- Email: [email protected]
Pulsatile administration of gonadotropin-releasing hormone for induction of ovulation
Pulsatile administration of gonadotropin -releasing hormone for induction of ovulation A. Loucopoulos, M. Fecin, R. L. Vande Wiele, I. Dyrenfurth, D. Linkie, M. Yeh, and R. Jewelewicz New York, New York Chronic pulsatile administration of gonadotropin-releasing hormone (GnRH) was used to induce ovulation in 12 women with various ovulatory disorders. In the first group of eight patients with normal to low baseline levels of gonadotropin, seven responded favorably to the treatment. Follicular maturation was observed in 57% of the treated cycles, and nonnal ovulatory cycles were induced in 24% of the patients. Two patients became pregnant. The intravenous route of administration was more effective than the subcutaneous one, possibly in response to the GnRH profile after each pulse. (The amplitude of GnRH peaks after an intravenous pulse was four times that seen after a subcutaneous one.) In contrast, follicular maturation and ovulation could not be induced in four women of a second group of patients with normal baseline levels of follicle-stimulating hormone but with high and frequent pulses of luteinizing hormone. The conclusion reached was that pulsatile administration of GnRH can be a new therapeutic tool in the treatment of ovulatory disorders in women who have an insufficient endogenous release of GnRH. (AM. J. OBSTET. GVNECOL. 148:895, 1984.)
Gonadotropin-releasing hormone (GnRH), when given in a pulsatile fashion, is known to stimulate the release of gonadotropin from the anterior pituitary gland. As a therapeutic modality for induction of ovulation, pulsatile administration of GnRH offers a major advantage over gonadotropin treatment because it leaves the feedback mechanism operational, thus avoiding the hyperstimulation syndrome and multiple pregnancy. Several investigators have reported successful induction of ovulation and pregnancy with pulsatile GnRH treatment in selected women with primary or secondary amenorrhea. I. 2 However, different treatment regimens were used in these studies, with sometimes conflicting results. In an attempt to establish criteria for the selection of patients who may benefit most from treatment, we carried out chronic pulsatile infusions of GnRH in 12 women with various ovulatory disorders. The responses to different doses, frequencies, and routes of administration were compared in the majority of patients, with each patient serving as her own control. Material and method Long-term pulsatile infusions of GnRH were administered to 12 women ages 22 to 41 years, with ovulatory From the Department of Obstetrics and Gynecology, Columbia University, College of Physicians and Surgeons. Supported fry Grant RR0064 5 from the National Institutes of Health. Presented by invitation at the Second Annual Meeting of the American Gynecological and Obstetrical Society, Phoenix, Arizona, September 7-10, 1983. Reprint requests: Dr. Raphael Jewelewicz, Department of Obstetrics and Gynecology, Columbia University, College of Physicians and Surgeons, 630 West 168th St., New York, New York 10032.
disorders, to induce an ovulatory cycle. After informed consent had been obtained, each patient was admitted to the Clinical Research Center for a baseline control gonadotropin study. Samples of blood were collected from an intravenous catheter at 10- to 20-minute intervals for a period of 6 hours (0900 to 1500 hours). Samples were later assayed for luteinizing hormone (LH) and follicle-stimulating hormone (FSH). This test allowed us to divide the patients into two groups, on the basis of mean baseline LH and FSH levels and the pattern of LH pulsatile secretion (see Results). GnRH treatment was then initiated, and the response was monitored by obtaining daily samples of blood for determinations of LH, FSH, estrogen, and progesterone and by ultrasonography in selected patients to assess follicular maturation. GnRH (Factrel, Ayerst Labs.) was infused in a pulsatile mode, by means of one of two portable pumps (Autosyringe, model AS, Hooksett, New Hampshire; Zyklomat, Ferring, West Germany). Doses varied from 2 to 20 J..tg/pulse (duration of the pulse: Zyklomat, 1 minute; Autosyringe, 15 seconds). The higher dose was used when follicular maturation failed to occur during the first 2 to 3 weeks of treatment. The frequency of pulsatile administration was set at 90 or 180 minutes, with the latter frequency being used mostly in patients of group 2. GnRH was administered via the subcutaneous route in 14 cycles and the intravenous route in 25 cycles. During intravenous administration, an intravenous catheter was in place for a period of 2 to 4 weeks. In two patients (one cycle each), infusion of GnRH was discontinued at the peak of follicular maturation, as determined by level of estrogen and sonography, and
895
896 Loucopoulos et al.
April I, 1984 Am. J. Obstet. Gynecol.
Table I. GnRH therapy in group 1 patients Basal levels of
GnRH treatment
Effectiveness of treatment
Route Patient No.
LH (mIU/ ml)
FSH (mIU/ ml)
41
4.0
4.2
6S
151
1.4
3.1
113
101
121
1.2
1.I
O.S
0.9
Estrogens (pg/ml)
66
15S
lSI
2.4
3.4
147
171
0.7
I.S
140
191
4.2
2.3
137
201
3.S
3.6
112
Dose (/Lg)
Frequeney (min)
2 2 2 5 20 20 2 20 20 5 10 20 20 5 20 20 20 5 20 10 10
90 ISO 90 90 90 90 90 90 ISO 90 90 90 90 90 90 90 90 90 90 90 90
Subcutaneous
I venous lntra-
Follicular maturation
Spontaneous
+ + + +
+ + +
+ +
+
+*
+ + +
+ +
+ +*
LH surge
I
Corpus luteum
Induced
Nor- Ilnademal quate
Pregnancy
+ + + + + +
+ + +
+
+ + + + + + + + + + + + + + +
+ + +
+
+
+ +t
+ +
+
Patients 41, 151, and 101 presented with primary amenorrhea; Patients 121 and lSI with secondary amenorrhea; and Patients 171, 191, and 201 with oligoamenorrhea. *hCG was given to maintain the corpus luteum. tTerm pregnancy.
