| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Prince Henry’s Institute of Medical Research (R.I.M., L.O., P.G.S., G.B., D.M.R.), Monash Medical Centre, Clayton, Victoria, Australia 3168; Institute of Reproduction and Development (D.M.d.K.), Monash University, Clayton, Australia 3168; and Department of Urology (M.F.), Monash University, Clayton, Australia 3168
Address all correspondence and requests for reprints to: Dr. Robert I. McLachlan, Prince Henry’s Institute of Medical Research, P.O. Box 5152, Clayton, Victoria 3168, Australia. E-mail: rob.mclachlan{at}med.monash.edu.au
Abstract
Testosterone (T) treatment suppresses gonadotropin levels and sperm counts in normal men, but the addition of a progestin may improve the efficacy of hormonal contraception. This study aimed to investigate the speed and extent of suppression of testicular germ cell number induced by T plus or minus progestin treatment and correlate these changes with serum gonadotropins and inhibin B levels, testicular androgens, and sperm output. Thirty normal fertile men (31–46 yr) received either testosterone enanthate (TE, 200 mg im weekly) alone or TE plus depot medroxyprogesterone acetate (DMPA, 300 mg im once) for 2, 6, or 12 wk (n = 5 per group) before vasectomy and testis biopsy. Five men (controls) proceeded directly to surgery. The inclusion of DMPA led to a more rapid fall in serum FSH/LH levels (time to 10% baseline: FSH; 12.6 ± 2.6 vs. 7.9 ± 1.4 d; LH, 9.9 ± 3.4 vs. 3.4 ± 1.7 d, TE vs. TE+DMPA, respectively, mean ± SD, both P < 0.0001), yet the mean time to reach a sperm count 10% of baseline was not different (23.7 ± 7.3 vs. 25.3 ± 13.9 d, NS). The maximum extent of FSH/LH suppression was identical at 12 wk (mean serum FSH 1.2 and 1.6%, and mean LH 0.3 and 0.2% of baseline: TE vs. TE+ DMPA, respectively) as was sperm count suppression (5 of 5 and 4 of 5 men, respectively, with sperm counts 
0.1 x 106/ml). Serum inhibin decreased to 55% control at 12 wk in the TE+DMPA group (P < 0.05) but was unchanged by TE treatment (86% control, NS). Testicular T levels declined to approximately 2% of control levels, but testicular dihydrotestosterone and 5
-androstane-3,17ß-diol (Adiol) levels were not different to control. Germ cell numbers as determined by stereological methods did not differ between TE and TE+DMPA except at 2 wk when type B spermatogonia and early spermatocytes were significantly lower in the TE+DMPA group (P < 0.05). In all groups, a marked inhibition of Apale
B spermatogonial maturation was seen along with a striking inhibition of spermiation. We conclude that: 1) the addition of DMPA hastens the onset of FSH/LH suppression, correlating with a more rapid impairment of spermatogonial development, but in the longer term, neither germ cell number nor sperm count differed; 2) testicular dihydrotestosterone and Adiol levels are maintained during FSH/LH suppression despite markedly reduced T levels suggesting up-regulation of testicular 5-reductase activity; and 3) spermatogonial inhibition is a consistent feature, but spermiation inhibition is also striking and is an important determinant of sperm output.
EXOGENOUS T ADMINISTRATION to normal men, either alone or in combination with a progestin, reduces the secretion of the pituitary gonadotropins, LH, and FSH, and thereby sperm production, and is a promising approach to male contraception. In Caucasian men, hormonal contraceptive formulations using testosterone enanthate (TE) alone induce azoospermia in approximately 70% of subjects while higher rates are seen in Asian men (1, 2).
The addition of a progestin to a regimen of T alone appears to improve the rate of azoospermia (3). In particular, the addition of a progestin to suboptimal doses of T alone promotes better spermatogenic suppression and allows a lower dose of T to be used (4). Nonetheless, regardless of contraceptive regimen, there remain some individuals in whom steroidal contraceptive administration does not adequately suppress sperm counts for effective contraception.
The basis for different responses of men to T-based contraception is unclear and not explicable on the basis of T pharmacokinetics (5) or anthropometric measures (6). Although sperm counts of less than 3 million per milliliter may provide adequate contraception (2), there is a general consensus that the reliable induction of azoospermia is important to ensure contraceptive efficacy and the widespread acceptance of male hormonal contraception. An understanding of the biological basis for the variable response is essential to reach this goal. The assessment of the degree of spermatogenic suppression in human studies is almost always based on sperm counts which, although representing a key end point for fertility potential, cannot contribute to the understanding of the process by which spermatogenesis is variably affected by T and/or progestin treatment.
Studies on the hormonal dependency of spermatogenesis in monkeys and man have produced evidence that both FSH and LH (via the stimulation of testicular T secretion) are required for quantitatively normal sperm output. In both species, spermatogonia are highly sensitive to gonadotropin withdrawal (7, 8, 9, 10); however, which spermatogonial subtypes are regulated by hormones is controversial (9, 10, 11, 12, 13). In a study of germ cell development in men receiving weekly injections of TE for 20 wk (14), the inhibition of development of type A to B spermatogonia was striking, but in addition the retention of mature spermatids in the epithelium strongly suggested a failure of sperm release (spermiation). We recently reached a similar conclusion in regard to chronically T-treated monkeys (15). Spermiation failure is also seen in the gonadotropin-deficient rat and appears to be regulated by both FSH and LH (16, 17, 18).
