Hormonal methods are the closest to meeting the requirement for a reliable, reversible, safe and acceptable male contraceptive. Although reliability is judged by the efficacy in preventing pregnancy in fertile female partners, because a hormonal male contraceptive aims to prevent pregnancy by reversible inhibition of sperm output, the suppression of spermatogenesis constitutes a useful surrogate marker for development and evaluation of prototype male contraceptive regimens. This makes defining the degree of suppression of sperm output required a key strategic issue in developing a hormonal male contraceptive (128).
Two landmark WHO studies, the first ever male contraceptive efficacy studies, involving 671 men from 16 centres in 10 countries established the proof of principle that hormonally-induced azoospermia provides highly reliable, reversible contraception (3, 4). Among the minority (~25%) of men who remained severe oligozoospermic (0.1-3 million sperm/ml) using weekly testosterone enanthate injections, contraceptive failure rate (~8% per annum) was directly proportional to their sperm output. Hence to achieve highly effective contraception, azoospermia is analogous to anovulation as a sufficient, but not necessary, requirement. Nevertheless, reliable contraception by modern standards (8) requires uniform azoospermia as the desirable target for male contraceptive regimens (129). No regimen yet achieves this consistently in all men, although in some Asian countries (e.g. China (3), Indonesia (130, 131)) an approximation to uniform azoospermia can be achieved by a variety of regimens. A study involving 308 Chinese men in 6 centres has shown that monthly injections of testosterone undecanoate provides highly effective and reversible contraception (132). No pregnancies were recorded among men who were azoospermic or severely oligozoospermic (<3 million sperm per mL) providing a 95% upper confidence limit of pregnancy (contraceptive failure) rate of 2.5% per annum. The overall failure rate based on suppression of spermatogenesis was <4%. The prototype regimen was well tolerated apart from injection site discomfort due to large oil injection volume (4 mL) and reversible androgenic effects (acne, weight gain, hemoglobin, lipids).
Despite these promising findings, non-Chinese men require combination hormonal regimens involving a 2nd gonadotropin suppressing agent, notably progestins, together with testosterone to ensure adequate spermatogenic suppression. Proof of principle for this combination approach was provided by a depot androgen/progestin regimen that observed no pregnancies among 55 couples during 35.5 person-years of exposure (95% upper limit of failure rate ~8%) in a study with satisfactory tolerability and reversibility for a prototype regimen (133). Hence, contraceptive efficacy studies show that highly effective contraception can be achieved with suppression of sperm output to near azoospermia (<1 million per mL) (129). Ultimately, the efficacy of male contraception must be established by enumerating pregnancies prevented, and not sperm counting. The present paucity of male contraceptive efficacy studies, for which placebo controls are effectively impossible (134), makes systematic evaluations comparing practical clinical regimens a task for the future (135).
The reversibility of hormonal male contraceptive regimens is clearly established by an integrated re-analysis which pooled primary data from over 90% of all hormonal male contraceptive studies reported to show that all regimens show full reversibility within a predictable time course (136). This comprehensive review of the recovery of 1549 healthy eugonadal men, aged 18-51 years, who underwent 1283.5 man-years of treatment and 705 man-years of post-treatment recovery, showed the median times for recover to sperm densities of 10 and 20 million per mL were 2.5 months (2.4-2.7) and 3.0 months (2.9-3.1), respectively. Covariables such as age, ethnicity and hormonal or sperm output kinetics had significant but minor influence on the rate, but not the extent, of recovery.
Acceptability of a hormonal male contraceptive is high across a wide range of countries and cultures. Willingness to use a hypothetical hormonal male contraceptive averaged 55% (range 29-71%) in an extensive, population representative survey of 9342 men aged 18-50 yr from 9 countries (4 Europe, 3 South America, Indonesia and USA) with consistency across a wide range of socio-economic and cultural settings (137, 138). Similar findings are reported in a 4 country study (UK, South Africa, Hong Kong, Shanghai) with 44-83% in each centre (139) as well as 75% in Australia (140) willing to try a hormonal male contraceptive. Female partners from a variety of cultures also indicate high acceptability in a survey of 1894 women in 4 countries, among whom 40-78% would support and trust their male partners in stable relationships to use a hormonal male contraceptive (141). Corroborating the acceptability of hormonal male contraception are findings from experimental studies of prototype regimens for up to 12 months usage in which most participants confirm high levels of satisfaction and willingness to try a commercial product (142, 143). Hence, prototype hormonal methods have proven reliability and reversibility and reasonable prospects for being well accepted and safe. Although they are the most likely opportunity in the foreseeable future to develop a practical contraceptive method for men, progress depends on pharmaceutical industry development, but the commitment of drug companies continues to languish (5).
