Predictive value of embryo grading for embryos with known outcomes

Transcription

1 Predictive value of embryo grading for embryos with known outcomes Vanessa N. Weitzman, M.D., Jennifer Schnee-Riesz, M.D., Claudio Benadiva, M.D., John Nulsen, M.D., Linda Siano, M.S., and Donald Maier, M.D. Center for Advanced Reproductive Services, University of Connecticut School of Medicine, Department of Obstetrics and Gynecology, Farmington, Connecticut Objective: To compare pronuclear morphology (Z-score), day 3 embryo grade, and day 3 cell number in the prediction of successful implantation rates (IRs), including cycles in which all or none of the embryos implanted. Design: Retrospective analysis. Setting: University-based IVF center. Patient(s): Four hundred twenty-six fresh IVF day 3 transfers of 82 embryos in women <36 years of age from January to December 3 in whom all or none of the embryos implanted. Main Outcome Measure(s): Evaluation of Z-scores, embryo morphology, cell number, and IR. Result(s): Day 3 parameters were more predictive than Z-scores. When early parameters were poor (Z-score) but late parameters were both good, the IR was 38%, compared with 4% when the Z-score was good but the late parameters were poor. Conclusion(s): Embryo grading systems are useful in the prediction of embryo implantation. In particular, cell number and embryo grade are more predictive than Z-scores. Therefore, late parameters have a better prognostic value than Z-scores when selecting embryos for transfer. (Fertil Steril Ò 1;93: Ó1 by American Society for Reproductive Medicine.) Key Words: Embryo grading, embryo score, pronuclear morphology, Z-score One of the most challenging clinical problems in IVF is the selection of the best embryo for transfer. In spite of multiple advances in the field of advanced reproductive technologies (ART), the efficiency of IVF remains low, with <% of all embryos transferred resulting in a clinical pregnancy (1 3). This low rate leads to the practice of transferring multiple embryos to achieve a good pregnancy rate; however, this practice then leads to an undesirable rate of multiple gestations. Therefore, a technique for embryo selection that would allow the reduction in the number of embryos transferred while increasing pregnancy rates is greatly needed. Many different techniques have been used to analyze and select the most viable embryos for transfer. These methods need to be noninvasive and not time-consuming. Embryo selection for day 3 transfers is generally based on the key morphological features of cleaving embryos that have been previously related Received 24, September 8; revised, January 9; accepted 2, February 9; published online May 1, 9. V.N.W. has nothing to disclose. J.S.-R. has nothing to disclose. C.B. has nothing to disclose. J.N. has nothing to disclose. L.S. has nothing to disclose. D.M. has nothing to disclose. Presented at the 62nd Annual Meeting of the American Society for Reproductive Medicine, which was held in Philadelphia, Pennsylvania, on October 16-21, 4. Reprint requests: John Nulsen, M.D., Center for Advanced Reproductive Services, Dowling South Building, 263 Farmington Avenue, Farmington, Connecticut (FAX: ; Nulsen@nso1. uchc.edu). to increased implantation (2). In addition, a pronuclear morphology scoring system was developed by Scott et al. and Scott and Smith (2, 4) based on previously published observations of zygote morphology and empirical observations correlating with pregnancy (2, 4). Tesarik and Greco () developed a modified grading system in which the nucleoli size, number, and distribution were used in a single observation scoring. On the third day of culture, embryo morphology was used to select the embryos for transfer. The results after a retrospective analysis found a strong correlation between zygote scoring and clinical pregnancy rates (2, ). The Tesarik zygote-grading system had the advantage of being a single observation and using fewer parameters for evaluation. Tesarik et al. later simplified their grading system from their original version into a single-observation zygote scoring system with either a normal (pattern ) or abnormal pronuclear stage pattern. They and others then reported that by using this modified scoring system and transferring at least one pattern embryo, significantly more pregnancies resulted than transfers without any pattern zygotes (6 8). In 6, Scott et al. published a modification of their earlier grading system (9) in an IVF program with extended culture and blastocyst transfer in % of ETs. These results suggest that zygote scoring can maintain pregnancy rates for day 3 and day transfers, increase implantation rates (IRs), and reduce the number of embryos needed to be transferred to achieve a pregnancy (2, 7). 