Article Blastocyst culture and transfer: lessons from an unselected, difficult IVF population

Transcription

1 RBMOnline - Vol 17. No Reproductive BioMedicine Online; on web 27 June 2008 Article Blastocyst culture and transfer: lessons from an unselected, difficult IVF population Dr Ariel Weissman graduated from the Hadassah-Hebrew University Medical School in Jerusalem in In 1994 he completed his residency in Obstetrics and Gynaecology at the Kaplan Medical Center, Rehovot, where he spent another 2 years working as a senior physician at the IVF unit. A 2-year research and clinical fellowship with Professor Bob Casper at the University of Toronto followed, where his research focused on transplantation of human ovarian tissue in immuno-deficient mice. Returning to Israel in 1998, he joined the IVF unit at the Wolfson Medical Center, Holon, Sackler Tel Aviv University School of Medicine, where he currently holds the position of Senior Lecturer. Dr Ariel Weissman Ariel Weissman 1,3, Galia Biran 1, Hana Nahum 1, Marek Glezerman 2, David Levran 1 1 IVF Unit, Department of Obstetrics and Gynecology, Wolfson Medical Centre, Holon, Israel; 2 Department of Obstetrics and Gynecology, Rabin Medical Centre, Beilinson Campus, Petah Tikva, Sackler Faculty of Medicine, Tel Aviv University, Israel 3 Correspondence: Tel: ; Fax: ; a_w@zahav.net.il Abstract Blastocyst-stage transfer has yielded excellent results in good prognosis IVF patients, but its efficacy in the general IVF population has not been clearly demonstrated. The objective of this study was to compare cleavage-stage and blastocyst-stage transfer in a mixed, general IVF population. In a prospective, quasi-randomized study, 152 patients underwent 164 treatment cycles. Patients were allocated to cleavage-stage (group 1; n = 94) or blastocyst-stage (group 2; n = 70) transfer. Main outcome measures included implantation, clinical pregnancy and live birth rates. Implantation (11.2% versus 15.5%), clinical pregnancy (34% versus 21%) and live birth rates per transfer (21.3% versus 13.8%) and per started cycle (21.3% versus 11.4%) were all comparable for groups 1 and 2, respectively. Logistic regression analysis revealed that blastocyst culture and transfer reduced the odds for pregnancy in the general IVF population and defined a good prognosis group for blastocyst transfer. Introducing blastocyst culture and transfer to all IVF patients is not advantageous. Blastocyst transfer should be offered primarily to good prognosis patients, and this group should be specifically defined in each clinical set-up. Keywords: blastocyst transfer, embryo transfer, infertility, IVF/ICSI outcome Introduction 220 Culture and transfer of human embryos at the blastocyst stage has gained widespread popularity during the last decade. The low implantation rates, in the range of 10 20%, commonly observed following transfer of cleavage-stage (day 2 or 3) embryos indicate that the efficiency of this conventional approach is rather low, since >80% of embryos never implant. In order to try to overcome the implantation barrier, it has been common in the past to transfer a high number of embryos to the uterus, resulting in an unacceptably high rate of multiple pregnancies, the major complication of assisted reproduction treatment. Advances in understanding the different metabolic needs of cleavage- and blastocyst-stage embryos have resulted in the development of sequential media systems, which have improved the capability of in-vitro blastocyst formation. Transfer at the blastocyst stage has been reported to improve success rates by identifying potentially superior quality embryos. By using sequential media, the reported rate of blastocyst formation varies between 33 and 93% (Jones et al., 1998; Plachot et al., 2000; Shapiro et al., 2000; Alper et al., 2001; Fisch et al., 2001; Rienzi et al., 2002). The possible advantages of culture and transfer at the blastocyst stage include: (i) selection of embryos with increased implantation potential after the initiation of embryonic genome activation (Braude et al., 1988); (ii) improved endometrial embryonic synchrony (Garcia-Velasco and Simon, 2001); (iii) transfer of fewer embryos leading to reduced probability of high-order multiple pregnancies; (iv) transfer of embryos with a greater likelihood of being chromosomally competent (Jones and Trounson, 1999; Magli et al., 2000); and (v) ample time and possibilities to perform preimplantation genetic diagnosis (Garcia-Velasco and Simon, 2008 Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB3 8DB, UK

