Evidence of luteinizing hormone secretion in hypothalamic amenorrhea associated with weight loss*

Transcription

1 FRTILITY AND STRILITY Copyright" 1990 The American Fertility Society Vol. 54, No.2, August 1990 Printed on ocid free poper in U.S.A. vidence of luteinizing hormone secretion in hypothalamic amenorrhea associated with weight loss* Alessandro D. Genazzani, M.D.t+ Felice Petraglia, M.D. t Giovanna Fabbri, M.D.t Anna Monzani, M.D. t Vanna Montanini, M.D. Andrea R. Genazzani, M.D.t University of Modena, Modena, Italy The lack of plasma luteinizing hormone (LH) pulsatile pattern or episodic LH secretory bursts during night have been demonstrated in hypothalamic amenorrhea. The availability of both sensitive and specific immunofluorimetric assay and algorithm for pulse detection enabled us to reanalyze the question of whether or not patients with hypothalamic amenorrhea secrete LH in a pulsatile fashion. Seven women with secondary amenorrhea associated with weight loss and four normally cycling women were studied, sampling every 5 minutes for 8 hours. Control subjects were studied during four different phases of the menstrual cycle. In all amenorrheic patients, a frequent LH pulsatile secretion, with pulses of low amplitude, was found (10.7 ± 1,4 peaks/8 h; mean ± SM). The pulse frequency was significantly higher (P < 0.05) than any phases of the control group (early follicular: 7 ± 0,4 peaks/8 h; late follicular: 6.8 ± 0.6 peaks/8 h; early luteal: 4.3 ± 0,4 peaks/8 h; late luteal: 7 ± 0.3 peaks/8 h). The LH pulsatile release in amenorrheic patients showed a mean pulse duration and amplitude shorter than in any phase of the menstrual cycle of the controls. In conclusion, in weight-loss-related-amenorrhea, the major change was not the absence of the LH pulsatile release but its increased frequency with reduced pulse amplitude. Fertil Steril 54:222, 1990 Hypothalamic amenorrhea is a complex syndrome under which several forms of secondary amenorrhea are described. The weight loss when associated with unbalanced dieting is responsible for the interruption of menstrual cycle. 1 An increased incidence of amenorrhea is also frequently described after discontinuing use of oral contraceptives in women who weigh less than the ideal body weight or who have lost weight while taking oral contraceptives. 2 Moreover, even in absence of weight loss, several long distance runners 3-5 pre- Received October 11, 1989; revised and accepted April 20, * Supported in part by the Ministero Pubblica Istruzione and Consiglio Nazionale delle Richerche, Rome, Italy. t Department of Obstetrics and Gynecology. :j: Reprint requests: Alessandro D. Genazzani, M.D., Depart ment of Obstetrics and Gynecology, University of Modena, Via del Pozzo 71, Modena, Italy. Department of ndocrinology. sent amenorrhea. The common feature of women with weight-loss amenorrhea is the hypogonadotropic or normogonadotropic circulating levels. A classical finding in male and female patients with hypogonadotropic hypogonadism is that they fail to show any detectable pulsatility of luteinizing hormone (LH) secretion. A small number showed pulsatile LH secretion during sleep. 6 Because both very sensitive and specific assays for measuring plasma LH7 and pulse detection algorithm have been developed 8 we reexamined if there is any pulsatility of LH in women with secondary amenorrhea associated with weight loss who were studied while not sleeping. Patients MATRIALS AND MTHODS Seven women with weight loss were studied after informed consent. They aged between 18 and Genazzani et al. LH pulsatility and amenorrhea

2 years. Criteria for the inclusion of the patients in our study were: (1) disappearance of menses;;:: 6 months before the study and no evidence of pregnancy; (2) low levels of LH, follicle-stimulating hormone, estradiol, and progesterone assayed in at least three different samples collected over a range of 45 days before the study; (3) plasma levels of cortisol, androstenedione, testosterone, dehydroepiandrosterone-sulfate, tiroxine, thyroid -stimulating hormone, prolactin, and growth hormone within the normal range; (4) normal computerized tomography of the sella turcica; (5) reduced LH response to a standard gonadotropin-releasing hormone (GnRH) test; (6) no psychiatric disease; and (7) presence of weight loss occurring in a range of 2 to 4 months before the disappearance