Pulsatile intravenous gonadotropin-releasing hormone for ovulation-induction in infertile women. II. Analysis of follicular and luteal phase responses*

Vol. 48, No. 1, July 1987

FERTILITY AND STERILITY Copyright " 1987 The American Fertility Society

Printed in U.S.A.

Pul sati l e i ntravenous gonadotropi n - re l eas i n g hormone for ovu lati o n - i n d ucti on in i n fe rti l e women. I I . Analys i s of fo l l icular and l uteal phase responses*

Robert P. S. Jansen, F.R.A.C.P., F.R.A.C.O.G.t* David J. Handelsman, Ph.D., F.R.A.C.P.§ Lyn M. Boylan, R.N.§ Ann Conway, F.R.A.C.P.§

Rodney P. Shearman, F.R.A.C.O.G.* Ian S. Fraser, F.R.A.C.O.G.* John C. Anderson, F.R.A.C.O.G.* II

Royal Prince Alfred Hospital, Camperdown, Sydney, Australia

Pulsatile intravenous gonadotropin releasing hormone (IV-GnRH) was used in 36 in­ fertile patients with primary amenorrhea (n = 5), secondary amenorrhea due to hypotha­ lamic chronic anovulation (HCA) (n = 22), hyperprolactinemia (n = 1) or polycystic ovary syndrome (PCOS) (n = 5), and oligomenorrhea (n = 3), using several dosage and timing regimens. Early follicular phase responses showed four patterns: type 1 consisted of a delayed follicle-stimulating hormone (FSH) peak and was seen with severe hypothalamic suppression (n = 4); type 2 consisted of a brisk and dominant FSH peak on the first day of treatment, and occurred with mild to moderate hypothalamic suppression (n = 19); type 3, which consisted of an FSH peak accompanied by an immediate and exaggerated luteiniz­ ing hormone (LH) rise, occurred with mild PCOS and some cases of HCA (n = 5); and type 4, in which LH levels were high to begin with and neither FSH nor LH levels rose with GnRH, occurred with severe PCOS (n = 2). Exaggerated estradiol responses within 24 hours of therapy were seen in eight cycles: in four cases no ovarian abnormality was apparent; in three cases a dominant follicle was already present; and in one case ovarian hyperstimulation was diagnosed ultrasonographically. With standard human chorionic gonadotropin luteal phase support, luteal phase defects were rare with HCA but common with PCOS. Fertil Steril 48:39, 1987

When patients with hypothalamic hypogonad­ ism are selected for treatment, follicular develop­ ment, ovulation, and a high expectation of preg­ nancy can accompany intravenous (IV) ad­ ministration of gonadotropin-releasing hormone (GnRH), regardless of the degree of hypothalamic

Received May 6, 1986; revised and accepted March 30, 1987. * Presented in part at the Seventh International Congress of Endocrinology, Abstract 897, Quebec City, Quebec, Canada, July 1 to 7, 1984. t Reprint requests : Robert P . S. Jansen, F . R . A. C . P . , F.R.A.C .O.G., The Fertility Laboratory, Royal Prince Alfred Hospital, Camperdown, Sydney 2050; Australia. :j: Department of Obstetrics and Gynaecology. § Department of Endocrinology. II Department of Diagnostic Ultrasound. Vol. 48, No. 1, July 1987

suppression (see accompanying article 1 for refer­ ences) . Pulsatile GnRH treatment thus offers a useful alternative to pituitary gonadotropin ther­ apy in patients with hypothalamic chronic anovu­ lation (HCA) resistant to clomiphene citrate (CC) . The role of pulsatile GnRH in other causes of anovulation, particularly the polycystic ovary syn­ drome (PCOS) , is less clear, because response to GnRH is unpredictable. 1 - 1 ° For example, in the pre­ ceding article, 1 we found an ovulation rate with IV­ GnRH of 97.5% for HCA compared with 50.0% for PCOS, and similar discrepancies recently have been reported for subcutaneous (SC) GnRH ther­ apy by Saffan and Seibel. 10 Thirty-six anovulatory infertile patients unre­ sponsive to CC were treated with pulsatile IV­ GnRH and their follicular and luteal phase reJansen et al.

