band was identified as IGFBP-4 by its comigration with sem- inal plasma IGFBP-4 (43). Ligand blots of follicular fluid from four patients are shown in Fig. 1.
0021-972X/97/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1997 by The Endocrine Society
Vol. 82, No. 4 Printed in U.S.A.
Adjunctive Growth Hormone during Ovarian Hyperstimulation Increases Levels of Insulin-Like Growth Factor Binding Proteins in Follicular Fluid: A Randomized, Placebo-Controlled, Cross-Over Study* JARON RABINOVICI, NICHOLAS A. CATALDO†, PRAMILA DANDEKAR, STEPHEN M. ROSENTHAL, SHARRON E. GARGOSKY, NEIL GESUNDHEIT, AND MARY C. MARTIN Departments of Obstetrics, Gynecology and Reproductive Sciences (J.R., N.A.C., P.D., M.C.M.), and Pediatrics (S.M.R.), University of California San Francisco, San Francisco, California 94143-0132; Department of Obstetrics and Gynecology (J.R.), Sheba Medical Center, Tel-Hashomer 52621, Israel; Department of Pediatrics (S.E.G.), Oregon Health Sciences University, Portland, Oregon 97201; and Department of Clinical Research (N.G.), Genentech Inc., South San Francisco, California 94080 ABSTRACT GH increases circulating insulin-like growth factor I (IGF-I), which can promote the growth and differentiated function of ovarian granulosa and theca cells. Reported studies of GH as an adjunct to menotropin stimulation in women, largely those with ovarian dysfunction, have not consistently shown a benefit of GH, despite increases in serum and follicular fluid IGF-I. We hypothesized that changes in intrafollicular IGF-binding proteins (IGFBPs), which can antagonize IGF actions on granulosa cells, may underlie the incon-
sistent effects of GH. In the present study of GH, administered in double-blind, placebo-controlled, cross-over fashion to regularly cycling women undergoing in vitro fertilization, we found that follicular fluid levels of IGFBP-1, -3, and -4 and serum levels of IGFBP-3, as well as follicular fluid and serum IGF-I, were significantly increased in the GH-treated cycles, when compared with the placebo cycle of the same patient. We suggest that the net increase in intrafollicular IGFBPs in GH cycles may mitigate the potential beneficial effect of increased IGF-I. (J Clin Endocrinol Metab 82: 1171–1176, 1997)
T
which IGF-I and IGF-II are bound with high affinity in serum and tissue fluids (reviewed in Ref. 7). In most IGF target tissues, including the ovary, IGFBPs are produced locally, are differentially regulated, and antagonize the actions of IGFs, consistent with sequestration of the growth factor from its receptor. In this model, only free, unbound IGFs are biologically active (1, 7). Human granulosa cells in culture express IGFBP-1, -2, -3, and -4 messenger RNA and release these proteins; all four have been found in follicular fluid (1, 8 –18). IGFBPs can decrease both steroidogenesis and mitosis in cultured granulosa cells, opposing the actions of IGFs and gonadotropins (18 –21). After initial favorable reports of adjunctive GH in hMG ovulation induction (22, 23), numerous subsequent studies, of mainly IVF cycles, have yielded mixed results: both a beneficial effect (24 –28) and no clinical effect of GH (29 –37) have been reported. These divergent clinical results were found despite consistent increases in serum and follicular fluid IGF-I levels (27–31). The majority of these studies included only poor responders to traditional ovarian stimulation protocols (25–34), and only some studies were randomized and blinded (27, 28, 30 –33). Only one study (35) examined unselected patients. It remains unclear how adjunctive GH influences ovarian function and why clinical efficacy has been so inconsistently found. We hypothesized that adjunctive GH may alter the balance between IGF-I and IGFBPs in the follicular fluid of women
HE INSULIN-LIKE growth factors, IGF-I and IGF-II, can modulate ovarian follicular function. In humans and other species, IGFs are produced within the follicle and can augment basal and gonadotropin-stimulated steroidogenesis and mitosis (1). These observations have led to an interest in GH as an adjunct to menotropins [human menopausal gonadotropin (hMG)] in ovulation induction regimens for both in vivo and in vitro fertilization (IVF). GH could influence ovarian function by one or more mechanisms. GH is unlikely to stimulate human ovarian IGF-I production directly, because only the theca expresses IGF-I (2), and this layer does not express GH receptors (3). GH could increase delivery to the ovary of IGF-I produced in the liver or other sites. GH could also act directly on granulosa cells, which express GH receptors (3) and show increased steroidogenesis in culture in response to GH (4 – 6). Finally, GH may affect ovarian IGF action indirectly by altering intrafollicular levels of one or more IGF-binding proteins (IGFBPs), members of a family of at least six proteins to Received April 17, 1996. Revision received December 3, 1996. Accepted December 18, 1996. Address all correspondence and requests for reprints to: Nicholas A. Cataldo, M.D., Reproductive Endocrinology Center, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California San Francisco, San Francisco, California 94143-0556. * Presented, in part, at the 50th Annual Meeting of The American Fertility Society, November 7–10, 1994. † Recipient of an American Fertility Society-Serono Laboratories, Inc. Research Grant in Reproductive Medicine.
