Effects of triploidy after intracytoplasmic sperm injection on in vitro fertilization cycle outcome Molina B. Dayal, M.D., M.P.H., Paul R. Gindoff, M.D., Abbaa Sarhan, M.D., Anil Dubey, Ph.D., Douglas Peak, B.S., and David Frankfurter, M.D. Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, The George Washington University, Washington, District of Columbia Objective: To evaluate the impact on the rates of clinical pregnancy and live birth of polyploidy after intracytoplasmic sperm injection (ICSI). Design: Retrospective cohort study. Setting: University-based IVF center. Patient(s): One hundred forty-three patients undergoing their first IVF embryo transfer cycle requiring ICSI. Intervention(s): None. Main Outcome Measure(s): Patients were divided into two groups on the basis of the proportion of post-icsi triploid fertilization that was observed at the time of fertilization assessment: group 1 included patients with %20% three-pronuclear (3PN) zygotes, and group 2 included patients with >20% 3PN zygotes. The primary outcomes compared between groups were pregnancy and implantation rates; secondary outcome measures included clinical-pregnancy rate and live-birth rate per embryo transfer. Result(s): Pregnancy, implantation, clinical-pregnancy, and live-birth rates were significantly higher in the cohort of patients who had %20% of embryos appearing triploid, compared with the group who had >20% of zygotes appearing triploid (relative risk [RR] for pregnancy, 2.4 [95% confidence interval {CI}, 1.22 4.77]; RR for implantation, 2.6 [95% CI, 1.17 5.56]; RR for clinical pregnancy, 2.8 [95% CI, 1.16 6.85]; and RR for live birth, 2.6 [95% CI, 1.06 6.38]). The proportion of 3PN zygotes after ICSI had a statistically significant inverse relationship to clinical-pregnancy rate. Conclusion(s): The proportion of triploid zygotes after ICSI serves as a negative prognostic indicator for IVF cycle outcome. This finding suggests that triploidy proportion is a surrogate marker of oocyte competence that represents the integrity of the oocytes in the entire recruited cohort. Such findings therefore may influence recommendations for embryo transfer number and freezing of supernumerary embryos. (Fertil Steril Ò 2009;91:101 5. Ó2009 by American Society for Reproductive Medicine.) Key Words: Polyploidy, triploidy, fertilization, ICSI, IVF outcome, oocyte competence, oocyte quality, live birth More than 3 million babies have been born since the pioneering efforts of Steptoe and Edwards (1). Although IVF initially was a highly inefficient process, 3 decades worth of technological advances have made it an increasingly useful therapy. However, most IVF-generated human embryos fail to implant (2). Because the morphological assessment of embryo competence remains limited and suboptimal, most IVF cycles fail to yield a pregnancy, and practitioners continue to rely on the transfer of multiple embryos (2). Noninvasive means have been proposed that are aimed at identifying competent embryos. Multiple paradigms that integrate cleavage rate, cell number and fragmentation, morphokinetics (3), and subcellular markers such as pronuclear morphology, have been used to predict chromosomal integrity (4). Although these strategies add insight, they do not reduce the need for additional cues predicting oocyte and embryo competence. Received August 3, 2007; revised and accepted November 14, 2007. Reprint requests: Molina B. Dayal, M.D., M.P.H., Fertility and IVF Center, George Washington Medical Faculty Associates, George Washington University, 2150 Pennsylvania Avenue, NW, 6th floor, Washington, District of Columbia 20037 (FAX: 202-741-2518; E-mail: mdayal@mfa.gwu. edu). It long has been observed that IVF can result in the formation of zygotes with three pronuclei (3PN). If this is encountered after oocytes are inseminated with sperm, it is believed that triploidy results either from polyspermic fertilization or from oocyte-derived meiotic failure (5 7). Although the use of intracytoplasmic sperm injection (ICSI) essentially eliminates dispermic triploidy, it does not prevent oocyte-induced 3PN formation. The latter form of triploidy results from a failure in meiosis II polar-body extrusion (8). Because IVF outcome is highly dependent on oocyte quality, a high incidence of triploidy in ICSI cycles may suggest an occult oocyte factor that can serve as a surrogate marker of oocyte competence and may serve as an independent predictor of IVF cycle outcome. Some investigators have suggested that the incidence of triploid fertilization after IVF is a result of advanced maternal age or severe sperm abnormalities (9). Although these inherent patient attributes may predispose them to triploid fertilization, other investigators have suggested that the propensity toward triploidy is a function of ovarian stimulation, as indicated by high peak E 2 levels, large oocyte yields, high gonadotropin doses, and lengthy stimulations (8, 10). Irrespective of the cause, the question remains as to whether the 0015-0282/09/$36.00 Fertility and Sterility â Vol. 91, No. 1, January 2009 101 doi:10.1016/j.fertnstert.2007.11.033 Copyright ª2009 American Society for Reproductive Medicine, Published by Elsevier Inc.
