- Email: [email protected]
Influence of the ovary on parameters of LH secretion during the recovery from buserelin-induced desensitization
European Journal of Obstetrics % Gynecology and Reproductive Biology 55 (1994) 187-192
ELSEVIER
Influence of the ovary on parameters of LH secretion during the recovery from buserelin-induced desensitization A. Cano*a,
R. Aliaga”, C. Puirtoias”,
M. Tortajadaa,
C. Arm~rob
“Department of Pediatrics, Obstetrics and Gynecology, Facultnd de Medicina - Hospital Clinico Uni~ersitario, Av. Biasco lb&e: 11, 46010 Vaiencia, Spain bDepartment of Statistics and Operations Research. University of Valencia, Valencia, Spain
Accepted 7 April 1994
Abstract
This study examined the effect of the ovary on LH pulsatiIity and on the secretory ~rfo~ance of gonadotrophes during the phase of recovery after treatment with buserelin, a GnRH analogue. We included 12 patients, who received busereiin (1.2 mglday, intranasally for 3 months) as a reductive therapy for uterine leiomyomatosis prior to hysterectomy. Six patients were oophorectomized and the other 6 patients had their ovaries preserved. LH was measured in samples taken basally up to 36 days after suppression of buserelin. LH pulsatility was studied on day 9 along a 24-h cycle, and the response of the hormone to a double-stimulus GnRH test on days 0, 9, 20, and 34. The concentration of LH reached normal premenopausal levels after an average of 2 weeks in women with ovaries but increased until 4-5 weeks in oophorectomized patients. The pulsatility of LH on day 9 was similar for both groups, but parameters related to LH amplitude or to baseline secretory activity of gonadotrophes were higher in the oophorectomized women. The response of LH to the GnRH tests was also significantly higher in the oophorectomized group from day 9. The conclusions are as follows. (1) At the early stage of recovery from desensitization, as represented by day 9, LH pulsatility was not substantially influenced by the presence or absence of the ovary. (2) There was an increase in parameters related to the amplitude of the LH bursts in the oophorectomized women. Although a higher amplitude of the endogenous GnRH pulses cannot be discarded, most probably that difference is due to a higher sensitivity at a pituitary level, as reflected by the GnRH stimulation tests. Keywords: LH; GnRH analogue; Desensitization;
Recovery; Oophorectomy
1. Introduction The analogues of GnRH induce pituitary desensitization, which is commonly used for many clinical purposes [If. During desensitization, the cellular machinery of the gonadotrophes, the pituitary cells involved in the synthesis and secretion of gonadotrophins, remains silent despite the availability of receptors for GnRH at the membrane [2], This lack of cellular response despite the availability of receptors suggests that the process of desensitization is linked to events mediated by GnRH at post-receptor level [3-51. * Corresponding
author.
~28-2243/94/$07.~ SSDI 0028-2243(94)0
0
The recovery from desensitization has received little attention in the literature. Studies performed in vivo in the human have shown that it lasts a variable period, which oscillates from 3 to 6 weeks, although the data were obtained from women with endocrine dysfunctions I&7]. This is in contrast to the in vitro situation, where the full functional activity of the gonadotroph recovers after 3-5 days [3]. The reason for that difference remains unclear, although that in vitro study was performed on gonadotrophes obtained from female weanling rats, where immaturity might influence the cell response. Beside the relatively sparse data in the human, the potential role that factors like the presence or absence of the ovary might have in modulating that process has not
1994 Elsevier Science Ireland Ltd. All rights reserved 1862-2
188
A. Cano et al. /Eur. J. Obstet. Gynecol. Reprod. Biol. SJ (1994) 187-192
been explored. Likewise, the data reported with regard to the behavior of the hypothalamus have been mostly obtained for GnRH antagonists and have concentrated on the ‘in use period’ [8,9]. In a recent study, pulsatility of LH was not altered in men despite the hypogonadal status induced by a GnRH antagonist, although the LH amplitude was reduced [lo]. Our investigation has focused on two points: (1) the behavior of the hypothalamic pacemaker in the early stage of the period of recovery from desensitization after use of buserelin, a GnRH analogue; and (2) the functional performance of the gonadotrophes along that period. For the first objective, we tested the hypothesis that, in the early phase after suppression of the analogue, the variable ‘ovary’ does not affect the pattern of GnRH secretion, although aspects related to the amplitude of the LH bursts might be influenced by another variable, the sensitivity of the gonadotrophes. To clarify that point, we added a second objective, which included the study of basal levels of LH and the response of the hormone to an overload of GnRH at different moments during that phase of recovery. The results for both objectives were analysed according to the presence or absence of the ovary. 2. Patients and methods 2.1. Patients Twelve premenopausal women, aged 42-46, volunteered to participate in the study after full explanation of its purpose and characteristics. The study had the approval of the research committee of the hospital. The premenopausal status of the patients was based on both the absence of clinical features of ovarian failure, such as hot flushes or irregularities in the cycle length, and a level of gonadotrophins and estradiol within the premenopausal range (4- 11 I.U./l for FSH and LH, and >35 pg/ml for estradiol, in the follicular phase). Most of the patients suffered from clinically disturbing uterine leiomyomatosis: hypermenorrhoea in 4 cases, abdominal discomfort in 5 patients, and difficulty in voiding the urine bladder in 1 case. The other 2 patients had a uterine size larger than 12 weeks of pregnancy. The patients were scheduled to follow a protocol which included a course of 3-month reductive treatment with buserelin (D-Ser,Tbu6, LHRH-EA”, HOE 766 Buserelin, Hoechst, Germany) prior to surgery. Each patient received 1.2 mg/day divided into four daily doses given intranasally, starting in the middle of the luteal phase of the cycle. Total hysterectomy was performed through an abdominal incision, and the ovaries were preserved or removed according to a decision taken in agreement between the patient and surgeon, after full explanation to the former of potential advantages and drawbacks. Six women, aged 44.8 f 0.4 (mean f S.E.M.), were oophorectomized. The age in the non-oophorectomized
group was 44.2 f 0.6. Buserelin was administered until 12 h before surgery and then suppressed. None of the patients required intra- or post-operative blood administration. 2.2. Blood sampling protocol The blood extractions were performed at 09:OO h. One sample of blood was obtained every third day from day 0 to 36 after surgery. On day 9, a 24-h protocol was performed on each woman. For that, blood was extracted every 20 min from the forearm vein through an indwelling 18-gauge i.v. cannula kept patent with an heparinized obturator in the between-extraction periods. At the end of the 24-h cycle, a bolus of 0.1 mg of synthetic GnRH (Luforan, Serono, Madrid, Spain) was injected through the cannula and samples taken at 15min intervals for 1 h. Then, a second I-mg GnRH bolus was injected and samples obtained at the same intervals for another hour. That 2-h protocol with GnRH was also performed during buserelin administration (the day before surgery, considered as day 0), and then repeated on the 20th and the 34th day after surgery. The blood was permitted to clot at room temperature, and the serum was separated and stored at -40°C until assay. 2.3. Hormone assays All samples from an individual subject were analysed in duplicate by RIA in the same assay with the use of a highly sensitive commercial kit based upon a doubleantibody method (Bio-Merieux, Madrid, Spain), where the first antibody was coated on the assay tube wall and the second was labelled with iodine-125. For LH measurements the reference standard was calibrated to the first IRP MRC 68/40. The cross reaction of this assay with human FSH, TSH or HCG was lower than 0.1% The detection limit was 0.4 I.U./l. To assess more precisely the degree of intra-assay measurement error for pulse analysis, 15 replicates from a pool prepared by removing 50 ~1 serum from each of the 72 individual samples from the 24-h cycle were assayed. A similar procedure was followed for each of the GnRH tests, where the pool was prepared by removing 0.25 ml from each of the 8 post-GnRH individual samples. In the oophorectomized group, the individual pool intra-assay coefficients of variation averaged 8.8 + 1.8 for the 24-h cycle samples, 9.3 f 1.7 for the rest of basal samples, and 10.1 f 2.3, 8.9 f 1.3, 10.9 + 2.2, and 12.8 f 2.0 for the day 0, day 9, day 20 and day 34 GnRH tests, respectively. In the non-oophorectomized women, the corresponding intra-assay coefficients of variation averaged 9.1 + 1.4 for the 24-h cycle samples, 10.3 f 1.8 for the rest of basal samples, and 10.8 f 1.6, 10.4 f 1.5, 11.0 f 1.9, and 9.9 f 2.1 for the corresponding GnRH tests. The interassay coefficient of variation was less than 15% in all the cases. Estradiol and progesterone were also measured by RIA (Bio-
