high dose r-hLH (375 IU, twice daily). Serum estradiol concentrations in the women receiving saline declined by the end of the 2-day ran- domization period.
0021-972X/99/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1999 by The Endocrine Society
Vol. 84, No. 1 Printed in U.S.A.
Ovarian Responses in Women to Recombinant FollicleStimulating Hormone and Luteinizing Hormone (LH): A Role for LH in the Final Stages of Follicular Maturation* MICHAEL W. SULLIVAN, ANN STEWART-AKERS, JOEL S. KRASNOW, SARAH L. BERGA, AND ANTHONY J. ZELEZNIK Departments of Obstetrics, Gynecology, and Reproductive Sciences (M.W.S., A.S.A., J.S.K., S.L.B., A.J.Z.) and Cell Biology and Physiology (A.J.Z.) and Magee Womens Research Institute, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15213 ABSTRACT During the follicular phase of the menstrual cycle, FSH stimulates follicular growth, granulosa cell aromatase activity, induction of LH receptors on the granulosa cell membrane, and estradiol secretion. As a result of negative feedback of estradiol on the pituitary, serum FSH concentrations decline. Despite the fall in FSH concentrations, the maturing follicle continues to develop to the preovulatory stage. In a prospective randomized trial, we tested the hypothesis that a key mechanism by which the dominant follicle continues to develop in the face of decreasing concentration of FSH is by acquiring LH responsiveness. In 24 women, pituitary gonadotropin secretion was downregulated with a GnRH agonist. Follicular growth was then stimulated with recombinant human FSH (r-hFSH) until a 14-mm follicle was identified by ultrasound. The women were then randomized to 1
of 4 groups for a 2-day period: continued r-hFSH treatment, substitution of r-hFSH with saline, low dose r-hLH (150 IU, twice daily), or high dose r-hLH (375 IU, twice daily). Serum estradiol concentrations in the women receiving saline declined by the end of the 2-day randomization period. In contrast, serum estradiol concentrations continued to rise in women receiving either r-hFSH or r-hLH compared with those in the saline-treated group (P , 0.05). Pregnancies occurred in each of the gonadotropin treatment groups. These findings indicate that once FSH initiates follicular growth, either FSH or LH is capable of sustaining follicular estradiol production. Extrapolating these findings to the normal menstrual cycle suggests that the maturing follicle may continue to develop in the presence of diminishing FSH concentrations by acquiring the capacity to respond to LH. (J Clin Endocrinol Metab 84: 228 –232, 1999)
HE SELECTION of a single ovulatory follicle during the menstrual cycle involves a process by which a maturing follicle inhibits the development of other follicles but does not succumb to its own inhibitory influences (1). The mechanism by which a follicle acquires dominance remains controversial. Studies in humans and subhuman primates have demonstrated convincingly that regulation of the absolute plasma concentrations of FSH during the follicular phase is crucial not only for the initiation of preovulatory follicular development, but also for the selective maturation of a single follicle (1, 2). Brown (3) demonstrated that the initiation of preovulatory follicular growth by FSH operates in a threshold manner, such that once adequate concentrations of FSH in blood are achieved, follicles are stimulated to advance from early antral stages to the fully mature preovulatory stage. In response to FSH stimulation, aromatase is induced in the granulosa cells of the maturing follicle, which results in the elevation of peripheral estradiol concentrations and feedback inhibition of FSH secretion (4 – 6). The inhibition of FSH secretion by the maturing follicle causes FSH concentrations in blood to fall below threshold
levels, thereby terminating the maturation of other less mature follicles (7–9) while the maturing follicle continues to develop in the presence of diminishing FSH concentrations (1, 10). Thus, as a direct consequence of stimulation by FSH, the follicle undergoes maturation-dependent changes, which diminish its dependence upon FSH (7). The final question that must be addressed to solve the problem of follicular selection is to elucidate the cellular mechanisms by which the developing follicle secures its diminished requirement for FSH. A hallmark of the cellular actions of FSH on the developing follicle is the induction of LH receptors on granulosa cells, which enables FSH-stimulated follicles to respond to LH (11). Although granulosa cells from early antral follicles respond only to FSH with increased cAMP production and steroid secretion, granulosa cells from mature follicles, which possess both FSH and LH receptors, are responsive to either FSH or LH (12, 13). Zeleznik and Hiller (14) proposed that the maturing follicle may become less dependent upon FSH because the presence of LH receptors on granulosa cells would enable it to respond to LH, whereas other lesser mature follicles whose granulosa cells lack sufficient LH receptors would not be protected from the fall in FSH. To date, this hypothesis has been untestable because preparations of pure FSH and LH have not been readily available for use in humans. The advent of recombinant gonadotropins has provided the reagents necessary to explore the individual roles of FSH and LH on follicular development in humans. In this study we report the use of recombinant gonadotropins to test
Received October 8, 1997. Revision received March 10, 1998. Rerevision received August 19, 1998. Accepted September 18, 1998. Address all correspondence and requests for reprints to: Michael W. Sullivan, M.D., State University of New York at Buffalo, Department of Gynecology and Obstetrics, Children’s Hospital, 219 Bryant Street, Buffalo, New York 14222. * This work was supported in part by NIH Grant HD-50748 and TAP Pharmaceuticals, Inc.