Testosterone provides both gonadotropin suppression and androgen replacement making it an obvious first choice as a single agent for a reversible hormonal male contraceptive. Although androgen-induced, reversible suppression of human spermatogenesis has long been known (144-147), systematic studies of androgens for male contraception began in the 1970's (148, 149). Feasibility and dose-finding studies (150), mostly using testosterone enanthate (TE) in an oil vehicle as a prototype, showed that weekly im injections of 100-200 mg TE induce azoospermia in most Caucasian men (151) but less frequent or lower doses fail to sustain suppression (152-155). The largest experience with an androgen alone regimen arises from the two WHO studies in which over 670 men from 16 centres in 10 countries received weekly injections of 200 mg TE. In these studies ~60% of non-Chinese and >90% of Chinese men became azoospermic and the remainder were severely oligozoospermic (3, 4). The high efficacy among Chinese men has also been replicated using monthly TU injections (132). Effective gonadotropin suppression is a prerequisite for effective testosterone-induced spermatogenic suppression in human (133, 156-160) and non-human primates (161, 162). However, the reasons for within and between population differences in susceptibility to hormonally-induced azoospermia remain largely unexplained (156). Possible factors include population differences in reproductive physiology of environmental (163, 164), genetic (165, 166) or uncertain (167, 168) origin that may lead to differences in suppressibility of circulating gonadotropins and/or depletion of intratesticular androgens. Limited invasive studies measuring intratesticular testosterone (and DHT) suggest that the degree of depletion may not predict reliably complete suppression of sperm output (169-171) but other more widely applicable, non-invasive markers of endogenous Leydig cell function such as circulating epitestosterone (172) or 17-hydroxyprogesterone (173) or non-steroidal testicular products such as INSL3 (174) may be more analytically informative as to the relative roles of gonadotropin suppression and intratesticular androgen depletion. Exogenous testosterone causes suppression of sperm output with an average of 13 weeks to reach severe oligozoospermia (<1 million per mL) or azoospermia with suppression being maintained consistently during ongoing treatment (175). Following cessation of treatment, sperm reappear within 3 months to reach sperm densities of 10 and 20 million per mL at an average of 11.5 and 13.6 weeks, respectively (175) with ultimately full recovery (136). Apart from intolerance of weekly injections, there were few discontinuations due to acne, weight gain, polycythemia or behavioral effects and these were reversible as were changes in hemoglobin, testis size and plasma urea. There was no evidence of liver, prostate or cardiovascular disorders (3, 4, 176).
The pharmacokinetics of testosterone products are crucial for suppressing sperm output. Oral androgens have major first-pass hepatic effects producing prominent route-dependent effects on hepatic protein secretion (eg SHBG, HDL cholesterol) and inconsistent bioavailability. Short-acting testosterone products requiring daily or more frequent administration (oral, transdermal patches or gels) which may be acceptable for androgen replacement therapy are not appropriate for hormonal contraception. Weekly TE injections required for maximal suppression of spermatogenesis (150) are far from ideal (177) and cause supraphysiological blood testosterone levels risking both excessive androgenic side effects and preventing maximal depletion of intratesticular testosterone for optimal efficacy (178, 179). Other currently available oil-based testosterone esters (cypionate, cyclohexane-carboxylate, propionate) are no improvement over the enanthate ester (180), and longer-acting depot preparations are needed. Subdermal testosterone pellets sustain physiological testosterone levels for 4-6 months (181) and the newer injectable preparations testosterone undecanoate (132), testosterone-loaded biodegradable microspheres (182) and testosterone buciclate (183) provide 2-3 months duration of action. Depot androgens suppress spermatogenesis faster, at lower doses and with fewer metabolic side effects than TE injections but azoospermia is still not achieved uniformly (184) although when combined with a depot progestin, this goal is achievable (133).
