Article Contribution of embryo cryopreservation to elective single embryo transfer in IVF ICSI

Transcription

1 RBMOnline - Vol 13. No Reproductive BioMedicine Online; on web 11 July 2006 Article Contribution of embryo cryopreservation to elective single embryo transfer in IVF ICSI Dominique Le Lannou is Professor in the Biology and Medicine of Reproduction at the Rennes University, France. He created CECOS (Centre d Étude et de Conservation des Oeufs et du Sperme) in Rennes Hospital in 1976 where he has worked on developing various assisted reproductive technologies: artificial insemination, IVF, ICSI, with a special interest in cryopreservation. Dr Dominique Le Lannou Dominique Le Lannou 1,3, Jean-François Griveau 1, Marie-Christine Laurent 2, Annie Gueho 1, Elisabeth Veron 1, Karine Morcel 2 1 Unité de Biologie de la Reproduction-CECOS, CHR Hotel-Dieu, Rennes, France; 2 Département de Gynécologie- Obstétrique, CHR Hôpital Sud, Rennes, France 3 Correspondence: dominique.lelannou@chu-rennes.fr Abstract Single embryo transfer is the best way to reduce the risk of multiple pregnancy in IVF intracytoplasmic sperm injection (ICSI). Between June 2002 and December 2004, all patients (first cycle, female age <38 years) were offered the choice between having one (SET) or two (DET) embryos transferred. Among 493 couples, 428 had at least two good quality embryos, and among them, 32% opted for SET. The SET and DET populations were not comparable (patients in the SET group were younger and had more oocytes retrieved), and therefore a paired, case control analysis was performed involving 130 SET couples and 130 DET couples, matched according to the female partners ages and the numbers of embryos available. All of the SET patients, and 82% of the DET group, had at least one embryo cryopreserved, (3.9 versus 2.8 embryos). The option of SET was continued for the frozen thawed embryo transfers. The pregnancy rate following embryo transfer was significantly lower after SET compared with DET for both fresh (27.6 versus 36.9%; P < 0.05) and frozen thawed (14.4 versus 23.5%) embryos. However, the cumulative live birth rates following the transfer of fresh and frozen embryos were identical between the two groups (43 versus 45%), with a high prevalence of twins following DET (34 versus 0%). Keywords: embryo cryopreservation, IVF, single embryo transfer 368 Introduction The prevalence of multiple pregnancies remains a major concern in the practice of IVF intracytoplasmic sperm injection (ICSI). Multiple pregnancy increases the risk to the child, in terms of prematurity, perinatal mortality and morbidity, and has important psychosocial and financial consequences for both the children and their parents (ESHRE Capri Workshop, 2000; De Sutter et al., 2002; Gerris et al., 2004). Northern European and Scandinavian countries such as Belgium, Finland and Sweden, recognized the seriousness of this problem several years ago and have pursued the strategy of single embryo transfer (SET), the validity of which has been demonstrated in numerous studies (Gerris et al., 1999; Vilska et al., 1999; Tiitinen et al., 2001). Recent reports on national results since the introduction of the SET policy have not revealed a reduction in the pregnancy rates, but the prevalence of twin pregnancies has fallen (Gordts et al., 2005; Saldeen and Sundstrom, 2005). The problem of multiple pregnancies has also been taken into consideration in France, and the national results showed a progressive decrease in the number of embryos being transferred, with the incidence of two-embryo transfers increasing relative to that of three-embryo transfers. However, this trend has stagnated over the past 2 years, and the latest registry report from 2003 shows that more than 50% of transfers still involved two embryos and only 4 5% of transfers were elective single embryo transfers (FIVNAT, 2003). As a consequence, the rate of multiple pregnancies has remained very high, with 25 30% of pregnancies being twin pregnancies. This situation is due to a fear of lower pregnancy rates with SET on the part of both the medical profession and the patients, and also to a growing banal acceptance of twin pregnancy, which is too often seen as

