The aging Leydig cell: 1. Testosterone and adenosine 3',5'-monophosphate responses to gonadotropin stimulation in rats

Plasma testosterone levels before and after a single injection of hCG were significantly lower in 24-month old rats than 60--90 day old animals (p less than 0.001). Even with repeated hCG administration for three weeks, plasma testosterone levels of old rats could not be restored to levels present in unstimulated young rats. In response to in vitro LH and 8-bromo-cyclic AMP stimulation, purified young Leydig cells produced significantly higher amounts of testosterone than Leydig cells from old rats. Maximal testosterone formation of the young Leydig cells in response to LH was 42.0 +/- 6.88 ng/10(6) cells, while cells from old rats produced only 16.8 +/- 3.69 ng/10(6) cells (p less than 0.01). However, the dose of LH at which one half maximal response (ED50) occurred was 0.1 mIU/ml for young Leydig cells and 0.05 mIU/ml for old Leydig cells. Basal and 1.0 mIU LH-stimulated cyclic AMP formation were comparable in both groups, but cyclic AMP formation in response to 10 mIU of LH was significantly less in the old rats (p less than 0.05). Present results demonstrate impaired steroidogenic capacity of old rats both in vivo and in vitro. Decreased testosterone response in old rats most likely is the consequence of understimulation of Leydig cells by gonadotropin; however, there appear to be additional intrinsic defects in old Leydig cells.     

Serum AFP, hCG and CEA in the management of patients with testicular, ovarian and extragonadal germ cell tumors.

Serum levels of AFP, hCG and CEA were initially and serially measured in 59 patients with testicular germ cell tumors, and serially in 37 with ovarian and 3 with extragonadal germ cell tumors. Patients with seminoma/dysgerminoma or mature teratoma had normal serum AFP and sporadically slightly elevated hCG. Some patients with embryonal carcinoma, pure or with admixture of seminoma, had serum AFP elevated to maximum 100 U/ml, yet its use for monitoring therapy was limited. Patients with yolk sac tumors had elevated AFP and sometimes CEA levels, those with choriocarcinoma had elevated hCG, and those with compound tumors had one or more of the markers highly elevated. High AFP and/or hCG levels indicated the presence of the relevant tumor cells both in the primary and in residual tumor and/or metastases, also those missed in histological material, and thus were useful in restaging. Unfortunately, their absence in serum did not exclude the presence of marker-negative subpopulations of tumor cells. Changes in marker values paralleled the effects of treatment: the level increasing from any nadir heralded recurrence in patients in remission; elevated or increasing levels during therapy implied resistance to the therapy; decreasing levels indicated regression even though a return to the normal range did not mean eradication of all tumor cells.     

Plasma leptin levels of elite endurance runners after heavy endurance training

A decrease in testosterone levels and an increase in cortisol levels are observed in male athletes with the overtraining syndrome (OTS). Cortisol causes blood leptin levels to rise and testosterone has an inverse relationship with blood leptin levels. Therefore, we hypothesized that the hormonal changes as a result of OTS induce an increase in leptin. To test this hypothesis, we examined the relationship among changes in leptin, testosterone and cortisol in thirteen male collegiate distance runners (aged 20.3+/-1.1 years) before and after an 8-day strenuous training camp. Runners ran 284.1+/-48.2 km during the training camp. Body fat percentages and plasma glucose concentrations decreased significantly after the training. Non-ester fatty acids and total cholesterol concentrations in blood were unchanged. Serum cortisol concentrations showed a significant increase after the training camp (from 11.82+/-2.00 microg/dl to 16.78+/-3.99 microg/dl), and serum testosterone decreased significantly (from 408.0+/-127.6 ng/dl to 265.2+/-97.6 ng/dl). The ratio of testosterone to cortisol (TCR) dropped by 50% after training (from 35.62+/-13.69 to 16.94+/-8.47). These results suggest that the subjects reached a state of the OTS. Contrary to our hypothesis, plasma leptin was not significantly changed (from 1.34+/-0.29 ng/ml to 1.49+/-0.18 ng/ml). Delta Plasma leptin was not significantly correlated with delta serum cortisol, delta TCR or delta fat percentage. However, delta serum testosterone was positively correlated with delta plasma leptin (r=596, p<0.05). Plasma leptin concentrations might modulate the secretion of testosterone in overtraining conditions. In conclusion, the change in blood leptin level is independent of the changes in cortisol, TCR and fat percentage in highly trained male athletes in the state of the OTS.     