human chorionic gonadotropin (hCG) (10,000 IU) was given intramuscularly in order to trigger ovulation. In two other patients, the infusion of GnRH was discontinued in midluteal phase, and hCG (2,500 IU) was given twice to maintain normal luteal function. Most patients were readmitted to the Clinical Research Center after the first week of GnRH treatment. Samples of blood were collected at 10-minute intervals during a period that included the administration of three GnRH pulses. LH and FSH responses to GnRH and concentrations of GnRH in the blood were determined. All samples of blood were centrifuged, and sera were stored at -10 C until radioimmunoassay was performed for LH,~ FSH,:l GnRH,4 estrogens,5 and progesterone. 5 0
Results
Attempts were made to induce ovulation in 12 women. All had normal levels of prolactin «25 ng/ml). The patients were divided into two groups on the basis of their baseline secretion patterns of gonadotropin. Group 1 consisted of eight women with low to normal levels of gonadotropin (LH, 2.3 ± 1.5 mIU/ml; FSH, 2.5 ± 1.2 mIU/ml; mean ± SD). Blood sampling at 20-minute intervals over a period of 6 hours indicated a
lack of distinct pulsatile release of LH. These women presented with ovulatory disorders that ranged from primary amenorrhea to oligoamenorrhea (Table I). These eight patients had 21 GnRH-treated cycles, 16 by the intravenous route, five by the subcutaneous route. The predominant frequency for the GnRH pulse was 90 minutes (19 of 21 cycles). In none of the treatment cycles during which GnRH was administered subcutaneously was there evidence of follicular maturation (as judged by increased secretion of estrogen). When GnRH was infused intravenously, follicular maturation was achieved in 12 cycles. In seven of these, a spontaneous surge of gonadotropin accompanied the estrogen peak. Ovulation and normal luteal function, indicated by increased secretion of progesterone (>4 ng/ml), were observed in five of these treatment cycles. In two of these five cycles, the patients received hCG (2,500 IU X 2) every other day to support the corpus luteum. In two of the seven treatment cycles, during which a spontaneous surge of LH occurred, an inadequate luteal phase ensued (progesterone <4 ng/ml; < 14 days' duration). In two of the other five cycles in which increased secretion of estrogen occurred, the pulsatile infusion of GnRH was terminated, and hCG (10,000 IU) was administered to induce ovulation, but in both, the luteal phase was inadequate.
Pulsatile GnRH for ovulation induction
Volume 148 Number 7
897
Table II. GnRH therapy in group 2 patients GnRH treatment
Basal levels of
Effectiveness of treatment
Route Patient No.
81
LH (mIUI ml)
FSH (mIUI ml)
Estrogen (pglml)
Dose (f.J.g)
5.1
2.5
140
8.9
4.7
204
2 2 20 20 20 2 20 2 2 20 20
91
5.5
3.4
205
141
9.0
2.6
223
10
2 2 2 2 2 2
Frequency (min)
90 180 180 90 540 180 180 90 180 180 90 180 180 90 180 90 180 90
Subcutaneous
I
lntravenous
Follicular maturation
Spontaneous
+
+
LH surge
I
Induced
Corpus luteum Nor-Ilnademal quate
Pregnancy
+ + + + + + + + + + + + + + + + + +
+
+
All patients were oligomenorrheic. Other basal levels: Testosterone levels in Patients 1, 91, 141 = 40, 44, and 76 ng/dl, respectively. Androstenedione levels in Patients 1,91, and 141 = 137, 183, and 138 ng/dl, respectively.
Two of the patients III group 1 conceived. One was delivered at term. The other pregnancy is still in progress. Group 2 consisted of four oligomenorrheic patients with normal FSH (3.2 ± 1.0 mIU/ml) and high LH (7.1 ± 2.1 mIUlml; mean ± SD). Multiple samplings of blood indicated frequent LH pulses at 40- to 80-minute intervals. In these four patients, GnRH was given during 18 cycles, nine by an intravenous route, nine by the subcutaneous route. A normal ovulatory cycle was observed in only one patient, whereas the others failed to show any response (Table II). Comment
Our results indicate that chronic pulsatile administration of GnRH for the treatment of ovulatory disorders was highly effective in women with deficient hypothalamic-gonadotropin function (group 1). In contrast, in women in whom ovulatory disturbances were accompanied by high hypothalamic-gonadotropin activity (group 2), exogenous administration of GnRH was unsuccessful. All but one woman in group 1 responded to the treatment. Follicular maturation occurred in 57% of the treated cycles, and corpus luteum formation followed in 24% of the treatment cycles. It is evident, therefore, that normal pituitary function can be restored in gonadotropin-deficient women when ade-
quate amounts of GnRH reach the gonadotropes, as has been shown in the monkey.6 Successful therapy depends on (1) the dose ofGnRH per pUlse, (2) the frequency of pulses, and (3) the route of administration. Insofar as frequency is concerned, all investigators agree that a gO-minute interval between pulses is as effective as a 2-hour interval. However, controversy exists in regard to the effectiveness of intravenous versus subcutaneous administration, as well as the amount of GnRH injected per pulse. In one recently reported case, the subcutaneous route was completely ineffective, whereas the same patient responded favorably and conceived after intravenous administration. 2 Other investigators also have reported less effectiveness via the subcutaneous route. 7 Our data confirm those reports. Almost all of our patients responded to the intravenous route of administration. The advantages of the intravenous route are most probably related to the levels of GnRH achieved in the blood. Two striking differences in blood concentrations of GnRH given by the intravenous and subcutaneous routes are shown in Fig. 1. First, peak levels of GnRH measured in blood were four to five times higher after intravenous than after subcutaneous administration. Second, the GnRH profile obtained after subcutaneous administration was sustained for a considerably longer period of time than that after intravenous administration. Since the subcutaneous GnRH
898
Loucopoulos et al.
April I, 1984 Am. J. Obstet. Gynecol.
2ug GnRH 280
20ug GnRH
~
~
-
240
E
~
S.C.
1200
I.V.
G)
I
200
"OJ Q.
1400
180
1000
~
JJ
I
800
~
"0
I
c:
800
120
"-
c:
CJ
to
80
40 0
)\"-.
~ 0
07~~i:::~~~~3
I .....