1171
1172
RABINOVICI ET AL.
with normal ovulatory function. Two groups reported no effect of GH on follicular fluid IGFBP-1 or -3 levels, but both included only poor responders to hMG (30, 31, 38). In the present study, we examined whether GH affects follicular fluid IGFBP and IGF-I levels, and thus IGF bioactivity, in women without ovulatory dysfunction undergoing ovarian hyperstimulation with hMG before IVF. Materials and Methods General study design All women under age 40, with tubal infertility, undergoing IVF at the University of California San Francisco, were eligible to participate in a double-blind, placebo-controlled study of adjunctive GH during ovarian hyperstimulation, approved by the University of California San Francisco Committee for the Protection of Human Subjects. Patients were excluded for ovulatory disorders, endocrine disease, severe endometriosis, or male factor infertility. Informed consent was obtained from all subjects. Participants were treated with a GnRH analog (leuprolide acetate or nafarelin acetate) starting in the luteal phase. After pituitary down-regulation, hMG (Pergonal, Serono, Norwell, MA) was begun at three ampules daily for 3 days. Further doses were individually determined by plasma estradiol (E2) levels and/or ovarian sonographic findings. In addition to hMG, patients received alternate-day injections, beginning on the first day of hMG, of either placebo or recombinant human GH (Protropin, Genentech) at 0.1 mg/kg (39), provided by Genentech in number-coded vials. Randomization was performed at Genentech. When three or more ovarian follicles 17 mm or larger were present and plasma E2 was between 900 and 2700 pg/mL, human CG (hCG, 10,000 IU) was administered, and ovarian follicles were aspirated under ultrasonographic guidance 36 h later. After removal of the oocyte-cumulus complex, fluid samples from individual follicles were centrifuged and frozen at 220 C before analysis. Blood samples were obtained at four time points during the cycle: after pituitary down-regulation with GnRH agonist and before hMG; on the 4th and 6th days of hMG; and on the day of hCG administration. Serum was separated by centrifugation and stored at 220 C. Fifteen patients underwent an initial cycle of treatment. Of these, the seven patients who requested to be entered in the study for a second cycle were crossed over in double-blind fashion to the opposite treatment; these patients are the subject of the present report. The treatment code was broken only after all patients had been studied. The present analysis includes all available follicular fluid samples (n 5 46) from both cycles in four of these patients and all available serum samples (n 5 42) from both cycles in five patients. Samples were excluded from analysis if paired samples were not available from both study cycles.
Ligand blot assay for IGFBPs Clear samples of follicular fluid (10 mL) were analyzed by electrophoresis on 10% SDS-polyacrylamide gels, transferred to nitrocellulose, and incubated with [125I]IGF-I (Amersham, Arlington Heights, IL) as described (13, 40). Autoradiograms were analyzed by integrated laser densitometry. All samples from each patient were analyzed on the same blot. Because of differences in signal intensities between blots, levels of each IGFBP species were normalized for each patient to the mean levels of that IGFBP in samples from her own placebo cycle.
Immunoprecipitation IGFBP-1 was identified in follicular fluid by immunoprecipitation and ligand blotting, as previously described (13), using a rabbit antiserum to IGFBP-1 from Upstate Biotechnology, Inc. (Lake Placid, NY), with less than 0.5% cross-reactivity with IGFBP-2, -3, -4, and -5.