68 Fertility and Sterility â Vol. 93, No. 2, January, 1-282/1/$36. Copyright ª1 American Society for Reproductive Medicine, Published by Elsevier Inc. doi:1.116/j.fertnstert

2 However, other studies have found that the use of the pronuclear morphology is not better than standard embryo morphology for predicting pregnancy outcome. Neither Nicoli et al. nor Payne et al. found any significant differences in embryo IR or pregnancy outcomes when using pronuclear morphology compared with standard morphological outcome (3, 7). This discrepancy in outcome may be the result of the different scoring systems used as well as the unknown outcomes of all embryos transferred. The purpose of our study was to compare pronuclear morphology (Z-score), day 3 embryo grade, and day 3 cell number in the prediction of successful IR. We included only cycles in which all or none of the embryos implanted to ensure that only embryos with known outcome were evaluated. MATERIALS AND METHODS This study included 426 fresh IVF day 3 transfers of 82 embryos in women <36 years of age who underwent controlled ovarian hyperstimulation (COH) for IVF at the Center for Advanced Reproductive Services at the University of Connecticut from January to December 3. In this retrospective review, only cycles in which two embryos were transferred and either both or none of the embryos implanted were included. The study was approved by the Institutional Review Board of the University of Connecticut. All the women in the study underwent one of two COH protocols, the long luteal phase GnRH agonist suppression protocol or the GnRH antagonist protocol, as determined by their primary physician depending on patient age, antral follicle count, and serum FSH levels. Patients who used the long GnRH agonist protocol started. mg of leuprolide acetate (Lupron; TAP Pharmaceuticals, North Chicago, IL) in the midluteal phase of the preceding cycle. A transvaginal ultrasound and serum E 2 assessment were then performed after the onset of menses to confirm pituitary suppression, as shown by the absence of follicular activity and a serum E 2 level of <7 pg/ml. Once pituitary suppression was achieved, controlled ovarian stimulation was initiated as described below, and the dose of lupron was then reduced to.2 mg daily and continued until the day of trigger of oocyte maturation. Patients who used the GnRH antagonist protocol were evaluated on day 2 of their menses, and gonadotropins were commenced if the ovaries were quiescent on ultrasound. Ganirelix acetate (Ganirelix; Organon Pharmaceuticals, Roseland, NJ) was initiated once the leading follicle was R14 mm or the E 2 level was R pg/ml and continued daily until the day of trigger. All the patients used either recombinant FSH (Gonal F; Serono, Inc., Rockland, MA) alone or in combination with purified urinary hmg (Repronex; Ferring Pharmaceuticals, Inc., Suffern, NY) for ovarian stimulation. The standard daily starting dose of gonadotropins was 4 IU depending on the patient s age, body mass index, antral follicle count, and basal serum FSH levels. Monitoring of follicular growth was achieved with serial ultrasound and serum E 2 measurements, and the dose of gonadotropins was adjusted, if necessary, according to follicular response. In both groups, 3 1, IU of hcg were administered SC depending on follicular response and serum E 2 levels (1) when at least two follicles reached R17 mm in diameter, followed hours later by ultrasound-guided transvaginal oocyte retrieval. The IVF and intracytoplasmic sperm injection (ICSI) procedures were performed as described elsewhere (1, 11); 18 hours after either insemination or ICSI, fertilization was assessed. Confirmation of fertilization