20
40
80
80
TIME
h~ ~ 'o~:::;::!~~ I
0
20
40
80
400
3
200
g:1J
0
80
( minJ
Fig. 1. Levels of GnRH in the blood after intravenous and subcutaneous administration of 2 and 20 f.Lg of GnRH per pulse. Note the different scale of the two panels.
profile would more closely mimic that of a continuous than that of a pulsatile infusion, a subcutaneous mode of administration may result in a suboptimal release of gonadotropins after "down regulation" of the gonadotrope. This was evident in one of our patients (Patient 41) in whom 2 Mg of GnRH given subcutaneously was ineffective, whereas the same dose given intravenously induced follicular maturation. The dose which gave the best results was 20 Mg of GnRH per pulse. This is an acceptable and highly effective dose. It produces levels of GnRH in the blood that do not exceed those found in human portal blood,s and thus simulates physiologic conditions. No side effects or evidence of hyperstimulation was noted in any of our patients. From our results in group 1 patients, it is also evident that women with a variety of ovulatory disorders, not just women with hypothalamic amenorrhea, can benefit from pulsatile administration of GnRH. Simulation of normal pituitary function by pulsatile administration of GnRH can be accomplished even in the presence of endogenous (but presumably not physiologic) GnRH release. Our inability to induce follicular maturation and ovulation in the women of group 2 may possibly have been related to the fact that the therapeutic GnRH regimen did not sufficiently reduce the high LHiFSH ratio observed in those patients. Persistently high levels of LH are known to stimulate the local production of androgen by the follicle and to impair follicular growth. 9 We attempted to lower the LH/FSH ratio by decreasing the GnRH pulse frequency from 90 to 180 minutes,
because it has been shown that a lower frequency of the GnRH pulse favors the secretion of FSH over that of LH in monkeys bearing arcuate lesions. 1o We were partially successful in this attempt in one patient (Patient 141), in whom 24-hour integrated changes in urinary gonadotropins were measured at various times during GnRH treatment. Integrated levels of LH decreased by 50% while the patient was on the 180minute frequency schedule, as compared to the 90minute frequency. Perhaps in reaction to the change in LH/FSH ratio, the patient successfully responded to the subsequent GnRH treatment cycle. Unfortunately, partial changes in the LH/FSH ratio may have been insufficient to overcome long-term effects of an abnormal ratio on ovarian function in other patients. Attempts to further modify LHiFSH ratios may be required. In conclusion, we have shown that chronic pulsatile administration of GnRH can be an effective alternative treatment for the induction of ovulation in women with a variety of ovulatory disorders manifested by low or normal secretion of gonadotropin. In our hands, a regimen of 20 Mg of GnRH given intravenously at 90minute intervals seems to have the highest efficacy and to be free of side effects and complications. More clinical trials with alternate regimens are needed in order to identify the dose and frequency most suitable for women with anovulatory disorders in whom endogenous hypothalamic-pituitary activity is high. We would like to thank Ayerst Labs. and Drs. R. H. Wehr and J. D. Stevens for their generous supply of
Volume 148 Number 7
GnRH (Factrel). We also wish to acknowledge the general clinical support of Dr. R. Canfield and the staff of the General Clinical Research Center. REFERENCES
1. Leyendecker. G., Wildt, L., and Hansmann, M.: Pregnancies following chronic intermittent (pulsatile) administration of GnRH by means of a portable pump ("Zyklomat"): A new approach to the treatment of infertility in hypothalamic amenorrhea, J. Clin. Endocrinol. Metab. 51:1214,1980. 2. Reid, L. R., Leopold, G. R., and Yen, S. S. C.: Induction of ovulation and pregnancy with pulsatile luteinizing hormone releasing factor: Dosage and mode of delivery, Ferti!. Steril. 36:553, 1981. 3. Midgley, A. R., Niswender, G. D., Gay, V. C., and Reichert, L. F.: Use of antibodies for characterization of gonadotropins and steroids, Recent Prog. Horm. Res. 27: 235, 1971. 4. Araki, S., Ferin, M., Zimmerman, E. A., and Vande Wiele, R. L.: Ovarian modulation of immunoreactive gonadotropin-releasing hormone (GnRH) in rat brain: Evidence for a differential effect on the anterior and mid hypothalamus, Endocrinology 96:644, 1975. 5. Abraham, G. E., Odell, W. D., Swerdloff, R. S., and Hopper, K.: Simultaneous RIA of plasma, FSH, LH, progesterone, 17-hydroxyprogesterone and estradiol-17,B during the menstrual cycle, J. Clin. Endocrinol. Metab. 34: 312,1972. 6. Knobil, E., Plant, T M., Wildt, L., Belchet, P. E., and Marshall, G.: Control of the rhesus monkey menstrual cycle: Permissive role of hypothalamic gonadotropinreleasing hormone, Science 207:1371,1980. 7. Leyendecker, G., and Wildt, L.: Induction of ovulation with chronic intermittent (pulsatile) administration of GnRH in hypothalamic amenorrhea, in Research on Steroids, Amsterdam, 1982, vol. 10, Excerpta Medica, p. 275. 8. Antunes, J. L., Carmel, P. W., Housepian, E. M., and Ferin, M.: Luteinizing hormone-releasing hormone in human pituitary blood, J. Neurosurg. 49:382, 1978. 9. Wu, Ch., Motohashi, T, Abdel-Rahman, H. A., and Mikhail, G.: Estrogen-androgen balance in anovulation, Fertil. Steril. 36:61,1981. 10. Wildt, L., Hausler, A., Marshall, G., Hutchinson, J. S., Plant, T M., Belchetz, P. E., and Knobil, E.: Frequency and amplitude of gonadotropin-releasing hormone stimulation and gonadotropin secretion in the rhesus monkey, Endocrinology 109:376, 1981.