RIAs for IGFBP-1, IGFBP-3, and IGF-I All samples of each type (follicular fluid or serum) were analyzed in a single assay run. IGFBP-1 was measured in follicular fluid with an immunoradiometric assay kit from Diagnostic Systems Laboratories, Inc. (Activey; Webster, TX). IGFBP-3 was measured in follicular fluid
JCE & M • 1997 Vol 82 • No 4
and serum by a previously described RIA employing a rabbit antiserum to glycosylated IGFBP-3 and tracer prepared by cross-linking radioiodinated IGF peptide to glycosylated IGFBP-3 (41). IGF-I was measured by double-antibody RIA after acid-ethanol extraction, using a modification of a published method (42). Samples diluted 1:50 in phosphatebuffered saline were extracted with 87.5% ethanol/12.5% 2 mol/L HCl, then incubated with a rabbit anti-IGF-I antibody prepared in our laboratory at a final concentration of 1:50,000. Samples were precipitated with a mixture of 1% goat antirabbit gamma globulin, 0.1% nonimmune rabbit serum, and 2% polyethylene glycol 6000. The IGF-I standard was from Chiron (Emeryville, CA). The assay sensitivity was 20 ng/mL, and intra- and interassay coefficients of variation were 8% and 15%, respectively.
Statistics Statistical comparisons between IGFBP or IGF-I levels in GH and placebo cycles were performed by t tests or ANOVA, with Scheffe’s F as a post hoc test. Differences among patients were examined by two-way ANOVA. IGF-I to IGFBP-3 ratios were log-transformed before comparison. P , 0.05 was taken as significant.
Results Follicular fluid IGFBP levels
Ligand blotting. Follicular fluid revealed a doublet of bands at 37– 43 kDa (IGFBP-3) and bands at 33 kDa (IGFBP-2), 27 kDa (IGFBP-1), and 24 kDa, consistent with a previous report (10). The 27-kDa band was identified as IGFBP-1 by immunoprecipitation and ligand blotting (data not shown). The 24-kDa band was identified as IGFBP-4 by its comigration with seminal plasma IGFBP-4 (43). Ligand blots of follicular fluid from four patients are shown in Fig. 1. By densitometric analysis, GH treatment significantly increased levels of IGFBP-3, IGFBP-1, and IGFBP-4 in follicular fluid, as indicated by a ratio greater than unity of the mean follicular fluid IGFBP level in GH cycles to the mean in the same patient’s placebo cycle. No statistically significant change in IGFBP-2 levels was noted with GH (Table 1). RIA. IGFBP-3 levels in follicular fluid from GH cycles were significantly (P , 0.05) greater than in fluid from placebo cycles in each of the four patients (Fig. 2A). Mean (6sd) IGFBP-3 in follicular fluid was 2940 6 499 ng/mL (n 5 22) in GH cycles and 2213 6 300 ng/mL (n 5 24) in placebo cycles (P , 0.0001). Significant differences (P , 0.02) also were found among patients in the relative increase with GH in follicular fluid IGFBP-3 levels. Like IGFBP-3, mean (6sd) IGFBP-1 in follicular fluid from GH cycles (83.2 6 18.3 ng/ mL) also was significantly greater than in placebo cycles (71.2 6 16.5 ng/mL; P , 0.03; Fig. 2B). Follicular fluid IGF-I levels
Follicular fluid IGF-I levels were significantly (P , 0.001) greater in GH cycles in each of the four patients (Fig. 2C). Mean (6sd) IGF-I was 178.2 6 50.9 ng/mL in GH cycles and 87.9 6 18.4 ng/mL in placebo cycles (P , 0.0001). Follicular fluid IGFBP-3 and IGF-I levels were significantly positively correlated (r 5 0.63, P , 0.0001) (Fig. 2D). The ratio of IGF-I to IGFBP-3 in follicular fluid by RIA was greater in the GH than the placebo cycle in three of the four patients; for all four patients, the geometric mean ratio (wt: wt) was 0.059 in GH cycles, compared with 0.039 in placebo cycles (P , 0.0001).
GH INCREASES OVARIAN FOLLICULAR IGFBPs
1173
TABLE 1. IGFBP levels in follicular fluid by ligand blotting
IGFBP-3 IGFBP-2 IGFBP-1 IGFBP-4
(37– 43 kDa) (33 kDa) (27 kDa) (24 kDa)
GH
Placebo
P
147.3 6 14.1 (21) 128.0 6 13.4 (14) 156.2 6 13.0 (14) 287.7 6 49.9 (14)
100 6 10.3 (19) 100 6 11.2 (13) 100 6 7.5 (13) 100 6 16.0 (13)
0.0115 0.1231 0.0011 0.0019
All follicular fluid samples from each patient were analyzed on the same ligand blot. Autoradiograms (exposure times 17 h to 9 days; see Fig. 1) were analyzed by integrated laser densitometry. For each IGFBP species, the mean of the densitometric signal intensities in the follicular fluid samples obtained from each patient’s placebo cycle was arbitrarily set at 100%, and the signal intensities of that IGFBP in follicular fluid from that same patient’s GH cycle were normalized to that 100% value, to allow comparison of results from samples from different patients on different blots. Shown for each IGFBP are the mean and SEM of all samples, with the number of samples in parentheses. The P values for the effect of GH treatment derived from two-way ANOVA are shown in the right column.