2 no different to a normal pregnancy. While published randomized studies have reported lower success rates in fresh cycles between single and double embryo transfer groups (Gerris et al., 1999; Martikainen et al., 2001; Thurin et al., 2004), several also demonstrated that the cumulative success rates were the same when the additional pregnancies resulting from the subsequent transfer of cryopreserved embryos were included in the analysis. However, the lower success rates seen with frozen embryos has led most groups to propose the transfer of two embryos post-thaw, even in the SET group. The aim of the present study was to investigate whether it might be possible, in France, to propose a policy of single fresh embryo transfer, in conjunction with the subsequent transfer of cryopreserved embryos, so as to evaluate the impact on the prevalence of multiple pregnancy in these patients. Materials and methods This was a prospective, non-randomized study. Study population Between June 2002 and December 2004, all couples receiving their first cycle of treatment in which the female partner was under 38 years of age were offered the choice of SET or double embryo transfer (DET). These patients were provided with clear information on their chances of pregnancy from one or two embryos, as well as the risks related to the outcome of twin pregnancy. They made their decision before oocyte retrieval. Ovarian stimulation and oocyte retrieval Patients were treated using a long down-regulation protocol for ovarian stimulation, except in patients with known risk of excessive response, where short protocol (Orgalutran; Organon, Paris, France) was used. Down-regulation was achieved using 1.5 mg of decapeptyl (Ipsen, Paris, France) intramuscularly on day 2 of the cycle, and suppression was confirmed 2 weeks later by ultrasonography and oestradiol concentrations. Ovarian stimulation was then initiated using 150 IU of recombinant FSH (Gonal F; Serono, Boulogne, France, or Puregon; Organon) administered subcutaneously. When at least three follicles were >18 mm, 10,000 IU of HCG was administered and transvaginal oocyte retrieval performed h later. Embryo transfer was performed on day 3 post-oocyte retrieval, and vaginal micronized progesterone (Utrogestran; Besins International, France) was used at 600 mg/day for luteal phase support. IVF ICSI procedures Cumulus oocyte complexes were isolated from the follicular aspirates and washed in Ferticult medium before being transferred individually into 25 μl droplets of Ferticult medium under mineral oil (Medicult, Copenhagen, Denmark) and incubated at 37 C in a humidified 5% CO 2 -in-air atmosphere. Motile spermatozoa were isolated from either fresh or cryopreserved semen using a two-step protocol. After processing on a 40/90% PureSperm discontinuous density gradient (Nidacon International, Göteborg, Sweden) the 90% fraction was washed in Ferticult medium (FertiPro; Beernem, Belgium), and the resulting sperm pellet was resuspended and transferred into the outer ring of a migration-sedimentation tube (Becton Dickinson, Franklin Lakes, CA, USA). After incubation at 37 C for a period of 1 4 h, an approximately 90% motile sperm suspension was harvested and used for standard IVF insemination or ICSI. For standard IVF, each oocyte was incubated with approximately 10,000 motile spermatozoa 3 5 h after oocyte retrieval and incubated overnight. Fertilization checks were performed after approximately 20 h incubation and all zygotes showing normal fertilization (i.e. two clearly visible pronuclei) were then transferred into one fresh drop of G1 medium (VitroLife, Göteborg, Sweden) and cultured for a further 48 h in a humidified atmosphere of 5% CO 2 /5% O 2 /90% N 2. For ICSI, after microinjection the oocytes were transferred separately into one drop of G1 medium, in which they were cultured for 72 h. Oocytes were examined for the presence of two pronuclei 20 h post-icsi. Embryo quality assessment Cleavage was assessed, and the embryos graded, on the morning of day 3, using the classification criteria of Scott et al. (1991): grade 1: blastomeres of equal size, no cytoplasmic fragments; grade 2: blastomeres of equal size, minor (<20%) cytoplasmic fragments; grade 3: blastomeres of distinctly unequal size, no cytoplasmic fragments; and grade 4: blastomeres of equal or unequal size, major (>20%) cytoplasmic fragments. Embryos were selected for transfer or cryopreservation if they had 6 8 regular-sized blastomeres and <20% fragmentation (i.e. grade 1 or 2). Embryo freezing and thawing Embryos were cryopreserved on day 3 using Embryo Freezing pack (Embryofreezing TM ; Medicult). In the SET group, all embryos were frozen singly in straws, whereas in the DET group up to two embryos were placed in each straw. The procedures for freezing and thawing were according to the manufacturer s instructions. Briefly embryos were placed in medium from vial 1 (phosphatebuffered saline, PBS) for 5 min at room temperature. They were then placed in medium from vial 2 containing 1.5 mol/ l propanediol for 10 min and then in medium from vial 3 containing 1.5 mol/l propanediol and 0.1 mol/l sucrose for 15 min. Medium from vial 3 was used to load embryos in 0.25 ml straws (Cryobiosystem; IMV Technologies, L Aigle, France) and embryos were frozen in a Minicool 40 (Air Liquide, Paris, France). Embryos were cooled from room temperature at a rate of 2 C/min to 7 C, and following a 5-min hold, they were manually seeded. Embryos were cooled at 0.3 C/min to 30 C and at 25 C/min to 140 C, and finally transferred and stored into liquid nitrogen at 196 C. Embryos were thawed using Embryo Thawing Pack (Embryothawing ΤΜ ; Medicult). Straws were removed from liquid nitrogen and kept at room temperature for 30 s then 369