Pulsatile GnRH ovulation-induction

39

sponses were analyzed. Nineteen patients with sec­ ondary amenorrhea due to HCA or PCOS were studied in detail to determine the basis for the fail­ ure of response that may be seen in patients with PCOS.

MATERIALS AND METHODS Gonadotropin-Releasing Hormone Doses and Regimens

GnRH (Gonadorelin, Relefact) was bought from Ayerst Laboratories Pty. Ltd. (Sydney, Australia) and Hoechst Australia Ltd. (Sydney, Australia) and was administered as previously described. 1 A pulse interval of 90 minutes was generally used, except that ten patients with secondary amenorrhea were randomized to receive GnRH at either 60- or 120-minute intervals for the first week of therapy of their first treatment cycle. In these patients, and in nine patients with 90-minute regi­ mens, measurements of serum follicle-stimulating hormone (FSH) and luteinizing hormone (LH) were made every 10 minutes from 20 minutes be­ fore until 20 minutes after a GnRH pulse, from the first to the fourth day of treatment. Other patients had daily measurements of serum FSH and LH, as well as estradiol (E 2 ) and progesterone (P), from initiation of therapy. Therapeutic protocols were approved by the Ethics Review Committee of Royal Prince Alfred Hospital, possible hazards of therapy were ex­ plained to all patients, and informed consent was obtained. Hormone Assays

Serum hormones were measured by the following standard radioimmunoassays: F S H ( B i o - RIA, Montreal, Canada), expressed in miU/ml, Medical Research Council (MRC) reference preparation 69/104; within-assay variability (WAV) < 5 % ; be­ tween-assay variability (BAV) = 4.3% at 8.2 miU/ ml; LH (Bioclone, Sydney, Australia) , MRC 68/40; WAV < 5 % ; BAV < 6.7%; P (Farmos, Turku, Fin­ land) , WAV < 5 % ; BAV < 7.7%; E 2 (Diagnostic Products, Los Angeles, CA) , WAV < 5 % ; BAV < 6.8% ; testosterone (T) (Mallinkrodt, Sydney, Australia) WAV < 5 % ; BAV < 8%; normal 0. 7 to 3.3 nmol/1 or 0.2 to 0.95 ng/ml. The detection limit of FSH and LH was 1.0 miU /ml. Serum unbound T (free-T; normal < 60 pmol/1 or 1 7 pg/ml) was as­ sayed after ultrafiltration. 40

Jansen et al.

Pulsatile GnRH ovulation-induction

Patients

Hypothalamic Hypogonadism

Hypothalamic hypogonadism (HH), with sexual infantilism prior to endocrine therapy, was diag­ nosed in four patients. Two of three patients with isolated gonadotropin deficiency were anosmic (Kallmann's syndrome ) . One patient was obese, but showed no other endocrine deficiency. The fourth patient, with short stature and delayed pu­ berty, had had a diagnosis of primary hypothyroid­ ism made in childhood. Hypothalamic Chronic Anovulation

Hypothalamic chronic anovulation (HCA) was diagnosed in 22 patients on the basis of low serum FSH and LH levels, serum T levels < 2 nmol/1 (0.58 ng/ml) and normal serum prolactin (PRL) concen­ trations. Three patients gave a history of previous hospital admission for anorexia nervosa, had par­ tially regained normal body weight, and were con­ sidered suitable for ovulation induction after psy­ chiatric assessment. Seven other patients were below optimal weight for height according to a Quetelet index (QI = weight [kg] /height2 [m2 ] 11) < 20; six displayed evidence of eating disorder and one was a marathon runner. Eleven patients were of normal weight for height (QI < 25) and one was above optimal weight. None had used oral contra­ ceptives for at least 12 months before treatment. Polycystic Ovary Syndrome

