This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Anderson, R. A. Articles by Baird, D. T. Search for Related Content PubMed PubMed Citation Articles by Anderson, R. A. Articles by Baird, D. T. Pubmed/NCBI databases Medline Plus Health Information Birth Control

Male Contraception

Medical Research Council Human Reproductive Sciences Unit (R.A.A.) and Contraceptive Development Network (D.T.B.), Centre for Reproductive Biology, University of Edinburgh, Edinburgh, Scotland EH16 4SB, United Kingdom

Correspondence: Address all correspondence and requests for reprints to: D. T. Baird, Center for Reproductive Biology, The University of Edinburgh Chancellor’s Building, 49 Little France Crescent, Old Dalkeith Road, Edinburgh EH16 4SB, United Kingdom. E-mail: dtbaird{at}ed.ac.uk

	Abstract

Top Abstract I. Introduction II. The Control of... III. Currently Available Male... IV. The Hormonal Approach... V. Nonhormonal Testicular and... VI. Acceptability of Male... VII. Conclusions References

 
The provision of safe, effective contraception has been revolutionized in the past 40 yr following the development of synthetic steroids and the demonstration that administration of combinations of sex steroids can be used to suppress ovulation and, subsequently, other reproductive functions. This review addresses the current standing of male contraception, long the poor relation in family planning but currently enjoying a resurgence in both scientific and political interest as it is recognized that men have a larger role to play in the regulation of fertility, whether seen in geopolitical or individual terms. Condoms and vasectomy continue to be popular at particular phases of the reproductive lifespan and in certain cultures. Although not perfect contraceptives, condoms have the additional advantage of offering protection from sexually transmitted infection. The hormonal approach may have acquired the critical mass needed to make the transition from academic research to pharmaceutical development. Greatly increased understanding of male reproductive function, partly stimulated by interest in ageing and the potential benefits of androgen replacement, is opening up other avenues for investigation taking advantage of nonhormonal regulatory pathways specific to spermatogenesis and the reproductive tract.

I. Introduction

II. The Control of Testicular and Epididymal Function

A. The regulation of gonadotropin secretion

B. Testosterone production

C. The regulation of spermatogenesis

D. The epididymis

E. Conclusion

III. Currently Available Male Contraceptive Methods

A. Condoms

B. Vasectomy

IV. The Hormonal Approach to Male Contraception

A. Testosterone alone: demonstration of contraceptive efficacy

B. Testosterone-progestogen combinations

C. Estrogens in male contraception

D. Testosterone with GnRH analogs

E. Adverse effects and long-term considerations

F. Future directions in androgen delivery

V. Nonhormonal Testicular and Posttesticular Agents

A. Immunological approach to male contraception

VI. Acceptability of Male Contraception

VII. Conclusions

	I. Introduction

Top Abstract I. Introduction II. The Control of... III. Currently Available Male... IV. The Hormonal Approach... V. Nonhormonal Testicular and... VI. Acceptability of Male... VII. Conclusions References

 
ACCESS TO EFFECTIVE contraception is a prerequisite of reproductive health (1). If the goal of ensuring that every birth results from a planned pregnancy is to be achieved, a wide range of methods of regulating fertility must be available. Because women literally are left "holding the baby," family-planning organizations have traditionally concentrated on female methods. New developments in the last 10 yr, including new formulations of the oral contraceptive pill, medicated intrauterine devices, and subdermal implants, have provided for women a wide range of contraceptive choice (Fig. 1). In contrast, advances in male-directed methods have been confined to refinements in the type of condom and technique of vasectomy. It has been argued that research on new male-directed methods is unnecessary and that resources would be better directed toward making existing methods more widely available (2). Yet despite their limitations, up to 30% of couples worldwide use a male method of contraception (Fig. 2). Moreover, recent research has demonstrated that, in many societies, men are prepared to share the responsibility of contraception more equally with their partners (3). An individual’s requirements for contraception differ depending on their changing social circumstances. It is likely that the method of contraception that meets the requirements of an adolescent in an early exploratory relationship will differ radically from that which is suitable for a stable couple that has completed its family. Thus, the development of new, effective methods of male contraception has been identified as a high priority by international organizations including the World Health Organization (WHO; Refs. 4 and 5).

View larger version (22K): [in this window] [in a new window]   Figure 1. Contraceptive prevalence by regions of the world. Data are percentages of married women currently using modern methods of contraception (male and female sterilization, intrauterine contraceptive device, the pill, injectables, hormonal implants, condoms, and female barrier methods) and vary from 8% in Western Africa to 80% in Eastern Asia, with narrower variation in the more developed regions. Modern methods account for approximately 90% of contraceptive usage in less developed areas but for 70% in more developed areas, with traditional methods, e.g., withdrawal and calendar rhythm methods, which require male involvement, accounting for 26% in developed regions compared with 8% in less developed areas. Data are from the United Nations Population Fund (323 ).

View larger version (31K): [in this window] [in a new window]   Figure 2. Distribution of contraceptive usage by method in developed vs. less-developed regions of the world. Seventy percent of users in the more developed areas rely on short-acting, reversible methods (condoms, pills, traditional methods), whereas in the less-developed areas, 70% use longer-acting, clinic-based methods (injectables, sterilization, intauterine contraceptive device). Data are from the United Nations Population Fund (319 ).

View larger version (38K): [in this window] [in a new window]   Figure 3. Potential targets for contraception in the male. The endocrine control of spermatogenesis is indicated, with contraceptive approaches in boxed text.

	II. The Control of Testicular and Epididymal Function

Top Abstract I. Introduction II. The Control of... III. Currently Available Male... IV. The Hormonal Approach... V. Nonhormonal Testicular and... VI. Acceptability of Male... VII. Conclusions References

 
This article is primarily concerned with the regulation of male fertility. Although some discussion of the physiological basis for male fertility is therefore required, the following section does not attempt to be comprehensive. The effects of abnormalities in gonadotropic regulation of testicular function have been recently reviewed (9), as has the regulation of FSH secretion (10). Focus will be on those aspects of male reproductive function that have been or may become targets for contraceptive action.

A. The regulation of gonadotropin secretion
The testis, like the ovary, has both endocrine and gametogenic functions and is totally dependent on pituitary gonadotropins. Gonadotropin secretion is under the overall stimulatory control of GnRH, with inhibitory inputs consisting of steroidal and peptide hormone feedback from the testes and local regulatory factors. Mechanisms and pathways involved in the regulation of GnRH secretion and action are thus central to the hormonal approach to male contraception, both by the use of GnRH analogs, in particular antagonists, and by the use of steroids to override physiological feedback signals. The pattern of both LH and FSH secretion in the peripheral circulation is pulsatile (11, 12) and, in experimental animals, secretion of LH has been demonstrated to directly parallel that of GnRH into portal blood (13, 14). The pulsatile nature of GnRH secretion is believed to be crucial for the maintenance of gonadotroph responsiveness (15), preventing receptor down-regulation and subsequent fall in LH secretion. However, the absence of the COOH-terminal tail in the mammalian GnRH receptor, compared with other species, results in a much slower rate of receptor internalization, which may be of importance in the design of novel ligands (16).