Discussion
DR. S. S. C. YEN, La Jolla, California. Within the past few years, several groups have demonstrated the effectiveness of pulsatile administration of GnRH in the activation of cyclic pituitary-ovarian axis. Induction of follicular maturation, ovulation, and pregnancy in a variety of patients with hypothalamic dysfunction, including congenital GnRH deficiency, functional hypothalamic amenorrhea, and panhypopituitarism due to previous stalk transaction, have been achieved. 1 Dr. Jewelewicz and his colleagues have presented their experience with pulsatile GnRH therapy and have confirmed the safety and ease with which ovulatory menstrual cycles can be established in hypogonadotropic amenorrheic patients (group 1). In our laboratory,
Pulsatile GnRH for ovulation induction
899
much smaller doses of GnRH per pulse (1 to 5 /Lg) have achieved ovulation and pregnancies. 2 It was anticipated (primarily on the basis of rhesus monkeys with arcuate nucleus lesions-the "hypothalamic clamp" model) that pulsatile administration of GnRH at a frequency of 60 minutes reinitiates ovulatory cycles with one dominant follicle, and that hyperstimulation should not occur because of the intrinsic feedback modulation of pituitary responsiveness to exogenous GnRH.3 This premise proves not to be the case, inasmuch as multiple pregnancies have already been reported; among 23 pregnancies induced by pulsatile GnRH, there have been two sets of twins and one set of triplets. 4 Moreover, we have deliberately induced multiple follicular maturations in women with normal cycles by using larger doses of GnRH pulses (10 /Lg).5 This finding may prove to be useful to the in vitro fertilization program. Thus, the internal ovarian feedback system can be overridden by excessive GnRH pulses. Nonetheless, pulsatile GnRH therapy will probably replace exogenous gonadotropin in the majority of anovulatory infertile patients. Pulsatile administration of GnRH has not been successful in anovulatory women with high LH/FSH ratios, as also shown by the data of Dr. Jewelewicz and associates. Under this condition, high levels of LH are likely to reflect increased GnRH receptors on the gonadotrope occasioned by enhanced GnRH secretion. Thus, the principle of up-regulation by constant GnRH pulses is not operative in these patients, since the GnRH-gonadotrope activities are already elevated. However, attenuation of this high functional activity by several pharmacologic maneuvers can potentially be achieved before initiation of pulsatile GnRH therapy. The subject addressed by Dr. Jewelewicz and his associates will undoubtedly improve further by detailed examination of the dose and frequency of GnRH pulses synchronized with the underlying GnRH-gonadotropic status. I have the following questions to ask of Dr. Jewelewicz. (1) Why was hCG used to "trigger ovulation"? One of the major advantages of pulsatile GnRH therapy is the ability of endogenous estradiol to initiate spontaneous midcycle surge and, consequently, provide for timely ovulation. (2) What was the rationale behind the use of pulsatile GnRH in a case with short luteal phase? (3) What explanation can you offer for the failure of response in patients with high LH/FSH ratios? (4) Finally, what are the bases for approximating that 20 /Lg per pulse is comparable to human portal blood GnRH values?
REFERENCES
1. Yen, S. S. C.: Clinical applications of gonadotropinreleasing hormone and gonadotropin-releasing hormone analogs, Ferti!. Steril. 39:257, 1983. 2. Miller, D. S., Reid, R. R., Cetel, N. S., Rebar, R. W., and Yen, S. S. c.: Conception following ovulation induction by pulsatile administration of low doses of gonadotropin-
900
Loucopoulos et al. Am.
releasing hormone (GnRH) in women with hypothalamic amenorrhea, JAMA. In press. 3. Knobil, E.: The neuroendocrine control of the menstrual cycle, Recent Prog. Horm. Res. 36:53, 1980. 4. Leyendecker, G., and Wildt, L.: Induction of ovulation with chronic-intermittent (pulsatile) administration of GnRH in hypothalamic amenorrhea, J. Reprod. Fertil. In press. 5. Liu, J. H., Durfee, R., Muse, K., and Yen, S. S. C.: Induction of multiple ovulation by pulsatile administration of gonadotropin-releasing hormone (GnRH), Fertil. Steril. 40: 18, 1983. DR. JEWELEWICZ (Closing). I will try to answer Dr. Yen's questions. First, why was hCG used to trigger ovulation? We us~d 10,000 IV of hCG in two patients to trigger ovulatIOn after they reached follicular maturation, in the same way that we do in Pergonal-treated cycles. Indeed, both patients ovulated, and had a normal luteal phase. This is not absolutely necessary. When pulsatile GnRH is continued, the patients will have an LH surge, ovulate, and have a normal luteal phase, as long as GnRH is continued, and adequate levels of LH and FSH are maintained. The reason we used hCG was to learn whether it can shorten the time that GnRH is given, and whether it can replace GnRH in maintenance of the luteal phase-and, indeed, it can. In two other patients, we discontinued GnRH in mid~uteal p~a~e a.nd supplemented luteal function by givmg two mJectlOns of 2,500 IV of hCG 48 hours apart, as was done by Dr. Yen's group (although they used only 1,500 IV of hCG times three) and by Dr. Leyendecker's group. .Second, What was the rationale for treating patients WIth a short luteal phase with GnRH? I regret to say that, up until now, we did not know exactly how to treat adequately and successfully patients with an inadequate luteal phase. There are several regimens of treatment, none of which is too successful. Our success r~te in treating the short luteal phase with clomiphene Citrate or gonadotropins is far from ideal. We decided to treat two patients with GnRH to learn whether we
J.
April I, 1984 Obstet. Gynecol.
could induce ovulation and improve luteal function. So far, we have not been successful. . Third.' Why did we fail to induce ovulation in patients with polycystic ovary syndrome who had high levels of LH and normal levels of FSH? The question is a most interesting and challenging one. These patients have ovulatory dysfunction because of the high LHI FSH. ratio, which is probably controlled centrally. A conSiderable degree of our understanding of the syndrome comes from Dr. Yen's work. We tried to change the LH/FSH ratio by decreasing the frequency of the GnRH pulses from 90 to 180 minutes, since Dr. Knobil showed, in rhesus monkeys with lesions of the arcuate nucleus, that a lower frequency of GnRH pulses resulted in an increased secretion of FSH over LH, and the LH/FSH ratio changed. However, that has not happened in our patients. In only one patient who received GnRH pulses every 180 minutes was a 50% decrease in urinary LH found over a 24-hour period, as co.mpared to the cycle when GnRH was given every 90 mmutes. However, this patient did not ovulate. I believe ~ha.t by more careful study with a different dosage a.nd tImmg and perhaps initial gonadotropin suppresSion, we may be a~le to manipulate and change the LH/FSH ratio and mduce ovulation and pregnancy in these patients. A review of the literature discloses that most ~nve.stigat~rs seem to have problems inducing ovulatIon m patients with polycystic ovary syndrome. Fourth, Why did we use a 20 ILg/pulse of GnRH? I agree with Dr. Yen that this is a relatively high dose, and, indeed, he is using very successfully 1 to 5 ILgl pulse. The literature reveals that Dr. Leyendecker uses from 2.5 to 20 ILg/pulse, Dr. Shoemaker 10 to 20 ILgl pulse, and up to 100 ILg/pulse. We used the same dose as Dr. Leyendecker (20 ILg/pulse), and since we had a pregnancy on this dose, we continued it. Twenty micrograms per pulse gives a blood level of GnRH comparable to the level found in portal blood (300 to 2,000 pg/ml). We are testing various doses (2, 5, 10, and 20 ILg/pulse), and I agree with Dr. Yen that smaller doses of GnRH may be as effective, if not more effective.