ministration. GH treatment during the stimulation cycle led to a significant overall increase in serum IGFBP-3 from 1468 6 177 to 1865 6 382 ng/mL (mean 6 sd; P , 0.01). IGFBP-3 levels at each time point during the stimulation cycle are shown in Fig. 3A. Serum IGF-I levels
As expected, GH treatment also increased serum IGF-I; mean (6sd) serum IGF-I during placebo-hMG cycles was 157 6 37 ng/mL and, during GH-hMG cycles, was 373 6 109 ng/mL (P , 0.0001). IGF-I levels at each time point during the stimulation cycle are shown in Fig. 3B. Clinical effects of GH treatment
There were no differences of clinically relevant magnitude and no statistically significant differences between GH and placebo cycles in duration of hMG treatment or total hMG used, peak plasma E2 level, number of retrieved oocytes or their maturity, fertilization and cleavage rates, or number of embryos transferred per cycle. There was one pregnancy, which occurred in a GH cycle and which ended in a secondtrimester spontaneous abortion of twins (Table 2). Discussion FIG. 1. Autoradiograms of ligand blots of follicular fluid samples. Follicular fluid (10 mL/lane) from GH (G) and placebo (P) cycles was subjected to SDS-10% PAGE and ligand blotting with [125I] IGF-I (13, 40). Molecular sizes were determined by the positions of prestained protein markers (in kDa, left). The identity of each IGFBP species has been reported previously (see Ref. 1). Because a 28-kDa glycosylated variant of IGFBP-4 has been reported in follicular fluid (see Ref. 13), the 27-kDa species was identified as IGFBP-1 by immunoprecipitation with a specific antiserum (data not shown). Panel A, Samples from Patients A and B, exposure times: 8 days (upper) and 17 h (lower); panel B, samples from Patient C, exposure times: 9 days (upper) and 3 days (lower); panel C, samples from Patient D, exposure time: 1 day.
Serum IGFBP-3 levels
IGFBP-3 levels were determined by RIA in serum samples drawn in the basal, GnRH-agonist-down-regulated state; on the 4th and 6th days of hMG; and on the day of hCG ad-
The present study of adjunctive GH in hMG ovulation induction differs from previous studies in including only women without ovulatory disorders. The double-blind, cross-over design allows an analysis of the interpatient differences in GH effects on follicular fluid or serum IGF-I and IGFBPs. We found that GH significantly increased IGFBP levels in follicular fluid at the time of oocyte aspiration. By ligand blotting, follicular fluid IGFBP-3, IGFBP-1, and IGFBP-4 levels were significantly higher in GH cycles than in matched placebo cycles. The IGFBP-3 and IGFBP-1 responses were confirmed by RIA. Follicular fluid IGF-I levels also were consistently higher in GH than placebo cycles, confirming previous findings in poor responders (30, 31). By RIA, the mean relative increase in IGF-I was greater than that of IGFBP-3. Although the ratio of IGF-I to total IGFBPs may more accurately reflect IGF-I bioactivity, it is difficult to
1174
RABINOVICI ET AL.
JCE & M • 1997 Vol 82 • No 4
FIG. 2. Follicular fluid IGFBP-3, IGFBP-1, and IGF-I levels by RIA. Panel A, Follicular fluid samples from four patients were subjected to RIA for IGFBP-3 (41). Shown are the mean 6 SD IGFBP-3 levels in follicular fluid samples from each patient (A–D) and combined data from all patients (ALL), obtained in GH (solid bars) and placebo (shaded bars) cycles. Panel B, The same samples subjected to immunoradiometric assay for IGFBP-1, labeled as in panel A; panel C, the same samples subjected to acid-ethanol extraction and RIA for IGF-I (see Ref. 42), labeled as in panel A; panel D, correlation between follicular fluid levels of IGF-I and IGFBP-3 by RIA.
