development (day 5) (Gardner et al., 2000). The media of choice were Blastassist System Medium 1 (MediCult, Copenhagen, Denmark) until day 3 and Blastassist System Medium 2 (MediCult) in case of extended culture to day 5. All oocytes or embryos were cultured in groups (at least 10 µl per conceptus) under oil. Affected oocytes (SER aggregation) were cultured individually under comparable conditions in order to facilitate exact tracing since no difference in outcome was reported as compared with singleton culture (Rijnders and Jansen, 1999). As recommended by Racowsky and associates (2000), the number of 8-cell embryos on day 3 was the key determinant for selecting day 3 or day 5 transfer. Based on selection criteria at different developmental stages, a maximum of two embryos/blastocysts were transferred, but where possible transfer of concepti deriving from dysmorphic oocytes (SER cluster) was avoided. However, four affected embryos were transferred after consultation with the patients (additional transfer of one unaffected embryo as well). All patients were placed in the lithotomy position during embryo replacement and neither sedation nor anaesthesia was used. Concepti were loaded into an Edwards-Wallace Catheter (Smiths Industries, Lancing, UK) using 10 µl of BlastAssist medium 2 and then expelled close to the fundus. Injections of 3000 IU HCG (Pregnyl) on days 2, 5 and 8 were to support luteal phase. Blood concentration of HCG was measured 17 days after intrauterine transfer. Clinical pregnancy was determined by visualization of at least one gestational

sac with positive heart activity 4 weeks after embryo transfer. Subclinical pregnancy showed no fetal heartbeat. Implantation rate was defined by ultrasound visualization of a gestational sac per transferred embryo (subclinical pregnancies were included). Medical files of the patients were requested from maternity clinics 1 month after the calculated date of birth in order to obtain insight into obstetric (placenta previa, premature rupture of the membranes, haemorrhage, gestosis) and neonatal data (weeks of gestation, birthweight, sex, minor and major malformations, transfer to neonatal intensive care unit). Twin pregnancies were not analysed in terms of obstetric and neonatal outcome in order to minimize any negative influence of multiple pregnancy. Comparison was performed using chi-squared and t-tests. Statistical significance was defined as P < 0.05. Results During the 12-month study period, 30 out of 487 ICSI cycles (6.2%) showed at least one oocyte with an aggregation of the SER. Overall, 3114 MII oocytes were checked for this specific anomaly, and as few as 56 (1.8%) were actually affected. However, within the group of SER cluster-positive patients, this percentage increased to 25% (56/224). Patients Table 1 indicates that SER cluster-positive and -negative patients did not differ in demographic data. However, the chance to detect at least one affected oocyte was significantly related to the duration of the stimulation (P < 0.001) as well as the required dose of gonadotrophins (P < 0.01). Hormonal concentrations at the time of ovulation induction showed no differences. Fertilization and further cleavage SER cluster-negative patients showed a 71.9% fertilization rate (2214/3078) which was comparable to SER cluster-positive women (163/224; 72.8%). However, within the latter group, fertilization rate was significantly reduced (P = 0.007) in affected gametes (58.9%; 33/56) as compared with unaffected sibling oocytes (77.4%; 130/168). Focusing on affected patients, there was a correlation between the diameter of the SER aggregation and the event of fertilization (P = 0.017). In detail, oocytes degenerating after ICSI (n = 5) showed the largest diameter (51.2 ± 18.5 µm) as compared with fertilizable gametes (22.6 ± 10.1 µm) (P = 0.0001). Eggs resulting in mono- or tri-pronuclear zygotes (34.4 ± 7.8 µm) and unfertilizable ova (29.9 ± 7.8 µm) were not distinguishable from fertilizable ones on the basis of dysmorphism diameter. On day 1, percentages of halo formation (60.6% versus 70.8%) and optimal pronuclear pattern 0 (36.4 versus 32.3) did not differ between zygotes from positive and negative gametes. Embryo morphology on days 2 (42.4% embryos with <15% fragments versus 42.3%) and 3 (42.2% versus 48.5%) were comparable in both groups. Out of those embryos considered for blastocyst culture (n = 66), significantly fewer (P = 0.0001) concepti reached blastocyst stage if they derived from SER aggregation-positive oocytes (11/25; 44.0%) as compared with unaffected ones (36/41; 87.8%). However, the quality of the blastocysts formed was not influenced since top quality blastocysts were found in 36.4% (SER cluster) and 41.7% of the cases, respectively. In terms of pregnancy outcome, Table 2 shows that neither number of positive βhcg nor clinical pregnancies, implantations or multiple gestations were significantly lower in patients with affected gametes. However, the rate of miscarriage was a b Figure 1. (a) Metaphase II oocyte showing a dominant aggregation of the smooth endoplasmic reticulum (38.9 µm). (b) Mature gamete showing two smaller clusters of smooth endoplasmic reticulum (25.0 and 11.1 µm). 115

