7α-Methyl-19-Nortestosterone (MENT) vs Testosterone in Combination With Etonogestrel Implants for Spermatogenic Suppression in Healthy Men

Authors

  • Melanie J. Walton,

    1. Division of Reproductive and Developmental Sciences, Centre for Reproductive Biology, The Queen's Medical Research Institute, University of Edinburgh, Edinburgh, United Kingdom
    Search for more papers by this author
  • Narender Kumar,

    1. Center for BioMedical Research, Population Council, New York, New York
    Search for more papers by this author
  • David T. Baird,

    1. Division of Reproductive and Developmental Sciences, Centre for Reproductive Biology, The Queen's Medical Research Institute, University of Edinburgh, Edinburgh, United Kingdom
    Search for more papers by this author
  • Helen Ludlow,

    1. School of Life Sciences, Oxford Brookes University, Oxford, United Kingdom.
    Search for more papers by this author
  • Richard A. Anderson

    Corresponding author
    1. Division of Reproductive and Developmental Sciences, Centre for Reproductive Biology, The Queen's Medical Research Institute, University of Edinburgh, Edinburgh, United Kingdom
    Search for more papers by this author

Centre for Reproductive Biology, The Queen's Medical Research Institute, The University of Edinburgh, 47 Little France Crescent, Edinburgh EH16 4TJ (e-mail: richard.anderson@ed.ac.uk).

Abstract

ABSTRACT: Testosterone with a progestogen can suppress spermatogenesis for contraception. The synthetic androgen 7α-methyl-19-nortestosterone (MENT) may offer advantages because it is resistant to 5α-reduction and is therefore less active at the prostate. This study aimed to investigate MENT implants in combination with etonogestrel on spermatogenesis, gonadotropins, and androgen-dependent tissues in comparison with a testosterone/etonogestrel regimen. Healthy men (n = 29) were recruited and randomized to receive 2 etonogestrel implants with either 600-mg testosterone pellets repeated every 12 weeks or 2 MENT implants for up to 48 weeks. Testosterone concentrations in the testosterone group remained in the normal range. Subjects with 2 MENT implants showed peak MENT levels at 4 weeks with testosterone concentrations of 2 nmol/L. Sperm concentrations fell rapidly to less than 1 × 106/mL at 12 weeks in 8 of 10 subjects in the MENT group and 13 of 16 subjects in the testosterone group with equally suppressed gonadotropins. Thereafter, suppression was not maintained in the MENT group, and 6 men noted loss of libido. Fourteen men completed 48 weeks of testosterone treatment, and all became azoospermic. Hemoglobin concentrations rose, and high density lipoprotein-cholesterol (HDL-C) fell in both groups. The MENT group showed a fall in prostate-specific antigen with no change in bone mass. MENT with a progestogen can achieve rapid suppression of spermatogenesis similar to testosterone, but this promising result was not sustained due to a decline in MENT release from the implants. This dose of testosterone, compared with previous studies using a lower dose with a higher dose of etonogestrel, had nonreproductive side effects without any increase in spermatogenic suppression. These data indicate the importance of the doses of progestogen and testosterone for optimum spermatogenic suppression while minimizing side effects.

Hormonal methods of contraception for men are based on suppression of gonadotropins and hence spermatogenesis. Administration of testosterone alone provides effective reversible contraception (World Health Organization Task Force on Methods for the Regulation of Male Fertility, 1990, 1996), but azoospermia is not universally achieved and the dose of testosterone produces unwanted side effects. Administration of a progestogen with testosterone increases spermatogenic suppression while allowing a reduction in the dose of testosterone toward that of physiologic replacement (Anderson and Baird, 2002; Kamischke and Nieschlag, 2004).

An alternative approach is the use of a synthetic androgen. Testosterone is a prohormone in many tissues, converted by 5α-reductase to dihydrotestosterone (DHT) (eg, in the prostate) and by aromatase to estradiol (eg, in bone). Tissue selectivity can thus be conferred by altered susceptibility to conversion by these enzymes. This is exemplified by 7α-methyl-19-nortestosterone (MENT), which is resistant to 5α-reduction but a substrate for aromatase (Agarwal and Monder, 1988; LaMorte et al, 1994). Theoretically, this will result in relative sparing of the prostate while maintaining other androgen-dependent functions (Sundaram et al, 1993). MENT Acetate (Ac) implants induce suppression of the hypothalamo-pituitary-testicular axis in healthy men (Noé et al, 1999; von Eckardstein et al, 2003), and data to support relative sparing of the prostate have been obtained in nonhuman primates (Cummings et al, 1998) and hypogonadal men (Anderson et al, 2003).