Gonadotropin-releasing hormone (GnRH), when given in a pulsatile fashion, is known to stimulate the release of gonadotropin from the anterior pituitary gland. As a therapeutic modality for induction of ovulation, pulsatile administration of GnRH offers a major advantage over gonadotropin treatment because it leaves the feedback mechanism operational, thus avoiding the hyperstimulation syndrome and multiple pregnancy. Several investigators have reported successful induction of ovulation and pregnancy with pulsatile GnRH treatment in selected women with primary or secondary amenorrhea. I. 2 However, different treatment regimens were used in these studies, with sometimes conflicting results. In an attempt to establish criteria for the selection of patients who may benefit most from treatment, we carried out chronic pulsatile infusions of GnRH in 12 women with various ovulatory disorders. The responses to different doses, frequencies, and routes of administration were compared in the majority of patients, with each patient serving as her own control. Material and method Long-term pulsatile infusions of GnRH were administered to 12 women ages 22 to 41 years, with ovulatory From the Department of Obstetrics and Gynecology, Columbia University, College of Physicians and Surgeons. Supported fry Grant RR0064 5 from the National Institutes of Health. Presented by invitation at the Second Annual Meeting of the American Gynecological and Obstetrical Society, Phoenix, Arizona, September 7-10, 1983. Reprint requests: Dr. Raphael Jewelewicz, Department of Obstetrics and Gynecology, Columbia University, College of Physicians and Surgeons, 630 West 168th St., New York, New York 10032.
disorders, to induce an ovulatory cycle. After informed consent had been obtained, each patient was admitted to the Clinical Research Center for a baseline control gonadotropin study. Samples of blood were collected from an intravenous catheter at 10- to 20-minute intervals for a period of 6 hours (0900 to 1500 hours). Samples were later assayed for luteinizing hormone (LH) and follicle-stimulating hormone (FSH). This test allowed us to divide the patients into two groups, on the basis of mean baseline LH and FSH levels and the pattern of LH pulsatile secretion (see Results). GnRH treatment was then initiated, and the response was monitored by obtaining daily samples of blood for determinations of LH, FSH, estrogen, and progesterone and by ultrasonography in selected patients to assess follicular maturation. GnRH (Factrel, Ayerst Labs.) was infused in a pulsatile mode, by means of one of two portable pumps (Autosyringe, model AS, Hooksett, New Hampshire; Zyklomat, Ferring, West Germany). Doses varied from 2 to 20 J..tg/pulse (duration of the pulse: Zyklomat, 1 minute; Autosyringe, 15 seconds). The higher dose was used when follicular maturation failed to occur during the first 2 to 3 weeks of treatment. The frequency of pulsatile administration was set at 90 or 180 minutes, with the latter frequency being used mostly in patients of group 2. GnRH was administered via the subcutaneous route in 14 cycles and the intravenous route in 25 cycles. During intravenous administration, an intravenous catheter was in place for a period of 2 to 4 weeks. In two patients (one cycle each), infusion of GnRH was discontinued at the peak of follicular maturation, as determined by level of estrogen and sonography, and
895
896 Loucopoulos et al.
April I, 1984 Am. J. Obstet. Gynecol.
Table I. GnRH therapy in group 1 patients Basal levels of
GnRH treatment
Effectiveness of treatment
Route Patient No.
LH (mIU/ ml)
FSH (mIU/ ml)
41
4.0
4.2
6S
151
1.4
3.1
113
101
121
1.2
1.I
O.S
0.9
Estrogens (pg/ml)
66
15S
lSI
2.4
3.4
147
171
0.7
I.S
140
191
4.2
2.3
137
201
3.S
3.6
112
Dose (/Lg)
Frequeney (min)
2 2 2 5 20 20 2 20 20 5 10 20 20 5 20 20 20 5 20 10 10
90 ISO 90 90 90 90 90 90 ISO 90 90 90 90 90 90 90 90 90 90 90 90
Subcutaneous
I venous lntra-
Follicular maturation
Spontaneous
+ + + +
+ + +
+ +
+
+*
+ + +
+ +
+ +*
LH surge
I
Corpus luteum
Induced
Nor- Ilnademal quate
Pregnancy
+ + + + + +
+ + +
+
+ + + + + + + + + + + + + + +
+ + +
+
+
+ +t
+ +
+
Patients 41, 151, and 101 presented with primary amenorrhea; Patients 121 and lSI with secondary amenorrhea; and Patients 171, 191, and 201 with oligoamenorrhea. *hCG was given to maintain the corpus luteum. tTerm pregnancy.