Although testosterone is the major steroid secreted by the testis, it has been long recognized that other steroids may be involved in the regulation of gonadotropin secretion (17). Such steroids, such as estradiol, might be secreted directly by the testis (18) or produced by conversion from testosterone in extraglandular tissues. Testosterone inhibits LH secretion by acting at both the hypothalamus and anterior pituitary gland (19, 20, 21, 22, 23, 24). Administration of dihydrotestosterone (DHT) has been widely used to investigate direct androgenic effects. High doses have generally been reported to result in suppression of LH concentrations (19, 25), whereas administration of more physiologically appropriate doses had little or no effect (17, 26, 27). The absence of an important effect of physiologically relevant amounts of endogenous DHT is suggested by the lack of effect of the 5-reductase (5R) inhibitor finasteride (28).

There is clear evidence that other steroids, including estradiol, are important in the regulation of gonadotropin secretion (17, 20). Estradiol infusion reduced LH secretion in response to pulsatile GnRH in men with idiopathic hypogonadotropic hypogonadism, whereas administration of testolactone, an aromatase inhibitor, caused both an increase in LH secretion when administered alone in this model and a reduction of the inhibitory effect of testosterone (26, 29). The relative contributions of androgen and estrogen have been recently reinvestigated by comparison of the effects of aromatase inhibition and of biochemical castration by administration of ketoconazole, inducing a fall in both testosterone and estradiol concentrations (30). These results indicate differential regulation of the two gonadotropins by testosterone and estradiol, with the effect of testosterone on FSH being largely mediated by aromatization.

Despite the use of progestogens to inhibit gonadotropin secretion in male contraceptive studies for several decades (31), surprisingly few studies have directly investigated the effects of these drugs in men. Progesterone receptors are present in the hypothalamus and anterior pituitary of male rats and rams (32, 33). An inhibitory effect of progesterone in castrated rams has recently been demonstrated when given with a low dose of testosterone, whereas progesterone given alone was ineffective (33). Thus, it is possible that the effect of progesterone on serum LH requires the presence of testosterone and/or estradiol, analogous to the effects of estradiol in the female. Increased LH concentrations in male progesterone receptor knockout mice suggest a possible physiological role (34). In normal men, the gestagen desogestrel caused a fall in LH, FSH, and testosterone concentrations over a 3-wk administration (35), with testosterone concentrations falling to approximately 35% of pretreatment values with 300 µg desogestrel. We have recently demonstrated that progesterone administration to normal men reduces both LH pulse frequency and amplitude and also reduces FSH secretion (36). These effects were similar to those observed with desogestrel administration, indicating that the effects of progestogens on gonadotropin secretion are not solely mediated by the androgen receptor but are, at least in part, mediated by the progesterone receptor, with evidence for both hypothalamic and pituitary sites of action.

In addition to steroidal feedback control of gonadotropin secretion, there is an important nonsteroidal gonadal contribution. This was recognized and given the name "inhibin" by McCullagh (37), who observed the inhibitory effect of an aqueous testicular extract on the formation of castrate cells in the pituitary gland. After the purification of inhibin from follicular fluid (38, 39) and the development of immunoassays specific for the dimeric forms of inhibin A and B, the presence of inhibin B, but not A, in the male was demonstrated. Blood concentrations are lower in men with testicular disorders, and an inverse relationship between inhibin B and FSH concentrations was confirmed (40, 41). This relationship between inhibin B and FSH was also observed across the physiological range in normal men (42). These data strongly suggest that inhibin B is an important component of the afferent arm of the feedback loop from the testis, selectively regulating FSH secretion (10).

B. Testosterone production
Testosterone is produced by the Leydig cells of the testis, under the stimulatory control of LH. The biosynthesis and metabolism of testosterone have been recently reviewed (43). Because the hormonal approach to male contraception involves administration of testosterone, the question of the appropriate dose arises. Some androgen-dependent functions, such as sexual behavior, are normalized with subphysiological testosterone concentrations and do not increase with supraphysiological doses (44, 45), whereas muscle mass and hemoglobin concentration continue to increase with increasing testosterone concentrations and high-density lipoprotein (HDL)-C continues to fall (46). Currently available preparations are limited, particularly the availability of long-acting formulations (Table 1). Most current regimens involve coadministration of a second agent such as a gestagen or GnRH analog to induce gonadotropin suppression, with replacement of testosterone to augment the suppressive effect on gonadotropin secretion and prevent hypogonadism. The dose required should therefore approximate physiological replacement. Derivation from measurement of the metabolic clearance rate of radiolabeled testosterone gives a production rate of 6–7 mg/d (47, 48), although a recent reinvestigation using stable isotope dilution indicated rather lower production rates of 3.7 ± 2.2 mg/d (49). In hypogonadal men, standard regimens include 250 mg testosterone esters every 3 wk, i.e., approximately 8 mg testosterone per day, but with a large differential across the injection interval. Testosterone pellet and transdermal regimens are similar at approximately 5 mg/d (50, 51), with the pellets proving relatively stable concentrations, whereas the patches may mimic the physiological diurnal variation in testosterone production (52), the significance of which is unknown. These doses normalize prostate volume and maintain bone mass in hypogonadal men (53, 54); thus, approximately 5 mg/d appears appropriate for physiological replacement in contraceptive regimens.