LH AND FOLLICULAR DEVELOPMENT
the hypothesis that the maturing follicle reduces its dependence on FSH by acquiring responsiveness to LH. Materials and Methods Gonadotropins Serono Laboratories, Inc. (Norwell, MA) generously provided recombinant human FSH (r-hFSH; Gonal-F) and r-hLH (LHadi). The specific activity of r-hFSH is 10,000 FSH IU/mg protein, and it has a terminal half-life of approximately 37 h (15). The specific activity of r-hLH is approximately 14,000 IU/mg, with a terminal half-life of approximately 10 h (Dr. Louis O’Dea, Serono Laboratories, Inc., personal communication). r-hFSH and r-hLH were packaged by the supplier in lyophilized form at 150 and 75 IU/vial, respectively. Hormones were reconstituted in sterile water within 30 min of sc injection.
Experimental design This prospective, randomized, double blinded study was approved by the institutional review board at Magee-Womens Hospital, University of Pittsburgh (Pittsburgh, PA). All participants were counseled regarding the nature and purpose of the study and signed a detailed consent form. Twenty-eight reproductive age women, ranging in age from 26 –36 yr with a mean age of 30.5 yr, were studied. All were ovulatory, of normal weight, and without evidence of tubal or ovarian disease. There was no history of ovarian or tubal surgery in any subject. All of the participants were nonsmokers. Three had previously been diagnosed with minimal endometriosis, 20 were diagnosed with unexplained infertility, and the partner of 1 participant suffered from male factor infertility. Women with evidence of anovulation, a past history of polycystic ovarian disease, an abnormal pelvic examination, or ultrasound evidence of ovarian cysts were excluded. Seventeen of the participants had been previously treated with clomiphene citrate, and 5 had been treated with gonadotropins before the study. All subjects were evaluated with physical exam and pelvic ultrasound scan before study participation. Before gonadotropin treatment, each participant self-administered leuprolide acetate (TAP Pharmaceuticals, Inc., Chicago, IL) 1 mg daily, sc, from menstrual day 21 throughout the study to minimize endogenous gonadotropin secretion. Serum samples, obtained by venepuncture, were assayed after 14 days of leuprolide acetate therapy for estradiol (E2) and LH. If the participant’s LH and/or E2 levels were above 2.5 IU/L or 20 pg/mL, respectively, the leuprolide therapy was continued for 7 more days, and serum E2 and LH were again assayed. Those women with serum LH levels greater than 2.5 IU/L after a 21-day period of leuprolide treatment were excluded from further study. Participants with serum E2 levels less than 20 pg/mL and LH levels less than 2.5 IU/L were then administered r-hFSH starting at 150 IU. r-hFSH was administered sc daily at 0730 h. After 4 days of r-hFSH treatment, serum was obtained by venepuncture for measurement of E2, androstenedione, FSH, and LH. If serum E2 levels were less than 100 pg/mL, the r-hFSH dose was increased to 225 IU. Follicle number and diameter were assessed daily with vaginal probe ultrasound (model RT3200 Advantage, 5-MHz vaginal probe, GE Medical Systems, Milwaukee, WI). Subjects with E2 levels greater than 250 pg/mL after 4 days of r-hFSH treatment were excluded from study. Participants with serum E2 concentrations below 250 pg/mL were maintained on r-hFSH at 150 IU/day or at 225 IU until the time of randomization. Daily blood samples were collected throughout the study. The serum was stored at 220 C for subsequent assay of steroids and gonadotropins. Stimulation with r-hFSH was continued until a 14-mm follicle (average of three dimensions) was identified by ultrasound. After a 14-mm