of menses. The weight of the patients was less than ideal body weight. These reductions ranged between 17% and 23%, with a mean of21%. Four normally cycling women were used as control group after they gave informed consent. No control subjects showed any weight loss in the last 6 months before the study, and their weight was 56 ± 3 kg (mean ± SM), 4.5% over the estimated ideal body weight. They were studied in four different phases of their menstrual cycles: assuming 0 as the day of ovulation, they were studied during the early follicular phase (between days -13 and -9), late follicular phase (between days -5 and -1), early luteal phase (between days +1 and +5), and late luteal phase (between days + 10 and + 14). All the subjects of the study were admitted to our department the day before the study. All subjects were sampled every 5 minutes for 8 hours. An intravenous heparin lock was inserted in an antecubital vein at 7 A.M., with the start of the sampling collection at 8 A.M. The plasma was immediately separated and frozen after sampling for subsequent determination of LH concentrations. The study protocol was approved by the Human Committee of the University of Modena. Assays Serum LH concentrations were determined using an immunofluorimetric assay (Pharmacia, Uppsala, Sweden). The assay was a time-resolved measurement that is a sandwich fluorescence technique with 2 monoclonal antibodies raised against different epitopes of a- and (j-subunits of LH. 7 All samples from the same subject were analyzed in duplicate in the same assay. To obtain a precise evalu- ation of the measurement error, we also assayed ;;::30 replicates from a serum pool from the same individual (obtained by combining 50 ILL aliquots from each of the samples collected). The plasma volume used for the assay was 50 ILL, and sensitivity of the assay expressed as minimal detectable dose was 0.11 miujml. Intra-assay and interassay coefficients of variation (CV s) based on 2 different quality control samples were 5.2% and 8%, respectively, whereas intra-assay and interassay (CVs) at 10%,50%, and 90% of total count over bound were 12.2%,5.1%, 14% and 16%, 8.4%, 18.4%, respectively. Pulse Detection The presence of LH pulses was determined using the program DTCT,s which is a software program for IBM PC-compatible computers, developed at the Laboratory of Theoretical and Physical Biology, National Institutes of Health, Bethesda, Maryland. DTCT looks for significant pulses with two types of logic: (1) analysis of first derivatives of data, for detection of rapid events; and (2) fitting of linear segments, for detection of slow events. Slowly rising or falling sections of data, with slope significantly different from 0, were detected. Data from each subject were analyzed using DTCT with two different values of P (P = and P = 0.01) for false positive peak detection. To examine the rate of false positive peaks and to assess what was the lowest change in plasma concentration detected as "false" peaks, we used DTCT to analyze the data of the plasma pools of each subject assayed together with the time series. The observed false positive rate was in good agreement with the expected value of 1 % for criteria at the P = 0.01 level. Time series were studied assuming as measurement error the percent of variability (percent of CV) computed on the serum pool of each subject and entered as a variance model for DTCT program analysis. 9 The percent of CVs varied between 2.8% and 7.4%, with a mean of 5.8%. An independent algorithm for pulse detection, CLUSTR method,lo was used to compare and validate results obtained with DTCT. RSULTS The characteristics of the LH pulses in healthy and amenorrheic women are shown in Table 1. Figure 1 shows the secretory patterns of three amenorrheic patients. All patients with weight-loss-re- VoL 54, No.2, August 1990 Genazzani et al. LH pulsatility and amenorrhea 223