While many studies have explored a range of preparations of testosterone and progestogens, there are limited data on the interaction between doses (Grimes et al, 2004). We have previously demonstrated that an implant formulation of the progestogen etonogestrel with long-acting testosterone pellets results in dose-dependent suppression of spermatogenesis with minimal nonreproductive effects (Anderson et al, 2002; Brady et al, 2004). The present study was designed to investigate the effectiveness and side effects of MENT as the androgen component of a prototype male contraceptive regimen involving 2 etonogestrel implants in comparison with a testosterone-based regimen. The dose of testosterone was slightly larger than that which we had previously shown to be highly effective at suppressing spermatogenesis when combined with an optimal number of etonogestrel implants (3; Brady et al, 2004).

Methods

Subjects

Twenty-nine healthy Caucasian men, mean age 34.1 years (range, 23–50 years), were recruited as previously described (Anderson et al, 2002; Kinniburgh et al, 2002; Brady et al, 2004). Inclusion criteria included good mental and physical health; body mass index (BMI) between 18 and 32 kg/m2; normal follicle stimulating hormone (FSH), luteinizing hormone (LH), and testosterone concentrations and standard biochemical and hematologic variables; and normal physical examination (Table 1). Blood pressure was measured in the sitting position after 5 minutes of rest using an automated device and was required to remain below 140 mm Hg systolic and below 90 mm Hg diastolic for the duration of the study. Two pretreatment semen samples were required with sperm concentration above 20 × 106/mL and with motility and morphology within local normal limits. All participants provided written informed consent. The study was approved by Lothian Research Ethics Committee.

Table 1. . Pretreatment characteristics of men in the 2 groups*
 MENT GroupTestosterone Group
  1. *Pretreatment values for all subjects; mean ± SEM. MENT group, n = 13; testosterone group, n = 16. MENT indicates 7α-methyl-19-nortestosterone; BMI, body mass index; LH, luteinizing hormone; and FSH, follicle stimulating hormone

Age (y)34.9 ± 1.532.2 ± 1.2
BMI (kg/m2)25.6 ± 0.824.7 ± 0.7
LH (IU/L)3.3 ± 0.43.7 ± 0.3
FSH (IU/L)2.0 ± 0.42.3 ± 0.4
Testosterone (nmol/L)17.6 ± 2.721.4 ± 2.5
Sperm concentration (× 106/mL)55.1 ± 9.158.7 ± 14.0

Study Design and Medication

The study was a randomized, open-label trial investigating MENT implants or testosterone in combination with etonogestrel implants, with both groups scheduled for treatment for 48 weeks. Following 2 screening visits, subjects were randomized using a computer-generated list and numbered sealed envelopes. Subjects in the MENT group were administered 2 implants each containing 135 mg of MENT acetate formulated as in previous studies to release about 400 μg of MENT Ac per implant per day (Anderson et al, 2003; von Eckardstein et al, 2003) subcutaneously with local anesthesia into the medial aspect of 1 arm, and 2 etonogestrel implants 4 cm long, each containing 68 mg of etonogestrel (Implanon; N.V. Organon, Oss, The Netherlands) into the other upper arm. The testosterone group was administered 600-mg testosterone pellets (3 × 200 mg, N.V. Organon) subcutaneously into the anterior abdominal wall, repeated at weeks 12, 24, and 36, with 2 etonogestrel implants at first administration of testosterone. The etonogestrel and MENT implants were removed at the end of the treatment period, at which time the subjects entered the recovery phase.

Subjects were reviewed at 4-week intervals. The number of episodes of sexual activity (sexual intercourse and masturbation) over the preceding 2 weeks was recorded by interview at 12-week intervals and at 16 weeks of the recovery phase, at which times physical examination was also performed and testicular volume was determined using the Prader orchidometer. All subjects were followed until 2 semen samples had been submitted with sperm concentrations above 20 × 106/mL.