follicle was detected, each subject was randomized in a double blind fashion to one of four groups (n 5 6/group) by random drawing. Group A subjects discontinued r-hFSH and received 2 cc saline at 0730 and 1930 h for 2 days. Group B subjects continued r-hFSH for 2 days (75 IU at 0730 h and 75 IU at 1930 h). Those subjects receiving 225 IU r-hFSH before randomization received 112.5 IU r-hFSH at 0730 h and 112.5 IU r-hFSH at 1930 h. Group C subjects discontinued r-hFSH and received 375 IU r-hLH at 0730 and 1930 h for 2 days. Group D subjects discontinued r-hFSH and received 150 IU r-hLH at 0730 and 1930 h for 2 days. Two days after randomization (24 h after the final saline or recombinant gonadotropin injection) subjects received 10,000 IU hCG if the serum E2 concentration was less than 2500 pg/mL (see Results). Serum pregnancy tests were obtained 14 days after hCG administration for those women who did not experience menses.
Hormone measurements E2, androstenedione, LH, and FSH levels were determined in duplicate for each sample using commercially available kits. Estradiol was measured using a solid phase RIA (Diagnostic Products Corp., Los Angeles, CA). The interassay coefficient of variation (CV) for this assay was 10.5%, and the intraassay CV was 3.5%. Androstenedione concentrations were measured in a single assay via solid phase RIA (Diagnostic Products Corp.) with an intraassay CV of 2.0%. The gonadotropins were measured by immunofluorometric assays (DELFIA hLH Spec and DELFIA hFSH kit, Wallac, Gaithersburg, MD). The LH assay had an interassay CV of 7.1% and an intraassay CV of 3.0%, and the FSH interassay and intraassay CVs were 5.8% and 3.9%, respectively.
Statistical analysis One-way ANOVA was used to determine whether differences existed between the groups for the parameters of age, body mass index, duration of infertility, and menstrual cycle length. The differences in FSH and LH concentrations across the groups, between the day of randomization (day zero) and the day of hCG administration (day 2), were analyzed by two-factor ANOVA with repeated measures in which one factor was the treatment group and the other factor (repeated measure) was the study day (time). A similar analysis was applied to the androstenedione concentrations. Post-hoc comparisons were examined using the StudentNewman-Keuls test (16). Because there were individual variations in the serum concentrations of E2 among women before randomization, the estradiol measurements are expressed as the percent change in the E2 level between day 0 and day 2 and were analyzed by one-way ANOVA; post-hoc comparisons were examined using the Student-Newman-Keuls test. The number of follicles 14 mm or greater were compared between groups on the day of hCG administration (day 2) using one-way ANOVA.
Results Study population
Table 1 shows the study population demographics. All subjects were similar with respect to age, duration of infertility, pregnancy history, menstrual cycle length, and body mass index. Twenty-eight women were recruited for the study, and 24 women were randomized and completed the study. Three were excluded because of an elevated serum E2 level after 4 days of r-hFSH treatment, and 1 woman was discontinued from the study because she failed to respond to
TABLE 1. Demographic characteristics of the treatment groups Group
Cycle length (days)
Duration of infertility (yr)
Saline (group A) r-hFSH (group B) r-hLH high (group C) r-hLH low (group D)
30.17 6 3.2 31.00 6 2.1 29.33 6 1.2 31.33 6 3.0
25.52 6 4.5 24.44 6 2.4 22.20 6 2.5 23.48 6 4.4
28.2 6 3.3 28.5 6 2.4 28.8 6 0.8 28.5 6 2.7
3.5 6 2.2 3.5 6 2.4 4.0 6 2.9 3.8 6 1.7
SULLIVAN ET AL.
