The number of eight-cell embryos is a key determinant for selecting day 3 or day 5 transfer

FERTILITY AND STERILITY VOL. 73, NO. 3, MARCH 2000 Copyright 2000 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. The number of eight-cell embryos is a key determinant for selecting day 3 or day 5 transfer Catherine Racowsky, Ph.D., Katharine V. Jackson, B.S., Natalie A. Cekleniak, M.D., Janis H. Fox, M.D., Mark D. Hornstein, M.D., and Elizabeth S. Ginsburg, M.D. Brigham and Women s Hospital, Harvard Medical School, Boston, Massachusetts Objective: To select patients for day 3 vs. day 5 embryo transfer. Design: Retrospective analysis of assisted reproduction technology (ART) cycles comparing outcomes of day 3 and day 5 transfers. Setting: ART program of Brigham and Women s Hospital. Patient(s): Patients with day 3 or day 5 embryo transfers (n 221 and 141, respectively). Intervention(s): Cycles with eight or more zygotes were stratified by the number of eight-cell embryos available on day 3 (none, one or two, or three or more). Main Outcome Measure(s): Number of blastocysts, implantation rates, ongoing pregnancy rates, and number of fetal heart beats. Result(s): With no eight-cell embryos on day 3, 0% and 33% pregnancies resulted from day 5 vs. day 3 transfers. With one or two eight-cell embryos on day 3, ongoing and high order multiple rates were not different between day 3 and day 5 transfers. With three or more eight-cell embryos, day 5 transfer resulted in a decrease in multiple gestations but no difference in ongoing pregnancy rates compared with day 3 transfer. Conclusion(s): With no eight-cell embryos on day 3, a day 3 transfer is warranted. With one or two eight-cell embryos, any benefit of day 5 transfer appears to be equivocal. With three or more eight-cell embryos, day 5 transfer is recommended. (Fertil Steril 2000;73:558 64. 2000 by American Society for Reproductive Medicine.) Key Words: Number of eight-cell embryos, day 3 vs. day 5 transfer, incidence of multiple gestation, blastocyst Received July 7, 1999; revised and accepted October 6, 1999. Reprint requests: Catherine Racowsky, Ph.D., Department of Obstetrics, Gynecology and Reproductive Biology, Brigham and Women s Hospital, Harvard Medical School, 75 Francis Street, ASB 1 3, Room 082, Boston, Massachusetts 02115 (FAX: 617-732-5570; E-mail: cracowsky@bics.bwh.harvard.edu). 0015-0282/00/$20.00 PII S0015-0282(99)00565-8 To overcome low implantation rates from IVF cycles, IVF programs developed protocols designed to stimulate follicles with the goal of obtaining several embryos for embryo transfer. Pregnancy rates were enhanced with the transfer of multiple embryos. This was accomplished, however, at the cost of increasing the incidence of multiple gestations. As stimulation protocols and laboratory techniques have improved, implantation rates have risen, further exacerbating the problem of multiple gestations. One recent strategy to reduce multiple gestation rates has been the reintroduction of extended embryo culture to the blastocyst stage, as originally proposed by Edwards and Hollands (1) and attempted by Bolton et al. (2). This strategy is based on the observations that a high proportion of cleavage-stage embryos undergo arrest in culture (3) and that this failed development in vitro is frequently correlated with aneuploidy (4). Accordingly, prolonging the duration of culture to day 5 should allow chromosomally competent embryos to develop to the blastocyst stage, thereby increasing the likelihood of identifying euploid, developmentally competent embryos for transfer. As such, this approach should enable the transfer of fewer but higher quality embryos, thereby maintaining high pregnancy rates while controlling the incidence of high-order multiple gestations. Initial attempts at blastocyst transfer, however, did not demonstrate an increase in either implantation or pregnancy rate because of poor embryo survival in inadequate extended culture conditions (2). Higher implantation rates had to await the development of better culture conditions to allow successful extended culture and acceptable rates of blastocyst formation (5). 558

