Human embryo fragmentation in vitro and its implications for pregnancy and implantation

Transcription

1 FERTILITY AND STERILITY VOL. 71, NO. 5, MAY 1999 Copyright 1999 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Human embryo fragmentation in vitro and its implications for pregnancy and implantation Mina Alikani, M.Sc., Jacques Cohen, Ph.D., Giles Tomkin, B.A., G. John Garrisi, Ph.D., Caryn Mack, M.S., and Richard T. Scott, M.D. Institute for Reproductive Medicine and Science of Saint Barnabas, West Orange, New Jersey Objective: To evaluate the effects of fragmentation and fragment removal in day 3 human embryos on implantation and pregnancy. Design: Retrospective analysis of ETs homogeneous with respect to embryo fragmentation. Setting: A program of IVF-ET. Patient(s): The study population consisted of 2,410 patients. Intervention(s): The degree and pattern of fragmentation were evaluated on days 2 and 3; microsurgical fragment removal was performed after assisted hatching on day 3. Main Outcome Measure(s): Clinical pregnancy and implantation rates. Result(s): The degree and pattern of fragmentation significantly impact pregnancy and implantation. With the application of microsurgical fragment removal before ET, embryos with 6% 35% fragmentation implant with similar frequency. The presence of large fragments (type IV) is detrimental to the developing embryo, whereas localized or small and scattered fragments do not significantly affect implantation. Conclusion(s): The potential of fragmented embryos for implantation is determined partly by the distribution of fragments. Adoption of an embryo classification system reflecting types of fragmentation is advisable. The use of microsurgical fragment removal significantly alters the course of development for some embryos and improves their implanting potential. (Fertil Steril 1999;71: by American Society for Reproductive Medicine.) Key Words: pattern, degree of fragmentation, fragment removal, implantation Received August 21, 1998; revised and accepted December 18, Reprint requests: Mina Alikani, M.Sc., 101 Old Short Hills Road, Suite 501, West Orange, New Jersey (FAX: ; mina.alikani@embryos.net) /99/$20.00 PII S (99) is a common feature of human embryos developing in vitro (1, 2). When exceeding 10% of the embryonic volume, it is thought to reflect a certain aberration in development. Although a study by Jurisicova et al. (3) clearly showed that programmed cell death or apoptosis is triggered in arrested fragmented human embryos, the exact manner in which fragments interfere with development of nonarrested embryos remains to be clarified. The current protocols assess the quality of an embryo based largely on the degree to which it has fragmented. This system of embryo grading reflects early knowledge of human embryo morphology (4). It is now inadequate because it distinguishes fragmented embryos only at a rudimentary level and may underestimate the developmental potential of certain types (5, 6). The largest studies published to date on the impact of fragmentation and the outcome of IVF are those of Staessen et al. (7), Giorgetti et al. (8), and Ziebe et al. (9). These investigators reported low implantation rates (approximately 5%) after transfer of embryos with 10% 50% fragmentation on day 2 of development. These reports do not address the differences among embryos within this wide range, nor do they consider the size and distribution of the fragments. Since August 1995, our record keeping of embryonic development and replacement has included detailed tracking of individual embryos. Therefore, we have been able to retrospectively analyze a large series of cases and establish apparent relationships between fragmentation and the incidence of pregnancy and implantation. We have considered specific de- 836

2 grees and patterns of fragmentation in this analysis and indicate their impact as well as the effects of fragment removal on the implanting potential of fragmented embryos. MATERIALS AND METHODS Patients and Ovarian Stimulation Institutional review board approval was not required for this study because the work is routine at this institution and based on a previously approved protocol. This involved 3,000 patients without obvious side effects (Cornell University Medical College, 1989). The study population consisted of patient for which pregnancy outcome was known (n 2,410). Thaw cycles were excluded. All patients undergoing IVF-ET were stimulated with use of either a standard down-regulation protocol or a modification of the microflare protocol (10) (for patients 40 and those patients in whom previous response to down-regulation was unsatisfactory). Patient Procedures With Homogeneous Transfers To eliminate or minimize correlative uncertainties when multiple embryos were transferred, only patient with homogeneous transfers were selected for these analyses. A transfer was considered homogeneous when more than one half of the embryos replaced were in the same category. A pool of 2,410 was searched for homogeneous transfers. For the degree of fragmentation, 1,727 were found to be homogeneous ( W groups). For the pattern of