r

FERTILITY AND STERILITY Copyright~

Vol. 55, No.1, January 1991

1991 The American Fertility Society

Printed on acid-free paper in U.S.A.

Preovulatory follicular fluid steroid levels in stimulated and unstimulated cycles triggered with human chorionic gonadotropin*

JaneL. Frederick, M.D.t:j: Mary M. Francis, B.S.§ Thelma M. Macaso, B.S.§

Roger A. Lobo, M.D.t§ Mark V. Sauer, M.D.t§ RichardJ. Paulson, M.D.t§ll

University of Southern California School of Medicine, California Medical Center, Los Angeles, California

The purpose of this study was to analyze follicular fluid (FF) samples for steroid levels from stimulated and unstimulated cycles triggered with human chorionic gonadotropin (hCG) and to assess the influence of controlled ovarian hyperstimulation and luteinizing hormone/hCG on these levels. Spontaneous ovulatory cycles were monitored with serial ultrasound examinations, and hCG 10,000 IU 'has given when the lead follicle was mature. Fourteen FF samples yielding fertilizable oocytes were compared with 13 FF samples from controlled ovarian hyperstimulation cycles. Progesterone (P) was higher in controlled ovarian hyperstimulation than in unstimulated cycles (9.0 ± 1.2 JLg/mL versus 4.4 ± 0.6 JLg/mL; mean ±SEM), whereas estradiol (E 2 ) was lower(0.8 ± 0.1 JLg/mL versus 1.3 ± 0.2 JLg/mL), resulting in a higher P:E 2 ratio (15.5 ± 3.3 versus 4.4 ± 0.7). Androstenedione (A), testosterone (T), and T:E 2 ratios were all higher in unstimulated than controlled ovarian hyperstimulation cycles. We conclude that controlled ovarian hyperstimulation is associated with increased FF P, decreased FF E 2 , T, and A levels, and decreased T:E 2 ratios, suggesting altered steroidogenesis and enhanced follicular aromatase activity. Fertil Steril55:44, 1991

Follicular fluid (FF) influences the microenvironment of the cumulus-oocyte complex through all antral stages of follicular development. It is formed by the exudate of the follicular cells and provides a hormonal microenvironment that is a key factor in determining the subsequent growth and maturation of the follicle. Peripheral steroid levels do not correlate with FF steroid levels, and Received May 15, 1990; revised and accepted October 18, 1990. * Presented at the 38th Annual Meeting of the Pacific Coast Fertility Society, Scottsdale, Arizona, April25 to 29, 1990. t Department of Obstetrics and Gynecology, University of Southern California School of Medicine. :j: Present address: University of California Irvine Medical Center, Division of Reproductive Endocrinology and Infertility, Orange, California. § California Reproductive Health Institute, California Medical Center. II Reprint requests: RichardJ. Paulson, M.D., Women's Hospital, 1240 North Mission Road, Room L-1022, Los Angeles, California 90033.

44

Frederick et al.

Follicular steroids in unstimulated IVF

some FF steroids may be present at concentrations that are several orders of magnitude greater than in plasma. 1 Changes in ovarian blood flow may affect the entry of blood borne substances into the follicle 2 and result in reduced levels of these substances, emphasizing the importance of measuring the FF directly. Analysis of FF steroid levels in stimulated and unstimulated cycles has been reported. 3 -5 However, previous studies have utilized FF from unstimulated cycles either before the onset of the luteinizing hormone (LH) surge 6 or after a variable period of LH exposure. 7 By contrast, FF from stimulated cycles was collected after a set period of time after administration of human chorionic gonadotropin (hCG ). Thus, the influence of stimulation with clomiphene citrate (CC) or human menopausal gonadotropin (hMG) could not be separated from the possible influence ofhCG. We recently have reported successful pregnancies after in vitro fertilization (IVF) in unstimuFertility and Sterility

lated cycles triggered with hCG. 8 This protocol provides a unique opportunity to assess the influence of controlled ovarian hyperstimulation on FF steroid levels as the follicles are not exposed to any stimulation until the administration of hCG, and the time from hCG injection to aspiration is the same as in stimulated cycles. Thus, a comparison between these FF samples and those obtained from stimulated cycles reflects the effect of controlled ovarian hyperstimulation while controlling for the important influence of hCG. The purpose of this study was to analyze FF samples from stimulated and unstimulated cycles triggered with hCG for steroid levels to assess the influence of controlled ovarian hyperstimulation on these levels while controlling for the effects of LH/hCG.