Interspecies differences in the effects of HCG on testicular function among rodents

Adult mice, rats and hamsters were injected with 0 or 0.3 IU hCG/g BW, 24 h before sacrifice. Basal LH receptor concentration was highest in rats and lowest in hamsters (rats greater than mice greater than hamsters). Injection of hCG caused LH receptor down-regulation in rats and mice, and up-regulation in hamsters. Basal plasma progesterone was highest in hamsters and lowest in rats (hamsters greater than mice greater than rats), however, hCG increased plasma progesterone levels in mice and rats, but not in hamsters. Mice had much higher plasma and testicular testosterone levels than other species, but hCG did not induce a relatively more dramatic increase in any species. When testes fragments were incubated with 0 or 12.5 mIU hCG/ml for 4 h, hCG increased media progesterone levels in rats and control mice, but not in hamsters and hCG-injected mice. Also, hCG elevated media testosterone levels in control but not in hCG-injected animals. Furthermore, addition of hCG in vitro partially prevented the elevation of media testosterone induced by in vivo hCG. The present results indicate that the mechanisms for the transduction of the gonadotropic signal by the Leydig cells are species-defined.     

Circulating plasma testosterone in the XYY male

A population of 6 XYY and 6 XY individuals participated in a study to compare the levels of free circulating testosterone. No significant differences were found in the amounts of free circulating testosterone or total plasma testosterone. The plasma testosterone levels of the twelve subjects were within the normal range of values (596ng/dl ± 232ng/dl) obtained in our laboratory. Estrogen levels were assayed to determine estrogen influence on testosterone levels. Estrogens were not significantly different in the two groups. Plasma estrogens were within the limits of a random normal population (15–55pg/nl).     

Saliva testosterone time-course response to hCG in adult normal men. Comparison with plasma levels

Testosterone time-course response to 5000 IU hCG was studied simultaneously in the saliva and the plasma of 13 adult normal men. Baseline levels in saliva and plasma were: 93 +/- 9 pg/ml (mean +/- SEM) and 4.9 +/- 0.3 ng/ml respectively. After hCG the same biphasic pattern was observed in both fluids with a similar early response but the delayed peak at 72 h was relatively higher in saliva than in plasma. Thus it was suggested to collect saliva instead of plasma for the evaluation of testicular secretion of testosterone after hCG administration.     

Plasma insulin-like growth factor-I,testosterone and morphological changes in the growth of captive agile gibbons ( Hylobates agilis) from birth to adolescence

We examined growth changes in concentrations of plasma insulin-like growth factor-1 (IGF-1) and testosterone, and somatometric parameters in two captive male agile gibbons from birth to about 4 years of age, to examine the evolution of growth patterns in primates. Plasma IGF-1 concentrations in agile gibbons generally increased with age with values ranging from 200 to 1,100 ng/ml. The growth profiles in plasma IGF-1 in the gibbons were similar to those reported for chimpanzees. The highest concentrations of plasma testosterone (230 and 296 ng/dl) were observed within the first 0.3 years from birth, then the concentrations rapidly decreased and fluctuated below 100 ng/dl. Continuously higher IGF-1 concentrations were observed after 2.6 and 3.5 years of age. The profiles of plasma testosterone in these gibbons also resembled those of other primates including humans. However, their plasma testosterone levels in both neonate and adult stages (60 ng/dl) were lower than those reported for macaques and chimpanzees of respective stages. The obtained growth profiles of plasma IGF-1 and testosterone suggest that the adolescent phase starts around 2.6 or 3.5 years of age in male agile gibbons. The growth trend in many morphological parameters including body weight showed a linear increase without a significant growth spurt at approximately the onset of puberty. Head length and first digit length had reached a plateau during the study period. Brachial index, which indicates the relative length of forearm to upper arm, significantly increased gradually through the growth period. This result indicates that forearm becomes relatively longer than the upper arm with growth, which may be an evolutionary adaptation for brachiation.     