Table 1. Demographic and stimulation data for cycles that were either positive or negative for smooth endoplasmic reticulum (SER). Parameter SER positive SER negative P-value Cycles (n) 30 457 Age (years) 33.0 ± 5.4 32.7 ± 4.7 NS Duration of infertility (years) 3.9 ± 3.2 4.3 ± 3.7 NS No. of previous cycles 1.3 ± 1.6 1.0 ± 1.3 NS Basal FSH (miu/ml) 7.8 ± 2.4 7.8 ± 2.8 NS Basal LH (miu/ml) 4.0 ± 2.3 5.7 ± 3.7 NS Basal oestradiol (pg/ml) 31.9 ± 10.2 35.9 ± 20.8 NS No. receiving long protocol (%) 10 (33.3) 104 (22.8) NS No. receiving antagonist protocol (%) 20 (66.7) 353 (77.2) NS Dose of gonadotrophins (IU/ml) 2190 ± 1000 1767 ± 701 <0.01 Duration of stimulation (days) 11.1 ± 2.5 9.6 ± 1.8 <0.001 Oestradiol at ovulation induction (pg/ml) 1397 ± 881 1270 ± 779 NS Progesterone at ovulation induction (ng/ml) 0.7 ± 0.3 0.7 ± 0.4 NS LH at ovulation induction (miu/ml) 2.2 ± 2.3 2.4 ± 2.5 NS Metaphase II oocytes 6.8 ± 4.0 6.7 ± 4.6 NS Values are mean ± SEM unless otherwise stated; NS = not statistically significant. Table 2. Pregnancy outcome for cycles that were either positive or negative for smooth endoplasmic reticulum (SER). Parameter SER positive SER negative P-value Cycles (n) 30 457 Positive βhcg 12 (40.0) 211 (46.2) NS Clinical pregnancy 8 (26.7) 188 (41.1) NS Miscarriage/biochemical 7/12 (58.3) 47/211 (22.3) <0.01 Take-home baby rate 5/12 (41.7) 164/211 (77.7) <0.01 Implantation rate 14/53 (26.4) 265/739 (35.9) NS Multiple pregnancy rate 2/12 (16.7) 54/211 (25.6) NS Singletons 6 111 Obstetric problems 2/6 (33.3) 6/111 (5.4) <0.01 Birthweight (g) a 2517 ± 887 3128 ± 621 <0.05 Weeks of gestation a 36.4 ± 4.0 39.0 ± 2.0 NS Sex (male/female) 3/3 (1.0) 57/61 (0.93) NS Neonatal deaths 2/6 (33.3) 0/111 <0.001 Malformations 1/6 (16.7) 5/111 (4.5) NS NICU 1/6 (16.7) 7/111 (6.3) NS Values are numbers (percent) unless otherwise stated. HCG: human chorionic gonadotrophin; NICU: neonatal intensive care unit; NS = not statistically significant. a Mean ± SEM. 116

definitely higher (P = 0.005) in the affected group (58.3%) as compared with healthy patients (22.3%). Consequently, takehome baby rate showed the opposite trend. Regarding birth, analysis of obstetric data (Table 2) revealed that SER cluster-positive women tended to deliver earlier (at 36.4 weeks of gestation) than healthy women (39.0 weeks). This is probably related to a higher percentage of obstetric complications during pregnancy (P = 0.008). As a consequence, birthweight in singletons was found to be significantly lower (P = 0.029) in this group (c. 2500 g versus 3100 g). More importantly, the rate of neonatal death was alarming in the affected patient cohort since two children died within 24 h after birth (both patients had had double embryo transfer including one SER-positive conceptus) whereas all babies survived in the normal control group. In terms of malformations, both patient groups had similar rates. In detail, one case of diaphragmatic hernia was diagnosed in the SER-positive group (1/6; 16.7%), whereas five malformations (Freeman-Sheldon syndrome, brain defect, renal agenesis, hydronephrosis, ovarian hernia) were observed in the unaffected counterparts (5/111; 4.5%). Discussion Empirical data show that oocytes with perfectly normal appearance can develop into normal-appearing embryos being developmentally incompetent. This implies that intrinsic oocyte-specific defects of important molecular and cellular activities do exist that are hardly detectable using conventional microscopic techniques. Electron microscopic studies (Van Blerkom, 1990) stressed that cytoplasmic organization of normal-appearing oocytes is characterized by an apparently random distribution of cell organelles. On the other hand, oocytes showing deviations in fertilization or further development often display atypical distributions of cellular components (Van Blerkom, 1990). A disk-like assemblage of the SER would represent such a cytoplasmic defect that could adversely influence competence (Van Blerkom and Henry, 1992). It is presumed that these clusters arise by dilatation and fusion of SER saccules during maturational process of the gamete (Van Blerkom, 1990). The relatively low frequency of this oocyte dysmorphism has kept several authors from analysing it separately (Serhal et al., 1997; Balaban et al., 1998). Others, however, scored between 1 3% of all analysed gametes to be SER cluster positive (Van Blerkom, 1990; Alikani et al., 1995; Corn et al., 