Transrectal ultrasound using a biplanar probe (Eccoccee; Toshiba, Stirling, United Kingdom) was used to measure total prostate volume (3.14/6 × anteroposterior × transverse × longitudinal measurements). The coefficient of variation (CV) for repeated measurement was 9.8%. Bone mineral density at the lumbar spine and hip was determined using a QDR-4500A (Hologic Inc, Bedford, Mass) with a CV of 1.3%.

Assays

Serum was stored at −20°C until assay. Testosterone, FSH, LH, and estradiol were measured by time-resolved immunofluorometric assay (DELFIA; Wallac, Beaconsfield, United Kingdom). Assay sensitivity was 0.3 nmol/L for testosterone, 0.05 IU/L for FSH and LH, and 50 pmol/L for estradiol. Intraassay CVs were less than 7%, and interassay CVs were less than 5% for FSH and LH and less than 10% for testosterone and estradiol. Inhibin B was assayed using a modification of the previously described assay (Groome et al, 1996; Anderson et al, 1998) using an improved primary antibody (46A/F). Results are quantified against the World Health Organization inhibin B standard and show a very high correlation with the previous method although are generally higher. Assay sensitivity is 7 pg/mL with interassay and intra-assay CVs of less than 10%. MENT in serum and in the implants following removal was measured by radioimmunoassay (Kumar et al, 1990; Suvisaari et al, 1997) with intra-assay CVs of 3.8%–7.9% and interassay CVs of 8.0%–12.3%. Due to cross-reactivity with testosterone and other serum factors, a value of 0.39 ± 0.01 nmol/L is seen in samples prior to MENT treatment. Samples were analyzed for biochemical and hematologic parameters by autoanalyzer at 12-week intervals.

Semen Analysis

Semen samples were submitted following 3–7 days of abstinence and sperm concentration determined (World Health Organization, 1999). Azoospermia was confirmed by thorough examination of the resuspended pellet following centrifugation at 3660 × g for 15 minutes.

Data Analysis

All results are presented as mean ± SEM. Serum hormone and biochemical data were log transformed before analysis by analysis of variance (ANOVA) for repeated measurements, and sperm concentrations were cube root transformed before ANOVA with Tukey's post hoc test. Proportions of men achieving thresholds for spermatogenic suppression were analyzed by the Fisher exact test. Sexual activity data were analyzed by nonparametric testing. For all comparisons, a P value of less than .05 was considered significant.

Results

Subjects, Withdrawals, and Adverse Events

Twenty-nine subjects were assigned to the treatment groups. Three subjects in the MENT group chose to leave the study after 8 weeks of treatment, 2 because of symptoms of low libido and erectile dysfunction and 1 for personal reasons. The 2 men who withdrew from the study because of reduced libido had serum MENT concentrations similar to the rest of the group. Two subjects in the testosterone group withdrew after 24 weeks of treatment, 1 for personal reasons and the other for symptoms of labile mood, sleep disturbance, and nocturia. Adverse events included reduced libido and erectile function in 4 additional subjects in the MENT group who completed treatment. In the testosterone group, single testosterone pellets were extruded in 2 subjects, both 1 week prior to scheduled readministration. One subject in the testosterone group reported increased libido and acne.

Due to the incidence of reports of low libido and early withdrawal in the MENT group, it was decided in consultation with the study Data Monitoring and Safety Committee to shorten the MENT treatment period to 24 weeks whereas men in the testosterone group completed 48 weeks of treatment.

MENT and Testosterone Concentrations

Serum MENT concentrations in that group demonstrated an initial peak at 4 weeks of 1.1 ± 0.1 nmol/L and then declined to 0.54 ± 0.05 nmol/L at week 24 (Figure 1a), similar to assay blank values. Analysis of MENT remaining in the implants after removal demonstrated that the implants had released 31% ± 2% of content over 24 weeks, giving an average release rate of 117 ± 6 μg per day per implant. Implants removed after shorter insertion periods (51–60 days; n = 3) showed a higher release rate at 329 ± 6 μg per day per implant. Testosterone concentrations in this group fell to 2.0 ± 0.4 nmol/L at 4 weeks (P < .0001 vs pretreatment; Figure 1b) and remained very low for the duration of treatment. Following removal of MENT and etonogestrel implants, testosterone concentrations rapidly returned to pretreatment values.