estimate this ratio from ligand blots because the relative contribution of each IGFBP cannot be determined. The correlation between follicular fluid IGFBP-3 and IGF-I levels is consistent with the observed stimulation by GH of serum levels of both proteins and suggests that both proteins in follicular fluid may be derived mainly from the circulation. In further support of this hypothesis, mean IGF-I levels in follicular fluid were lower than those in serum, as noted previously (1, 44). Because some IGF-I of thecal origin (2) may act on the granulosa without reaching the antral compartment, follicular fluid IGF-I levels may not accurately reflect IGF-I action on the granulosa. It is uncertain what role GH plays in regulating production of each IGFBP. IGFBP-1 levels in follicular fluid greatly exceed those in serum (11). Granulosa cells produce IGFBP-1, with a large increase after luteinization (8, 17). The increase in follicular fluid IGFBP-1 with GH may result from direct stimulation of IGFBP-1 production (6) or from accelerated luteinization mediated by increased IGF-I. IGFBP-3, the major serum carrier of IGFs, is produced principally in the liver under positive modulation by GH (45). Our finding that GH stimulates both follicular fluid and serum IGFBP-3 supports the hypothesis that follicular IGFBP-3 is derived largely from the circulation and that its increase in GH cycles reflects a hepatic action of GH. Published reports differ on whether human granulosa cells produce IGFBP-3 (1, 15–17). IGFBP-4 messenger RNA and protein are produced by human granulosa cells (1, 15–17). Whereas no effect of GH on granulosa cell IGFBP-4 accumulation was noted (15), follicular fluid IGFBP-4 levels in spontaneous cycles are regulated by a protease (46), and the effect of GH treatment on follicular fluid IGFBP-4 in the present study may reflect decreased protease activity. In a double-blind study of the effects of GH on IGFBPs in
poor responders, Huang et al. noted that GH increased serum IGFBP-3 and IGF-I in parallel. Follicular fluid IGF-I, but not IGFBP-3, was significantly increased by GH, and follicular fluid IGF-I and IGFBP-3 were correlated only in placebotreated, not in GH-treated patients. GH did not affect the follicular fluid IGFBP profile, as determined by ligand blotting (30, 38). In another study, serum IGF-I and IGFBP-3 and follicular fluid IGF-I rose with GH treatment, but neither IGFBP-3 nor IGFBP-1 in follicular fluid was significantly altered (31). Our results may differ from these findings because GH may have a greater effect on follicular fluid IGFBP levels in normally ovulatory women than in poor responders, or possibly because of differences in the IGFBP-3 assay. Although the stimulation of IGF-I and IGFBPs confirms the endocrine effects of adjunctive GH, this treatment did not produce significant clinical benefits. Although we cannot exclude a Type II statistical error because of our small patient number, our results are consistent with the only other reported randomized, blinded study of GH in unselected patients with mechanical factor infertility undergoing a similar IVF stimulation protocol (35). We propose that the lack of clinical efficacy of GH in most studies could result from the parallel increases in IGFBPs, along with IGF-I, which may blunt IGF and gonadotropin bioactivity. The significant interpatient differences in the follicular fluid IGFBP-3 responses to GH may additionally offer an explanation for the differences in clinical response to GH among published series. It is possible that only some, and not all poor responders to hMG (for example, those with decreased GH reserve (47) or an underlying ovulatory disorder) may benefit from adjunctive GH, and further studies to define this subgroup would be of interest. It is also possible that differences in intrafollicular IGF-I levels and availability may not result in
GH INCREASES OVARIAN FOLLICULAR IGFBPs
1175
action of GH on intrafollicular IGFBP levels, but the parallel increases of intrafollicular and circulating IGF-I and IGFBP-3 suggest that these GH effects are mediated at the hepatic level. References
FIG. 3. Serum IGFBP-3 and IGF-I levels by RIA throughout the stimulation cycle. Panel A, The mean 6 SD serum IGFBP-3 levels from all patients for whom paired cycles were available at each of four time points in the treatment cycle: Basal, after pituitary down-regulation with GnRH agonist but before hMG; Day 4, the 4th day of hMG; Day 6, the 6th day of hMG; hCG, the day of hCG administration. GH treatment resulted in higher IGFBP-3 levels (P , 0.006, two-way ANOVA). No differences were found in serum IGFBP-3 levels among the three time points in the cycle after hMG 6 GH was started. Panel B, The mean 6 SD serum IGF-I levels at the same four time points. GH treatment resulted in higher IGF-I levels (P , 0.0001, two-way ANOVA). No differences were found in serum IGF-I levels among the three time points in the cycle after hMG 6 GH were started. TABLE 2. Clinical outcome of study cycles
Number of subjects Age (yr) Days (hMG) Ampules (hMG) Peak E2 (pg/mL) Oocytes retrieved % Mature oocytes % Fertilization % Cleavage Embryos transferred Pregnancies Term pregnancies
GH
Placebo
P
7 37.7 6 2.4 10.6 6 1.3 36.6 6 12.7 2065 6 720 9.6 6 6.6 73 6 23 63 6 33 92 6 10 5.0 6 3.2 1 0
7 38.4 6 1.6 11.0 6 1.2 39.3 6 10.6 2275 6 599 9.7 6 2.8 68 6 21 75 6 19 96 6 7 5.7 6 1.0 0 0
0.35 0.53 0.43 0.16 0.94 0.68 0.20 0.17 0.58 1.00
Shown are the stimulation, retrieval, and embryologic outcomes of IVF cycles with adjunctive GH, compared with placebo cycles in the same patients. Mean differences are expressed as the value in the GH cycle minus that in the same patient’s placebo cycle. There were no statistically significant differences between treatments in any of the outcome variables shown.