A. Cuno et al. /Eur.
J. Obster. Gynecol. Reprod. Biol. 55 (1994) 187-192
Merieux, Madrid, Spain), where the hormone in the sample competed with a fixed amount of ‘251-labelled steroid for a number of specific anti-steroid antibodies coated in the inner wall of the tube. The coefficients of variation for the range of concentration in our samples were 9.9% and 8.8% (intra-assay) and 16.2% and 13.5% (interassay) for estradiol and progesterone, respectively.
189
under each 24-h individual LH profile), mean peak area (expressed as the product of the mean peak amplitude that is above the lower of the two flanking nadirs times the peak duration), incremental peak amplitude (expressed as the algebraic difference between the maximal peak height and the prepeak nadir value), mean peak height as percentage increase (defined by the percentage increase of the mean peak amplitude above preceding nadir), and the average valley mean level (defined as the overall mean value within the entire valley, including the flanking nadirs that immediately precede or follow peaks).
2.4. Analyses of LH pulses Each LH data series was analysed for significant pulse parameters (amplitude and frequency) using the cluster algorithm [ 111. This computerized pulse detection program searches for increases and decreases in a hormone series by applying a pooled t statistic to hormone concentrations (clusters). To be detected as a pulse, the peak had to be preceded by a significant up-stroke and followed by a significant down-stroke. We chose a cluster configuration of 2 x 1 (2 data points for the nadir and 1 for the peak) and a I statistic of l/l. This criterion yields the highest score for sensitivity and positive accuracy for the sampling intensity (20-min) of our protocol [12]. In addition to the amplitude and frequency of the LH pulses, the cluster algorithm determines the following parameters: total area value (defined as the area
2.5. Statistics The results are presented as the mean f S.E.M. The possible ovarian influence on specific LH pulse properties was tested between both groups of women by the non-parametric Mann-Whitney U-test. This test was also used to detect differences in the rest of comparisons between groups. Differences between groups during the GnRH tests were evaluated by analysis of variance for repeated measurements, by use of the computer package SPSS. Significance was construed for P < 0.05.
3.500
J
V
3.000
-A 7.000
2.500
6.000
2.000
5.000
=i E
I 500
4.000
3 E
1.000 a.500
I J
0.000
0.000 a boo
a ,200
0.400
0.600
0.800
1.000 x
E 3
a g
L 0.000
1.200
0.2aa
0.400
0.600
0.800
1.000
lo+03
I I 400 I 10+m
1 200
24.000
6.000
B
Vlf-V-Xf
20.000
4.000
16.000
3 000
12.000
2.000
8.000
1.000
4.000
0.000
0.000 b.bbO
a ,200
0.400
0.600
0.800
1.000
W-W-UD
V
5.000
I 0.000
1.280
0.200
0.400
0.600
0.800
1.000
1 2oa
J I 423
x 10 ‘03
x 10 +03
TIME
.
(MINI
Fig. I, Representative individual profiles of immunoreactive LH in women with (A and B) and without (C and D) ovaries on day 9 after suppression of i.n. buserelin. Each subpanel shows the time-course (horizontal axis. min) of the mean serum LH concentrations (ml.U./ml) assayed in duplicate for every repetitive sampling session. Individually significant LH pulses are depicted by the continuous schematized line in the uppermost portion of each subpanel. as detected by computer analysis.
190
A. Cam
et al. f Eur.
J. Obsrer.
Gynecof.
3. Results 3.1. Clinical and hormonal response to buserelin The size of the uterus was reduced by 37.3 f 6.80/o,as assessed by ultrasound, after the third month of therapy. Maximal desensitization, as defined by stable low LH and estradiol values, was obtained after 3 weeks of initiating therapy. 3.2. Individual LH profiles Fig. 1 shows representative protiles of LH values obtained in serum for some of the women on day 9 after suppression of buserelin. Patterns of LH were clearly pulsatile in all the cases and no substantial difference occurred between the diurnal and nocturnal patterns. The baseline level, as defined by the value obtained from the 50-~1 aliquot pool, was already higher for the oophorectomy group and so was the total area value of 8638 f 2822, which was significantly higher (P < 0.05) than the corresponding value of 3564 f 735 in the non-castrated women (min. III/I). The mean peak area in the oophor~tomized patients (208 f 123) demonstrated a significant increase when compared with the value of 95 f 38, found in women with preserved ovaries (P < 0.05, min - III/l). Those differences between both groups of women occurred in spite of similar values for estradiol (23.2 f 3.6 for women with ovaries, and 21.2 f 1.6 for oophorectomized women). 3.3. Pulsatile LH The parameters of LH pulsatility are summarized in Table 1. There was no difference between both groups for parameters like the frequency of the pulses or the pulse interval. Some properties related with the LH pulse amplitude, like the incremental (IUII) or the mean peak height as percentage increase, were already different in both groups, and the magnitude of that difference almost reached statistical difference (P < 0.06); the
~gprod.
BittI.