The early fall in sperm counts (<4 wk) reported in some human contraceptive trials is not consistent with a lesion in the earlier stages of spermatogenesis (i.e. spermatogonial development) but rather with an effect on a late stage of spermatogenesis (i.e. spermiogenesis or spermiation). Furthermore, although no relationship has been reported between the induction of azoospermia and the extent of decline in FSH/LH levels, existing gonadotropin assays lack the sensitivity to confidently make this conclusion when serum FSH levels fall below 5% of baseline. In this study we sought to describe the time course of suppression in germ cell development induced by exogenous T administration, alone or in combination with a progestin, in normal men, to better understand how spermatogenic cell populations are affected by contraceptive treatment. We further sought to relate these changes to those in sperm output, serum gonadotropins (using assays of increased sensitivity) (19), serum inhibin B, and testicular androgen levels.
Materials and Methods
Subjects
Thirty-five normal men presenting for vasectomy were recruited for the study. These men underwent hormone treatment along with blood and semen testing before a testicular biopsy at the time of surgery for the evaluation of testicular morphology. The men were all Caucasian, aged 31–46 yr, all had fathered children, and had no history of significant illness or medication. General and testicular examination were normal and testicular volume was determined using a Prader orchidometer. The men had at least two normal semen analyses using World Health Organization criteria (2) and normal serum T, LH, FSH, and PRL levels, fasting lipids, liver function tests, prostate-specific antigen, and full blood examination. The study was approved by the Human Research and Ethics Committee of the Monash Medical Center, and informed consent was obtained from each subject.
Design
Groups of five men received either TE [Primotestin depot, Schering AG Berlin, Germany, 200 mg (0.8 ml im weekly)] for 2, 6, or 12 wk before surgery (groups TE2, TE6, and TE12) or the same dose and duration of TE but also including a single im injection of depot medroxyprogesterone acetate (DMPA, 300 mg in 2 ml, Upjohn Pharmaceuticals, Rydalmere, Australia) at wk 0 (groups TE +DMPA 2, 6, and 12). The remaining five men served as controls and proceeded directly to surgery.
Semen analyses were performed after abstinence from ejaculation for 2–5 d and were obtained weekly during TE treatment, the last being 1 wk before surgery (i.e. after 1, 5, and 11 wk of treatment). In all 12-wk subjects, sperm density was reduced to either azoospermia or virtual azoospermia (occasional sperm seen in a concentrated sample, estimated to be <0.1 million per milliliter). Gonadotropin levels were assessed on d 1, 2, 4, and 7 in the first week of treatment and then weekly until the day of surgery. Inhibin B was measured at baseline and at biopsy in all men. Serum T was measured at baseline in all groups and was measured weekly in the 12-wk TE and TE+DMPA groups only. All blood samples were taken at random times in relation to TE injections.
Endocrine hormone assays
Serum LH and FSH were measured using commercial immunofluorometric assays (Delfia, Wallac, Inc., Turku, Finland) that were modified to increase assay sensitivity (19). The modifications consisted of an increase in serum volume, the inclusion of gonadotropin-depleted serum to offset matrix effects, and longer incubation times. The sensitivity of the FSH assay was 0.010 IU/liter with a between-assay variation of 10%. The FSH standard preparation was calibrated in terms of second international reference preparation of pituitary FSH/LH (78/549). The sensitivity of the LH assay was 0.005 IU/liter with a between-assay variation of 10%. The hLH kit standard preparation was calibrated in terms of the second international standard for pituitary LH (80/552).
Inhibin B ELISA
An inhibin B ELISA method (20) was employed without modifications using kit reagents and inhibin B standard provided by Oxford Bio-innovation Ltd. (Oxon, UK). The assay sensitivity was 4 pg/ml and the between-assay variation was 15%.
Serum T
Serum T was measured by a commercial immunofluorometric procedure (ACS:180, Bayer, East Walpole, MA) with between-assay variation of 5.5% and a working range of 0.3–58 nmol/liter. Serum E2 was measured by a commercial kit RIA from Diagnostic Products Corp. (Los Angeles, CA) with between-assay variation of 15% and a working range of 20–1840 pmol/liter.