degree of triploidy observed after ICSI predicts the quality of the resultant embryo cohort. It has been shown that the presence of triploid embryos negatively correlates with IVF-ET pregnancy outcome. However, it remains to be determined whether the relative proportion of 3PN zygotes impacts upon the likelihood of IVF-ET success. In the current study, we attempt to assess whether triploid frequency after ICSI correlates with IVF outcome independently of traditional markers of oocyte competence in patients undergoing IVF-ICSI. MATERIALS AND METHODS A retrospective review of IVF embryo transfer (ET) cycles at the Fertility and IVF Center of the George Washington Medical Faculty Associates assessed all initial IVF-ET cycles that used fresh, nondonated oocytes and that required ICSI between January 1, 2000 and December 31, 2005. Intracytoplasmic sperm injection was used to minimize fertilization failure in those couples for whom an abnormal parameter (count, motility, or morphology) was noted on semen analysis, as defined by the World Health Organization and Kruger s strict criteria. Couples requiring use of either testicular or epididymal sperm were excluded from this study. All patients underwent ET 3 days after oocyte retrieval. To better elucidate the potential independent effect of degree of triploidy on live-birth rate, a receiver operating characteristic curve for the 143 cycles was generated to determine the optimal triploidy percentage for data analysis. Degree of triploid fertilization was grouped in 1% increments. The point that was furthest from baseline, 20% triploidy, was selected as the best discriminatory point for data analysis. Accordingly, group allocation was based on on the degree of triploidy (3PN) that was observed after ICSI at the time of fertilization check (18 20 h after ICSI). Group 1 included those cycles that had %20% 3PN zygotes present 18 20 hours after ICSI, whereas group 2 included cycles that had >20% 3PN zygotes observed after ICSI. This study was approved by the George Washington University Medical Center Office of Human Research Institutional Review Board. Ovarian Stimulation and Gamete Retrieval Controlled ovarian stimulation was accomplished by using either pituitary suppression in the luteal phase with a GnRH agonist (leuprolide acetate; TAP Pharmaceuticals, Lake Forest, IL) or midfollicular pituitary suppression with a GnRH antagonist (Antagon; Organon Pharmaceuticals, Roseland, NJ). Split doses of recombinant FSH (rfsh, Serono Inc, Rockland, MA or rfsh, Organon Pharmaceuticals) or purified hmg (Menopur or Repronex; Ferring Pharmaceuticals, Suffern, NY) were used to induce multiple follicle development. Stimulation then was tailored in accordance with follicular response, as determined by ultrasound monitoring and E 2 levels. When at least two follicles were noted to have R18 mm mean diameter on ultrasound, 10,000 IU of hcg (Profasi; Serono Inc., Rockland, MA) or 250 mg of hcg (Ovidrel; Serono Inc.) was administered, followed by oocyte retrieval 36 hours later. Luteal support was accomplished via vaginal P (Prometrium, 200 mg 3 times per d; Solvay Pharmaceuticals, Baudette, MN), beginning the morning after oocyte retrieval. Intracytoplasmic Sperm Injection After egg retrieval, oocyte cumuli were removed by exposure to hyaluronidase (80 miu/ml). Sperm with grossly normal morphology and progressive motility were selected for injection. After immobilization with the micropipette at the 9-o clock position, each mature oocyte was oriented such that the polar body was at the 6-or 12-o clock position. The injecting pipette then was gently advanced through the zona and the oolemma until the pipette was beyond the center of the oocyte, and the sperm was gently injected into the cytoplasm of the oocyte. Embryo Culture Embryos were cultured individually in 10-mL droplets of global media (LifeGlobal, Guilford, CT), supplemented with 10% human serum albumin (LifeGlobal), and overlaid with mineral oil (Sigma