55 (1994)
187-192
absolute (maximal) LH peak amplitude, however, was significantly higher in the oophorectomized group (P c 0.05). This result is consistent with the differences found for a parameter like the interpulse valley mean LH concentration (P < 0.02), which reflects increments in the baseline secretory activity of the gonadotroph for the oophorectomized group. 3.4. Basal levels of LH and responses to GnRH The mean basal morning values of LH obtained at 3day intervals stabilized within the normal range for premenopausal women after the second week (time-point corresponding to day 18) in the non-castrated women. This means that the mean level of LH ful~lled three conditions: it reached a value within the premenopausal range (4-11 W/I for the follicular phase in our hands), it did not decrease below that range subsequently, and the mean value of all the samples corresponding to the 3 subsequent time-points was not significantly higher, Increases in the concentration of estradiol were only apparent after the second week in that group, with levels within the range of an early/mid proliferative phase. Three women underwent a pre-ovulatory pattern (i.e. discharges of LH coexisting with estradiol increments) on day 27, two more on day 30, and the other woman on day 33. Positive values were obtained for progesterone in the subsequent blood samples for each case. The oophorectomized group suffered a maintained increase in the level of LH along the successive timepoints, which seemed to curve between 30 and 36 days, whereas the level of estradiol remained low.
4
Table 1 Parameters (mean f S.E.M.) of 24-h pulsatile LH secretion on day 9 post-buserelin in women whose ovaries were either preserved or ablated on the same day of interruption of the analogue
Pulse frequency Pulse amplitude (LUJl) Mean peak height as o/u increase Mean peak width (min) lnterpulse interval (mm) lnterpulse valley mean concentration (LUA)
Non-castrated
Castrated
7.1 f 1.6 4.7 * 1.6 223.1 zt 35.6
8.5 f 2.3 9.9 * 2.Y 179.4 f 21.1
62.3 f 4.9 119.5 f 17.9 2.6 f 0.7
73.5 * 3.1 108.6 f 2.2 7.1 f 1.2**
*P < 0.05 compared with corresponding women. **P c 0.02 compared with corresponding women.
values in the non-castrated values in the non-castrated
34
DAYS
Fig. 2. Response (mean f S.E.M.) of serum LH to stimulation with a O.l-mg i.v. bolus/h of synthetic GnRH administered in two consecutive doses, in women in which the ovaries were ablated (solid circles) or preserved (solid triangles). The tests were performed during buserelin treatment (day 0) and on days 9,20, and 34 after interrupting the drug. The analysis of variance for repeated measurements demonstrated that the response was different in either group (P < 0.001). The LH concentration has been represented according to a semi-logarithmic scale.
A. Cuno et al. /Eur.
J. Ubstet.
Gynecol.
The responses of LH to GnRH stimulation during buserelin administration and during the recovery period are shown in Fig. 2. As expected, the gonadotropin response to GnRH during administration of buserelin was inhibited in both groups, thus confirming the effective desensitization induced by the anaiogue. The recovery of responsiveness to GnRH was already evident for both groups by day 9 after buserelin interruption, with a response that was higher for the oophorectomized women at certain time-points, in addition to the basal. The magnitude of LH response increased again for both groups by day 20, where a difference in favor of the oophorectomized women was already clear. The increased response in the castrated patients was further pronounced on day 34 where, in contrast, the women with preserved ovaries exhibited a lower magnitude in LH secretion than that for day 20. That different response of LH to GnRH in both groups was statistically significant (P < 0.001). 4. Discussion Our experimental design provides a model for understanding some aspects of the time-course of the reactivation of the hypothalamic-pituitary unit after desensitization in the human within a context where the function of the hypothalamic level remains preserved, as well as a possibility of studying the ovarian influence on that process. Using this model we were able to observe that, in presence of a functional ovary, the reactivation of the gonadotroph proceeds slowly along 2 weeks and therefore, was slower than anticipated from the in vitro data. In the oophorectomized group, in contrast, the lack of the modulating action of the ovary was accompanied by an increased functional activity of the gonadotrophes until at least 1 month from suppression of the analogue. The presence of the ovary also determined that the morning basal levels of LH were higher for the oophorectomized group, an increase that was attained even in the absence of differences for estradiol. This suggests that perhaps other nonestrogenic factors of ovarian origin participate in regulating the circulating levels of LH, something that has also been observed in menopausal women subjected to hormonal therapy, where levels of estradiol similar to those found in an advanced follicular phase are unable to abate the concentration of LH to premenopausai ranges [ 131. We have also obtained information about specific properties of pulsatile LH release. For that, we chose day 9, a moment in which the possibility of an effect of the previous anesthesia or surgery-induced stress is unlikely [14]. Additionally, the in vitro data show that on that day the functional performance of the gonadotroph is already recovered and, according to our results, the output of estradiol by the ovary in the women who had it preserved does not determine a systemic concentration
Reprod. Biol. 55 (1994)
187-192
191
substantially different from the oophorectomized group. In this way we could gather information about the pulsatile behavior of the hypothalamic pacemaker within an endocrine scenario similar to that of desensitization in the presence of the analogue. Additionally, we could envision a possible early influence of the ovary in the functional response of the hypothalamic-pituitary unit. We analysed our data through the cluster algorithm under conditions (cluster configuration of 2 x 1 and a i statistic of 1) which offer levels of sensitivity and positive accuracy comparable with more frequent sampling protocols [12]. Our data showed that 9 days after buserelin the presence or absence of the ovary was irrelevant regarding the frequency of the pulses. This reflects that the hypothalamic pacemaker still operated in concordance for both groups at this stage and that, most likely, its pattern of secretion was not substantially modified from the period in which the analogue was being used. The similitude between both groups also involved some parameters of the LH pulse amplitude, like the incremental or the fractional LH peak amplitudes. The presence of the ovary, in contrast, was associated to a lower interpulse valley mean LH concentration, an indication in favour of a role for the ovary in modulating the baseline secretory activity of the gonadotroph. This is also supported by the higher values found for the mean peak area and the integrated serum LH concentration in the oophorectomized women, and by the significant increase of the absolute LH peak amplitude. Globally, our data suggest that during the process of recovery from desensitization, the presence of the ovary resulted in a relative decline in either the efficiency of the cellular machinery in response to the GnRH stimuli or the magnitude, rather than the frequency, of the secretion of GnRH itself or both. Our data obtained with the GnRH tests on the same day, 9, support the hypothesis of an already higher sensitivity to the decapeptyde by the gonadotrophes in oophorectomized women. This was obvious even in response to a dose of 0.1 mg GnRH, which is clearly supramaximal. The difference with the ovarian-bearers at day 9 was amplified with the GnRH tests performed subsequently, due to a maintained increase in response in the oophorectomized women until day 34, while the other group stabilized by day 20. Among the factors with potential regulatory effect on the expression of gonadotropin subunit genes, estradiol, some gonadal peptides, and GnRH itself are relevant. Estradiol acts on the gonadotroph at posttranslational level, and exerts both positive and negative effects on gonadotropin secretion and synthesis [ 15,161. The facilitatory action of estradiol on the GnRH-stimulated secretion of LH has been well established for both immunoactive and bioactive forms of the hormone [17]. That effect consists of enhancement of the pituitary response to successive GnRH stimuli, an effect mediated
A. Cano et al. /Eur. J. Obstet. Gynecol. Reprod. Biol. 55 (1994) 187-192
192
by the augmentation of the duration of the secretory burst [ 181. That progressive response could be clearly observed after the second GnRH bolus in our noncastrated group. That increase, however, was insuflicient to compensate for the high response found after the first injection in the castrated women. The role of the gonadal peptides has not been adequately explained at present although experimental data have implicated inhibin in the process [19,20]. Finally, there is abundant evidence that GnRH does increase the synthesis of LH. The rapid increase of LH in normal women after castration is well documented in the literature and, at a molecular level, it has been shown that the postcastration level of gonadotropin mRNAs declined in the rat after treatment with a GnRH antagonist [21]. Despite the similar pulse frequency observed on day 9, it is possible that a higher secretory activity of GnRH is already present in the oophorectomized women since the early moments of the recovery. In summary, our study has analysed in detail the phase of recovery from buserelin-induced desensitization. Our data show that pulsatility of LH is not modified by the ovary on day 9 of that period, but that there is an increase in parameters related with the amplitude of the LH bursts in the oophorectomized women. Although a higher amplitude of the endogenous GnRH pulses cannot be discarded, most probably that difference is due to a higher sensitivity at a pituitary level, as reflected by the GnRH stimulation tests.