Oral synthetic 17-a alkylated androgens such as methyltestosterone (185), fluoxymesterone (186), methandienone (187) and danazol (188, 189) suppress spermatogenesis but azoospermia is rarely achieved and the inherent hepatotoxicity of the 17-a alkyl substitutent (190) renders them unsuitable for long-term use. Athletes self-administering supratherapeutic doses of androgens also exhibit suppression of spermatogenesis (187, 191). Synthetic androgens lacking the 17-a alkyl substituent have been little studied although injectable nandrolone esters produce azoospermia in 88% of European men (192, 193) whereas oral mesterolone is ineffective (194). On the other hand, nandrolone hexyloxyphenylpropionate alone was unable to maintain spermatogenic suppression induced by a GnRH antagonist (195) in a prototype hybrid regime (where induction and maintenance treatment differ) whereas testosterone appears more promising (196). A 7-methyl derivative of nandrolone (MENT), which is partly aromatisable but resistant to 5α reductive amplification of androgenic potency, has been studied as a non-oral androgen for hormonal male contraceptive regimens (197). While it is prostate-sparing (198), dose titration to achieve essential androgen replacement at each relevant tissue is more complex than for testosterone and may be difficult to achieve (199). More potent, synthetic androgens lacking 17-a alkyl groups (200, 201) remain to be fully evaluated.
Antiandrogens have been used to selectively inhibit epididymal and testicular effects of testosterone without impeding systemic androgenic effects (202). Cyproterone acetate, a steroidal antiandrogen with progestational activity, suppresses gonadotropin secretion without achieving azoospermia but leads to androgen deficiency when used alone (203). In contrast, pure non-steroidal antiandrogens lacking androgenic or gestagenic effects such as flutamide, nilutamide and casodex fail to suppress spermatogenesis when used alone (204, 205). Two studies evaluating the hypothesis that incomplete suppression of spermatogenesis is due to persistence of testicular DHT have reported no additional suppression from administration of finasteride, a type II 5a reductase inhibitor (206, 207); however as testes express predominantly the type I isoforms (208), further studies are required to conclusively test this hypothesis using an inhibitor of type I 5-a reductase (209).
The safety of androgen administration concerns mainly potential effects on cardiovascular and prostatic disease. As the explanation for the higher male susceptibility to cardiovascular disease is not well understood, the risks of exogenous androgens are not clear (210, 211). In clinical trials, lipid changes are minimal with depot (non-oral) hormonal regimens (133, 172, 184, 212). Changes in blood cholesterol fractions observed during high hepatic exposure to testosterone and/or progestins, due to either oral first pass effects or high parenteral doses, have unknown clinical significance but, in any case, maintenance of physiological blood testosterone concentrations is the prudent and preferred objective. The real cardiovascular risks or benefits of hormonal male contraception will require long-term surveillance of cardiovascular outcomes (213).
The long-term effects of exogenous androgens on the prostate also require monitoring since prostatic diseases are both age and androgen-dependent. Exposure to adult testosterone levels is required for prostate development and disease (214-216). The precise relationship of androgens to prostatic disease and in particular any influence of exogenous androgens remains poorly understood. There is little direct relationship between blood testosterone levels and the occurrence of prostatic disease in prospective studies of adults (217). A genetic polymorphism, the CAG (polyglutamine) triplet repeat in exon 1 of the androgen receptor, is an important determinant of prostate sensitivity to circulating testosterone with short repeat lengths leading to increased androgen sensitivity (218), however the relationship of the CAG triplet repeat length polymorphism to late-life prostate diseases remains unclear (219). Among androgen deficient men, prostate size and PSA concentrations are reduced and returned towards normal by testosterone replacement without exceeding age-matched eugonadal controls (218, 220-222). Even self-administration of massive androgen over-dosage does not increase total prostate volume or PSA in anabolic steroid abusers although central prostate zone volumes increases (223). In-situ prostate cancer is common in all populations of older men whereas rates of invasive prostate cancer differ many-fold between populations despite similar blood testosterone concentrations. This suggests that early and prolonged exposure to androgens may initiate in-situ prostate cancer but later androgen-independent environmental factors promote the outbreak of invasive prostate cancer. Therefore it is prudent to maintain physiological androgen levels with exogenous testosterone, which then might be no more hazardous than exposure to endogenous testosterone. Prolonged surveillance comparable with that for cardiovascular and breast disease in users of female hormonal contraception would be equally essential to monitor both cardiovascular and prostatic disease risk in men receiving exogenous androgens for hormonal contraception.