Indeed, with contemporary sequential culture media, clinical pregnancy rates as high as 60% have been reported, with a significant reduction in the incidence of high-order multiple pregnancies (5). Despite the considerable progress made with extended culture, questions arise as to whether day 5 transfer is appropriate for all patients and whether extending the duration of culture of some embryos for an additional 48 hours might compromise their implantation potential. In other words, are there some patients for whom the uterus is a better embryo incubator than that provided by our current technologies of extended culture in vitro? There are at least two reasons why these issues are particularly relevant. First, all recent reports of day 5 transfers relate to selected populations of patients who are at greatest risk of high-order multiple gestations (5); would day 5 pregnancy rates be similar in poor prognosis patients? Second, there are some patients for whom no blastocysts have formed on day 5, even after extended culture in the most advanced media currently available (6). Would these embryo cohorts be doomed to result in a failed conception even if the transfer had been performed on day 3? The purpose of the present retrospective analysis was to identify a clinical protocol for prospective selection of the optimum day of embryo transfer for a patient based on the morphologic appearance of her embryos on day 3. The hypothesis tested was that the number of eight-cell embryos on day 3 provides an independent marker for identifying patients who will benefit from transfer on day 5. The number of eight-cell embryos was chosen on the basis of data showing higher day 3 pregnancy rates when at least one eight-cell embryo is available (7), as well as a positive correlation between the number of eight-cell embryos on day 3 and blastocyst formation rate on day 5 (8). Therefore, this study evaluated the number of eight-cell embryos as the sole determinant among all patients for identifying the optimum day of embryo transfer. MATERIALS AND METHODS IVF Program Entry Criteria Entry criteria for women presenting for IVF were age 44 years, day 3 FSH level 15 miu/ml, and a normal uterine cavity on the basis of hysteroscopy or hysterosalpingography before initiation of treatment. Women 40 years of age underwent clomiphene challenge testing with 100 mg of clomiphene citrate daily on cycle days 5 to 9. A day 10 FSH level 15 miu/ml was also mandatory for program entry for this group. Study Cycles and Criteria for Acceptance to a Day 5 Embryo Transfer A retrospective analysis of 369 cycles with at least eight zygotes at the fertilization check was undertaken, after appropriate institutional review board approval for chart review was obtained. The inclusion criterion of a minimum of eight zygotes was based on a blastocyst formation rate of 30% 40%, thereby maximizing the probability of having at least two blastocysts available for transfer on day 5. Of these 369 cycles, 221 were historical controls performed from January 1, 1998 to May 1, 1998, resulting in day 3 embryo transfers. The remaining 148 cycles, destined for day 5 transfer, were performed between May 5, 1998 (when day 5 transfer was introduced into our program) and December 29, 1998. Day 3 transfers performed during this latter time period were excluded from the analysis to avoid possible patient selection bias. Of the 148 cycles destined for day 5 transfer, 141 had at least morulae available for transfer on day 5, while 7 had all embryos arrested in development at or before the eight-cell stage. In 3 of these cleavage-arrested cycles, the patients still elected to have embryo transfers, and none of them conceived. Data from the 7 cycles involving arrested embryos were excluded from subsequent analyses. Patients were offered a day 5 transfer if they had eight or more zygotes on day 1 (day 0 day of insemination) and were extensively counseled concerning the potential benefits and risks of extending culture of their embryos for an additional 48 hours to day 5. Stimulation Protocols Controlled ovarian hyperstimulation was performed with luteal downregulation with 1.0 to 0.5 mg of leuprolide acetate (Lupron; TAP Pharmaceuticals, Deerfield, IL). The standard IVF protocol was as follows: Leuprolide acetate (1.0 or 0.5 mg depending on prior gonadotropin response), was begun either a week after documentation of urinary LH surge or the day after a midluteal progesterone determination and