6
7
8
9
10
11
I2
13
14
Acknowledgements
We are most grateful for the assistance provided by Dr J. Veldhuis, who generously supplied us with the computer program for pulse analysis. This work was supported by grant 90/0455 from the Fondo de Investigaciones Sanitarias, and grant 90-0069 from Direccion General de Investigation Cientilica y Ttcnica.
15
16 17
References Conn PM, Crowley WF. Gonadotropin-releasing hormone and its analogues. N Engl J Med 1991; 324: 93-103. Gorospe WC, Conn PM. Agents that decrease gonadotropinreleasing hormone (GnRH) receptor internalization do not inhibit GnRH-mediated gonadotrope desensitization. Endocrinology 1987; 120: 222-229. Jinnah HA, Conn PM. GnRH-stimulated LH release from rat anterior pituitary cells in culture: refractoriness and recovery. Am J Physiol (Endocrinol Metab) 1985; 249: E619-E625. Conn PM. Staley DD, Yasumoto T et al. Homologous desensitization with gonadotropin-releasing hormone (GnRH) also diminishes gonadotrope responsiveness to maitotoxin: a role for the GnRH receptor-regulated calcium ion channel in mediation of cellular desensitization. Mol Endocrinol 1987; 1: 154- 159. Uemura T, Yanagisawa T, Shirasu K et al. Mechanisms involved in the pituitary desensitization induced by gonadotropin-releasing hormone agonists. Am J Obstet Gynecol 1992; 167: 283-291.
18
19
20
21
Calogero AE, Macchi M, Montanini Vet al. Dynamics of plasma gonadotropin and sex steroid release in polycystic ovarian disease after pituitary-ovarian inhibition with an analog of gonadotropin-releasing hormone. J Clin Endocrinol Metab 1987; 64: 980-985. MacLeod AF, Wheeler MJ, Gordon P et al. Effect of long-term inhibition of gonadotrophin secretion by the gonadotrophinreleasing hormone agonist, buserelin, on sex steroid secretion and ovarian morphology in polycystic ovary syndrome. J Endocrinol 1990; 125: 317-325. Hall JE, Whitcomb RW, Rivier JE et al. Differential regulation of luteinizing hormone, follicle-stimulating hormone, and free alpha-subunit secretion from the gonadotrope by gonadotropinreleasing hormone (GnRH): evidence from the use of two GnRH antagonists. J Clin Endocrinol Metab 1990; 70: 328-335. Pavlou SN, Veldhuis JD, Lindner J et al. Persistence of concordant luteinizing hormone (LH), testosterone, and alpha-subunit pulses after LH-releasing hormone antagonist administration in normal men. J Chn Endocrinol Metab 1990; 70: 1472-1478. Salameh W, Bhasin S, Steiner BS et al. Effect of improved assay sensitivity on luteinizing hormone pulse detection after gonadotropin-releasing hormone antagonist treatment in man. J Clin Endocrinol Metab 1992; 75: 1479-1483. Veldhuis JD, Johnson ML. Cluster analysis: a simple, versatile, and robust algorithm for endocrine pulse detection. Am J Physiol 1986; 250: E486-493. Urban RJ, Johnson ML, Veldhuis JD. In vivo biologycal validation and biophysical modeling of the sensitivity and positive accuracy of endocrine peak detection. 1. The LH pulse signal. Endocrinology 1989; 124: 2541-2547. Chetkowski RJ, Meldrum DR, Steingold KA et al. Biologic effects of transdermal estradiol. N Engl J Med 1986; 314: 16151620. Shelley HJ. The use of chronically catheterized foetal lambs for the study of foetal metabolism. In: Foetal and neonatal physiology. Proceedings of the Sir Joseph Barcroft Centenary Symposium. Cambridge UK: Cambridge University Press 1973; 360381. Veldhuis JD, Evans WS, Rogol AD et al. The pituitary selfpriming actions of gonadotropin-releasing hormone: kinetics of estradiol’s potentiating effects on GnRH-facilitated LH and FSH release in healthy post-menopausal women. J Clin Invest 1986; 77: 1849-1856. Gharib SD, Wierman ME, Shupnik MA et al. Molecular biology of the pituitary gonadotropins. Endocr Rev 1990; 1 I: 177-199. Urban RJ, Veldhuis JD, Dufau ML. Estrogen regulates the gonadotropin-releasing hormone-stimulated secretion of biologically active luteinizing hormone. J Clin Endocrinol Metab 1991; 72: 660-666. Quyyumi SA, Pinkerton JV, Evans WS et al. Estradiol amplifies the amount of luteinizing hormone (LH) secreted in response to increasing doses of gonadotropin-releasing hormone by specilically augmenting the duration of evoked LH secretory events and hence their mass. J Clin Endocrinol Metab 1993; 76: 594-600. Culler MD, Negro-Vilar A. Endogenous inhibin suppresses only basal follicle-stimulating hormone secretion but suppresses all parameters of pulsatile luteinizing hormone secretion in the diestrous female rat. Endocrinology 1989; 124: 2944-2953. Muttukrishna S. Knight PG. Effects of crude and highly purified bovine inhibin (M, 32 000 form) on gonadotrophin production by ovine pituitary cells in vitro: inhibin enhances gonadotrophinreleasing hormone-induced release of LH. J Endocrinol 1990; 127: 149-159. Wierman ME, Rivier J, Wang C. Gonadotropin-releasing hormone (GnRH) dependent regulation of gonadotropin subunit mRNA levels in the rat. Endocrinology 1989: 124: 272-278.