3 Table 1 Characteristics of the LH Secretory Pattern in Controls and in Amenorrheic Subjects No. of peaks/ 8 h Duration Amplitude min miu/ml Controls (n = 4) 7.0 ±OA 50.0± ±0.8 arly follicular phase Late follicular phase 6.8 ± ± ± 0.8 b arly luteal phase 4.3 ± 004" ± ±0.9 b Late luteal phase 7.0± ± ± 0.5 Amenorrhea (n = 7) 10.7 ± lad 28.9± 3.5" 0.7 ±0.26 Values are means ± SM. b P < 0.05 versus amenorrheic patients. " P < 0.05 versus early follicular phase, late follicular phase, and late luteal phase. d P < 0.05 versus controls (all phases). " P < 0.01 versus controls (all phases). lated amenorrhea were characterized by a frequent LH pulsatile release. No significant difference was found when analyzing the time series with two different P levels (0.005 and 0.01) for false positive peak detection. When the time series were analyzed at P = 0.01 level, the frequency of LH pulsatile release was higher (P < 0.05) than any phase of the control group (10.7 ± 1.4 peaks/8 h; duration 28.9 ± 3.5 minutes; interpulse interval 49.2 ± 5.9 minutes; mean ± SM) (Table 1). Similar results were observed at P = All the patients showed LH pulses of low amplitude with a quick onset and relative slow downstroke. These results were common to all patients with no relation to mean plasma LH levels, the time of amenorrhea, and the entity of the weight loss. The LH secretory pattern of the controls changed from a frequency of 7 ± 0.4 peaks/8 h (mean ± SM) and a duration of 50 ± 6.6 minutes during the early follicular phase to a similar frequency (6.8 ± 0.6 peaks/8 h) and duration (55.5 ± 10 minutes) during the late follicular phase. In the early luteal phase, plasma LH showed a significantly lower (P < 0.05) pulse frequency (4.3 ± 0.4 peaks/8 h) with a higher duration (104 ± 23 minutes). In the late luteal phase, LH pulse frequency increased (7 ± 0.3 peaks/8 h), and pulse duration decreased (50.3 ± 6 minutes) in comparison with the early luteal phase. Pulse amplitude of amenorrheic patients was reduced significantly from controls when compared with the late follicular phase and early luteal phase (P < 0.05) (Table 1). When time series were analyzed with CLUSTR method, LH pulse frequency was equal to the estimation of DTCT. In particular, the 224 Genazzani et al. LH pulsatility and amenorrhea amenorrheic patients showed a frequency of 11.2 ± 0.5 peaks/8 h (mean ± SM). DISCUSSION This study showed the presence of a frequent pulsatile release of LH in patients with weightloss-related amenorrhea. Previous studies demonstrated the absence of any pulsatile release - :::> " -:::> -:::> I.' PATINT # 1 :.:, ', , o no loo lig ' ao u-,, PATINT # 2 u , o so eo to ' UO 270 sao no 310 leo H PATINT # 3 3.', ,,, TIM Figure 1 Luteinizing hormone secretory pattern in three of seven patients studied. Arrows indicate significant pulses found by DTCT program (P < 0.01).

4 from the gonadotropes during the day, whereas some groups of subjects showed a night-related pulsatile secretory pattern. In all these previous reports, plasma LH levels were measured by radioimmunoassay, and the mean LH levels in these patients are usually close to the lower detection limits of these systems. Recently, new assay methodologies have been assessed improving sensitivity and specificity of hormonal measurements. The availability of highly sensitive immunofluorimetric assay,7 the high sampling frequency (5 minutes), and the use of a sensitive peak detection program8.17 allowed us to demonstrate an LH pulsatile pattern in all patients studied. They showed a pulsatile secretion with a frequency significantly higher than the control group in any of the four phases of the menstrual cycle investigated. Recently, it has been reported that in hypothalamic amenorrhea, central regulation of pituitary hormone secretion is modified; therefore "a costellation of neuroendocrine aberrations" in this syndrome may be suggested.18 vidences exist that patients with weight loss show abnormal thermoregulation,19 reduced and delayed hormonal response to GnRH and thyrotrop(h)in-releasing hormone TRH standard tests,20 and lack of diurnal variation of cortisol.16 The abnormality of these findings is directly correlated with the severity of weight loss.20 A more profound hypothalamic dysfunction seems to occur in women with anorexia nervosa.20 Anorexic patients are characterized by LH pulsatile release reduced in frequency or completely absent, probably because of a GnRH deficiency. This represents the reversion to a prepubertal pattern of hypothalamic activity. 