clinically significant differences in responsiveness to hMG stimulation or IVF. In summary, adjunctive GH treatment of normally cycling women during ovarian hyperstimulation with hMG produces a net increase in follicular fluid IGFBPs, as well as IGF-I. Our data are consistent with either a direct or indirect
1. Giudice LC. 1992 Insulin-like growth factors and ovarian follicular development. Endocr Rev. 13:641– 669. 2. El-Roeiy A, Chen X, Roberts V, LeRoith D, Roberts CJ, Yen S. 1993 Expression of insulin-like growth factor-I (IGF-I) and IGF-II and the IGF-I, IGF-II, and insulin receptor genes and localization of the gene products in the human ovary. J Clin Endocrinol Metab. 77:1411–1418. 3. Tamura M, Sasano H, Suzuki T, et al. 1994 Immunohistochemical localization of growth hormone receptor in cyclic human ovaries. Hum Reprod. 9:2259 –2262. 4. Mason H, Martikainen H, Beard R, Anyaoku V, Franks S. 1990 Direct gonadotrophic effect of growth hormone on oestradiol production by human granulosa cells in vitro. J Endocrinol. 126:R1–R4. 5. Barreca A, Artini P, Del MP, et al. 1993 In vivo and in vitro effect of growth hormone on estradiol secretion by human granulosa cells. J Clin Endocrinol Metab. 77:61– 67. 6. Ovesen P, Ingerslev H, Ørskov H, Ledet T. 1994 Effect of growth hormone on steroidogenesis, insulin-like growth factor-I (IGF-I) and IGF-binding protein-1 production and DNA synthesis in cultured human luteinized granulosa cells. J Endocrinol. 140:313–319. 7. Rechler MM. 1993 Insulin-like growth factor binding proteins. Vitam Horm. 47:1–91. 8. Seppa¨la¨ M, Wahlstrom T, Koskimies A, et al. 1984 Human preovulatory follicular fluid, luteinized cells of hyperstimulated preovulatory follicles, and corpus luteum contain placental protein 12. J Clin Endocrinol Metab. 58:505–510. 9. Jalkanen J, Suikkari AM, Koistinen R, et al. 1989 Regulation of insulin-like growth factor-binding protein-1 production in human granulosa-luteal cells. J Clin Endocrinol Metab. 69:1174 –1179. 10. Giudice L, Farrell E, Pham H, Rosenfeld R. 1990 Identification of insulin-like growth factor-binding protein-3 (IGFBP-3) and IGFBP-2 in human follicular fluid. J Clin Endocrinol Metab. 71:1330 –1338. 11. Hartshorne GM, Bell SC, Waites GT. 1990 Binding proteins for insulin-like growth factors in the human ovary: identification, follicular fluid levels and immunohistological localization of the 29 –32 kd type 1 binding protein, IGFbp1. Hum Reprod. 5:649 – 660. 12. Dor J, Costritsci N, Pariente C, et al. 1992 Insulin-like growth factor-I and follicle-stimulating hormone suppress insulin-like growth factor binding protein-1 secretion by human granulosa-luteal cells. J Clin Endocrinol Metab. 75:969 – 671. 13. Cataldo N, Giudice L. 1992 Insulin-like growth factor binding protein profiles in human ovarian follicular fluid correlate with follicular functional status. J Clin Endocrinol Metab. 74:821– 829. 14. San Roman GA, Magoffin DA. 1992 Insulin-like growth factor binding proteins in ovarian follicles from women with polycystic ovarian disease: cellular source and levels in follicular fluid. J Clin Endocrinol Metab. 75:1010 –1016. 15. Cataldo N, Woodruff T, Giudice L. 1993 Regulation of insulin-like growth factor binding protein production by human luteinizing granulosa cells cultured in defined medium. J Clin Endocrinol Metab. 76:207–215. 16. Zhou J, Bondy C. 1993 Anatomy of the human ovarian insulin-like growth factor system. Biol Reprod. 48:467– 482. 17. El-Roeiy A, Chen X, Roberts V, et al. 1994 Expression of the genes encoding the insulin-like growth factors (IGF-I and II), the IGF and insulin receptors, and IGF-binding proteins 1– 6 and the localization of their gene products in normal and polycystic ovary syndrome ovaries. J Clin Endocrinol Metab. 78:1488 –1496. 18. Iwashita M, Adachi T, Katayama E, Kudo Y, Takeda Y. 1994 Regulation and physiological role of insulin-like-growth-factor-binding protein-1 in human granulosa cells. Horm Res. [Suppl 1]41:22–28. 