3 370 placed into vial 1 containing 1 mol/l propanediol and 0.2 mol/l sucrose in PBS for 5 min. Embryos were then placed into vial 2 containing 0.5 mol/l propanediol and 0.2 mol/l sucrose for 5 min and then placed into vial 3 containing 0.2 mol/l sucrose for 5 min. Embryos were rinsed for 5 min in PBS before being placed in G1 culture medium (Vitrolife) until the transfer. Embryos were thawed on the scheduled day of transfer; only embryos with at least 50% intact blastomeres were transferred. Cryopreserved embryos were transferred in cycles controlled using oestradiol valerate (6 mg/day) and intravaginal micronized progesterone (600 mg/day). In the event of pregnancy, the hormone treatment was continued: 6 mg oestradiol valerate per day for 60 days, and 800 mg of progesterone for 90 days after embryo transfer. Outcomes Serum β-hcg concentrations were measured 12 and 14 days after embryo transfer. If positive (>20 IU), clinical pregnancy was confirmed by ultrasonography 30 days post-embryo transfer. Pregnancy results were expressed in terms of live births. Implantation rates were calculated only on clinical pregnancies that had been confirmed by ultrasound, regardless of whether they were ongoing or resulted in an early miscarriage or were ectopic. Cumulative live births were total live births resulting from fresh and frozen thawed embryo transfers per oocyte retrieval. Statistical analyses The chi-squared test was used to compare the outcomes between groups. Differences of P < 0.05 were considered as significant. Results Total patient population Of the 493 couples who met the study s clinical selection criteria (i.e. female partners under 38, first treatment cycle), 428 (87%) had at least two good quality embryos and were eligible for inclusion in the study. After the information session, 32% of these couples chose to have a single embryo transferred ( tset = prevalence of single embryo transfer among the total population), and the other 68% chose to have two embryos transferred ( tdet = prevalence of double embryo transfer among the total population). Those who chose tset were slightly younger (30.1 ± 3.7 versus 31.8 ± 3.2 years, P < 0.05) and had higher numbers of oocytes retrieved and embryos produced (9.4 ± 4.7 versus 7.9 ± 4.1 oocytes retrieved, P < 0.05; also see Table 1). While it was, in general, younger women with better prognoses who chose SET, the couples decisions evolved with the centre s experience: from 20% SET in the first year of the study to 45% in the third year. The pregnancy rates in the fresh transfer cycles were equivalent between the two groups. However, the cumulative pregnancy rate (i.e. fresh + subsequent frozen transfer cycles) was higher in the tset group, and the multiple pregnancy rate was higher in the tdet group (30 versus 0%). In the tset group, 39 pregnancies were obtained following the transfer of a single fresh embryo. There were two early miscarriages, resulting in the live birth of 37 singletons. The subsequent transfer of cryopreserved embryos generated a further 32 pregnancies, which included six early miscarriages, one ectopic pregnancy and one death in utero, and hence an additional 24 live births of singletons. The cumulative pregnancy rate in the tset group was therefore 51% of couples, with a 44% live birth rate. In the tdet group, 89 pregnancies were obtained following the transfer of two fresh embryos in each treatment cycle. There were 11 early miscarriages as well as two ectopic pregnancies and two therapeutic abortions, resulting in 74 deliveries, which included 51 singletons, 22 pairs of twins and one set of triplets. A further 34 pregnancies were obtained following the subsequent transfer of cryopreserved embryos in this group, which, after eight early miscarriages, resulted in 26 deliveries comprising 18 singletons and eight pairs of twins. The cumulative pregnancy rate in the tdet group was therefore 42% of couples, with a 34% live birth rate including 30% multiples. Paired case control study Since the patient populations who opted for SET or DET were not comparable, a further analysis was performed by matching each SET case with a DET case, i.e. a case control pairing. The selection was performed using a computer to search for a DET case that matched each SET case using the following selection criteria (in decreasing order of