Figure 1.

. Serum concentrations of (a) 7α-methyl-19-nortestosterone (MENT), (b) testosterone, (c) follicle stimulating hormone (FSH), (d) luteinizing hormone (LH), (e) estradiol, and (f) inhibin B before treatment and during treatment and recovery phases. MENT group (open squares), n = 13; testosterone group (filled circles), n = 19; mean ± SEM. The MENT group received drug treatment for 24 weeks and the testosterone group for 48 weeks. Testosterone was administered at weeks 0, 12, 24, and 36. In panel b, the normal range is indicated by the broken lines.

Serum testosterone concentrations in the testosterone group were significantly greater than in the MENT group at all time points during treatment (P < .0001; Figure 1b). There was some fluctuation in keeping with the schedule of testosterone administration at 12-week intervals. Trough and peak concentrations rose during the study, indicating some accumulation. After an initial fall over the initial 12 weeks of treatment (P < .001), testosterone concentrations were similar to pretreatment at 24 weeks but at 28 and 40 weeks were significantly higher than at pretreatment (both P < .001). However, average testosterone concentrations over both the first and second 24-week periods were not significantly different than pretreatment (22.6 ± 1.9 vs 20.2 ± 1.4 vs 26.9 ± 1.5 nmol/L, pretreatment, 4–24 weeks, and 28–48 weeks, respectively).

Sperm Concentrations

Both groups showed profound suppression of spermatogenesis initially (Figure 2). At 12 weeks, 8 of 10 subjects (80%) in the MENT group and 13 of 16 (81%) in the testosterone group demonstrated suppression to less than 1 × 106/mL, with 3 and 11 subjects, respectively, being azoospermic. Thereafter, suppression was inconsistent in the MENT group. Only 4 men maintained less than 1 × 106/mL until 24 weeks, whereas the others showed partial recovery. Overall the mean sperm concentration rose to 11.5 × 106/mL ± 5.2 × 106/mL at that time.

Figure 2.

. Sperm concentrations during treatment and recovery phases. 7α-Methyl-19-nortestosterone (MENT) group (open squares), n = 13; testosterone group (filled circles), n = 19; mean ± SEM. The MENT group received drug treatment for 24 weeks and the testosterone group for 48 weeks.

In contrast, the testosterone group showed maintained suppression. At 24 weeks, 14 of 16 subjects (88%) were azoospermic, and by the end of the treatment period all subjects demonstrated azoospermia (n = 14). One of the 2 men not azoospermic at 24 weeks withdrew from the study for personal reasons, and the other became azoospermic at week 44. The range of time to achieve azoospermia was therefore 4–44 weeks (median, 12 weeks). Once azoospermic, sperm were detected only in the centrifuged pellet of the ejaculate in 1 individual on 1 occasion.

Recovery was rapid following implant removal in the MENT group, with 9 of 10 subjects achieving sperm concentration of more than 20 × 106/mL within 16 weeks. Slower recovery was seen in the testosterone group, with a median duration of 28 weeks. One subject continued to show sperm concentrations of 10 × 106/mL–15 × 106/mL after more than 1 year of follow-up, with normal total sperm number, motility and morphology, and gonadotropin concentrations. A second subject underwent vasectomy before his sperm concentration had recovered to more than 20 × 106/mL, 64 weeks after removal of etonogestrel implants.

Other Reproductive Hormones

Initial suppression of both FSH and LH was rapid and profound in both treatment groups (Figure 1). In the MENT group, mean FSH was 0.18 ± 0.03 IU/L after 4 weeks. However, progressive partial escape from suppression was evident with a rise to 1.29 ± 0.21 IU/L at 24 weeks (P = .04, ANOVA of treatment values). Concentrations were similar to pretreatment in the recovery phase. In the testosterone group the initial suppression was well maintained (Figure 1) with minor rises at times of trough testosterone concentrations that did not reach statistical significance. Two men showed FSH rises to more than 1 IU/L at 24 weeks, none at 36 weeks, and 1 at 48 weeks of treatment. FSH showed an overshoot in the recovery phase (P < .01).

