Effect of collection timing on selectivity index in unstimulated adrenal venous sampling: discovery and validation

This study is to determine peripheral sample collection timing impact on selectivity index (SI) and success of unstimulated adrenal venous sampling (AVS) for primary aldosteronism (PA) subtyping. In this study, a retrospective discovery and a prospective validation arm were conducted. 74 patients undergoing AVS before and after-ACTH stimulation were reviewed. Discovery dataset was divided into 1) "pre" group, peripherals collected 30-min before AVS and 2) "post" group, peripherals collected 1-min after AVS. SIs were calculated using 30-min-pre and 1-min-post values. Patients with samples having SIs < 2 and SIs ≥ 5 on before and after-ACTH were classified as false negative and those with SI ≥ 2 and SI ≥ 5 on before and after-ACTH as true positive. Data was analyzed using Chi-squared test. For validation, 27 patients were enrolled prospectively as a paired group. In each, two peripherals were collected approximately 30 min before and 1 min after AVS. Cortisol was compared using Wilcoxon matched-pair signed rank test. Retrospectively, 38% of "30 min-pre" patients had SI < 2 in right adrenals, compared to 14% in "1 min-post" patients (P = 0.007). For the left, 45.9% of "30 min pre" patients had SIs < 2 compared to 13.5% in "1 min-post" patients, (P = 0.002). Prospectively, peripheral cortisol 1 min post decreased by 28% compared to 30-min pre (median, 6.4 to 4.6 μmol/L; P < 0.001). SIs increased 40% bilaterally (P < 0.001). In conclusion, in unstimulated AVS, collecting peripheral samples after sampling the adrenal veins is more likely to give more accurate SI than before sampling the adrenal veins.