was continued until at least day 2 of menses. Baseline ultrasonography and blood testing were then performed to document appropriate downregulation. If these criteria were met, gonadotropin therapy was begun. The most common gonadotropin used was either purified FSH (Fertinex; Serono Laboratories, Norwell, MA) or recombinant FSH (Follistim; Organon, West Orange, NJ, or Gonal-F; Serono Laboratories, Norwell, MA). Standard stimulation was generally divided daily dosing of a total of 3 4 ampules/d for women 30 40 years of age. Women 40 years or with a history of low gonadotropin response were given a general daily maximum of 8 ampules, administered in divided doses. Women 30 years, or with a history of high gonadotropin response, were given a single daily injection of 2 3 ampules. Follicle growth monitoring, achieved with the use of ultrasonography and measurement of serum estradiol, was begun on stimulation day 6 and was then performed every 1 3 days, as indicated. A dose of 10,000 IU of hcg (Profasi; Serono) was administered intramuscularly when two follicles reached a maximal diameter 20 mm (mean, 16.5 mm) with an estradiol level 500 pg/ml. Transvaginal oocyte retrieval was performed 36 hours after hcg administration. Luteal progesterone supplementa- FERTILITY & STERILITY 559

tion was initiated the day after oocyte retrieval and was achieved by one of three regimens: [1] daily intramuscular progesterone (50 mg), [2] daily vaginal gel (8% progesterone [Crinone; Wyeth-Ayerst, Philadelphia, PA]); or [3] twicedaily vaginal progesterone suppositories (50 to 100 mg). Embryo transfer was performed 3 or 5 days after retrieval with a Wallace catheter (Marlow/Cooper Surgical, Shelton, CT). For difficult transfers, a Marrs No. 4 or Marrs No. 5 embryo transfer catheter (Cook Ob/Gyn, Indianapolis, IN) was occasionally used. The incidence of such cases was 9.7% and was not significantly different across all patient groups. Assays Estradiol and progesterone were assayed with the Immuno-1 method (Bayer Corp., Tarrytown, NY). For estradiol, the intraassay and interassay coefficients of variation were, respectively, 7.6% and 13.0% ( 50 pg/ml), 2.3% and 3.4% (600 pg/ml), and 2.1% and 4.3% (1,200 pg/ml). The lower limit of the progesterone assay was 0.1 ng/ml, and the intraassay and interassay coefficients of variation were, respectively, 6.1% and 14% ( 1.0 ng/ml), 2.0% and 2.3% (5.0 ng/ml), and 2.3 % and 5.6% (18.0 ng/ml). Oocyte Fertilization, Embryo Culture, and Transfer The basic culture system employed Scandinavian IVF Science (Gothenburg, Sweden) IVF-500 culture medium to support embryo growth from days 1 3 (day 1 being the day of the fertilization check), followed by S2 (Scandinavian IVF Science) for embryo culture from days 3 5. S2 was marketed for use sequentially with S1 (Scandinavian IVF Science ) for culture from days 1 3. However, our early experience with S1 and S2 showed that lower quality blastocysts formed compared with IVF-500/S2 media. Accordingly, we elected to use the IVF-500/S2 sequential media as a standard system for culturing embryos in our day 5 transfer program. Two hundred seventy-two cycles of IVF treatment had oocytes (3 5 per dish) inseminated with 400,000 motile spermatozoa in 1.0 ml IVF-500 culture medium overlaid with 1.0 ml Ov-Oil-150 (Scandinavian IVF Science). Sperm inseminate preparations were performed by using a single density 90% Isolate column, or a mini Isolate (Irvine Scientific, Santa Ana, CA) method, depending upon initial semen parameters. Ninetyseven ICSI cycles had mature metaphase II oocytes injected with a single spermatozoon (9) because of either insufficient sperm to perform standard IVF or unexpected failure to fertilize in a previous IVF cycle. Immediately after ICSI, oocytes were cultured in 25- L microdrops of IVF-500, overlain with Ov-Oil-150 in Falcon 1007 culture dishes (Becton Dickinson Labware, Franklin Lakes, NJ). All cultures were maintained at 37 C in a humidified atmosphere of 5% CO 2 in air. Fertilization checks were performed 15 18 hours after insemination or ICSI, and oocytes exhibiting two pronuclei were rinsed through three 50- L drops of IVF-500 before being set up for culture. Regardless of day of transfer, embryos were cultured individually in 25- L drops of medium (IVF-500: days 1 3; S2: days 3 5) under 