JCE & M • 1999 Vol 84 • No 1
r-hFSH (serum E2 level of , 100 pg/mL after 5 days of r-hFSH at 225 IU/day). hCG was withheld from 2 women who were at increased risk for ovarian hyperstimulation based on follicle number and serum estradiol levels greater than 2500 pg/mL. Gonadotropin and E2 concentrations
Figures 1 and 2 illustrate serum FSH (Fig. 1) and LH (Fig. 2) concentrations in the four study groups on the day of randomization (day 0) and the day of hCG administration (day 2). As expected, there were no significant differences between the groups with respect to serum concentrations of FSH or LH on the day of randomization (day 0). On the day of hCG administration (day 2), serum FSH concentrations in the subjects maintained on r-hFSH (Fig. 1, group B) were significantly greater (P , 0.05) than those in the salinetreated group (group A), the high dose r-hLH-treated group (group C), and the low dose r-hLH-treated group (group D). Likewise, serum LH concentrations in the subjects receiving r-hLH during the study period (Fig. 2, groups C and D) were significantly (P , 0.05) greater than those receiving saline or r-hFSH (Fig. 2, groups A and B, respectively). The mean LH concentrations in group C (high dose r-hLH) were significantly greater than those in the low dose r-hLH (group D). Figure 3 illustrates serum E2 concentrations in the individual treatment groups. Serum E2 concentrations in women receiving saline (group A) increased the day after randomization (day 1) and then decreased by the day of hCG administration (day 2). In contrast, the serum estradiol levels of women in the groups receiving either r-hFSH or r-hLH (groups B, C, and D) rose throughout the study period. The percent change in E2 concentrations between day 2 and day 0 was 116.8 6 0.87% for group A, 1122.6 6 1.8% for group B, 1163.0 6 2.4% for group C, and 1112.8 6 4.2% for group D. The changes in E2 differed significantly (P , 0.05) between the saline group (group A) and the groups receiving either
FIG. 2. Results show the mean serum LH concentration (international units per L) 6 SEM of the four treatment groups (A–D) on the day of randomization (day 0; black bars) and the day of hCG administration (day 2; gray bars). Note that the mean LH concentrations of the groups receiving r-hLH (groups C and D) were greater than those of the groups not receiving r-hLH (groups A and B; P , 0.05).
FIG. 3. Results show the mean serum E2 concentrations (picograms per mL) 6 SEM of the four treatment groups (A–D) throughout the study period (days 0, 1, and 2). Note that the serum E2 concentrations increased throughout the study period in the groups receiving gonadotropin (groups B, C, and D), whereas serum E2 concentrations decreased toward the baseline in the saline-treated group (group A).
r-hFSH or r-hLH (groups B, C, and D; P , 0.05). Serum androstenedione concentrations ranged between 0.18 – 0.28 ng/mL; there were no statistically significant differences in the various treatment groups at any point measured (data not shown). Pregnancy results FIG. 1. Results show the mean serum FSH concentration (international units per L) 6 SEM of the four treatment groups (A–D) on the day of randomization (day 0; black bars) and the day of hCG administration (day 2; gray bars). The mean concentration of FSH was maintained in the group receiving r-rFSH (group B) as the mean concentration of FSH fell in groups A, C, and D (P , 0.05).