fragmentation, 570 were homogeneous ( T groups). Procedures common to W and T groups were homogeneous in both degree and pattern of fragmentation and were analyzed separately. The degree of fragmentation in replaced embryos in groups W1 W5 was 0% 5%, 6% 15%, 16% 25%, 26% 35%, and 35%, respectively. The pattern of fragmentation in groups T1 T5 was types I V, respectively, and follows the definitions below. Because only six patients were found to be in group T5, this group was excluded from all analyses. Embryo Culture and Assessment of Embryos were cultured individually in L droplets of human tubal fluid medium (HTF) (11), under mineral oil (Squibb, Princeton, NJ) and supplemented with heatinactivated maternal serum. On day 2 and day 3 of development in vitro, embryos were evaluated with use of the 40 objective plus the 1.5 magnifying lens of an Olympus IX70 inverted microscope (Olympus America, Melville, NY), equipped with Hoffman Modulation Optics (Narishige, Japan). Total magnification was 600. The degree of fragmentation was expressed as a percentage and defined as the volume of the perivitelline space FIGURE 1 Day 3 fragmented human embryos with fragmentation patterns I, II, III, IV, and V. Left: Embryos on day 3 of development before assisted hatching and fragment removal. Right: The same embryos after assisted hatching and fragment removal. Note distinct differences among the organization and appearance of the remaining blastomeres after removal of different types of fragments. and/or cleavage cavity occupied by anucleate cytoplasmic fragments. Based on distribution and size of the fragments, five patterns of fragmentation were defined (12, 13). Embryos with these patterns are presented in Figure 1, before and after fragment removal. FERTILITY & STERILITY 837

3 FIGURE 2 The percentage of fragments removed from fragmented embryos versus the degree of fragmentation (in percent). The highly fragmented embryos had the most fragments removed. of type I is minimal in volume, and the fragments are associated typically with only one blastomere. Type II fragments are localized, and the fragments predominantly occupy the perivitelline space. Small, scattered fragments comprise type III fragmentation; these fragments may be in the cleavage cavity or peripherally positioned. Pattern IV has large fragments, sometimes resembling whole blastomeres. These fragments often are distributed randomly and are associated with uneven cells. Type V fragments appear necrotic, with a characteristic granularity and cytoplasmic contraction within the intact blastomeres. When embryonic fragmentation did not fit any of these patterns, it was described as no distinct pattern. Assisted Hatching and Fragment Removal All embryos fragmented beyond 5% arriving to day 3 of development and selected for replacement were subjected to selective assisted hatching (14), followed by microsurgical removal of some or all fragments (see Fig. 1 and 2). Micromanipulation were performed on the heated stage of an Olympus IX70 inverted microscope, using Narishige hydraulic micromanipulators (Narishige). A microneedle measuring m in outer diameter attached to a mouth suction unit was used for the procedure. Fragment removal was attempted at the highest magnification and with continuous refocusing and rotation of the embryo to avoid damaging intact blastomeres and to provide access to more fragments. The needle was moved towards the contralateral side of the aperture in the zona or placed between blastomeres to remove fragments. Medium was injected occasionally in the intercellular spaces to facilitate fragment aspiration. Any blastomeres lysed accidentally during the procedure were removed completely. Data and Statistical Analysis The EggCyte database was used to collect all data (Egg- Cyte, ART Institute of NY and NJ, Livingston, NJ, ). To find the relationship between fragmentation and pregnancy and implantation, variables represented as proportions (pregnancy and implantation rates) were analyzed by means of logistic regression. Tukey s multiple comparisons test was used to compare mean cell numbers among different fragmentation patterns. RESULTS Degree of and Its Impact on Implantation and Pregnancy To measure the impact of fragmentation on implantation and pregnancy, we considered the degree and type of fragmentation as evaluated on the morning of developmental day 3. Data for the homogeneous transfer groups in the degree of fragmentation ( W groups, n 1,727) are presented in Table 1. During the 1,727 reviewed, a total of 30, Alikani et al. Human embryo fragmentation in vitro Vol. 71, No. 5, May 1999

