In this talk we integrate a number of discoveries made in recent years: (1) Normal androgen levels are crucial for follicle development at small growing follicle stages, with poor androgen levels causing developmental arrest and apoptosis, and in surviving follicles poor egg quality. The decisive androgen is testosterone, acting via the androgen receptor on granulosa cells by sensitizing granulosa cells to FSH. (2) Androgen production in women is shared between ovaries (theca) and adrenals (zona reticularis). Just as the hyperandrogenemia of PCOS has for decades been recognized as ovarian and/or adrenal in nature, so can hypoandrogenemia be ovarian and/or adrenal in origin; (3) Low functional ovarian reserve (LFOR), whether occurring at young ages (called premature ovarian aging, POA, or occult primary ovarian insufficiency, OPOI) or due to physiologic ovarian aging, is almost universally characterized by relative hypoandrogenemia (and therefore relatively high SHBG).
If DHEAS is low, the hypoandrogenemia is likely at least to a significant degree adrenal in nature since DHEAS is almost exclusively produced by adrenals. (4) Hypoandrogenemia leads to poor follicle growth, which leads to low granulosa cell mass, which leads to low estradiol levels, which raises FSH levels and lowers AMH, thus creating a typical LFOR phenotype, generally assumed to be ovarian in nature. (5) When investigating DHEAS levels and adrenal function via morning cortisol and ACTH levels (and ACTH stimulation tests), it, however, becomes apparent that in a quite substantial minority of cases the woman's hypoandrogenism is really primarily adrenal in nature, and in some cases even the outright consequence of clinical adrenal insufficiency (we have diagnosed primary, secondary and tertiary forms of adrenal insufficiency in women previously characterized as POA/OPOI). (6) This observation, thus, establishes a new diagnosis in reproductive medicine by defining these women's ovarian insufficiency as a "secondary" ovarian insufficiency, with insufficiency of the adrenal zona reticularis being the primary insufficiency which induces this secondary ovarian insufficiency. (7) Reaching this diagnosis is important for patient prognosis because if the observed LFOR phenotype is adrenal in origin, the patients ovarian function (theca) is often fully intact (like in PCOS, there, of course, can be combined cases). Adrenal supplementation, including with androgens, will, therefore, reestablish normal follicle development at small growing follicle stages, increase granulosa cell mass, decrease FSH and increase AMH levels, and offer such patients an excellent pregnancy prognosis. We, indeed, have seen rare individual cases where menopausal FSH levels almost normalized with appropriate androgen supplementation over 6-12 weeks.
Related to these observation is what we have come to call the "burning out" PCOS: PCOS is, of course, at young ages characterized by hyperandrogenemia. PCOS is, however, not a static condition and, as we recently discovered, is at least in non-obese PCOS patients characterized by more rapid androgen than AMH decreases with advancing age. Non-obese PCOS patients presenting for infertility care are, therefore, frequently not recognized as former PCOS patients, even though they can be recognized by still disproportionally high AMH levels (for age and in comparison to their FSH levels). These patients, not only present with relative hypoandrogenemia but their ovarian tissue receptors from younger years are actually used to abnormally high androgen levels. Their relative hypoandrogenemia is, therefore, even more pronounced than in non-PCOS patient. Women with "burning out" PCOS, therefore, are in even more dire need for androgen supplementation, and will with 6-12 weeks of appropriate supplementation greatly improve egg/embryo quality and pregnancy chances. Like all hypoandrogenic infertility patients, they should be investigates for the origin of their hypoandrogenemia to rule out an adrenal component.
We supplement hypoandrogenemic women with micronized DHEA, 25mg TID, and follow their free and total testosterone levels, DHEAS and SHBG. A small number of women do not convert DHEA to testosterone well (we recently reviewed the causes (Shohat-Tal et al. Nat Rev Endorcinol 2015 doi:10.1038/nrendo.2015.64), and they require direct testosterone supplementation.
The Center for Human Reproduction, New York, N.Y. 10021