5 The present report shows that in hypothalamic amenorrhea associated with weight loss, the major abnormality in the hypothalamic-hypopheseal-gonadal axis is not the lack of the pulsatile secretion but the modification of its secretory pattern, which is characterized by a frequent release, with pulses of low amplitude and short duration. At present time, it is not yet possible to distinguish whether the small and frequent LH peaks are dependent on small residual bursts of GnRH or on an intrinsic LH pulsatile release from the gonadotropes. This last hypothesis might be supported by recent reports21.22 that demonstrated the presence of an intrinsic pulsatile release of LH from human pituitary in an in vitro perfused system. Acknowledgment. We thank David Rodbard, M.D., Laboratory of Theoretical and Physical Biology, National Institutes of Vol. 54, No.2, August 1990 Child Health and Development, National Institutes of Health, Bethesda, Maryland, for his helpful suggestions and comments as well as for providing the program DTCT. RFRNCS 1. Knuth UA, Hull MGR, Jacobs HS: Amenorrhea and loss of weight. Br J Obstet Gynaecol 84:801, Fries H, Nillius SJ: Dieting, anorexia nervosa and amenorrhea after oral contraceptive treatment. Acta Psychiatr Scand 49:669, Dale, Gerlach DH, Wilhite AL: Menstrual dysfunction in distance runners. Obstet Gyneco154:47, Veldhuis JD, vans WS, Demers LM, Thorner MO, Wakat D, Rogol AD: Altered neuroendocrine regulation of gonadotropin secretion in women distance runners. J Clin ndocrinol Metab 61:557, Clayton RN: Gonadotrophin-releasing hormone: from physiology to pharmacology. Clin ndocrinol (OxO 26:361, Crowley WF, Filicori M, Spratt DI, Santoro NF: The physiology of gonadotropin-releasing hormone (GnRH) secretion in men and women. Recent Prog Horm Res 41:473, Lovgren T, Hemmila I, Petterson K, Halonen P: Time-resolved fluorometry in immunoassay. In Alternative Immunoassays, dited by WP Collins. Boston, John Wiley and Sons, 1985, p Oerter K, Guardabasso V, Rodbard D: Detection and characterization of peaks and estimation of instantaneous secretory rate for episodic pulsatile hormone secretion. Comput Biomed Res 19:170, Guardabasso V, Oerter K, Iademarco MF, Veldhuis JD, Rodbard D: DTCT's user guide. dited by Laboratory of Theoretical and Physical Biology, National Institutes of Health, Bethesda, Maryland, Veldhuis JD, Johnson ML: Cluster analysis: a simple, versatile, and robust algorithm for endocrine pulse detection. Am J Physiol 250:486, Genazzani AD, Rodbard D, Genazzani AR: Unpublished data 12. Santoro N, Filicori M, Crowley WF: Hypogonadotropic disorders in men and women: diagnosis and therapy with pulsatile gonadotropin-releasing hormone. ndocr Rev 7:11, Khoury SA, Reame N, Kelch RP, Marschall JC: Diurnal patterns of pulsatile luteinizing hormone secretion in hypothalamic amenorrhea: reproducibility and responses to opiate blockade and an a2-adrenergic agonist. J Clin ndocrinol Metab 64:755, Urban RJ, vans WS, Rogol AD, Kaiser DL, Johnson ML, Veldhuis JD: Contemporary aspects of discrete peak-detection algorithms. I. The paradigm of the luteinizing hormone pulse signal in men. ndocr Rev 9:3, Petterson K, Siitari H, Hemmila I, Soini, Lovgren T, Hanninen V, Tanner P, Stenman UH: Time-resolved fluoroimmunoassay of human choriogonadotropin. Clin Chem 29:60, Wennik JMB, Delemarre VanDe Waal HA, Van Kassel H, Mulder GH, Foster JP, Schoemaker J: Luteinizing hormone secretion patterns in boys at the onset of puberty measured using a highly sensitive immunoradiometric assay. J Clin ndocrinol Metab 67:924, 1988 Genazzani et a1. LH pulsatility and amenorrhea 225

5 17. Genazzani AD, Guardabasso V, Rodbard D: valuation of methods for detection of pulsatile hormone secretion: sensitivity vs. specificity. (Abstr.) Presented at the Seventy First ndocrine Society, Seattle, Washington, June 21 to 24, Published by the ndocrine Society, Bethesda, Maryland, 1989, p Berga SL, Mortola SF, Girton L, Suh B, Laughlin G, Pham P, Yen SSC: Neuroendocrine aberrations in women with functional hypothalamic amenorrhea. J Clin ndocrinol Metab 68:301, Vigersky RA, Andersen A, Thompson RH, Lauriaux DL: Hypothalamic dysfunction in secondary amenorrhea asso- ciated with simple weight loss. New ngl J Med 297:1141, Vigersky RA, Lauriaux DL, Andersen A: Anorexia nervosa: behavioral and hypothalamic aspects. Clin ndocrinol (Dxf) 5:517, Rasmussen DD: New concepts in the regulation of the hypothalamic gonadotropin releasing hormone (GnRH) secretion. J ndocrinol Invest 9:427, Gambacciani M, Liu JH, Swartz WH, Tueros VS, Yen SSC, Rasmussen DD: Intrinsic pulsatility ofluteinizing hormone release from pituitary in vitro. Neuroendocrinology 45:402, Genazzani et al. LH pulsatility and amenorrhea