application): (i) age of the female partner, within ±1 year if no exact match was available; (ii) number of embryos obtained, within ±2 embryos if no exact match was available; and (iii) the date of oocyte retrieval, matched as closely as possible. Of the 138 SET couples, eight had to be excluded because the female partners were aged between 21 and 23 years and had not been able to choose the number of embryos that were transferred (given their young age, a single embryo transfer had been imposed on clinical grounds). These eight couples had achieved three pregnancies following the transfer of single fresh embryos and a further three pregnancies following the subsequent transfer of cryopreserved embryos, all of which had resulted in live singleton births. There were no patients aged between 21 and 23 years in the DET group. These two paired groups of patients (pset and pdet respectively) were identical in terms of female partner s age and the numbers of oocytes retrieved and embryos produced, and that 76 and 73% of the embryos were deemed to be utilizable (i.e. were graded as grade 1 or grade 2, and hence were suitable for transfer or freezing) (Table 2). Obviously more embryos were cryopreserved in the pset group than the pdet group (averages of 3.9 embryos/case in the pset group compared with 2.8 in the pdet group, P < 0.001). The ICSI/IVF ratio was identical in the two groups (76 and 70%). There were no statistically significant differences in embryo quality between IVF and ICSI (data not shown). The fresh embryo transfer results for the paired groups are shown in Table 3. While the implantation rates were equivalent between the two groups (27.6 compared with 23.8%), the

4 pregnancy rate was lower in the pset group (27.6 versus 36.9%, P < 0.05) with a high prevalence of multiple pregnancies arising from the transfer of two embryos (pdet group = 37%, versus zero in the pset group). There was a non-significant trend of an increased miscarriage rate in the pdet group (14 versus 5%). The transfer of cryopreserved embryos was offered to those patients who either did not become pregnant in the fresh transfer cycle or had a miscarriage: this was possible for 100% of those from the pset group and 82% from the pdet group (see Table 4). The criterion for embryo cryosurvival and transfer was >50% intact blastomeres post-thaw, and this included over 80% of the thawed embryos. In the pset group, 91 of the 96 patients who were not pregnant had at least one first thawed single embryo transfer, with 13 pregnancies and 10 deliveries of singletons. The live birth in the pset group after the first (fresh) and the second (frozen) single embryo transfer (i.e ) was = 33%, almost identical to that of the pdet group after the first fresh double embryo transfer (i.e ), which was 41 = 31.5%. Although all patients in the pset group were advised to have a single thawed embryo transferred, some patients did not accept this, in particular after failure of two or three thawed embryo transfers, and so a single thawed embryo was transferred in 177 cycles, and two embryos were transferred in 33 cycles, resulting in 29 pregnancies (representing a cumulative pregnancy rate of 30% per patient treated) and 22 singleton deliveries. Patients in the pdet group elected to have two embryos transferred in their frozen thawed cycles if possible, although only one embryo was available for transfer post-thaw in 24/65 cycles, and four of them failed to have at least one frozen embryo transfer due to lack of viable embryos after thawing. Among the resulting 23 pregnancies (i.e. 32% per patient treated), there were 17 deliveries, comprising 12 singletons and five pairs of twins. The pset patients had a higher number of frozen thawed embryo transfers compared with the pdet group (2.19 versus 1.22 transfer/patient, P < 0.01). Although the pregnancy rate per transfer with cryopreserved embryos was lower when one embryo was transferred compared with two (14.4 versus 23.5%, P < 0.05), overall, the total number of pregnancies obtained was higher in the pset group (n = 29) than in the pdet group (n = 23). In the pset group, the transfer of cryopreserved embryos with 100% intact blastomeres post-thaw resulted in an implantation rate of 18.5% per embryo transferred, compared with only 10% if 50 90% of the blastomeres were intact (P < 0.05; data not shown). The cumulative results for the transfer of fresh and frozen embryos revealed no significant difference in the live birth rate between the pset and pdet groups (43 and 45%; see Table 5). Higher birth rates were obtained using fresh embryos in the pdet group (32%) than in the pset group (26%), although not with frozen thawed embryos (13 and 17% respectively; P < 0.05). There was a 34% twin pregnancy rate in the