8 ml of Ov-Oil-150 in Falcon 1007 culture dishes. All embryos were evaluated for cell number, degree of fragmentation, and extent of symmetry on the morning of day 3 (mean time, 66.6 hours postinsemination). Embryos cultured to day 5 were rinsed sequentially through three 50- L drops of S2 before being placed in 25- L S2 culture drops. On the morning of day 5, embryos were graded as follows: grade 1, a cleavage-arrested embryo; grade 2, a cavitating cleavage-stage embryo, but with no evidence of blastomere compaction; grade 3, a morula, with clearly defined compaction evident; grade 4, an early blastocyst surrounded by a thick zona pellucida, with a distinct, but small, blastocoelic cavity and differentiating outer trophectoderm and inner cell mass cells; grade 5, an expanding blastocyst with thinning zona pellucida, enlarging blastocoelic cavity, and differentiated outer trophectoderm and inner-cell-mass cells; grade 6: a fully expanded blastocyst with thin zona pellucida, an accentric inner cell mass, and an enlarged blastocoelic cavity enclosed by a clearly defined, multicellular trophectoderm composed of polygonal cells. For day 3 transfers, embryos with the highest cell number, the lowest fragmentation score, and greatest symmetry in a given cohort were selected for transfer. For day 5 transfers, the most expanded blastocysts with the greatest number of cells in the inner cell mass and trophectoderm were selected. The number of embryos transferred was determined by the day of transfer, the number of embryos available, the patient s age, and the patient s previous clinical history. For day 3 transfers, three to four embryos were typically transferred, whereas two embryos were generally transferred for day 5 transfers. However, if the most developmentally advanced embryo available was a morula or if a patient had failed to conceive with at least three prior attempts, three or, in one case, four embryos were transferred on day 5. For the day 3 transfer group, high quality eight-cell embryos in excess of those transferred on day 3 were cryopreserved by using Freedman et al. s one-step procedure (10). If perfect six-cell embryos were present, these were also frozen. All remaining cleavage-stage embryos with at least six cells were transferred into S2 for extended culture. Any fully expanded blastocysts developing from this residual cohort of day 3 embryos were cryopreserved on day 5 by using Menezo et al. s slow cooling protocol with glycerol as the cryoprotectant (11). Likewise, for the day 5 transfer group, any fully expanded blastocysts that were not selected for a day 5 transfer were cryopreserved by using the procedure of Menezo et al. Clinical pregnancies were identified by the presence of a gestational sac on ultrasonongraphy performed 5 weeks after oocyte retrieval. The implantation rate was calculated as the number of fetal heartbeats present, divided by the number of embryos transferred, multiplied by 100. 560 Racowsky et al. Number of eight-cell embryos determines day of ET Vol. 73, No. 3, March 2000

TABLE 1 Patient and embryo characteristics for day 3 versus day 5 embryo transfers. No. of eight-cell embryos on day 3 Charactristic Day of embryo transfer 0 8 cells 1 or 2 8 cells 3 8 cells No. of patients 3 27 77 117 5 14 51 83 Age (y) 3 34.5 4.2 34.9 3.5 35.2 3.5 5 34.2 3.6 33.7 3.6 33.7 3.7 Attempt no. 3 1.9 1.2 1.8 1.0 1.6 0.9 5 1.7 1.0 2.2 1.5 1.6 0.9 No. of oocytes 3 17.0 7.8 17.5 6.7 20.8 6.4 5 16.4 5.3 18.6 6.9 22.2 8.9 No. of mature oocytes 3 15.4 7.2 15.9 6.4 18.8 6.0 5 14.4 3.7 16.5 5.6 20.3 7.8 No. of 2 PN embryos 3 9.9 3.6 10.8 4.0 13.4 4.3 5 9.5 3.9 11.4 3.5 14.4 5.4 No. of eight-cell embryos 3 0 1.6 0.5 6.0 2.9 5 0 1.6 0.5 5.2 2.4 Note: Values are expressed as means SD. No statistically significant difference was noted between the day 3 and day 5 transfer patients within any of the eight-cell embryo groups for any of the variables tested, except for age in the group with three or more eight-cell embryos. (P.006). Statistical Analysis Data were compared after stratification of patient cycles by transfer day (day 3 or day 5) and were further classified into three groups according to the number of eight-cell embryos present on day 3 (none, one or two, or three or more eight-cell embryos). This classification was undertaken without regard