2 2001). The potential benefits and drawbacks of blastocyst transfer have been recently reviewed (Quea et al., 2007). While early studies following blastocyst culture and transfer in good prognosis patients reported excellent outcomes (Gardner et al., 1998a; Jones et al., 1998; Behr et al., 1999), subsequent trials that have looked at implementation of blastocyst transfer programmes in the general IVF population have yielded conflicting results (Scholtes and Zeilmaker, 1998; Marek et al., 1999; Coskun et al., 2000; Hsieh et al., 2000; Huisman et al., 2000; Plachot et al., 2000; Karaki et al., 2002). Like others, the authors of this report have also found the concept of extended culture and transfer at the blastocyst stage transfer attractive. Therefore, a prospective study was conducted in order to compare the efficiency of blastocyst-stage transfer with transfer of cleavage-stage embryos in the general IVF population. An additional search was carried out for predictors of blastulation and pregnancy in a prospective series of IVF patients undergoing blastocyst culture and transfer. Materials and methods All patients undergoing IVF treatment during an 8-month period in the IVF unit who had more than three cleavage-stage embryos on day 2 or 3 after oocyte retrieval were allocated into one of two groups. Patients who had oocyte retrieval on Sundays, Mondays and Wednesdays had cleavage-stage (day 2 or 3) embryo transfers, and those who had oocyte retrieval on Tuesdays, Thursdays and Fridays had their embryos cultured to the blastocyst stage, and transferred on day 5 or 6 after oocyte retrieval. This policy was adopted in order to avoid Saturday transfers, which are not possible in the current clinic set-up, and also to reduce workload for the laboratory personnel (avoiding culture media changes) on Saturdays. Since both approaches for embryo transfer were common practice in Israel by the time the study was conducted, institutional review board approval was not considered necessary and was therefore not obtained. A total of 152 patients underwent 164 treatment cycles (12 patients had two cycles each): 94 cycles with day 2 3 transfer (cleavage-stage group), and 70 cycles with day 5 6 transfer (blastocyst group). The difference in group size can be explained by the fact that fewer retrievals take place on Fridays, as it is already part of the weekend and a short work day. Patients underwent controlled ovarian stimulation with a gonadotrophin-releasing hormone (GnRH) analogue and recombinant FSH/human menopausal gonadotrophin with the use of either conventional down-regulation or flare protocols. Human chorionic gonadotrophin (HCG; 10,000 IU; Chorigon; Teva, Nethanya, Israel) was administered i.m. when at least two follicles were 18 mm in diameter, and serum oestradiol concentrations were within the acceptable range as determined by the number of follicles present. Oocyte retrieval was performed 36 h after HCG administration by transvaginal ultrasound-guided needle aspiration under general anaesthesia. Retrieved oocytes were fertilized by either conventional IVF or intracytoplasmic sperm injection (ICSI), as described in detail elsewhere (Levran et al., 1998). Fertilization was assessed h following insemination, and considered normal when two distinct pronuclei were present. The embryos were cultured in P1 medium (Irvine Scientific, Santa Ana, CA, USA) supplemented with 10% synthetic serum substitute (SSS;Irvine Scientific) at 37 C in 5% CO 2 up to day 3. Embryos designated for extended culture were transferred to blastocyst medium (Irvine Scientific) supplemented with 20% SSS. Embryo cleavage and quality were assessed and recorded at h post-retrieval. Criteria for embryo grading were based on gross morphological characteristics under light microscopy, according to regularity and symmetry of blastomeres and degree of fragmentation (Scott et al., 1991). Blastocyst quality was determined on the day of blastocyst transfer according to Gardner s criteria (Gardner et al., 2000). In both groups, transfer was performed using Cook catheter (Soft Pass, J-SPPE; Cook Ob/Gyn, Spencer, IN, USA) using a standard technique. Cleavage-stage transfers were performed preferably on day 3 after oocyte retrieval, unless day 3 coincided with the weekend, in which case day 2 transfers were performed. Blastocyst transfers were performed on day 5 after oocyte retrieval unless blastocysts were available only on day 6. The number of embryos transferred was individualized based on the age of the patient, number of previous failed attempts and embryo/blastocyst availability. Patients in both groups were administered luteal phase support by either transvaginal micronized progesterone 200 mg three times daily (Uterogestan; Besins Iscovesco, Paris, France) and progesterone in oil 50 mg i.m. on alternate days when peak oestradiol concentrations exceeded 2500 pg/ml, or by i.m. HCG 2500 IU (Chorigon) given twice 4 and 6 days after oocyte retrieval when oestradiol concentrations were lower than 2500 pg/ml. Pregnancy was diagnosed by presence of a gestational sac on ultrasound 5 weeks from oocyte retrieval. Implantation rate was determined by dividing the number of gestational sacs by the number of embryos transferred. Data analysis The data recorded and analysed were: the age of the patient, day 3 FSH concentration, number of previous IVF attempts, main diagnosis of infertility, type of ovarian stimulation protocol, serum oestradiol concentration on the day of HCG administration, number of oocytes retrieved and fertilized, number of developing embryos, number of 6- to 8-cell embryos and 8-cell embryos on day 3, number of grade I embryos, number of embryos transferred, number of blastocysts, number of blastocysts transferred, number of high-quality blastocysts, pregnancy rate, implantation rate and pregnancy outcome. Analysis of data was carried out using SPSS statistical analysis software (SPSS Inc., Chicago, IL, USA). Descriptive statistics for continuous variables such as age, number of embryos and cell number were calculated and are reported as mean ± SD. Distributions of continuous variables were tested for normality using the Kolmogorov Smirnov method. Variables with distributions not differing significantly from normal were compared using the t-test for independent samples. Variables with distributions significantly differing from normal were 221

