4 The role of treatment with growth hormone in infertile patients ZEEV SHOHAM ROY HOMBURG E L I Z A B E T H J. O W E N G E R A R D S. C O N W A Y HANNE OSTERGAARD H O W A R D S. JACOBS

The optimal hormonal milieu and sequential events regulating granulosa cell maturation during follicular growth are still largely undetermined. Efforts have concentrated on elucidating the action of gonadotrophins and the major ovarian steroids on the processes of follicular maturation. The role of gonadotrophins and oestradiol (E2) in the development of the antral follicle have been well characterized (Veldhuis et al, 1982). Folliclestimulating hormone (FSH) is the major physiological regulator of E2 biosynthesis through stimulation of cytochrome P-450 aromatase (Erickson et al, 1979). Involvement of luteinizing hormone (LH) in follicle formation, ovulation and corpus luteum formation, is also well established (McNeilly et at, 1982). Involvement of growth hormone (GH) (Jia XC et al, 1986), insulin (Veldhuis et al, 1983) and insulin-like growth factor-I (IGF-I) or somatomedin-C (Adashi et al, 1985) have, however, only recently been examined. Two lines of evidence suggest a role for GH in the reproductive process. The first is the observation that lowering GH levels in female rats led to delayed puberty and decreased ovarian steroidogenesis in response to stimulation by gonadotrophins (Advis et al, 1981). It has also been noted that puberty is delayed and prolonged in children with Laron-type dwarfism, that is resistance to the action of GH (Laron et al, 1980), whereas puberty and gonadal maturation could be induced by the administration of GH to hypophysectomized animals (Shiekholislam and Stempfel, 1972). The second line of evidence suggesting a role for GH in the reproductive process is more deductive. Thus, in reviewing the endocrine events of the human ovulation cycle, one notes during the follicular phase a striking contrast between the explosive increase in E2 secretion, reflecting the exponential increase in follicular growth and the very modest increase in the concentrations of immunoreactive FSH. This difference in the pattern of E2 and FSH suggests the possibility of ovarian paracrine factors that amplify the Bailli~re' s Clinical Obstetrics and Gynaecology--

Vol. 6, No. 2, June 1992 ISBN 0-7020--1633-0

267 Copyright © 1992, by Bailli~re Tindall All rights of reproduction in any form reserved

268

Z . S H O H A M ET A L

effect of FSH on follicle growth. Adashi et al (1985a, 1985b) recently summarized information on the paracrine control of follicular activity: it appears that, in addition to gonadotrophins which regulate granulosa cell proliferation, intraovarian peptides also play a part in local control of follicular development (Hammond, 1981). It seems that insulin-like growth factors are particularly relevant in this matter. Their synergistic effects with FSH on granulosa cell differentiation have been demonstrated through stimulation of aromatase activity, induction of LH receptors, synthesis of progestins and proteoglycans, and cyclic adenosine monophosphate accumulation (Adashi et al, 1985a). In particular, IGF-I has bee1) shQwn to synergize with FSH in the induction of rat granulosa cell activity, ~;uggesting that peptide growth factors may contribute to promotion of follictilar development (Adashi et al, 1985b). It was also found that IGF-I enhances replication of cultured mammalian granulosa cells, and can induce doserelated increases in production of progesterone, pregnenolone, and 20 alpha-hydroxy-pregn-4-en-3-one by cultured granulosa cells (Savion et al, 1981; Barananao and Hammond, 1984; Veldhuis and Furlanetto, 1985). It was also shown in rat'granulosa culture systems, that purified IGF-I synergized with FSH to stimulate release of E2 into the cultured medium (Adashi et al, 1985b; 1986). In addition, there is some recent evidence that IGF-I is synthesized by mature rat ovarian follicles in which RNA transcripts have been identified (Hernandez et al, 1989; Oliver et al, 1989). At about the time the results of these in vitro studies were appearing, we noted that a number of our patients who were undergoing induction of ovulation were resistant to treatment with gonadotrophins. We hypothesized that by co-treatment with GH we might increase endogenous IGF-I concentrations and thereby sensitize the ovaries to stimulation by gonadotrophins. At that time, we have in mind that the putative action of GH would operate through a paracrine mechanism--that is to say, that treatment with GH would increase production of IGF-I within the ovarian follicle itself. Accordingly we started a series of clinical studies: the first was open but thereafter we moved to randomized placebo-controlled studies for induction of ovulation for in vivo and in vitro fertilization. The latter study was designed to gain access to follicular fluid from GH-treated cycles, but we also used the experimental set-up to determine whether, for the same dose of gonadotrophins, women who had been co-treated with GH would yield more follicles and more oocytes than women who received placebo. The first pilot study of co-treatment with growth hormone to facilitate induction of ovulation

