Lowering the Adrenal:Peripheral Cortisol Ratio Cutoff for Determining Successful Adrenal Venous Sampling for Primary Aldosteronism May Improve Sensitivity While Maintaining Specificity and Accuracy

Presentation Number: SUN 513
Date of Presentation: April 2nd, 2017

Andrew Paul Demidowich*1, Naris Nilubol2, Nicket Dedhia1, Fady Hannah-Shmouni1, Charalampos Lyssikatos1, Elena Belyavskaya1, Richard Chang3 and Constantine A Stratakis1
1Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), National Institutes of Health (NIH), Bethesda, MD, 2National Cancer Institute, NIH, Bethesda, MD, 3Radiology and Imaging Sciences, National Institutes of Health, Bethesda, MD


Background: Adrenal venous sampling (AVS) is performed in the work up of primary aldosteronism (PA) to identify those patients with unilateral disease, who consequently would benefit from surgical intervention. Lateralization ratio (LR) cutoffs of 2:1 at baseline or 4:1 after cosyntropin administration suggest laterality.(1) The adrenal:peripheral (A:P) cortisol ratio is used to confirm proper cannulation of each gland; however, controversy exists over which A:P cutoff value to use at the baseline time points.(1,2)

Methods: Adult subjects referred to the National Institutes of Health Clinical Center since 2002 were included in the analysis. PA was confirmed with the intravenous saline suppression test or salt-loading test prior to AVS. During AVS, samples were routinely taken in duplicate prior to (-5min & 0min) and after (+10min & +15min) cosyntropin administration.

Results: 104 AVS procedures were performed in 91 subjects (mean age±SD: 51.9±10.4, 53% male) with 89 having at least one successful AVS. Failed AVS cannulation included poor right-sided (n=8) and left-sided (n=2) sampling. ROC analysis of baseline A:P ratios to determine cannulation success, as defined by post-cosyntropin A:P ≥ 5:1, demonstrated an AUC of 98.6% (95% CI: 96.5-100%). At 100% specificity, a baseline A:P ratio of 1.4 provided the highest sensitivity at 92.9% (95% CI: 88.4-96.1%). In contrast, a baseline A:P ratio of 2.0 provided a sensitivity of only 63.1% (95%CI: 56.0-69.9%) with the same specificity. After cosyntropin administration, excess aldosterone secretion localized to the left in 38.2% (n=34), right in 22.5% (n=20), and bilaterally in 31.5% (n=28) of the cases, with 7.9% (n=7) cases deemed “indeterminate” (LR 3.0-4.0). Of the 54 lateralized cases, 82% left-lateralizing and 90% right-lateralizing cases underwent unilateral adrenalectomy; favorable response with improvements in blood pressure, hypokalemia, and/or decreased fasting aldosterone level was seen in all individuals both in the immediate post-op setting and in those who had follow up. Of the seven “indeterminate” cases, two had a LR ≥ 2:1 at baseline and underwent successful unilateral adrenalectomy with excellent response post-operatively. Accuracy of baseline lateralization (LR ≥ 2:1) in agreeing with post-cosyntropin lateralization (LR ≥ 4:1) was similar in patients deemed successfully cannulated using the 1.4 or 2.0 A:P cutoff: 82.5% (n=66/80) vs 84% (n=42/50). Most discrepancies occurred with right-sided disease.

Conclusions: In our series of patients with PA, a baseline A:P ratio of 1.4 was better than 2.0 at confirming successful AVS cannulation, while maintaining similarly high specificity and accuracy of lateralization conclusions. In cases deemed “indeterminate” by the post-cosyntropin LR, the baseline LR may help identify additional patients who would benefit from unilateral adrenalectomy.


Nothing to Disclose: APD, NN, ND, FH, CL, EB, RC, CAS