Testicular steroids
Intratesticular T (iT), dihydrotestosterone (iDHT) and 5-androstane-3,17ß-diol (iAdiol) were measured in testis homogenates (17–87 mg tissue/sample) using HPLC and RIA, as validated previously for rats (21). To follow steroid recoveries throughout sample processing, 9,000–12,000 cpm of radiolabeled 3H-T (1,2,6,7,16,17-3H-T, 121Ci/mmole), 3H-dihydrotestosterone (3H-DHT) (1,2,3,4,5,6-3H-DHT, 110Ci/mmole) (NEN Life Science Products, Boston, MA), and 3H-Adiol (1,2,3,4,5,6-3H-Adiol) were added to each tube after homogenization. 3H-Adiol was prepared in-house from 3H-DHT using 3-hydroxysteroid dehydrogenase (Sigma, St. Louis, MO) with separation of the products by HPLC. The specific activities of all three tritiated tracers were determined by RIA and shown to be 66–110Ci/mmole. A single assay for each of T, DHT, and Adiol was run using unlabeled T, DHT, and Adiol as standard, respectively. All assays used 125I-testosterone as tracer. Recoveries of added 3H-steroid tracers at the final RIA stage were T: 51.6% (SD: 1.8%, n = 38), DHT: 34.9% (SD 5.8%, n = 38), and Adiol: 52.9% (SD 4.3%, n = 38). The intraassay variations (n = 3) from parallel fragments from the same testis sample were 14% for T, 18% for DHT, and 21% for Adiol. All samples were assayed in the one assay. The sensitivity of the combined extraction, HPLC, and RIA components of the assay was calculated from the sensitivity of the RIA, the average recoveries of tritiated steroid and the average tissue mass extracted in the assay. These values were 0.16 ng/g testis for T, 0.70 ng/g testis for DHT, and 0.46 ng/g for Adiol. Serum samples from control men (n = 5) and men given TE for 6 wk (n = 5) were also processed using this method to measure serum 5-reduced androgens.
Stereology
Testicular biopsy and tissue processing. A single testicular biopsy was obtained from one testis following incision of the tunica. Great care was taken to avoid compression of the testis to minimize disruption of the testicular architecture. Approximately 100 mg of biopsy tissue was frozen in liquid nitrogen and stored at -80 C for the determination of testicular androgen concentrations. The remainder of the biopsy was immersion fixed in Bouin’s fluid for about 3–5 h and then observed under a dissecting microscope. Under direct vision the biopsy was divided in two along a plane lying at right angles to the dominant tubule orientation. One piece was then further processed and embedded in hydroxyethyl-methacrylate resin (Technovit 7100, Kulzer and Co., Friedrichsdorf, Germany) according to the manufacturer’s instructions and 3 x 25 µm sections were cut, with a space of 100 µm in between, using a Supercut microtome (Reichert Jung 2050, Nossloch, Germany). The sections were then stained with periodic acid-Schiff’s using a commercial Schiff’s reagent (Amber Scientific, Belmont, Western Australia, Australia) and counterstained with hematoxylin or stained with conventional hematoxylin and eosin.
The estimation of germ and Sertoli cell number
Germ cells and Sertoli cells were identified based on their nuclear morphology (22, 23, 24) (Fig. 1
). Type A dark (Ad) and A pale (Ap) spermatogonia were present in all stages of the spermatogenic cycle and were distinguished as follows: Ad spermatogonia had nuclei that were usually broadly applied to the plasma membrane; had heterogenous, deeply stained chromatin; and usually contained one or more clear vacuoles. There was considerable heterogeneity in terms of morphology within the Ad classification (see Fig. 1). Ap spermatogonial morphology showed less variation, the nuclei contained one or two nucleoli attached to the nuclear membrane, and the nucleus was usually close to, but not in contact with, the basement membrane. Ap nuclei were more lightly stained with hematoxylin, had a more homogenous chromatin pattern, and were slightly larger, compared with Ad.
|
Germ cells were grouped as follows: Ad spermatogonia; Ap spermatogonia; type B spermatogonia; preleptotene + leptotene + zygotene spermatocytes; pachytene spermatocytes; step 1–2 round spermatids (14), which equate to the Sa class described by Clermont (22, 23, 24); steps 3–6 elongating spermatids, which equate to Sb + Sc classes; and step 7–8 elongated spermatids, which equate to the Sd class of spermatids (see Fig. 1
).
Statistics
Data are shown as mean ± SD for all parameters except germ cell data, which is shown as mean ± SEM. For germ cells, n = 5 per group is shown; however, for serum hormones and semen, data were pooled from the 2-, 6-, and 12-wk groups where appropriate. Statistical comparisons were performed using GB Stat (Dynamic Systems Inc., Silver Spring, MD). All data were log transformed before analysis and the variances analyzed for homogeneity. Nonparametric statistics were used when equal variance tests failed. Baseline characteristics were analyzed by one-way ANOVA and ED90 values were analyzed by unpaired t test. Serum hormone and semen data in treated men were analyzed using repeated-measures ANOVA or nonparametric Friedman ANOVA, followed by paired t test or Wilcoxon signed-rank test, to investigate differences, compared with baseline values. Differences among control, TE, or TE+DMPA groups at various time points were analyzed by unpaired t test or Mann-Whitney U test. Germ cell data were analyzed by one-way ANOVA and Newman-Keuls post hoc comparison for differences between groups.
Results
Endocrinology
There were no significant differences in age, sperm output, and reproductive hormones among the control, TE, and TE+DMPA treatment groups at baseline (Table 1). In response to TE and TE+DMPA treatment, serum T levels increased by the end of wk 1 to 207% ± 39% and 196% ± 43% baseline and serum E2 to 264% ± 37% and 229% ± 71% baseline, respectively, and remained at that level until 12 wk (Table 1). At 6 wk of TE treatment, serum DHT levels tended to increase above that in control men (8.91 vs. 4.06nmol/liter, NS), and serum Adiol levels were also not altered by this treatment (5.47 vs. 5.75 nmol/liter, NS).
