Very high serum estradiol levels are not detrimental to clinical outcome of in vitro fertilization

Transcription

1 FERTILITY AND STERILITY Copyright 990 The American Fertility Society Printed on acid-free paper in U.S.A. Very high serum estradiol levels are not detrimental to clinical outcome of in vitro fertilization Philip E. Chenette, M.D. *t Mark V. Sauer, M.D.*:j: RichardJ. Paulson, M.D.*:j: University of Southern California School of Medicine and California Medical Center, Los Angeles, California The use of gonadotropin-releasing hormone agonists as adjuncts to ovulation induction for in vitro fertilization (IVF) has resulted in increases in oocyte recovery rates. Along with increased oocyte number, greatly increased estradiol (E 2 ) levels have been found. We sought to determine the clinical effect of very high E 2 levels on the outcome of IVF cycles. Estradiol levels were measured in 4 patients undergoing controlled ovarian hyperstimulation with leuprolide acetate and human menopausal gonadotropin for IVF. Whereas the number of oocytes recovered and fertilized and the number of embryos available for cryopreservation were directly proportional to the E 2 level, the fertilization rate and embryo cleavage rates were unrelated to the E 2 level. When the patients were grouped in thirds according to E 2 levels, pregnancy rate (PR) was highest in the patients with the highest E 2 levels (E 2 > 2, 777 pg/ml, PR = 37% ). One mild, one moderate, and one severe case of ovarian hyperstimulation syndrome occurred in patients with E 2 2': 3,000 pg/ml (n = 2), but in general, high E 2 levels were attained with few complications. We conclude that high E 2 levels are not detrimental to the pregnancy outcome of IVF. Our experience further suggests that cycles with E 2 levels of s5,000 pg/ml need not be canceled and can proceed to oocyte recovery and embryo transfer. Fertil Steril 54:858, 990 Oocyte recovery for in vitro fertilization (IVF) has steadily improved as a result of changes in ovulation induction protocols. Pituitary down regulation with gonadotropin-releasing hormone agonist (GnRH-a) before administration of human menopausal gonadotropins (hmg) has improved follicle recruitment and has resulted in an increase in the numbers of oocytes retrieved. 2 3 Consequently, routine GnRH-a pretreatment before hmg stimulation has been advocated. 4 Nevertheless, increases in oocyte numbers have been accompanied by in- Received March 2, 990; revised and accepted July, 990. *Department of Obstetrics and Gynecology, Division of Reproductive Endocrinology, University of Southern California. t Present address: Department of Obstetrics and Gynecology, Northwestern University School of Medicine, Chicago, Illinois. :j: California Reproductive Health Institute, California Medical Genter. Reprint requests: Richard J. Paulson, M.D., 240 North Mission Road, Room L022, Los Angeles, California creases in other follicular products as well, and the effect of these secretory products on IVF outcome has not been determined. Our group has previously reported data suggesting that controlled ovarian hyperstimulation is associated with decreased embryo implantation in a mouse embryo donation model 5 as well as human IVF. 6 Because estradiol (E 2) has been used successfully as an interceptive agent, 7 and because controlled ovarian hyperstimulation is associated with high E 2 levels, it has been suggested that IVF cycles with high E 2 levels would have decreased success because of abnormal embryo implantation. Additionally, concerns about poor embryo quality and ovarian hyperstimulation syndrome have led some to advocate cancellation or early administration of human chorionic gonadotropin (hcg) in cycles with high E 2 4 However, along with increased E 2, other secretory products are also increased; the clinical significance of high E 2 levels in IVF cycles has not been adequately assessed. 858 Chenette et al. Effect of high E 2 on IVF outcome Fertility and Sterility