ELSEVIER
Influence of the ovary on parameters of LH secretion during the recovery from buserelin-induced desensitization A. Cano*a,
R. Aliaga”, C. Puirtoias”,
M. Tortajadaa,
C. Arm~rob
“Department of Pediatrics, Obstetrics and Gynecology, Facultnd de Medicina - Hospital Clinico Uni~ersitario, Av. Biasco lb&e: 11, 46010 Vaiencia, Spain bDepartment of Statistics and Operations Research. University of Valencia, Valencia, Spain
Accepted 7 April 1994
Abstract
This study examined the effect of the ovary on LH pulsatiIity and on the secretory ~rfo~ance of gonadotrophes during the phase of recovery after treatment with buserelin, a GnRH analogue. We included 12 patients, who received busereiin (1.2 mglday, intranasally for 3 months) as a reductive therapy for uterine leiomyomatosis prior to hysterectomy. Six patients were oophorectomized and the other 6 patients had their ovaries preserved. LH was measured in samples taken basally up to 36 days after suppression of buserelin. LH pulsatility was studied on day 9 along a 24-h cycle, and the response of the hormone to a double-stimulus GnRH test on days 0, 9, 20, and 34. The concentration of LH reached normal premenopausal levels after an average of 2 weeks in women with ovaries but increased until 4-5 weeks in oophorectomized patients. The pulsatility of LH on day 9 was similar for both groups, but parameters related to LH amplitude or to baseline secretory activity of gonadotrophes were higher in the oophorectomized women. The response of LH to the GnRH tests was also significantly higher in the oophorectomized group from day 9. The conclusions are as follows. (1) At the early stage of recovery from desensitization, as represented by day 9, LH pulsatility was not substantially influenced by the presence or absence of the ovary. (2) There was an increase in parameters related to the amplitude of the LH bursts in the oophorectomized women. Although a higher amplitude of the endogenous GnRH pulses cannot be discarded, most probably that difference is due to a higher sensitivity at a pituitary level, as reflected by the GnRH stimulation tests. Keywords: LH; GnRH analogue; Desensitization;
Recovery; Oophorectomy
1. Introduction The analogues of GnRH induce pituitary desensitization, which is commonly used for many clinical purposes [If. During desensitization, the cellular machinery of the gonadotrophes, the pituitary cells involved in the synthesis and secretion of gonadotrophins, remains silent despite the availability of receptors for GnRH at the membrane [2], This lack of cellular response despite the availability of receptors suggests that the process of desensitization is linked to events mediated by GnRH at post-receptor level [3-51. * Corresponding
author.
~28-2243/94/$07.~ SSDI 0028-2243(94)0
0
The recovery from desensitization has received little attention in the literature. Studies performed in vivo in the human have shown that it lasts a variable period, which oscillates from 3 to 6 weeks, although the data were obtained from women with endocrine dysfunctions I&7]. This is in contrast to the in vitro situation, where the full functional activity of the gonadotroph recovers after 3-5 days [3]. The reason for that difference remains unclear, although that in vitro study was performed on gonadotrophes obtained from female weanling rats, where immaturity might influence the cell response. Beside the relatively sparse data in the human, the potential role that factors like the presence or absence of the ovary might have in modulating that process has not
1994 Elsevier Science Ireland Ltd. All rights reserved 1862-2
188
A. Cano et al. /Eur. J. Obstet. Gynecol. Reprod. Biol. SJ (1994) 187-192
been explored. Likewise, the data reported with regard to the behavior of the hypothalamus have been mostly obtained for GnRH antagonists and have concentrated on the ‘in use period’ [8,9]. In a recent study, pulsatility of LH was not altered in men despite the hypogonadal status induced by a GnRH antagonist, although the LH amplitude was reduced [lo]. Our investigation has focused on two points: (1) the behavior of the hypothalamic pacemaker in the early stage of the period of recovery from desensitization after use of buserelin, a GnRH analogue; and (2) the functional performance of the gonadotrophes along that period. For the first objective, we tested the hypothesis that, in the early phase after suppression of the analogue, the variable ‘ovary’ does not affect the pattern of GnRH secretion, although aspects related to the amplitude of the LH bursts might be influenced by another variable, the sensitivity of the gonadotrophes. To clarify that point, we added a second objective, which included the study of basal levels of LH and the response of the hormone to an overload of GnRH at different moments during that phase of recovery. The results for both objectives were analysed according to the presence or absence of the ovary. 2. Patients and methods 2.1. Patients Twelve premenopausal women, aged 42-46, volunteered to participate in the study after full explanation of its purpose and characteristics. The study had the approval of the research committee of the hospital. The premenopausal status of the patients was based on both the absence of clinical features of ovarian failure, such as hot flushes or irregularities in the cycle length, and a level of gonadotrophins and estradiol within the premenopausal range (4- 11 I.U./l for FSH and LH, and >35 pg/ml for estradiol, in the follicular phase). Most of the patients suffered from clinically disturbing uterine leiomyomatosis: hypermenorrhoea in 4 cases, abdominal discomfort in 5 patients, and difficulty in voiding the urine bladder in 1 case. The other 2 patients had a uterine size larger than 12 weeks of pregnancy. The patients were scheduled to follow a protocol which included a course of 3-month reductive treatment with buserelin (D-Ser,Tbu6, LHRH-EA”, HOE 766 Buserelin, Hoechst, Germany) prior to surgery. Each patient received 1.2 mg/day divided into four daily doses given intranasally, starting in the middle of the luteal phase of the cycle. Total hysterectomy was performed through an abdominal incision, and the ovaries were preserved or removed according to a decision taken in agreement between the patient and surgeon, after full explanation to the former of potential advantages and drawbacks. Six women, aged 44.8 f 0.4 (mean f S.E.M.), were oophorectomized. The age in the non-oophorectomized
group was 44.2 f 0.6. Buserelin was administered until 12 h before surgery and then suppressed. None of the patients required intra- or post-operative blood administration. 2.2. Blood sampling protocol The blood extractions were performed at 09:OO h. One sample of blood was obtained every third day from day 0 to 36 after surgery. On day 9, a 24-h protocol was performed on each woman. For that, blood was extracted every 20 min from the forearm vein through an indwelling 18-gauge i.v. cannula kept patent with an heparinized obturator in the between-extraction periods. At the end of the 24-h cycle, a bolus of 0.1 mg of synthetic GnRH (Luforan, Serono, Madrid, Spain) was injected through the cannula and samples taken at 15min intervals for 1 h. Then, a second I-mg GnRH bolus was injected and samples obtained at the same intervals for another hour. That 2-h protocol with GnRH was also performed during buserelin administration (the day before surgery, considered as day 0), and then repeated on the 20th and the 34th day after surgery. The blood was permitted to clot at room temperature, and the serum was separated and stored at -40°C until assay. 2.3. Hormone assays All samples from an individual subject were analysed in duplicate by RIA in the same assay with the use of a highly sensitive commercial kit based upon a doubleantibody method (Bio-Merieux, Madrid, Spain), where the first antibody was coated on the assay tube wall and the second was labelled with iodine-125. For LH measurements the reference standard was calibrated to the first IRP MRC 68/40. The cross reaction of this assay with human FSH, TSH or HCG was lower than 0.1% The detection limit was 0.4 I.U./l. To assess more precisely the degree of intra-assay measurement error for pulse analysis, 15 replicates from a pool prepared by removing 50 ~1 serum from each of the 72 individual samples from the 24-h cycle were assayed. A similar procedure was followed for each of the GnRH tests, where the pool was prepared by removing 0.25 ml from each of the 8 post-GnRH individual samples. In the oophorectomized group, the individual pool intra-assay coefficients of variation averaged 8.8 + 1.8 for the 24-h cycle samples, 9.3 f 1.7 for the rest of basal samples, and 10.1 f 2.3, 8.9 f 1.3, 10.9 + 2.2, and 12.8 f 2.0 for the day 0, day 9, day 20 and day 34 GnRH tests, respectively. In the non-oophorectomized women, the corresponding intra-assay coefficients of variation averaged 9.1 + 1.4 for the 24-h cycle samples, 10.3 f 1.8 for the rest of basal samples, and 10.8 f 1.6, 10.4 f 1.5, 11.0 f 1.9, and 9.9 f 2.1 for the corresponding GnRH tests. The interassay coefficient of variation was less than 15% in all the cases. Estradiol and progesterone were also measured by RIA (Bio-