Polycystic ovary syndrome (PCOS) or PCO-like syndrome was diagnosed in five patients, all with secondary amenorrhea of more than 12 months' duration, on the basis of clinical evidence of hyper­ androgenism (acquired hirsutism and/or acne) ac­ companied by objective evidence of bilateral ovar­ ian enlargement; these patients all had a random serum T level > 2 nmol/1. All except one were of normal weight for height. Hyperprolactinemia

Hyperprolactinemia (serum PRL range, 60 to 90 ng/ml) and a high serum T level (5.5 nmol/1) coex­ isted in one patient. Although ovulatory cycles and a laparoscopically documented corpus luteum ap­ parently had occurred with bromocriptine therapy up to 60 mg/day, serum PRL had not fallen below 60 ng/ml. Transphenoidal pituitary fossa explora­ tion for a suspected microadenoma had been negaFertility and Sterility

tive. Serum-free T was considerably elevated at 122 pmol/1 (normal, 0 to 60) . GnRH therapy was cho­ sen because of lack of results with bromocriptine therapy, which was discontinued during two cycles of GnRH therapy.

m i U/ 1 20

A

10 0

RESULTS Follicular Phase Analysis: Gonadotropin Levels

Nineteen patients had measurements of serum FSH and serum LH every 10 minutes for 20 min­ utes before treatment (day 0) and for 40 minutes during administration of a GnRH pulse on days 1 to 4 of treatment. Four patterns of response were recognized (Fig. 1) and daily blood samples on re­ maining patients in their first treatment cycle al­ lowed classification of all responding patients into one of these response types. A type 1 response (Fig. 1A) was seen in patients with substantial hypothalamic suppression, in­ cluding those with primary amenorrhea and HH; several days elapsed before serum FSH response was maximal. This response was seen in 2 of 4 patients with HH and 2 of 22 patients with HCA, but was not seen with PCOS or luteal phase defi­ ciency (LPD). All patients with this response ovu­ lated consistently. Mean follicular phase length was 16.0 ± 1 . 1 days (±SEM). A type 2 response (Fig. 1B) was brisker, with peak FSH levels reached after only 24 hours GnRH therapy. LH response was greater or less than FSH response, but the LH/FSH ratio remained below 2. This response was seen in 2 of 4 patients with HH, 16 of 22 patients with HCA, 1 patient with PCOS, and in the patient with hyperprolactinemia. Each patient with this response ovulated consistently with therapy. Mean follicular phase length was 12.8 ± 0.4 days, significantly shorter than with type 1 responses (P = 0.01, t-test) . A type 3 response (Fig. 1 C ) , in which there was evidence of LH fluctuation unaccompanied by FSH changes before therapy but in which baseline FSH and LH levels were otherwise normal, showed an immediately exaggerated LH rise in response to therapy. De spite this elevation, gonadotropin levels, especially LH, rose further in response to exogenous GnRH pulses. This response was seen in two patients with PCOS and. three patients with HCA, all of whom ovulated. Mean follicular phase length was 12.5 ± 0.61 days, not significantly dif­ ferent from the type 2 response. A type 4 response (Fig. 1D) was characterized by Vol. 48, No. 1, July 1987

t

�

t

� � :::: � � ......

L.U

L.U..LJ

L.U..LJ

L.L.LJ..J

L.L.LJ..J

20

B

10 0 20

c

10 0

......

M��� L.LJ

D

L..L.I

30 20 10 0

� ... L.J...I

d ay

L.L.LJ..J

L.U..LJ

U-!..J..J

0

L.L.LJ..J

L.U..LJ

�v ...... ..... L.U..LJ

L...L.L..LJ

u..J..J..I

� ..

u..J..J..I

u...u.J

L.I...L.L.J

2

3

4

Patte rns of serum FSH ( e e) and LH 0) response to GnRH therapy estimated at 10-minute

Figure 1 -

�

J � ] J G, ....- ..... ..... .....

