Significance of basal follicle-stimulating hormone levels in women with one ovary in a program of in vitro fertilization*

Transcription

1 FERTILITY AND STERILITY Copyright e 1992 The American Fertility Society Printed on acid-free paper in U.S.A. Significance of basal follicle-stimulating hormone levels in women with one ovary in a program of in vitro fertilization* Emad Khalifa, M_D. James P. Toner, M.D., Ph.D.t Suheil J. Muasher, M.D. Anibal A. Acosta, M.D. The Jones Institute for Reproductive Medicine, Department of Obstetrics and Gynecology, Eastern Virginia Medical School, Norfolk, Virginia Objective: To understand the impact of having a single ovary on basal follicle-stimulating hormone (FSH) level and its diagnostic and prognostic usefulness in in vitro fertilization (IVF). Design: All IVF cases from July 1987 to June 1990 with known basal FSH (n = 1,272) were divided into those with one and those with two ovaries to compare outcomes based on basal FSH levels. Setting: Tertiary care academic center with a large IVF practice. Main Outcome Measures: Basal FSH, age, and IVF outcomes including peak estradiol, numbers of follicles aspirated, oocytes retrieved, fertilized, and transferred, and pregnancies (clinical and ongoing). Results: In women with only one ovary, basal FSH was increased, and IVF outcomes were poorer. The rise in FSH was able, in large part, to account for the diminished performance in the single ovary cases. Conclusions: Women with only one ovary have higher basal FSH levels than those with two ovaries, and this rise can be used to predict their IVF performance. Fertil Steril 1992;57:835-9 Key Words: Ovary, oophorectomy, basal follicle-stimulating hormone/in vitro fertilization Many women participating in in vitro fertilization (IVF) programs have significant pelvic disease; a number have had unilateral oophorectomy (with or without ovarian cystectomy in the remaining ovary) (1). The impact of having a single ovary on IVF outcome has been investigated by a number of authors. Most of these studies compare cancellation rates, peak estradiol (E 2) levels, number of preovulatory oocytes retrieved, fertilized, and transferred, and pregnancy rates (PRs) in normal women versus Received September 20, 1991; revised and accepted December 3,1991. * Presented at the 47th Annual Meeting of The American Fertility Society, Orlando, Florida, October 21 to 24, t Reprint requests: James P. Toner, M.D., Ph.D., Jones Institute for Reproductive Medicine, Department of Obstetrics and Gynecology, Eastern Virginia Medical School, Norfolk, Virginia, patients with unilateral oophorectomy (2-4). Our own results in the Norfolk IVF program have been published (5), but were based on stimulations before gonadotropin-releasing hormone agonists were in common use. In general, these reports show trends toward poorer performance in the group with one ovary, but small sample size has limited the significance of these observations. Animal studies show that there is compensatory ovarian hypertrophy of the remaining ovary after unilateral oophorectomy (6). This compensatory hypertrophy might be attributed to the loss of negative feedback on the pituitary, resulting in an increase of follicle-stimulating hormone (FSH) secretion. This reduced ovarian feedback may involve steroidal and/or nonsteroidal factors (7-9). Serum levels of FSH increase as ovarian function declines (10), an increase that can be detected years Khalifa et al. Basal FSH in cases of one ovary 835

2 before menopause (11). The value of basal (day 3) FSH levels during the recruitment phase as a predictor of ovarian reserve and IVF results has recently been established, and useful cutoffs have been established (12). The impact of unilateral oophorectomy on basal FSH levels in women undergoing IVF has not been directly investigated. Further, if oophorectomy alters FSH levels, it is not clear how to interpret such alterations. Are they still predictive of IVF performance? Are the cutoffs for normal and poor IVF performance different from women with two ovaries? These were the issues addressed in the current study. Further, because we were able to study a larger number of patients and cycles than previously reported, our study had more power to detect differences in the overall IVF performance of women with one versus two ovaries. Subjects MATERIALS