2 The purpose of this study was to examine the effect of high serum E 2 levels on oocyte retrieval, fertilization, embryo cleavage, and pregnancy outcome in IVF cycles stimulated with GnRH-a and hmg. Furthermore, we wished to ascertain whether cycles with exaggerated E 2 levels should be canceled or avoided either because of poor outcome or the high risk of ovarian hyperstimulation syndrome. MATERIALS AND METHODS Between July 987 and April989, 4 patients underwent oocyte retrieval for IVF after controlled ovarian hyperstimulation with GnRH-a and hmg. The GnRH -a leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, North Chicago, IL) was started in the midluteal phase at a dose of mg/d subcutaneous (SC) and continued for 2 weeks or until serum E 2 levels were <30 pg/ml and no follicular activity was noted by transvaginal ultrasound (US) examination. At that time, hmg (Pergonal; Serono, Randolph, MA) was started in an initial dose of three to four ampules per day, and the LA dose decreased to 0.5 mg/d SC. The hmg dose was titrated to the ovarian response, and hmg was continued with daily transvaginal US and serum E 2 determination until the lead follicles achieved diameters of 8 to 22 mm and E 2 levels reached 0 pg/ml per codominant follicle. Human chorionic gonadotropin, 0,000 IU intramuscular, was then given and 36 hours later transvaginal US-guided oocyte retrieval was performed under intravenous sedation. All oocytes were inseminated and embryo culture continued for 48 hours as previously described. Up to 5 cleaving embryos were transferred at one time. Standard luteal phase support in the form ofim progesterone in oil (25 mg/d from day of transfer) was given. Nine and 4 days after embryo transfer (ET), serum {3-hCG levels were determined; US was performed at 5 and 8 weeks after ET to confirm gestation. Cases of ovarian hyperstimulation were classified according to Rabau et al. 8 Clinical pregnancy was defined as detection of a sac on US in conjunction with rising {3-hCG levels. Ongoing pregnancy was defined as a visible heartbeat within the fetus 8 weeks after ET. Biochemical pregnancies were not included in the analysis of pregnancy rates (PRs). Serum E 2 and {3-hCG were determined using radioimmunoassay (Tandem; Hybritech, San Diego, CA). Intra-assay and interassay variability was 0% and 8%, respectively, for E 2 and 7% and 5% for {3-hCG. Serum E 2 levels and oocyte and embryo numbers were compared using linear regression, and pregnancy outcome compared using x 2 analysis and Fisher's exact test. RESULTS In the study group (n = 4), maximum E 2 (E 2 max) level occurred on the day after hcg administration in 84.3% and on the day ofhcg administration in 5.6%. Estradiol max values showed a lognormal distribution and ranged from 282 to 2,534 pg/ml with a mean of,908 ± 22 (mean ± SEM) pg/ml. Twenty-one patients had E 2 max value ;;::: 3,000 pg/ml. The number of oocytes retrieved per cycle correlated with E 2 max values (Fig. A, r = 0.855, P < 0.0). The number of oocytes that fertilized also correlated with the E 2 max value (Fig. B, r = 0.76, P < 0.0). The fertilization rate was unrelated to the E 2 max value (Fig. C, r = 0.045), as was the percentage of oocytes that cleaved (Fig. D, r = 0.000). Maximum E 2 values correlated with the number of pre-embryos frozen in each cycle, (Fig. E, r = 0.570, P < 0.0). When divided into percentile ranges, an increase in PR was observed in patients who developed the highest E 2 levels (Fig. 2). To control for the number of embryos transferred, because patients with high E 2 levels tended to have more embryos, the group of patients who had five embryos transferred (n = 64) was studied separately in analyzing pregnancy outcome. In this group, the highest third of E 2 max values (>2,777 pg/ml) was associated with the highest clinical PR (PR = 45%, P = 0.02) compared with the rate in patients with E 2 max in the mid (,722 to 2,777 pg/ml, PR = 24%) and lower ( <,722 pg/ml, PR = 4%, not significant) thirds. In the group of patients with E 2 > 3,000 pg/ml (n = 2), all but 3 had five embryos transferred. The three who did not had severe male factor; two of these had no embryos, the other had one embryo and did not conceive. Thus to control for the number of embryos transferred, cycles with five embryos transferred were also examined and grouped according to E 2 max by thousands (Table ). The highest ongoing PR occurred in cycles with E 2 max of 3,000 to 4,000 pg/ml (n =, PR = 45%). One case of moderate ovarian hyperstimulation syndrome occurred in a patient who achieved pregnancy; she was managed with bed rest. The next highest ongoing PR was achieved in the 4,000 to 5,000 pg/ml group (n = 4, PR = 25%). No ongoing Chenette et al. Effect of high E 2 on IVF outcome 859