4 TABLE 1 Pregnancy and implantation in five transfer groups according to the degree of fragmentation. range (%) group Total no. of patients with positive B-hCG with FHB 0 Clinical pregnancy embryos replaced sacs FHBs Implantation 0 5 W1 1, ,651 1,312 1, W , W W W Total 1,727 1,207 1, ,916 1,998 1, Note: FHB fetal heart beat. oocytes were retrieved, 5,916 embryos were replaced of which 2,105 had fragment removal. The average fragmentation rate for all day 3 embryos and for those replaced on day 3 was 15.4% and 8.6%, respectively. The incidence of clinical pregnancy, defined as fetal heart beat detected on ultrasound, was 59.8% (1,033 of 1,727). Implantation rate, defined as fetal heart beat per embryo replaced, was 29.9% (1,766 of 5,916). The average ( SD) maternal age in this population was years. A significant decrease in implantation and pregnancy occurs as fragmentation increases in degree (P.001). This significance is largely due to the drop in pregnancy and implantation beyond 35% fragmentation. Microsurgical Fragment Removal After microsurgical fragment removal, none of the transfers in the study involved embryos with 25% fragmentation. Clearly, the most fragments were removed from embryos with 35% fragmentation (Fig. 2). The damage rate was calculated on the basis of the number of embryos in which one or more blastomeres lysed as a result of contact with the assisted hatching needle. The overall damage rate was 6.33% (255 of 3,775), and it ranged from 1.7% (11 of 654) in embryos 0% 5% fragmented to 11.4% (22 of 193) in embryos with 35% fragments. The pattern of fragmentation also affected the degree to which cells were damaged accidentally. Embryos with no distinct pattern of fragmentation were those most frequently damaged (27 of 223; 12.1%), followed by types V, IV, III, II, and I (4 of 190; 2.1%). Pattern of and Its Impact on Implantation and Pregnancy Data for the T groups are presented in Table 2. There is a general decline in implantation and pregnancy in successive groups T1 T4, with the largest decrease in group T4 (P.001). A meaningful trend was observed when transfers homogeneous in both degree and pattern of fragmentation were considered. Compared with the implantation rate of embryos with 6% to 35% fragmentation (without consideration of the TABLE 2 Pregnancy and implantation in four transfer groups according to the pattern of fragmentation. type group Mean percent day 3 fragmentation* (SD) Total no. of patients with positive -hcg with FHB 0 Clinical pregnancy embryos replaced sacs FHBs Implantation T1 5.9 (2.91) T (6.77) T (5.89) , T (10.43) Total , Note: FHB fetal heart beat. * Before fragment removal. P.001. P.001. FERTILITY & STERILITY 839

5 FIGURE 3 The incidence of implantation in transfer groups T2 ( ), T3 (Œ), and T4 (X) compared with implantation of embryos according to the degree of fragmentation and without consideration of the pattern of fragmentation (E). Homogeneous transfer of type IV fragmented embryos led to the lowest rate of implantation. pattern of fragmentation), embryos with 6% to 35% type IV fragments implanted least frequently (Fig. 3). Cell Number and Patterns of These data are presented in Table 3. Embryos with localized fragments (type II) had an average ( SD) of cells on day 2. This was significantly lower compared with all other fragment types (P.05). On day 3, both types II ( cells) and IV ( cells) embryos had fewer cells than others (P.05). DISCUSSION TABLE 3 Average cell number on days 2 and 3 of development according to the four patterns of fragmentation. Fragment type Day 2 cell number Day 3 cell number I II 3.21* 5.98* III IV * * Significantly different from cell number of other types on that day (P.05). What is evident from years of recorded observations on fragmentation is that it does limit development of the human embryo in vitro. The specifics of this developmental limitation, however, have been far from clear. Comparisons among programs have been uncertain because of inadequate descriptions of morphology or indirect evaluations such as assignment of grades or scores. This has hampered the development of an acceptable embryo classification system that, for instance, would serve as a general aid in embryo selection for replacement. The present study focuses on an important aspect of embryo morphology, namely, fragmentation, from an angle other than that used traditionally. This work establishes a relationship between the degree and pattern of fragmentation on day 3 of development and the incidence of implantation and pregnancy after selective assisted hatching and fragment removal. Our findings indicate that not all fragmentation, even at high levels, is detrimental to the embryo. The pattern of fragmentation has a profound effect on the developmental fate of the embryo, thus must be considered in its evaluation. Of the three mechanisms proposed to generate these patterns (13), the loss of large volumes of cytoplasm appears detrimental to the embryo. Exclusive replacement of embryos with large fragments, designated here as type IV pattern of fragmentation, produced an implantation rate (18%) significantly lower than that of other types. In addition, a significant decrease in clinical pregnancy rate occurred after homogeneous transfer of such embryos (40% compared with an overall pregnancy rate of 58.5%). Electron microscopic evidence suggests that large fragments 840 Alikani et al. Human embryo fragmentation in vitro Vol. 71, No. 5, May 1999

