Adrenal androgen measurement for assessing the selectivity of adrenal venous sampling in primary aldosteronism

Adrenal venous sampling (AVS) is the reference standard for primary aldosteronism (PA) subtyping. Cortisol is widely used to assess the success of selective AVS, but it is not always reliable. The aim of the present study was to investigate the usefulness of adrenal androgens, compared to cortisol, in assessing the selectivity of AVS. We consecutively recruited 37 patients with PA undergoing AVS. AVS procedures were performed with the bilateral simultaneous technique without cosyntropin stimulation. We collected two baseline blood samples from 10 patients with a 15-min interval between time-15 (t-15) and time 0 (t0) to measure the variability in the levels of the tested hormones over time. Cortisol, androstenedione, dehydroepiandrosterone (DHEA), and DHEA sulfate (DHEAS) concentrations were measured in adrenal and peripheral venous (AV and PV, respectively) samples. The selectivity index (SI) values for androstenedione and DHEA were on average 3.0- and 2.0-fold higher, respectively, than those for cortisol, respectively (P < 0.05). However, the SI for DHEAS was 5.0-fold lower than the SI for cortisol (P < 0.05). Plasma androstenedione and DHEA concentrations positively correlated with cortisol and aldosterone concentrations in AV samples (P < 0.01). Compared to cortisol, the variation ratio of AV androstenedione and DHEA decreased from t-15 to t0 (0.23 and 0.43 vs. 0.52; P = 0.001 and P = 0.061, respectively). Using receiver operating characteristic curves, an SI ≥ 3 for androstenedione or DHEA provided optimal sensitivity and specificity in AVS. Given the much larger AV/PV ratios and reduced variability compared to cortisol, the adrenal androgens androstenedione and DHEA are useful for assessing the selectivity of AVS without cosyntropin stimulation and may be superior analytes in conditions with marked variability of cortisol levels or with adrenocortical tumors cosecreting cortisol and aldosterone.