The Accuracy of Simple and Adjusted Aldosterone Indices for Assessing Selectivity and Lateralization of Adrenal Vein Sampling in the Diagnosis of Primary Aldosteronism Subtypes

Abstract

ObjectiveThis study aimed to evaluate the reliability of simple and corrected aldosterone indices for assessing the selectivity and lateralization of adrenal vein sampling (AVS) in patients with primary aldosteronism.MethodsData of all consecutive patients with primary aldosteronism who underwent AVS for subtype diagnosis, followed at two Italian referral centers, were analyzed retrospectively.ResultsAVS achieved bilateral selectivity in 112/144 patients. Unilateral disease was diagnosed in 60 cases (53.6%) and idiopathic hyperaldosteronism in 52 individuals (46.4%). The aldosterone index (aldosterone ratio between an adrenal vein and the inferior vena cava) showed a high accuracy in predicting selectivity, compared to a cortisol selectivity index of 1.1, and a moderate accuracy, compared to cortisol cut-offs of 2 and 3. The simple aldosterone index showed a moderate accuracy in predicting ipsi/contralateral aldosterone hypersecretion, while lesion side- and hypokalemia-corrected aldosterone index revealed a significant improvement in predicting ipsi/contralateral disease. Moreover, the comparative aldosterone index (aldosterone ratio in the dominant vs the non-dominant adrenal vein) revealed a high accuracy in predicting unilateral primary aldosteronism. For an immediate clinical application of our results, the adjusted cut-offs were calculated, according to the Youden’s criterion and to a pre-established specificity of 90%, for all possible combinations of lesion side at imaging and presence/absence of hypokalemia.ConclusionsThis study demonstrated the diagnostic accuracy of simple and clinical-/imaging-corrected aldosterone indices for adrenal vein sampling in subtype diagnosis of primary aldosteronism and suggests the potential application of these tools to select patients for adrenalectomy when standard indices cannot be performed.