Benefit of intracytoplasmic sperm injection in patients with a high incidence of triploidy in a prior in vitro fertilization cycle

IN VITRO FERTILIZATION Benefit of intracytoplasmic sperm injection in patients with a high incidence of triploidy in a prior in vitro fertilization cycle Sunny H. Jun, M.D., a Thomas O Leary, B.S., b Katharine V. Jackson, B.S., b and Catherine Racowsky, Ph.D. b a Department of Obstetrics and Gynecology, Stanford University Medical Center, Stanford, California; and b Department of Obstetrics and Gynecology, Brigham and Women s Hospital, Boston, Massachusetts Objective: To test the hypothesis that intracytoplasmic sperm injection (ICSI) overcomes a high incidence of tripronucleate zygotes resulting from standard insemination in a previous cycle. Design: A retrospective analysis of matched-pair cycles. Setting: Assisted reproductive technologies (ART) program of Brigham and Women s Hospital. Patient(s): Ninety-five patients with a 20% incidence of tripronucleate zygotes in an IVF cycle with use of ICSI in a subsequent attempt. Intervention(s): Cycles with either standard insemination or ICSI. Main Outcome Measure(s): Incidence of diploid (2pn) and triploid (3pn) zygotes and number and quality of embryos obtained. Result(s): Patient age, ampules of gonadotropin used, peak E 2, number of follicles at hcg trigger, and total number of oocytes were all significantly higher in the ICSI cycles, but the number of mature oocytes did not differ. After ICSI, the percentage of 2pn was higher (65.0% vs. 34.1%) and the percentage of 3pn was lower (5.0% vs. 33.9%) than after IVF, and more diploid embryos were obtained with ICSI (5.5 3.7 vs. 3.4 2.2 [mean SD]). There was no difference in embryo quality between the two groups. Conclusion(s): ICSI appears beneficial in women with a high 3pn occurrence from IVF because it increases the number of diploid zygotes without affecting embryo quality. (Fertil Steril 2006;86:825 9. 2006 by American Society for Reproductive Medicine.) Key Words: 3pn, 2pn, triploidy, ICSI, IVF, insemination, embryo, zygote, dispermy The incidence of triploid or tripronucleate (3pn) human embryos ranges from 2% 10% after IVF (1 5). Several mechanisms may give rise to this abnormal condition, including [1] dispermy (fertilization of a haploid oocyte with two haploid sperm), [2] diplospermy (fertilization of a haploid oocyte and a diploid sperm), [3] digyny (fertilization of a diploid oocyte by a haploid sperm), [4] retention of the second polar body within the oocyte, and [5] pronuclear fragmentation (abnormal splitting of a single pronucleus). Of the above possible underlying causes of triploidy, dispermy is considered the most common (1). Therefore, in patients with IVF who experience a high incidence of 3pn zygotes after standard insemination, intracytoplasmic sperm injection (ICSI) may be a useful tool to overcome this abnormal condition in a subsequent treatment cycle. Our study was performed to assess this possibility by comparing Received October 13, 2005; revised and accepted March 13, 2006. Reprint requests: Catherine Racowsky, Department of Obstetrics and Gynecology, Brigham and Women s Hospital, 75 Francis St., ASB 1 3, Room 082, Boston, MA 02115 (FAX: 617-232-6346, E-mail: cracowsky@ partners.org). diploid and triploid fertilization rates for patients who had a high triploidy rate from standard insemination ( 20%) and who then proceeded to an ICSI cycle. The ultimate goal of this analysis was to determine whether subsequent ICSI would result in an increased number of normally fertilized zygotes and embryos for potential transfer. MATERIALS AND METHODS Patient and Cycle Selection The study protocol was approved by the Institutional Review Board of Brigham and Women s Hospital and the Partners Healthcare System. Review of the database of all assisted reproduction cycles conducted at Brigham and Women s Hospital between 1998 and 2004 (n 9,804), was performed to identify patients who had a 20% incidence of 3pn zygotes in an IVF cycle, followed by a cycle using exclusively ICSI. Donor egg, GIFT, zygote intrafallopian transfer (ZIFT), freeze all, and gestational carrier cycles were excluded from the analyses. Ninety-five patients met inclusion criteria. If a patient underwent more than one 0015-0282/06/$32.00 Fertility and Sterility Vol. 86, No. 4, October 2006 doi:10.1016/j.fertnstert.2006.03.043 Copyright 2006 American Society for Reproductive Medicine, Published by Elsevier Inc. 825