19. Bicsak T, Shimonaka M, Malkowski M, Ling N. 1990 Insulin-like growth factor-binding protein (IGF-BP) inhibition of granulosa cell function: effect on cyclic adenosine 39,59-monophosphate, deoxyribonucleic acid synthesis, and comparison with the effect of an IGF-I antibody. Endocrinology. 126:2184 –2189. 20. Angervo M, Koistinen R, Suikkari A, Seppa¨la¨ M. 1991 Insulin-like growth factor binding protein-1 inhibits the DNA amplification induced by insulinlike growth factor I in human granulosa-luteal cells. Hum Reprod. 6:770 –773. 21. Mason H, Willis D, Holly J, Cwyfan-Hughes S, Seppa¨la¨ M, Franks S. 1992 Inhibitory effects of insulin-like growth factor-binding proteins on steroidogenesis by human granulosa cells in culture. Mol Cell Endocrinol. 89:R1–R4. 22. Homburg R, Eshel A, Abdalla H, Jacobs H. 1988 Growth hormone facilitates ovulation induction by gonadotrophins. Clin Endocrinol. 29:113–117. 23. Homburg R, West C, Torresani T, Jacobs H. 1990 Cotreatment with human
1176
24.
25. 26.
27.
28. 29.
30.
31. 32.
33.
34. 35.
RABINOVICI ET AL.
growth hormone and gonadotropins for induction of ovulation: a controlled clinical trial. Fertil Steril. 53:254 –260. Blumenfeld Z, Amit T, Barkey R, Lunenfeld B, Brandes J. 1991 Synergistic effect of growth hormone and gonadotropins in achieving conception in “clonidine-negative” patients with unexplained infertility. Ann NY Acad Sci. 626:250 –265. Volpe A, Coukos G, Barreca A, et al. 1989 Ovarian response to combined growth hormone-gonadotropin treatment in patients resistant to induction of superovulation. Gynecol Endocrinol. 3:125–133. Ibrahim Z, Matson P, Buck P, Lieberman B. 1991 The use of biosynthetic human growth hormone to augment ovulation induction with buserelin acetate/human menopausal gonadotropin in women with a poor ovarian response. Fertil Steril. 55:202–204. Owen E, Shoham Z, Mason B, Ostergaard H, Jacobs H. 1991 Cotreatment with growth hormone, after pituitary suppression, for ovarian stimulation in in vitro fertilization: a randomized, double-blind, placebo-control trial. Fertil Steril. 56:1104 –1110. Owen E, West C, Mason B, Jacobs H. 1991 Co-treatment with growth hormone of sub-optimal responders in IVF-ET. Hum Reprod. 6:524 –528. Hughes S, Huang Z, Matson P, Buck P, Lieberman B, Morris I. 1992 Clinical and endocrinological changes in women following ovulation induction using buserelin acetate/human menopausal gonadotrophin augmented with biosynthetic human growth hormone. Hum Reprod. 7:770 –775. Huang Z, Baxter R, Hughes S, Matson P, Lieberman B, Morris I. 1993 Supplementary growth hormone treatment of women with poor ovarian response to exogenous gonadotrophins: changes in serum and follicular fluid insulinlike growth factor-1 (IGF-1) and IGF binding protein-3 (IGFBP-3). Hum Reprod. 8:850 – 857. Bergh C, Hillensjo¨ T, Wikland M, Nilsson L, Borg G, Hamberger L. 1994 Adjuvant growth hormone treatment during in vitro fertilization: a randomized, placebo-controlled study. Fertil Steril. 62:113–120. Hughes S, Huang Z, Morris I, Matson P, Buck P, Lieberman B. 1994 A double-blind cross-over controlled study to evaluate the effect of human biosynthetic growth hormone on ovarian stimulation in previous poor responders to in-vitro fertilization. Hum Reprod. 9:13–18. Dor J, Seidman D, Amudai E, Bider D, Levran D, Mashiach S. 1995 Adjuvant growth hormone therapy in poor responders to in-vitro fertilization: a prospective randomized placebo-controlled double-blind study. Hum Reprod. 10:40 – 43. Levron J, Lewit N, Erlik Y, Itskovitz-Eldor J. 1993 No beneficial effects of human growth hormone therapy in normal ovulatory patients with a poor ovarian response to gonadotropins. Gynecol Obstet Invest. 35:65– 68. Younis J, Simon A, Koren R, Dorembus D, Schenker J, Laufer N. 1992 The effect of growth hormone supplementation on in vitro fertilization outcome: a
36.
