Endocrine Abstracts

Introduction: Primary aldosteronism (PA) is one of the most common causes of secondary hypertension, leading to a higher risk of patient morbidity and mortality. Through collaboration with the University of Dresden and New South Wales (NSW) Health Pathology, we have developed a LC-MS/MS method for measurement of aldosterone and various metabolites to streamline identification and subtype classification of PA utilising previously validated machine learning (ML) models. This has been shown to enable early identification of PA patients with KCNJ5 mutations to enable swifter adrenalectomy, without necessitating the need for adrenal vein sampling or other confirmation.

Methods: Analysis of 18-hydroxycortisol, 18-oxocortisol, DHEA, 17-hydroxyprogesterone, androstenedione, corticosterone, 11-deoxycortisol, 21-deoxycortisol, cortisone, cortisol, aldosterone, 11-deoxycorticosterone and 11-dehydrocorticosterone was performed in a single LC-MS/MS method after supported liquid extraction (SLE) of plasma samples. Analytical and ML comparisons were undertaken between MFT, Dresden and NSW using fully validated methods.

Results: Validation parameters were acceptable. There was clear numerical correlation for the analytes, analysis highlighted a negative bias for 18-hydroxycortisol which was minimised after a change in calibration material. Results were processed through ML programs and outcomes compared.

Discussion: We have developed an LC-MS/MS assay and proven the results generated are comparable with the groups in Dresden and NSW. To our knowledge, we are the only UK laboratory involved in this project and hope to develop our in-house method further, with a view to implementing this approach in the future for diagnostic work up of PA patients.