3 tested using the Mann Whitney non-parametric U-test. Both methods used day 2 3 embryo transfer versus blastocyst transfer as the independent variable. Categorical variables such as the presence or absence of specific diagnoses and the insemination method (IVF/ICSI) and treatment protocol were described using frequency distributions and percentages. Chisquared test (with Yates correction as needed) was used to detect differences in categorical variables by day 2 3 embryo transfer versus blastocyst transfer. Logistic regression analysis was used to model pregnancy and, separately, clinical pregnancy using a forward, stepwise approach. All tests were two-tailed and considered significant at P < Results Both groups were comparable in terms of patient age, main indication for IVF treatment, number of previous IVF attempts, and day 3 FSH concentrations (Table 1). Regarding cycle characteristics (Table 2), serum oestradiol concentrations on the day of HCG administration were significantly higher in the blastocyst group, and the number of oocytes retrieved and fertilized was also higher in that group (P = , P = and P = 0.005, respectively). Fertilization rate was comparable for the two groups. Day 3 parameters for both groups are presented in Table 2. The number of 8-cell embryos was significantly higher (P = 0.005) in the blastocyst group but the proportion of high-quality embryos (grade I) was comparable. The overall blastulation rate was 32.7% (222/678), and in 58/70 (82.9%) of the cycles there was at least one blastocyst available for transfer. Twelve patients (17.1%) had no blastocyst available for transfer. The mean number of transferred blastocysts was 2.1 ± 1.3, significantly lower than the number of embryos transferred on day 2 3 (3.3 ± 1.1; P < 0.001). The mean number of high-quality blastocysts (3AA/4AA) was 1.8 ± 1.92 per cycle in the blastocyst transfer group, and these comprised 51.7% of blastocysts formed. Pregnancy rates and outcome are presented in Table 3. Implantation rate was higher in the blastocyst group (15.2% versus 11.5%) but this difference did not reach statistical significance. There were 32 clinical pregnancies (34.0%) following cleavage stage transfers and 15 (21.4%) with blastocyst transfers. Multiple pregnancy rates were also similar for the two groups. There was only one triplet pregnancy in the entire study group (0.6% per cycle), and this occurred following cleavage-stage transfer. There were three ectopic pregnancies in the blastocyst transfer group and one ectopic pregnancy following cleavage-stage transfer. Logistic regression was used to model blastulation and pregnancy. In the total study population, each additional year of age reduced the odds of pregnancy by 11% [odds ratio (OR) 0.89, 95% confidence interval (CI) , P = ]. The long protocol more than tripled the odds of pregnancy compared with the short protocol (OR 3.4, 95% CI , P = 0.038). Being in the blastocyst group reduced the odds of pregnancy by nearly 52% (OR 0.48, 95% CI , P = 0.047). This model correctly classified pregnancy in 65.2% of cycles. Predictors for blastocyst formation included the following; (i) number of retrieved oocytes: each additional oocyte increased the chance for blastocyst formation by 28% (OR 1.28, CI 95%, , P = 0.006). Odds for blastocyst formation increased significantly when the number of oocytes retrieved was 8 (OR 7.4, 95% CI , P = 0.008). This model correctly classified 82.6% of the cases; (ii) each additional fertilized oocyte increased the odds of blastocyst formation by 66% (OR 1.67, CI , P = 0.03); (iii) each additional embryo available on day 2 3 increased the odds of blastocyst formation (OR 1.66, 95% CI , P = 0.03). When five or more embryos were available, the odds of blastocyst formation were significantly increased (OR 26.5, 95% CI , P < Table 1. Clinical characteristics of the patients in the two study groups. Characteristic Cleavage-stage Blastocyst group (n = 94) group (n = 70) Age (years) 32.8 ± ± 5.1 Main indication for treatment Tubal 22 (31.4) 27 (29) Male 33 (47.1) 39 (41.9) Anovulation 4 (5.7) 4 (4.3) Endometriosis 1 (1.4) 2 (2.2) Uterine factor 1 (1.4) 3 (3.2) Low response 0 (0) 1 (1.1) Unexplained 9 (12.9) 17 (18.3) No. of previous treatment attempts 3.8 ± ± 4.3 Day 3 FSH (IU/l) 7.3 ± ± Values are means ± SD or number (%). There were no statistically significant differences between the two groups.