|
, and the absolute values in control, TE, and TE+DMPA men at biopsy are shown in Table 1. The combination of TE+DMPA significantly accelerated the rate of decline in both gonadotropins when compared with TE alone, serum LH levels falling to 10% of baseline levels by 3.4 ± 1.7 vs. 9.9 ± 3.3 d (P < 0.0001) and serum FSH by 7.9 ± 1.4 vs. 12.6 ± 2.6 d (P < 0.0001), respectively. LH levels, expressed as percent of baseline, remained significantly lower in the TE+DMPA group until wk 6 of treatment, with both groups showing a similar extent of suppression thereafter (Fig. 2A). LH levels in both groups were consistently below 1% of baseline after 6 wk, and by 12 wk LH was 0.2% and 0.3% of baseline in the TE+DMPA and TE groups, respectively, with 5 of 5 men in the TE+DMPA group and 2 of 5 men in the TE group having at least one undetectable value.
|
and Table 1). In contrast to LH, FSH levels remained above 1% of baseline in both groups throughout the treatment period and remained consistently detectable, such that at 12 wk only 1 of 5 men in each group showed undetectable values. The maximal extent of FSH suppression at 12 wk was 1.6% and 1.2% of baseline in the TE+DMPA and TE groups, respectively.
Serum inhibin B levels (Fig. 2C) did not alter at 2 wk and showed a modest decline to approximately 55% baseline in the TE+DMPA groups at 6 and 12 wk (P < 0.05 and <0.01, respectively, compared with baseline values) but not in the TE group (86% baseline at 12 wk, NS). Inhibin B levels were significantly lower in the TE+DMPA-treated men, compared with the TE-treated men at 12 wk (P < 0.05, n = 5/group) but not at the other time points (Fig. 2C and Table 1).
The rate and extent of sperm count suppression in both TE- and TE+DMPA-treated men were similar (Fig. 2D and Table 1). Sperm counts first showed a significant decrease, compared with baseline, after 2 wk in the TE-treated men (P < 0.05, n = 10) and 3 wk in the TE+DMPA-treated men (P < 0.05, n = 10) (Fig. 2D). There was no difference in the time taken to suppress sperm counts to 10% of baseline when comparing the two groups (23.7 ± 7.3 vs. 25.3 ± 13.9 in TE vs. TE+DMPA groups, n = 10/group, NS), and there were no differences between the groups at any time during the treatment period. In the 6- and 12-wk TE-treated groups, a sperm count less than 1 million per milliliter was achieved in 9 of 10 men after 3–10 wk (mean 5.4 wk), and in the TE+DMPA groups, 8 of 10 men achieved this threshold after 3–11 wk of treatment (mean 5.8 wk) (data not shown).
Testicular steroid levels
In control men, iT levels (2231 ± 403 nmol/liter) were 100-fold higher than those of iDHT and iAdiol (22.0 ± 8.3 and 23.6 ± 10.9 nmol/liter, respectively, Fig. 3). The iT levels fell to 5.2% (P < 0.05) and 2.2% (P < 0.05) of the control group by 2 wk of TE or TE+DMPA treatment, respectively. The iT levels were significantly lower in the TE+DMPA men, compared with the TE men at this time (P < 0.05, n = 5/group, Fig. 3A) as were serum LH levels (Fig. 2A). By 6 and 12 wk, iT levels were approximately 2% of control in both treatments and were not different between the TE- vs. TE+DMPA-treated groups.
|
-reduced androgens DHT and Adiol (NS, Fig. 3), either compared with control or between TE vs. TE+DMPA. The levels of DHT and Adiol in the testes of control men were significantly (P < 0.05) greater than those measured in serum [testis DHT (nmol/liter): 22.02 ± 8.22; serum DHT: 4.06 ± 3.65 (P < 0.05); testis Adiol: 23.60 ± 10.38; serum Adiol 5.47 ± 4.96 (P < 0.05, mean, SD, n = 5)] (see also Fig. 3). After treatment with TE alone for 6 wk, iDHT and Adiol levels were similar to serum levels [testis DHT (nmol/liter) 11.87 ± 5.99; serum DHT 8.91 ± 2.48 (NS); testis Adiol 9.41 ± 5.16, serum Adiol 5.47 ± 4.96 (NS, mean, SD, n = 5)]. Because no differences were observed between serum and testicular 5-reduced androgens at 6 wk in the TE group, we did not investigate 5-reduced androgens in the 12-wk groups. Germ cell numbers
Mean data.
Germ cell data for control and hormone-treated men are expressed as number per Sertoli cell (Table 2) and as a percentage of that seen in the five control subjects for the different time points (Fig. 4).
|
|
) or as a percentage of the control group (Fig. 4). When comparing the treated groups with the control men, most germ cell numbers were similar with the exception in TE+DMPA men in whom the number of Ad spermatogonia were significantly elevated (130%), but the number of B spermatogonia were significantly lower (60%), and the number of pachytene spermatocytes and round spermatids seemed to be higher than the control group at 2 wk (Table 2).
Six weeks.
Both treatment groups showed similar patterns of spermatogenic suppression, with no significant differences between the TE and TE+DMPA groups (Table 2 and Fig. 4). When compared with the control group, significant reductions were seen in germ cells from type B spermatogonia through to pachytene spermatocytes (Table 2). However, at this time, round, elongating, and elongated spermatids were not significantly different from the control group. Notably, elongated spermatids (step 7–8) were not different from control in either group (Fig. 4, Table 2) despite the fact that sperm counts in the week before biopsy were less than 10% of the control group (Fig. 4).