AND METHODS One thousand two hundred seventy-two consecutive IVF cycles in which basal FSH values were known (Norfolk series 28 to 39; July 1987 to June 1990) were retrospectively studied; there were 162 cycles in women with one ovary and 1,110 cycles in patients with two ovaries. All male factor cases (concentration <20 X 1Q6/mL, motility <30%, and/ or morphology <4% normal forms by strict morphological criteria) (13) have been excluded from the study. Patients with cycle day 3 E2 levels> 70 pg/ml were also excluded (9 cycles in the I-ovary group [5.6%] and 44 in the 2-ovary group [3.9%]) because prematurely elevated E2 values suppress basal FSH levels and indicate a premature follicular recruitment common in the perimenopause. Another 13 cycles (8.5%) in the one-ovary group and 81 cycles (7.6%) in the two-ovaries group were canceled before oocyte retrieval because of inadequate response. Canceled cycles were excluded from the statistical analysis, leaving 140 cycles in the one-ovary group and 985 cycles in the two-ovaries group appropriate for statistical analysis. In 5 cycles (3.3%) in the oneovary group and 30 cycles (3.1 %) in the two-ovaries group, despite follicle aspiration, no oocytes were retrieved. Additionally, transfer was not performed in 18 cycles (11.8%) in the one-ovary group and in 76 cycles of the two-ovaries group (7.7%) (P < 0.02) because of failed or abnormal fertilization or degeneration of fertilized oocytes. Indications for unilateral oophorectomy group were: ovarian endometriosis in 72 cases, pelvic in- 836 Khalifa et al. Basal FSH in cases of one ovary flammatory disease in 34, benign ovarian cysts in 24, and ectopic pregnancy in 23. In addition, 14 patients had ovarian cystectomy performed in the remaining ovary. Because these 14 cases showed no statistically significant differences in any parameter when compared with the rest of the single ovary group, they were not excluded from the evaluation. Basal FSH values from cycle day 3 in the cycle of stimulation were used for analysis except when leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, Abbott Park, IL) pretreatment had been employed; in such cases, a basal FSH value from a previous cycle within 3 months of the IVF attempt was applied. This FSH radioimmunoassay kit (Leeco Diagnostics, Inc, Southfield, MI) uses the 2nd International Reference Preparation as its standard and had 9.1 % total assay variation. The E2 kit (direct method; Pantex, Inc, Santa Monica, CA) had a total assay variation of 9.2%. Protocol for IVF In most cases, gonadotropins were used alone (human menopausal gonadotropin or FSH [Pergonal and Metrodin; Serono Laboratories, Randolph, MA], alone or in combination intramuscularly [1M]), beginning on the 3rd day of the menstrual cycle. When LA was used adjunctively, it was begun either on the 2nd day of the menstrual cycle in a "flare-up" protocol (follicular LA) (1 mg subcutaneously [SC] days 2 to 5 then 0.5 mg/d thereafter), or in the midluteal phase of the preceding cycle in a suppression protocol (luteal LA) (0.5 mg/d SC). In both protocols LA was continued until the day of human chorionic gonadotropin (10,000 IV; Steris Laboratories, Phoenix, AZ) administration. Leuprolide acetate was used in 28.8% of cycles in the one-ovary group (26 cycles [16.9%] as follicular LA and 18 [11.8%] as luteal LA) and in 38% of cycles in the two-ovary group (all cases as luteal LA). Human chorionic gonadotropin was administered 1M when two or more follicles were 16 mm in diameter. Transvaginal follicular aspiration was performed 34 to 36 hours.later. Gamete processing, cell culture technique, and transfer procedures were as previously described (14). Oocyte maturity was classified by the criteria of Veeck et al (15). Transfer of four or fewer pre-embryos from preovulatory eggs was performed 2 days later. Pre-embryos in excess of four were cryopreserved. Progesterone (Schein Pharmaceuticals Incorporation, Port Washington, NY) supplementation (25 mg/d 1M) was administered during the luteal phase. If pregnancy was confirmed, 17-hydroxyprogesterone (Steris Laboratories) 250 mg/wk 1M was administered until the 14th Fertility and Sterility