3 A Oocytes Retrieved B Fertilized 0 c 00.me % Oocytes Fertilized D %Cleaved 40 lldiiid [J D E Embryos Frozen 0 I E2 (pg/ml) Figure Linear regression of E 2 max and the following parameters: (A), oocytes retrieved at aspiration; (B), number of embryos obtained; (C), fertilization rate; (D), embryo cleavage rate; and (E), number of embryos frozen. pregnancies occurred in patients with E 2 levels > 5,000 pg/ml. The outcome of cycles with E 2 max > 4,000 pg/ml is detailed in Table 2. As noted, two instances of ovarian hyperstimulation syndrome occurred in this group. One was moderate and was managed with bed rest and fluids. The other developed severe ovarian hyperstimulation syndrome. This particular patient had polycystic ovarian syndrome and was essentially amenorrheic. After GnRH-a and hmg, E 2 levels rose rapidly and reached 5,7 pg/ml on the loth day of hmg therapy. Even though follicle size criteria for maturity were only marginally satisfied, hcg was given, and the E 2 rose to 2,534 pg/ml on the following day Pregnancy 30 Rate (%) All Patients * 5 Embryos Transferred < n >2m Figure 2 Pregnancy occurrence by E 2 max range in all patients and those who had five embryos transferred, grouped by E 2 percentile ranges divided into thirds. The lower third encompasses E 2s of <, 722 pg/ml, middle third, 722 to 2, 777 pg/ml, and upper third> 2,777 pg/ml. *,Significant difference from clinical pregnancy rate of other groups, P = * Extreme enlargement of the ovaries made the aspiration of all follicles technically impossible, nonetheless, 58 oocytes were identified. Four days after retrieval (6 days after hcg administration), the patient required hospitalization for management of ascites and enlarged ovaries. She subsequently recovered completely. DISCUSSION Enhanced oocyte recovery has resulted from improvements in the technique of controlled ovarian hyperstimulation. Down regulation with GnRH -a results in greater follicular recruitment and prevents the occurrence of luteinizing hormone surges, permitting greater ovarian stimulation before follicle aspiration. This greater degree of ovarian response results in an increase in other follicular secretory products, and although the effects of several of these secretory products on the endometrium individually have been studied, there is little information on the overall effect of high E 2 levels on PRs in IVF. We found that the absolute number of oocytes obtained was proportional to the E 2 max, as was the total number of fertilized embryos. However, the oocyte fertilization rate and embryo cleavage rate did not vary with the E 2 max, suggesting that oocyte and embryo quality are independent of the E 2 level. These findings are consistent with the re- 860 Chenette et al. Effect of high E 2 on IVF outcome Fertility and Sterility

4 Table Pregnancy Outcome in Patients With Five Embryos Transferred" No. of patients No. of clinical pregnancy No. of No. of No. of ovarian term spontaneous hyperstimulation pregnancy abortion syndrome E 2 range (pg/ml) Oto (8)b ()b 000to 00 8 (28) 3 (7) 00to (34) 6 (27) 3000to 4000 (7) 5 (45) 4000to (6) 2 (50) 5000 to (5) 0 (0) ~6000 () 0 (0) Total 64 (00) 7 (27) a Grouped by E 2 level in the,000s. 0 (O)b ()b 0 2 () (6) 0 5 (23) (5) 0 5 (45) 0 (0) moderate (25) (25) mild 0 (0) 0 (0) 0 0 (0) 0 (0) severe 3 () 4 (6) 3 b Values in parentheses are percents. sults of Levran et al. 9 who found oocyte retrieval and fertilization to be proportional to E2 levels. Even abnormal changes in E 2 may not affect oocytes and embryos because Diamond et al. 0 found no difference in fertilization or cleavage rates when an abnormal E 2 response occurred. Thus oocytes do not appear to be affected by the absolute E 2 level per se. The presumption of an adverse effect of high E 2 on endometrial receptivity originates in the use of estrogen (E) as an interceptive agent. However, there are several difficulties with using exogenously administered E as a model for ovarian hyperstimulation. Oral and injectable Es act by multiple mechanisms, such as inducing premature luteolysis, inhibiting tubal motility, and altering tubal mucus, in addition to affecting endometrial histology, biochemistry, and receptivityy- 3 Effective interception with E 2 alone requires large doses of ethinyl E2, resulting in serum levels as high as 0,000 pg/ml. 4 In addition, the endogenous secretion of multiple hormones that occurs during controlled ovarian hyperstimulation is different from exogenous administration of E alone. There does exist indirect evidence that controlled ovarian hyperstimulation has an adverse effect on fertility, however. Our group has previously shown that in a mouse embryo donation model fewer embryos implant in uteri of mice stimulated with gonadotropins.5 Additionally, by comparing embryo implantation data in oocyte donation and standard IVF cycles, we have shown that controlled ovarian hyperstimulation is associated with up to a 60% decrease in endometrial receptivity. 6 In the human, scanning electron microscopy has revealed endometrial abnormalities in uteri of stimulated cycles/5 and exogenous E 2 has also been demonstrated to be luteolytic. 