subsequent ICSI cycle, the cycle with a retrieval date closest to that of the IVF cycle was selected for analysis. Study Design The 95 IVF and 95 ICSI cycles were compared with respect to maternal age, total ampules of gonadotropins used, number of follicles on day of hcg administration, E 2 level on day of hcg administration, number of oocytes retrieved, number of metaphase II (MII) oocytes identified, and total motile sperm in the raw sample. Numbers and percentages of 2pn and 3pn zygotes obtained from MII oocytes, as well as number of 2pn zygotes that cleaved, were used to evaluate fertilization outcomes between the two groups. Quality of embryos (i.e., number of 8-cell embryos with 10% fragmentation and perfect symmetry) also was assessed. Stimulation Protocols Patients with normal results from clomiphene citrate challenge testing (generally FSH levels 10 miu/ml) underwent controlled ovarian stimulation with luteal down-regulation using leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, Deerfield, IL). Leuprolide acetate was begun either a week after documentation of urinary LH surge or the day after a midluteal P determination and was continued until at least day 2 of menses. Baseline ultrasonography and blood testing were then performed to document that no cysts 3 cm were present, E 2 was 50 pg/ml, and P was 1.5 ng/ml. Alternatively, in patients with histories of poorer gonadotropin responses or FSH levels 10 miu/ml, poor responder protocols were used. The most usual protocol was either a microdose LA protocol (0.05 mg LA SC two times per day) started cycle day 1 of a period after oral contraceptive (OC) pill lead-in and baseline ultrasound testing performed on day 2 or, alternatively, a GnRH antagonist protocol using an OC pill for 3 weeks, then baseline ultrasound testing on cycle day 2, with GnRH antagonist initiation at a dose of 0.25 mg/day SC starting stimulation day 6. When baseline criteria were met, gonadotropin therapy (either Gonal-F; Serono Laboratories, Inc., Rockland, MA, or Follistim; Organon, Roseland, NJ) with or without hmg (Humegon, Organon; Pergonal, Serono; or Repronex, Ferring Pharmaceuticals Inc., Suffern, NY) was begun. Stimulation generally was achieved with use of divided daily dosing of a total of 6 8 ampules/day to maximize follicular recruitment in patients of this age. Monitoring of follicle growth was achieved with use of ultrasound, and serum E 2 levels were measured starting on stimulation day 6 and then every 1 3 days as indicated. A dose of 10,000 IU of hcg (Profasi; Serono) was administered IM when two follicles reached a maximum diameter of 20 mm (mean 16.5 mm) and the E 2 concentration was 500 pg/ml. Transvaginal oocyte retrieval was performed 36 hours after hcg administration in the standard fashion with intravenous general anesthesia or, in some cases, spinal anesthesia as indicated. Oocyte Fertilization and Embryo Culture For the standard IVF insemination, oocytes were inseminated between 4 and 6 hours of retrieval with an average of 100,000 motile sperm in groups of three to five in 1 ml Ham s F-10 supplemented with 5% human serum albumin. For the ICSI procedure, oocytes were injected with a single sperm between 3 and 5 hours of the retrieval. Mature oocytes were identified by the presence of the first polar body in the ICSI cycles after removal of the corona cells following exposure to hyaluronidase and in the IVF cycles at the fertilization check. At the fertilization check, which was performed 16 18 hours after insemination or ICSI, the unfertilized oocytes and the haploid and triploid zygotes were identified, counted, and then set aside for ultimate discard at the completion of the cycle. Diploid zygotes were cultured individually into 25 L droplets of growth medium overlaid with 8 ml oil in Falcon 1007 culture dishes (Becton Dickinson Labware, Franklin Lakes, NJ). During the 6-year study period, four growth media were used in our laboratory: P1 (Irvine Scientific, Santa Ana, CA), IVF500, G1.2, or G1.3 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden). However, within a patient, the same type of growth medium was used for both her IVF and ICSI cycles. All cultures were maintained at 37 C in a humidified atmosphere of 5% CO 2 in air. On day 3, the morphology of each embryo was assessed with the use of standard criteria (5) 68 72 hours after insemination. Fragmentation was graded as 10%, 10% 25%, and 25% of the blastomere volume; asymmetry was graded according to uniformity in size and shape of the blastomeres as no asymmetry, moderately