3 ovary 2 ovaries FSH.25 o FSH 20- ~ FSH 15- o FSH 10- II!!I FSH.10 regression analyses in which the FSH levels were considered before evaluating any residual effect of ovary number. When controlling for the FSH level eliminated the importance (i.e., significant difference) of the one- versus two-ovary distinction, we concluded that the rise in FSH was a sufficient predictor of that IVF outcome in cases with one ovary. Graphically, this control for FSH has been illustrated by stratification of cases with one and two ovaries according to the basal FSH level «10, 10 to 14, 15 to 19, 20 to 24, >24 lull). When the oneand two-ovary cases showed the same performance at each level of FSH, we affirmed that the FSH level in one-ovary cases can be used to predict IVF outcome just as it is used in cases with two ovaries. Figure 1 Distribution of basal FSH levels according to the number of ovaries. week of pregnancy. Clinical pregnancies were identified by visualizing fetal heart motion on ultrasound. Chemical pregnancies were excluded from all analyses, and PRs were based on the number of aspiration attempts. Statistics Overall differences between single and double ovary cases were evaluated by unpaired, two-tail t tests. To understand whether the increased basal FSH in single ovary cases was an accurate predictor of the diminished IVF performance, the effect of altered FSH levels in the one-ovary cases on IVF outcome was controlled in a series of stepwise RESULTS The basal FSH levels in women with single ovaries (17.3 ± 9.1 lull) were significantly higher than in women with two ovaries (12.1 ± 3.3 lull) (P < ). Figure 1 shows the distribution of basal FSH levels in both conditions. More women with one ovary had high basal FSH levels (> 15 lull) (12) than women with two ovaries (54.9% of I-ovary cases versus 24.9% of 2-ovary cases; P < 0.001). The higher FSH levels in women with one ovary were not because of higher age; there was no difference in ages (34.6 ± 3.9 in I-ovary cases versus 34.9 ± 3.5 in 2-ovary cases; P = 0.26). The IVF performance in cases with one ovary was in most respects poorer than that observed in women with two ovaries (Table 1). All measures of ovarian responsiveness (peak E 2, number of total or mature Table 1 IVF Outcomes in Cases With One and Two Ovaries * IVF outcome 1 ovary Probability 1 versus 2 OvarY no. FSH after 2 ovaries ovariest after FSH:j: ovary no. Cancellation rate (%) 8.5 Peak E2 (pg/ml) 456 ± 3431f Follicles aspirated 5.9 ± 4.2 Oocytes inseminated 4.0± 3.0 Oocytes fertilized 3.2 ± 2.5 Metaphase eggs retrieved 4.3 ± 2.9 Metaphase eggs fertilized 3.8± 2.9 Pre-embrYos transferred 2.4 ± 1.5 Clinical pregnancy/aspiration (%) 17 Ongoing pregnancy/aspiration (%) NSII NS NS 657 ± ± ± NS ± NS ± NS ± NS ± NS NS NS NS * Male factor cases excluded. t Independent two-tailed t-tests comparing overall 1- and 2- ovary conditions. :j: Stepwise regression analyses evaluating role of ovary number in IVF outcomes after controlling for influence of basal FSH levels. Stepwise regression analyses evaluating role of basal FSH number in IVF outcome after controlling for influence of ovary number. II NS, not significant. 1f Values are means ± SD. Khalifa et al. Basal FSH in cases of one ovary 837

4 eggs retrieved or fertilized, number of pre-embryos transferred) showed statistically lower outcomes in cases with only one ovary. This resulted in lower PRs; the difference was statistically significant for clinical PR but not for ongoing PRs (though 5% fewer ongoing pregnancies were observed). None of the 19 women with one ovary and basal FSH level >20 IV/L had an ongoing pregnancy. To understand whether the decreased IVF performance could be estimated from the increased basal FSH that followed oophorectomy, a series of stepwise regression analyses was performed in which ovary number and basal FSH were considered simultaneously. According to these analyses, the difference in performance between one and two ovaries was largely explained by their different FSH levels. As seen in Table 1, all outcomes related to oocyte numbers and PRs could be explained based solely on their FSH differences (see last column); the independent effect of ovary number was insignificant for these outcomes (see second to last column). Only for the peak E2 and the number of follicles aspirated was ovary number an important variable after accounting for basal FSH levels. Figure 2 illustrates the relationships between ovary number and FSH for peak E2 (Fig. 2A), the number of preovulatory oocytes