There remain uncertainties regarding the importance of FSH in the maintenance of adult spermatogenesis (10). Inactivating mutations of the FSH ß-subunit and FSH receptor have been identified in men (9) and in knockout models developed in mice (57, 58). Men with inactivating FSH receptor mutations showed qualitatively normal spermatogenesis and were in some cases fertile (59), whereas men with FSHß mutations had a more marked phenotype and were azoospermic (60, 61), although there may have been coexisting abnormalities of Leydig cell function. The mouse knockout models for both the FSH ß-subunit and receptor also showed qualitatively normal spermatogenesis (58, 62), whereas in the LH receptor knockout mouse, spermatogenesis was arrested at the round spermatid stage (63). In men who were administered supraphysiological doses of testosterone resulting in suppression of gonadotropins and spermatogenesis to azoospermia, subsequent administration of FSH resulted in a resumption of spermatogenesis (64). Conversely, administration of human chorionic gonadotropin or LH to men during testosterone-induced suppression (i.e., a model of selective FSH suppression) also resulted in reinitiation of spermatogenesis to sperm concentrations within the normal range, although lower than pretreatment concentrations for those men, despite FSH concentrations remaining suppressed (65, 66). Overall, these data indicate that FSH is not absolutely required for adult spermatogenesis, but it is required for quantitatively normal spermatogenesis to proceed. The testosterone regimen used in that and related studies investigating the effect of LH/human chorionic gonadotropin administration only induces azoospermia in a modest majority of Caucasian men (67), possibly related to the high testosterone concentrations achieved, which may directly support spermatogenesis in some men (65, 68). Whether FSH would induce spermatogenesis in a model with more complete intratesticular testosterone deprivation is uncertain. Although data from nonhuman primates support the existence of an important role for FSH in the regulation of spermatogonial replication in the presence of well-maintained intratesticular testosterone concentrations (69), it is unlikely that a contraceptive approach based on selective FSH withdrawal would be successful. The present availability of recombinant gonadotropin preparations and potent GnRH antagonists opens this area to further detailed study. The recent development of a transgenic model based on the gonadotropin-deficient hypogonadal (hpg) mouse with transgenic FSH expression provides an opportunity to study the effects of FSH in isolation from LH (70): in the absence of FSH, the germinal epithelium is disorganized with no tubular lumen and no spermatogonial progression beyond the pachytene stage, whereas in the presence of FSH such mice show near-complete spermatogenesis with a large number of round spermatocytes but few elongate spermatids, indicating lack of completion of spermiogenesis. This is similar to the testicular phenotype of the LH receptor knockout mouse (63), indicating the importance of FSH in the completion of germ cell meiosis.

A related finding is the effect of contraceptive steroid administration on inhibin B concentrations, as inhibin B is a Sertoli cell product reflecting both Sertoli cell number and the resident population of germ cells (71). Although initial studies using either testosterone alone (42) or testosterone with levonorgestrel (LNG; Ref. 41) demonstrated a fall in inhibin B concentrations, other studies using a range of testosterone/progestogen combinations have demonstrated that inhibin B concentrations can be maintained even in the face of induced azoospermia (72, 73). Despite FSH concentrations being generally suppressed to the limit of detection by such regimens, these data appear to suggest that Sertoli cell function is largely maintained. The predominant site of spermatogenic arrest, itself variable under these conditions (74), is at a relatively late stage of spermatogenesis. This has been recently investigated using stereological analysis of testicular biopsies from men receiving testosterone alone or with depot medroxyprogesterone acetate (DMPA) (75). Two major effects were noted, inhibition of maturation of Apale to B spermatogonia and a striking inhibition of spermiation, of which the latter is probably a mechanism for the early fall in sperm concentrations in clinical studies. Maturing germ cells were however present at approximately 20% of normal in treated men, indicating that qualitatively normal spermatogenesis continues in the testis despite most men having sperm concentrations of less than 0.1 x 10⁶/ml in the ejaculate, with gonadotropins suppressed to minimal concentrations.

The concentration of testosterone within the testis is undoubtedly much higher than in the circulation, although the quantification of intratesticular testosterone concentrations has been the subject of considerable debate (76). The functional significance of such high concentrations remains unclear, as the androgen receptor present in the testis is believed to be identical with that expressed throughout the body (77). Of relevance to the present discussion, however, is the repeated observation that Leydig cells continue to produce some testosterone after hypophysectomy (78, 79). Although testosterone concentrations in the interstitial fluid are less than 5% of normal under those conditions, this remains similar to physiological concentrations in peripheral blood. Contraceptive approaches whose mechanism is predominantly through gonadotropin suppression are therefore unlikely to reduce intratesticular concentrations much lower than this. Such low concentrations may support spermatogenesis in some men, as antiandrogen administration results in a further decline in spermatogenesis (80). There are very limited human data on intratesticular testosterone concentrations under such circumstances, but weekly administration of testosterone propionate caused a fall in intratesticular testosterone concentrations to only double the total testosterone concentration in plasma (81). Similar data have recently been obtained in men undergoing testicular biopsy after treatment with testosterone enanthate (200 mg/wk) alone or with DMPA (75). Testicular testosterone concentrations declined to 2% of normal after 6 wk of treatment and were then similar to circulating concentrations of total testosterone. Because of the fall in sex hormone-binding globulin concentrations resulting from testosterone administration, the rise in free testosterone is considerably greater than that of total testosterone (68). The potential for a reverse gradient between the peripheral circulation and the testis therefore exists when high peripheral concentrations are achieved by exogenous steroid administration in the presence of suppressed LH secretion. An alternative approach to this question is the measurement of epitestosterone (EpiT; 17-hydroxyandrost-4-en-3-one), a natural epimer of testosterone secreted predominantly by the testis (82, 83). Its excretion is suppressed by exogenous testosterone administration (84, 85) to approximately 10% of normal. EpiT remains detectable in all men during testosterone treatment, with concentrations severalfold higher than those found in hypogonadal men (83). These data suggest that that there remains a low rate of steroidogenesis within the testis during testosterone treatment, consistent with data from hypophysectomized rats (78, 79).

Analysis of the effects and mechanism of action of testosterone within the testis is further complicated by the potential for conversion to other steroids, particularly DHT and estradiol. The ability of rat testicular tissue to convert testosterone to DHT in vitro has long been recognized (86, 87) and is differentially distributed with greater activity in the seminiferous tubules than in the interstitium (86). 5R activity has also been demonstrated in the human testis (88, 89), and there appeared to be an increase in activity at the expected time of puberty (90). The presence of mRNA for both isoenzymes of 5R has been demonstrated in the human male reproductive tract (91, 92). Although mRNA and enzyme activity levels were very low in human testis, the presence of enzyme activity at pH 5.0, but not 7.0, is consistent with the presence of the type 2 enzyme (91). Conversely, the type 1 isoenzyme is the predominant isoform in the rat testis (92, 93). There may therefore be significant species difference in the testicular expression of 5R isoenzymes. DHT can quantitatively support spermatogenesis in rats at lower doses than are required with testosterone (94, 95). Under physiological conditions, i.e., in the presence of an apparent vast excess of testosterone within the testis, what the potential role of the amplification of testosterone action by conversion to DHT might be is unclear. However, when intratesticular testosterone concentrations are low, such as during gonadotropin suppression, the amplification of androgen signaling may become of importance in supporting spermatogenesis. This has been investigated using the 5R inhibitor L675-272 (80, 96) in rats after induction of hypogonadotrophism by administration of testosterone and estradiol implants. In that experimental model, testosterone administration results in dose-dependent stimulation of spermatogenesis. Coadministration of the 5R inhibitor resulted in a reduction of the ability of lower testosterone doses to restore spermatogenesis. Detailed morphological analysis demonstrated a reduction in the progression of round spermatids through midspermiogenesis as well as in the number of elongate spermatids produced (80). The two isoforms of 5R also appear to be differentially regulated in the testis, with 5R1 negatively regulated by testosterone and 5R2 positively regulated by FSH (97). Although data regarding 5R isoforms in the testis are not available for the human, it has been demonstrated that intratesticular DHT concentrations do not fall after gonadotropin suppression in men (75). This further illustrates the potential importance of 5R1 in conditions of reduced intratesticular testosterone concentrations, although attempts to exploit this using the 5R inhibitor finasteride have been unsuccessful (98, 99).