asymmetric, and severely asymmetric (6). Statistical Analyses Continuous variables were analyzed with the use of Wilcoxon s signed rank test for matched pairs. Differences in proportions were analyzed with 2 or Fisher s exact test, as appropriate. In all cases, P.05 was considered statistically significant. RESULTS Age, total number of ampules of gonadotropins, E 2 level, number of follicles, and total number of oocytes on day of hcg were all significantly higher in the ICSI group compared with the IVF group. The number of mature (MII) oocytes and total motile sperm did not differ between the two groups (Table 1). The fertilization outcomes are shown in Table 2. Comparison between the ICSI and IVF groups revealed that the percentage of 2pn zygotes from MII oocytes was significantly higher after ICSI (65.0% vs. 34.1%; P.0001). Conversely, the percentage of 3pn zygotes was significantly lower in the ICSI group compared with the IVF group (5.0% vs. 33.9%; P.0001). These outcomes were reflected in the 826 Jun et al. ICSI for high incidence of 3pn zygotes Vol. 86, No. 4, October 2006

TABLE 1 Cycle characteristics between IVF and ICSI groups. Cycle characteristics IVF group ICSI group P value Age (y) 36.9 3.9 37.4 4.0.0001 Total no. of ampules of gonadotropins 66.4 23.5 70.9 24.4.02 E 2 level on day of hcg (pg/ml) 1660.5 714.4 2004.9 939.5.0005 No. of follicles on day of hcg 9.8 4.4 11.6 6.3.002 Total motile sperm (millions) 142.4 132.2 129.5 102.5 NS No. of oocytes 10.3 5.3 12.1 7.3.003 No. of MII oocytes 7.9 4.5 8.6 5.3 NS Note: Values are means SD. NS; Not significant. mean numbers of both 2pn (5.5 4.0 vs. 2.9 2.3 [mean SD], ICSI vs. IVF respectively; P.0001) and 3pn (0.5 1.1 vs. 2.6 1.7, ICSI vs. IVF respectively; P.0001) zygotes obtained, as well as the mean number of zygotes that underwent cleavage (5.2 3.8 vs. 2.8 2.4, for ICSI vs. IVF respectively; P.0001; Table 2). A subanalysis was performed to assess the distribution of ICSI cycles according to the extent to which the incidence of triploidy was eliminated. There was 100% elimination of 3pn zygotes in 68 of 95 (71.6%) cycles, and less than 100% elimination but greater than 0% in 13 of 95 (13.7%) cycles. In 9 ICSI cycles (9.5%), there was no elimination of 3pn zygotes, and in 5 cycles (5.3%), the incidence was actually higher than that observed in the corresponding IVF cycle (Fig. 1). Overall, cleavage rates were similar between the two groups (data not shown). We further examined the quality of embryos having at least eight cells with respect to the degree of fragmentation and symmetry. There were no significant differences between the two groups (Table 3). Compared with the IVF group, the ICSI group had a significantly increased viable pregnancy rate (22.1% [21/95] vs. 5.2% [5/95]; P.0007) and improved implantation rates (number of viable fetuses at 8 weeks of number of embryos transferred: 6.8% [23/339] vs. 2.3% [5/213]; P.02). However, both the pregnancy and implantation rates in the ICSI subgroup were significantly lower than those overall in ICSI cycles performed in our program during the study period (pregnancy rates: 22.1% vs. 40.7%, P.001; implantation rates: 6.8% vs. 12.1%, P.01). DISCUSSION Our study shows that when patients have a high incidence of 3pn zygotes after a standard IVF cycle, use of ICSI ameliorates this abnormal condition in a subsequent cycle. Indeed, FIGURE 1 Distribution of cycles according to the extent to which the incidence of triploidy was eliminated after ICSI (cross-hatched bars). TABLE 2 Comparison of fertilization outcomes between IVF and ICSI groups. Fertilization outcome IVF group a ICSI group a % 2pn of MII oocytes 34.1 19.7 65.0 23.3 % 3pn of MII oocytes 33.9 13.8 5.0 10.1 No. of 2pn zygotes 2.9 2.3 5.5 4.0 No. of 3pn zygotes 2.6 1.7 0.5 1.1 No. of embryos 2.8 2.4 5.2 3.8 Note: Values are means SD. a P.0001 (IVF group vs. ICSI group for all values). Fertility and Sterility 827