to fragmentation or symmetry scores of eight-cell embryos. Data were analyzed by using 2 and Fisher s exact tests, the Mann-Whitney U test, or analysis of variance with the Bonferroni correction for P value, as appropriate, A P value.05 was considered statistically significant. RESULTS Patient and embryo characteristics for each eight-cell group within each transfer day subset are shown in Table 1. No statistically significant differences were noted for any eight-cell group between day 3 and day 5 transfer patients for cycle attempt number, total number of oocytes or number of mature oocytes retrieved, number of 2PN zygotes, or number of eight-cell embryos available on day 3. Although patient age did not differ between day 3 and day 5 transfers for patients with either zero or one or two eight-cell embryos, patients with three or more eight-cell embryos who underwent day 5 transfer were significantly younger (P.006). Cleavage arrest was observed on day 5 in 3 of 14 (21.4%) cycles that had no eight-cell embryos on day 3; embryo transfer was cancelled for all 3 cycles. Although it was not significantly different from the percentage of cycles that exhibited cleavage arrest when one or two eight-cell embryos were available on day 3 (4 of 51 [7.8%]; P.192), this incidence was significantly greater than that for cycles having three or more eight-cell embryos (0 of 83 [0%]; P.002). Likewise, the percentage of cycles exhibiting cleavage arrest when one or two eight-cell embryos were available on day 3 (4 of 51 [7.8%]) was also significantly greater than that for cycles with three or more eight-cell embryos (0 of 83 [0%]; P.019). Three of the four patients with arrested embryos in this group selected to undergo embryo transfer, although none conceived. Overall, significantly fewer embryos were transferred on day 5 as compared with day 3, regardless of the eight-cell embryo group (P.0001 for all three groups; Table 2). When no eight-cell embryos were available on day 3, no pregnancies occurred when a day 5 transfer was performed. Indeed, day 3 embryo transfers resulted in significantly higher clinical and ongoing pregnancy rates compared with those achieved after transfer on day 5 (P.01 in both cases). However, implantation rates were not different between the two transfer days (7.9% vs. 0%). In cycles with one or two eight-cell embryos available on day 3, the day of transfer did not significantly affect the clinical or ongoing pregnancy rates or the implantation rate. In contrast, in cycles with more than three eight-cell embryos on day 3, day 5 transfers resulted in an increased implantation rate (35.9% vs. 24.2%; FERTILITY & STERILITY 561

TABLE 2 Cycle outcomes in day 3 versus day 5 transfers stratified by number of 8-cell embryos on day 3 post-retrieval. No. of eight-cell embryos on day 3 Variable Day of embryo transfer 0 8 cells 1 or 2 8 cells 3 8 cells No. of transfers 3 27 77 117 5 11 47 83 No. of embryos transferred 3 4.7 2.0 4.0 0.9 3.7 0.9 5 1.6 1.1* 2.2 0.8* 2.2 0.5* No. of embryos implanted/total transferred (%) 3 10/126 (7.9) 54/306 (17.6) 105/434 (24.2) 5 0 26/110 (23.6) 66/184 (35.9) No. of clinical pregnancies (% of ETs) 3 10 (37.0) 38 (49.4) 67 (57.3) 5 0 20 (42.6) 52 (62.7) No. of spontaneous abortions (% of clin pregs) 3 1 (10.0) 10 (26.3) 7 (10.4) 5 N/A 1 (5.0) 10 (19.2) No. of ongoing pregnancies (% of ETs) 3 9 (33.3) 28 (36.4) 60 (51.3) 5 0 19 (40.4) 42 (50.6) Note: No statistically significant differences were noted between day 3 and day 5 transfer patients for any of the groups of 8-cell embryos except where indicated. ET embryo transfer. * P.0001. P.003. P.01. P.003) but no difference in either the clinical or the ongoing pregnancy rate. The outcome of pregnancies according to the number of fetal heartbeats for each eight-cell embryo group is shown in Table 3. There was no statistical difference between day 3 and day 5 transfers in the incidence of singleton or twin pregnancies for any of the eight-cell embryo groups. However, significantly fewer triplet pregnancies resulted from day 5 transfers in the group with three or more eight-cell embryos (4.3% vs. 17.7%, P.034), despite the occurrence of two sets of monochorionic twins in association with singletons, each of which arose following the transfer of two blastocysts. The average number