Testosterone prevents complete suppression of spermatogenesis in the gonadotropin-releasing hormone antagonist-treated nonhuman primate (Macaca fascicularis).

We studied the effects of administration of a GnRH antagonist combined with testosterone (T) as an approach to male contraception as well as the role of intratesticular androgens in spermatogenesis using a nonhuman primate model. Three groups of five adult cynomolgus monkeys (Macaca fascicularis) received daily sc injections of 420-460 micrograms/kg GnRH antagonist ([Ac-D2Nal1,D4ClPhe2,DPal3,Arg5,DGlu6(AA), DALa10]GnRH) for a period of 15 weeks. T supplementation, commencing on the first day of GnRH antagonist administration, was provided by single im injection of 40 mg (group 2) or 200 mg (group 3) of the long-acting testosterone ester testosterone-trans-4-n-butylcyclohexancarboxylate (20-Aet-1). Serum LH bioactivity was undetectable within 1 week of GnRH antagonist administration in all monkeys. GnRH antagonist administration alone (group 1) reduced serum T levels into the castrate range. Forty milligrams of 20-Aet-1 maintained serum T levels in the upper range of normal monkeys, while 200 mg 20-Aet-1 maintained serum T levels about 1.5-fold above normal. The response to electroejaculation was fully maintained in all T-treated monkeys. Sperm counts in the ejaculates dropped to zero among group 1 animals within 7-10 weeks of GnRH antagonist administration. In groups 2 and 3 consistent azoospermia could not be induced, and the sperm counts were significantly (P less than 0.05) higher in group 3 than in group 2. Histologically, spermatogenesis in group 1 was arrested at the spermatogonial level in 75% of seminiferous tubules. In group 2, spermatogenesis proceeded to spermatocytes in 50% of tubules and to elongated spermatids in 10% of tubules, while in group 3 elongation of spermatids occurred in 75% of tubules. The mean T and dihydrotestosterone concentrations in baseline testicular biopsies (n = 15) were 43.8 +/- 6.8 (+/- SE) and 5.7 +/- 1.5 ng/g, respectively. After GnRH antagonist with or without T administration, the mean (n = 15) intratesticular T and dihydrotestosterone levels were reduced to 20.3 +/- 4.9 and 3.2 +/- 0.5 ng/g, respectively, and differed little among the three groups. No correlation, however, could be established between testicular androgen levels and spermatogenic status (P greater than 0.30) or sperm counts (P greater than 0.60). These results demonstrate that administration of a GnRH antagonist in the presence of constant serum T levels does not induce consistent azoospermia, and that the supporting effects of T on spermatogenesis cannot be explained exclusively on the basis of the testicular androgen concentrations.