37. 38. 39.
40.
41.
42. 43. 44. 45. 46.
47.
JCE & M • 1997 Vol 82 • No 4
prospective randomized placebo-controlled double-blind study. Fertil Steril. 58:575–580. Tapanainen J, Martikainen H, Voutilainen R, Orava M, Ruokonen A, Ro¨nnberg L. 1992 Effect of growth hormone administration on human ovarian function and steroidogenic gene expression in granulosa-luteal cells. Fertil Steril. 58:726 –732. Levy T, Limor R, Villa Y, et al. 1993 Another look at co-treatment with growth hormone and human menopausal gonadotrophins in poor ovarian responders. Hum Reprod. 8:834 – 839. Huang Z, Matson P, Lieberman B, Morris I. 1994 Insulin-like growth factor binding proteins in serum and follicular fluid from women undergoing ovarian stimulation with and without growth hormone. Hum Reprod. 9:1421–1426. Ross J, Olson K, Geier M, O’Connor J, Jones A. 1986 Recombinant DNA synthesis of human growth hormone. In: Raiti S, Tolman R, eds. Human growth hormone. New York and London: Plenum Medical Book Company, 241–256. Hossenlopp P, Seurin D, Segovia-Quinson B, Hardouin S, Binoux M. 1986 Analysis of serum insulin-like growth factor binding proteins using Western blotting: use of the method for titration of the binding proteins and competitive binding studies. Anal Biochem. 154:138 –143. Gargosky S, Pham H, Wilson K, Liu F, Giudice L, Rosenfeld R. 1992 Measurement and characterization of insulin-like growth factor binding protein-3 in human biological fluids: discrepancies between radioimmunoassay and ligand blotting. Endocrinology. 131:3051–3060. Breier BH, Gallaher BW, Gluckman PD. 1991 Radioimmunoassay for insulinlike growth factor I: solutions to some potential problems and pitfalls. J Endocrinol. 128:347–357. Rosenfeld RG, Pham H, Oh Y, Lamson G, Giudice LC. 1990 Identification of insulin-like growth factor-binding protein-2 (IGF-BP-2) and a low molecular weight IGF-BP in human seminal plasma. J Clin Endocrinol Metab. 70:551–553. Rabinovici J, Dandekar P, Angle M, Rosenthal S, Martin M. 1990 Insulin-like growth factor I (IGF-I) levels in follicular fluid from human preovulatory follicles: correlation with serum IGF-I levels. Fertil Steril. 54:428 – 433. Baxter R, Martin J. 1986 Radioimmunoassay of growth hormone-dependent insulin-like growth factor binding protein in human plasma. J Clin Invest. 78:1504 –1512. Chandrasekher YA, Van Dessel HJHM, Fauser BCJM, Giudice LC. 1995 Estrogen- but not androgen-dominant human ovarian follicular fluid contains an insulin-like growth factor binding protein-4 protease. J Clin Endocrinol Metab. 80:2734 –2739. Menashe Y, Lunenfeld B, Pariente C, Frenkel Y, Mashiach S. 1990 Can growth hormone increase, after clonidine administration, predict the dose of human menopausal hormone needed for induction of ovulation? Fertil Steril. 53:432– 435.