The second major metabolic pathway of testosterone metabolism is conversion to estradiol by the enzyme aromatase. Aromatase activity is high in the Sertoli cells of the immature testis but decreases thereafter (100) and is also present in both Leydig and germ cells (101, 102). Human testicular venous blood contains higher estradiol concentrations than are found in the peripheral circulation (18). The study of the role of estradiol in the testis has received a considerable stimulus by the identification of a second estrogen receptor, ERß, and the demonstration of the expression of ERß receptors by many cell types, including germ cells, within the male reproductive tract (reviewed in Ref. 102). Analysis of the effects of ER knockout mice gave rise to the novel finding of a major physiological role for estrogen/ER in the regulation of epididymal fluid transport, whereas ERß knockout mice have apparently normal reproductive function. Analysis of the physiological role of ERß is complicated by the recent identification of several variants of the receptor, some of which may not be active as transcription activators and are expressed by human testicular germ cells (103). Further study of ERß receptor function may give rise to novel insights on the steroidal control of spermatogenesis.

D. The epididymis
The essential role of the epididymis is the maturation of spermatozoa, including the capacity for motility and fertilization. Although spermatozoa can be used for in vitro fertilization after surgical recovery before passage through the epididymis, fertilization rates are low without intracytoplasmic sperm injection. The epididymis secretes proteins that modulate spermatozoal function (104) and has absorptive and secretory functions. The major problem with the consideration of the epididymis in this regard is the identification of cellular processes specific to the epididymis that could be used as potential targets without toxicity in other organs. Advances in molecular biology and proteomics are likely to improve identification of epididymis-specific regulatory pathways (105, 106, 107).

E. Conclusion
In summary, the adult testis is controlled by a feedback system involving the hypothalamus, anterior pituitary, and testis. The main testicular components of the feedback loop involve testosterone, inhibin B, and estradiol, secreted directly or arising by aromatization of testosterone in peripheral tissues. Testosterone and possibly DHT have important effects on spermatogenesis directly within the testis as well as systemic effects maintaining libido and sexual function. Testosterone suppresses LH secretion by acting directly at both the hypothalamus and anterior pituitary. FSH secretion is mainly controlled by the action of inhibin and estradiol. Progesterone and synthetic gestagens suppress the concentrations of FSH and LH, although the role, if any, of progesterone in the physiological regulation of gonadotropin secretion in men is unknown. Because spermatogenesis is dependent on gonadotropins, an obvious contraceptive approach is suppression with exogenous steroids.

	III. Currently Available Male Contraceptive Methods

Top Abstract I. Introduction II. The Control of... III. Currently Available Male... IV. The Hormonal Approach... V. Nonhormonal Testicular and... VI. Acceptability of Male... VII. Conclusions References

 
A. Condoms
Condoms have been in use since antiquity. Their initial use was predominantly to provide some protection from sexually transmitted disease (STD), an issue that has come full circle with the emergence of HIV. Condom usage is the only method, other than lifelong mutual monogamy, that can reduce the risk of HIV infection and other STDs, with other contraceptive methods, particularly female hormonal methods, possibly increasing susceptibility to STD acquisition (108, 109, 110). In the United States, one in five adults has an STD, and many go untreated; thus, approximately 15 million new sexually transmitted infections occur annually in the United States (111). The great majority of condoms are made from latex rubber and undergo testing for water leakage, tensile strength, and longevity. Possible improvements to these tests include testing with viral particles, which have demonstrated the potential for viral penetration in approximately 2% of condoms (112). However, the volume of semen contamination from such holes is very low, orders of magnitude lower than for not using the condom and probably of little significance compared with slippage and breakage.

The effectiveness of condoms is influenced by the nature of the product; variation in use of the product by individual users, including variation between users, as to risk of pregnancy or STD acquisition; and characteristics of the population being studied, including background prevalence of STDs. There are two pertinent aspects of the nature of condoms: 1) by providing a physical barrier to semen, the effectiveness of condoms will depend on the proportion of acts of intercourse during which they are used (i.e., correctly and consistently); and 2) they are prone to physical complications, particularly breakage and slippage. The term "efficacy" is used to denote the protection afforded by usage under ideal conditions, and "effectiveness"’ is the term used to describe the protection afforded under real conditions. Effectiveness therefore includes the contribution of the user as well as that of the device. The great majority of studies that have investigated the effectiveness of condoms have been observational in design with inherent risk of confounding bias. Another major source of bias is the necessity for reliance on self-report for much of the information to be gathered, including the occurrence of slippage and breakage, although more objective tests are being developed (113, 114, 115). These factors have all contributed to uncertainties in establishing whether condom usage really provides protection against STDs.

1. Prevalence of condom usage.
The usage of all methods of contraception is limited by availability. This in itself is not only a societal issue but is relevant at the level of the individual. Thus, many aspects of contraceptive use and failure are partly determined by relative poverty and its associated barriers (116), in addition to inherent method differences. Condoms are very widely used by men at some point in their lives, with up to 90% of respondents in a survey across different cultures reporting usage of the condom at a rate higher than that for any other method (3). There have also been large increases in the use of condoms over the last two decades; these increases are associated with increased awareness of HIV and public information campaigns (117, 118). Usage by women in the United States aged 15–44 yr increased from 12% in 1982 to 20% in 1995 (119), with higher usage in the young and unmarried. Reported usage of condoms by men aged 15–19 in all acts of intercourse also increased from 33% in 1988 to 45% in 1995, with a halving to 9.5% in the proportion of men reporting never having used condoms (120). The prevalence of HIV infection in some high-risk groups, however, remains very high (121), resulting in concern that the perceived risk of HIV has diminished, perhaps as a result of the increasing effectiveness of antiretroviral therapy.

Recent data from the United Kingdom indicate that the proportion of family-planning clinic attenders using condoms rose from 6% to 35% over the yr 1975 to 2000–2001, whereas the proportion using the combined contraceptive pill fell from 70% to 42%. Condoms were the most widely used method of contraception by partners of girls under age 16 yr, with more than 50% of those attending family-planning clinics using this method. The proportion using the pill was, however, greater than that using condoms in all other than the youngest age group and was highest in the 20- to 24-yr age group (122).