retrieved (Fig. 2B), and ongoing PRs (Fig. 2C). Miscarriage rates did not differ between conditions. Not surprisingly, many measures of IVF performance declined as basal FSH rose, whether the patient had one or two ovaries. In single ovary cases, the peak E2 for FSH < 10 IV /L was significantly higher compared with all other FSH groups, which, in turn, were not different from one another. The number of preovulatory oocytes retrieved in women with one ovary was higher for FSH < 15 IV /L versus FSH > 24 IV /L. Follicle-stimulating hormone> 24 IV /L was also a sign of fewer preovulatory oocytes fertilized and fewer pre-embryos transferred in cases with one ovary versus other FSH groups. DISCUSSION The value of basal FSH as a predictor of IVF performance is clear (16, 17). Elevated basal FSH levels on cycle day 3 (> 15 IV /L and especially ;:::0:25 IV /L) are associated with poorer IVF outcomes, from peak E2 to oocytes retrieved and fertilized to pregnancy outcomes (12, 17). The current study examined the effect of oophorectomy on basal FSH levels and IVF outcomes and explored the inter-relationship between FSH and oophorectomy as predictors of IVF performance. We A 800 =~. 1 U B Metapha.. Oocytea Retrieved c Ongoing Pregnancy Rate ('10) ~ (lull ~~ \11] IlIlIJ Il ~11l ~ IJ 0 < Bua) FSH (lull) Figure 2 In vitro fertilization outcome according to number of ovaries and basal FSH levels. (A), Peak E 2 levels. (B) Number of metaphase oocytes retrieved. (C), Ongoing PRs. suspected that FSH would be higher after oophorectomy but were uncertain whether this elevation would accurately reflect the diminished IVF performance. All relationships were considered possible. Some believed an FSH rise would overestimate the compromise in IVF performance because those with one ovary might enter IVF at a younger age and may have fewer but healthier eggs. Some felt an FSH rise would underestimate the compromised IVF performance because there is often significant pelvic disease surrounding or involving the remaining ovary, which consequently may not perform well. Last, some guessed FSH would rise in proportion to the actual compromise. The last hypothesis appears to be closest to the truth. Our results indicate that the basal FSH levels are indeed higher in patients with unilateral oophorectomy compared with patients with two ovaries. Importantly, this was not an age effect because women with single ovaries tended to be younger than those with two ovaries. Several investigators have studied the impact of unilateral oophorectomy on the ovarian response and pregnancy outcome in IVF; the cancellation rates (4), the peak E2 levels (5, 18), the number of follicles aspirated (5), the number of oocytes retrieved (4, 5, 18), and the number of pre-embryos transferred (3, 5) were all adversely affected by removing one ovary. However, in none of the earlier studies was the pregnancy outcomes statistically different when compared with women with two ovaries. Because of the larger number of IVF cycles examined in the current work, we have been able to demonstrate statistically fewer clinical pregnancies per 838 Khalifa et al. Basal FSH in cases of one ovary Fertility and Sterility

5 T aspiration attempt in women with a single ovary. Ongoing PRs, though 5% lower in the one-ovary group, were not statistically different; a difference this small would require nearly 800 pregnancies to achieve a power of 80%. When the difference in basal FSH levels was taken into account when comparing one- and two ovary conditions, in general the differences between one and two ovaries vanished. This indicates that the rise in FSH levels accurately predicts the decline in IVF performance. Thus the FSH rise is neither an overestimation or underestimation of IVF performance. Conversely, the value of FSH could not be explained by the number of ovaries. The only exceptions to this general rule were the peak E2 levels and the number of follicles aspirated, which showed a dependence on both basal FSH and ovary number. Clinically, this means that one can use the basal FSH in women with one or two ovaries in exactly the same manner: performance declines as FSH rises > 15 IU jl and falls off precipitously with elevations >25 IUjL. Though women with one ovary are certainly more likely to be in the higher FSH brackets, their performance at a given level of FSH is remarkably