No pregnancies occurred in the women receiving saline (group A). There was a twin pregnancy delivered without complication at 34 weeks gestation in group B (r-hFSH). One woman in group C (375 IU r-hLH, twice daily) delivered triplets at 32 weeks gestation without complication, and
LH AND FOLLICULAR DEVELOPMENT
there was a singleton pregnancy delivered at 37 weeks gestation in group D (150 IU r-hLH, twice daily). Discussion
During the mid- through late follicular phase of the menstrual cycle, preovulatory follicular development proceeds despite a progressive fall in serum FSH concentrations. We initiated the current studies to determine whether LH plays a role in maintaining follicular development as serum FSH concentrations decline. The rationale for this hypothesis is that the FSH-stimulated induction of LH receptors on granulosa cells could enable the maturing follicle to respond to LH and thereby continue to mature in the presence of continuously declining FSH concentrations. We tested this hypothesis in 24 reproductive aged women using recombinant human gonadotropins, r-hFSH and r-hLH, after pituitary down-regulation with leuprolide acetate. The results of this study demonstrate that LH sustains follicular E2 production in the presence of falling serum FSH concentrations. Serum E2 concentrations were used as an index of follicular development because analysis of the number of follicles 14 mm or more in diameter measured before and after the study period revealed no significant differences in any of the treatment groups. In retrospect, this finding is not surprising, as it has been shown by others that follicles do not decrease in size or collapse 1–2 days after gonadotropin withdrawal (17). In group A, substitution of r-hFSH with saline resulted in reductions in serum E2 concentrations by 48 h. The serum E2 concentrations of women receiving r-hLH (groups C and D) increased throughout the study period despite the fact that FSH concentrations declined to levels seen in the control group. The increases in serum E2 in the women receiving r-hLH were not statistically different from those in women continuing r-hFSH treatment (group B). These results indicate that although follicles 14 mm or more in diameter require gonadotropic stimulation for continued E2 production, they are responsive to either FSH or LH. It is generally accepted that E2 production by the maturing follicle occurs by way of the two-cell, two-gonadotropin model (18, 19). In this model, theca cells produce androstenedione and testosterone under LH stimulation, and FSH induces granulosa cell aromatase, thus enabling the thecally derived androgens to be metabolized to E2. Assuming the validity of this model in humans (20, 21), our results indicate that thecal androgen production is exquisitely sensitive to LH, as a plasma LH concentration of 1.5 IU/L was sufficient to maintain E2 production (day 0; groups A–D) as well as plasma androstenedione concentrations (data not shown). Our observation of E2 production despite very low serum LH concentrations is in agreement with other published data showing that women treated with GnRH agonists to suppress gonadotropin secretion maintain E2 production in the presence of very low levels of serum LH (,0.5 IU/L) (22). Our current study also indicates that although LH concentrations of approximately 1.5 IU/L are able to sustain thecal androgen production, these levels of LH are unable to maintain granulosa cell aromatase activity when FSH concentrations decline. Thus, serum E2 concentrations decreased in the control group (group A) despite serum concentrations
of androstenedione equal to those in the women receiving either r-hFSH or r-hLH. This result suggests that the decrease in E2 secretion in the control group was due to a deficiency of aromatase activity rather than to androgen precursors. Our results also indicate that serum concentrations of LH of 2.5 IU/L or more were sufficient to maintain adequate androgen substrate and maintain follicular aromatase activity as FSH concentrations fell, because the serum E2 concentrations increased in the two groups receiving r-hLH (groups C and D). Comparison of the plasma concentrations of FSH and LH observed in our study with those observed during the spontaneous menstrual cycle gives further merit to the hypothesis that LH protects the maturing follicle from declining FSH concentrations. Using the same FSH and LH assays that we used, Saketos et al. (23) noted that serum FSH concentrations during the spontaneous menstrual cycle fall from approximately 10 IU/L in the early follicular phase to approximately 4 –5 IU/L during the mid- through late follicular phase. The decline in FSH concentrations during the spontaneous follicular phase is similar to the decline in FSH concentrations seen in our study subjects who discontinued FSH at the time of randomization (groups A, C, and D). Without additional treatment with r-hLH, these FSH concentrations (4 –5 IU/L) were unable to sustain follicular E2 production (group A). However, the study by Saketos et al. (23) also revealed that as FSH concentrations declined during the mid- through late follicular phase, plasma LH concentrations were maintained at approximately 4 IU/L, similar to the mean concentration of LH (3.6 IU/L) observed in our subjects that received r-hLH after randomization (groups C and D). Follicular E2 production continued in our subjects when LH concentrations were maintained at levels seen during the spontaneous menstrual cycle (groups C and D), whereas the lower concentration of LH (,1.5 IU/L) observed in group A was unable to sustain E2 production. Our current results, therefore, are consistent with the idea that the absolute concentration of LH present during the mid- through late follicular phase of the spontaneous menstrual cycle (4 –5 IU/L) may play an important role in maintaining preovulatory folliculogenesis as FSH concentrations decline. Although our data show that LH is capable of maintaining the maturation of follicles with diameters of 14 mm, the actual stage of follicular development when LH can sustain follicular development in the presence of declining serum FSH concentrations is not known. The occurrence of multiple pregnancies in two of the three subjects in the present study indicates that a number of follicles reached the preovulatory stage in response to LH. The most likely explanation for the exaggerated ovarian response is that the duration of the stimulation by FSH was sufficient, such that a number of follicles acquired LH responsiveness and therefore were maintained by LH. Thus, FSH treatment was arbitrarily discontinued when one or more follicles reached a diameter of 14 mm. If follicles less than 14 mm in diameter had acquired LH responsiveness, a greater number of follicles would have been able to respond to LH and reach the preovulatory stage. Although studies in humans have indicated that LH receptors are present on the granulosa cells by the midfollicular phase (day 7) and increased throughout the late follicular phase (24), the exact size at which a follicle becomes LH
SULLIVAN ET AL.
