Endocrine Abstracts

Unilateral forms of primary aldosteronism (PA) should, by inference, have a contralaterally normal and therefore suppressed adrenal zona glomerulosa. However, there is no agreed definition of expected adrenal suppression. We hypothesized two biochemical definitions of adrenal suppression based upon measurements done during adrenal vein sampling. These were then applied to a PA–AVS-outcomes database to determine whether either definition proved useful for interpretation of AVS and/or prediction of hypertension resolution in operated cases. Definition 1 was the proportional suppression of the uninvolved/lowest adrenal_{aldo/cortisol}:IVC_{aldo/cortisol} ratio both pre and post cortrosyn stimulation. Definition 2 was the absolute decrease in the uninvolved adrenal_{aldo/cortisol} ratio post cortosyn. Applied to a cohort of 99 confirmed PA cases, Definition 1 proved highly predictive of lateralization when using post cortrosyn values with an ROC AUC of 0.958, P<0.0001 for adrenal_{aldo/cortisol} : IVC_{aldo/cortisol}<1.4 as the optimal criterion (sensitivity 90% and specificity 94%). Definition 2 was less predictive of lateralization unless a stringent criterion of lowest-adrenal_{aldo/cortisol} decrease to<1.0 was used, in which case specificity was 100% but sensitivity just 66%. For blood pressure outcomes in the surgical subgroup (n=52), hypertension resolution was most commonly seen among subjects with adrenal suppression by both definitions although there was significant overlap with subjects requiring ongoing medication. In conclusion, post cortrosyn suppression of the uninvolved adrenal:IVC ratio is a highly useful definition of adrenal suppression that accurately predicts unilateral PA. This may be particularly useful in a case where AVS fails to catheterize one of the adrenal veins but suppression is seen on the other side. Adrenal suppression may also predict blood pressure outcome however a much larger PA database is likely necessary to determine the relative contribution of this predictor.