Twelve weeks.
Both treatment groups showed similarly profound reductions in all germ cell types (Fig. 4). Germ cells from Ap spermatogonia through to elongated (step 7–8) spermatids were significantly below the control group (Table 2). In both groups, elongated spermatids were present in the testes at 17–23% of control despite the fact that their sperm counts were less than 2% of the control group in the week before biopsy (Fig. 4).
Parameters predictive of a rapid fall to sperm count 1 million per milliliter or less
The data from men in the 6- and 12-wk groups were considered on the basis of whether they achieved a sperm count of 1 million per milliliter or less within 6 wk (n = 12) or did not reach this level of suppression within 6 wk (n = 8). Differences were sought among serum LH, FSH, and inhibin B (from baseline through to 6 wk) to address the issue of whether circulating hormones predict which men will achieve a more rapid suppression of sperm count. None of these endocrine parameters were related to the rapid fall in sperm count. Similarly, germ cell numbers at the time of biopsy were not indicative of the speed of onset of severe oligospermia in the preceding weeks.
Presence of elongated spermatids in the testes of azoospermic men
Elongated spermatids were frequently seen to be retained in men whose sperm count was zero or extremely low. An example is shown (Fig. 5) in which clusters of retained spermatids are apparent in the testicular biopsy, yet this individual achieved a sperm count of less than 1 million per milliliter 3 wk earlier.
|
In this study we report the time course of the suppression of gonadotropins, serum, and testicular androgens; inhibin B; sperm counts; and germ cell numbers during T treatment, alone or in combination with a long-acting depot progestin, to better understand the effects of male hormonal contraception. We have shown that the addition of DMPA accelerated the rate of decline in both serum FSH and LH but that the extent of suppression was similar after 12 wk of treatment. The more rapid decline in gonadotropins with DMPA was associated with a more marked impairment of type B spermatogonia and early spermatocyte maturation, but the extent of germ cell suppression at later time points did not differ from that seen with TE alone. The addition of DMPA did not hasten the onset of suppression of sperm counts, and the level of sperm count suppression achieved did not differ in the long term. A striking finding with all treatments was the profound inhibition in sperm release such that severe oligospermia was seen within 6 wk of treatment at a time when elongated spermatid number remained at control levels.
Similar to our previous study (14), long-term gonadotropin suppression induced marked suppression of sperm counts (9 of 10 men in the 12-wk groups achieving 0.1 million per milliliter or less), yet germ cell development from type B spermatogonia onward was maintained at approximately 20–30% of the control group. Thus, this study indicates that the addition of a progestin to a suppressive dose of T results in a more rapid induction of gonadotropin suppression and defects to germ cell development but does not change the pattern of germ cell development during long-term treatment. Furthermore, the data suggest that when gauging the contraceptive effectiveness of potential male hormonal contraceptive regimens by considering sperm output, both spermatogonial inhibition and defective spermiation must be taken into account.
FSH/LH
Highly sensitive gonadotropin assays (19) allowed us to determine that although serum LH is suppressed to exceedingly low levels (
0.3% of baseline), FSH remains readily detectable in the serum of most men. Whether these low residual levels of FSH (1–5% baseline) have any biological effect to maintain germ cell development is unclear. Recent studies in monkeys have provided evidence that inadequate suppression of FSH can be associated with a failure to fully inhibit spermatogenesis. We recently reported that T-treated monkeys achieving consistent azoospermia had significantly lower levels of circulating FSH (both immuno- and bioactive FSH) (27) than animals with residual sperm output, although no differences were seen between groups in terms of serum bioactive LH, serum T, or testicular androgen levels. Similarly, Weinbauer et al. (28) recently reported that a transient rise in serum FSH, but not other reproductive hormones, was associated with a partial restoration of spermatogenesis in T-treated monkeys. The importance of FSH suppression for the effectiveness of human male contraceptive regimens has also been emphasized (29).
Because all men receiving treatment for 12 wk showed suppression of sperm counts to very low levels, we were unable to test the hypothesis that residual FSH could account for the failure of some men to suppress their sperm counts to levels consistent with contraceptive efficacy. Previous studies using less sensitive assays have consistently failed to show this relationship (6, 30). Handelsman et al. (6) assessed serum gonadotropin levels in azoo- and oligospermic men participating in the World Health Organization trials and, although no differences were found, it is notable that the FSH and LH assays were performed in each different center and details of the methods were lacking. This conclusion was unaffected when these samples were rerun using more sensitive methods but assay performance data were not presented. Interestingly, the only differences in serum gonadotropins were higher baseline FSH and LH levels in men later rendered azoospermic by the treatment. We were unable to confirm this observation in our small group of men in whom severe oligospermia was consistently induced. Because we have shown that serum LH levels are exceedingly low, the issue of whether the suppression of serum FSH levels is more profound in azoospermic, compared with oligospermic men, will require further examination in a large number of men undergoing contraceptive treatments.
Inhibin B
Inhibin B levels were suppressed only by TE+DMPA treatment to approximately 55% of baseline, with this decline being slow and lagging behind changes in sperm output and spermatogenesis. This modest decline is similar to that observed previously in men with congenital or experimental gonadotropin deficiency (31) and suggests a major component of inhibin B in previously normal male volunteers is gonadotropin independent. It is curious that the decrease in inhibin B at 12 wk in the TE+DMPA group was not seen in the TE group even though germ cell number and gonadotropin suppression was similar at this time.