MATERIALS AND METHODS

The unstimulated IVF protocol has previously been described. 8 •9 Briefly, normally ovulatory patients undergo serial ultrasound and serum estradiol (E 2 ) monitoring. Human chorionic gonadotropin 10,000 IU was administered when follicles reach 20 mm and E 2 levels were >200 pg/mL. If an LH surge is not detected by urinary enzyme immunoassay, then follicle aspiration is performed 36 hours later. The program has preliminarily established an ongoing pregnancy rate of 13% per retrieval and 16% per embryo transfer. 8 From the first 25 aspirations, we were successful in obtaining an oocyte from the dominant follicle, which fertilized and cleaved in 20 cases. Of these, 14 FF samples were free of blood and culture medium and were used in the following analysis. All of the oocytes from these follicles were successfully fertilized, cleaved, and transferred. The clinical results of this series have been reported elsewhere. 8 For the purpose of comparison, we also analyzed FF samples from 13 individuals undergoing 13 stimulated cycles. The details of these regimens have been previously described. 10•11 Briefly, the first regimen consisted of (CC) 100 mg/d, menstrual cycle day 3 to 7, with 75 IU of hMG given concurrently and continued until at least two follicles were > 18 mm by ultrasound measurement, when hCG 10,000 IU was given (CC-hMG, n = 5). The second regimen consisted of leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, North Chicago, IL) 1 mg/d starting in the midluteal phase and continued for 2 weeks or until down regulation was achieved. The initial dose of hMG 225 IU/d was titrated to follicular response after 4 days of Vol. 55, No.1, January 1991

stimulation and hCG 10,000 IU was given when follicle maturity criteria were achieved (diameter, 18 mm; E 2 , 200 pg/mL/codominant follicle; LA-hMG, n = 8). In all cases, if an LH surge was not detected by urinary enzyme immunoassay, then follicle aspiration was performed 36 hours after hCG administration. After microscopic inspection of the aspirated follicular contents, the fluid was immediately centrifuged at 300 X g. Cell-free FF was frozen at -20°C until analyzed for steroid content. To control for follicle maturity, only FF samples yielding oocytes judged mature by morphological criteria and that fertilized in vitro were considered. No significant difference between the ages of the patients in the stimulated and the unstimulated groups was noted. Estradiol, progesterone (P), testosterone (T), and androstenedione (A) were measured by direct radioimmunoassay (RIA; Diagnostic Products, Los Angeles, CA). Statistical significance was determined by the two sample t-test. RESULTS

The results of the steroid RIA are depicted in Table 1. Fourteen FF samples from unstimulated cycles were compared with 13 FF samples from controlled ovarian hyperstimulation cycles. The data are described separately for each stimulation regimen and for the two stimulated regimens taken together. No differences were noted between the CC-hMG and LA-hMG groups in any of the steroids measured. For the purpose of comparing controlled ovarian hyperstimulation versus unstimulated, the controlled ovarian hyperstimulation groups were considered together. Follicular fluid P levels in the unstimulated group were significantly lower than in the controlled ovarian hyperstimulation group (P < 0.05), whereas E 2 levels were higher (P < 0.05). Testosterone and A levels in the unstimulated group were higher than in the controlled ovarian hyperstimulation group (P < 0.05). As a result of lower P and higher E 2 levels in the unstimulated group, P:E 2 ratios were also lower than in the controlled ovarian hyperstimulation group (P < 0.05). Although T was higher in the unstimulated than in the controlled ovarian hyperstimulation group, E 2 levels were also increased, and whereas the T:E 2 ratios were more than fourfold higher, statistical significance was marginal (P"' 0.08). In examining each stimulation group separately, both LA-hMG and CC-hMG groups had results that were signifiFrederick et al.

Follicular steroids in unstimulated IVF

45

Table 1

Follicular Fluid Steroid Levels a Unstimulated cyclesb

Steroid P (~tg/mL) E 2 (~tg/mL) T (ng/mL) A (ng/mL) P:E2 ratio T:E 2 ratio (xl0- 3 )

Controlled ovarian hyperstimulation c

CC-HMGd

LA-HMG•

9.0 ± 1.2 0.8 ± 0.1 7.3 ± 0.8 15.2 ± 1.5 15.5 ± 3.3 1.1 ± 0.2

7.9 ± 2.4 1.0 ± 0.2 7.9 ± 0.8 18.2 ± 2.1 8.0 ± 1.4 1.2 ± 0.2

9.6 ± 1.4 0.6 ± 0.1 6.9 ± 1.3 13.3 ± 1.7 20.1 ± 4.7 1.0 ± 0.3

4.4 ± 0.6' 1.3 ± 0.2g 13.3 ± 2.3h 52.1 ± 11.7h 4.4 ± 0.7' 5.1 ± 2.1j

Values are means ± SEM. Unstimulated cycles triggered with hCG; n = 14. c Total average of all stimulations; n = 13. dn = 5. •n = 8. f Progesterone levels in unstimulated cycles group are significantly lower than in LA-hMG and controlled ovarian hyperstimulation (P < 0.05 for each). g Estradiol levels in unstimulated cycles group are significantly higher than in LA-hMG and controlled ovarian hyperstimulation (P < 0.05 for each).