Usage and attitudes about condoms vary greatly among different societies. As part of a recent survey of attitudes regarding male contraception, condom usage was found to differ by more than 3-fold among men in Cape Town and in Hong Kong. Men in Hong Kong were found the most likely to currently use condoms (62% of subjects; Ref. 3) and, compared with men in other locations, appeared to rate the convenience of condoms highly while being least likely to think that they provided effective protection against pregnancy; they were also the least enthusiastic about novel male methods.

2. Protection from pregnancy.
Estimates of the pregnancy rate during condom usage vary greatly according to the population studied. With near-perfect use, pregnancy rates as low as 3% have been reported, although national data probably more closely reflecting typical use show a rate of 14% in the first year (116). Condom usage is generally highest in the young, which, by their high fecundity, accentuates the problems of the learning curve: the pregnancy rate is approximately 50% lower in the second year of use than in the first (123). This is paralleled in the data for slippage and breakage, which are also related to user experience and knowledge (124). Data from recent prospective studies (125, 126, 127) indicate that slippage occurs on 0.6–1.3% of occasions and breakage on 0.4–2.3% of occasions. Although these data may be taken to reflect the overall difficulties with the method (i.e., the relative ease of misuse and nonuse), they also show that very good contraceptive protection can be obtained when condoms are used consistently and correctly (128).

3. Protection from STDs.
STDs include bacterial, viral, and parasitic infections, may be ulcerative or nonulcerative discharge diseases, and may be clinically overt or asymptomatic. A large number of these and other microbiological, physiological, and behavioral factors contribute to the risk of disease acquisition, including the coexistence of other STDs (129) and, for the female partner, the hormonal status of the vaginal and cervical epithelium. These and the methodological issues mentioned previously complicate interpretation of the available data. However, the results of meta-analysis of the 12 studies, which were regarded as sufficiently informative, of the potential protective effect of condom usage on HIV transmission clearly show a reduction in risk of infection with condom usage of approximately 87% (130). Among those reporting consistent usage of the condom, HIV incidence was 0.9 per 100 person-years, compared with 6.8 per 100 person-years for male-to-female transmission and 5.9 per 100 person-years for female-to-male transmission in those who reported never using condoms.

The data on protection against other STDs are more limited, and a recent workshop concluded that the data showing a protective effect was only consistent for HIV for both men and women and for gonorrhea for men (131). The available epidemiological studies pertaining to other infections were either inconsistent or regarded as methodologically flawed; thus, it was considered impossible to give an accurate estimate of the degree of potential protection offered by condom usage. However, the panel also concluded that there was strong laboratory-based evidence for protection against gonorrhea for women and against chlamydia and trichomoniasis, and that condom use might reduce the risk of human papillomavirus-associated diseases including genital warts and cervical neoplasia. Since that report, additional prospective studies have indicated protective effects of condoms against transmission of herpes simplex virus type 2 from men to women but not from women to men (132) and against transmission of several STDs in a study of Kenyan prostitutes (110).

4. New developments in barrier male contraceptives.
Developments in condom manufacture include the use of polyurethane, styrene ethylene butylene styrene (125, 133), and hypoallergenic latex (134). These are useful for those with latex allergy, and although most men found the polyurethane condom gave increased sensitivity, it had higher slippage and breakage rates (relative risks of 6.0 and 6.6, respectively) than latex (125). In a randomized trial comparing latex with two new materials, two thirds of both male and female participants preferred one of the synthetic condoms (133). No information is available regarding protection from STDs, and there are only limited data on protection from pregnancy (126).

A separate line of development is in the microbicidal coating of condoms. At present, many condoms are coated with lubricant containing the nonionic detergent nonoxynol-9 (N9). This was originally introduced as a spermicide, but it became apparent that this and similar compounds had antiinfective (including antiviral) activity by disrupting cell membranes (135). N9 decreased the rate of simian immunodeficiency virus transmission to monkeys (136) and has also been suggested to reduce HIV transmission in women in one epidemiological study (137). However, by virtue of the same cell membrane activity responsible for its antimicrobial activity, it also has irritant activity on the epithelia of the penis and vagina with changes in approximately 50% of women administered N9 suppositories 4 times per day for 2 wk (138). It has been suggested that the use of N9 and related compounds might actually increase the risk of HIV infection; evidence to support this was found in a study of female sex workers in Kenya who used vaginal sponges containing N9 (139), but additional large studies are required for confirmation. There is, however, probably a dose threshold below which N9 has no significant detrimental effect, and there is no evidence that this effect is relevant to the doses involved in condom use. A recent randomized, controlled study comparing condom use with and without N9 in more than 1000 women at high risk for STD (but excluding sex workers) showed no difference in the rate of urogenital gonorrhea and chlamydial infection (140). There is currently much interest in the development of combined spermicides and microbicides that may come to have a major role in both contraception and prevention of STDs. Although these are generally considered in terms of female application, they may also be used in condom lubricants.

B. Vasectomy
Division and/or occlusion of the vas deferens (vasectomy) is a highly effective method of contraception that has been shown to be extremely cheap and cost effective (141). Between 40 and 60 million couples (about 7%) in the world depend on vasectomy as their method of contraception (Ref. 142 and Fig. 2). Although it is usually regarded as permanent, the pregnancy rate reported after reversal by trained surgeons using microsurgical techniques is as high as 50% (143). The international incidence of vasectomy varies greatly and within each country by ethnic origin, socioeconomic status, age, and marital status. For example, in United States, the incidence of vasectomy rises from 1% in men aged 20–24 yr to 20% in men aged over 40 yr (144), although it has been becoming relatively less popular than female sterilization over the last four decades (145). In Great Britain in 1992, nearly 30% of couples over 35 yr old were using vasectomy as their contraceptive method compared with approximately 20% choosing female sterilization (146). Vasectomy was found to be more popular in men who are better educated, more affluent, and who are currently married. Cultural factors also influence the popularity of vasectomy: in Europe, less than 1% of French men are vasectomized (147), whereas it is particularly common in New Zealand (148). Thus, it would appear that cultural and socioeconomic factors are more important than concerns about safety and efficacy in determining the popularity of vasectomy.