Testicular androgens
The level of iT in control testes was 100-fold higher than that in serum (2231 vs. 17.3 nmol/liter) and was similar to that reported by Morse et al. (32) (1920 nmol/liter) and Huhtaniemi et al. (33). Gonadotropin withdrawal resulted in a 98% reduction in iT, which was also similar to that reported following T-administration (32) or GnRH agonist treatment (33). We attribute the more rapid fall in iT at 2 wk when DMPA was included to the more rapid suppression of serum LH. However, after this time no difference was seen in iT levels between the TE ± DMPA groups, despite the fact that differences in LH were observed until wk 7 of treatment. The decline in iT levels during TE ± DMPA treatment is substantially greater than those we (10) and others (7, 8) have reported in long-term gonadotropin-suppressed monkeys. In our recent primate study (Narula, A., Y. P. Gu, L. O’Donnell, P. G. Stanton, D. M. Robertson, R. I. McLachlan, and W. J. Bremner, submitted for publication), using similar steroid extraction and assay methods, we clearly showed that iT did not decline below 85% of baseline, implying that some androgen synthesis continued in the virtual absence of LH. These data imply a difference in Leydig cell regulation between man and monkeys. Indeed, the absolute levels of T in serum of men in this study were similar to their iT levels leading to the possibility that iT could derive from passive diffusion from the circulation rather than local production.
We have found that testicular levels of 5-reduced androgens are 100-fold lower than that of T in normal men. However, after long-term TE ± DMPA treatment, the levels of 5-reduced androgens (iDHT, iAdiol) were not significantly affected, compared with the control group, by either treatment. Thus, iDHT levels became approximately 25% of iT, compared with 1% of iT in the control testes. Given the higher affinity and receptor-bound half-life of DHT (34), this suggests that iDHT is at least as important as iT in mediating any residual androgen effects in this setting. An increased production of 5-reduced androgens was seen after exogenous T-induced gonadotropin suppression in men, and this increase was greater in men rendered oligo- rather than azoospermic, implying that DHT formation may permit the maintenance of low levels of spermatogenesis (35, 36). From a mechanistic viewpoint, this may be owing to the up-regulation of testicular 5-reductase type 1 during gonadotropin suppression as seen in rats (37). The fact that serum DHT levels are similar to iDHT levels in this study suggests that iDHT levels can be accounted for by passive diffusion from the circulation and thus does not support the local production of DHT within the testis during TE ± DMPA treatment; however, studies on human testicular 5-reductase are required to confirm this finding. The inclusion of 5-reductase inhibitors active against both isoenzymes in T-based contraceptive regimens is worthy of further study.
It is unclear whether these residual levels of T and/or DHT are responsible for the maintenance of meiotic and postmeiotic germ cell development at approximately 20–30% of normal. Exceedingly high levels of serum T can maintain spermatogenesis in monkeys (7) but this direct-action hypothesis has not been proven in humans. Nonetheless, use of the lowest effective dose of T to suppress gonadotropin levels, but maintain peripheral androgen action, is appropriate so as to minimize this potential anticontraceptive effect and also to not provide excess substrate for 5-reduction in the testis or elsewhere. Data from rats indicate that even when iT levels are markedly reduced (1–3% of baseline), further inhibition of spermatogenesis can be seen when an androgen receptor antagonist is administered (38, 39), implying that low levels of androgens still have a biological effect in the absence of serum LH. Inhibition of residual androgen action within the testis may account for the reported enhanced effectiveness of a hormonal contraceptive regimen which included the antiandrogenic progestin, cyproterone acetate (40).
Germ cell development
Determination of germ cell populations after 2 wk of contraceptive treatment allowed us to study the onset of spermatogenic suppression in humans. The more rapid decline in gonadotropins induced by the addition of progestin caused a more rapid onset of germ cell suppression at 2 wk, as indicated by the lower numbers of type B spermatogonia and, subsequently, preleptotene to zygotene spermatocytes. Therefore, type B spermatogonia were the first cells to decrease after contraceptive treatment in humans. This is consistent with our previous study in long-term TE-treated men, in which B spermatogonia were the primary lesion to spermatogenesis (14). However, in our recent study in cynomolgus monkeys (15), 2 wk of T administration resulted in a significant decrease in type Ap but not type B spermatogonia. Thus, it seems that during acute gonadotropin suppression, Ap spermatogonia are the first cells to be affected in monkeys, whereas B spermatogonia are the first to be affected in humans. There is evidence for hormonal regulation of both spermatogonial subtypes in primates (7, 9, 12, 13, 15) and therefore the differences noted between primates and humans may be owing to different sensitivities of the subtypes to FSH and/or androgens.
It is also likely that Ap spermatogonial mitosis was affected in the short term (at both the 2- and 6-wk time points) in the current study because, although changes in Ap numbers were not seen, increases in Ad spermatogonia were noted. It is generally considered that when Ap spermatogonia cease to proliferate, they differentiate without division into mitotically inactive Ad spermatogonia (41, 42). Our data thus suggest that when gonadotropins are acutely suppressed in humans, Ap spermatogonia cease to proliferate into B spermatogonia and instead differentiate into Ad spermatogonia, leading to a decrease in the number of type B spermatogonia and a corresponding increase in the number of type Ad. Direct effects of gonadotropin suppression on the type B spermatogonial population (e.g. by the induction of apoptosis) are also to be expected because these cells are well known to be affected by hormones and are particularly sensitive to FSH (10, 12, 13, 15).