Patients and study design Four patients undergoing induction of ovulation for in vivo and three patients for in vitro fertilization, age 30 to 39 years, who had all previously proved resistant to gonadotrophin therapy, were recruited to the study. All patients were treated with a combination of intramuscular (i.m.) injections of biosynthetic natural sequence growth hormone (GH; Norditropin, Novo

GROWTH HORMONE TREATMENT

269

Nordisk A/S Denmark) 20 IU on alternate days for 2 weeks (up to a total of 120 IU) in combination with hMG (Pergonal, Serono, Welwyn Garden City, Herts, UK). In the cycle of treatment for in vivo fertilization, hMG was given using an individually adjusted dose scheme which entailed increasing the number of ampoules given per day, every 5-7 days, until an ovarian response was obtained. This 'daily effective dose' was then contained until one or more follicles attained a size of 18mm or more, whereupon human chorionic gonadotrophin (hCG: Profasi, Serono, Welwyn Garden City, Herts, UK) was given in a single injection of 10 000 IU, to induce the final stage of ovulation. For induction of ovulation in the in vitro cycles, a fixed dose of hMG (usually 2 ampoules per day) was given.

Results The four patients who were treated for in vivo fertilization received a total of 21 cycles of hMG therapy alone and 14 cycles of co-treatment with GH. Comparing results of the different modes of treatment, we found that the required gonadotrophin dose fell significantly when treatment with GH was added compared with treatment with hMG alone (P=0,005); the daily effective dose and the duration of treatment was also reduced ( P = 0.001, P = 0.043, respectively) when treatment with G H was added. In the three patients who were treated for in vitro fertilization, again the gonadotrophin requirements and duration of treatment fell significantly when GH was added to the treatment ( P = 0.03, P17 mm in diameter. Human chorionic gonadotrophin, 10 000 IU, was administered to induce ovulation. Growth hormone, 24IU, was administered i.m. on alternate days to a total dose of 144 IU or until the day of hCG administration, starting on the first day of the hMG therapy. Patients who needed treatment for more than 12 days continued with hMG alone as the GH treatment was stopped after six injections. Placebo was adminstered in the same way as the GH. The mean ages of patients in the GH and the placebo group were 31.9 (range 23-40 years) and 30.1 (range 26-34 years) respectively, with no statistically significant difference. The mean serum (+SD) of serum LH, FSH, and E2 concentrations were 2.7_+1.7IU/1, 1.6_+1.5IU/1 and 62.5 _+35 pmol/1, respectively. The various clinical diagnoses were equally distributed between the two treatment groups and, in particular, there were two hypophysectomized patients in the placebo group and three in the GH-treated group. A comparison of the results of the last prestudy treatment cycle of the patients who subsequently received placebo with those of the last prestudy cycle of the patients who subsequently received GH revealed no significant differences. Results Three patients conceived in their first study cycle: one in the placebo and two in the GH group. During the GH treatment, the number of ampoules of hMG required to induce follicular development was reduced (P = 0.008), the duration of treatment was shorter (P=0.011) and the 'daily effective dose' was lower (P = 0.035) compared with the placebo cycles (Table 1). The number of follicles induced to a diameter of >I 17 mm was marginally higher in the placebo (2.4 + 1.1) than in the GH group (1.4 + 0.7) ( P = 0.05). The number of cohort follicles (14-16mm in diameter) was not significantly different in the placebo (2.9 + 2.2) and GH groups (2.4 + 1.4) (P = 0.58).

GROWTH HORMONE TREATMENT

271

Table 1. Comparison of the number of ampoules, days of treatment and daily effective dose of hMG in placebo and GH groups (values are mean +SD). Variable