2005; Ebner et al., 2006b). These figures may be somewhat underestimated, since it became apparent from the findings of Otsuki et al. (2004) that smaller clusters (e.g. 2 5 µm) cannot be detected without the help of an electron microscope. These smaller aggregations were exclusively found in presumed SER-negative oocytes (on the basis of light-microscopic analysis) of SER cluster-positive women (Otsuki et al., 2004), and thus no bias was introduced into statistical analysis, as probably no SER-positive patients were missed. In addition, these smaller aggregations were exclusively situated in the cortical area of the ooplasm and it is not clear whether they have the same detrimental effect as larger accumulations near the centre of the egg. It should be noted that the smallest aggregation detected in this prospective study was 5.1 µm in size and found in an egg with three clusters. Otsuki et al. (2004) also showed that SER clusters are an exclusive anomaly of MII oocytes (not detected in prophase or metaphase I gametes) and tend to grow in diameter before disappearing approximately 16 20 h post-icsi. This phenomenon may indeed explain the observed decrease in fertilization rate in the present study since pronuclear formation might be constrained by larger aggregations of the SER, a situation similar to that found in vacuolized oocytes (Ebner et al., 2005). Once the oocyte is fertilized, no further limitation in development (cleavage, compaction) was recognizable up to day 3 (see also Otsuki et al., 2004) and day 4. On day 5, however, significantly fewer embryos reached blastocyst stage if they derived from SER cluster-positive ova as compared with their SER-negative counterparts, which favourably supports the low blastocyst formation rate (18.2%) seen in previous reports (Otsuki et al., 2004). It can be assumed that this is the manifestation of an intrinsic oocyte defect eventually caused by a suboptimal ovarian stimulation (Van Blerkom, 1990). Ultimately, present data indicate that the risk of having at least one affected gamete is positively related to the duration and dosage of the stimulation. This raises the question about whether at least some patients are over-stimulated in IVF, as suggested by Sathananthan et al. (1988). Further evidence comes from the findings of Japanese colleagues (Otsuki et al., 2004) who could demonstrate a positive correlation between the occurrence of SER clusters and the serum oestradiol concentration on the day of ovulation induction. Previous data (Ebner et al., 2006b) are in line with this hypothesis, since patients with higher concentrations of anti-müllerian hormone, and consequently with higher concentrations of oestradiol, were shown to produce a higher percentage of gametes with SER aggregations (6.4%) as compared with patients with anti-müllerian hormone within normal ranges (1.7%). A prospective analysis of gynaecological and obstetric data revealed that patients with one or more gametes showing an aggregation of the SER have to face a higher risk of miscarriage (P < 0.01). Even though they may reach birth, there is also tendency towards premature delivery resulting in a significantly lower birthweight (P < 0.05). An alarming finding is the observation of two stillbirths in affected patients while no such outcome was noted in the unaffected control group. Keeping in mind the pivotal role of SER in calcium storage and release, it may be postulated that Ca 2+ balance might be affected in SER-positive oocytes and the corresponding embryos (Otsuki et al., 2004). Since developmental failures appear to originate from subtle anomalies in cell structure or function, oocytes displaying gross deviations from normal patterns of SER distribution might be most affected. However, to date, the molecular and cellular activities that lead to gametes with an impaired SER are still unknown. The present phenomenon that even transfer of presumed normal embryos (without SER cluster in the corresponding egg) from affected patients may lead to a detrimental effect on obstetric and neonatal outcome may be explained by the possibility that, in fact, oocytes were not devoid of organelle aggregation but much rather smaller aggregations could have been below light-microscopic resolution. An additional explanation comes from Meriano and co-workers (2001) who suggested that an 117

