FERTILITY AND STERILITY

Vol. 53, No.3, March 1990

Copyright 0 1990 The American Fertility Society

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

Comparison of urinary human follicle-stimulating hormone and human menopausal gonadotropin for ovarian stimulation in polycystic ovarian syndrome* Torben Larsen, M.D. t:f:§ J0rgen Falck Larsen, M.D.t Vibeke Schi0ler, Ph.D. II

Erik Bostofte, M.D. t Christine Fehling, M.D.t

Herlev University Hospital, Copenhagen, Denmark

A randomized, double-blind, crossover study was carried out to compare purified urinary follicle-stimulating hormone (FSH) and human menopausal gonadotropin (hMG) for ovarian stimulation in polycystic ovarian syndrome (PCOS). Twelve patients were stimulated with FSH and hMG in three alternate cycles. FSH, luteinizing hormone (LH), estradiol, dihydroepiandrosterone sulphate, free and total testosterone, ~5 -androstenedi­ one, sex hormone binding globulin, and ovarian volume were monitored during the stimulation. There was no difference between the dose of FSH and hMG necessary to induce preovulatory follicles in the individual patients. The mean increase of ovarian volume during stimulation with FSH and hMG was 120% and 129% respectively (no significant difference). Two patients became pregnant in the first cycle. Two other patients had delayed bleeding and positive serum-human chorionic gonadotropin. No significant difference was found in the endocrine changes during the two different stimulation methods. The LH/FSH ratio was normalized after a few days of treatment regardless of the type of stimulation. The size of the material does not permit a comparison of the efficacy of the two treatment schedules. Our clinical and ultrasonic observations do not support the theory that treatment of infertility in PCOS with FSH is more safe than with hMG. Fertil Steril53:426, 1990

In the treatment of infertility in patients with polycystic ovarian syndrome (PCOS), clomiphene citrate (CC) is the drug of first choice. In some cases, however, this treatment does not lead to pregnancy. This may be explained by ovulation failures, luteal phase defects or different types of luteinized unruptured syndrome (LUF). Human menopausal gonadotropin (hMG) in combination with human chorionic gonadotropin (hCG) has been used to induce ovulation in those Received June 21, 1989; revised and accepted November 8, 1989. • Presented in part at the XII World Congress of Gynecology and Obstetrics, Rio de Janeiro, Brazil, October 23 to 28, 1988. t Department of Obstetrics and Gynecology. :1: Department of Ultrasound §Reprint requests: Torben Larsen, M.D., Department of Obstetrics and Gynecology, Herlev University Hospital, DK-2730 Herlev/Copenhagen, Denmark. II Department of Clinical Chemistry.

426

Larsen et al. Purified urinary FSH and hMG in PCO

cases in which CC has failed. In many cases the treatment with hMG/hCG is difficult because patients with PCOS seem to have highly sensitive ovaries and have a significant risk of superovulation and hyperstimulation. The PCOS patient already has an elevated endogenous level of luteinizing hormone (LH), and it has been suggested that the treatment with hMG containing additional LH causes the risk of hyperstimulation.1 Therefore, the use of "pure" folliclestimulating hormone (FSH) has been advocated for the treatment of PCOS. 2 In 1977 Raj et al. 3 observed ovulation in 14 out of 18 cycles and achieved two pregnancies using an FSH preparation obtained from human pituitary extracts. It has been shown that in some cases ovulation can be induced with purified pituitary FSH without the addition of LH or hCG. 4 Clinical trials have demonstrated that a purified FSH preparation obtained from human menopausal urine, "uroFertility and Sterility

follitropin" FSH, can be used for ovulation induction in PCOS patients.5- 7 It was claimed that the incidence of hyperstimulation and the potential for multiple births appeared lower with FSH alone than with hMG in combination with hCG. 7 In patients with PCOS, levels of the androgens testosterone (T) and a5 -androstenedione are increased. Although the total circulating androgen pool may not be above the normal range, the biologically active (free) concentrations are elevated because of a reduction in sex hormone binding globulin (SHBG). The increased production of androgens is believed to be induced by the elevated levelofLH. The total production of estrone is markedly increased because of extraglandular (nongonadal) aromatization of androgens, principally a5 -androstenedione. The use of a purified FSH preparation should, theoretically, correct the biochemical imbalance and normalize the LH/FSH ratio. 8 If this is the case, the abnormal concentrations of androgens and estrone should also be corrected The purpose of this study was to compare the clinical effects and endocrine changes during stimulation with FSH and hMG. The hypothesis was that FSH is as effective as hMG in inducing ovulation in patients with PCOS, that the risk of hyperstimulation is less with FSH, and that the abnormal levels of hormones are corrected during treatment with FSH. MATERIALS AND METHODS