4 Table 2. Clinical and laboratory cycle characteristics in the two study groups. Characteristic Cleavage-stage Blastocyst P-value group (n = 94) group (n = 70) Protocol Long 78 (83.0) 61 (87.1) Short 16 (17.0) 9 (12.9) NS Peak oestradiol concentration (pg/ml) ± ± No. of oocytes retrieved 11.0 ± ± No. of oocytes fertilized 7.7 ± ± Fertilization rate, % NS ICSI cycles 65 (69) 50 (71.4) NS No. of embryos 7.5 ± ± No. of 6- to 8-cell embryos 2 ± ± 2.3 NS No. of 8-cell embryos 1.5 ± ± No. of grade I embryos 2 ± ± 2 NS No. of embryos transferred 3.3 ± ± 1.3 <0.001 Values are means ± SD or number (%), unless otherwise indicated. ICSI = intracytoplasmic sperm injection; NS = not statistically significant. Table 3. Pregnancy rates and outcomes in the two study groups. Parameter Cleavage-stage Blastocyst group (n = 94) group a (n = 70) Pregnancy 36 (38.3) 20 (28.5) (34.5) Clinical pregnancy 32 (34.0) 15 (21.4) (25.8) Live birth 20 (21.3) 8 (11.4) (13.8) Multiple pregnancies 6 (18.8) 5 ( ) (33) Miscarriage 8 (25.0) 3 ( ) (20) Ectopic pregnancy 1 (2.8) 3 ( ) (15) Implantation rate 41/357 (11.5) 22/141 ( ) (15.2) Values are number (%), unless otherwise indicated. There were no statistically significant differences between the groups. a Values in parentheses are percentages per cycle and per transfer, respectively. The per transfer values exclude data for 12 patients who had no blastocyst available for transfer ); and (iv) the number of 8-cell embryos present on day 3 also improved the odds of blastocyst formation. Each additional 8-cell embryo doubled the chances for blastocyst (OR 2.3, 95% CI , P = 0.01). When each parameter was analysed separately in the logistic regression model for pregnancy in the blastocyst group, each additional year in the age of the patient decreased the chances for pregnancy by about 27% (OR 0.74, 95% CI , P = 0.02). Each previous failed IVF attempt decreased the chance for pregnancy by 57% (OR 0.43, 95% CI , P = 0.003). Each additional blastocyst that was formed and each additional blastocyst that was transferred increased the odds of pregnancy (OR 1.28, 95% CI , P = 0.02, and OR 1.95, 95% CI , P = 0.02, respectively). The final model included age (OR 0.78, 95% CI , P = 0.008) and the number of previous IVF attempts (OR 0.47, 95% CI , P = 0.009) and correctly classified pregnancy in 83% of the cases. Discussion The results of this study demonstrate the lack of superiority of blastocyst-stage transfer over cleavage-stage transfer in the unselected IVF population in terms of implantation, pregnancy and live birth rates. Nevertheless, even in mixed and difficult IVF populations such as this, by statistical modelling of the data it was possible to define the patient population that is likely both to produce blastocysts and to achieve pregnancy following blastocyst culture and transfer. Mixed IVF populations vary by country and even by clinic. The unselected IVF population of this study is rather difficult, consisting of couples with a mean of 3.5 previous failed cycles. This is certainly different than the general population presenting for IVF in other Western countries, where patients often elect not to have multiple IVF attempts simply because of the financial, 223

5 224 physical and emotional burdens, and choose other options such as adoption or gamete donation. In Israel, full coverage for IVF is provided by the national health insurance programme and couples are entitled to an almost unlimited number of IVF attempts, as long as embryos are available for transfer, and up to the birth of two children. Consequently, patients with multiple failed attempts are common in many Israeli clinics. Therefore, the results of this study are population specific, and should be viewed as such. It has been demonstrated that pregnancy rates do not change over the first three IVF embryo transfer cycles but decrease by 40% for four or more prior failed attempts (Templeton and Morris, 1998). Blastocyst culture and transfer has been proposed as one of the strategies to overcome the challenge of achieving implantation and pregnancy in patients with multiple failed attempts. In a small retrospective cohort study, Cruz et al. (1999) reported that transfer at the blastocyst stage resulted in significantly higher implantation and pregnancy rates as compared with conventional day-3 embryo transfer in patients with multiple failed IVF attempts. Levitas et al. (2004) reported the results of a prospective, randomized study on cleavagestage versus blastocyst-stage transfer in patients with three or more failed IVF attempts. Implantation rate was significantly higher following blastocyst-stage transfer (15.2%) as compared with cleavage-stage transfer (11.5%). Clinical pregnancy rate per oocyte retrieval and the multiple pregnancy rate were comparable for both groups. Guerif et al. (2004) allowed patients who failed to conceive after two or more cleavagestage transfers to choose between day-2 embryo transfer and blastocyst transfer in the subsequent cycle. The live birth rates per cycle (27.9% versus 19.7%) and implantation rates per cycle (25.4% versus 12.4%) were higher in the blastocyst group compared with the cleavage-stage group, respectively. Recently, Barrenetxea et al. (2005) have also suggested that blastocyst transfer appears to be a successful and improved alternative for patients with multiple (more than three) failed IVF attempts. The authors have previously failed to observe any benefit from culture and transfer at the blastocyst stage in patients with repeated implantation failure (Levran et al., 2002). In the present study, it was once again observed that each previous failed IVF attempt decreased the chance for pregnancy by 57%. Support