Testicular responsiveness to hCG of the ovine fetus in the last third of gestation

The in vivo and in vitro testicular responsiveness to hCG of hemicastrated lamb fetuses 95-99, 110-118 and 130-141 days of gestational age was studied. Basal plasma testosterone (T) levels were similar at all ages (less than 0.25 ng/ml), while the mean testicular concentrations of dehydroepiandrosterone sulfate (DHA-S), 17 alpha-hydroxyprogesterone (17-OHP) and T were higher in 95- to 99-day-fold fetuses. Plasma T levels and the concentration of T, DHA-S, 17-OHP, androstenedione (A) and cyclic adenosine 3'5'-monophosphate (cAMP) were increased by hCG in the hemicastrated animal at all ages. cAMP and T production by enriched preparations of dispersed interstitial cells from control testes was increased by hCG in all groups. In fetuses pretreated with hCG in vivo the addition of hCG in vitro failed to modify cAMP and T production. 100 micrograms of LHRH to a 130-day-old fetus increased plasma LH and T levels. From these experiments, it is suggested that the low plasma LH and T levels found throughout the last trimester of fetal life reflect a relative lack of endogenous LHRH synthesis and/or release, rather than reduced testicular steroidogenic capacity.     

Partial deficiency in 17-ketosteroid reductase presenting as gynecomastia

We report a 14 year-old male with severe, long-lasting gynecomastia. Baseline serum androstenedione levels were elevated compared to testosterone levels (330 ng/dl vs 28 ng/dl). In order to evaluate testosterone biosynthesis by this patient in more detail, androstenedione, testosterone, dehydroepiandrosterone (DHEA) and estradiol responses to a single dose of hCG were measured. The responses observed were different from those reported in normal males in two respects: 1) there was no immediate rise in testosterone two to four hours after the injection of hCG, and 2) levels of androstenedione and estradiol at 24, 36 and 48 hours after injection were much higher than expected. We postulate that a partial defect in testicular 17-ketosteroid reductase activity was responsible for the abnormal androstenedione to testosterone ratio in our patient. This, in turn, lead to an increased peripheral synthesis of estrogens and marked gynecomastia.     

Testosterone metabolic clearance rate and production rate in the male infant rhesus monkey

The male infant rhesus monkey (Macaca mulatta) undergoes a period of testicular activation similar to that seen in the human infant. Plasma testosterone (T) concentrations rise after birth, reaching levels of about 500 ng/dl at 1-3 mo of age and then fall to approximately 50 ng/dl at 60 mo. The plasma T metabolic clearance rates (MCRT) and production rates (PRT) were measured in two rhesus infants at 1 and 6 mo of age to determine the mechanism of the observed increase in plasma T. While there was little change in the MCRT between 1 and 6 mo, PRT was much higher at 1 mo than at 60 mo of age. These observations are consistent with the hypothesis that the increased plasma testosterone levels in infant rhesus monkeys reflect an increased production of testosterone rather than an altered metabolic disposition of the hormone.     

Daily variations of plasma sex hormone-binding globulin binding capacity, testosterone and luteinizing hormone concentrations in healthy rested adult males

In this study the daily variations of plasma sex hormone-binding globulin (SHBG) binding capacity were measured together with plasma testosterone and luteinizing hormone (LH) concentrations in 7 healthy rested adult males. Plasma SHBG-binding capacity demonstrated a significant circadian rhythm (acrophase = 2.06 p.m.; mesor = 0.35 +/- 0.6 ng testosterone bound/100 ml; amplitude = 17% of the mesor). Plasma testosterone also showed a circadian rhythm (acrophase = 7.02 a.m.; mesor = 4.38 +/- 0.67 ng/ml; amplitude = 18% of the mesor). The free testosterone index (or the ratio between plasma testosterone and SHBG-binding capacity) was not correlated with plasma LH levels. In our hands this last parameter did not vary according to a circadian pattern. These data are discussed in terms of a feedback mechanism controlling the pituitary-testis axis regulation.     

Effect of oestradiol and tamoxifen on the testosterone response in male rats to a single injection of hCG