was made by noting the presence or absence of two pronuclei (2PN). At this stage, the embryos underwent pronuclear scoring. The Z-score scoring system was based on the work by Scott et al. and Tesarik and Greco (2, ). The zygotes are assigned a Z-score based on the alignment, number, and size of nucleoli found at this 2PN stage. Z1 includes zygotes in which the nucleolar precursor bodies (NPBs) were aligned in each pronucleus. There were at least three NPBs in each pronucleus, and there was no difference of more than three NPBs between the two pronuclei. Z2 corresponded to embryos in which the NPBs were also equal, with at least three in each pronucleus; however, the NPBs were scattered throughout and not aligned. Z3 corresponded to embryos that were of different morphologies with either an unequal number of NPBs or different positions (aligned in one pronuclei while scattered in the other). Z4 included pronuclei that were separated, of different sizes, or displaced to the periphery. On day 3, all embryos were evaluated for cell number and embryo morphology. Each embryo transferred was evaluated for blastomere size and fragmentation. Embryos exhibiting equal blastomere size and no fragmentation were considered G1. G2 embryos had blastomeres of equal size with slight fragmentation (<%), while G3 embryos had blastomeres of unequal size but no fragmentation. G4 embryos had blastomeres of equal or unequal size and moderate fragmentation (% %), and G grade embryos were those with unrecognizable blastomeres and severe fragmentation (>%). In all cases, two embryos, chosen by the embryologist, were transferred on the third day after retrieval. All patients received mg P in oil IM daily for luteal support, which was initiated the day after oocyte retrieval. This was continued until a negative serum pregnancy test (bhcg) was achieved. If the serum pregnancy test was positive (bhcg > IU/L), P was continued at least until a transvaginal ultrasound confirmed fetal heart activity. Clinical pregnancy was defined as a normal gestational sac measured with a transvaginal ultrasound after weeks, and an ongoing pregnancy was defined as a clearly visible fetal pole with a normal fetal heart rate observed after 8 weeks. Statistical analysis was done using c 2 and Fisher s exact test methods. P<. was considered statistically significant. RESULTS Our study included a total of 426 fresh IVF transfers from January to December 3. We analyzed a total of 82 Fertility and Sterility â 69

3 embryos that were fertilized by either IVF or ICSI where all or none of the embryos implanted (negative pregnancy test or twin pregnancy). All embryos were classified by standard parameters, Z-score, cell number, and embryo grade, and were transferred on day 3; 144 transfers resulted in twin gestation. Mean maternal age was noted to be 31.3 years, with no statistically significant differences between any of the groups analyzed. On analysis of early morphologic features, the majority of the zygotes analyzed were noted to be Z1 (4.8%, 338 embryos). Z2, Z3, and Z4 accounted for 31.3%, 26.%, and 1.4% of all embryos, respectively. IRs were Z1, 38.8%; Z2, 33.46%; Z3, 28.6%; and Z4, 36.36%, with no statistically significant difference (Fig. 1). When evaluating cell grade, G2 was the largest group (4 embryos, 49.3%). Grades 1, 3, 4, and accounted for 19.9%, 8.2%, 22.3%, and.2% of the embryos, respectively. IRs per grade were statistically significant between each of the five groups and were 41.76% for G1, 39.7% for G2, 28.7% for G3, 18.42% for G4, and % for G (P<.1; Fig. 2). Analysis of cell number at day 3 was the final evaluation performed. When analyzed, the majority of embryos fell into the category of 6 8 cells on day 3 (722 embryos, 84.%). Cell number was divided into three categories, % cells, 6 8 cells, and R9 cells; 1.4% of all embryos accounted for the % cell group, while 4.8% accounted for the R9 cell group. IRs were 7.9% for the % cell group, 37.3% for the 6 8 cell group, and 29.3% for the R9 cell group and were significantly different between all three groups (P<.1; Fig. 3). Outcomes were then combined, and the results were analyzed to differentiate between early scoring outcomes versus late scoring outcomes. Z scores of 1 and 2 were considered good early outcomes while grades 1 and 2 were considered good late outcomes. When early outcomes were compared, good early outcomes resulted in a 36.