It is well established that valproate increases hypothalamic concentrations of gamma-aminobutyric acid (GABA). Although little research has been done on the role of GABA in the control of pulsatile luteinizing hormone (LH) secretion in humans, our group recently found that administration of valproate had no significant effect on pulsatile LH secretion in late follicular and mid-late luteal phase normal women. However, the results of several studies of rats suggest that GABAergic regulation of LH secretion may depend on steroid levels. The objective of this work was to determine whether regular administration of sodium valproate inhibits pulsatile LH secretion in ovariectomized women. Twelve women who had undergone ovariectomy for causes other than malignant tumors were each studied in two 8 h sessions, in each of which blood samples were taken every 5 min. The first session was the control; for the second. 400 mg of sodium valproate was administered every 8 h during the seven preceding days and at 08.00 h and 14.00 h on the day of the study session. Serum valproate was determined by repolarization fluorescence spectrophotometry, and LH, estradiol and progesterone by radioimmunoassay. The serum LH series were subjected to a deconvolution procedure to reconstruct the pattern of pituitary LH secretion. Luteinizing hormone pulses were identified by the authors' non-parametric method. Control and post-valproate results were compared with regard to number of pulses, pulse duration, the quantity of LH secreted in each pulse, interpulse interval and mean serum LH level. There was no statistically significant difference between control and post-valproate results for any of the variables considered. It is concluded that sustained serum valproate levels do not alter pulsatile secretion of LH in ovariectomized women. This implies that, in humans, GABA is probably not a decisive factor in the regulation of the GnRH pulse generator.

Patients with diabetes insipidus (DI) are at high risk for dehydration, hypernatremia and hemodynamic instability. Ovarian hyperstimulation syndrome (OHSS) likewise has negative effects on electrolyte balance. If ovulation induction is required in a patient with DI, there is thus a strong argument for avoiding techniques that increase the risk of OHSS. This paper reports the results of ovulation induction in a patient with anovulation of hypothalamic origin and central (vasopressin deficit) DI. Ovulation was induced with pulsatile gonadotropin-releasing hormone (GnRH), which induces single-ovule cycles and which does not increase the risk of OHSS. The patient successfully achieved pregnancy after seven ovulatory cycles.

Differentiation between hypothalamic and pituitary amenorrhea is generally based on the luteinizing hormone-releasing hormone (LHRH) test (whether as a single dose, two consecutive doses, or pulsatile over 5-10 days), together with high-resolution imaging (computed tomography or magnetic resonance) of the sellar region. Long-term administration of gonadotropin-releasing hormone (GnRH) is generally used only for ovulation induction, and not for diagnostic purposes. Here, we report the results of long-term administration of GnRH to 19 women initially diagnosed as suffering from hypothalamic amenorrhea on the basis of LHRH testing and computed tomography imaging. During treatment, subjects received 20-micrograms pulses of GnRH every 90 min, subcutaneously from a portable infusion pump. Fourteen subjects responded (i.e. ovulated) during the first treatment cycle; one subject menstruated but did not ovulate during the first cycle, and the dropped out of the study; the remaining four subjects did not ovulate or menstruate despite at least three treatment cycles. Magnetic resonance imaging of the sellar region of these four subjects revealed pituitary lesions (partially empty sella in three cases, microadenoma in one case) which had not been detected by computed tomography. By contrast, no such abnormalities were detected in the nine responders who agreed to undergo magnetic resonance imaging. These findings suggest that long-term administration of GnRH is of value not only for ovulation induction but also for diagnostic purposes. Specifically, an initial diagnosis of hypothalamic amenorrhea is confirmed if there is a positive ovulation response after two GnRH treatment cycles; otherwise, pituitary amenorrhea should be suspected. Our results also suggest that magnetic resonance imaging is more effective than computed tomography for the detection of partially empty sella.