Group

No. of cycles

Prestudy cycle

Treatment cycle

No. of hMG ampoules

Placebo GH

8 8

Days of treatment

Placebo GH

8 8

Daily effective dose of hMG

Placebo GH

8 8

43.6 + 13.1 34.4 + 7.8 P = 0.11 19.3 + 3.7 17.3 + 2.3 P = 0.22 3.13 + 0.83 2.88 + 0.64 P = 0.51

42.5 + 13.1 24.5 + 9.7 P = 0.008 18.5 + 3.8 13.4 + 3.2 P = 0.011 3.38 + 0.74 2.50 + 0.76 P = 0.035

T h e r a n g e of s e r u m E2 c o n c e n t r a t i o n s was 8 5 0 7 3 4 0 pmol/l a n d there was n o significant difference b e t w e e n the groups. A l l p l a c e b o cycles a n d all b u t o n e of the G H - t r e a t e d cycles were o v u l a t o r y , as d e t e c t e d b y u l t r a s o u n d s c a n n i n g and by midluteal serum progesterone concentrations. Serum I G F - I concent r a t i o n s rose d u r i n g G H t r e a t m e n t a n d p e a k e d b e t w e e n the s e c o n d a n d third i n j e c t i o n at a m e a n of m o r e t h a n twice the u p p e r limit of n o r m a l , falling b a c k i n t o n o r m a l r a n g e within o n e w e e k of the last G H i n j e c t i o n ( F i g u r e 1). T h e r e was n o c h a n g e in s e r u m I G F - I I c o n c e n t r a t i o n s . N e i t h e r of the growth factor c o n c e n t r a t i o n s c h a n g e d significantly d u r i n g p l a c e b o cycles.

Comments T h e results of this p r o s p e c t i v e , d o u b l e - b l i n d a n d r a n d o m i z e d study confirmed the findings i n t h e initial o p e n trial that c o - t r e a t m e n t with G H

90 80 70 60

~

_ - --

GH Placebo

IGF~ 50 nmol/I 40 I 30 ~!!!~i~i~i~~i!!~i~i~i!~!!~::ii~i~::i~i~ii~!*!*~!~:~ i ~ ~ ~::!~:::::~ ~ :I:~~!~!~:#::! 0

::::::~:+:::::

,0ot

:~::::::::

:::::::+~"

tttttt I

I

I

i

I

I

0

5

10

15

20

25

Days

Figure 1. Serum IGF-I concentrations in hMG + GH and hMG + placebo treatment cycles. The arrows indicate the timing of the injections of GH and placebo. The shaded areas indicate the normal range.

272

z . S H O H A M ET A L

sensitized the human ovary to stimulation by gonadotrophins. The results did, however, raise an important question--whether the effect of GH we observed was exerted directly on the ovary or mediated through IGF-I. To address this question, we studied patients undergoing treatment by in vitro fertilization (IVF) and embryo transfer (ET) as the IVF protocol offered us access to human follicular fluid (FF).

A prospective, randomized, double-blind, placebo-controlled trial, in patients undergoing IVF-ET following pituitary suppression

Study design and patients Patients undergoing IVF-ET at Hallam Medical Centre, London, between January and December 1989, were candidates for the study. We studied those who were less than 38 years of age and who had undergone one or more IVF-ET cycle(s) in which ovarian stimulation had been carried out using the combined regimen of luteinizing hormone-releasing hormone agonist (LHRHa) and hMG and in which the response had been considered 'suboptimal'. A suboptimal response was defined as one in which fewer than six oocytes were collected from which fewer than four embryos developed. Twenty-five patients were recruited into a randomized double-blind trial of co-treatment with GH compared with both the results of previous treatment and with co-treatment with placebo. All patients had a pretreatment ultrasound scan of the ovaries to determine ovarian morphology. According to the ovarian scan the patients were divided into two groups, those with normal ovaries and those with ultrasound diagnosed PCO. This distinction was based on the criteria of Adams et al (1985). Patients in both groups were then randomized to receive GH or placebo, in addition of course to their standard treatment for IVF. Thirteen patients were allocated to receive GH (24 units per injection given i.m.), and 12 to receive placebo injections, starting on the first day of hMG treatment. The GH or placebo was given on alternate days until the administration of hCG or for a maximum period of 2 weeks. At the completion of the cycle of treatment the assignment code was broken. Those who had received GH were considered to have completed the study, those who had received placebo entered an open study in which they received GH, so that they were not deprived of a potentially beneficial therapy. The results of the open study are not, however, discussed in this report. In all cases, an interval of 2 months was allowed to elapse between cycles of treatment. All patients had had at least one previous IVF-ET attempt using pituitary gonadotrophin suppression (buserelin, Suprefact, Hoechst, Hounslow, UK) 2001~g subcutaneously. In all treatment cycles, the analogue was administered daily from the first day of the menstrual period, for a minimum of 14 days (Figure 2). When ovarian suppression was confirmed (serum E2 concentrations 14mm in diameter were detected on ultrasound scan of the ovaries, with at least one _>17mm in diameter, in the presence of serum E2 concentration >1500pmol/1. Oocyte recovery was performed 35h later using transvaginal ultrasound-directed follicle aspiration. The technique of IVF, culture of oocytes and embryos, fertilization, and ET were as described by Owen et al (1989). Follicular fluid for analysis of IGF-I was collected by carefully selecting the clear portion of the aspirate using three separate tubes. The fluid was subsequently centrifuged, separated from cellular particles, and the supernatant stored frozen at -20°C. Only FF obtained from follicles containing an oocyte were used for measurement of IGF-I concentrations. In total, 44FF were analysed: 26 from GH augmented cycles and 18 from placebo cycles. During the period in this study, up to four embryos that had shown evidence of normal cleavage were transferred to the uterus after 48 h culture. Remaining embryos of sufficient quality were cryopreserved. Luteal phase support was given to all patients in the form of hCG 5000 IU on the day of ET and again 3 days later. The mean (+SD) age of the patients was 32.4 +__3.0 years in the GH group, which was not significantly different from that in the placebo group (33.5 + 2.6years). The mean body mass index (BMI = weight(kg)/height [m]2) was also not,significantly different (22.6 + 4.6 in the GH group and 22.0+1.8 in the placebo group) between the two groups. The various