Our study paradigm does not allow us to identify the relative contributions of the reduction in intratesticular T or circulating FSH to the acute impairment of spermatogonial populations. The limited number per group precluded detailed correlation analysis, although we did note that the number of B spermatogonia at the 2-wk time point showed a positive correlation with FSH but not testicular androgens, suggesting that FSH may be the primary factor causing the initial decline in type B spermatogonial populations.
After 6 wk of gonadotropin suppression, both groups of men showed decreases in germ cell populations from type B spermatogonia through to pachytene spermatocytes. The duration of one spermatogenic cycle in the human is 16 d (43), and it takes two cycles for a cell to proceed from a type B spermatogonia to the end of the pachytene spermatocyte phase. Given that the primary decrease was in type B spermatogonial numbers, it is expected that round and elongated spermatid populations had not yet been affected by this time. The fact that both the TE alone and the TE+DMPA groups had similar patterns of germ cell development at the 6-wk time point suggests that the onset of disruption to spermatogonia in the TE-alone group must have occurred only a matter of days after the TE+DMPA group, probably reflecting the 5- to 6-d difference in the onset of gonadotropin suppression between the two treatments.
After 12 wk of treatment, both groups showed decreases in germ cell numbers from type Ap spermatogonia through to mature elongated spermatids. Thus, longer-term treatment produced decreases in the Ap spermatogonial population, supporting our earlier contention that Ap spermatogonia are also targets for gonadotropins. However, type B populations were still lower than Ap when expressed as a percentage of the control group, further suggesting that B spermatogonia are the primary lesion to human spermatogenesis after contraceptive treatment. Our previous study in men given TE for 20–24 wk (14) did not subdivide Ad and Ap spermatogonia but showed that although B spermatogonia were markedly suppressed, the entire A spermatogonial population was unchanged. This was also noted in the TE-alone group in the current study, in that total A spermatogonial numbers were not different to control, yet type B’s were markedly reduced. Total A spermatogonial numbers were, however, significantly lower than the control group in the TE+DMPA group at 12 wk, suggesting that the addition of progestin produces slightly more marked effects on spermatogonia. Nonetheless, addition of the progestin did not affect other germ cell populations, suggesting that in the long term, the overall pattern of germ cell development is similar during TE or TE+DMPA contraception.
After long-term gonadotropin suppression, it was clear that spermatogenesis from type B spermatogonia onward was maintained at approximately 20% of the control group. The maintenance of qualitatively normal meiotic and postmeiotic germ cell development, albeit at reduced levels, was also noted in our previous study (14) and suggests that large numbers of germ cells are not lost by apoptosis during meiosis and spermiogenesis.
Spermiation
The failure of spermiation was a feature of both acute and chronic spermatogenic suppression and was characterized by the presence of large numbers of elongated spermatids before spermiation and of retained spermatids after spermiation in the seminiferous epithelium in the presence of severe oligospermia. The maintenance of normal numbers of mature elongated spermatids in the testis at 6 wk, yet marked suppression of sperm counts, to less than 1 million per milliliter in the majority of cases, in the preceding week suggests near-complete spermiation failure within 5 wk of contraceptive treatment. Clinical contraceptive trials add strong evidence for an inhibition of spermiation wherein sperm counts have been observed to drop below 1 million per milliliter within 4 wk (3). Taking into account epididymal transit time and the kinetics of spermatogenesis, these data suggest acute defects in spermiation, highlighting the importance of spermiation failure for the onset of contraceptive suppression. Spermiation failure is also important for the maximal suppression of sperm counts in the longer term because both the current and our previous (14) study demonstrated mature spermatids in the testis at a level 20% that of the control group yet reductions in sperm counts to less than 0.1 million per milliliter in the majority of cases. Our previous study in monkeys (15) suggested a relationship between the degree of spermiation failure and the extent of sperm count suppression supporting our contention that spermiation failure is an important feature of chronic contraceptive treatment. Sperm release in rodents is known to be dependent on both FSH and T (16, 18); however, the molecular mechanisms regulating sperm release are unclear (17, 44). Spermatids that fail to be released from the seminiferous epithelium during spermiation failure are retained by the Sertoli cells and subsequently phagocytosed (17).
Mitosis of type B spermatogonia (45) and spermiation occur in stages I-II of the human spermatogenic cycle. The fact that spermatogonial development and spermiation are the two major defects in human spermatogenesis induced by contraceptive treatment suggests these stages may be more hormone sensitive than others. The cellular associations in stages I-II in the human are analogous to stages VII-VIII in the rat, which are well known to be sensitive to hormone suppression (46). Immunohistochemical analysis of androgen receptor showed that staining was more intense in stages I-III than at later stages suggesting that stage-specific variation in hormone sensitivity may also apply to humans (47).
