Abstract
Primary aldosteronism (PA) is a cause of secondary hypertension. Accurate diagnosis is important to direct treatments: surgical for aldosterone-producing adenoma (APA) or mineralocorticoid receptor antagonist therapy for bilateral adrenal hyperplasia (BAH). The PA diagnosis is sequential: positive screening (raised aldosterone:renin ratio, ARR) is followed by confirmatory testing (e.g. salt-loading protocols) and subtype diagnosis (computerized tomography [CT] scanning and adrenal vein sampling, AVS). These techniques are invasive, costly and often limited to specialist centres.
The steroidogenic intermediates 18-hydroxycorticosterone (18OHB), 18-hydroxycortisol (18OHF) and 18-oxocortisol (18oxoF), primarily synthesized by aldosterone synthase, are raised in PA. This study assessed serum 18OHB and serum and urinary 18OHF and 18oxoF concentrations in PA diagnosis in 143 ARR-positive patients off interfering medications. Salt-loading classified 62 low-renin essential hypertension (LREH) and 81 PA cases. Of confirmed PA, CT and AVS classified 61 BAH and 20 APA cases. PA patients excreted higher urinary 18OHF and 18oxoF (median excretion 206 and 4.5 μg/day) than LREH patients (94 and 2.2 μg/day, P< 0.001). APA patient excretion (407 and 9.3 μg/day) was higher than BAH (160 and 4.3 μg/day, P< 0.001). Serum results were similar to urine, although a greater overlap between groups occurred. Salt-loading reduced serum 18OHB, 18OHF and 18oxoF across groups. Nonetheless, PA and LREH and APA and BAH groups could still be separated. Serum 18OHB:aldosterone ratio doubled in LREH, while being unchanged in PA.
In conclusion, this study demonstrates that serum 18OHB, urinary 18OHF and urinary 18oxoF in patients with raised ARR correlate with confirmatory PA tests. Receiver-operator curve analysis for urinary 18OHF in PA diagnosis demonstrated better performance than ARR (area under the curve, 0.82 versus 0.61, P< 0.05). Its measurement could have prevented 36.4% of salt-loading procedures and 7.3% of AVS. However, a number of patients were not classified by steroid measurement alone, and would require further testing. Nonetheless, results are promising and would benefit from larger studies and steroid measurement standardization.