LH was also profoundly suppressed in both treatment groups initially (Figure 1). In the MENT group there was a small rise during continuing treatment, but this did not reach statistical significance (Figure 1). In the testosterone group the suppression was more complete and consistent for the duration of treatment, with only 1 man showing any detectable recovery of LH to more than 1 IU/L, at 48 weeks of treatment.

Estradiol concentrations demonstrated a marked fall in the MENT group at 12 and 24 weeks of treatment (P < .001; Figure 1). In the testosterone group the results paralleled the testosterone concentrations with a small fall seen at 12 weeks (P < .001), recovering by 24 weeks (nonsignificant [ns] vs pretreatment, P < .01 vs 12 weeks). There was a significant difference in estradiol concentrations between the 2 groups at 24 weeks (P = .01).

Inhibin B concentrations showed a significant decline during treatment in both groups but differed between groups (P = .01). In the testosterone group, inhibin B concentrations declined during the first 24 weeks of treatment, with little fall thereafter. There was, however, a significant further fall during the recovery phase (P < .0001 vs week 48), with a nadir at 12 weeks of recovery. There was an inverse relation between inhibin B and FSH during the recovery phase (P = .004 at 16 weeks). In the MENT group, inhibin B concentrations were only significantly lower than pretreatment at 8 weeks (P = .001).

Hematology and Lipids

Hemoglobin concentrations were significantly increased in the MENT group at 12 weeks, but this did not persist at 24 weeks (Table 2). A similar pattern was also seen in hematocrit, but this did not reach statistical significance. In the testosterone group a slower progressive rise in hemoglobin concentration (P = .0006) was observed that only became significantly at 48 weeks. There was also a significant overall rise in hematocrit (P = .009) in the testosterone group, although none of the individual treatment time points were significantly different from pretreatment (Table 2).

Table 2. . Serum values of hematology, lipid, and prostate-specific antigen results*
 Pretreatment12 Weeks24 Weeks36 Weeks48 WeeksRecovery
  1. *Mean ± SEM. MENT group, n = 13; testosterone (T) group, n = 16. MENT indicates 7α-methyl-19-nortestosterone; HDL-C, high density lipoprotein-cholesterol; and LDL-C, low density lipoprotein-cholesterol

  2. P < .05

  3. P < .01 vs pretreatment. There were no significant differences between the MENT and testosterone groups in any of these variables

Hemoglobin      
    T group (g/L)152 ± 2.5149 ± 2.2153 ± 2.1154 ± 2.5157 ± 2.5‡154 ± 2.5
    MENT group (g/L)149 ± 2.9154 ± 3.3†148 ± 2.4148 ± 2.0
Hematocrit      
    T group0.45 ± 0.010.44 ± 0.010.45 ± 0.010.46 ± 0.010.46 ± 0.010.46 ± 0.01
    MENT group0.44 ± 0.010.46 ± 0.010.43 ± 0.010.43 ± 0.01
HDL-C      
    T group (nmol/L)1.4 ± 0.11.2 ± 0.1†1.2 ± 0.1‡1.2 ± 0.01‡1.1 ± 0.1‡1.3 ± 0.1
    MENT group (nmol/L)1.4 ± 0.11.2 ± 0.2‡1.2 ± 0.1‡1.3 ± 0.1
Cholesterol      
    T group (nmol/L)4.8 ± 0.24.4 ± 0.24.6 ± 0.24.5 ± 0.24.7 ± 0.25.0 ± 0.2
    MENT group (nmol/L)4.9 ± 0.24.6 ± 0.24.4 ± 0.2†4.9 ± 0.2
LDL-C      
    T group (nmol/L)2.8 ± 0.22.6 ± 0.22.9 ± 0.22.9 ± 0.23.0 ± 0.23.1 ± 0.2
    MENT group (nmol/L)2.9 ± 0.22.9 ± 0.22.5 ± 0.22.9 ± 0.2
Triglycerides      
    T group (nmol/L)1.3 ± 0.21.4 ± 0.21.3 ± 0.31.1 ± 0.11.5 ± 0.21.4 ± 0.2
    MENT group (nmol/L)1.5 ± 0.21.3 ± 0.21.7 ± 0.21.5 ± 0.2

There were significant falls in HDL-C concentrations in both groups during treatment, with return to pretreatment values in the recovery period (Table 2). In both groups HDL-C was significantly lower than pretreatment at all time points during treatment. Cholesterol concentrations in the MENT but not the testosterone group also showed a fall during treatment (Table 2). There were no significant changes in triglyceride or low density lipoprotein-cholesterol (LDL-C) concentrations in either group. Neither group demonstrated any significant changes in any biochemical variables throughout the study.