Vasectomy almost always involves occlusion and/or division of the vas under local anesthesia (149). The vas can be accessed either by traditional surgical incision or by the "no-scalpel" technique using a specially designed sharp, pointed forceps (150). Occlusion of the ends of the tube by cautery, sclerosing agent, or interposition of fascia are more effective than simple division and ligation. A number of modifications of the technique, including possible reversible methods, have been investigated, including insertion of rubber plugs into the vas or injection of styrene polymer, the latter having the advantage of administration by injection (151). Assessment of efficacy is not easy. Short-term failure is usually defined as the presence of sperm in the ejaculate at some arbitrary time after operation (3–6 months) or after 23–25 ejaculates (141). There are always some sperm present in the initial ejaculates, although after 4 wk the number and quality in the majority of men is probably insufficient to achieve fertilization. In practice, a significant proportion of men fail to provide postvasectomy ejaculates for examination, and failures are identified only after an unexpected pregnancy in the partner. Late failure can occur at any time after vasectomy and is thought to be due to recanalization of the vas. The pregnancy rate in partners whose fertility status is unknown is only an indirect measure of efficacy. The apparent failure rate is inversely related to the age of the partner because of the marked decline in fertility of older women. In one study, the cumulative pregnancy rate after 10 yr was only 1.9 per 100 when the woman was over 40 yr at the time of vasectomy, as compared with 12 in 100 cases when the wife was 25–29 yr (152). Common sense dictates that vasectomy should be more successful when performed by experienced surgeons using a technique that involves occlusion and division of the vas. Case series by individual surgeons report failure rates of 0–2%. In one prospective series, 2250 men were followed up for at least 1 yr after at least two postvasectomy samples had no sperm (153). In the first year, 15 men had sperm in the ejaculate, 4 in the second, and only 1 in the third. All the men had sperm counts less than 0.1 million/ml, and there were no pregnancies reported. The same clinic reported only 9 failures that resulted in pregnancy in more than 30,000 vasectomies performed between 1970 and 1999. The low failure rate (1 in 2000) is probably an underestimate due to underreporting. However, there are no published prospective studies equivalent to those for female sterilization [e.g., Collaborative Review of Sterilization (CREST); Ref. 154 ] in which the incidence of failure can be assessed in relation to these factors. It is likely that the effectiveness, as judged by the number of unplanned pregnancies, is lower than the results of these individual series would suggest.

Vasectomy under local anesthesia has few serious side effects. Perioperative complications include bleeding and hematoma, the prevalence of which is related to the experience of the surgeon and the type of procedure (155). Incisional vasectomy is associated with a higher complication rate than no-scalpel techniques (156). Although sperm granulomas at the site of occlusion occur in 15–40% of vasectomies, they are usually asymptomatic. Although long-term complications such as chronic pain and epididymitis are rare (157), it is important that prevasectomy counseling includes information about these risks.

The majority of men develop antisperm antibodies that persist in the circulation for several years (158, 159) and may lead to continuing infertility even when the patency of the vas has been reestablished by surgical reversal (160). The interval between vasectomy and reversal is also an important predictor of success, as are the techniques of both surgical procedures. Initial reports of a high incidence of atherosclerosis in monkeys after vasectomy have not been confirmed (161, 162). Epidemiological surveillance fails to demonstrate any increase in cardiovascular disease in vasectomized men (163). Case control and cohort studies investigating the incidence of carcinoma of the prostate and testis in vasectomized men have given conflicting results (141, 164, 165). Interpretation of these studies is complicated by the difficulty of removing confounding factors, especially detection bias. A recent review (165) concluded overall that there was no association between vasectomy and cancer of the prostate or testis.

In summary, vasectomy is one of two existing methods of contraception available to men. It is more effective than the condom and has the advantage that it does not rely on compliance at the time of coitus to be effective. However, it provides no protection against STD, and reversal is expensive and only partially successful. The fact that men in many countries choose vasectomy reflects their commitment to sharing the burden of fertility control with their partners and is an indication of the potential demand for new methods of contraception.

	IV. The Hormonal Approach to Male Contraception

Top Abstract I. Introduction II. The Control of... III. Currently Available Male... IV. The Hormonal Approach... V. Nonhormonal Testicular and... VI. Acceptability of Male... VII. Conclusions References

 
Whatever the precise mechanisms of actions of the two gonadotropins and their relative importance in maintaining spermatogenesis in the adult may be, it is clear that suppression of gonadotropin secretion will result in a fall in sperm production, which is the basis of the hormonal approach to male contraception. Suppression of testicular steroidogenesis is therefore also a consequence of endocrinological male contraception, requiring coadministration of androgen to prevent the symptoms and consequences of hypogonadism. Suppression of gonadotropin secretion can be achieved by overriding the physiological negative feedback control mechanisms at the hypothalamus and pituitary gland by administration of exogenous steroids, or perhaps more directly, the effect of GnRH on the pituitary may be prevented by administration of a GnRH analog or by a combination of such agents. The requirement for androgen to provide replacement for the secondary hypogonadism will also provide a physiological feedback signal at the hypothalamus, preventing increased GnRH secretion, which may increase the effectiveness of a coadministered GnRH analog. The major issues are the need for rapid, consistent, and sustained suppression of spermatogenesis to a level that will give adequate contraceptive efficacy, potential adverse effects of administered steroids or other agents, and the need for appropriately acceptable drug formulations. Conversely, there is the potential for noncontraceptive health benefits as well as risks from such alterations in the hormonal milieu, as with the female combined contraceptive pill.

A. Testosterone alone: demonstration of contraceptive efficacy
The potential of this concept is far from new; the demonstration that administration of testosterone resulted in suppression of spermatogenesis dates back to 60 yr ago (7, 8, 166). These initial observations and subsequent studies (Ref. 167 and Fig. 4) demonstrated that testosterone could induce fully reversible azoospermia using the short-acting ester testosterone propionate. The development of the longer-acting testosterone enanthate (TE) allowed investigation of the effect of varying dosage and administration frequency, particularly to reduce exposure to supraphysiological doses of testosterone, and introduced a distinction between frequent administration for induction followed by a reduced-frequency maintenance phase (168). These and similar studies involving varying injection intervals (169, 170) were encouraging, with near-azoospermia maintained during TE injections at 10- to 12-d intervals, although longer injection intervals resulted in partial recovery.

View larger version (11K): [in this window] [in a new window]   Figure 4. Illustration of reversible, testosterone-induced azoospermia in normal men, whom in this study were treated with 25 mg testosterone propionate im daily, as indicated. [Derived from Ref. 167 .]

The relationship between spermatogenic suppression and contraceptive efficacy raises the issue of the degree of efficacy required from a hormonal method. All existing methods of contraception have a failure rate, and although some recently introduced methods are more effective than sterilization, the range is wide. Condoms have an important place in contraceptive practice despite their evident shortcomings. Thus, it cannot be assumed that a male contraceptive that does not offer near-100% contraceptive efficacy will have no place in global provision. However, the skepticism still associated with the introduction of a hormonal male method (2) is such that the first method to be introduced should be as efficacious as possible.