Chemical Co., St. Louis, MO) in a 6% humidified CO 2 incubator at 37 C. Selection of embryos for transfer was based on morphologic criteria and on the rate of embryo cleavage on day 3 after egg retrieval. Identification of Polyploid Zygotes Assessment of fertilization and of ploidy was performed 18 20 hours after performance of ICSI. Triploidy was defined as the presence of 3PN at the time of fertilization assessment. Statistical Analysis Statistical comparisons between group 1 (cycles with %20% 3PNs) and group 2 (cycles with >20% 3PNs) were performed regarding pregnancy and implantation rates (primary outcomes) and clinical-pregnancy and live-birth rates (secondary outcomes). Relative risks, with 95% confidence intervals (CIs), were calculated accordingly, and statistical significance was confirmed when the CI did not overlap unity. Continuous variables (age, day 3 FSH, peak E 2, number of oocytes, number of embryos transferred, sperm concentration, and total number of motile sperm) were assessed for normalcy by using Kolmogorov-Smirnov testing and were analyzed by using the Kruskal-Wallis test for nonnormally distributed data, with a P value of <.05 considered statistically significant. For those variables found to be statistically significant, logistic regression was performed for categorical outcomes, whereas linear regression was used to analyze continuous variables. Pregnancy was defined as a serum b-hcg level of >5 miu/ ml, 14 days after egg retrieval. Implantation rate was defined 102 Dayal et al. Triploidy after ICSI impacts IVF outcome Vol. 91, No. 1, January 2009
TABLE 1 Cycle-specific parameters associated with post-icsi triploidy. Parameter Group 1 %20% 3PN Group 2 >20% 3PN No. of patients 94 49 Mean age (y) 36.5 3.4 37.4 3.9 Mean day 3 FSH (miu/ml) 5.8 2.2 6.6 2.7 Mean peak E 2 (pg/ml) 2,230 912 1,677 868 Mean length of stimulation (d) 11.6 2.3 12.6 2.8 Mean no. of oocytes 16.9 7.2 a 10.0 7.4 Mean no. of embryos transferred 3.1 1.2 2.7 1.3 Mean sperm concentration (10 6 ) 59 39 59 40 Mean no. of motile sperm (10 6 ) 90 90 91 83 Note: Data are mean standard deviation (SD), unless otherwise indicated. a Statistically significant difference between groups, with P value of <.05. Dayal. Triploidy after ICSI impacts IVF outcome. Fertil Steril 2009. as the number of gestational sacs that were seen on transvaginal ultrasound at 4 weeks after ET, divided by the total number of embryos transferred. Clinical-pregnancy rate was defined as an intrauterine gestational sac and fetal heart motion noted on transvaginal ultrasound 4 weeks after ET. Live-birth rate was defined as the total number of pregnancies yielding a live-born infant beyond 24 weeks of gestational age, per initiated cycle. RESULTS Between January 1, 2000 and December 31, 2005, a total of 143 initial IVF-ET cycles were identified with triploid fertilization after ICSI. Ninety-four cycles were found to have %20% of their zygotes triploid (group 1), whereas 49 cycles had >20% of zygotes with 3PN (group 2). The distribution of cycle-specific parameters between groups is described in Table 1. For the entire study, the proportion of 3PN zygotes after ICSI was inversely related to clinical-pregnancy rate, with a P value for this trend of.013 (correlation coefficient, r ¼ 0.20). Pregnancy, implantation, clinical-pregnancy, and live-birth rates were significantly higher in the cohort of patients with %20% of embryos appearing triploid, compared with the case of the group with >20% of zygotes appearing triploid (Table 2). Implantation rates in cycles with <20% triploid zygotes were more than two and a half times higher than those in cycles with <20% triploid zygotes (RR, 2.6; 95% CI, 1.17 5.56). Clinical-pregnancy and live-birth rates were similarly higher in group 1 vs. group 2 (RR, 2.8; 95% CI, 1.16 6.85 and RR, 2.6; 95% CI, 1.06 6.38, respectively). Regression modeling did not affect the outcomes analysis (pregnancy, implantation, clinical-pregnancy, and live-birth