40

(0

�

-

intervals in relation to IV -GnRH pulses in amenorrheic patients resistant to CC. (A) HCA without detectable baseline gonado­ tropin activity; serum FSH response maximal after 72 hours of GnRH therapy. (B) HCA with detectable baseline gonadotropin activity; serum FSH response maximal after 24 hours of GnRH therapy. (C) Mild PCOS with ovulatory cycles on therapy, showing normal baseline FSH /LH ratio but with evidence of endogenous LH pulses and an immediately exaggerated LH re­ sponse to GnRH. (D) severe PCOS with elevated baseline LH/ FSH ratio and evidence on day 1 of endogenous LH pulse unre­ lated to exogenous GnRH.

endogenous LH overactivity and a substantial tonic rise in LH levels, but there was no acute im­ pact of exogenous GnRH pulses on LH or FSH levels. This response was seen in two of five PCOS Jansen et al.

Pulsatile GnRH ovulation-induction

41

patients, neither of whom ovulated in response to GnRH. Two gonadotropin patterns were seen in unre­ sponsive patients. A patient with primary amenor­ rhea, short stature, obesity, and a small pituitary fossa on roentgenography (features suggestive of the Lawrence-Moon-Biedl syndrome, but without retinitis pigmentosa) , showed no response to ther­ apy, despite measurable baseline levels of FSH and LH (doses up to 100 �g/pulse at 120-minute inter­ vals were used) . A patient aged 40 years with a LPD developed postmenopausal elevations of FSH and LH within 3 days of commencement of therapy: gonadotropin levels were still high 6 months later. Neither the duration, the magnitude of the initial FSH rise, nor the length of the follicular phase seemed to depend on the dose or pulse frequency of IV -GnRH. Initial E 2 levels also were unrelated to subsequent FSH response. Follicular Phase Analysis: Exaggerated Estradiol Responses

In eight cycles, a rapid rise of E 2 occurred within 24 hours of initiation of therapy. An exaggerated E 2 response was defined as levels increasing more than 2-fold every 24 hours, over at least 2 days, and was classified into three circumstances. A type A exaggerated E 2 response was unaccom­ panied by sonographic indications of abnormal ovarian reaction, and GnRH therapy was contin­ ued as E 2 levels declined spontaneously, to be fol­ lowed later by ovulation, in most cases, of a single follicle. This response was seen in four patients: one patient with HCA and a type 2 gonadotropic pattern and three patients with PCOS and a type 3 (n = 2) or type 4 gonadotropic pattern; in each case, sonography showed no abnormal ovarian enlarge­ ment, GnRH was continued, and E 2 levels fell spontaneously before rising again prior to ovula­ tion. A type B exaggerated E 2 response occurred in association with the early sonographic appearance of a single dominant follicle and was followed within several days by an LH surge and ovulation. This response was seen in two patients with a type 2 gonadotropic pattern and in one patient with a type 3 pattern. A type C exaggerated E 2 response was considered to be indicative of hyperstimulation, with simulta­ neous development of multiple follicles and ovar­ ian enlargement on sonography. One suspected type C response occurred in a patient with PCOS 42

Jansen et al.

Pulsatile GnRH ovulation-induction

and a type 4 gonadotropic pattern: further GnRH was withheld. Luteal Phase Analysis

Luteal phase length and midluteal serum P levels correlated better with diagnosis than with early follicular gonadotropin response patterns. Mean luteal length was 16.8 ± 0.8 days for HH, 16.5 ± 0.6 days for HCA (not significant), and 14.4 ± 1 .8 days for PCOS (P < 0.05, t-test) . Midluteal serum P levels were below 2 standard deviations below the population mean in 2 of 10 ovulatory cycles in PCOS (mean P 43.3 nmol/1) , compared with 0 of 67 ovulatory cycles in HCA (mean P 66.2 nmol/1, P < 0.02, t-test). Mean midluteal P levels and luteal lengths accompanying type 1, type 2, and type 3 responses, on the other hand, were not significantly different. DISCUSSION

