2003.-Studies using pharmacological gonadotropin stimulation suggest that ovarian steroidogenesis is abnormal in the polycystic ovary syndrome (PCOS). We assessed ovarian steroid secretion in response to near-physiological gonadotropin stimuli in 12 ovulatory controls and 7 women with PCOS. A gonadotropin-releasing hormone-receptor antagonist (ganirelix, 2 mg sc) was given to block endogenous LH secretion, followed by dexamethasone (0.75 mg orally) to suppress adrenal androgen secretion. After ganirelix injection (12 h), intravenous infusions of recombinant human LH (0, 10, 30, 100, and 300 IU; each over 8 min) were administered at 4-h intervals in a pseudorandomized (highest dose last) manner. Plasma LH, 17-hydroxyprogesterone (17-OHP), androstenedione, and testosterone were measured concurrently. LH dose-steroid response relationships (mean sex-steroid concentration vs. mean LH concentration over 4 h postinfusion) were examined for each subject. Linear regression of 17-OHP on LH yielded a higher (mean Ϯ SE) slope in PCOS (0.028 Ϯ 0.010 vs. 0.005 Ϯ 0.005, P Ͻ 0.05), whereas extrapolated 17-OHP at zero LH was similar. The slopes of other regressions did not differ from zero in either PCOS or controls. We conclude that near-physiological LH stimulation drives heightened 17-OHP secretion in patients with PCOS, suggesting abnormalities of early steps of ovarian steroidogenesis. With the exception of 17-OHP response in PCOS, no acute LH dose-ovarian steroid responses were observed in controls or PCOS. Defining the precise mechanistic basis of heightened precursor responsiveness to LH in PCOS will require further clinical investigation. ovarian steroidogenesis; hyperandrogenism; androstenedione; testosterone; 17-hydroxyprogesterone; luteinizing hormone THE ETIOLOGY OF OVARIAN HYPERANDROGENEMIA in the polycystic ovary syndrome (PCOS) remains enigmatic. Previous studies have emphasized the relative roles of neuroendocrine abnormalities leading to persistent and excessive LH secretion (11,14,20,30,34) and the ovarian actions of hyperinsulinemia, a consequence of insulin resistance (4,27,35,40). Additional evidence suggests that anomalous ovarian steroidogenesis is a primary abnormality in PCOS. For example, theca cell cultures derived from women with PCOS secrete androgens excessively (13), even after propagation for three to four passages in LH-free media (22).In vivo studies of PCOS demonstrate characteristically abnormal ovarian steroid responses to acute administration of either a potent gonadotropin-releasing hormone (GnRH) agonist (2, 7, 15, 31, 38) or a high dose of human chorionic gonadotropin (hCG; see Ref. 12,15,19). Specifically, these stimuli elicit exaggerated secretion of 17-hydroxyprogesterone (17-OHP) and, to a lesser degree, androstenedione (⌬4A), suggesting abnormal ovarian steroidogenesis. However, these paradigms involve pharmacological ovarian stimulation and do not reproduce physiological LH pulsatility. For instance, with acute GnRH agonist administration, plasma LH approaches or exceeds 100 IU/l ...

Women with polycystic ovary syndrome (PCOS) have reduced GnRH sensitivity to suppression by ovarian steroids, which can be ameliorated by androgen blockade. We studied nine PCOS women and nine controls to determine whether metformin could change feedback inhibition by estradiol (E(2)) and progesterone (P). LH was measured every 10 min, and FSH, E(2), P, and testosterone (T) were measured every 2 h. Frequently sampled iv glucose tolerance test was performed at the end of each admission. After the first admission, metformin (500 mg, three times a day) was started. The second admission occurred on d 8-11 of the next menstrual cycle in controls and on d 28 in PCOS patients. Patients subsequently took E(2) and P for 1 wk until the third admission. At baseline, PCOS women had higher T, free T, androstenedione, and estrone. After 4 wk of metformin, controls had a slight reduction in total T, but free T was unchanged. However, PCOS patients had reduced insulin, T, and E(2), and increased LH mean/amplitude and FSH. After ovarian steroids, controls had a greater reduction in LH pulse frequency than PCOS (61 vs. 25%). These results suggest that the beneficial effects of metformin on ovulatory function in obese PCOS women are probably not mediated by enhanced hypothalamic sensitivity.