pdet group. Among the patients who have delivered, 217 embryos remain in cryostorage from the pset group, and 177 embryos remain for the pdet group of patients. These embryos should permit between 20 and 25 further pregnancies for each group. Table 1. Summary results for total patient population. tset tdet P-value No. couples Female age, years (range) (21 37) 31.8 (24 37) <0.05 Oocytes retrieved (mean ± SD) 9.4 ± ± 4.1 <0.05 Embryos obtained (grade 1 or 2) a <0.05 Embryos frozen a <0.01 Live births from fresh embryos (% per couple) 37 (27) 74 (26) NS Live births from frozen thawed embryos (% per couple) 24 (17) 26 (9) Cumulative live births (% per couple) 61 (44) 100 (34) <0.05 Singletons Twins 30 <0.01 Triplets 1 Evolution of patient choice (%) a Mean values. tset = prevalence of single embryo transfer among the total population; tdet = prevalence of double embryo transfer among the total population; NS = not statistically significant. 371

5 Table 2. Characteristics of case control study population. pset pdet P-value No. couples NS Female age (mean ± SD) 30.5 ± ± 3.4 NS Mature oocytes (n) NS Embryos obtained (mean per couple) 843 (6.5) 855 (6.6) NS Fertilization rate (%) NS No. grade 1 or 2 embryos (mean per couple) 643 (4.9) 623 (4.8) NS No. frozen embryos (mean per couple) 513 (3.9) 363 (2.8) P < pset = paired-single embryo transfer; pdet = paired-double embryo transfer; NS = not statistically significant. Table 3. Case control analysis: results using fresh embryos. pset pdet P-value No. transfers Pregnancies (% of transfers) 36 (27.6) 48 (36.9) P < 0.05 Implantation rate % NS Live birth rate (% of transfers) 34 (26.1) 41 (31.5) NS Singletons Twins 0 15 P < pset = paired-single embryo transfer; pdet = paired-double embryo transfer; NS = not statistically significant. Table 4. Case control analysis: results using frozen thawed embryos. pset pdet Total No. couples not pregnant No. couples with embryos frozen (%) 96 (100) 73 (82) No. frozen embryos No. thawed embryos No. transferred embryos Transfers of one embryo Pregnancies (% of transfers) 24 (13.5) 5 (20.8) 29 (14.4) Transfers of two embryos Pregnancies (% of transfers) 5 (15.2) 18 (27.7) 23(23.5) Live births Singletons Twins 0 5 pset = paired-single embryo transfer; pdet = paired-double embryo transfer. 372

6 Table 5. Case control analysis: cumulative live birth rates. pset pdet P-value No. couples Live births from fresh embryos P < 0.05 Live births from frozen thawed embryos Cumulative live birth rate (%) 56 (43) 58 (45) NS Singletons Twins (%) 0 (0) 20 (34) P < 0.01 No. frozen embryos remaining pset = paired-single embryo transfer; pdet = paired-double embryo transfer; NS = not statistically significant. Discussion Couples presenting for IVF typically request the transfer of two embryos, as it is considered the best means of increasing the chance of a pregnancy. The risk of multiple pregnancy, prematurity, and neonatal mortality and morbidity is minimized by the majority of patients (Porter and Bhattacharya, 2005) and is generally accepted, even wished. Because of this, the concept of being able to transfer just a single embryo places couples in a novel, unexpected situation, and necessitates an inherently difficult decision (Blennborn et al., 2005). They are faced with two fundamentally opposing principles: on the one hand wanting to increase the chance of a pregnancy, while on the other hand wanting to reduce the risk of multiple pregnancy. The information provided by the medical profession will clearly influence the couple s decision. The medical profession, for good reasons, wants to maintain the highest success rates, and the fear of seeing their pregnancy rates fall remains an obstacle to the routine application of SET. This is particularly true in certain countries such as France and the USA, whereas in other countries such as Sweden, Finland and Belgium, the paediatric risks associated with multiple pregnancy have, for several years now, led clinicians to evaluate the effects of transferring fewer embryos. It has been observed that the results achieved in the routine practice of SET in a centre progressively reassure the medical staff, and so the information they provide to patients becomes more positive. This matches current experience, with an increasing proportion of patients choosing SET throughout a 3-year period, from 20% in 2002 to 45% in 2004 and, in 2005, almost 70% of couples opted for SET. This has been reported in other studies which found that when patients are given the choice between one or two embryos, 40 60% choose to have one embryo transferred (De Neubourg et al., 2002; Blennborn et al., 2005), and these patients are, in general, the youngest ones. The first publication on the use of elective SET, from Finland (Vilska et al., 