of blastocysts (grades 4, 5, and 6 combined) and fully expanded blastocysts (grade 6) that developed in the group with three or more eight-cell embryos were both significantly greater than those in the other two eight-cell embryo groups (Table 4). Consistent with these observations, 67.5% of cycles in the group with three or more eight-cell embryos had at least one expanded blastocyst transferred; this value was significantly higher than that in the group with no eight-cell embryos (18.2%; P.003) or with one or two eight-cell embryos (48.9%; P.042). DISCUSSION The present study provides the first analysis to establish an independent marker for identifying patients who will benefit from embryo transfer on day 5: the number of eightcell embryos available on day 3. Our data show that day 5 transfers are beneficial for patients with a superior cohort of embryos (i.e., three or more eight-cell embryos on day 3), as evidenced by maintenance of a high ongoing pregnancy rate combined with a significant reduction in the incidence of high-order multiple gestations. Nevertheless, any benefit to day 5 transfers appears to be equivocal for patients with embryo cohorts of intermediate quality (one or two eight-cell embryos on day 3) and, strikingly, appears to be negative for patients with cohorts of poor quality (no eight-cell embryos on day 3). Collectively, the results of this analysis show that, at least with the culture system that we used, the number of eight-cell embryos on day 3 provides an appropriate criterion for prospective identification of the optimum day of embryo transfer. Stratification of cycles to identify the quality of embryo cohorts for this analysis was based on previous observations that cycle outcome is affected as much by the quality of the overall cohort as by the quality of the embryos transferred (12). Our findings show that when stratified by the number of eight-cell embryos available on day 3, embryo cohorts of varying quality were indeed distinguishable. This is evidenced by an increased frequency of cleavage arrest in an entire embryo cohort in which there were no eight-cell embryos on day 3, compared with a decreased frequency in intermediate cohorts with one or two eight-cell embryos. 562 Racowsky et al. Number of eight-cell embryos determines day of ET Vol. 73, No. 3, March 2000

TABLE 3 Pregnancy outcomes in day 3 versus day 5 transfers, stratified by number of eight-cell embryos on day 3 post-retrieval. No. of eight-cell embryos on day 3 Day of embryo transfer No. of transfers No. of patients with fetal hearts No. of fetal hearts 1 2 3 0 3 27 9 8 (88.9) 1 (11.1) 0 (0) 5 11 0 0 0 0 1 or 2 3 77 33 15 (45.4) 15 (45.4) 3 (9.1) 5 47 19 12 (63.1) 7 (36.8) 0 (0) 3 3 117 62 31 (50.0) 20 (32.3) 11 (17.7) 5 83 46 28 (60.9) 16 (34.8) 2 (4.3)* Note: Numbers in parentheses are % of the number of all patients with fetal hearts in that group. No statistically significant differences were noted between the day 3 and day 5 for any of the groups of 8-cell embryos except the incidence of pregnancies with 3 fetal hearts. * P.034. Notably, complete cleavage arrest was not observed when three or more eight-cell embryos were present. The underlying cause of cleavage arrest is unknown. However, because activation of the embryonic genome occurs on day 3, between the four- and eight-cell stage (13), such arrest probably relates to intrinsic abnormalities (14) for all embryos within that particular cohort. The data in Table 2 show that poor-prognosis patients (i.e., those with no eight-cell embryos on day 3) do not benefit from day 5 transfers. Indeed, 0% of these patients achieved a pregnancy after a day 5 transfer, in contrast to a 33% ongoing pregnancy rate after transfer on day 3. These observations indicate that the extended culture conditions used were inadequate to support normal development of embryos in this suboptimal, slower-cleaving cohort. In contrast, the uterus was apparently able to rescue some of the embryos in this cohort, because pregnancies were established in 33% of cases. Together, these findings indicate that further refinements of even the most recent extended culture media are needed to meet the metabolic demands of more slowly developing embryos. Until this occurs, day 3 transfer is preferable for this patient group. For patients with