% IR, compared with a 28.6% IR for poor early outcomes, which was not significantly different. However, when good late outcomes were compared with poor late outcomes, IRs were found to 39.% versus 21%, which was statistically significant (P<.1). When combined, results were compiled into four categories, good early and late outcomes, group 1; good early outcomes with poor late outcomes, group 2; poor early outcomes and good late outcomes, group 3; and poor early and late outcomes, group 4. When analyzed, group 1 was noted to have the highest IR, at 41.8%; group 2 had an IR of 23.4%; group 3 had an IR of 34.4%; and group 4 had an IR of 17%. These differences were statistically significant (P<.1; Fig. 4). When adding the cell number into the analysis as well, those embryos with good outcomes in Z-score, embryo grade as well as cell number, were noted to have the highest IR, at 43%. When early parameters were poor (Z-score) but late parameters were noted to be good, both grade and cell number, the IR was 38%, compared with 4% when the Z-score was good but the late parameters were poor (P<.1; Fig. ). FIGURE 1 IR (%) based on Z-score Z1 Z2 Z3 Z4 Implantation rate % Weitzman. Predictive value of embryo grading. Fertil Steril 1. DISCUSSION The goal of embryo grading at any stage is to select the embryos with the highest implantation potential. By having this ability to predict which embryos are more ideal to replace, we can increase implantation and pregnancy rates while decreasing the risk of multiples. Traditionally, embryo quality has been evaluated on the basis of day 2 or 3 morphology and embryo development. In the past, increasing the number of embryos transferred has been a way to increase pregnancy rates; however, this came at the cost of increasing multiple pregnancy rates. With the ability to extend culture, better embryo selection has reduced the multiple gestation rates while improving pregnancy rates (12, 13). The most current guidelines from the American Society for Reproductive Medicine suggest replacing one and no more than two embryos in women under age who have a favorable prognosis (14) and two in women aged 37. This continues to apply pressure on centers to define which embryos are best for transfer to achieve the highest IR. FIGURE 2 IR (%) based on day 3 grade G1 G2 G3 G4 G Implantation Rate % Weitzman. Predictive value of embryo grading. Fertil Steril Weitzman et al. Predictive value of embryo grading Vol. 93, No. 2, January, 1

4 FIGURE 3 IR (%) based on day 3 cell number to 8 9 Implantation rate % Weitzman. Predictive value of embryo grading. Fertil Steril 1. Fertilization is a process that culminates in the union of the maternal and paternal pronuclei, leading to the formation of new, unique individuals (). Multiple complex events must occur for this process to be successful. These events include sperm penetration, oocyte activation, development of male and female pronuclei, and migrations to the center of the oocyte. Apposition of the pronuclei at the center of the oocyte is a prerequisite for paternal and maternal chromosome assembly (). It has also been suggested that the alignment of the NPBs in the zygote may be a predictor of implantation (2,, 6, 16). Various forms of pronuclear morphology assessment have been used (2, 4,, 9, 17), and these studies have reported different predictive abilities (3). Tesarik and Greco were the first to classify zygotes into patterns ( ) based on size, number, and distribution of nucleoli (). In the process, they found a strong correlation between implantation and FIGURE 4 IR (%) with combined outcomes. 4 1 Combined outcomes GG GP PG PP Implantation rates %- GG=Good early and late outcomes, GP=Good early Poor late outcomes, PG=Poor early and Good late outcomes, PP= Poor early and late outcomes Weitzman. Predictive value of embryo grading. Fertil Steril 1. FIGURE IRs with combination of all scoring systems. 