1999), has been followed by numerous other studies. In observational studies, SET and DET have given identical pregnancy rates, although in these studies the two groups of patients were not comparable: the younger women, or those with better quality embryos, received SET, while those with a poorer prognosis preferred DET (Gerris et al., 2004; Bergh, 2005). This was also seen in the authors own general patient population (Table 1), where the live birth rates were 27 and 26% in the SET and DET groups respectively, confirming that in good prognosis patients the practice of SET provides a drastic reduction in the twin pregnancy rate while not impacting the pregnancy rate. In the paired, case control analysis (Tables 2 and 3), a reduction in the live birth rate was seen in the SET group compared with the DET group (26.1 compared with 31.5%), findings that are identical to those in the four published randomized studies that included comparable groups of patients (Gerris et al., 1999; Martikainen et al., 2001; Gardner r et al., 2004; Thurin et al., 2004). Combining these results reveals that an acceptable average delivery rate of 28% can be achieved using SET, although this is significantly lower than for the practice of DET (42%). Moreover, in all cases the use of SET achieved a drastic reduction in the multiple pregnancy rate (Gerris et al., 2004; Bergh, 2005). It can be concluded that the transfer of a single fresh embryo does reduce the pregnancy rate compared with transferring two embryos. In the present study, the intention was not to be too selective, and hence all embryos of grade 1 or grade 2 were included, and the selection of patients between SET or DET was made before oocyte retrieval. The majority (87%) of couples where the female partner was under 38 years of age having their first IVF ICSI cycle had at least two good embryos and were eligible for inclusion, and they achieved a delivery rate of 26% after SET. In comparison, in the majority of published studies the proportion of SET-eligible couples was much lower: 22% in Van Montfort et al. (2005a); 25% for Gerris et al. (2004); 34% for Thurin et al. (2004); 39% for Lukassen et al. (2005); and 46% for Martikainen et al. (2001). It is therefore evident that the stricter the selection criteria for SET, the higher the pregnancy rate for SET will be, but at the expense of excluding increasing numbers of couples who could have benefited from this modality of treatment. In a study comparing non-selected and good prognosis patient populations, Van Montfort et al. (2005b) reported that the pregnancy rate obtained after SET 373

7 374 was lower in the non-selected population (31.5 versus 21.4%, P < 0.05). To reduce the risk of occurrence of twin pregnancies, the populations of patients who are at risk need to be defined. However, while this risk is very high when two embryos are transferred in good prognosis patients (e.g. 33% for Thurin et al., 2004; or 40.6% for Gerris et al., 2002; and 60% for Gardner et al., 2004), it is lower in patients with a poorer prognosis, but still not negligible (e.g. 16.9% for Gerris et al., 2002 and 20% for Van Montfort t et al., 2005b). Furthermore, if the primary goal is to reduce the prevalence of twin pregnancies as much as possible, the practice of SET must be extended to a large population of patients, accepting the principle that there will be a decrease in the pregnancy rate following fresh embryo transfer. Obviously, any policy limiting the number of fresh embryos transferred must also consider the outcome from cryopreservation of the supplementary embryos. In the present study, all couples who had SET had at least one embryo cryopreserved (by definition), and in the DET group 82% of the non-pregnant patients also had at least one embryo frozen. A total of 22 additional deliveries were obtained in the pset group, and 17 in the pdet group. Combining the results from the fresh and frozen embryo transfers increased the live birth rate in the pset group from 26 to 43%, and in the pdet group from 31.5 to 45%. In the SET group, the reduction in the number of births obtained after the transfer of a single embryo was fully compensated by the number of births resulting from frozen embryos transferred. The cumulative rates were therefore not different between pset and pdet, while there was a twin pregnancy rate of 34% in the pdet group, compared with zero in the pset patients. Larger studies of sufficient power should be performed to confirm these results. Only one multicentre randomized study has compared the transfer of two fresh embryos with the transfer of a single fresh embryo followed by a frozen thawed embryo (Thurin et al., 2004), and there was no difference in the live birth rates between the two groups (43.3 compared with 