A single s.c. injection of hCG (100 i.u.) produced a biphasic serum testosterone response in adult male rats, peaks being noted at 2 h (24 ng/ml) and 3 days (16 ng/ml). The levels fell to control during the intervening interval (8 ng/ml), although there were elevated levels of serum hCG. Maintenance of high oestradiol levels by a s.c. injection of 50 micrograms oestradiol benzoate given on Day 2 after the initial hCG injection failed to prolong the refractory period and the secondary peak of testosterone (16 ng/ml) occurred on Day 3. Administration of the antioestrogen, tamoxifen (2 mg or 3 micrograms), 24 h before or simultaneously with hCG did not prevent testicular refractoriness in vivo because serum testosterone levels still declined after 2 h to reach a nadir at 2 days. The basal in-vitro testosterone production by decapsulated testes from animals injected with hCG was enhanced at 2 h. Stimulation by hCG increased the amount of testosterone produced (X 1.5 that in controls). By 12 h basal production decreased and there was no further increment in testosterone in the presence of hCG. This refractoriness to further hCG stimulation prevailed until Day 3, but the total production of testosterone fell so that at 24 h and 2 days testes were producing basal amounts of testosterone. Testes recovered from refractoriness at 4 and 5 days, when basal and stimulated testosterone production were greater than in controls. Injection of 50 micrograms oestradiol benzoate at 2 days did not prolong the in-vitro refractory period and 2 mg or 3 micrograms tamoxifen had no effect on the in-vitro steroidogenic activity, since testes were still refractory to further hCG stimulation from 12 h to 3 days. The results of the present study do not support the hypothesis that oestradiol is involved in the hCG-induced refractoriness of the Leydig cell. The nadir between the peaks of serum testosterone in vivo corresponds to the period during which the testis is refractory to in-vitro stimulation by hCG.     

Differential effects of follicle-stimulating and luteinizing hormones on testosterone production by mouse testes

In adult mice, direct intratesticular injection of ovine follicle-stimulating hormone (o-FSH-13; AFP 2846-C, from NIAMDD, less than 1% LH contamination) at 10, 100 or 1000 ng significantly elevated concentrations of testosterone (T) within the testis. These effects were rapid, with peak values attained by 15 min, and transient, with return to values comparable to that in the contralateral, saline-injected testis within 90 min. Intratesticular injection of FSH (1 microgram) significantly increased testicular T levels in 15- and 60-day old mice. This contrasted with the effects of intratesticular administration of human chorionic gonadotropin (hCG), which stimulated T production significantly at 30 days of age through adulthood. In adult mice, the equivalent LH to the possible contamination in the FSH preparation (1 ng) had no effect. Intratesticular injection of 10 ng LH produced comparable stimulation to that by 100 ng FSH (approximately 7-fold). Systemic pre-treatment with a charcoal-treated porcine follicular fluid (PFF) extract for 2 days reduced plasma FSH levels [86 +/- 17 (5) vs 700 +/- 8 (6); P less than 0.05], but had no effect on plasma LH. Twenty-four hours after the last treatment, the response to intratesticular injection of hCG (2.5 mIU), FSH (100 ng) or LH (10 ng) was also significantly attenuated in these mice. Intratesticular injection of PFF had no direct effect on testicular T levels. In vitro T production in the presence of hCG, LH or FSH were differentially affected by the concentrations of calcium (Ca2+) or magnesium (Mg2+) in the incubation media. The stimulatory effects of FSH were apparent at significantly lower levels of Ca2+ or Mg2+, than were those of LH or hCG. The results of these studies indicate that FSH is capable of stimulating testicular T production. Furthermore, the responsiveness to FSH is qualitatively different than that to LH/hCG in terms of the age pattern, as well as the dependence on Ca2+ or Mg2+. In addition, plasma FSH levels appear to influence testicular responsiveness to direct exogenous administration of gonadotropins. These studies indicate that FSH stimulation of T production can be differentiated from those of LH, and that these effects of FSH can be observed under physiological conditions.     

Interspecies differences in testicular LH receptors and in vitro testosterone production among rodents

Testes from rats, mice and hamsters were incubated for 4 h with 0, 3.125 or 12.5 mIU hCG/ml. The LH receptor concentration in incubated testes of rats and mice was higher than that observed in hamsters. Testosterone levels in incubation media were significantly different among species (mice greater than rats greater than hamsters). During the incubation, hCG caused an increase in testosterone levels in all three species, but produced no significant changes in LH receptor concentration. Furthermore, a correlation between LH receptor concentration and testosterone only in hamsters is observed. The efficiency of the LH receptor-steroidogenesis interaction was estimated from the ratio of testosterone levels to receptor concentration under basal conditions and was found to differ among species (mice greater than hamster greater than rats). The levels of PGE and PGF in incubation media were higher in mice than in rats or hamsters, and hCG did not alter prostaglandin levels in any of the species. The present results indicate that acute in vitro hCG stimulation of testosterone synthesis does not involve appreciable changes in testicular LH receptor levels.     