h Testosterone and A in unstimulated cycles group are significantly higher than in LA-hMG, CC-hMG, and controlled ovarian hyperstimulation (P < 0.05 for each). ' Progesterone:E 2 ratio in unstimulated cycles group is significantly lower than in CC-hMG, LA-hMG, and controlled ovarian hyperstimulation (P < 0.05 for each). j Differences in the T:E 2 ratio in the unstimulated cycles and controlled ovarian hyperstimulation groups did not reach significance (P = 0.08).

cantly different from those of the unstimulated group, except in the T:E 2 ratio. The CC-hMG group, perhaps because of the smaller size, also did not achieve statistical significance for P or E 2 levels, although the trend in the differences was the same as in the LA-hMG group (Table 1).

oocyte obtained. No correlation was found between FF E 2 levels and the stage of oocyte maturation. The mean P levels were higher and T and A concentrations were lower in FF containing metaphase I and metaphase II oocytes than in FF containing germinal vesicle oocytes. The P:E 2 and A: E 2 ratios were higher in FF containing metaphase I and metaphase II oocytes compared with FF containing oocytes with germinal vesicles. Since the stage of meiosis of the aspirated oocyte is dependent on the time of the LH exposure, this data suggest that the time ofLH exposure has profound influence on FF steroid levels. Our unstimulated IVF protocol allowed us to control for the LH surge by triggering ovulation with hCG administration and then recovering FF from stimulated and unstimulated cycles in the same time interval. We found that P levels in the controlled ovarian hyperstimulation group were greater than in the unstimulated group, suggesting that controlled ovarian hyperstimulation, independent of hCG, results in increased follicular steroidogenesis. Estradiol levels were lower in the controlled ovarian hyperstimulation group and as a result, the P:E 2 ratios were higher in the controlled ovarian hyperstimulation group. Whereas P:E 2 ratios have been suggested to correlate best with oocyte maturity and morphology, 6 and since all of the oocytes in this study were morphologically mature and fertilized, our findings suggest a further modulation of this ratio by controlled ovarian hyperstimulation. Androgens were decreased in the controlled ovarian hyperstimulation group and T:E 2 ratios were also lower in controlled ovarian hyperstimu-

a

b

DISCUSSION

During the follicular phase of the human menstrual cycle, the ovary is found to contain several follicles in various stages of development. In most cases, of these antral follicles, only one dominates and proceeds to ovulation. MeN a tty et al. 12 studied the human preovulatory FF and concluded that the hormonal microenvironment of the graafian follicle was a key factor in the subsequent preovulatory growth and development of the follicle and of the cumulus oocyte complex. Previous studies have shown that FF steroid content is modulated and altered by exposure to gonadotropic stimuli. 13 Lobo et al. 6 compared FF steroids in mature follicles from untreated patients with those of patients undergoing controlled ovarian hyperstimulation with CC-hMG. They found that the untreated follicles had lower E 2 levels and higher P:E 2 ratios than the controlled ovarian hyperstimulation follicles. In studies using bovine follicles, Dielman et al. 14 suggested that the changes in follicular steroid levels were regulated by the preovulatory LH peak. More recently, Seibel et al. 7 obtained FF during laparoscopy at various time intervals after the onset of an endogenous LH surge and correlated FF steroid levels with the stage of maturation of the 46

Frederick et al.

