Steroid hormones for contraception in men

Male hormonal contraception has been an elusive goal. Administration of sex steroids to men can shut off sperm production through effects on the pituitary and hypothalamus. However, this approach also decreases production of testosterone, so "add-back" therapy is needed.

To summarize all randomized controlled trials of male hormonal contraception.

We searched the computerized databases Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, Popline, and LILACS (each from inception to February, 2003) for randomized controlled trials of hormonal contraception in men. We wrote to authors of identified trials to seek unpublished or published trials that we might have missed.

We included all randomized controlled trials in any language that compared a steroid hormone with another contraceptive. We excluded non-steroidal male contraceptives, such as gossypol. We included both placebo and active-regimen control groups. All trials identified included only healthy men with normal semen analyses.

Azoospermia (absence of spermatozoa on semen examination) was the primary outcome measure. Data were insufficient to examine pregnancy rates and side effects.

The proportion of men who achieved azoospermia varied widely in reports to date. Few significant differences emerged from these trials. Levonorgestrel implants combined with injectable testosterone enanthate (100 mg IM) was significantly more effective than was levonorgestrel 125 mcg by mouth daily combined with testosterone patches (10 mg/d) (OR for azoospermia with the oral levonorgestrel regimen 0.03; 95%CI 0.00-0.29). The addition of levonorgestrel 500 mcg by mouth daily improved the effectiveness of testosterone enanthate 100 mg IM weekly by itself (OR for azoospermia with the combined regimen 4.0; 95%CI 1.00-15.99). Several regimens, including testosterone alone and GnRH agonists and antagonists, had disappointing results.

No male hormonal contraceptive is ready for clinical use. All trials published to date have been small exploratory studies. As a result, their power to detect important differences has been limited and their results imprecise. In addition, the definition of oligospermia has been imprecise or inconsistent in many reports. To avoid bias, future trials need more attention to the methodological requirements for randomized controlled trials. Trials with adequate power would also be helpful.