6 originate from cells of two- or four-cell embryos, because these fragments have mitochondria that are more electrondense than in later stage cells (15). The release of large fragments at an early stage may deplete the embryo of essential organelles, such as mitochondria, or structures, such as pinocytotic caveolae, that are involved in uptake of exogenous protein (16). Moreover, the portion of the cell that retains the nucleus may actually arrest after such a great loss. The minimum cytoplasmic volume required by a cell at various stages to undergo further division is not known at present. The average degree of fragmentation on day 3 was higher (25%) in group T4 (type IV replaced exclusively), raising the possibility that this was the cause of diminished implantation. Thus, we considered implantation in transfers homogeneous for both degree and pattern of fragmentation. Although statistical significance could not be established, a trend is evident in Figure 3. Exclusive replacement of type IV fragmented embryos, with 6% 35% fragmentation, led to implantation rates lower than their type II or III counterparts or the implantation rate of fragmented embryos regardless of the pattern of fragmentation. In addition, implantation in group W3 with 16% 25% fragmentation is, in fact, identical to that of group W2 with 6% 15% fragmentation. It is thus more likely that the difference in implantation of type II and type IV embryos is a result of fragmentation type rather than degree alone. In contrast with large fragments, small, scattered fragments (type III) do not affect appreciably cell number during the 3 days in culture; neither do they appear to pose a serious threat to further development. On the basis of the appearance of these fragments, it is likely that they are generated during successive divisions. They may not point to a specific anomaly but simply to imperfect cytokinesis. The distribution of fragments is also of significance, as in the case of localized fragments (type II), which appear to have resulted from the complete fragmentation of one or more cells. Despite lower cell numbers both on day 2 and day 3 of development, the implantation rate of embryos with type II fragmentation is nearly 34%. Elimination of selected blastomeres by the embryo may reflect its effort in restoring or maintaining viability when anomalies affect particular blastomeres. For instance, limited diploid mosaicism is not uncommon among day 3 embryos (17). Early loss of such cells obviously prevents their later contribution to the inner cell mass or the trophectoderm of the blastocyst. There was no correlation between the appearance of any pattern and maternal age, indicating that ovarian response to exogenous gonadotropins or oocyte quality probably are not significant factors in the development of fragmentation patterns. Although implantation is best when embryos are not fragmented, moderately fragmented embryos (6% 35% by volume) implant with similar frequency (28% 23%). This is consistent with cytogenetic data indicating that approximately 60% of highly fragmented embryos (up to 40%) are chromosomally normal (18). However, our data are in sharp contrast with previous reports that estimate the implantation rate of such embryos at 5% (7, 9). In the current study, a low implantation rate (6%; all delivered) occurred only when replaced embryos had 35% fragmentation before fragment removal. This interesting observation may be attributed to microsurgical fragment removal, because it was applied routinely during the course of this study. The aggressive application of fragment removal is based on our previous finding of a 4% overall increase in implantation rate when assisted hatching and fragment removal were applied simultaneously; this was in comparison with embryos that only were zona drilled (19). We propose two hypotheses, one or both of which may explain this effect. The first is that removal of extracellular fragments restores the spatial relationship of cells within the embryo. Fragments positioned in the cleavage cavity (see Fig. 1III) may cause distortion of division planes or interfere with normal cell-cell contact, leading to abnormal compaction, cavitation, and blastocyst formation. Furthermore, fewer cells may be allocated to the inner cell mass, because this depends on contact between sister blastomeres (20). Preliminary experiments in our laboratory indicate that blastocysts formed after fragment removal are better organized than their unmanipulated counterparts. The latter blastocysts often form multiple cavities with numerous excluded fragments and cells in the perivitelline space (unpublished observations). A second, and probably concurrent, explanation for the high rate of implantation of fragmented embryos after fragment removal may be the prevention of secondary degeneration. On the transmission electron microscope, blastomeres immediately adjacent to fragments show early signs of vacuolar degeneration (2). Removal of fragments and cell degradation products may stop this deterioration of the remaining cells. Perhaps a lesson is to be learned from lower organisms, such as Caenorhabditis elegans, in which at least eight genes are dedicated to the function of cell corpse recognition, engulfing, and degradation by neighboring cells (21). Because blastomeres of early human embryos do not appear to have such phagocytic activity, necrotic fragments and cell remains are not effectively removed from the developing embryo. In addition to fragments and remains of cells, the perivitelline space in fragmented embryos often appears to be filled with other debris. Some of this debris is traced to the polar body or corona cell processes (1). However, it is possible that debris also is generated by fragment components that do not remain confined within their membrane. During artificial removal, fragments, cells completely devoid of cytoplasm, as well as structures resembling extra- FERTILITY & STERILITY 841