human chorionic gonadotropin (hCG) (10,000 IU) was given intramuscularly in order to trigger ovulation. In two other patients, the infusion of GnRH was discontinued in midluteal phase, and hCG (2,500 IU) was given twice to maintain normal luteal function. Most patients were readmitted to the Clinical Research Center after the first week of GnRH treatment. Samples of blood were collected at 10-minute intervals during a period that included the administration of three GnRH pulses. LH and FSH responses to GnRH and concentrations of GnRH in the blood were determined. All samples of blood were centrifuged, and sera were stored at -10 C until radioimmunoassay was performed for LH,~ FSH,:l GnRH,4 estrogens,5 and progesterone. 5 0
Results
Attempts were made to induce ovulation in 12 women. All had normal levels of prolactin «25 ng/ml). The patients were divided into two groups on the basis of their baseline secretion patterns of gonadotropin. Group 1 consisted of eight women with low to normal levels of gonadotropin (LH, 2.3 ± 1.5 mIU/ml; FSH, 2.5 ± 1.2 mIU/ml; mean ± SD). Blood sampling at 20-minute intervals over a period of 6 hours indicated a
lack of distinct pulsatile release of LH. These women presented with ovulatory disorders that ranged from primary amenorrhea to oligoamenorrhea (Table I). These eight patients had 21 GnRH-treated cycles, 16 by the intravenous route, five by the subcutaneous route. The predominant frequency for the GnRH pulse was 90 minutes (19 of 21 cycles). In none of the treatment cycles during which GnRH was administered subcutaneously was there evidence of follicular maturation (as judged by increased secretion of estrogen). When GnRH was infused intravenously, follicular maturation was achieved in 12 cycles. In seven of these, a spontaneous surge of gonadotropin accompanied the estrogen peak. Ovulation and normal luteal function, indicated by increased secretion of progesterone (>4 ng/ml), were observed in five of these treatment cycles. In two of these five cycles, the patients received hCG (2,500 IU X 2) every other day to support the corpus luteum. In two of the seven treatment cycles, during which a spontaneous surge of LH occurred, an inadequate luteal phase ensued (progesterone <4 ng/ml; < 14 days' duration). In two of the other five cycles in which increased secretion of estrogen occurred, the pulsatile infusion of GnRH was terminated, and hCG (10,000 IU) was administered to induce ovulation, but in both, the luteal phase was inadequate.
Pulsatile GnRH for ovulation induction
Volume 148 Number 7
897
Table II. GnRH therapy in group 2 patients GnRH treatment
Basal levels of
Effectiveness of treatment
Route Patient No.
81
LH (mIUI ml)
FSH (mIUI ml)
Estrogen (pglml)
Dose (f.J.g)
5.1
2.5
140
8.9
4.7
204
2 2 20 20 20 2 20 2 2 20 20
91
5.5
3.4
205
141
9.0
2.6
223
10
2 2 2 2 2 2
Frequency (min)
90 180 180 90 540 180 180 90 180 180 90 180 180 90 180 90 180 90
Subcutaneous
I
lntravenous
Follicular maturation
Spontaneous
+
+
LH surge
I
Induced
Corpus luteum Nor-Ilnademal quate
Pregnancy
+ + + + + + + + + + + + + + + + + +
+
+
All patients were oligomenorrheic. Other basal levels: Testosterone levels in Patients 1, 91, 141 = 40, 44, and 76 ng/dl, respectively. Androstenedione levels in Patients 1,91, and 141 = 137, 183, and 138 ng/dl, respectively.
Two of the patients III group 1 conceived. One was delivered at term. The other pregnancy is still in progress. Group 2 consisted of four oligomenorrheic patients with normal FSH (3.2 ± 1.0 mIU/ml) and high LH (7.1 ± 2.1 mIUlml; mean ± SD). Multiple samplings of blood indicated frequent LH pulses at 40- to 80-minute intervals. In these four patients, GnRH was given during 18 cycles, nine by an intravenous route, nine by the subcutaneous route. A normal ovulatory cycle was observed in only one patient, whereas the others failed to show any response (Table II). Comment
Our results indicate that chronic pulsatile administration of GnRH for the treatment of ovulatory disorders was highly effective in women with deficient hypothalamic-gonadotropin function (group 1). In contrast, in women in whom ovulatory disturbances were accompanied by high hypothalamic-gonadotropin activity (group 2), exogenous administration of GnRH was unsuccessful. All but one woman in group 1 responded to the treatment. Follicular maturation occurred in 57% of the treated cycles, and corpus luteum formation followed in 24% of the treatment cycles. It is evident, therefore, that normal pituitary function can be restored in gonadotropin-deficient women when ade-
quate amounts of GnRH reach the gonadotropes, as has been shown in the monkey.6 Successful therapy depends on (1) the dose ofGnRH per pUlse, (2) the frequency of pulses, and (3) the route of administration. Insofar as frequency is concerned, all investigators agree that a gO-minute interval between pulses is as effective as a 2-hour interval. However, controversy exists in regard to the effectiveness of intravenous versus subcutaneous administration, as well as the amount of GnRH injected per pulse. In one recently reported case, the subcutaneous route was completely ineffective, whereas the same patient responded favorably and conceived after intravenous administration. 2 Other investigators also have reported less effectiveness via the subcutaneous route. 7 Our data confirm those reports. Almost all of our patients responded to the intravenous route of administration. The advantages of the intravenous route are most probably related to the levels of GnRH achieved in the blood. Two striking differences in blood concentrations of GnRH given by the intravenous and subcutaneous routes are shown in Fig. 1. First, peak levels of GnRH measured in blood were four to five times higher after intravenous than after subcutaneous administration. Second, the GnRH profile obtained after subcutaneous administration was sustained for a considerably longer period of time than that after intravenous administration. Since the subcutaneous GnRH
898
Loucopoulos et al.
April I, 1984 Am. J. Obstet. Gynecol.
2ug GnRH 280
20ug GnRH
~
~
-
240
E
~
S.C.
1200
I.V.
G)
I
200
"OJ Q.
1400
180
1000
~
JJ
I
800
~
"0
I
c:
800
120
"-
c:
CJ
to
80
40 0
)\"-.
~ 0
07~~i:::~~~~3
I .....
20
40
80
80
TIME
h~ ~ 'o~:::;::!~~ I
0
20
40
80
400
3
200
g:1J
0
80
( minJ
Fig. 1. Levels of GnRH in the blood after intravenous and subcutaneous administration of 2 and 20 f.Lg of GnRH per pulse. Note the different scale of the two panels.