Diurnal variations of androgens in sexually mature male Bolivian squirrel monkeys (Saimiri sciureus) during the breeding season

To assess diurnal fluctuations of serum androgens and cortisol in adult male Bolivian squirrel monkeys, these steroids were measured at predetermined times (0300, 0900, and 2300 hours) during two separate 24-hour periods in the breeding season (January 1983 and late November 1983). A significant diurnal change in serum cortisol was noted, with a nadir of 99.9 ± 11.9 μg/dl (x? ± SEM) at 2300 hours and a peak of 168.9 ± 7.8 μg/dl at 0900 hours. Conversely, a nadir in serum testosterone was noted at 0900 hours (117 ± 26.5 ng/ml) increasing to a peak of 328.5 ± 57.9 ng/ml at 0300 hours. Serum androstenedione and dehydroepiandrosterone followed a pattern similar to testosterone, with a serum androstenedione (176.4 ± 34.9 ng/ml) and dehydroepiandrosterone (11.7 + 1.8 ng/ml) nadir at 0900 hours and a plasma androstenedione (494.5 ± 55.4 ng/ml) and dehydroepiandrosterone (32.5 ± 4.1 ng/ml) peak at 0300 hours. Parallel changes of testosterone, androstenedione, and dehydroepiandrosterone suggest a significant contribution of all three androgens from a common site, the testes. In contrast to old world primates and humans, serum androstenedione levels exceeded serum testosterone levels in this species.     

Male pseudohermaphroditism due to testicular 17-ketosteroid reductase deficiency.

A new case of testicular 17 ketosteroid reductase (17 KSR) deficiency without gynecomastia was investigated. Delta4 androstenedione (15.6 ng/ml) was ten times the normal range, unchanged after dexamethasone administration. In contrast, plasma testosterone (4.1 ng/ml) was in the low normal male range and plasma dehydroepiandrosterone (4.2 ng/ml) was normal. Plasma luteinizing hormone and follicle-stimulating hormone were increased (162 and 470 ng/ml LER 907 respectively). After adrenal suppression and human chorionic gonadotropin stimulation, the increase of delta4 androstenedione was in contrast with the inertia of testosterone. In spermatic venous plasma delta4 androstenedione level (293.2 ng/ml) was very high and testosterone level (7.1 ng/ml) a hundred times below the normal mean. Plasma estrone (124 pg/ml) was increased and estradiol (22 pg/ml) was normal. In spermatic venous plasma estrone was elevated and estradiol very low (1380 and 32 pg/ml respectively). It is the third case of 17 KSR deficiency where the lack of E2 increase explains the absence of gynecomastia.     

Effects of hyper- and hypoprolactinemia on gonadotropin secretion, rat testicular luteinizing hormone/human chorionic gonadotropin receptors and testosterone production by isolated Leydig cells

The effect of prolactin (Prl) on gonadotropin secretion, testicular luteinizing hormone (LH)/human chorionic gonadotropin (hCG) receptors, and testosterone (T) production by isolated Leydig cells has been studied in 60-day-old rats treated for 4 days, 4 and 8 weeks with sulpiride (SLP), a dopaminergic antagonist, or for 4 days and 4 weeks with bromocriptine (CB), a dopaminergic agonist. Plasma Prl concentrations were significantly greater in the SLP groups (204 +/- 6 ng/ml) and lower in the CB groups (3.0 +/- 0.2 ng/ml) than those measured in the control groups (54 +/- 6 ng/ml). The plasma concentrations of gonadotropin were not affected by a 4-day treatment with SLP or CB, nor were they after a 4-week treatment with CB. However, the hyperprolactinemia induced by an 8-week treatment with SLP was associated with a reduced secretion of gonadotropin (LH, 16 +/- 4 vs. 35 +/- 6 ng/ml; FSH, 166 +/- 12 vs. 307 +/- 14 ng/ml). In SLP-induced hyperprolactinemia, a 30% increase in the density of the LH/hCG testicular binding sites was observed (178 +/- 12 fmol/mg protein), whereas a 60% decrease was measured in hypoprolactinemia (55 +/- 5 vs. control 133 +/- 5 fmol/mg protein). Plasma T levels were increased in 4-day and 4-week hyperprolactinemic animals (4.3 +/- 0.4 and 3.9 +/- 0.4 ng/ml, respectively), but returned to normal levels in the 8-week group (3.0 +/- 0.5 vs. C: 2.3 +/- 0.2 ng/ml). No T modifications were observed in hypoprolactinemic animals. Two distinct populations of Leydig cells (I and II) were obtained by centrifugation of dispersed testicular cells on a 0-45% continuous Metrizamide gradient. Both possess LH/hCG binding sites. However, the T production from Leydig cells of population II increased in the presence of hCG, whereas that of cell population I which also contain immature germinal cells did not respond. The basal and stimulated T secretions from cell populations I and II obtained from CB-treated animals were similar to controls, whereas from 4 days to 8 weeks of hyperprolactinemia, basal and hCG induced T productions from cell population II decreased progressively. These data show that hyperprolactinemia causes, in a time-dependent manner, a trophic effect on the density of LH/hCG testicular receptors; reduces basal and hCG-stimulated T production from isolated Leydig cells type II; and results in an elevated plasma T concentration which decreases with time. The latter suggests a slower T catabolism and/or an impaired peripheral conversion of T into 5 alpha-dihydrotestosterone (DHT). Although hypoprolactinemia is associated with a marked reduction in testicular LH receptors, it does not affect T production.     