Blood Pressure

A significant rise in systolic blood pressure was observed in the MENT group throughout the treatment period (P = .02; Figure 3) with no significant change in diastolic blood pressure. Systolic blood pressure was not significantly different than pretreatment during the recovery phase. There were no significant changes observed in the testosterone group.

Figure 3.

. Systolic and diastolic blood pressure during treatment and recovery phases. 7α-Methyl-19-nortestosterone (MENT) group (open squares), n = 13; testosterone group (filled circles), n = 19; mean ± SEM. The MENT group received drug treatment for 24 weeks and the testosterone group for 48 weeks.

Testes and Prostate

Both treatment groups showed decreases in testicular volume during the study period. This fall was only transient in the MENT group. In the testosterone group mean testis volume fell to a nadir at 48 weeks (P < .001).

There was no significant change in prostate volume over the course of the study in the MENT group (Table 3). However, in the testosterone group a small but statistically significant increase in prostate volume was seen at the end of the treatment period (P = .007), and it remained slightly elevated after 16 weeks in the recovery phase (P < .05 vs pretreatment, ns vs week 48). Serum prostate-specific antigen (PSA) concentration demonstrated a significant fall in the MENT group (Table 3) at both 12 and 24 weeks of treatment but was unchanged in the testosterone group.

Table 3. . Weight, prostate volume and prostate-specific antigen (PSA), testis volume, sexual activity, and bone density data in the 2 treatment groups during 7α-methyl-19-nortestosterone (MENT)/testosterone plus etonogestrel treatment*
 Pretreatment12 Weeks24 Weeks36 Weeks48 WeeksRecovery
  1. *Mean ± SEM. MENT group, n = 13; testosterone (T) group, n = 16. BMD indicates bone mineral density

  2. P < .05

  3. P < .01 vs pretreatment. There were no significant differences between the MENT and testosterone groups in any of these variables

Weight      
    T group, kg80.8 ± 3.880.6 ± 3.482.2 ± 3.583.6 ± 4.184.2 ± 3.9†84.1 ± 3.8
    MENT group, kg80.5 ± 3.083.0 ± 3.681.5 ± 3.781.3 ± 3.3
Prostate volume      
    T group (mL)14.7 ± 1.316.2 ± 0.917.4 ± 0.9‡16.9 ± 1.2
    MENT group (mL)16.3 ± 1.415.8 ± 2.316.4 ± 1.4
PSA      
    T group (ng/mL)0.8 ± 0.10.87 ± 0.10.84 ± 0.10.73 ± 0.10.72 ± 0.10.68 ± 0.1
    MENT group (ng/mL)1.1 ± 0.20.77 ± 0.1‡0.74 ± 0.1‡0.80 ± 0.1
Testis volume      
    T group (mL)18.6 ± 0.913.7 ± 0.9‡12.4 ± 0.6‡12.6 ± 1.2‡11.5 ± 1.1‡19.0 ± 0.9
    MENT group (mL)18.1 ± 0.912.9 ± 1.0‡15.4 ± 1.118.6 ± 1.1
Sexual activity      
    T group7.1 ± 0.98.7 ± 1.57.8 ± 0.97.6 ± 1.87.4 ± 1.97.1 ± 1.2
    MENT group6.9 ± 0.95.9 ± 1.36.4 ± 1.37.9 ± 2.1
    T group (g/m2)1.03 ± 0.021.03 ± 0.031.08 ± 0.021.01 ± 0.06
    MENT group (g/m2)1.05 ± 0.041.04 ± 0.041.01 ± 0.04
Total hip BMD      
    T group (g/m2)1.08 ± 0.021.08 ± 0.021.10 ± 0.031.08 ± 0.03
    MENT group (g/m2)1.01 ± 0.051.01 ± 0.051.01 ± 0.05

Body Composition, Sexual Behavior, and Bone Mineral Density

There was no change in weight in the MENT group but an increase in the testosterone group (P = .02; Table 3). There were no significant changes in sexual activity or bone mineral density in either group (Table 3).