38.8%), although there was a 33% twin pregnancy rate in the DET group. In the present study, the majority of the SET patients pursued the transfer of a single frozen thawed embryo. Certainly, the pregnancy rates per transfer were significantly lower after the transfer of one versus two cryopreserved embryos (14.4 compared with 23.5%), but the number of frozen thawed embryo transfers after SET was much greater than after DET (2.1 versus 1.0 transfers), and the cumulative pregnancy rates were equivalent between the two groups (30 versus 32% per non-pregnant patient who had at least one embryo cryopreserved). Moreover, over 25% of the pregnancies following the transfer of two frozen thawed embryos were twin pregnancies. The goal of transferring only one frozen thawed embryo is therefore possible, without reducing the chances of pregnancy. Few studies have taken the outcome from the supplementary cryopreserved embryos into account, and most of those were only observational studies on small numbers of patients, and in the majority of cases two frozen thawed embryos were transferred at a time. Overall, the implantation rate for frozen thawed embryos is always significantly lower than that of fresh embryos, and hence most workers usually transfer two thawed embryos, even in cases where the women had a fresh SET. Notwithstanding this, Martikainen et al. (2001) reported that the pregnancy rate was increased from 32 to 47% in their SET group (n = 74) and from 47 to 58% in their DET group by using cryopreserved embryos; but, in their SET group, only 26/84 cycles (31%) were performed using just a single embryo. In the study by Tiitinen et al. (2001), the delivery rate for their SET group was increased from 26.8 to 52.8%, but again, 93/129 (72%) thawed embryo cycles involved the transfer of two embryos, and there was a twin pregnancy rate of 7.6%. The present study has shown that it is quite practicable to pursue a policy of SET in combination with the outcome from cryopreserved embryos. A recent retrospective analysis of 1647 cycles of frozen thawed embryos also concluded that SET could be proposed as a means of reducing the multiple pregnancy rate (Hyden-Granskog et al., 2005). Clinically, it was clear that the cumulative success rates are identical in the SET and DET frozen embryo groups. However, it was also apparent that in those patients who had more embryos cryopreserved, the repetition of several cycles of SET with relatively poor outcome (14%) can result in a certain disillusionment, which leads to its abandonment. Among the patient population, 10% (representing 15% of cryopreservation cycles) abandoned SET in favour of DET after two or three cycles of a single frozen thawed embryo transfer. To reduce this number of cycles, one could apply selection criteria to the cryopreserved embryos. Embryos that have all their blastomeres intact after thawing have a much higher implantation rate than partially lysed embryos (18.5 versus 10.5%, P < 0.05). Other selection criteria were proposed, as resumption of mitosis after thawing (Gabrielsen, 2006). In all the cases if the selection of thawed embryos can reduce the number of frozen thawed transfer cycles, and thus can increase the pregnancy rate per transfer, on the other hand, these not selected embryos exist, and those pregnancies that might be obtained by transferring them represent a real value to the patients, especially when the number of IVF cycles is restricted (e.g. four cycles in France). Clearly some difficult decisions need to be made in regard to these embryos and, after giving them all the available information, the decision to transfer or not must rest with the couple. Another approach to reducing the number of cycles in a SET protocol is to extend the embryo culture to day 5 and to transfer a single blastocyst. Papanikolaou (2006) has shown that the rate of delivery was higher in the group undergoing transfer of a single fresh blastocyst-stage embryo than in the group with a transfer of a single fresh cleavage-stage embryo (32.0 versus 21.6%). However, in that study significantly fewer embryos were cryopreserved in the blastocyst-stage group than in the cleavage-stage group (2.2 versus 4.2, P < 0.001). Unfortunately, there are no studies that have compared the cumulative live birth results for the transfer of fresh and frozen embryos at day 2 3 or at day 5 6. The efficacy of blastocyst culture and cryopreservation is still controversial (Anderson, 2005). In conclusion, while twin pregnancies are not totally adverse outcomes in IVF ICSI, it is possible to reduce their occurrence considerably by pursuing a policy of single embryo transfer that can be applied to a large majority of patients, but only if it is accompanied by widespread embryo cryopreservation.