TABLE 3 Embryo qualities for IVF and ICSI groups. Embryo quality IVF (n 271) a ICSI (n 526) a 8-cell, 10%fragmentation 45/271 (16.6%) 95/526 (18.1%) 8-cell, 10%fragmentation, with perfect symmetry 17/271 (6.3%) 44/526 (8.4%) 8-cell, 10%fragmentation, with moderate symmetry 27/271 (10%) 44/526 (8.4%) 8-cell, 10%fragmentation, with severe asymmetry 1/271 (0.4%) 7/526 (1.3%) Note: a P NS (IVF group vs. ICSI group for all values). the mean diploid fertilization rate of 65.0% after ICSI in the study population is close to the mean 72% diploid fertilization rate in our overall ICSI program. Furthermore, application of ICSI resulted in complete elimination of triploidy in slightly more than 70% of patients and a decrease in the incidence of triploid embryos in close to 95% of our study population. These observations, combined with the fact that embryos resulting from ICSI were of comparable quality with those arising from IVF, indicate that ICSI may be a beneficial treatment option when patients have a high incidence of triploidy after standard IVF insemination. Among the several underlying causes for triploidy, dispermy is recognized to be the most common (1). Possible explanations for the occurrence of polyspermy have been associated with the concentration of inseminated sperm and the timing of insemination (7 9). Although in vivo only several hundred sperm may reach the oocyte, in IVF as many as 100,000 motile sperm may be used to inseminate each egg. Wolf et al. (8) reported a greater than threefold increase in the incidence of polyspermy when the inseminated sperm concentration was doubled from 50,000 motile sperm/ml. Trounson et al. (9) found a 30% occurrence of polyspermy when oocytes were inseminated immediately versus no occurrence when the insemination was delayed 5 to 6 hours after retrieval. Immediately after egg retrieval, the cortical granules are thought still to be migrating toward the oocyte surface. Thus, after fertilization, there may be a delay in the release of cortical granule contents, which subsequently may affect the zona reaction to block further sperm entry (10). It is noteworthy that 100,000 sperm/ml is the optimum concentration of sperm for insemination in our culture system, and we routinely perform inseminations between 4 and 6 hours after retrieval. With these findings taken together, therefore, it seems likely that a portion of the oocytes in our study population were prone to an abnormal kinetics of cortical granule exocytosis that, in turn, resulted in dispermy. Because dispermy is considered to be the most common contributor to triploidy resulting from standard insemination, we used ICSI in an attempt to overcome the high incidence of 3pn zygotes observed from a previous IVF cycle. To our knowledge, there are no previous studies in the literature assessing a potential benefit of using ICSI in this particular population of patients with infertility. Before undertaking this study, we recognized that cycle characteristics between our IVF and ICSI groups were unlikely to be similar in most aspects because of an obvious age increase in the follow-on ICSI cycle and our clinical practice of being more aggressive in a subsequent attempt after a cycle failure. This was confirmed by our results when variables such as age, total amount of gonadotropins, E 2 level, and numbers of follicles and oocytes on day of hcg were observed to be significantly different. Nevertheless, among all the variables investigated, we believe that showing no difference in the mean number of MII oocytes was the most important because we derived our fertilization outcomes solely from this subcohort of retrieved oocytes (Table 1). When our patients underwent ICSI in their subsequent cycles, the incidence of 2pn zygotes nearly doubled from 34.1% to 65.0%. Conversely, the occurrence of 3pn zygotes decreased from 33.9% in the IVF group to 5% in the ICSI group. Moreover, 70% of our patient population had 100% elimination of 3pn zygotes after ICSI, indicating that, in this population at least, the formation of 3pn zygotes after IVF insemination was, indeed, most likely caused by dispermy. In addition to the patients with complete 3pn elimination after ICSI, a further 14% experienced some reduction in the incidence of 3pn zygotes with ICSI, although 5% of patients showed an increased incidence compared with their corresponding IVF cycle. The mechanisms underlying the failure of ICSI to overcome the triploid condition in some oocytes are unknown but may have resulted from either fragmentation of the pronuclei, retention of the second polar body, or both. Abnormal spindle formation or stabilization may underpin polar body retention (11, 12). Alternatively, defects associated with the centrosomes and motor proteins leading to altered control mechanisms during chromosome segregation may be involved (13). Finally, the ICSI procedure per se may contribute to polar body retention, particularly in those few cases in which there was an increased 3pn occurrence after sperm injection. Indeed, ICSI has been reported to exert hydrostatic pressure in the oocyte, which may disrupt the microtubules of the oocyte spindle and cause an abnormal 828 Jun et al. ICSI for high incidence of 3pn zygotes Vol. 86, No. 4, October 2006