Discussion

These data further confirm that administration of sex steroids can suppress spermatogenesis sufficiently and reversibly to allow development as a hormonal contraceptive for men (Anderson and Baird, 2002; Nieschlag et al, 2003; Kamischke and Nieschlag, 2004). What is less clear is how to optimize the rate, extent, and interindividual consistency of spermatogenic suppression while minimizing the potentially adverse nonreproductive effects of treatment, which largely reflect administration of supraphysiologic doses of testosterone or other androgens. Administration of a progestogen allows considerable dose sparing of testosterone, and studies using long-duration formulations suggest that this may also allow a reduction in dose while maintaining efficacy (Handelsman et al, 1996; Anderson et al, 2002; Brady et al, 2004).

The synthetic androgen MENT may offer advantages over testosterone for replacement therapy both in hypogonadal men and as a component of a hormonal male contraceptive (Anderson et al, 1999; von Eckardstein et al, 2003). However, for the latter, the dose of MENT required was associated with undesirable nonreproductive effects indicating supraphysiologic androgen action (ie, increases in hemoglobin and hematocrit) (Noé et al, 1999; von Eckardstein et al, 2003), as with testosterone alone. We have previously demonstrated that 2 MENT implants formulated as in the present study appeared to provide replacement in hypogonadal men (Anderson et al, 2003). Similarly, the dose of etonogestrel used here when combined with replacement doses of testosterone resulted in marked but not maximal suppression of spermatogenesis (Anderson et al, 2002). We therefore investigated this dose of etonogestrel with MENT in comparison with a slightly larger dose of testosterone.

There was rapid spermatogenic suppression during the initial weeks of treatment, with approximately 80% of both groups achieving less than 1 × 106/mL at 12 weeks. This is comparable to the most effective regimens previously reported (Kamischke et al, 2002; Turner et al, 2004), indicating the potential value of MENT in male contraception. Thereafter, however, suppression in the MENT group was inconsistent and some subjects complained of reduced interest in sex. Serum MENT concentrations were similar to those previously reported using this implant formulation (Anderson et al, 2003), but determination of MENT in the implant after removal confirmed that the release rate was low. It therefore appears that a greater sustained release rate than achieved here is necessary for continuing spermatogenic suppression and support of sexual interest in healthy men.

The inconsistent suppression of spermatogenesis in the MENT group was accompanied by an increase in FSH with no significant change in LH or testosterone concentrations during the second half of the treatment period. Measurement of intratesticular testosterone concentrations and of the specific testicular androgen epitestosterone has shown that testicular steroidogenesis is incompletely suppressed during administration of male contraceptive regimens (McLachlan et al, 2002; McLachlan et al, 2004; Walton et al, 2006). It appears that the rise in FSH seen here with low but detectable LH concentrations is sufficient to support spermatogenesis in some men, emphasizing the need for maximal suppression of both gonadotropins and thus depletion of intratesticular testosterone for optimal spermatogenic suppression.

The differential effects on spermatogenic suppression were also evident in the divergent patterns of inhibin B, which fell only transiently in the MENT group. A more marked fall was demonstrated in the testosterone group, as previously reported (Anderson et al, 1998; Brady et al, 2004). However, a striking further decline during the early stages of recovery was detected. The basis for this is unclear, although it accounts for the overshoot in FSH during the recovery phase. It is possible that the rapid restoration of LH and thus intratesticular testosterone and of FSH delays reinitiation of spermatogenesis in some men, as demonstrated in rats following radiotherapy (Meistrich et al, 1997; Shetty et al, 2006).

The men who withdrew from the study because of reduced sexual interest did so after only 8 weeks treatment; thus, even the peak serum MENT concentrations were insufficient in these healthy men. This contrasts with other evidence that the MENT dose was not insufficient, because hemoglobin concentration was increased and HDL-C was reduced at 12 weeks, whereas in the testosterone group both these were unchanged with no evidence of inadequate behavioral support. This discrepancy may indicate that a relatively higher dose of MENT is required for behavioral support in healthy men compared with hypogonadal men (Anderson et al, 1999, 2003) and also compared with trophic effects on the bone marrow and liver, which may reflect its restricted metabolism.