In conclusion, we have shown that the addition of DMPA hastens the onset of FSH/LH suppression, correlating with a more marked impairment of spermatogonial development, but in the longer term, neither germ cell number nor sperm count differed. Although spermatogonial inhibition is a consistent feature, spermiation inhibition is also striking and is an important determinant of sperm output. Testicular DHT and Adiol levels are maintained during FSH/LH suppression despite markedly reduced T levels suggesting up-regulation of 5-reductase activity, although there was no direct evidence that their synthesis was occurring in the testis. The degree to which these effects vary between individuals must be explored further because they may be able to be exploited to enhance contraceptive efficacy for individual patients.
Acknowledgments
The excellent assistance of the following is acknowledged: clinical research nurses Sue Morton and Maree Sperling for the performance of the study; Nic Balasz and Michael Daskalakis of the Chemical Pathology Department, Southern Health Pathology, Monash Medical Center, for the performance of selected reproductive hormone assays; Fiona McLean for intratesticular steroid assays; Enid Pruysers for the LH, FSH, and inhibin B assays; Peter Royce and Dr. Nigel Wreford for aspects of the sample collection and preparation; Wallac, Inc. (Turku, Finland) for the provision of reagents for the LH and FSH assays; Bio-Innovations Ltd. (United Kingdom) for the provision of reagents for the inhibin B assay; and Sue Panckridge for help with the preparation of the figures.
Footnotes
This work was supported by an Australian NH&MRC Program Grant (973218) and a Wellcome Trust UK Populations Studies Grant (053741).
Abbreviations: Ad, Type A dark; Adiol, 17ß-diol; Ap, type A pale; DMPA, depot medroxyprogesterone acetate; iAdiol, intratesticular 17ß-diol; iDHT, intratesticular dihydrotestosterone; iT, intratesticular T; TE, testosterone enanthate.
Received June 22, 2001.
Accepted October 31, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  S. M. Ruwanpura, R. I. McLachlan, K. L. Matthiesson, and S. J. Meachem Gonadotrophins regulate germ cell survival, not proliferation, in normal adult men Hum. Reprod., February 1, 2008; 23(2): 403 - 411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Dias, S. Meachem, E. Rajpert-De Meyts, R. McLachlan, U. Manuelpillai, and K.L. Loveland Activin receptor subunits in normal and dysfunctional adult human testis Hum. Reprod., February 1, 2008; 23(2): 412 - 420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Walton, N. Kumar, D. T. Baird, H. Ludlow, and R. A. Anderson 7{alpha}-Methyl-19-Nortestosterone (MENT) vs Testosterone in Combination With Etonogestrel Implants for Spermatogenic Suppression in Healthy Men J Androl, September 1, 2007; 28(5): 679 - 688. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. T. Page, T. F. Kalhorn, W. J. Bremner, B. D. Anawalt, A. M. Matsumoto, and J. K. Amory Intratesticular Androgens and Spermatogenesis During Severe Gonadotropin Suppression Induced by Male Hormonal Contraceptive Treatment J Androl, September 1, 2007; 28(5): 734 - 741. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Wang, Y.-G. Cui, X.-H. Wang, Y. Jia, A. Sinha Hikim, Y.-H. Lue, J.-S. Tong, L.-X. Qian, J.-H. Sha, Z.-M. Zhou, et al. Transient Scrotal Hyperthermia and Levonorgestrel Enhance Testosterone-Induced Spermatogenesis Suppression in Men through Increased Germ Cell Apoptosis J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3292 - 3304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. T. Page, J. K. Amory, B. D. Anawalt, M. S. Irwig, A. T. Brockenbrough, A. M. Matsumoto, and W. J. Bremner Testosterone Gel Combined with Depomedroxyprogesterone Acetate Is an Effective Male Hormonal Contraceptive Regimen and Is Not Enhanced by the Addition of a GnRH Antagonist J. Clin. Endocrinol. Metab., November 1, 2006; 91(11): 4374 - 4380. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. L. Matthiesson, R. I. McLachlan, L. O'Donnell, M. Frydenberg, D. M. Robertson, P. G. Stanton, and S. J. Meachem The Relative Roles of Follicle-Stimulating Hormone and Luteinizing Hormone in Maintaining Spermatogonial Maturation and Spermiation in Normal Men J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 3962 - 3969. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Beardsley, D. M Robertson, and L. O'Donnell A complex containing {alpha}6{beta}1-integrin and phosphorylated focal adhesion kinase between Sertoli cells and elongated spermatids during spermatid release from the seminiferous epithelium. J. Endocrinol., September 1, 2006; 190(3): 759 - 770. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Walton, R. A. L. Bayne, I. Wallace, D. T. Baird, and R. A. Anderson Direct Effect of Progestogen on Gene Expression in the Testis during Gonadotropin Withdrawal and Early Suppression of Spermatogenesis J. Clin. Endocrinol. Metab., July 1, 2006; 91(7): 2526 - 2533. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 |
) for 2–12 wk. All data are expressed as a percentage of the subjects’ baseline value and shown as mean ± SD. A-C, d 1–14, n = 15; wk 3–6, n = 10; wk 7–12, n = 5. D, d 7, n = 15; wk 2–5, n = 10; wk 6–11, n = 5. Note the log axis for gonadotropins and sperm count data. *, P < 0.05 between TE and TE+DMPA groups by unpaired t test or Mann-Whitney U test.
, P less than 0.05, compared with the TE-alone group at the same time point.