274

Z . SHOHAM ET A L

clinical diagnoses were equally distributed between the two groups (11 patients had tubal damage, 7 had unexplained infertility, 3 had oligospermia, 2 had male antibodies and 2 failed donor insemination). Of the 25 patients, 18 were diagnosed as having PCO based solely on ovarian morphology according to the ultrasound criteria of Adams et al (1985). Pretreatment serum LH concentrations were not significantly different between the patients with ultrasound diagnosed PCO (medium of 6.2 IU/1, range 4.0-25.8 and those with ultrasound findings of normal ovaries (median of 5.5IU/1, range 3.%6.8) group. Results Comparing the results of the prestudy cycles in patients who went on to receive placebo or GH (Table 2), we found no significant difference between the groups in the dose of hMG used, number of days of treatment or in the number of follicles which developed. There were also no significant differences in the number of oocytes which fertilized or in the number of embryos which cleaved. There were, however, more oocytes collected in the

Table 2. Group comparison of the prestudy cycles of those who subsequently received GH or placebo. Values are medians with ranges in parentheses.

Days of hMG HMG/cycle (ampoules) Follicles ->14 mm (day of hCG) Oocyte collected Oocyte fertilized Embryos cleaved Embryos replaced

Prestudy cycles Placebo group (n = 12)

Prestudy cycles GH group (n = 13)

P

8.0 (6-12) 33.5 (18--64) 4.0 (2-6)

8.0 (7-10) 28.0 (14-60) 4.0 (2-7)

0.81 0.15 0.29

3.5 (2-6) 1.5 (0-3) 1.0 (0-3) 1.0 (0-3)

6.0 (0.8) 2.0 (0--4) 1.0 (0.3) 1.0 (0--3)

0.03 0.52 0.40 0.76

Table 3. Results of ovarian stimulation comparing GH and placebo cycles. Values are medians with ranges in parentheses.

Days of hMG HMG/cycle (ampoules) Follicles ->14 mm (day of bCG) Oocyte collected Serum E2 (pmol/l) Oocyte fertilized Oocyte cleaved Embryos replaced No. of pregnancies

Study cycle with placebo

Study cycle with GH

(n = 12)

(n = 13)

t'

8.5 (6-12) 36.0 (24-100) 5.0 (2-9)

8.0 (6-13) 28.0 (20-85) 8.0 (3-19)

0.73 0.05 0.15

5.0 (2-13) 3139 (1495-8440) 3.0 (0-6) 2.0 (0-5) 2.5 (0M) 1

11.0 (2-16) 3804 (1397-8960) 5.0 (0-11) 4.0 (0-10) 4.0 (0-4) 4

0.11 0.27 0.04 0.06

275

GROWTH HORMONE TREATMENT

Table 4. Group comparison of placebo with GH-treated cycles in patients with polycystic ovaries. Values are medians with ranges in parentheses.

HMG/cycle (ampoules) Follicles _>14 mm (day of hCG) Serum E2 (pmol/l) Oocyte collected Oocyte fertilized Oocyte cleaved Embryos replaced No. of pregnancies

Placebo (n=8)

GH ( n = 10)

P

36.0 (24-100) 6.5 (4-9)

27.5 (20--42) 8.5 (4-19)

0.01 0.05

3991 (1613--8440) 7.0 (2-13) 3.5 (0--6) 3.0 (0-5) 3.5 (0-4) 1

5810 (2355-8960) 11.5 (6-16) 6.0 (5-11) 5.5 (3-10) 4.0 (3-4) 4

0.17 0.03 0.004 0.02

prestudy cycles of the patients who went on to receive GH (median 6, range 0-8) than in those who went on to receive placebo (median 3.5, range 2-6), P