7 cellular vesicles, occasionally break as they come in contact with the needle. The close proximity of any of these components to the surrounding cells can cause further damage, leading to the demise of an otherwise viable embryo. This was demonstrated in the mouse, where the presence of deliberately lysed cells along with healthy cells led to reduced blastocyst expansion and hatching (22). This work is partly an effort toward expanding the focus of embryo morphology assessment to include the size and distribution of fragments. This is a different concept from the traditional view, indicated only by blastomere size. These data also demonstrate the impact of fragment removal on the implanting potential of partially fragmented embryos. Acknowledgments: The authors thank Eurof Walters, Ph.D., and Ms. Nury Steurwald for statistical evaluation of these data. They also thank embryologists Toni Ferrara-Congedo, Adrienne Reing, Marlena Blake, Sasha Sadowy, Rene Walmsley, and Elena Kissin for their skilled technical contributions. References 1. Trounson A, Sathananthan AH. The application of electron microscopy in the evaluation of two- to four-cell human embryos cultured in vitro for embryo transfer. J In Vitro Fert Embryo Transf 1984;1: Sathananthan AH, Bongso A, Ng SC, Ho J, Mok H, Ratnam S. Ultrastructure of preimplantation human embryos co-cultured with human ampullary cells. Hum Reprod 1990;5: Jurisicova A, Varmuza S, Casper RF. Programmed cell death and human embryo fragmentation. Mol Hum Reprod 1996;2: Veeck LL. Oocyte assessment and biological performance. Ann NY Acad Sci 1988;541: Erenus M, Zouves C, Rajamahendran P, Leung S, Fluker M, Gomel V. The effect of embryo quality on subsequent pregnancy rates after in vitro fertilization. Fertil Steril 1991;56: Shulman A, Ben-Nun I, Ghetler Y, Kaneti H, Shilon M, Beyth Y. Relationship between embryo morphology and implantation rate after in vitro fertilization treatment in conception cycles. Fertil Steril 1993; 60: Staessen C, Janssenwillen C, Van Den Abbeel E, Devroey P, Van Steirteghem AC. Avoidance of triplet pregnancies by elective transfer of two good quality embryos. Hum Reprod 1993;8: Giorgetti C, Terriou P, Auquier P, Hans E, Spach JL, Salzmann J, et al. Embryo score to predict implantation after in vitro fertilization: based on 957 single embryo transfers. Hum Reprod 1995;10: Ziebe S, Petersen K, Lindenberg S, Andersen AG, Gabrielsen A, Andersen AN. Embryo morphology or cleavage stage: how to select the best embryos for transfer after in vitro fertilization. Hum Reprod 1997;12: Scott RT, Navot D. Enhancement of ovarian responsiveness with micro-doses of GnRH-agonist during ovarian induction for in vitro fertilization. Fertil Steril 1994;61: Quinn P, Warnes GM, Kerin JF, Kirby C. Culture factors affecting the success rate of IVF and embryo transfer. Ann NY Acad Sci 1985;442: Alikani M, Cohen J. Patterns of cell fragmentation in the human embryo. J Assist Reprod Genet 1995;12:28S. 13. Warner CM, Cao W, Exley GE, McElhinny AS, Alikani M, Cohen J, et al. Genetic regulation of egg and embryo survival. Hum Reprod 1998; 13: Cohen J, Alikani M, Trowbridge J, Rosenwaks Z. Implantation enhancement by selective assisted hatching using zona drilling of human embryos with poor prognosis. Hum Reprod 1992;7: Sathananthan H, Ng SC, Bongso A, Trounson A, Rtnam S. Visual atlas of early human development for assisted reproductive technology. Singapore: National University of Singapore, Sathananthan AH, Wood C, Leeton J. Ultrastructural evaluation of 8 16 cell human embryos developed in vitro. Micron 1982;13: Munné 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: Munné S, Cohen J. Chromosome abnormalities in human embryos. Hum Reprod Update 1999; In press. 19. Cohen J, Alikani M, Ferrara T, Munné S, Reing A, Schatman G, et al. Rescuing abnormally developing embryos by assisted hatching. In: Mori T, Aono T, Tominaga T, Hiroi M, eds. Frontiers in endocrinology: perspectives on assisted reproduction. Vol. 4. Rome: Ares Serono Symposia, 1994: Edwards RG, Beard HK. Oocyte polarity and cell determination in early mammalian embryos. Mol Hum Reprod 1997;3: Ellis RE, Jacobson DM, Horvitz HR. Genes required for the engulfment of cell corpses during programmed cell death in Caenorhabditis elegans. Genetics 1991;129: Alikani M, Olivennes F, Cohen J. Microsurgical correction of partially degenerate mouse embryos promotes hatching and restores their viability. Hum Reprod 1993;8: Alikani et al. Human embryo fragmentation in vitro Vol. 71, No. 5, May 1999