distribution of the maternal chromosomes into two separate sets after extrusion of the first polar body (14). Use of ICSI resulted in a significant increase in the mean number of embryos from 3.4 to 5.5, without any compromise in embryo quality. Similar percentages of embryos with 8-cell, 10% fragmentation and degree of symmetry were obtained for both groups (Table 3). The number of goodquality embryos on day 3 has been reported to be of strong predictive value for both pregnancy and implantation rates (15). Volpes et al. demonstrated a pregnancy rate of 39.4% when three good-quality embryos (i.e., embryos having 8 cells with 20% fragmentation) were obtained versus 64.3% when 3 embryos were generated (15). In summary, the results of our study support the hypothesis that ICSI can overcome a high incidence of tripronucleate zygotes resulting from standard insemination in a previous cycle. The data show that ICSI appears beneficial in such patients, because it increases the number of diploid zygotes without affecting overall embryo quality. Indeed, this nonconventional application of ICSI to overcome the occurrence of triploidy from dispermy after IVF significantly increases the likelihood of achieving a successful outcome in a subsequent assisted reproductive technology attempt. Nevertheless, the clinical outcomes in the ICSI subgroup were significantly lower than those in our overall ICSI population during the study period. Therefore, even though normal fertilization is achieved after ICSI in this subgroup, the implantation potential of these normally fertilized embryos may be unusually low. REFERENCES 1. Boyers SP, Diamond MP, Lavy G, Russell JB, DeCherney AH. The effect of polyploidy on embryo cleavage after in vitro fertilization in humans. Fertil Steril 1987;48:624 7. 2. Fishel SB, Cohen J, Fehilly C, Purdy JM, Walters DE, Edwards RG. Factors influencing human embryonic development in vitro. Ann NY Acad Sci 1985;442:342 56. 3. Marrs RP, Saito H, Yee B, Sato F, Brown J. Effect of variation of in vitro culture techniques upon oocyte fertilization and embryo development in human in vitro fertilization procedures. Fertil Steril 1984; 41:519 23. 4. Mettler L, Michelmann HW. Chromosome studies of early human embryos: proof of fertilization in uncleaved human oocytes. Ann NY Acad Sci 1985;442:458 65. 5. Mahadevan M, Baker G. Assessment and preparation of semen for in vitro fertilization. In: Wood C, Trounson A, eds. Clinical in vitro fertilization. Berlin: Springer-Verlag, 1984:83. 6. Racowsky C, Combelles CMH, Nureddin A, Pan Y, Finn A, Miles L, et al. Day 3 and day 5 morphological predictors of embryo viability. Reprod Biomedicine Online 2002;6(2):76 84. 7. Diamond MP, Rogers BJ, Webster BW, Vaughn WK, Wentz AC. Polyspermy: effect of varying stimulation protocols and inseminating sperm concentrations. Fertil Steril 1985;43:777 80. 8. Wolf DP, Byrd W, Dandekar P, Quigley MM. Sperm concentration and the fertilization of human eggs in vitro. Biol Reprod 1984;31:837 48. 9. Trounson AO, Mohr LR, Wood C, Leeton JF. Effect of delayed insemination on in-vitro fertilization, culture and transfer of human embryos. J Reprod Fertil 1982;64:285 94. 10. Dandekar PV, Martin MC, Glass RH. Polypronuclear embryos after in vitro fertilization. Fertil Steril 1990;53:510 4. 11. Combelles CMH, Albertini DF, Racowsky C. Distinct microtubule and chromatin characteristics of human oocytes after failed in-vivo and in-vitro meiotic maturation. Hum Reprod 2003;10:2124 30. 12. Windt ML, Coetzee K, Kruger TF, Marino H, Kitshoff MS, Sousa M. Ultrastructural evaluation of recurrent and in-vitro maturation resistant metaphase I arrested oocytes. Hum Reprod 2001;16:2394 8. 13. Hodges CA, Ilagan A, Jennings D, Keri R, Nilson J, Hunt PA. Experimental evidence that changes in oocyte growth influence meiotic chromosome segregation. Hum Reprod 2002;17:1171 80. 14. Macas E, Imthurn B, Rosselli M, Keller PJ. The chromosomal complements of multipronuclear human zygotes resulting from intracytoplasmic sperm injection. Hum Reprod 1996;11:2496 501. 15. Volpes A, Sammartano F, Coffaro F, Mistretta V, Scaglione P, Allegra A. Number of good quality embryos on day 3 is predictive for both pregnancy and implantation rates in in vitro fertilization/intracytoplasmic sperm injection cycles. Fertil Steril 2004;82:1330 6. Fertility and Sterility 829