6 The precise clinical effect of these alterations has been only partially defined, though levels of E2 were higher among cycles of IVF without conception in one studyp Advanced endometrial maturation has been demonstrated in stim- Table 2 Pregnancy Outcome in Patients With High Levels of E 2 " No. of oocytes No. fertilized No. cleaved Pregnancy Complications Patient ,278 4,359 4,470 4,695 5,025 5,30 5,78 2, Biochemical Spontaneous abortion Term Biochemical Mild ovarian hyperstimulation syndrome Severe ovarian hyperstimulation syndrome c a E 2 max of >4,000 pg/ml. b Zygotes with 3 pronuclei. c Because of technical reasons, only follicles from one ovary and a portion of the second ovary were able to be aspirated. Approximately follicles were not punctured. Chenette et al. Effect of high E 2 on IVF outcome 86

5 ulated cycles, but this appeared to improve pregnancy outcome rather than diminish it. 8 The composite effect of these changes on human implantation remains unexplored. We noted a high PR in our patients with very high E 2 levels. Among patients who received five embryos at transfer, the best PR was achieved in the moderately high percentiles of E 2 max. Pregnancy has previously been shown to occur even when a patient develops high E 2 values. Jones et al. 9 were the first to report that IVF PRs were correlated with certain patterns of E 2 rise. However, they did not report the actual range of E 2 values encountered. Okamoto et al. developed 95% confidence limits for E 2 in patients who conceived with IVF. Despite their suggestion that there is a normal range for E 2, nearly 5% of the patients who conceived had at least one E 2 above the 95th percentile during their cycle, and eliminating these patients would have significantly reduced their PR. Haning et al. 2 were unable to show a correlation between E 2 levels and pregnancy or implantation rate by more sophisticated mathematical techniques, though the maximum E 2 encountered by them is not clear. Forman et al. 22 demonstrated a reduced PR in patients with moderately elevated E 2 and postulated an antinidatory effect of E, but inhibition of implantation was inconsistent and occurred only when one or two embryos were transferred. Dor et al. 23 in comparing conception and matched nonconception cycles, found no difference in E 2 levels, though a fall in E 2 levels occurred more frequently in nonconception groups. The notion of high E 2 inhibiting pregnancy during IVF is thus supported only by indirect evidence. Meldrum et al. 4 observed a high rate of oocyte retrieval and pregnancy in patients with moderately high E 2 levels using GnRH-a down regulation; however, they canceled cycles where E 2 rose above 3,000 pg/ml before achieving follicle maturity. 4 Our results confirm these findings and suggest that excellent PRs occur in cycles with even higher E 2 The safety of IVF in patients with high E 2 levels is an area of significant concern. In our series of 4 women, one patient experienced mild and one patient moderate ovarian hyperstimulation syndrome. Only one experienced severe hyperstimulation (0.7%), a rate that is well within the range for hmg ovulation induction as reviewed by Schenker and Weinstein. 24 Diamond et al. 25 in a review of IVF, cycles with high E 2 values reported no instances of ovarian hyperstimulation syndrome requiring hospitalization, though they did not detail how high E 2 values rose. It has been suggested that follicle aspiration may protect against hyperstimulation syndrome by reducing the volume of tissue in the luteal ovary. Our study can neither confirm nor refute this suggestion; however, our one patient with severe ovarian hyperstimulation syndrome did not have most of the follicles in one ovary aspirated because of technical difficulties. In summary, we found that patients with high E 2 max levels had both excellent oocyte recovery and absolute numbers of embryos. High E 2 max levels were not detrimental to oocyte fertilization, embryo cleavage, or implantation. The incidence of ovarian hyperstimulation syndrome was low. Our experience suggests that IVF cycles with E 2 levels of ::;5,000 pg/ml need not be canceled when maturity criteria are satisfied. We conclude that among IVF patients undergoing controlled ovarian hyperstimulation with GnRH-a and hmg stimulation the PR is not impaired in cycles with high E 2 levels. REFERENCES. Paulson RJ, Marrs RP: Ovulation stimulation and monitoring for in vitro fertilization. Curr Probl Obstet Gynecol Fertil 9:50, Porter RN, Smith W, Craft JL, Abdulwahid NA, Jacobs HS: