ECEESPE2025 Oral Communications Oral Communications 13: Adrenal and Cardiovascular Endocrinology Part 2 (6 abstracts)
1Haukeland University Hospital, Department of Medicine, Bergen, Norway; 2University of Bristol, Henry Wellcome Laboratories for Integrative Neuroscience and Endocrinology, Translational Health Sciences, Faculty of Health Sciences, Bristol, United Kingdom; 3University of Birmingham, Centre for Systems Modelling and Quantitative Biomedicine, School of Medical Sciences, Birmingham, United Kingdom; 4University of Bergen, Haukeland University Hospital, Department of Clinical Science, Faculty of Medicine, Bergen, Norway; 5Haukeland University Hospital, University of Bergen, Department of Medicine, Bergen, Norway; 6Evangelismos Hospital, Department of Endocrinology, Athens, Greece; 7Karolinska University Hospital, Karolinska Institute, Department of Endocrinology, Stockholm, Sweden
JOINT561
Background: Primary aldosteronism (PA) is common, affecting 5-20% of the hypertensive population and is associated with an increased cardiovascular risk and incidence of metabolic disease when compared with primary hypertension. To date, establishing a correct diagnosis is hampered by laborious and time-consuming protocols unable to capture the pulsatile variability of aldosterone across the day. In this study we applied a dynamic hormone profiling technique to study the dynamics of aldosterone and other corticosteroids in patients with established PA to lay the foundation for new diagnostic procedures and criteria.
Methods: Sixty PA patients were recruited, 26 unilateral, 24 bilateral, and 10 with undetermined PA subtype. Ambulatory 24-hour profiles of corticosteroids including aldosterone and the hybrid steroids 18-hydroxycortisol and 18-oxocortisol were quantified in subcutaneous tissue microdialysate by liquid chromatography tandem mass spectrometry and compared with 215 healthy participants. Specific dynamic features were extracted and used to develop a machine learning model to distinguish PA profiles from healthy normal variation.
Results: PA profiles under unstimulated conditions were characterised by a pulsatile secretion and diurnal rhythmicity of aldosterone, cortisol, intermediary steroids and hybrid steroids. Nocturnal and early morning hypersecretion of aldosterone, 18-hydroxycortisol and 18-oxocortisol was prominent in unilateral PA. 18-oxocortisol was seen almost exclusively in unilateral PA, both with and without somatic KCNJ5-mutations. 24-hour cortisol concentrations did not differ between PA and healthy participants. Applying dynamic hormone profiling with a machine learning classifier algorithm resulted in a diagnostic sensitivity of 88% and specificity of 80% in predicting PA. Normalisation of aldosterone hypersecretion was observed after adrenalectomy for unilateral PA.
Conclusions: 24-hour tissue corticosteroid profiling in the home setting is achievable and reveals new insights into the pathophysiology of PA, characterised by nocturnal hypersecretion of aldosterone and hybrid steroids. The technique holds promise for faster and more accurate PA diagnosis with identification of unilateral cases.