rates) between groups. Regarding traditional cycle measures, the mean (SD) age of patients with %20% triploid fertilization (36.5 3.4 y) was similar to those with >20% triploid fertilization (37.4 3.9 y; P>.05). The duration of stimulation did not differ between group 1 (11.6 2.3 d) and group 2 (12.6 2.8 d; P>.05). Although the mean number of oocytes obtained TABLE 2 Effect of post-icsi triploidy on IVF outcome. Parameter Group 1 (%20% 3PN) Group 2 (>20% 3PN) Relative risk (95% CI) No. of patients 94 49 NA Pregnancy rate per ET (%) 40 16 2.4 (1.22 4.77) Clinical-pregnancy rate per ET (%) 29 10 2.8 (1.16 6.85) Implantation rate (%) 14 5 2.6 (1.17 5.56) Live-birth rate per ET (%) 27 10 2.6 (1.06 6.38) Note: NA ¼ not applicable. Dayal. Triploidy after ICSI impacts IVF outcome. Fertil Steril 2009. Fertility and Sterility â 103
was higher in group 1 vs. group 2 (16.9 vs. 10.0; P<.01), a yield of 10 oocytes is not conventionally believed to confer a poor prognosis. No significant difference was noted between groups with regard to other cycle-specific parameters, such as day 3 FSH, peak E 2 levels, number of embryos transferred, mean sperm concentration, or mean total motile sperm. In addition, the number of so-called good-quality embryos (defined as nonfragmented 6- to 8-cell cleavage-stage embryos) transferred did not differ between group 1 (2.8 1.1) and group 2 (2.4 1.7; P¼.09). Type of stimulation protocol, the proportion of patients using either recombinant or purified hmg, and cause of infertility (tubal disease, unexplained infertility, diminished ovarian reserve, ovulatory dysfunction, or male factor) did not differ between groups (data not shown). DISCUSSION Triploidy results from an additional complement of maternal (digynic) or paternal (diandric) chromosomes (11). Diandric triploidy is observed with spontaneous as well as conventional IVF and is assumed to be the most common form of triploidy (5). Intracytoplasmic sperm injection, by its virtue of injecting a single sperm into a single oocyte, negates the potential for dispermic triploidy. Digynic triploidy occurs secondary to failed meiosis II expulsion of the second polar body and subsequent fertilization of a diploid oocyte. Cytogenetic evidence of 3PN post-icsi zygotes that result from failed extrusion of the second polar body has been described by Tsuchiya et al. (12). Mechanistically speaking, this may occur with a damaged metaphase plate or oocyte cytoskeleton (13), after abnormal spindle formation or increased female age (12, 14). The latter etiologies may represent an occult oocyte factor (12, 14). The presence of a high percentage of 3PN zygotes after ICSI also may reflect a globally dysfunctional oocyte cohort. As seen in this report, there was an inverse correlation between percentage triploid and IVF outcomes. According to our findings, the greater the proportion of affected embryos, the worse the clinical outcome. Correspondingly, cycles with a high proportion of triploid zygotes (group 2) displayed lower pregnancy and live-birth rates despite having morphologically normal-appearing embryos for transfer. Our study suggests that the degree of triploidy after ICSI serves as a marker of oocyte cohort competence and as such, helps predict IVF outcome. Other investigators have suggested that IVF stimulation specific parameters, such as prolonged duration of stimulation, elevated peak E 2 levels, and increased number of oocytes, may induce the appearance of triploid zygotes (10). In contrast, we observed that couples with %20% triploid zygotes had higher E 2 levels at the time of hcg trigger (2,230 912 vs. 1,677 868 pg/ml) and had a higher number of oocytes (16.9 7.2 vs. 10.0 7.4), indicating that the presence of 3PNs was not associated with hyperresponse but may in fact be reflective of a compromised oocyte cohort. Rosen et al. (8), via statistical modeling, investigated the effects of cycle-specific parameters on the incidence of triploidy and concluded that both the