4 1 GG PG GP Implantation rate % GG= Good Z score, Grade and Cell number, PG= Poor Z score Good Grade and Cell number, GP= Good Z score Poor Grade and Cell number Weitzman. Predictive value of embryo grading. Fertil Steril 1. pattern. At the same time, Scott et al. classified zygotes into 4 grades (Z1 Z4) based on the distribution and size and nucleoli within each nucleus (4). They found that a greater number of zygotes in the Z1 and Z2 grades went on to develop into blastocysts. Z-scoring has also been associated with rates of aneuploidy in one paper that found that Z1 and Z4 scores correlated with the expected embryo outcomes and preimplantation genetic diagnosis findings (18). At this time, it is still controversial as to which criteria are best for choosing optimal embryos as different investigators (12, 13, 19, ) have reported a correlation between blastocyst formation and zygote morphology, while others (3, 7, 21, 22) found no significant differences in embryo quality or implantation and pregnancy rates in the different zygote grades. When the Z-score is combined with growth rate and grade, a number of investigators (12, 13, ) have also found a correlation between blastocyst formation and increases in IR. A number of difficulties exist in the interpretation of these studies as the data collected by the different investigators have not been uniform. Final outcomes measured in these different papers range from blastocyst development to pregnancy rates (3). One major confounder in a majority of these studies is the practice of transferring multiple embryos with different grades, which does not allow for the identification of the actual embryo responsible for the establishment of the pregnancy. To remove this confounder, we analyzed results from cycles in which the outcome for all the embryos was known. All cycles had two embryos transferred, and only those that resulted in either no pregnancy or a twin pregnancy were analyzed, with all singleton and monochorionic twin pregnancies excluded. Our results show that, individually, embryo grade and cell number better predict implantation when compared with the Z-score alone. When the results of the scoring systems are combined, it is seen that when early embryo scoring was poor but late parameters were good, implantation was better than when early parameters were good but late parameters were not. Therefore, combined late scoring systems (grade Fertility and Sterility â 661

5 and cell number) were more predictive than the Z-score for predicting outcomes. Interestingly, the results showed that individually the grade and cell number were as predictive as when the two scores were combined. When all three scoring systems were good, the outcomes were the highest observed. While it is intuitive that embryos with better scores would tend toward better outcomes, this study allows quantification of the relative contribution of different aspects of embryo morphology at different stages. Being able to predict with relative accuracy those embryos that have higher IR will allow us to limit the numbers of embryos transferred per cycle, thereby decreasing the chance of high-order multiple gestation pregnancies while increasing overall pregnancy rates. Along with the weakness inherent in a retrospective study, there is also the inability to exclude cycles in which a single embryo split creates identical twins. While none of these twin pregnancies were monochorionic in nature, some may have been dichorionic and diamniotic. However, these pregnancies are rare,.4% in our center, and most likely have little impact on the outcomes noted here. This study was limited to patients under the age of 36 to eliminate the influence of the known decrease in IR in an older population. Therefore, the results found here may not be applicable to an older population in which implantation and pregnancy rates are known to be significantly lower owing to the higher rate of embryo aneuploidy. This may be an explanation for the lack of significance in the outcomes of the Z-scores in our patient population. Ultimately, our results demonstrate that embryo grading systems can assist in selecting superior quality embryos when all three systems are combined. Z-score alone was not superior to the standard morphologic criteria used for selecting embryos on day 3, but a combination of both early and late parameters did make a