OR02-03 Twenty-four-hour Tissue Microdialysis Of Aldosterone And Metabolites In Primary Aldosteronism

Abstract

Abstract Disclosure: M.A. Grytaas: None. P. Methlie: None. I. Marinelli: None. E. Zavala: None. M. Øksnes: None. T. Upton: None. K. Simunkova: None. D.A. Vassiliadi: None. K. Løvås: None. S. Bensing: None. I. Botusan: None. K. Berinder: None. G.M. Russell: None. G. Ueland: None. O. Kampe: None. S. Tsagarakis: None. S.L. Lightman: None. E.S. Husebye: None. Background: Primary aldosteronism (PA) is the most common cause of secondary hypertension but is grossly underdiagnosed. Early detection, accurate discrimination between unilateral and bilateral PA and appropriate treatment are vital to prevent cardiovascular and renal complications. The current workup however is a cumbersome, multistep process that may not detect dynamic aldosterone variability or nocturnal hypersecretion. Furthermore, subtype determination with adrenal venous sampling is invasive and technically challenging. Novel dynamic 24-hour steroid profiling may represent a major step forward in the diagnostics and subtyping of PA. Aim: To assess 24-hour rhythmicity of aldosterone and metabolites in PA, and to develop a novel dynamic diagnostic tool to discriminate PA from healthy subjects (HS), using tissue microdialysis. Material and methods: Twenty-minute frequency microdialysis fractions were collected over 24 hours from 64 ambulatory patients with PA (26 bilateral, 24 unilateral, 14 with undetermined subtype) using the U-Rhythm sampler followed by multiplex hormone assay liquid chromatography tandem mass spectrometry (LC-MS/MS). Sixteen patients subsequently treated with adrenalectomy for unilateral PA had additional postoperative samplings. Microdialysis samplings from 214 HS were used as controls. We calculated statistical distributions of dynamic biomarkers of abnormality, and developed a machine learning classifier that discriminates PA from HS. Results: Although there were large interindividual variations, PA patients did show a disturbed circadian rhythmicity of aldosterone compared with HS, including some with multiple very high ultradian spikes. Unilateral PA had higher aldosterone spikes and more disturbed rhythmicity compared with bilateral PA. A profound reduction of aldosterone levels and normalisation of rhythmicity was seen post adrenalectomy. Applying a Random Forest classifier in a subgroup of 20 PA patients, 82% specificity and 85% sensibility were achieved to discriminate PA from HS. Our analysis also suggests 18-hydroxycortisol as the most discriminative dynamic biomarker. Conclusions: We demonstrate for the first time circadian and ultradian rhythms of aldosterone in a large cohort of PA compared with HS. Based on our preliminary results, dynamic 24-hour samplings may provide a novel method for diagnosing and subtyping PA, and assess biochemical cure after adrenalectomy. Presentation: Thursday, June 15, 2023