profile would more closely mimic that of a continuous than that of a pulsatile infusion, a subcutaneous mode of administration may result in a suboptimal release of gonadotropins after "down regulation" of the gonadotrope. This was evident in one of our patients (Patient 41) in whom 2 Mg of GnRH given subcutaneously was ineffective, whereas the same dose given intravenously induced follicular maturation. The dose which gave the best results was 20 Mg of GnRH per pulse. This is an acceptable and highly effective dose. It produces levels of GnRH in the blood that do not exceed those found in human portal blood,s and thus simulates physiologic conditions. No side effects or evidence of hyperstimulation was noted in any of our patients. From our results in group 1 patients, it is also evident that women with a variety of ovulatory disorders, not just women with hypothalamic amenorrhea, can benefit from pulsatile administration of GnRH. Simulation of normal pituitary function by pulsatile administration of GnRH can be accomplished even in the presence of endogenous (but presumably not physiologic) GnRH release. Our inability to induce follicular maturation and ovulation in the women of group 2 may possibly have been related to the fact that the therapeutic GnRH regimen did not sufficiently reduce the high LHiFSH ratio observed in those patients. Persistently high levels of LH are known to stimulate the local production of androgen by the follicle and to impair follicular growth. 9 We attempted to lower the LH/FSH ratio by decreasing the GnRH pulse frequency from 90 to 180 minutes,
because it has been shown that a lower frequency of the GnRH pulse favors the secretion of FSH over that of LH in monkeys bearing arcuate lesions. 1o We were partially successful in this attempt in one patient (Patient 141), in whom 24-hour integrated changes in urinary gonadotropins were measured at various times during GnRH treatment. Integrated levels of LH decreased by 50% while the patient was on the 180minute frequency schedule, as compared to the 90minute frequency. Perhaps in reaction to the change in LH/FSH ratio, the patient successfully responded to the subsequent GnRH treatment cycle. Unfortunately, partial changes in the LH/FSH ratio may have been insufficient to overcome long-term effects of an abnormal ratio on ovarian function in other patients. Attempts to further modify LHiFSH ratios may be required. In conclusion, we have shown that chronic pulsatile administration of GnRH can be an effective alternative treatment for the induction of ovulation in women with a variety of ovulatory disorders manifested by low or normal secretion of gonadotropin. In our hands, a regimen of 20 Mg of GnRH given intravenously at 90minute intervals seems to have the highest efficacy and to be free of side effects and complications. More clinical trials with alternate regimens are needed in order to identify the dose and frequency most suitable for women with anovulatory disorders in whom endogenous hypothalamic-pituitary activity is high. We would like to thank Ayerst Labs. and Drs. R. H. Wehr and J. D. Stevens for their generous supply of
Volume 148 Number 7
GnRH (Factrel). We also wish to acknowledge the general clinical support of Dr. R. Canfield and the staff of the General Clinical Research Center. REFERENCES
1. Leyendecker. G., Wildt, L., and Hansmann, M.: Pregnancies following chronic intermittent (pulsatile) administration of GnRH by means of a portable pump ("Zyklomat"): A new approach to the treatment of infertility in hypothalamic amenorrhea, J. Clin. Endocrinol. Metab. 51:1214,1980. 2. Reid, L. R., Leopold, G. R., and Yen, S. S. C.: Induction of ovulation and pregnancy with pulsatile luteinizing hormone releasing factor: Dosage and mode of delivery, Ferti!. Steril. 36:553, 1981. 3. Midgley, A. R., Niswender, G. D., Gay, V. C., and Reichert, L. F.: Use of antibodies for characterization of gonadotropins and steroids, Recent Prog. Horm. Res. 27: 235, 1971. 4. Araki, S., Ferin, M., Zimmerman, E. A., and Vande Wiele, R. L.: Ovarian modulation of immunoreactive gonadotropin-releasing hormone (GnRH) in rat brain: Evidence for a differential effect on the anterior and mid hypothalamus, Endocrinology 96:644, 1975. 5. Abraham, G. E., Odell, W. D., Swerdloff, R. S., and Hopper, K.: Simultaneous RIA of plasma, FSH, LH, progesterone, 17-hydroxyprogesterone and estradiol-17,B during the menstrual cycle, J. Clin. Endocrinol. Metab. 34: 312,1972. 6. Knobil, E., Plant, T M., Wildt, L., Belchet, P. E., and Marshall, G.: Control of the rhesus monkey menstrual cycle: Permissive role of hypothalamic gonadotropinreleasing hormone, Science 207:1371,1980. 7. Leyendecker, G., and Wildt, L.: Induction of ovulation with chronic intermittent (pulsatile) administration of GnRH in hypothalamic amenorrhea, in Research on Steroids, Amsterdam, 1982, vol. 10, Excerpta Medica, p. 275. 8. Antunes, J. L., Carmel, P. W., Housepian, E. M., and Ferin, M.: Luteinizing hormone-releasing hormone in human pituitary blood, J. Neurosurg. 49:382, 1978. 9. Wu, Ch., Motohashi, T, Abdel-Rahman, H. A., and Mikhail, G.: Estrogen-androgen balance in anovulation, Fertil. Steril. 36:61,1981. 10. Wildt, L., Hausler, A., Marshall, G., Hutchinson, J. S., Plant, T M., Belchetz, P. E., and Knobil, E.: Frequency and amplitude of gonadotropin-releasing hormone stimulation and gonadotropin secretion in the rhesus monkey, Endocrinology 109:376, 1981.