intermediate embryo quality (i.e., one or two eight-cell embryos on day 3), we found that pregnancy rates after day 5 transfer were no different from those after transfer on day 3. Additionally, the results did not show a statistically significant reduction in the incidence of highorder multiple gestations in this patient group (9.1% for day 3 transfers vs. 0% for day 5 transfers; Table 3). Although it remains to be determined whether a larger sample size would reveal significance, the present data suggest that day 5 transfers may not provide any additional benefit over day 3 transfers to patients with intermediate embryo quality. Indeed, in light of the additional laboratory work, expense, and TABLE 4 The incidence of blastocyst formation on day 5 according to the number of eight-cell embryos available on day 3. No. of eight-cell embryos on day 3 Variable 0 8 cells 1 or 2 8 cells 3 8 cells No. of blastocysts 2.4 2.5* 2.8 2.1 5.4 3.7 (range) (0 8) (0 7) (0 16) % blastocysts of total embryos 21.0 21.0 25.6 19.8 38.6 20.5 No. of expanded blastocysts 0.4 0.7 1.0 1.2* 2.4 2.8 (range) (0 2) (0 4) (0 18) % expanded blastocysts of total embryos 2.6 5.1 9.3 12.6 16.9 18.4 Note: Values are expressed as means SD. Compared with the 3 8-cell embryo group, significantly fewer blastocysts and expanded blastocysts were noted. * P.01. P.001. P.05. FERTILITY & STERILITY 563

expertise involved in extending culture to day 5, it may seem prudent to perform exclusively day 3 transfers in these patients. All patients with a superior cohort of embryos (i.e., three or more eight-cell embryos on day 3) and a day 5 transfer had embryos transferred that had progressed to or beyond the morula stage. No patient had to cancel her transfer because of complete cleavage arrest. Furthermore, pregnancy rates were similar to those of previous reports for day 5 transfers in which patients were selected on the basis of the presence of numerous follicles or zygotes (16). Importantly, the data provide additional support for the true benefit of day 5 transfer: that is, to reduce the incidence of high-order multiple pregnancies in patients at high risk for this undesirable outcome. Nevertheless, this benefit is tempered by the two cases of monozygotic twinning that we observed in the pregnant patients, representing an incidence of 4.3%. The underlying cause of this phenomenon is unknown but may relate to culture-induced increased hardening of the zona pellucida (17), which may be exacerbated by extending the duration of culture to day 5. Alternatively, the sequential culture media used may have induced perturbations in intercellular adhesion within the inner cell mass, resulting in splitting of this cell lineage. The varying quality in embryo cohorts across the three strata was also reflected in the incidence of overall blastocyst formation and the proportion of fully expanded blastocysts that developed. The positive impact on successful outcome of transferring fully expanded blastocysts was observed in an earlier study in this field (15). However, it remains to be seen whether fully expanded blastocysts grown in contemporary media are indeed developmentally superior to slower-developing blastocysts. Regardless, it is clear that such developmentally advanced embryos are not required for establishing viable pregnancies because in one case, a viable, ongoing pregnancy resulted from the transfer of morulae. Nevertheless, the overall implantation rate for morulae was very low compared with that for fully expanded blastocysts (4.5% vs. 44.2%). Collectively, the results of this study indicate that given our current technologies, there is a role for both day 3 and day 5 transfers in IVF, but that each is appropriate for specific patient populations. Therefore, there is a need to establish a clinical protocol that enables prospective identification of the most appropriate day for embryo transfer. The present study demonstrates that consideration of the number of eightcell embryos on day 3 provides the basis for such a protocol. Accordingly, we are currently implementing the following algorithm. For patients with at least eight embryos, transfer day is identified early in the morning of day 3 immediately after embryo assessment, which occurs 65 67 hours postinsemination. Patients with more than three