responsive is not known. Defining a cut-off point below which LH or hCG will not maintain follicular growth could be helpful to control the number of preovulatory follicles in ovulation induction protocols. Theoretically, a sequential FSH and LH ovarian stimulation protocol could be used to limit follicular recruitment, thereby reducing the complications now associated with ovulation induction protocols. In addition, hCG is usually used as a surrogate to LH to stimulate ovulation in most infertility treatment protocols. Inevitably, at the time of hCG administration multiple follicles in various stages of development are present (2). In view of our current results, administration of hCG to initiate ovulation of the leading follicles may also have unintended consequences by maintaining the growth of smaller follicles and therefore may contribute to the undesirable effect of multiple pregnancies and/or ovarian hyperstimulation. In summary, our results are consistent with the hypothesis that the maturing follicle continues to develop in the presence of diminishing FSH concentrations because it acquires the capacity to respond to LH. If so, this would indicate that the FSH-mediated induction of aromatase and LH receptors on granulosa cells are the principal components of the process of follicle selection. By acquiring aromatase, the maturing follicle produces E2, which inhibits FSH secretion and terminates the maturation of less mature follicles, while at the same time the concomitant induction of LH receptors enables maturing follicles to continue to develop despite the fall in FSH concentrations. FSH-dependent preovulatory follicular development is associated with other changes in the follicle, such as angiogenesis and the production of autocrine and paracrine growth factors that have been shown in vitro to modify the actions of FSH and LH (1–3, 25). However, our results indicate that in the absence of an adequate gonadotropin stimulus in vivo, the various locally produced paracrine and autocrine factors are unable to sustain follicular development in the presence of declining serum FSH concentrations. Acknowledgments We thank Dr. Louis O’Dea, Ms. Kate Banks, and Ms. Marsha Yarkosky from Serono Laboratories, Inc. for generously providing the recombinant gonadotropins, and TAP Pharmaceuticals, Inc. for providing Lupron. We are very grateful for the technical advice of Ms. Tammy Daniels, and the clinical assistance of Ms. Beth Mastascusa.
References 1. Zeleznik AJ, Fairchild-Benyo D. 1994 Control of follicular development, corpus luteum function, and the recognition of pregnancy in higher primates. In: Knobil E, Neill JD, eds. The physiology of reproduction, 2nd Ed. New York: Raven Press; vol 2:751–782. 2. Fauser BC, Van Heusden AM. 1997 Manipulation of human ovarian function: physiological concepts and clinical consequences. Endocr Rev. 18:71–106.