significant difference. Therefore, by being able to recognize the embryos with the highest likelihood of implantation, physicians will not only be able to better counsel patients on the number of embryos to transfer but will also be able to select the best embryos to achieve a pregnancy as well as decrease the likelihood of multiple gestations. REFERENCES 1. Edwards RG, Beard HK. Blastocyst stage transfer: pitfalls and benefits. Is the success of human IVF more a matter of genetics and evolution than growing blastocysts? Hum Reprod 1999;14: Scott L, Alvero R, Leondires M, Miller B. The morphology of human pronuclear embryos is positively related to blastocyst development and implantation. Hum Reprod ;: Nicoli A, Valli B, Di Girolamo R, Di Tommaso B, Gallinelli A, La Sala GB. Limited importance of pre-embryo pronuclear morphology (zygote score) in assisted reproduction outcome in the absence of embryo cryopreservation. Fertil Steril 7;88: Scott LA, Smith S. The successful use of pronuclear embryo transfers the day following oocyte retrieval. Hum Reprod 1998;13: Tesarik J, Greco E. The probability of abnormal preimplantation development can be predicted by a single static observation on pronuclear stage morphology. Hum Reprod 1999;14: Tesarik J, Junca AM, Hazout A, Aubriot FX, Nathan C, Cohen-Bacrie P, et al. Embryos with high implantation potential after intracytoplasmic sperm injection can be recognized by a simple, non-invasive examination of pronuclear morphology. Hum Reprod ;: Payne JF, Raburn DJ, Couchman GM, Price TM, Jamison MG, Walmer DK. Relationship between pre-embryo pronuclear morphology (zygote score) and standard day 2 or 3 embryo morphology with regard to assisted reproductive technique outcomes. Fertil Steril ;84: Wittemer C, Bettahar-Lebugle K, Ohl J, Rongiere C, Nisand I, Gerlinger P. Zygote evaluation: an efficient tool for embryo selection. Hum Reprod ;: Scott L, Finn A, O Leary T, McLellan S, Hill J. Morphologic parameters of early cleavage-stage embryos that correlate with fetal development and delivery: prospective and applied data for increased pregnancy rates. Hum Reprod 7;22: Schmidt DW, Maier DB, Nulsen JC, Benadiva CA. Reducing the dose of human chorionic gonadotropin in high responders does not affect the outcomes of in vitro fertilization. Fertil Steril 4;82: Van Steirteghem AC, Nagy Z, Joris H, Liu J, Staessen C, Smitz J, et al. High fertilization and IR after intracytoplasmic sperm injection. Hum Reprod 1993;7: Lan K, Huang F, Lin Y, Kung F, Hsieh C, Huang H, et al. The predictive value of using a combined Z-score and day 3 embryo morphology score in the assessment of embryo survival on day. Hum Reprod 3;18: De Placido G, Wilding M, Strina I, Alviggi E, Alviggi C, Mollo A, et al. High outcome predictability after IVF using a combined score for zygote and embryo morphology and growth rate. Hum Reprod 2;17: ASRM Practice Committee. Guidelines on number of embryos transferred. Fertil Steril 6;86:s1 2.. Plachot M. Fertilization. Hum Reprod ;: Chen C, Kattera S. Comparison of pronuclear zygote morphology and early cleavage status of zygotes as additional criteria in the selection of day 3 embryos: a randomized study. Fertil Steril 6;8: Scott L. Pronuclear scoring as a predictor of embryo development. Reprod Biomed Online 3;6: Edirisinghe WR, Jemmott R, Smith C, Allan J. Association of pronuclear Z score with rates of aneuploidy in in vitro-fertilised embryos. Reprod Fertil Dev ;17: Nagy Z, Dozortsev D, Diamond M, Rienzi L, Ubaldi F, Abdelmassih R, et al. Pronuclear morphology evaluation with subsequent evaluation of embryo morphology significantly increases implantation rates. Fertil Steril 3;8: Zollner U, Zollner ZP, Hartle G, Dietl J, Steck T. The use of a detailed zygote score after IVF/ICSI to obtain good quality blastocysts: the German experience. Hum Reprod 2;17: Salumets A, Hyden-Granskog C, Suikkari AM, Titinen A, Tuuri T. The predictive value of pronuclear morphology of zygotes in the assessment of human embryo quality. Hum Reprod 1;16: James AN, Hennessy S, Reggio B, Wiemer K, Larsen F, Cohen J. The limited importance of pronuclear scoring of human zygotes. Hum Reprod 6;21: Weitzman et al. Predictive value of embryo grading Vol. 93, No. 2, January, 1