A. Cuno et al. /Eur.
J. Obster. Gynecol. Reprod. Biol. 55 (1994) 187-192
Merieux, Madrid, Spain), where the hormone in the sample competed with a fixed amount of ‘251-labelled steroid for a number of specific anti-steroid antibodies coated in the inner wall of the tube. The coefficients of variation for the range of concentration in our samples were 9.9% and 8.8% (intra-assay) and 16.2% and 13.5% (interassay) for estradiol and progesterone, respectively.
189
under each 24-h individual LH profile), mean peak area (expressed as the product of the mean peak amplitude that is above the lower of the two flanking nadirs times the peak duration), incremental peak amplitude (expressed as the algebraic difference between the maximal peak height and the prepeak nadir value), mean peak height as percentage increase (defined by the percentage increase of the mean peak amplitude above preceding nadir), and the average valley mean level (defined as the overall mean value within the entire valley, including the flanking nadirs that immediately precede or follow peaks).
2.4. Analyses of LH pulses Each LH data series was analysed for significant pulse parameters (amplitude and frequency) using the cluster algorithm [ 111. This computerized pulse detection program searches for increases and decreases in a hormone series by applying a pooled t statistic to hormone concentrations (clusters). To be detected as a pulse, the peak had to be preceded by a significant up-stroke and followed by a significant down-stroke. We chose a cluster configuration of 2 x 1 (2 data points for the nadir and 1 for the peak) and a I statistic of l/l. This criterion yields the highest score for sensitivity and positive accuracy for the sampling intensity (20-min) of our protocol [12]. In addition to the amplitude and frequency of the LH pulses, the cluster algorithm determines the following parameters: total area value (defined as the area
2.5. Statistics The results are presented as the mean f S.E.M. The possible ovarian influence on specific LH pulse properties was tested between both groups of women by the non-parametric Mann-Whitney U-test. This test was also used to detect differences in the rest of comparisons between groups. Differences between groups during the GnRH tests were evaluated by analysis of variance for repeated measurements, by use of the computer package SPSS. Significance was construed for P < 0.05.
3.500
J
V
3.000
-A 7.000
2.500
6.000
2.000
5.000
=i E
I 500
4.000
3 E
1.000 a.500
I J
0.000
0.000 a boo
a ,200
0.400
0.600
0.800
1.000 x
E 3
a g
L 0.000
1.200
0.2aa
0.400
0.600
0.800
1.000
lo+03
I I 400 I 10+m
1 200
24.000
6.000
B
Vlf-V-Xf
20.000
4.000
16.000
3 000
12.000
2.000
8.000
1.000
4.000
0.000
0.000 b.bbO
a ,200
0.400
0.600
0.800
1.000
W-W-UD
V
5.000
I 0.000
1.280
0.200
0.400
0.600
0.800
1.000
1 2oa
J I 423
x 10 ‘03
x 10 +03
TIME
.
(MINI
Fig. I, Representative individual profiles of immunoreactive LH in women with (A and B) and without (C and D) ovaries on day 9 after suppression of i.n. buserelin. Each subpanel shows the time-course (horizontal axis. min) of the mean serum LH concentrations (ml.U./ml) assayed in duplicate for every repetitive sampling session. Individually significant LH pulses are depicted by the continuous schematized line in the uppermost portion of each subpanel. as detected by computer analysis.
190
A. Cam
et al. f Eur.
J. Obsrer.
Gynecof.
3. Results 3.1. Clinical and hormonal response to buserelin The size of the uterus was reduced by 37.3 f 6.80/o,as assessed by ultrasound, after the third month of therapy. Maximal desensitization, as defined by stable low LH and estradiol values, was obtained after 3 weeks of initiating therapy. 3.2. Individual LH profiles Fig. 1 shows representative protiles of LH values obtained in serum for some of the women on day 9 after suppression of buserelin. Patterns of LH were clearly pulsatile in all the cases and no substantial difference occurred between the diurnal and nocturnal patterns. The baseline level, as defined by the value obtained from the 50-~1 aliquot pool, was already higher for the oophorectomy group and so was the total area value of 8638 f 2822, which was significantly higher (P < 0.05) than the corresponding value of 3564 f 735 in the non-castrated women (min. III/I). The mean peak area in the oophor~tomized patients (208 f 123) demonstrated a significant increase when compared with the value of 95 f 38, found in women with preserved ovaries (P < 0.05, min - III/l). Those differences between both groups of women occurred in spite of similar values for estradiol (23.2 f 3.6 for women with ovaries, and 21.2 f 1.6 for oophorectomized women). 3.3. Pulsatile LH The parameters of LH pulsatility are summarized in Table 1. There was no difference between both groups for parameters like the frequency of the pulses or the pulse interval. Some properties related with the LH pulse amplitude, like the incremental (IUII) or the mean peak height as percentage increase, were already different in both groups, and the magnitude of that difference almost reached statistical difference (P < 0.06); the
~gprod.
BittI.