Low endogenous GnRH is assumed to underlie the amenorrhea of HH 1 2 and HCA, 13- 1 5 including exercise and weight-loss amenorrhea/6• 17 and an­ orexia nervosa. 18 •19 When patients with amenor­ rhea due to absent or reduced endogenous GnRH activity are selected for treatment, high ovulation rates result from almost all IV -GnRH and most SC-GnRH regimens. 1 •3 • 20 In PCOS, however, evidence indicates that en­ dogenous GnRH production is probably increased in frequency and amount as a manifestation of heightened positive estrogen feedback. 2 1 - 2 4 The grounds for expecting a response to further GnRH administered exogenously would be the hope that short-cycle periodic or tonically elevated endoge­ nous GnRH would be suppressed or overridden by slower exogenous pulses. 6 The empirical basis for this hope is limited. In various series, ovulation rates have been well below 90% whenever the diag­ nosis of PCOS has been implied by the presence of endogenous LH pulses,17 a high baseline LH/FSH ratio,5-8• 10•19 or the existence of multiple small cysts < 4 mm diameter in association with stromal thickening on ultrasound scanning of the ovaries. 3 Bringer et al.,8 for example, found that, with IV­ GnRH, ovulation occurred in 85% HCA cycles, in 50% PCOS cycles, and in 4 1 % cycles in an inter­ mediate group characterized by high baseline LH/ FSH ratios-results similar to ours. 1 We did not find elevated androgens per se to associate with a poor response to GnRH therapy: one such patient ovulated normally twice and conFertility and Sterility

ceived despite an elevated free-T level more than twice normal. Conversely, those patients with ele­ vated baseline serum LH activity over-responded to exogenous GnRH therapy, producing large ele­ vations of serum LH in comparison with FSH (Fig. 1 ) . When baseline LH levels were moderately ele­ vated, response to GnRH occurred and cycles were generally ovulatory, though with variable luteal phases (see subsequent discussion) . When baseline LH levels were considerably elevated, the exagger­ ated tonic response to GnRH was unaccompanied by an immediate LH response to treatment pulses, and not one of these cycles was ovulatory. Al­ though we conclude that hypothalamic factors rather than direct ovarian effects of androgens un­ derlie the refractoriness of PCOS to GnRH ther­ apy, these observations have not been tested pro­ spectively and should be regarded as tentative pre­ dictions of response in PCOS. The length of the follicular phase with treatment seems to depend less on the dose or pulse frequency of IV -GnRH than it does on the degree of gonado­ tropic suppression at the start of therapy. 20 After averaging follicular phase lengths for each patient, the median time from initiation of treatment to ovulation was 12.5 days for secondary hypotha­ lamic anovulation, but 19.5 days for primary amen­ orrhea (P < 0.025 ) . Neither the pulse frequency nor the pulse dose appeared to influence the time course of early FSH response or follicular phase length. Peak serum FSH levels (reached before ovarian response produced negative feedback on FSH levels) were most often found a day after starting therapy, but when initial gonadotropin levels were especially low and E 2 responses were delayed, peak FSH levels occurred on the third or fourth day. Initial E 2 levels were unrelated to subsequent FSH response, in contrast to other reports. 25 In six cycles, acute rises of serum E 2 to four to ten times basal levels within 24 hours of therapy did not stop satisfactory FSH responses and subsequent follicu­ lar development and ovulation. No ultrasound evi­ dence of rapid follicular growth or hyperstimula­ tion accompanied these E 2 rises. GnRH therapy was stopped once ovulation had been confirmed. Luteal function was stimulated with hCG. Other studies have shown that pulsatile GnRH can be continued through the cycle and lu­ teal function maintained,6•26 but this offers no clin­ ical advantage and is less convenient for patients. Mean luteal phase length was the same for primary amenorrhea with HH ( 16.5 days) and secondary Vol. 48, No. 1, July 1987