for these findings can be found in the study of Shapiro et al. (2001), who observed dramatic declines in pregnancy and implantation rates in repeated IVF cycles with blastocyst transfer, following one or more unsuccessful cycles with blastocyst transfer. Twelve patients (17.1%) had no blastocyst available for transfer. This is a well-known risk of extended culture to the blastocyst stage that is extremely devastating for the couples, and of which patients should be properly informed. According to the recent Cochrane review (Blake et al., 2007), failure to transfer any embryos per couple is indeed significantly higher in the blastocyst group (OR 2.85, 95% CI ; day2/3 2.8% versus day 5/6 8.9%). More cycles (n = 94) were included in the early cleavage transfer group as compared with the blastocyst transfer group (n = 70). This reflects the policy to keep the number of Friday retrievals at a minimum in order to reduce the workload during the weekend. The concept of randomization according to the weekday of oocyte retrieval has been used in the past (Scholtes and Zeilmaker, 1996; Huisman et al., 2000; Utsunomiya et al., 2002) and is not believed to introduce a selection bias in the current set-up, as a team approach is used, with strict criteria for HCG administration. Theoretically, however, the physician writing orders and scheduling a patient for oocyte retrieval could consciously or unconsciously influence the inclusion of a given patient in a specific group, and randomization according to the weekday is far from being ideal. As such a method of randomization is not supported by CONSORT standards, the present study cannot be considered as a randomized controlled trial. It is difficult to explain why, despite comparable baseline clinical characteristics (Table 1), mean serum oestradiol concentrations and number of oocytes retrieved and fertilized, as well as the total number of embryos and 8-cell embryos available on day 3 (Table 2), were all significantly higher in the blastocyst-transfer group. Nevertheless, while all of the above differences should be considered advantageous, being good prognostic factors for both blastocyst formation and for pregnancy (Jones et al., 1998; Racowsky et al., 2000), this potential advantage did not translate into higher implantation, pregnancy and delivery rates in the blastocyst transfer group. Again, the inclusion of patients with multiple failed attempts in the study population may serve as a partial explanation for this observation. While several groups of investigators have reported increased implantation rates following blastocyst transfer to unselected IVF patients, others have failed to observe any benefit compared with cleavage-stage embryo transfers. High implantation rates after blastocyst transfer were presented in comparative or retrospective studies (Alves da Motta et al., 1998; Gardner et al., 1998b; Marek et al., 1999; Milki et al., 1999, 2000; Balaban et al., 2001a; Langley et al., 2001; Yoon et al., 2001; Wilson et al., 2002), but randomized controlled trials have yielded conflicting results. Some of these randomized trials (Scholtes and Zeilmaker, 1996; Coskun et al., 2000; Hsieh et al., 2000; Huisman et al., 2000; Plachot et al., 2000; Karaki et al., 2002; Levron et al., 2002; Rienzi et al., 2002; Van der Auwera et al., 2002; Emiliani et al., 2003; Pantos et al., 2004) are summarized in Table 4. A closer look at these trials reveals a considerable heterogeneity in almost every variable that was studied. Studies vary in terms of patient inclusion criteria, timing of cleavage-stage and blastocyst transfers, culture media used as well as outcome variables. A recent Cochrane review (Blake et al., 2007) of evidence-based data from randomized controlled trials concluded that there is a significant difference in pregnancy and live birth rates in favour of blastocyst transfer in good prognosis patients, and those with high numbers of 8-cell embryos on day 3 are the most favoured subgroup. No clear recommendation for blastocyst culture and transfer in the general IVF population was given. The blastocyst formation rate (33%) in group 1 is in accordance with figures recently reported elsewhere for mixed IVF populations (Coskun et al., 2000; Racowsky et al., 2000; Karaki et al., 2002; Levron et al., 2002). Several authors have previously reported on clinical and laboratory indicators of subsequent blastocyst development. It has been demonstrated that early pronuclear embryo assessment can predict goodquality blastocyst development (Scott et al., 2000; Balaban et al., 2001b; Neuber et al., 2003). Others have found the

6 Table 4. Summary of prospective randomized controlled studies comparing cleavage-stage with blastocyst-stage transfer. Study Day 2 3 transfer (day; no. of cycles) Day 5 6 transfer (day; no. of cycles) Inclusion criteria Blastulation rate (%) Cycles without transfer in blastocyst group (%) Implantation rate (%) Clinical pregnancy rate/ transfer (%) Ongoing/ delivery rate (%) Multiple pregnancy rate (%) Remarks Van der Auwera et al. (2002) Emiliani et al. (2003) Plachot et al. (2000) Karaki et al. (2002) Rienzi et al. (2002) Pantos et al. (2004) Huisman et al. (2000) Coskun et al. (2000) Levron et al. (2002) Scholtes and Zeilmaker (1996) Hsieh et al. (2000) 2; 63 5; 66 No preselection vs vs vs vs 37.5 Only five 2PN cultured in day 2 ET group 2; 94 5; 99 4 previous IVF cycles, <39 years old, 4 zygotes on day 1 2; 60 5; 50 3 embryos on day 2 3; 82 5; 80 5 zygotes on day 1 3; 48 5; 50 <38 years old, ICSI only, 8 zygotes on day 1 2; 81, 3; 81 3; 590, 4; 488 6; years old, 3 previous IVF cycles vs vs vs vs vs vs 36 In-house sequential medium NA 25.8 vs vs vs 29 NA 48 vs NA 35 vs vs vs 48 NA (day 2) vs 16 (day 3) vs (day 6) 5; 709 No preselection (62% of transfers included a well developed blastocyst) 14.4 (day 3) vs 14.7 (day 4) vs 15.5 (day 5) (day 2) vs (day 3) vs (day 6) 25.3 (day 3) vs 25.8 (day 4) vs 27.8 (day 5) (day 2) vs 43.2 (day 3) vs (day 6) 21.7 vs 21.4 vs (day 2) vs (day 3) vs (day 6) 25.8 vs 28.8 vs ; 101 5; zygotes vs vs vs vs 38 3; 44 5; 46 <38 years old, <5 previous IVF cycles, >5 zygotes on day vs vs 18.6 NA 40 vs 50 Day 6 blastocyst transfers Single nonsequential medium 3; 233 5; 410 No preselection NA NA 13 vs vs 25 NA 15 vs 20 Single nonsequential medium 2; 158 5; 201 Age<40 years, 4 zygotes on day NA 10.8 vs vs vs 32.3 NA ET = embryo transfer; NA = not applicable; 2PN = embryo with two distinct pronuclei; vs = versus. 225