118 underlying adverse factor might impair the entire follicular cohort, regardless of detectability of a dysmorphism by means of a light microscope. These authors (Meriano et al., 2001) also noticed a repetitive nature of SER clusters from one cycle to another. This is in line with observations that, out of all non-pregnant patients in the study group (n = 13), five (38.5%) showed repetitive aggregation of the SER in the following cycle (Ebner et al., unpublished data). This brings us to the question of whether a change in stimulation protocol (e.g. from antagonist to long protocol or vice versa) and/or dosage would reduce the possibility of producing oocytes with suboptimal maturation, e.g. resulting in organelle clustering. Indeed, the frequency of this anomaly was three times higher using a short protocol as compared with the long protocol (Otsuki et al., 2004). Although these data did not reach statistical significance because of the relatively small sample size, some stimulation regimens could be worse than others. It should be kept in mind that SER clusters could also be present in natural conception and, thus, be responsible for miscarriage and other obstetric problems in unstimulated cycles. In summary, lower fertilization and blastulation rates in oocytes with SER clusters as well as bad obstetric and neonatal outcomes lead us to suggest that transfer of embryos/blastocysts deriving from SER cluster-positive eggs should be strictly avoided. References Alikani M, Palermo G, Adler A et al. 1995 Intracytoplasmic sperm injection in dysmorphic human oocytes. Zygote 3, 283 288. Balaban B, Urman B 2005 Effect of oocyte morphology on embryo development and implantation. Reproductive BioMedicine Online 12, 608 615. Balaban B, Urman B, Sertac A et al. 1998 Oocyte morphology does not affect fertilization rate, embryo quality and implantation rate after intracytoplasmic sperm injection. Human Reproduction 13, 3431 3433. Borini A, Lagalla C, Cattoli M et al. 2005 Predictive factors for embryo implantation potential. Reproductive BioMedicine Online 10, 653 668. Corn CM, Hauser-Kronberger C, Moser M et al. 2005 Predictive value of cumulus cell apoptosis with regard to blastocyst development of corresponding gametes. Fertility and Sterility 84, 627 633. De Sutter P, Dozortsev D, Qian C et al. 1996 Oocyte morphology does not correlate with fertilization rate and embryo quality after intracytoplasmic sperm injection. Human Reproduction 11, 595 597. Ebner T, Moser M, Tews G 2006a Is oocyte morphology prognostic of embryo developmental potential after ICSI? Reproductive BioMedicine Online 12, 507 512. Ebner T, Sommergruber M, Moser M et al. 2006b Basal level of anti- Müllerian hormone is associated with oocyte quality in stimulated cycles. Human Reproduction 21, 2022 2026. Ebner T, Moser M, Sommergruber M et al. 2005 Occurrence and developmental consequences of vacuoles throughout preimplantation development. Fertility and Sterility 83, 1635 1640. Ebner T, Yaman C, Moser M et al. 2001 A prospective study on oocyte survival rate after ICSI: influence of injection technique and morphological features. Journal of Assisted Reproduction and Genetics 18, 601 606. Gardner DK, Lane M, Stevens J et al. 2000 Blastocyst score affects implantation and pregnancy outcome: towards a single blastocyst transfer. Fertility and Sterility 73, 1155 1158. Loutradis D, Drakakis P, Kallianidis K et al. 1999 Oocyte morphology correlates with embryo quality and pregnancy rate after intracytoplasmic sperm injection. Fertility and Sterility 72, 240 244. Meriano JS, Alexis J, Visram-Zaver S et al. 2001 Tracking of oocyte dysmorphisms for ICSI patients may prove relevant to the outcome in subsequent patient cycles. Human Reproduction 16, 2118 2123. Mikkelsen AL, Lindenberg S 2001 Morphology of in-vitro matured oocytes: impact on fertility potential and embryo quality. Human Reproduction 16, 1714 1718. Otsuki J, Okada A, Morimoto K et al. 2004 The relationship between pregnancy outcome and smooth endoplasmic reticulum clusters in MII human oocytes. Human Reproduction 19, 1591 1597. Racowsky C, Jackson KV, Cekleniak NA et al. 2000 The number of eight-cell embryos is a key determinant for selecting day 3 or day 5 transfer. Fertility and Sterility 73, 558 564. Rijnders PM, Jansen CAM 1999 Influence of group culture and culture volume on the formation of human blastocysts: a prospective randomized study. Human Reproduction 14, 2333 2337. Sathananthan AH, Ng SC, Ratnam SS et al. 1988 Are we overstimulating in IVF? Journal of Obstetrics and Gynecology (Singapore) 19, 83 88. Serhal PF, Ranieri DM, Kinis A et al. 1997 Oocyte morphology predicts outcome of intracytoplasmic sperm injection. Human Reproduction 12, 1267 1270. Van Blerkom J 1990 Occurrence and developmental consequences of aberrant cellular organization in meiotically mature human oocytes after exogeneous ovarian hyperstimulation. Journal of Electron Microscopical Technics 16, 324 346. Van Blerkom J, Henry G 1992 Oocyte dysmorphisms and aneuploidy in meiotically mature human oocytes after ovarian stimulation. Human Reproduction 7, 379 390. Declaration: The authors report no financial or commercial conflicts of interest. Received 11 July 2007; revised and resubmitted 31 August 2007; refereed 14 September 2007; accepted 1 October 2007.