Induction of ovulation for in-vitro fertilization using buserelin and gonadotropins. Lancet 2:284, Neveu S, Hedon B, Bringer J, Chinchole J-M, Arnal F, Humeau C, Cristo! P, VialaJ-L: Ovarian stimulation by a combination of a gonadotropin-releasing hormone agonist and gonadotropins for in vitro fertilization. Fertil Steril 4 7:639, Meldrum DR, Wisot A, Hamilton F, Gutlay AL, Kempton W, Huynh D: Routine pituitary suppression with leuprolide before ovarian stimulation for oocyte retrieval. Fertil Steril 5:455, Fossum GT, Davidson A, Paulson RJ: Ovarian hyperstimulation prevents embryo implantation in the mouse. J In Vitro Fert Embryo Transfer 6:7, Paulson RJ, Sauer MV, Lobo RA: Embryo implantation after human in vitro fertilization: importance of endometrial receptivity. Fertil Steril53:870, Morris JM, van W agenen G: Interception: the use of postovulatory estrogens to prevent implantation. Am J Obstet Gynecol5:0, Rabau E, David A, Serr DM, Mashiach S, Lunenfeld B: Human menopausal gonadotropins for anovulation and sterility: results of 7 years of treatment. Am J Obstet Gynecol 98:92, Levran D, Lopata A, Nayudu PL, Martin MJ, McBain JC, Bayly CM, Speirs AL, Johnston WIH: Analysis of the outcome of in vitro fertilization in relation to the timing of human chorionic gonadotropin administration by the duration of estradiol rise in stimulated cycles. Fertil Steril 44: 335, Diamond MP, Webster BW, Garner CH, Vaughn WK, Maxson WS, Herbert CM, Osteen KG, Rogers BJ, Wentz 862 Chenette et al. Effect of high E 2 on IVF outcome Fertility and Sterility

6 AC: Selection of superior stimulation protocols for follicular development in a program of in vitro fertilization. Fertil Steril43:25, 985. Morris JM: Mechanisms involved in progesterone contraception and estrogen interception. Am J Obstet Gynecol 7:67, Pauerstein CJ, Fremming BD, Martin JE: Estrogen-induced tubal arrest of ovum: antagonism by alpha adrenergic blockade. Obstet Gynecol35:67, Ling WY, Robichaud A, Zayid I, Wrixon W, MacLeod SC: Mode of action of DL-norgestrel and ethinylestradiol combination in postcoital contraception. Fertil Steril 32:297, Back DJ, Bolt HM, Breckenridge AM, Crawford FE, Orme MLE, Rowe PH, Schindler AE: The pharmacokinetics of a large (3 mg) oral dose of ethinyl estradiol in women. Contraception 2:45, Martel D, Frydman R, Glissant M, Maggioni C, Roche D, Psychoyos A: Scanning electron microscopy of postovulatory human endometrium in spontaneous and stimulated cycles. J Endocrinol4:39, Schoonmaker JN, Bergman KS, Steiner RA, Karsch FJ: Estradiol-induced luteal regression in the rhesus monkey: evidence for an extraovarian site of action. Endocrinology 0:708, Gidley-Baird AA, O'Neill C, Sinosich MJ, Porter RN, Pike IL, Saunders DM: Failure of implantation in human in vitro fertilization and embryo transfer patients: the effects of altered progesterone/estrogen ratios in humans and mice. Fertil Steril 45:69, Garcia JE, Acosta AA, Hsiu J-G, Hones HW, Jr: Advanced endometrial maturation after ovulation induction with human menopausal gonadotropin/human chorionic gonadotropin for in vitro fertilization. Fertil Steril4:3, Jones HW, Jr, Acosta A, Andrews MC, Garcia JE, Jones GS, Mantzavinos T, McDowell J, Sandow B, Veeck L, Whibley T, Wilkes C, Wright G: The importance of the follicular phase to success and failure in in vitro fertilization. Fertil Steril40:37, 983. Okamoto S, Healy DL, Howlett DT, Rogers PAW, Leeton JF, Trounson AL, Wood EC: An analysis of plasma estradiol concentrations during clomiphene citrate-human menopausal gonadotropin stimulation in an in vitro fertilization -embryo transfer program. J Clin Endocrinol Metab 63:736, Haning RV, Jr, Boehnlein LM, Carlson IH, Zweibel WJ: Diagnosis-specific serum 7!'l-estradiol (E 2) upper limits for treatment with menotropins using a 25 I direct E2 assay. Fertil Steril42:882, Forman R, Fries N, Testart J, Belaisch-Allart J, Hazout A, Frydman R: Evidence for and adverse effect of elevated serum estradiol concentrations on embryo implantation. Fertil Steril49:8, Dor J, Rudak E, Mashiach S, Nebel L, Serr DM, Goldman B: Periovulatory 7!'l-estradiol changes and embryo morphologic features in conception and nonconceptional cycles after human in vitro fertilization. Fertil Steril45:63, Schenker JG, Weinstein D: Ovarian hyperstimulation syndrome: a current survey. Fertil Steril30:255, Diamond MP, Buchholz T, Boyers SP, Lavy G, Shapiro BS, DeCherney AH: Super high estradiol response to gonadotropin stimulation in patients undergoing in vitro fertilization. J In Vitro Fert Embryo Transfer 6:8, 989 Chenette et al. Effect of high E 2 on IVF outcome 863