All men in the testosterone group eventually became azoospermic, with similar suppression to the same dose of etonogestrel with a lower dose of testosterone (Anderson et al, 2002). This contrasts with improved suppression using the lower dose of testosterone but higher dose of etonogestrel (Brady et al, 2004). In these earlier studies there was no evidence of inadequate androgen replacement, and there were only minimal nonreproductive side effects. Together these data illustrate dosage effects for both the testosterone and progestogen component and demonstrate the importance of the progestogen dose over the testosterone dose in maximizing spermatogenic suppression. Disadvantageous effects of this increased dose of testosterone (600 mg per 12 weeks) were exemplified by the increase in hemoglobin and decrease in HDL-C concentrations despite average testosterone concentrations being similar to pretreatment over the second 24 weeks of treatment, which were not seen in our previous studies using a lower dose of testosterone (400 mg per 12 weeks). These results are in keeping with the physiologic replacement dose of testosterone being nearer 5 than 7 mg per day (Walton et al, 2006) and highlight the importance of measuring the response of androgen-dependent variables such as hemoglobin concentration and hematocrit to determine the optimum replacement dose rather than solely serum testosterone, which will be overestimated before treatment if only morning sampling is used (Anderson and Baird, 2002).

The major potential advantage of MENT is that it is resistant to 5α-reduction and may therefore relatively spare the prostate. The present data also support this, with a fall in serum PSA with a nonsignificant fall in prostate volume. These results are similar to those in similarly aged hypogonadal men (Anderson et al, 2003). This interpretation must be tempered, however, by the above reservations regarding the overall adequacy of this dose of MENT.

Maintenance of bone mass in men is dependent on serum testosterone and local conversion to estradiol (Vanderschueren et al, 2004). MENT is also a substrate for aromatase, and in an aged orchidectomized rat model MENT was effective in maintaining bone mass (Venken et al, 2005). In hypogonadal men, 2 MENT implants did not appear sufficient to maintain lumbar spine bone mass (Anderson et al, 2003). The present data do not show any loss of bone mass in the MENT group over this relatively short duration of treatment although it is possible that with longer duration of treatment bone mass may not be adequately supported. This may reflect differences between healthy and hypogonadal men or the additional administration of the progestogen.

There was a small but significant elevation of systolic blood pressure in the MENT group with no change in the testosterone group. A similar finding was reported in a previous study investigating MENT alone (von Eckardstein et al, 2003). Systolic blood pressure may reflect arterial stiffness and is increasingly recognized to be a strong cardiovascular risk factor (Oliver and Webb, 2003). Arterial stiffness is inversely related to testosterone concentrations in older men (Hougaku et al, 2006) and was increased by induced hypogonadism in men with prostate cancer (Smith et al, 2001). During MENT administration both testosterone and estradiol concentrations are low. Recent data suggest that endogenous estradiol may be vasculoprotective (Arnlov et al, 2006). Although MENT is aromatized to an active estrogen (LaMorte et al, 1994), the low serum concentrations both of MENT and therefore of potential active metabolites may contribute to increased arterial stiffness and a rise in systolic blood pressure, which therefore might actually be less affected by administration of a higher, more effective dose.

In conclusion, this study demonstrates that the combination of MENT and etonogestrel results in effective spermatogenic suppression. Formulation of MENT to give consistent release at a dose similar to that in the initial weeks of this study may be a promising approach for hormonal male contraception although nonreproductive effects were detected. The testosterone group also showed rapid spermatogenic suppression, and it is now clear that it is of greater value to raise the dose of progestogen than that of the testosterone component to optimize suppression and minimize nonreproductive androgenic effects.

Acknowledgement

We are grateful to Nick Malone for assistance with subject recruitment and management, Ian Swanston for hormone analysis, and Nigel Groome for supplying the inhibin B assay. We also gratefully acknowledge support from Organon for the provision of Implanon and testosterone pellets.

Footnotes

  1. This study was supported by a grant from the Medical Research Council and Department for International Development (G9523250).

Ancillary