Utility of the post-cosyntropin urinary metabolomic profile for the subtype diagnosis of primary aldosteronism

Objective: To analyze if differences between patients with unilateral (UPA) and bilateral primary aldosteronism (BPA) in baseline and cosyntropin-stimulated urinary steroid metabolomic profile (USMP), might serve to identify patients in whom adrenal venous sampling (AVS) could be avoided because of a low probability of UPA. Methods: A prospective, multicenter study of patients consecutively diagnosed with primary aldosteronism with a known source of aldosterone based on AVS and/or surgical outcomes. Thirty-three urinary steroid metabolite levels were measured before and after 250 μg intravenous cosyntropin. Results: Thirty-one patients with UPA and 20 with BPA were enrolled. Patients with UPA had a shorter duration of hypertension ( P = 0.048) and a tendency towards a higher prevalence of hypokalemia ( P = 0.056). The absolute increase of cortisol after cosyntropin infusion ( P = 0.130) tended to be lower in UPA cases than in those with BPA. In the post-cosyntropin USMP, cortisone levels were lower in UPA than BPA (median: 289 vs. 395 μg/24 h, P = 0.034). The predictive model that combines post- cosyntropin 24 h urinary cortisone levels, hypokalemia, and Δserum cortisol post- cosyntropin provided a diagnostic accuracy of 73% for differentiating UPA and BPA [Area under the ROC curve (AUC) 0.728, 95% confidence interval (CI) 0.565-0.840]. The probability of UPA was low (17.3%) in patients with normokalemia, post-cosyntropin cortisone levels greater than 384 μg/24 h, and serum cortisol greater than 14 μg/dl. Conclusion: The probability of UPA is lower than 18% in patients with normokalemia, post-cosyntropin 24 h urinary cortisone greater than 384 μg/24 h and serum cortisol increase with cosyntropin greater than 14 μg/dl. Thus, AVS might be avoided in these patients, and targeted medical therapy can be initiated with no further procedures.