Ovarian and adrenal vein steroids in patients with nonneoplastic hyperandrogenism: selective catheterization findings

Abstract

Standardized bilateral ovarian-adrenal vein catheterization was utilized to assess directly glandular steroid release in 60 androgenized women without evidence of a functional neoplasm. Testosterone (T), dihydrotestosterone (DHT), androstenedione ( Δ 4 A), dehydroepiandrosterone (DHEA), DHEA sulfate (DHEA-S), 17-hydroxyprogesterone (17-OHP), and Cortisol (F) were measured by radioimmunoassay in samples obtained from a peripheral vein and the four glandular veins (all values are given as nanograms per milliliter, mean ± standard deviation). Peripheral values were as follows: T, 0.68 ± 0.43; DHT, 0.32 ± 0.13; Δ 4 A, 2.2 ± 2.0; DHEA, 8.8 ± 8.9; DHEA-S, 3137 ± 1774; 17-OHP, 2.0 ± 3.0; and F, 216 ± 121. Peripheral elevations of at least one androgen were found in 80% of the 60 cases (T, 38%; DHT, 18%; Δ 4 A, 50%; DHEA, 45%; and DHEA-S, 37%). Ovarian-peripheral vein gradients (OPGs) and adrenal-peripheral vein gradients (APGs) served as semiquantitative estimates of glandular secretion. OPGs were as follows: T, 0.4 ± 1.1; DHT, 0.1 ± 02; Δ 4 A, 3.4 ± 7.0; DHEA, 14.6 ± 100; DHEA-S, −288 ± 523; 17-OHP, 4.5 ± 8.4; and F, −35 ± 47. APGs were as follows: T, 0.88 ± 1.3; DHT, 1.1 ± 0.9; Δ 4 A, 14.4 ± 38.4; DHEA, 327 ± 367; DHEA-S, 854 ± 1223; 17-OHP, 20.8 ± 41.3; and F, 1252 ± 2023. Excess ovarian and/or adrenal androgen output was assumed in a given individual when one or more of the respective T, DHT, Δ 4 A, DHEA, and DHEA-S gradients exceeded the upper 95% confidence limits of normal previously established in this laboratory. Combined hypersecretion (41%) was the most frequent cause; purely ovarian (27%) or adrenal overproduction (12%) were identified less often; normal glandular androgen output was found in 20% of patients. It is suggested that catheterization should be reserved for patients with peripheral T and DHEA-S levels compatible with a neoplasm, i.e., concentrations exceeding the upper 95% confidence limits of this cohort (T > 1.5 ng/ml, DHEA-S > 6700 ng/ml).