8 Acknowledgements We thank David and Sharon Mortimer, Oozoa Biomedical Inc., Canada, for their friendly assistance with the translation of the manuscript. References Anderson A 2005 Reduction of high order multiple in frozen embryos transfers. Reproductive BioMedicine Online 10, Bergh C 2005 Single embryo transfer: a mini review. Human Reproduction 20, Blennborn M, Nilsson S, Hillervik C, Hellberg D 2005 The couple s decision-making in IVF: one or two embryos at transfer? Human Reproduction 20, De Neubourg D, Mangelschots K, Van Royen E et al Impact of patients choice for single embryo transfer of a top quality embryo versus double embryo transfer in the first IVF/ICSI cycle. Human Reproduction 17, De Sutter P, Gerris J, Dhont M 2002 A health-economic decisionanalytic model comparing double with single embryo transfer in IVF/ICSI. Human Reproduction 17, ESHRE Capri Workshop 2000 Multiple gestation pregnancy. Human Reproduction 15, FIVNAT (French In Vitro National Database) 2003 Dossier FIVNAT Bilan de l année Gabrielson A 2006 Parameters predicting the implantation rate of thawed IVF/ICSI: a retrospective study. Reproductive BioMedicine Online 12, Gardner DK, Surrey E, Minjarez D et al Single blastocyst transfer: a prospective randomized trial. Fertility and Sterility 81, Gerris J, De Sutter P, De Neubourg D et al A real-life prospective health economic study of elective single embryo transfer versus two-embryo transfer in first IVF/ICSI cycles. Human Reproduction 19, Gerris J, De Neubourg D, Mangelschots K et al Elective single day 3 embryo transfer halves the twinning rate without decrease in the ongoing pregnancy rate of an IVF/ICSI programme. Human Reproduction 17, Gerris J, De Neubourg D, Mangelschots K et al Prevention of twin pregnancy after in-vitro fertilization or intracytoplasmic sperm injection based on strict embryo criteria: a prospective randomized clinical trial. Human Reproduction 14, Gordts S, Campo R, Puttemans P et al Belgian legislation and the effect of elective single embryo transfer on IVF outcome. Reproductive BioMedicine Online 10, Hyden-Granskog C, Unkila-Kallio L, Halttunen M, Tiitinen A 2005 Single embryo transfer is an option in frozen embryo transfer. Human Reproduction 20, Lukassen HG, Braat DD, Wetzels AM et al Two cycles with single embryo transfer versus one cycle with double embryo transfer: a randomized controlled trial. Human Reproduction 20, Martikainen H, Tiitinen A, Tomas C et al Finnish ET study group. One versus two embryo transfer after IVF and ICSI: a randomized study. Human Reproduction 16, Papanikolaou Z, Camus M, Kolibianakis M et al In vitro fertilization with single blastocyst-stage versus single cleavagestage embryos. New England Journal of Medicine 354, Porter M, Bhattacharya S 2005 Investigation of staff and patients opinions of a proposed trial of elective single embryo transfer. Human Reproduction 20, Saldeen P, Sundstrom P 2005 Would legislation imposing single embryo transfer be feasible way to reduce the rate of multiple pregnancies after IVF treatment. Human Reproduction 20, 4 8. Thurin A, Hausken J, Hillensjo T et al Elective single-embryo transfer versus double-embryo transfer in in vitro fertilization. New England Journal of Medicine 351, Titinen A, Halttunen M, Harkki P 2001 Elective single embryo transfer: the value of cryopreservation. Human Reproduction 16, van Montfoort AP, Dumoulin JC, Land JA et al. 2005a Elective single embryo transfer (eset) policy in the first three IVF/ICSI treatment cycles. Human Reproduction, 20, Van Montfort AP, Fiddelers AA, Janssen JM et al. 2005b In unselected patients, elective single embryo transfer prevent all multiple, but results in significantly lower pregnancy rates compared with double embryo transfer: a randomized controlled trial. Human Reproduction, advance access, published November 30, Vilska S, Tiitinen A, Hyden-Granskog C, Hovatta O 1999 Elective transfer of one embryo results in an acceptable pregnancy rate and eliminates the risk of multiple birth. Human Reproduction 14, Received 30 March 2006; refereed 12 April 2006; accepted 23 May