dose and duration of exposure to gonadotropins are independent predictors of post-icsi triploidy. They further reported that implantation rates decreased when 3PN fertilization was noted (8). Our data complement and elaborate on these findings. Because we realized that 3PN fertilization after ICSI has been associated with a reduction in implantation rate (8), we sought to establish whether the degree of 3PN after ICSI fertilization predicts outcome. In doing so, we determined that the lowest degree of triploid fertilization beyond which IVF clinical outcomes were adversely affected was 20%. In our study, observations of traditional predictors of IVF outcomes were similar between groups, indicating that the triploidy rate was an independent and distinct predictor of IVF outcome. At a threshold level of 20% triploidy, we showed that the clinical-pregnancy rate is reduced by >50% when compared with the cycles with %20% triploidy. An even greater impact was seen regarding implantation rates and live-birth rates. These findings suggest that the proportion of triploidy may be a marker of the competence of the total oocyte cohort that is yielded during an IVF cycle. These findings therefore may influence recommendations on the number of embryos to transfer and on the utility of freezing supernumerary embryos. Clearly, before such recommendations can be made, further studies are needed to better elucidate both the causes and the effects of triploid fertilization on IVF outcome. REFERENCES 1. Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet 1978;2:366. 2. Hu Y, Maxson WS, Hoffman DI, Ory SJ, Eager S, Dupre J, et al. Maximizing pregnancy rates and limiting higher-order multiple conceptions by determining the optimal number of embryos to transfer based on quality. Fertil Steril 1998;69:650 7. 3. Hsu MI, Mayer J, Aronshon M, Lanzendorf S, Muasher S, Kolm P, et al. Embryo implantation in in vitro fertilization and intracytoplasmic sperm injection: impact of cleavage status, morphology grade, and number of embryos transferred. Fertil Steril 1999;72:679 85. 4. Boyers SP, Diamon MP, Lavy G, Russell JB, DeCherney AH. The effect of polyploidy on embryo cleavage after in vitro fertilization in human. Fertil Steril 1987;48:624 7. 5. Jun SH, O Leary T, Jackson KV, Racowsky C. Benefit of intracytoplasmic injection in patients with a high incidence of triploidy in a prior in vitro fertilization cycle. Fertil Steril 2006;86:825 9. 6. McFadden DE, Jiang R, Langloise S, Robinson WP. Dispermy origin of diandric triploidy. Hum Reprod 2002;17:3037 8. 7. Egozcue S, Blanco J, Vidal F, Egozcue J. Diploid sperm and the origin of triploidy. Hum Reprod 2002;17:5 7. 8. Rosen MP, Shen S, Dobson AT, Fujimoto VY, McCulloh CE, Cedars MI. Triploidy formation after intracytoplasmic sperm injection may be a surrogate marker for implantation. Fertil Steril 2006;85:384 90. 9. Macas E, Imthurn B, Keller PJ. Increased incidence of numerical chromosome abnormalities in spermatozoa injected into human oocytes by ICSI. Hum Reprod 2001;16:115 20. 10. Sachs AR, Politch JA, Jackson KV, Racowsky C, Hornstein MD, Ginsburg ES. Factors associated with the formation of triploid zygotes after intracytoplasmic sperm injection. Fertil Steril 2000;73: 1109 14. 104 Dayal et al. Triploidy after ICSI impacts IVF outcome Vol. 91, No. 1, January 2009
11. McFadden DE, Jiang R, Langloise S, Robinson WP. Dispermy origin of diandric triploidy: brief communication. Hum Reprod 2002;17: 3037 8. 12. Tsuchiya K, Kamiguchi Y, Sengoku K, Ishikawa M. A cytogenetic study of in-vitro matured murine oocytes after ICSI by human sperm. Hum Reprod 2002;17:420 5. 13. Kimura Y, Yanagimachi R. Intracytoplasmic sperm injection in the mouse. Biol Reprod 1995;52:709 20. 14. Spandorfer SD, Avrech OM, Colombero LT, Palermo GD, Rosenwaks Z. Effect of parental age on fertilization and pregnancy characteristics in couples treated by intracytoplasmic sperm injection. Hum Reprod 1998;13:334 8. Fertility and Sterility â 105