Extensive experience with testosterone in doses equivalent to replacement therapy in normal men indicates minimal effects on mood or behavior (3, 4, 150, 224-226). A careful placebo-controlled, cross-over study showed that a 1000 mg TU injection in healthy young men produces minor mood changes without any detectable increase in self or partner-reported aggressive, non-aggressive or sexual behaviors (227). By contrast, extreme androgen doses used experimentally in healthy men can produce idiosyncratic hypomanic reactions in a minority (228). Aberrant behaviour in observational studies of androgen-abusing athletes or prisoners are difficult to interpret particularly to distinguished genuine androgen effects from the influence of self-selection for underlying psychological morbidity (229).
Combination steroid regimens use non-androgenic steroids (estrogens, progestins) to suppress gonadotropins, in conjunction with testosterone for androgen replacement, have shown the most promising efficacy with enhanced rate and extent of spermatogenic suppression compared with androgen alone regimens (172, 230, 231). Synergistic combinations reduce the effective dose of each steroid and minimising testosterone dosage could enhance spermatogenic suppression if high blood testosterone levels counteract the necessary maximal depletion of intratesticular testosterone (232, 233) as well as reducing androgenic side-effects.
Progesterone is a key precursor and steroidogenic intermediate for all bioactive natural steroids and the progesterone receptors A and B are structurally and evolutionarily the closest members of the nuclear receptor superfamily to the androgen receptor. Yet, although progesterone has crucial gestational and lactational roles in female reproductive physiology, it has no well established role in male reproductive physiology apart from a possible role in sperm function (234), possibly via non-genomic rather than a classically genomic mechanism (235). Nevertheless functional nuclear progesterone receptors are expressed in male brain, smooth muscle and reproductive, but not most non-reproductive tissues (236). Synthetic progestins, steroidal structural agonistic analogs of progesterone, are potent inhibitors of pituitary gonadotropin secretion used widely for female contraception and hormonal treatment of disorders such as endometriosis, uterine myoma and mastalgia. Used alone, progestins suppress spermatogenesis but cause androgen deficiency including impotence (237, 238) so androgen replacement is necessary. Non-human primate studies indicate that this is mediated via a central hypothalamic-pituitary site of action rather than direct effects on the testis (239).
Extensive feasibility studies concluded that progestin-androgen combination regimens had promise as hormonal male contraceptives if more potent and durable agents were developed (150, 240). The most detailed information on androgen/progestin regimens derives from studies with medroxyprogesterone acetate (MPA) combined with testosterone. Monthly injections of both agents or daily oral progestin with dermal androgen gels produce azoospermia in ~60% of fertile men of European background with the remainder having severe oligozoospermia and impaired sperm function (150, 240, 241). Nearly uniform azoospermia is produced in men treated with depot MPA and either of two injectable androgens in Indonesian men (130, 131) or testosterone depot implants in Caucasian men (172). Smaller studies with other oral progestins such as levo-norgestrel (230, 242, 243) and norethisterone (244, 245) combined with testosterone demonstrate similar efficacy to oral MPA whereas cyproterone acetate with its additional anti-androgenic activity has higher efficacy in conjunction with TE (231, 246) but not oral testosterone undecanoate (247). Promising findings of highly effective suppression of spermatogenesis are reported with depot progestins in the form of non-biodegradable implants of norgestrel (248-250) or etonorgestrel (251, 252) or depot injectable medroxyprogesterone acetate (133, 172, 253, 254) or norethisterone enanthate (255, 256) coupled with testosterone. The pharmacokinetics of the testosterone preparation is critical to efficacy of spermatogenic suppression with long-acting depots being most effective while transdermal delivery is less effective than injectable testosterone (248). Progestin side-effects are few and sexual function is maintained by adequate androgen replacement dosage. The metabolic effects depend on specific regimen with oral administration and higher testosterone doses exhibiting more prominent hepatic effects such as lowering SHBG and HDL cholesterol. After treatment ceases with depletion or withdrawal of hormonal depots, spermatogenesis recovers completely but gradually consistent with the time-course of the spermatogenic cycle (136).
Estradiol augments testosterone-induced suppression of primate spermatogenesis (257) and fertility (258) but estrogenic side-effects (gynecomastia) and modest efficacy at tolerable doses make estradiol-based combinations impractical for male contraception (259). The efficacy and tolerability of newer estrogen analogs in combination with testosterone remain to be evaluated.