55 (1994)
187-192
absolute (maximal) LH peak amplitude, however, was significantly higher in the oophorectomized group (P c 0.05). This result is consistent with the differences found for a parameter like the interpulse valley mean LH concentration (P < 0.02), which reflects increments in the baseline secretory activity of the gonadotroph for the oophorectomized group. 3.4. Basal levels of LH and responses to GnRH The mean basal morning values of LH obtained at 3day intervals stabilized within the normal range for premenopausal women after the second week (time-point corresponding to day 18) in the non-castrated women. This means that the mean level of LH ful~lled three conditions: it reached a value within the premenopausal range (4-11 W/I for the follicular phase in our hands), it did not decrease below that range subsequently, and the mean value of all the samples corresponding to the 3 subsequent time-points was not significantly higher, Increases in the concentration of estradiol were only apparent after the second week in that group, with levels within the range of an early/mid proliferative phase. Three women underwent a pre-ovulatory pattern (i.e. discharges of LH coexisting with estradiol increments) on day 27, two more on day 30, and the other woman on day 33. Positive values were obtained for progesterone in the subsequent blood samples for each case. The oophorectomized group suffered a maintained increase in the level of LH along the successive timepoints, which seemed to curve between 30 and 36 days, whereas the level of estradiol remained low.
4
Table 1 Parameters (mean f S.E.M.) of 24-h pulsatile LH secretion on day 9 post-buserelin in women whose ovaries were either preserved or ablated on the same day of interruption of the analogue
Pulse frequency Pulse amplitude (LUJl) Mean peak height as o/u increase Mean peak width (min) lnterpulse interval (mm) lnterpulse valley mean concentration (LUA)
Non-castrated
Castrated
7.1 f 1.6 4.7 * 1.6 223.1 zt 35.6
8.5 f 2.3 9.9 * 2.Y 179.4 f 21.1
62.3 f 4.9 119.5 f 17.9 2.6 f 0.7
73.5 * 3.1 108.6 f 2.2 7.1 f 1.2**
*P < 0.05 compared with corresponding women. **P c 0.02 compared with corresponding women.
values in the non-castrated values in the non-castrated
34
DAYS
Fig. 2. Response (mean f S.E.M.) of serum LH to stimulation with a O.l-mg i.v. bolus/h of synthetic GnRH administered in two consecutive doses, in women in which the ovaries were ablated (solid circles) or preserved (solid triangles). The tests were performed during buserelin treatment (day 0) and on days 9,20, and 34 after interrupting the drug. The analysis of variance for repeated measurements demonstrated that the response was different in either group (P < 0.001). The LH concentration has been represented according to a semi-logarithmic scale.
A. Cuno et al. /Eur.
J. Ubstet.
Gynecol.
The responses of LH to GnRH stimulation during buserelin administration and during the recovery period are shown in Fig. 2. As expected, the gonadotropin response to GnRH during administration of buserelin was inhibited in both groups, thus confirming the effective desensitization induced by the anaiogue. The recovery of responsiveness to GnRH was already evident for both groups by day 9 after buserelin interruption, with a response that was higher for the oophorectomized women at certain time-points, in addition to the basal. The magnitude of LH response increased again for both groups by day 20, where a difference in favor of the oophorectomized women was already clear. The increased response in the castrated patients was further pronounced on day 34 where, in contrast, the women with preserved ovaries exhibited a lower magnitude in LH secretion than that for day 20. That different response of LH to GnRH in both groups was statistically significant (P < 0.001). 4. Discussion Our experimental design provides a model for understanding some aspects of the time-course of the reactivation of the hypothalamic-pituitary unit after desensitization in the human within a context where the function of the hypothalamic level remains preserved, as well as a possibility of studying the ovarian influence on that process. Using this model we were able to observe that, in presence of a functional ovary, the reactivation of the gonadotroph proceeds slowly along 2 weeks and therefore, was slower than anticipated from the in vitro data. In the oophorectomized group, in contrast, the lack of the modulating action of the ovary was accompanied by an increased functional activity of the gonadotrophes until at least 1 month from suppression of the analogue. The presence of the ovary also determined that the morning basal levels of LH were higher for the oophorectomized group, an increase that was attained even in the absence of differences for estradiol. This suggests that perhaps other nonestrogenic factors of ovarian origin participate in regulating the circulating levels of LH, something that has also been observed in menopausal women subjected to hormonal therapy, where levels of estradiol similar to those found in an advanced follicular phase are unable to abate the concentration of LH to premenopausai ranges [ 131. We have also obtained information about specific properties of pulsatile LH release. For that, we chose day 9, a moment in which the possibility of an effect of the previous anesthesia or surgery-induced stress is unlikely [14]. Additionally, the in vitro data show that on that day the functional performance of the gonadotroph is already recovered and, according to our results, the output of estradiol by the ovary in the women who had it preserved does not determine a systemic concentration
Reprod. Biol. 55 (1994)
187-192
191
substantially different from the oophorectomized group. In this way we could gather information about the pulsatile behavior of the hypothalamic pacemaker within an endocrine scenario similar to that of desensitization in the presence of the analogue. Additionally, we could envision a possible early influence of the ovary in the functional response of the hypothalamic-pituitary unit. We analysed our data through the cluster algorithm under conditions (cluster configuration of 2 x 1 and a i statistic of 1) which offer levels of sensitivity and positive accuracy comparable with more frequent sampling protocols [12]. Our data showed that 9 days after buserelin the presence or absence of the ovary was irrelevant regarding the frequency of the pulses. This reflects that the hypothalamic pacemaker still operated in concordance for both groups at this stage and that, most likely, its pattern of secretion was not substantially modified from the period in which the analogue was being used. The similitude between both groups also involved some parameters of the LH pulse amplitude, like the incremental or the fractional LH peak amplitudes. The presence of the ovary, in contrast, was associated to a lower interpulse valley mean LH concentration, an indication in favour of a role for the ovary in modulating the baseline secretory activity of the gonadotroph. This is also supported by the higher values found for the mean peak area and the integrated serum LH concentration in the oophorectomized women, and by the significant increase of the absolute LH peak amplitude. Globally, our data suggest that during the process of recovery from desensitization, the presence of the ovary resulted in a relative decline in either the efficiency of the cellular machinery in response to the GnRH stimuli or the magnitude, rather than the frequency, of the secretion of GnRH itself or both. Our data obtained with the GnRH tests on the same day, 9, support the hypothesis of an already higher sensitivity to the decapeptyde by the gonadotrophes in oophorectomized women. This was obvious even in response to a dose of 0.1 mg GnRH, which is clearly supramaximal. The difference with the ovarian-bearers at day 9 was amplified with the GnRH tests performed subsequently, due to a maintained increase in response in the oophorectomized women until day 34, while the other group stabilized by day 20. Among the factors with potential regulatory effect on the expression of gonadotropin subunit genes, estradiol, some gonadal peptides, and GnRH itself are relevant. Estradiol acts on the gonadotroph at posttranslational level, and exerts both positive and negative effects on gonadotropin secretion and synthesis [ 15,161. The facilitatory action of estradiol on the GnRH-stimulated secretion of LH has been well established for both immunoactive and bioactive forms of the hormone [17]. That effect consists of enhancement of the pituitary response to successive GnRH stimuli, an effect mediated
A. Cano et al. /Eur. J. Obstet. Gynecol. Reprod. Biol. 55 (1994) 187-192
192
by the augmentation of the duration of the secretory burst [ 181. That progressive response could be clearly observed after the second GnRH bolus in our noncastrated group. That increase, however, was insuflicient to compensate for the high response found after the first injection in the castrated women. The role of the gonadal peptides has not been adequately explained at present although experimental data have implicated inhibin in the process [19,20]. Finally, there is abundant evidence that GnRH does increase the synthesis of LH. The rapid increase of LH in normal women after castration is well documented in the literature and, at a molecular level, it has been shown that the postcastration level of gonadotropin mRNAs declined in the rat after treatment with a GnRH antagonist [21]. Despite the similar pulse frequency observed on day 9, it is possible that a higher secretory activity of GnRH is already present in the oophorectomized women since the early moments of the recovery. In summary, our study has analysed in detail the phase of recovery from buserelin-induced desensitization. Our data show that pulsatility of LH is not modified by the ovary on day 9 of that period, but that there is an increase in parameters related with the amplitude of the LH bursts in the oophorectomized women. Although a higher amplitude of the endogenous GnRH pulses cannot be discarded, most probably that difference is due to a higher sensitivity at a pituitary level, as reflected by the GnRH stimulation tests.