Follicular steroids in unstimulated IVF

Fertility and Sterility

lation perhaps due to increased aromatase activity. This is consistent with the known ability offolliclestimulating hormone to stimulate aromatase activity in granulosa cells in vitro. 15 Interestingly, studies in bovine preovulatory follicles have found that granulosa cells obtained before the LH surge had a significantly greater capacity to aromatize exogenous T to E 2 than did cells obtained after the LH surge in vitro, 16 implying that LH stimulation and/ or luteinization results in relatively lower aromatase activity. Our findings of decreased androgen and E 2 levels and increased P:E 2 ratios in the controlled ovarian hyperstimulation group contrast with those of Lobo et al. 6 who reported that controlled ovarian hyperstimulation led to similar androgen and P levels, increased E 2 levels, and decreased P:E 2 ratios as compared to the unstimulated cycles. Even though Lobo et al. 6 controlled for oocyte maturity by examining FF only from follicles that contained a fertilizable oocyte, the unstimulated group did not receive hCG and the follicles were exposed to endogenous LH stimulation only. The difference in these findings, therefore, can be attributed to the hCG stimulation administered to our unstimulated group. Thus, it appears that triggering of ovulation by exogenous hCG, as compared with endogenous LH, results in a relative decrease in P levels and increase in E 2 and androgen levels, implying altered steroid metabolism. From our data, the following general statements may be made. When controlling for the hCG exposure, controlled ovarian hyperstimulation, during the follicular phase, alters steroidogenesis in FF and appears to enhance aromatase activity. As compared with previous studies that used endogenous LH exposure, hCG has a profound effect on the steroid environment of the follicular fluid. We speculate that this difference in steroid levels may be reflected in qualitative differences in the cumulus oocyte complex. REFERENCES 1. Hartshone GM: Preovulatory follicular fluid: relationships to ovarian stimulation protocol, fertilization, and sperm penetration in vitro. Fertil Steril52:998, 1989 2. Frederick JL, Shimanuki T, diZerega GS: The initiation of

Vol. 55, No.1, January 1991

angiogenesis by human follicular fluid. Science 224:389, 1984 3. diZerega GS, Campeau JD, Ujita EL, Kling DR, Marrs RP, Lobo RA, Nakamura RM: The possible role for a follicular protein in the intraovarian regulation of folliculogenesis. Sem Rep rod Endocrinoll:309, 1983 4. diZerega GS, Campeau JD, Nakamura RM, Ujita EL, Lobo RA, Marrs RP: Activity of a human follicular fluid protein in spontaneous and induced ovarian cycles. J Clin Endocrinol Metab 57:838, 1983 5. Marrs RP, Lobo RA, Campeau JD, Nakamura RM, Brown J, Ujita EL, diZerega GS: Correlation of human follicular fluid inhibin activity with spontaneous and induced follicle maturation. J Clin Endocrinol Metab 58:704, 1984 6. Lobo RA, diZerega GS, Marrs RP: Follicular fluid steroid levels in dysmature and mature follicles from spontaneous and hyperstimulated cycles in normal and anovulatory women. J Clin Endocrinol Metab 60:81, 1985 7. Seibel MM, Smith D, Olugi AM, Lesque L: Periovulatory follicular fluid hormone levels in spontaneous human cycles. J Clin Endocrinol Metab 68:1073, 1989 8. Paulson RJ, Sauer MV, Francis MM, Macaso TM, Lobo RA: In vitro fertilization in unstimulated cycles: a clinical trial utilizing hCG for timing of follicle aspiration. Obstet Gynecol 76:788, 1990 9. Paulson RJ, Sauer MV, Lobo RA: In vitro fertilization in unstimulated cycles: a new application. Fertil Steril 51: 1059, 1989 10. Paulson RJ, Marrs RP: Ovulation stimulation and monitoring for IVF. Curr Probl Obstet Gynecol Fertil 9: 498, 1986 11. Paulson RJ, Sauer MV, Lobo RA: Embryo implantation after human in vitro fertilization: importance of endometrial receptivity. Fertil Steril 53:870, 1990 12. McNatty KP, Smith DM, Madris A, Osathanondh R, Ryan KJ: The microenvironment of the human antral follicle: interrelationships among the steroid levels in antral fluid, the population of granulosa cells, and the status of the oocyte in vivo and in vitro. J Clin Endocrinol Metab 49: 851, 1979 13. Ainsworth L, Tsang BK, Downey BR, Marcus GJ, Armstrong DT: Interrelationships between follicular fluid steroid levels, gonadotrophic stimuli, and oocyte maturation during preovulatory development of porcine follicles. Biol Reprod 23:621, 1980 14. Dielman SJ, Kruip TAM, Fontijine P, de Jong WH, van der W eyden GC: Changes in estradiol, progesterone and testosterone concentrations in follicular fluid and in the micromorphology of preovulatory bovine follicles relative to the peak of LH. J Endocrinol 97:31, 1983 15. Armstrong DT, Papkoff H: Stimulation of aromatization of exogenous and endogenous androgens in ovaries of hypophysectomized rats in vivo by follicle-stimulating hormone. Endocrinology 99:1144, 1976 16. Fortune JE, Hansel W: Effect of LH surge on steroid secretion by theca and granulosa cells of bovine preovulatory follicles. Biol Reprod 20(suppl):46A, 1979

Frederick et al.

Follicular steroids in unstimulated IVF

47