The pivotal role of GnRH in the hormonal control of testicular function makes it an attractive target for biochemical regulation of male fertility. Blockade of GnRH action by GnRH receptor blockade with synthetic analogs or GnRH immunoneutralization would eliminate LH and testosterone secretion requiring testosterone replacement. Many superactive GnRH agonists are used to induce reversible medical castration for androgen-dependent prostate cancer by causing a sustained, paradoxical inhibition of gonadotropin and testosterone secretion and spermatogenesis due to pituitary GnRH receptor downregulation. When combined with testosterone, GnRH agonists suppress spermatogenesis but rarely achieve azoospermia (232, 233, 260) being less effective than androgen/progestin regimens. By contrast, pure GnRH antagonists create and sustain immediate competitive blockade of GnRH receptors (261, 262) and, in combination with testosterone, are highly effective at suppressing spermatogenesis. Early hydrophobic GnRH antagonists were difficult to formulate and irritating, causing injection site mast cell histamine release. Newer more potent but less irritating GnRH antagonists produce rapid, reversible and complete inhibition of spermatogenesis in monkeys (263-265) and men (266, 267) when combined with testosterone. The striking superiority of GnRH antagonists over agonists may be due to more effective and immediate inhibition of gonadotropin secretion and thereby more effective depletion of intratesticular testosterone. Due to their highly specific site of action, GnRH analogs have few unexpected side-effects. Depot GnRH antagonist plus testosterone formulations suitable for administration at up to 3 month intervals could be promising as a hormonal male contraceptive regimen. Whether GnRH antagonists are more cost-effective than progestins as the second, non-androgenic component of combination male hormonal contraceptive regimens remains to be established (268) (195) (170) (254). The drawback of high cost might be overcome by hybrid regimens using GnRH antagonists to initiate suppression followed by a switch to more economical steroids for maintenance of spermatogenic suppression in humans (196). A GnRH vaccine could intercept GnRH in the pituitary-portal bloodstream preventing its reaching pituitary GnRH receptors. Gonadotropin-selective immunocastration would require androgen replacement in men (269) and pilot feasibility studies in advanced prostate cancer are underway (270) but the prospects for acceptably safe application for male contraception remain doubtful (271). By contrast there are growing applications for anti-hormonal contraceptive vaccines in control of companion (pet), agricultural, zoo, feral and wild animal populations (272, 273).
Selective FSH blockade theoretically offers the opportunity to reduce spermatogenesis without inhibiting endogenous testosterone secretion. FSH action could be abolished by selective inhibition of pituitary FSH secretion with inhibin (274) or novel steroids (275), by FSH vaccine (276) or by FSH receptor blockade with peptide antagonists (277). Although FSH was considered essential to human spermatogenesis, spermatogenesis and fertility persist in rodents (278-280) and humans (281) lacking FSH bioactivity. Even complete FSH blockade alone might produce insufficient reduction in sperm output and function required for adequate contraceptive efficacy (282). In addition to the usual safety concerns of contraceptive vaccines including autoimmune hypophysitis, orchitis or immune-complex disease, an FSH vaccine might be overcome by reflex increases in pituitary FSH secretion. Hence, FSH suppression is a necessary but not sufficient for a hormonal male contraceptive regimen.
Figure 1. Pooled summary of contraceptive efficacy from two WHO male contraceptive efficacy studies (3, 4)where contraceptive failure rate (pregnancy rate) is plotted against the current sperm concentration in the ejaculate. This illustrates a summation of all data pooled from both studies. Data comprise monthly observations of the mean sperm concentration (averaging monthly sperm counts) and whether a pregnancy occurred in that month or not. Pregnancy rate (per 100 person-years, Pearl index) on the y-axis is plotted against the cumulative sperm density (in million sperm per ml) indicating that contraceptive failure rates are proportional to sperm output. The inset is the same data re-plotted in discrete sperm concentration bands rather than cumulatively. For comparison, the average contraceptive failure rates in the first year of use (8, 10) of modern reliable contraceptive methods are indicated by diamond symbols.
Figure 2. Recovery of spermatogenesis after cessation of treatment with hormonal male contraceptive regimens modified after an integrated re-analysis of over 90% of all reported studies (136). Data is plotted as a Kaplan-Meier survival plot of the increasing proportion of men recovering to various thresholds over time since last treatment. The data of last treatment is defined as the time elapsed from the end of the last treatment cycle that is the latest date of the first missed treatment dose. The thresholds are a sperm concentration of 3, 10 or 20 million sperm per mL in the ejaculate or a return to their own pre-treatment baseline sperm concentration. The median time to achieving each threshold is tabulated together with its 95% confidence interval.