These WHO studies also identified significant ethnic differences in the proportion of men who achieved azoospermia, being greater in Chinese (91%) than Caucasian (60%) men (171). High azoospermia rates in Asian men have been confirmed in subsequent studies using both testosterone alone (175, 176) and androgen/progestogen combinations (177). The basis for this is uncertain and could not be explained on the basis of body size or pretreatment endocrine or seminal differences in data from the WHO study (178). Overall, men achieving azoospermia had slightly higher pretreatment FSH concentrations and showed a gonadotropin rebound in the recovery phase (178).

On the basis of investigation of men taking part in the second WHO study (172), it had been suggested that the maintenance of oligozoospermia was associated with higher 5R activity than in men achieving azoospermia (48, 179). As discussed above, inhibition of 5R activity in the rodent testis has been demonstrated to reduce the supportive effect of low-dose testosterone on spermatogenesis (80, 96), and intratesticular DHT concentrations are maintained during gonadotropin suppression (75). Differences in 5R activity between Caucasian and Chinese men and women have been identified (180) in parallel to the greater suppression of spermatogenesis in Asian men. The possible importance of 5R in supporting low rates of spermatogenesis in some men has been investigated in two studies involving administration of the 5R inhibitor finasteride. One protocol involved selective administration of finasteride to men once incomplete suppression had been identified (98), and the second protocol involved administration to men from the initiation of suppression using a submaximal regimen of testosterone with the oral progestogen desogestrel (99). In both studies, no additional effect of finasteride was demonstrated. However, because it is likely that 5R type 1 is the predominant isoform in the testis and finasteride has preferential activity on the type 2 isoform, it is possible that more selective inhibition of the type 1 isoform may be more effective.

Other reproductive differences between Caucasian and Asian men include differences in feedback sensitivity to testosterone (181) and in rates of germ cell apoptosis (182). A recent analysis of differences in androgen production and metabolism between Caucasian and Chinese men concluded that, although there were differences, they were largely due to dietary/environmental rather than genetic factors (183). Although there are several possible contributory factors to the increased sensitivity of Asian men to steroidal suppression of spermatogenesis, none have been clearly demonstrated to be of direct importance.

These two WHO studies (171, 172) demonstrated conclusively that hormonal suppression of spermatogenesis sufficient for contraceptive efficacy was possible. The main drawbacks of the TE preparation used were the need for frequent injection; the relatively high proportion of men not achieving azoospermia and, thus, the need for identifying nonresponders; and side effects due to the high dose of testosterone administered. The influence of the testosterone formulation has been demonstrated using testosterone pellets. These pellets consist of fused crystalline testosterone, which is usually inserted sc using a trocar, with the patient under local anesthesia, into the lower abdominal wall at a dose of 800 mg (4 x 200 mg pellets, each releasing 1.3 mg/d), providing hypogonadal replacement for approximately 5 months with good acceptability and reproducibility (184). Although the pellets may occasionally be extruded, their long duration of action with near-zero-order pharmacokinetics make them an excellent prototype preparation for long-acting injectable preparations with similar properties that are yet to be developed. Their use allows considerable reduction in the dose of testosterone required for gonadotropin and spermatogenic suppression (98, 185), resulting in a reduced prevalence of side effects while maintaining similar efficacy to TE. The absence of supraphysiological testosterone concentrations, while also avoiding diurnal fluctuations, may also contribute to spermatogenic suppression (68). We have recently investigated the effect of repeated administration of testosterone pellets in combination with a gestagen; the findings in these studies are discussed in Section IV.B.

A variety of other testosterone preparations, particularly longer-acting injections and, more recently, transdermal formulations, have also been investigated. The data from investigations in which these preparations have been in combination with other suppressive agents are discussed below. Injectable testosterone preparations are 17ß-esters, slowly absorbed from the site of injection then rapidly hydrolyzed in the circulation (186), with the speed of absorption being related to the length and hydrophobicity of the side chain (187).

The longest-acting ester tested as a potential contraceptive thus far is testosterone buciclate, which has a duration of action of 3–4 months (188). Single administration of 1200 mg to a small group of men resulted in encouraging spermatogenic suppression, with three men becoming azoospermic (189), but this androgen has not been available for further investigation due to difficulties with formulation and potential toxicity. More information is available regarding injectable preparations of testosterone undecanoate (TU), different preparations of which have been developed in China and Europe. Pharmacokinetic analysis of the Chinese preparation, in which the TU is dissolved in tea seed oil, indicated that administration of 500 mg provides replacement for 6–8 wk (190). A subsequent study investigating contraceptive potential demonstrated that 500 mg every 4 wk resulted in azoospermia in 11 of 12 men, and a higher dose of 1000 mg per 4 wk resulted in azoospermia in all 12 men, although there was some accumulation of testosterone (176). A large efficacy study using this preparation involving 308 men demonstrated that only 9 failed to suppress spermatogenesis to a sperm concentration of 3 x 10⁶/ml or lower within 6 months, and that there were no pregnancies among partners of men whose sperm concentration remained below that threshold for an additional 6 months (191). As with the WHO studies, most contraceptive exposure was with azoospermia, and pregnancies were reported in partners of the small number of men whose sperm concentration showed a partial recovery to above 3 x 10⁶/ml. The 8-ml injection volume appeared to be well tolerated: although it is reduced in the preparation developed by the pharmaceutical company Jenapharm (Jena, Germany) in which the drug is dissolved in caster oil with improved pharmacokinetics (192), the diluent volume required for 1000 mg TU remains 4 ml. This preparation has also been investigated as a potential contraceptive: 1000 mg TU administered every 6 wk resulted in azoospermia in 7 of 14 men (193). It appears from studies of repeated administration in hypogonadal men (194) that this injection interval can be significantly increased, and this is therefore a promising development in testosterone therapy with contraceptive potential, particularly when combined with another gonadotropin-suppressing agent.

Transdermal testosterone preparations including scrotal and nonscrotal patches and gels are becoming widely available (Refs. 51, 195, 196 and Table 1). Because their main value is in maintaining physiological testosterone concentrations, data so far on their potential in contraception have been as components of a combination preparation (72, 197, 198), in which the degree of spermatogenic suppression achieved has been surprisingly low. This may, in part, reflect poor compliance because the transdermal patches currently available are irritant to the skin (199, 200).

In summary, the studies described illustrate the development from proof of concept that testosterone-based regimens could provide efficacious contraception in at least a proportion of men to newer preparations with improved pharmacokinetics allowing for significant dosage reductions. However, the long-term side effects (and low efficacy in Caucasian populations) associated with this approach make it unlikely that testosterone alone will be developed as a male contraceptive.