7 226 morphology and cell number of day-3 embryos to be predictive of blastocyst development (Racowsky et al., 2000; Shapiro et al., 2000). In contrast, others have found the predictive value of day 3 embryo morphology for subsequent blastocyst formation and quality rather limited (Rijnders and Jansen, 1998; Graham et al., 2000). A more complex approach, integrating nucleolar alignment along the pronuclear axis, regular cleavage and degree of fragmentation at the first cell division, and cell number and morphology on day-3, also known as the graduated embryo score (Fisch et al., 2001), has also been suggested to predict which cleaved embryos will form blastocysts. The graduated embryo score has been found to be more predictive for cycle outcome than a single day 3 morphological evaluation (Fisch et al., 2003). In the authors hands and experience, the number of oocytes retrieved (Jones et al., 1998) and the number of embryos (in particular 8-cell embryos) present on day 3 (Jones et al., 1998; Racowsky et al., 2000), were best predictors of blastocyst formation. With fewer than eight oocytes retrieved and/or fewer than five embryos formed there was a significantly lower chance for blastocyst formation. Patient age and the number of previous failed IVF attempts were negatively correlated with the likelihood of pregnancy in the blastocyst-transfer group, while each additional available blastocyst and each additional transferred blastocyst increased the odds of pregnancy. It has been shown that success rates for the development of viable human blastocysts, pregnancy and implantation decline significantly in women 40 years old (Pantos et al., 1999; Shapiro et al., 2002). Indeed, each additional year in patient age in the present series decreased the chance of pregnancy following blastocyst transfer by 27%. The decline in female fertility with age appears to be the result of a reduced number of oocytes and the inability of fertilized oocytes to develop to the blastocyst stage (Shapiro et al., 2002). On the other hand, patients over 40 years of age who have multiple good quality embryos can achieve favourable pregnancy rates following blastocyst transfer (Milki et al., 2002). The relatively low clinical pregnancy and live birth rates observed in both study arms reflect the difficult histories of patients who participated in the study. Patients in both groups had a mean of more than three previous failed IVF cycles and, as previously mentioned, multiple failures reduce the odds for implantation and pregnancy. The high multiple pregnancy rate in the blastocyst transfer group (33.3%, all twins) deserves further attention. Embryo transfer policy should be individualized, and the authors currently offer single embryo transfer of cleavage-stage embryos or blastocysts to good prognosis patients. The occurrence of three ectopic pregnancies in the blastocyst-transfer group (15.0% of pregnancies in this group) is most likely incidental; what appears to be an extremely high ectopic pregnancy rate may be explained by the small numbers involved. In summary, the results of this prospective evaluation suggest that in the mixed IVF population studied there are no evident differences in implantation, clinical pregnancy and live birth rates between cleavage-stage and blastocyst-stage transfers. There is no advantage in introducing blastocyst culture and transfer to the general IVF population as routine, since allocation to the blastocyst group significantly reduced the odds for pregnancy. On the other hand, by statistical analysis and modelling of the data, it was possible to define the patient population that is likely both to produce blastocysts and to achieve pregnancy following blastocyst culture and transfer. This should be regarded as the major achievement of this study, since it should now be possible to make better selective use of this important clinical tool. It appears that young patients, without multiple previous failed IVF attempts, who produce multiple oocytes and good-quality cleavage-stage embryos, are the best candidates for blastocyst culture and transfer. Obviously, this is also the group at highest risk for multiple gestations, and therefore the ability to restrict the number of blastocysts transferred without compromising the chances for pregnancy in this selective good prognosis group is a major advantage (Gardner et al., 2000). The findings of this study, along with the conflicting results of the additional randomized trials presented above, suggest that a more realistic approach to the role of blastocyst-stage transfer in assisted reproduction treatment should be adopted. The question raised by Alper et al. (2001), to blastocyst or not to blastocyst?, has a different answer for every clinical set-up. Since the success of blastocyst-stage transfer differs considerably between countries, clinics and individual patient populations, it is in the hands of every team to determine for