Effect of fetal sex on maternal and fetal human chorionic gonadotropin levels and comparison of their levels in paired umbilical arteries and veins

There were discordant results regarding the effect of fetal sex on human chorionic gonadotropin (hCG) concentrations in maternal or fetal circulation and regarding whether the levels in umbilical arteries are equal to those in umbilical veins. Totally, 188 singleton pregnancies at 36 to 42 weeks of gestation without any obstetrical or medical complication were studied. The hCG levels were measured by radioimmunoassay specific for hCG using Sb3 antibody raised against beta-subunit of hCG. The maternal ages and parities between those who gave birth to a male or a female baby were not different statistically. The birth weights between male and female babies were also not different. The serum hCG levels had a wide range in maternal circulation (200-75,200 mIU/ml) and their distribution was positively skewed. The mean value (geometric mean, G.M.) in maternal circulation for those who carried a female fetus (11,500 mIU/ml) was significantly higher than that for those carrying a male fetus (6,470 mIU/ml) (P less than 0.001, Student's t-test). The hCG concentrations in umbilical veins of female fetuses were also higher than in those of male fetuses (G.M., 26.8 vs 19.5 mIU/ml, P less than 0.01, Student's t-test). Umbilical arterial hCG levels (G.M., 10.05 mIU/ml) were statistically not different from umbilical venous levels (G.M., 10.92 mIU/ml) (paired t-test).(ABSTRACT TRUNCATED AT 250 WORDS)     

Testicular responsiveness following chronic administration of hCG (1500 IU every six days) in untreated hypogonadotropic hypogonadism

The observation that the testosterone (T) response to a single intramuscular injection of hCG is prolonged suggests that currently used regimens (2-3 injections per week) to stimulate endogenous androgen secretion in hypogonadotropic hypogonadism (HH) patients have to be reassessed. Moreover, during the last few years, Leydig cell steroidogenic desensitization has been found after massive doses of hCG. The aim of the present investigation, carried out in 6 HH patients who showed no signs of puberty, was to study the effect of 1500 IU hCG administered every six days over a period of one year to induce the onset of pubertal development. To evaluate the kinetics of the response of T, 17 alpha-hydroxyprogesterone (17 alpha-OHP) and 17 beta-oestradiol (E2), blood samples were taken basally and 1, 2, 4 and 6 days after drug injection. This dynamic study was performed after the first injection and after the 4th and 12th month of treatment. During this one year time period, a progressive increase in testicular size was observed. Comparing plasma T levels (mean +/- SE) before the first injection (11.2 +/- 4.7 ng/dl) with the corresponding values at the 4th (38.7 +/- 10.5 ng/dl) and 12th months (99.5 +/- 19.9 ng/dl) of therapy, a progressive and significant increase was observed. T reached a maximum elevation 58 hours after hCG injection at the 4th month (198.3 +/- 42 ng/dl; P less than 0.01) and at the 12th month (415.6 +/- 62.6 ng/dl; P less than 0.05), whereas it remained unchanged following the first hCG injection.(ABSTRACT TRUNCATED AT 250 WORDS)