similar to the two-ovary performance at that same FSH level. To date, no ongoing pregnancies in the single ovary cases have been observed with FSH >20 IU jl, but the numbers of cases are relatively small, so we are unwilling to emphasize this as an important distinction. In vitro fertilization performance in both one- and two-ovary conditions declined as FSH levels rise. Though not surprising, this is the first report of this effect for the one- and two-ovary conditions separately. The present study differs from prior reports of single ovary performance in IVF because of its large size and the inclusion of basal FSH as a prospective marker of performance. Single ovary performance was shown to be poorer than that observed when two ovaries are present. Further, the clearly higher FSH levels in those with only one ovary was a good predictor of the degree of diminished IVF performance and will surely have value to clinicians who must advise patients with single ovaries of their real chances of success in IVF. Acknowledgment. We thank Ms. Debi Jones for providing the data analyzed. REFERENCES 1. Laufer N, Tarlatzis BC, Naftolin F. In vitro fertilization: state of the art. Semin Reprod Med 1984;2: Alper MM, Seibel MM, Oskowitz SP, Smith BD, Ransil BJ, Taymor ML. Comparison of follicular response in patients with one or two ovaries in a program of in vitro fertilization. Fertil Steril 1985;44: Dodds WG, Chin N, Awadalla SG, Miller F, Friedman C, Kim M. In vitro fertilization and embryo transfer in patients with one ovary. Fertil Steril1987;48: Dodson MG, Young RL, Poindexter AN, Gibbons WE, Rossavik IK, Findley WE. Comparison of in vitro fertilization results in women with one or two ovaries. J Reprod Med 1987;32: Boutteville C, Muasher SJ, Acosta AA, Jones HW Jr, Rosenwaks Z. Results of in vitro fertilization attempts in patients with one or two ovaries. Fertil Steril1987;47: Biggers JD, Finn CA, McLaren A. Long-term reproductive performance of female mice. I. Effect of removing one ovary. J Reprod Fertil1962;3: Gibson WR, Ingram BW, Lee VWK. Can reduced consumption of gonadotropins account for ovarian compensation in unilateral ovariectomized, immature mice injected with gonadotropins? J Reprod FertiI1979;57: Butcher RL. Changes in gonadotropins and steroids associated with unilateral ovariectomy of the rat. Endocrinology 1977;101: Cameron IT, O'Shea FC, Rolland JM, Hughes EG, dekretser DM, Healy DL. Occult ovarian failure: a syndrome of infertility, regular menses and elevated FSH concentration. J Clin Endocrinol Metab 1988;67: Reyes FI, Winter JSD, Fairman C. Pituitary ovarian relationships preceding the menopause. Am J Obstet Gynecol 1977;129: Lenton EA, Sexton L, Lee S, Cooke ID. Progressive changes in LH and FSH and LH:FSH ratio in women throughout reproductive life. Maturitas 1988;10: Toner JP, Philput CB, Jones GS, Muasher SJ. Basal folliclestimulating hormone level is a better predictor of in vitro fertilization performance than age. Fertil Steril 1991;55: Oehninger S, Acosta AA, Morshedi M, Veeck L, Swanson RJ, Simmons K, et al. Corrective measures and pregnancy outcome in in vitro fertilization in patients with severe sperm morphology abnormalities. Fertil Steril1988;50: Muasher SJ, Garcia JE, Rosenwaks Z. The combination of follicle-stimulating hormone and human menopausal gonadotropin for the induction of multiple follicular maturation for in vitro fertilization. Fertil Steril 1985;44: Veeck LL, Wortham JWE, Witmeyer J, Sandow BA, Acosta AA, Garcia JE, et al. Maturation and fertilization of morphologically immature human oocytes in a program of in vitro fertilization. Fertil Steril 1983;39: Muasher SJ, Oehninger S, Simonetti S, Matta J, Ellis LM, Liu H -C, et al. The value of basal and/or stimulated serum gonadotropin levels in prediction of stimulation response and in vitro fertilization outcome. Fertil Steril1988;50: Scott RT, Toner JP, Muasher SJ, Oehninger S, Robinson S, Rosenwaks Z. Follicle-stimulating hormone levels on cycle day 3 are predictive of in vitro fertilization outcome. Fertil Steril1989;51: Diamond MP, Wentz AC, Herbert CM, Pittaway DE, Maxon WS, Daniell JF. One ovary or two: differences in ovulation induction, estradiol levels, and follicular development in a program for in vitro fertilization. Fertil Steril 1984;41: Khalifa et a1. Basal FSH in cases of one ovary 839