which patients, and under which circumstances, blastocyst culture and transfer should be offered. For example, the Brussels group (Papanikolaou et al., 2005) has recently reported that a threshold of four good embryos on the third day of embryo culture appears to indicate that the patient will benefit from embryo transfer at the blastocyst stage, and this group have a better chance of achieving a live delivery than with cleavagestage embryo transfer. Furthermore, several groups have recently demonstrated that in good prognosis patients delivery rates are significantly higher following single blastocyst transfer versus transfer of a single cleavage-stage embryo (Papanikolaou et al., 2006; Zech et al., 2007). At the current status of knowledge, it is premature to assume that failure to reach the blastocyst stage indicates that the embryo never had implantation potential (Racowsky et al., 2000; Alper et al., 2001). While only a fraction of cleavage-stage embryos reaches the blastocyst stage, it is currently unknown how many of the other embryos could potentially result in a live birth if transferred at the cleavage stage. Therefore, the incorporation of blastocyst-stage transfer into routine clinical practice should be done on a selective clinic- and patient-specific basis, and should certainly not be offered to all patients. References Alper MM, Brinsden P, Fischer R, Wikland M 2001 To blastocyst or not to blastocyst? That is the question. Human Reproduction 16, Alves da Motta EL, Alegretti JR, Baracat EC et al High implantation and pregnancy rates with transfer of human blastocysts developed in preimplantation stage one and blastocyst media. Fertility and Sterility 70, Balaban B, Urman B, Alatas C et al. 2001a Blastocyst-stage transfer of poor-quality cleavage-stage embryos results in higher implantation rates. Fertility and Sterility 75, Balaban B, Urman B, Isiklar A et al. 2001b The effect of pronuclear morphology on embryo quality parameters and blastocyst transfer outcome. Human Reproduction 16, Barrenetxea G, Lopez de Larruzea A, Ganzabal T et al Blastocyst culture after repeated failure of cleavage-stage embryo transfers: a comparison of day 5 and day 6 transfers. Fertility and Sterility 83,

8 Behr B, Pool TB, Milki AA et al Preliminary clinical experience with human blastocyst development in vitro without co-culture. Human Reproduction 14, Blake DA, Farquhar CM, Johnson N, Proctor M 2007 Cleavage stage versus blastocyst stage embryo transfer in assisted conception. Cochrane Database System Reviews, CD Braude P, Bolton V, Moore S 1988 Human gene expression first occurs between the four- and eight-cell stages of preimplantation development. Nature 332, Coskun S, Hollanders J, Al-Hassan S et al Day 5 versus day 3 embryo transfer: a controlled randomized trial. Human Reproduction 15, Cruz JR, Dubey AK, Patel J et al Is blastocyst transfer useful as an alternative treatment for patients with multiple in vitro fertilization failures? Fertility and Sterility 72, Emiliani S, Delbaere A, Vannin AS et al Similar delivery rates in a selected group of patients, for day 2 and day 5 embryos both cultured in sequential medium: a randomized study. Human Reproduction 18, Fisch JD, Sher G, Adamowicz M, Keskintepe L 2003 The graduated embryo score predicts the outcome of assisted reproductive technologies better than a single day 3 evaluation and achieves results associated with blastocyst transfer from day 3 embryo transfer. Fertility and Sterility 80, Fisch JD, Rodriguez H, Ross R et al The graduated embryo score (GES) predicts blastocyst formation and pregnancy rate from cleavage-stage embryos. Human Reproduction 16, Garcia-Velasco, JA, Simon C 2001 Blastocyst transfer: does it really affect the outcome? Current Opinion in Obstetrics and Gynecology 13, Gardner DK, Lane M, Stevens J et al Blastocyst score affects implantation and pregnancy outcome: towards a single blastocyst transfer. Fertility and Sterility 73, Gardner DK, Schoolcraft WB, Wagley L et al. 1998a A prospective randomized trial of blastocyst culture and transfer in in-vitro fertilization. Human Reproduction 13, Gardner DK, Vella P, Lane M et al. 1998b Culture and transfer of human blastocysts increases implantation rates and reduces the need for multiple embryo transfers. Fertility and Sterility 69, Graham J, Han, T, Porter R et al Day 3 morphology is a poor predictor of blastocyst quality in extended culture. Fertility and Sterility 74, Guerif F, Bidault R, Gasnier O, Couet ML et al Efficacy of blastocyst transfer after implantation failure. Reproductive BioMedicine Online 9, Hsieh YY, Tsai HD, Chang FC 2000 Routine blastocyst culture and transfer: 201 patients experience. Journal of Assisted Reproduction and Genetics 17, Huisman GJ, Fauser BC, Eijkemans MJ, Pieters MH 2000 Implantation rates after in vitro fertilization and transfer of a maximum of two embryos that have undergone three to five days of culture. Fertility and Sterility 73, Jones GM, Trounson AO 1999 Blastocyst stage transfer: pitfalls and benefits. The benefits of extended culture. Human Reproduction 14, Jones GM, Trounson AO, Lolatgis N, Wood C 1998 Factors affecting the success of human blastocyst development and pregnancy following in vitro fertilization and embryo transfer. Fertility and Sterility 70, Karaki RZ, Samarraie SS, Younis NA et al Blastocyst culture and transfer: a step toward improved in vitro fertilization outcome. Fertility and Sterility 77, Langley MT, Marek DM, Gardner DK et al Extended embryo culture in human assisted reproduction treatments. Human Reproduction 16, Levitas E, Lunenfeld E, Har-Vardi I et al Blastocyst-stage embryo transfer in patients who failed to conceive in three or more