eight-cell embryos are offered a day 5 embryo transfer. Those patients with fewer than three eight-cell embryos are encouraged to undergo a day 3 transfer. This schedule provides sufficient time for calling patients, answering questions, decision making, and traveling that day if a day 3 transfer is chosen. We find that adherence to this protocol necessitates streamlined communication among embryologists, clinical staff, and patients. Patients must be educated about the advantages of day 5 transfer but must be able to adjust their schedules with great flexibility. Effective implementation of this protocol is laborintensive and requires great elasticity among all members of the IVF team. However, the findings of this study indicate that the appropriate application of day 5 transfer maximizes the likelihood of success for all patients while minimizing the possibility of a high-order multiple pregnancy. References 1. Edwards RG, Hollands P. New advances in human embryology: implications of the preimplantation diagnosis of genetic disease. Hum Reprod 1988;3:549 56. 2. Bolton VN, Wren ME, Parsons JH. Pregnancies after in vitro fertilization and transfer of human blastocysts. Fertil Steril 1991;55:830 2. 3. Dawson KJ, Conaghan J, Ostera R, Winston RML, Hardy K. Delaying transfer to the third day post-insemination, to select non-arrested embryos, increases development to the fetal heart stage. Hum Reprod 1995;10:177 82. 4. Munne S, Alikani M, Tomkin G, Grifo J, Cohen J. Embryo morphology, developmental rates, and maternal age are correlated with chromosome abnormalities. Fertil Steril 1995;64:382 91. 5. Gardner DK, Vella P, Lane M, Wagley L, Schlenker T, Schoolcraft WB. Culture and transfer of human blastocysts increases implantation rates and reduces the need for multiple embryo transfers. Fertil Steril 1998;69:84 8. 6. Jones GM, Trounson AO, Gardner DK, Kausche A, Lolatgis N, Wood C. Evolution of a culture protocol for successful blastocyst development and pregnancy. Hum Reprod 1998;64:1125 9. 7. Carrillo AJ, Lane B, Pridham DD, Risch P, Pool TB, Silverman IH, et al. Improved clinical outcomes for in vitro fertilization with delay of embryo transfer from 48 to 72 hours after oocyte retrieval: use of glucose- and phosphate-free media. Fertil Steril 1998;69:329 34. 8. Jones GM, Trounson AO, Lolatgis N, Wood C. Factors affecting the success of human blastocyst development and pregnancy following in vitro fertilization and embryo transfer. Fertil Steril 1998;70:1022 9. 9. Palermo G, Joris H, Derde MP, Camus M, Devroey P, Van Steirteghem A. Sperm characteristics and outcome of human assisted fertilization by subzonal insemination and intracytoplasmic sperm injection. Fertil Steril 1993;59:826 35. 10. Freedman M, Farber M, Farmer L, Leibo SP, Heyner S, Rall WF. Pregnancy resulting from cryopreserved human embryos using a onestep in situ dilution procedure. Obstet Gynecol 1988;72:502 5. 11. Menezo Y, Nicollet B, Herbaut N, Andre D. Freezing cocultured human blastocysts. Fertil Steril 1992;58:977 80. 12. Jackson KV, Ginsburg ES, Hornstein MD, Rein MS, Clarke, RN. Multinucleation in normally fertilized embryos is associated with an accelerated ovulation induction response and lower implantation and pregnancy rates in in vitro fertilization-embryo transfer cycles. Fertil Steril 1998;70:60 6. 13. Braude PR, Bolton V, Moore S. Human gene expression first occurs between the four and eight cell stages of preimplantation development. Nature 1988;332:459 61. 14. Plachot M, Junca AM, Mandelbaum J, de Grouchy J, Salat-Barous J, Cohen J. Chromosome investigations in early life. II. Human preimplantation embryos. Hum Reprod 1987;2:29 35. 15. Dokras A, Sargent IL, Barlow DH. Human blastocyst grading: an indicator of developmental potential? Hum Reprod 1993;8:2119 27. 16. Gardner DK, Schoolcraft WB, Wagley L, Schlenker T, Stevens J, Hesla J. A prospective randomized trial of blastocyst culture and transfer in human in vitro fertilization. Hum Reprod 1998;13:3434 40. 17. Cohen J, Elsner C, Kort H, Malter H, Massey J, Mayer MP, et al. Impairment of the hatching process following IVF in the human and improvement of implantation by assisting hatching using micromanipulation. Hum Reprod 1990;5:7 13. 564 Racowsky et al. Number of eight-cell embryos determines day of ET Vol. 73, No. 3, March 2000