Twelve patients with PCOS were included in the study. All couples had been through a complete infertility work-up. All women were anovulatory in all or most of the observed cycles. Enlarged ovaries with small cysts had been diagnosed by laparoscopy, laparotomy or ultrasound scanning. Tubal patency was ascertained by hysterosalpingography and/or laparoscopy. The LH/FSH ratio was elevated. They had hirsutism and/or increased free T. All had normal serum prolactin (PRL). The patients had failed to conceive during treatments with CC. Three patients did not ovulate when treated with CC. The other patients had ovulation in some CC cycles but did not conceive. All husbands had a normal semen analysis except in one case in which artificial insemination with donor was used. Prior to the treatment, baseline values were established for FSH, LH, estradiol (E2), estrone, diVol. 53, No.3, March 1990

hydroepiandrosterone sulphate (DHAS), total and free T, a5 -androstenedione, SHBG, and PRL. The ovarian volume and the size of the adrenal glands were evaluated by ultrasound. If pregnancy had not occurred in the first or second cycle, it was planned to induce ovulation in three cycles in each patient. The stimulation started on the 4th day after the onset of a spontaneous bleeding, or in patients with amenorrhea or oligomenorrhea, on the 5th day after the last tablet of 5 mg medroxyprogesterone given twice a day for 4 days to induce a withdrawal bleeding. Using a double-blind design, each patient was randomized to different stimulation schedules using either hMG (Pergonal; Serono, Rome, Italy) or purified human urinary FSH (FSH) (Metrodin; Serono). Each patient was treated with both methods in different cycles. In this way the patients served as their own controls. The randomization was performed by Serono (Rome, Italy). Each ampule was marked with a code for the patient and the cycle in which it was to be used. Thus, the type of hormone stimulation was unknown to the clinician, who controlled the daily dosage of gonadotropin. The gonadotropins were given intramuscularly, and the dosage was adjusted individually according to the response as judged by the increase in serum E 2 and by the growth of the follicles as measured by ultrasonography. The starting dose was 75 IU for the first 3 days because of the known risk of hyperstimulation in PCOS. The dose was increased by 75 IU until rise in serum E 2 and follicular growth measured by ultrasound. The maximum daily dose was 225 IU. The dosage was then maintained until the serum E 2 had reached a level between 1 and 2 nmol/L and the leading follicle measured> 18 rom in diameter. Then the patient was given an injection with 5,000 IU ofhCG (Profasi; Serono). Ultrasound examinations of the ovary and the endometrium were carried out daily, as were assays for serum E 2 and serum LH. These results were available for the treating gynecologist. Serum was stored for future assays of FSH, androgens, estrone, and SHBG. Ovulation was detected by ultrasound, and serum progesterone (P) was measured one week after the hCG injection or spontaneous LH peak. If a vaginal bleeding was not observed 2 weeks after the injection of hCG, a pregnancy test was performed. In case of a positive test, quantitative serum hCG and ultrasound examinations were done weekly. Larsen et al.