Discussion
DR. S. S. C. YEN, La Jolla, California. Within the past few years, several groups have demonstrated the effectiveness of pulsatile administration of GnRH in the activation of cyclic pituitary-ovarian axis. Induction of follicular maturation, ovulation, and pregnancy in a variety of patients with hypothalamic dysfunction, including congenital GnRH deficiency, functional hypothalamic amenorrhea, and panhypopituitarism due to previous stalk transaction, have been achieved. 1 Dr. Jewelewicz and his colleagues have presented their experience with pulsatile GnRH therapy and have confirmed the safety and ease with which ovulatory menstrual cycles can be established in hypogonadotropic amenorrheic patients (group 1). In our laboratory,
Pulsatile GnRH for ovulation induction
899
much smaller doses of GnRH per pulse (1 to 5 /Lg) have achieved ovulation and pregnancies. 2 It was anticipated (primarily on the basis of rhesus monkeys with arcuate nucleus lesions-the "hypothalamic clamp" model) that pulsatile administration of GnRH at a frequency of 60 minutes reinitiates ovulatory cycles with one dominant follicle, and that hyperstimulation should not occur because of the intrinsic feedback modulation of pituitary responsiveness to exogenous GnRH.3 This premise proves not to be the case, inasmuch as multiple pregnancies have already been reported; among 23 pregnancies induced by pulsatile GnRH, there have been two sets of twins and one set of triplets. 4 Moreover, we have deliberately induced multiple follicular maturations in women with normal cycles by using larger doses of GnRH pulses (10 /Lg).5 This finding may prove to be useful to the in vitro fertilization program. Thus, the internal ovarian feedback system can be overridden by excessive GnRH pulses. Nonetheless, pulsatile GnRH therapy will probably replace exogenous gonadotropin in the majority of anovulatory infertile patients. Pulsatile administration of GnRH has not been successful in anovulatory women with high LH/FSH ratios, as also shown by the data of Dr. Jewelewicz and associates. Under this condition, high levels of LH are likely to reflect increased GnRH receptors on the gonadotrope occasioned by enhanced GnRH secretion. Thus, the principle of up-regulation by constant GnRH pulses is not operative in these patients, since the GnRH-gonadotrope activities are already elevated. However, attenuation of this high functional activity by several pharmacologic maneuvers can potentially be achieved before initiation of pulsatile GnRH therapy. The subject addressed by Dr. Jewelewicz and his associates will undoubtedly improve further by detailed examination of the dose and frequency of GnRH pulses synchronized with the underlying GnRH-gonadotropic status. I have the following questions to ask of Dr. Jewelewicz. (1) Why was hCG used to "trigger ovulation"? One of the major advantages of pulsatile GnRH therapy is the ability of endogenous estradiol to initiate spontaneous midcycle surge and, consequently, provide for timely ovulation. (2) What was the rationale behind the use of pulsatile GnRH in a case with short luteal phase? (3) What explanation can you offer for the failure of response in patients with high LH/FSH ratios? (4) Finally, what are the bases for approximating that 20 /Lg per pulse is comparable to human portal blood GnRH values?
REFERENCES
1. Yen, S. S. C.: Clinical applications of gonadotropinreleasing hormone and gonadotropin-releasing hormone analogs, Ferti!. Steril. 39:257, 1983. 2. Miller, D. S., Reid, R. R., Cetel, N. S., Rebar, R. W., and Yen, S. S. c.: Conception following ovulation induction by pulsatile administration of low doses of gonadotropin-
900
Loucopoulos et al. Am.
releasing hormone (GnRH) in women with hypothalamic amenorrhea, JAMA. In press. 3. Knobil, E.: The neuroendocrine control of the menstrual cycle, Recent Prog. Horm. Res. 36:53, 1980. 4. Leyendecker, G., and Wildt, L.: Induction of ovulation with chronic-intermittent (pulsatile) administration of GnRH in hypothalamic amenorrhea, J. Reprod. Fertil. In press. 5. Liu, J. H., Durfee, R., Muse, K., and Yen, S. S. C.: Induction of multiple ovulation by pulsatile administration of gonadotropin-releasing hormone (GnRH), Fertil. Steril. 40: 18, 1983. DR. JEWELEWICZ (Closing). I will try to answer Dr. Yen's questions. First, why was hCG used to trigger ovulation? We us~d 10,000 IV of hCG in two patients to trigger ovulatIOn after they reached follicular maturation, in the same way that we do in Pergonal-treated cycles. Indeed, both patients ovulated, and had a normal luteal phase. This is not absolutely necessary. When pulsatile GnRH is continued, the patients will have an LH surge, ovulate, and have a normal luteal phase, as long as GnRH is continued, and adequate levels of LH and FSH are maintained. The reason we used hCG was to learn whether it can shorten the time that GnRH is given, and whether it can replace GnRH in maintenance of the luteal phase-and, indeed, it can. In two other patients, we discontinued GnRH in mid~uteal p~a~e a.nd supplemented luteal function by givmg two mJectlOns of 2,500 IV of hCG 48 hours apart, as was done by Dr. Yen's group (although they used only 1,500 IV of hCG times three) and by Dr. Leyendecker's group. .Second, What was the rationale for treating patients WIth a short luteal phase with GnRH? I regret to say that, up until now, we did not know exactly how to treat adequately and successfully patients with an inadequate luteal phase. There are several regimens of treatment, none of which is too successful. Our success r~te in treating the short luteal phase with clomiphene Citrate or gonadotropins is far from ideal. We decided to treat two patients with GnRH to learn whether we
J.
April I, 1984 Obstet. Gynecol.
could induce ovulation and improve luteal function. So far, we have not been successful. . Third.' Why did we fail to induce ovulation in patients with polycystic ovary syndrome who had high levels of LH and normal levels of FSH? The question is a most interesting and challenging one. These patients have ovulatory dysfunction because of the high LHI FSH. ratio, which is probably controlled centrally. A conSiderable degree of our understanding of the syndrome comes from Dr. Yen's work. We tried to change the LH/FSH ratio by decreasing the frequency of the GnRH pulses from 90 to 180 minutes, since Dr. Knobil showed, in rhesus monkeys with lesions of the arcuate nucleus, that a lower frequency of GnRH pulses resulted in an increased secretion of FSH over LH, and the LH/FSH ratio changed. However, that has not happened in our patients. In only one patient who received GnRH pulses every 180 minutes was a 50% decrease in urinary LH found over a 24-hour period, as co.mpared to the cycle when GnRH was given every 90 mmutes. However, this patient did not ovulate. I believe ~ha.t by more careful study with a different dosage a.nd tImmg and perhaps initial gonadotropin suppresSion, we may be a~le to manipulate and change the LH/FSH ratio and mduce ovulation and pregnancy in these patients. A review of the literature discloses that most ~nve.stigat~rs seem to have problems inducing ovulatIon m patients with polycystic ovary syndrome. Fourth, Why did we use a 20 ILg/pulse of GnRH? I agree with Dr. Yen that this is a relatively high dose, and, indeed, he is using very successfully 1 to 5 ILgl pulse. The literature reveals that Dr. Leyendecker uses from 2.5 to 20 ILg/pulse, Dr. Shoemaker 10 to 20 ILgl pulse, and up to 100 ILg/pulse. We used the same dose as Dr. Leyendecker (20 ILg/pulse), and since we had a pregnancy on this dose, we continued it. Twenty micrograms per pulse gives a blood level of GnRH comparable to the level found in portal blood (300 to 2,000 pg/ml). We are testing various doses (2, 5, 10, and 20 ILg/pulse), and I agree with Dr. Yen that smaller doses of GnRH may be as effective, if not more effective.