6
7
8
9
10
11
I2
13
14
Acknowledgements
We are most grateful for the assistance provided by Dr J. Veldhuis, who generously supplied us with the computer program for pulse analysis. This work was supported by grant 90/0455 from the Fondo de Investigaciones Sanitarias, and grant 90-0069 from Direccion General de Investigation Cientilica y Ttcnica.
15
16 17
References Conn PM, Crowley WF. Gonadotropin-releasing hormone and its analogues. N Engl J Med 1991; 324: 93-103. Gorospe WC, Conn PM. Agents that decrease gonadotropinreleasing hormone (GnRH) receptor internalization do not inhibit GnRH-mediated gonadotrope desensitization. Endocrinology 1987; 120: 222-229. Jinnah HA, Conn PM. GnRH-stimulated LH release from rat anterior pituitary cells in culture: refractoriness and recovery. Am J Physiol (Endocrinol Metab) 1985; 249: E619-E625. Conn PM. Staley DD, Yasumoto T et al. Homologous desensitization with gonadotropin-releasing hormone (GnRH) also diminishes gonadotrope responsiveness to maitotoxin: a role for the GnRH receptor-regulated calcium ion channel in mediation of cellular desensitization. Mol Endocrinol 1987; 1: 154- 159. Uemura T, Yanagisawa T, Shirasu K et al. Mechanisms involved in the pituitary desensitization induced by gonadotropin-releasing hormone agonists. Am J Obstet Gynecol 1992; 167: 283-291.
18
19
20
21
Calogero AE, Macchi M, Montanini Vet al. Dynamics of plasma gonadotropin and sex steroid release in polycystic ovarian disease after pituitary-ovarian inhibition with an analog of gonadotropin-releasing hormone. J Clin Endocrinol Metab 1987; 64: 980-985. MacLeod AF, Wheeler MJ, Gordon P et al. Effect of long-term inhibition of gonadotrophin secretion by the gonadotrophinreleasing hormone agonist, buserelin, on sex steroid secretion and ovarian morphology in polycystic ovary syndrome. J Endocrinol 1990; 125: 317-325. Hall JE, Whitcomb RW, Rivier JE et al. Differential regulation of luteinizing hormone, follicle-stimulating hormone, and free alpha-subunit secretion from the gonadotrope by gonadotropinreleasing hormone (GnRH): evidence from the use of two GnRH antagonists. J Clin Endocrinol Metab 1990; 70: 328-335. Pavlou SN, Veldhuis JD, Lindner J et al. Persistence of concordant luteinizing hormone (LH), testosterone, and alpha-subunit pulses after LH-releasing hormone antagonist administration in normal men. J Chn Endocrinol Metab 1990; 70: 1472-1478. Salameh W, Bhasin S, Steiner BS et al. Effect of improved assay sensitivity on luteinizing hormone pulse detection after gonadotropin-releasing hormone antagonist treatment in man. J Clin Endocrinol Metab 1992; 75: 1479-1483. Veldhuis JD, Johnson ML. Cluster analysis: a simple, versatile, and robust algorithm for endocrine pulse detection. Am J Physiol 1986; 250: E486-493. Urban RJ, Johnson ML, Veldhuis JD. In vivo biologycal validation and biophysical modeling of the sensitivity and positive accuracy of endocrine peak detection. 1. The LH pulse signal. Endocrinology 1989; 124: 2541-2547. Chetkowski RJ, Meldrum DR, Steingold KA et al. Biologic effects of transdermal estradiol. N Engl J Med 1986; 314: 16151620. Shelley HJ. The use of chronically catheterized foetal lambs for the study of foetal metabolism. In: Foetal and neonatal physiology. Proceedings of the Sir Joseph Barcroft Centenary Symposium. Cambridge UK: Cambridge University Press 1973; 360381. Veldhuis JD, Evans WS, Rogol AD et al. The pituitary selfpriming actions of gonadotropin-releasing hormone: kinetics of estradiol’s potentiating effects on GnRH-facilitated LH and FSH release in healthy post-menopausal women. J Clin Invest 1986; 77: 1849-1856. Gharib SD, Wierman ME, Shupnik MA et al. Molecular biology of the pituitary gonadotropins. Endocr Rev 1990; 1 I: 177-199. Urban RJ, Veldhuis JD, Dufau ML. Estrogen regulates the gonadotropin-releasing hormone-stimulated secretion of biologically active luteinizing hormone. J Clin Endocrinol Metab 1991; 72: 660-666. Quyyumi SA, Pinkerton JV, Evans WS et al. Estradiol amplifies the amount of luteinizing hormone (LH) secreted in response to increasing doses of gonadotropin-releasing hormone by specilically augmenting the duration of evoked LH secretory events and hence their mass. J Clin Endocrinol Metab 1993; 76: 594-600. Culler MD, Negro-Vilar A. Endogenous inhibin suppresses only basal follicle-stimulating hormone secretion but suppresses all parameters of pulsatile luteinizing hormone secretion in the diestrous female rat. Endocrinology 1989; 124: 2944-2953. Muttukrishna S. Knight PG. Effects of crude and highly purified bovine inhibin (M, 32 000 form) on gonadotrophin production by ovine pituitary cells in vitro: inhibin enhances gonadotrophinreleasing hormone-induced release of LH. J Endocrinol 1990; 127: 149-159. Wierman ME, Rivier J, Wang C. Gonadotropin-releasing hormone (GnRH) dependent regulation of gonadotropin subunit mRNA levels in the rat. Endocrinology 1989: 124: 272-278.