amenorrhea with HCA ( 1 6.5 days) , but shorter ( 14.4 days) for PCOS (P < 0.05). Although single midluteal P determinations give only a rough indi­ cation of luteal phase adequacy, mean midluteal P concentrations were significantly lower among ovulatory cycles in PCOS (43.3 nmol/1) than in HCA (66.2 nmol/1) (P < 0.02 ) . Although hyperstimulation syndrome has been reported after GnRH therapy, 2 7 our study shows that not all acute elevations of serum E 2 with ther­ apy are attributable to it. We found that an imme­ diate and rapid E 2 rise: ( 1 ) could also reflect stimu­ lation and growth of a pre-existing single dominant follicle, or (2) could be transient. The diagnosis of hyperstimulation, which implies a need to suspend GnRH treatment, should therefore be made only if there is ultrasound confirmation of ovarian en­ largement due to multiple follicular development. We conclude that IV -GnRH is safe and predict­ ably effective for ovulation induction in women presumed to have GnRH deficiency, whereas the role of GnRH in PCOS is doubtful, probably be­ cause exogenous GnRH must compete with inap­ propriate endogenous GnRH in this disorder. REFERENCES 1 . Jansen RPS, Handelsman DJ, Boylan LM, Conway A, Shearman RP, Fraser IS: Pulsatile intravenous gonadotro­ pin-releasing hormone for ovulation-induction in infertile women. I. Safety and effectiveness with outpatient therapy. Fertil Steril 48:33, 1987 2. Coelingh Bennink HJT, Weber HW, Alsbach HPJ, Thijssen JHH: Induction of ovulation by pulsatile intravenous ad­ ministration of GnRH in polycystic ovarian disease (Abstr). Fertil Steril 41:34s, 1984 3. Mason P, Adams J, Morris DV, Tucker M, Price J, Voul­ garis Z, van der Spuy ZM, Sutherland I, Chambers GR, White S, Wheeler MJ, Jacobs HS: Induction of ovulation with pulsatile luteinising hormone releasing hormone. Br Med J 288:181, 1984 4. Loucopoulos A, Ferin M, Vande Wiele RL, Dyrenfurth I, Linkie D, Yeh M, Jewelewicz R: Pulsatile administration of gonadotropin-releasing hormone for induction of ovulation. Am J Obstet Gynecol 148:895, 1984 5. Hurley DM, Brian R, Outch K, Stockdale J, Fry A, Hack­ man C, Clarke I, Burger HG: Induction of ovulation and fertility in amenorrheic women by pulsatile low-dose gonad­ otropin-releasing hormone. N Engl J Med 310: 1069, 1984 6. Burger CW, van Kessel H, Schoemaker J: Induction of ovulation by prolonged pulsatile administration of luteiniz­ ing hormone releasing hormone (LRH) in patients with clo­ miphene-resistant polycystic ovary-like disease. Acta En­ docrinol (Copenh) 104:357, 1983 7. Molloy BG, Hancock KW, Glass MR: Ovulation induction in clomiphene nonresponsive patients: the place of pulsatile gonadotropin-releasing hormone in clinical practice. Fertil Steril 43:26, 1985 Jansen et al.