day 2 3 embryo transfer cycles: a prospective, randomized study. Fertility and Sterility 81, Levran D, Farhi J, Nahum H et al Prospective evaluation of blastocyst stage transfer vs. zygote intrafallopian tube transfer in patients with repeated implantation failure. Fertility and Sterility 77, Levran D, Mashiach S, Dor J et al Zygote intrafallopian transfer may improve pregnancy rate in patients with repeated failure of implantation. Fertility and Sterility 69, Levron J, Shulman A, Bider D et al A prospective randomized study comparing day 3 with blastocyst-stage embryo transfer. Fertility and Sterility 77, Magli MC, Jones GM, Gras L et al Chromosome mosaicism in day 3 aneuploid embryos that develop to morphologically normal blastocysts in vitro. Human Reproduction 15, Marek D, Langley M, Gardner DK et al Introduction of blastocyst culture and transfer for all patients in an in vitro fertilization program. Fertility and Sterility 72, Milki AA, Hinckley MD, Behr B 2002 Comparison of blastocyst transfer to day 3 transfer with assisted hatching in the older patient. Fertility and Sterility 78, Milki AA, Hinckley MD, Fisch JD et al Comparison of blastocyst transfer with day 3 embryo transfer in similar patient populations. Fertility and Sterility 73, Milki AA, Fisch JD, Behr B 1999 Two-blastocyst transfer has similar pregnancy rates and a decreased multiple gestation rate compared with three-blastocyst transfer. Fertility and Sterility 72, Neuber E, Rinaudo P, Trimarchi JR, Sakkas D 2003 Sequential assessment of individually cultured human embryos as an indicator of subsequent good quality blastocyst development. Human Reproduction 18, Pantos K, Makrakis E, Stavrou D et al Comparison of embryo transfer on day 2, day 3, and day 6: a prospective randomized study. Fertility and Sterility 81, Pantos K, Athanasiou V, Stefanidis K et al Influence of advanced age on the blastocyst development rate and pregnancy rate in assisted reproductive technology. Fertility and Sterility 71, Papanikolaou EG, Camus M, Kolibianakis EM et al In vitro fertilization with single blastocyst-stage versus single cleavagestage embryos. New England Journal of Medicine 354, Papanikolaou EG, D Haeseleer E, Verheyen G et al Live birth rate is significantly higher after blastocyst transfer than after cleavage-stage embryo transfer when at least four embryos are available on day 3 of embryo culture. A randomized prospective study. Human Reproduction 20, Plachot M, Belaisch-Allart J, Mayenga JM et al Blastocyst stage transfer: the real benefits compared with early embryo transfer. Human Reproduction 15 (Suppl. 6), Quea G, Romero K, Garcia-Velasco JA 2007 Extended embryo culture to increase implantation rate. Reproductive BioMedicine Online 14, Racowsky C, Jackson KV, Cekleniak NA et al The number of eight-cell embryos is a key determinant for selecting day 3 or day 5 transfer. Fertility and Sterility 73, Rienzi L, Ubaldi F, Iacobelli M et al Day 3 embryo transfer with combined evaluation at the pronuclear and cleavage stages compares favourably with day 5 blastocyst transfer. Human Reproduction 17, Rijnders PM, Jansen CA 1998 The predictive value of day 3 embryo morphology regarding blastocyst formation, pregnancy and implantation rate after day 5 transfer following in-vitro fertilization or intracytoplasmic sperm injection. Human Reproduction 13, Scholtes MC, Zeilmaker GH 1998 Blastocyst transfer in day-5 embryo transfer depends primarily on the number of oocytes retrieved and not on age. Fertility and Sterility 69, Scholtes MC, Zeilmaker GH A prospective, randomized study of embryo transfer results after 3 or 5 days of embryo culture in in vitro fertilization. Fertility and Sterility 65, Scott L, Alvero R, Leondires M, Miller B 2000 The morphology of human pronuclear embryos is positively related to blastocyst development and implantation. Human Reproduction 15,

9 Scott RT, Jr, Hofmann GE, Veeck LL et al Embryo quality and pregnancy rates in patients attempting pregnancy through in vitro fertilization. Fertility and Sterility 55, Shapiro BS, Richter KS, Harris DC, Daneshmand ST 2002 Influence of patient age on the growth and transfer of blastocyst-stage embryos. Fertility and Sterility 77, Shapiro BS, Richter KS, Harris DC, Daneshmand ST 2001 Dramatic declines in implantation and pregnancy rates in patients who undergo repeated cycles of in vitro fertilization with blastocyst transfer after one or more failed attempts. Fertility and Sterility 76, Shapiro BS, Harris DC, Richter KS 2000 Predictive value of 72-hour blastomere cell number on blastocyst development and success of subsequent transfer based on the degree of blastocyst development. Fertility and Sterility 73, Templeton A, Morris JK 1998 Reducing the risk of multiple births by transfer of two embryos after in vitro fertilization. New England Journal of Medicine 339, Utsunomiya T, Naitou T, Nagaki M 2002 A prospective trial of blastocyst culture and transfer. Human Reproduction 17, Van der Auwera I, Debrock S, Spiessens C et al A prospective randomized study: day 2 versus day 5 embryo transfer. Human Reproduction 17, Wilson M, Hartke K, Kiehl M et al Integration of blastocyst transfer for all patients. Fertility and Sterility 77, Yoon HG, Yoon SH, Son WY et al Alternative embryo transfer on day 3 or day 5 for reducing the risk of multiple gestations. Journal of Assisted Reproduction and Genetics 18, Zech NH, Lejeune B, Puissant F et al Prospective evaluation of the optimal time for selecting a single embryo for transfer: day 3 versus day 5. Fertility and Sterility 88, Declaration: The authors report no financial or commercial conflicts of interest. Received 2 January 2008; refereed 21 February 2008; accepted 17 April