B. Testosterone-progestogen combinations
The ability of progesterone and synthetic progestogens to suppress spermatogenesis has long been recognized, with azoospermia being achieved in all 19 men administered 50 mg progesterone im or 30 mg daily of one of three synthetic progestogens administered by mouth (31). The administration of the progestogens alone resulted in loss of libido; thus, these potent suppressors of gonadotropin secretion have been widely investigated in combination with testosterone to allow a reduction in testosterone dosage while potentially augmenting the degree of spermatogenic suppression (Fig. 5). In addition to suppression of gonadotropin secretion, progestogens may have additional intratesticular effects. Inhibitory effects on Leydig cell steroidogenesis, specifically on 17ß-hydroxysteroid dehydrogenase activity, have been identified (201), as well as inhibitory effects on LH receptor expression and function (202), possibly mediated by a nonclassical progesterone receptor also identified in spermatozoa (203). Both effects might be important in a gonadotropin-depleted state, by further reducing Leydig cell steroidogenesis and, thus, intratesticular testosterone concentrations.

View larger version (28K): [in this window] [in a new window]   Figure 5. Comparison of spermatogenic suppressive efficacy in a number of recent studies of testosterone-only and combination regimens. Data indicate percentages of subjects achieving azoospermia (solid columns) and oligozoospermia (open/shaded columns). Open columns indicate threshold of less than 3 x 106/ml, and shading indicates less than 1 x 106/ml. Numbers within base of columns indicate numbers of subjects. T, Testosterone (precise preparation specified at base of figure); DSG, desogestrel (oral); ETO, etonogestrel (implant). Horizontal lines indicate groups of columns to which the text refers. Numbers immediately below columns indicate doses (see text for details). Reference numbers are indicated at the base of the figure below the appropriate columns.

MPA can also be administered orally and has been investigated in combination with early percutaneous testosterone preparations (215, 216). Although effective suppression of spermatogenesis could be achieved (216), absorption of the testosterone by the female partners was troublesome (217). This would not be anticipated using current transdermal testosterone preparations, although their use results in a very high incidence of allergic reaction (199), which can be severe (200). Transdermal testosterone with oral LNG or desogestrel induced dose-dependent suppression of spermatogenesis (72, 197), although this was not complete, and significant numbers of subjects withdrew because of skin irritation.

2. Cyproterone acetate (CPA).
CPA is an orally active antiandrogen and progestogen that is widely used in the treatment of hirsutism in women and prostate cancer in men. Its potential use in male contraception is therefore perhaps surprising, but some very informative data have been obtained. Initial studies demonstrated that high doses (200 mg/d) resulted in marked suppression of spermatogenesis, with five of six men reaching sperm concentrations less than 1 x 10⁶/ml, two of whom became azoospermic (218). Gonadotropin secretion in that study was estimated using urinary analysis. The lack of change in gonadotropin excretion was interpreted to indicate a direct effect within the testis, and although subsequent studies of the effects of lower doses (5–30 mg daily; Refs. 219 and 220) using more accurate gonadotropin assays demonstrated that gonadotropin secretion was indeed suppressed, the possibility that intratesticular effects contribute to the efficacy of CPA remains.

The high prevalence of side effects of even the low doses of CPA (219, 220, 221), including loss of libido and fatigue, precluded further studies without androgen supplementation. A series of studies have more recently been carried out reinvestigating oral CPA with TE (222, 223). These studies, although involving small numbers of men in each group, suggested that the combination of CPA in doses of 25–100 mg/d with 100 mg TE per week resulted in the rapid onset of azoospermia in all subjects, whereas that dose of TE alone was less effective. The antiandrogenic effect of CPA appeared to be reflected in a dose-dependent fall in hemoglobin concentration and hematocrit and also in body weight, despite the mildly supraphysiological dose of testosterone. CPA with oral TU offers the potential for complete self-administration; suppression of spermatogenesis was achieved in all eight subjects, and one became azoospermic despite the relatively low dose of 12.5 mg CPA (224). Although this approach remains attractive, it is likely to require novel oral androgens to become a more realistic method.

It is possible that the high efficacy of CPA may reflect antagonism of the effect of residual testosterone concentrations within the testis. The available data on residual androgen concentrations in the human testis during gonadotropin withdrawal suggest that intratesticular testosterone concentrations are similar to physiological peripheral concentrations during testosterone administration (75, 81), and measurement of the testicular steroid EpiT indicates that there continues to be a low rate of testicular steroidogenesis (84, 85). The potential enhancement of spermatogenic suppression with other antiandrogenic compounds as part of a contraceptive regimen therefore remains an attractive area for further investigation.

3. LNG.
LNG, as with the other progestogens previously discussed in this review, is widely used in female contraception both orally and as the Norplant device (Wyeth, Philadelphia, PA). Initial studies using LNG alone or in combination with testosterone showed only modest suppression of spermatogenesis (225, 226), with no subjects achieving azoospermia when administered 500 µg LNG orally with 200 mg TE monthly. Oral LNG at that dose does induce improved suppression with a higher dose of TE (100 mg/wk), with 12 of 18 men becoming azoospermic, compared with 6 of 18 with TE alone, with more rapid onset of azoospermia (211). Titration of the dose of LNG downward to 125 µg daily demonstrated the continuance of good, although incomplete, spermatogenic suppression but with reduced metabolic effects (227). The importance of the dose of testosterone is demonstrated by comparison of these studies involving weekly administration of TE with the earlier study (226) involving less frequent administration of TE and with a recent study of the combination of LNG with transdermal testosterone patches in which only 2 of 11 men became azoospermic (72); thus, maintenance of serum testosterone concentrations appears crucial to prevent escape of gonadotropin secretion and, thus, spermatogenesis. Conversely, when high doses of testosterone as injectable testosterone undecanoate are administered with oral LNG (228), no additive effect of the gestagen on suppression of spermatogenesis was observed, with only 50% of subjects achieving azoospermia, although suppression of serum lipoproteins was greater in the combination group.

LNG is also formulated as an implant. The potential advantage of sustained-release preparations include avoidance of reliance on the subject’s compliance, and as with the testosterone pellets (185), a dose-sparing effect may be evident. Administration as a two-rod implant, each rod containing 75 mg LNG, with TU (250 mg/month) resulted in azoospermia in 6 of 16 men in a Chinese study, but sperm concentrations remained in the normal range in 4 men (229). Higher doses of either the progestogen or androgen may improve on these results (198).

4. Desogestrel and etonogestrel.
Promising results have been obtained with the oral progestogen desogestrel (35, 73, 230, 231, 232). Desogestrel is a potent progestogen converted to the active agent, etonogestrel, by first-pass metabolism (233). In one study, eight of eight men became azoospermic with 300 µg oral desogestrel daily in combination with 50 mg im TE weekly (35). This study also demonstrated the apparent narrow dose-response relationship with this combination: either decreasing the dose of desogestrel to 150 µg or a higher dose of testosterone (100 mg/wk)