Pulsatile GnRH ovulation-induction

43

·

8. Bringer J, Hedon B, Jaffiol C, Nicolau S, Gibert F, Cristol P, Orsetti A, Viala J-L, Mirouze J: Influence of the frequency of gonadotropin-releasing hormone (GnRH) administration on ovulatory responses in women with anovulation. Fertil Steril 44:42, 1985 9. Ory SJ, London SN, Tyrey L, Hammond CB: Ovulation induction with pulsatile gonadotropin-releasing hormone administration in patients with polycystic ovarian syn­ drome. Fertil Steril 43:20, 1985 10. Saffan D, Seibel MM: Ovulation induction with subcutane­ ous pulsatile gonadotropin-releasing hormone in various ovulatory disorders. Fertil Steril 45:475, 1986 1 1 . Llewellyn-Jones D, Abraham SF: Quetelet index in diag­ nosis of anorexia nervosa. Br Med J 288: 1800, 1984 12. Spitz IM, Almaliach U, Rosen E, Polishuk W, Rabinowitz D: Dissociation of prolactin responsiveness to TRH and chlorpromazine in women with isolated gonadotropin defi­ ciency. J Clin Endocrinol Metab 45: 1 173, 1977 13. Lachelin GCL, Yen SSC: Hypothalamic chronic anovula­ tion. Am J Obstet Gynecol 130:825, 1978 14. Santen RJ, Friend JN, Trojanowski D, Davis B, Samojlik E, Bardino CW: Prolonged negative feedback suppression after estradiol administration: proposed mechanism of eu­ gonadal secondary amenorrhea. J Clin Endocrinol Metab 47:1220, 1978 15. Quigley ME, Sheehan KL, Casper RF, Yen SSC: Evidence for increased dopaminergic and opioid activity in patients with hypothalamic hypogonadotropic amenorrhea. J Clin Endocrinol Metab 50:949, 1980 16. Vigersky RA, Andersen AE, Thompson RH, Loriaux DL: Hypothalamic dysfunction in secondary amenorrhea asso­ ciated with simple weight loss. N Engl J Med 297: 1141, 1977 17. McArthur JW, Bullen BA, Beitins IZ, Pagano M, Badger TM, Klibanski A: Hypothalamic amenorrhea in runners of normal body composition. Endocr Res Commun 7:13, 1980 18. Vigersky RA, Loriaux DL, Andersen AE, Mecklenburg RS, Vaitukaitis JL: Delayed pituitary hormone response to LRF and TRF in patients with anorexia nervosa and with sec-

44

Jansen et al.

Pulsatile GnRH ovulation-induction

19.

20.

21.

22.

23.

24.

25.

26.

27.

ondary amenorrhea associated with simple weight loss. J Clin Endocrinol Metab 43:893, 1976 Marshall JC, Kelch RP: Low dose pulsatile gonadotropin­ releasing hormone in anorexia nervosa: a model of human pubertal development. J Clin Endocrinol Metab 49: 7 1 2 , 1979 Leyendecker G, Wildt L: Induction of ovulation with chronic intermittent (pulsatile) administration of Gn-RH in women with hypothalamic amenorrhoea. J Reprod Fertil 69:397 1983 ' Rebar R, Judd HL, Yen SCC, Rakoff J, Vandenberg G, Naftolin F: Characterization of the inappropriate gonado­ tropin secretion in polycystic ovary syndrome. J Clin Invest 57:1320, 1976 Quigley ME, Rakoff JS, Yen SSC: Increased luteinizing hormone sensitivity to dopamine inhibition in polycystic ovary syndrome. J Clin Endocrinol Metab 52:231, 1981 El Sheikh MAA, Molloy BG, Oakey RE, Chapman C, Han­ cock KW, Glass MR: LH pulse frequency in relation to positive oestrogen-gonadotrophin feedback in amenor­ rhoea. Clin Endocrinol 18:517, 1983 Ravnikar VA, Elkind-Hirsch K, Schiff I, Tulchinsky D, Ryan KJ: Peripheral immunoreactive-LHRH concentra­ tions in polycystic ovarian disease (Abstr 268). Presented at the 30th Annual Meeting of the Society for Gynecologic Investigation, Washington, DC, March 1 7-20, 1983 Marshall JC, Case GD, Valk TW, Corley KP, Sauder SE, Kelch RP: Selective inhibition of follicle-stimulating hor­ mone secretion by estradiol: mechanism of modulation of gonadotropin responses to low dose pulses of gonadotro­ pin-releasing hormone. J Clin Invest 71:248, 1983 Valk TW, Marshall JC, Kelch RP: Simulation of the follicu­ lar phase of the menstrual cycle by intravenous administra­ tion of low-dose pulsatile gonadotropin-releasing hormone. Am J Obstet Gynecol 141:842, 1981 Geisthovel F, Peters F, Breckwoldt M: Ovarian hyperstim­ ulation due to long-term pulsatile intravenous GnRH treat­ ment. Arch Gynecol 236:255, 1985

Fertility and Sterility