JCE & M • 1999 Vol 84 • No 1
3. Brown J. 1978 Pituitary control of ovarian function: concepts derived from gonadotropin therapy. Aust NZ J Obstet Gynecol. 18:47–54. 4. Steinkampf MP, Mendelson CR, Simpson ER. 1987 Regulation by follicle stimulating hormone of the synthesis of aromatase P-450 in human granulosa cells. Mol Endocrinol. 1:465– 471. 5. Zeleznik AJ. 1981 Premature elevation of systemic estradiol reduces serum levels of follicle stimulating hormone and lengthens the follicular phase of the menstrual cycle in rhesus monkeys. Endocrinology. 109:352–355. 6. Zeleznik AJ, Hutchison JS, Schuler HM. 1987 Passive immunization with anti-oestradiol antibodies during the luteal phase of the primate menstrual cycle potentates the perimenstrual rise in serum gonadotropin concentrations and stimulates follicular growth in the cynomolgus monkey. J Reprod Fertil. 80:403– 410. 7. Zeleznik AJ, Kubik CJ. 1986 Ovarian responses in macaques to pulsatile infusion of follicle stimulating hormone and luteinizing hormone: increased sensitivity of the maturing follicle to FSH. Endocrinology. 119:2025–2032. 8. Baird DT. 1987 A model for follicular selection and ovulation: lessons from super ovulation. J Steroid Biochem. 27:15–23. 9. Fauser BC, Donderwinkel P, Shoot DC. 1993 The step-down principle in gonadotropin treatment and the role of GnRH analogues. Bailliere Clin Obstet Gynecol. 7:309 –330. 10. vanSantbrink EJ, Hop WC, van Dessel TJ, de Jong FH, Fauser BC. 1995 Decremental follicle-stimulating hormone and dominant follicle development during the normal menstrual cycle. Fertil Steril. 64:37– 43. 11. Zeleznik AJ, Midgley AR, Reichert LE. 1974 Granulosa cell maturation in the rat: increased binding of human chorionic gonadotropin following treatment with follicle stimulating hormone in vivo. Endocrinology. 95:818 – 825. 12. Hillier SG, Zeleznik AJ, Ross GT. 1978 Independence of steroidogenic capacity and luteinizing hormone receptor induction in developing granulosa cells. Endocrinology. 102:937–946. 13. Richards JS, Jahnsen T, Hedin L, Lifka J, Ratoosh S, Durcia JM, Goldring NB. 1987 Ovarian follicular development: from physiology to molecular biology. Recent Prog Horm Res. 43:231–276. 14. Zeleznik AJ, Hillier SG. 1984 The role of gonadotropins in the selection of the preovulatory follicle. Clin Obstet Gynecol. 27:927–940. 15. Loumaye E, Campbell R, Salat-Baroux J. 1995 Human follicle-stimulating hormone produced by recombinant DNA technology: a review for clinicians. Hum Reprod. 1:188 –199. 16. Miller Jr RG. 1981 Simultaneous statistical inference. New York: SpringerVerlag; 81–90. 17. Shoot DC, Coelingh Bennink HJ, Mannaerts BM, Lamberts SW, Bouchard P, Fauser BC. 1992 Human recombinant follicle-stimulating hormone induces growth of preovulatory follicles without concomitant increase in androgen and estrogen biosynthesis in a woman with isolated gonadotropin deficiency. J Clin Endocrinol Metab. 74:1471–1473. 18. Falck B. 1959 Site of production of oestrogen in rat ovary as studied in micro-transplants. Acta Physiol. 193(Suppl):1–101. 19. Armstrong DT, Papkoff H. 1976 Stimulation of aromatization of exogenous and endogenous androgens in ovaries of hypophysectomized rats in vivo by follicle stimulating hormone. Endocrinology. 99:1144 –1151. 20. Baird DT, Fraser IS. 1975 Concentration of oestrone and oestradiol in follicular fluid and ovarian venous blood of women. Clin Endocrinol (Oxf). 4:259 –266. 21. Hillier S, DerBoogaard AV, Reichert L, VanHall E. 1980 Intraovarian sex steriod hormone interactions and the regulation of follicular maturation: aromatization of androgens by human granulosa calls in vivo. J Clin Endocrinol Metab. 50:640 – 664. 22. Loumaye E, Engrand P, Howles CM, O’Dea L. 1997 Assessment of the role of serum luteinizing hormone and estradiol response to follicle-stimulating hormone in vitro fertilization treatment outcome. Fertil Steril. 67:889 – 899. 23. Saketos M, Sharma N, Adel T, Raghuwanshi M, Santoro N. 1994 Timeresolved immunoflorometric assay and specimen storage conditions for measuring urinary gonadotropins. Clin Chem. 40:749 –753. 24. Yamoto M, Shima K, Nakano R. 1992 Gonadotropin receptors in human ovarian follicles and corpra lutea throughout the menstrual cycle. Horm Res. 37:5–11. 25. Zeleznik, AJ. 1993 Dymanics of primate follicular growth: a physiologic perspective. In: Adashi E, Leung PKC, eds. The ovary. New York: Raven Press; 41– 45.