Purified urinary FSH and hMG in PCO

427

The hormone tests were carried out using commercial radioimmunoassays (RIA). A kit from Nordic-Lab, Oulu, Finland, was used for serum E 2 • Serum P was measured with a kit from Hoechst Beringwerke, Marburg, West Germany. For deter~ mination of FSH and LH, reagent kits from Amersham Int., Little Chalfont, Amersham, England were used (reference standards WHO 2 IRP 78/549, MCR 69/104 and WHO 1 IRP 68/40). Serum estrone was measured with a kit from Wien Laboratories, Succasunna, NJ. The kit for PRL was supplied by Amersham Int. (reference standard WHO 1 IRP 75/504). The androgens and SHBG were measured with RIA at the State Serum Institute, Copenhagen, Denmark. Free T was assayed using an RIA after dialysis at Medicinsk Laboratorium, Copenhagen, Denmark. Ultrasound scannings of the ovaries and endometrium were performed daily by the same gynecologist using dynamic ultrasound equipment (Picker LSC 7000; Hitachi Medical Cooperation, Tokyo, Japan; SSD-280; Aloka, Tokyo, Japan or Briiel & Kjaer 1845-8529; Briiel & Kjaer, Naerum, Denmark). During stimulation the following variables were recorded: 1. Size of the ovaries. The ovarian volume (v) was expressed in milliliters (mL) and calculated as v = 1r X a X b, where a is the long axis of the ovary, and b is the diameter perpendicular to the long axis. 2. Number of growing follicles. The follicles were measured in three diameters, and the mean follicle diameter was calculated. 3. Thickness of the endometrium. On the days following the hCG injection, the ovaries were examined with ultrasound to detect ovulation or a possible luteinized unruptured follicle (LUF) syndrome. All experiments and calculations were completed before the code was revealed to the clinical investigators. Wilcoxon's test for pair differences was used to evaluate the possible difference in the effect by stimulation using the two different hormone preparations. The project was approved by the Committee on Medical Ethics for Copenhagen County, and all patients gave informed consent.

RESULTS Clinical Observations

Twelve patients were included in the study. A total of 32 series were started. Two stimulations were discontinued, one because of bleeding and one because of ultrasonic signs ofhyperstimulation with a 428

Larsen et al.

Purified urinary FSH and hMG in PCO

110 100 90

80 70

_60 E

50 40 30 20 10 0

huFSH

~ Before Day 0 Before Day 0 stimulation (HCG) stimulation (HCG)

Figure 1 Change in ovarian volume during the two types of stimulation.

total ovarian volume of about 175 mL. This patient was stimulated with hMG. Thus, 30 cycles were completed. The gonadotropins used for stimulation were FSH in 15 cycles and hMG in 15 cycles. Two patients became pregnant during the first cycle and delivered normal infants at term. One patient was stimulated with hMG and the other patient with FSH. The average dose of gonadotropin necessary to induce preovulatory follicles was 1,190 IU using FSH and 980 IU in the cycles in which hMG was used. There was no significant difference between the doses of FSH and hMG necessary to induce preovulatory follicles in the individual patients. Injection of hCG was not given in three cycles with spontaneous LH peaks. In a fourth cycle the serum LH analysis showed that an LH peak had occurred before the injection of hCG. Biochemical pregnancies occurred in two cycles in which FSH was used and in one cycle with hMG. In these cases the bleeding was delayed for 2 to 4 weeks, and the serum hCG was elevated, but fetal heart activity was not observed by ultrasound. Clinical hyperstimulation syndrome was not observed in any of the 30 completed cycles. One of the patients conceived after stimulation with FSH after this study was completed. Ultrasonic Findings

The sonograms of the ovaries showed considerable variations in the different cycles. In most cases a large number of follicles was stimulated. Figure 1 shows the increase in ovarian volume in the individual treatment cycles. The average increase in ovarian volume was 129% when hMG was used and 120% in the FSH cycles. Two patients experienced a much larger increase in ovarian volume Fertility and Sterility

Serum LH miU/ml

30 25

i:!: SEM

20 15 10 5

0 -6

Figure 2 ulation.

-5

0 -4 ·1 ·3 ·2 Days from spontaneous LH-peak or injection of hCG

Changes in serum LH during the two types of stim-

during stimulation with hMG than with FSH. No significant difference was found when the increase in ovarian volume produced by the two different treatments was analyzed in the other patients. Although the increase of ovarian volume assessed by sonography was considerable and occurred during a few days, clinical signs of hyperstimulation were sparse. The ovaries were examined for ultrasonic evidence9 of ovulation. Definitive evidence of ovulation was observed in 16 cycles. Serum Prose to normalluteal values in 14 of these cycles. Convincing ultrasonic evidence of ovulation was missing in 14 cycles. Collapse of follicles was not observed in these cases, and, in many, follicles showed continued growth subsequent to a spontaneous LH peak or injection of hCG. This pattern was found in 8 cycles in which FSH was used and in 6 hMG cycles. The serum P was normal in all these cases. At the time ofthe injection of hCG or occurrence of a spontaneous LH peak, the endometrium measured 10 mm or more except in three cases in which the thickness was between 7 and 9 mm. The serum E 2 was