The diagnostic utility of late night salivary cortisol (LNSF) and cortisone (LNSE) in Cushing's syndrome.

Abstract

Measurement of late night salivary cortisol (LNSF) is useful in the identification of cyclical Cushing's syndrome (CS); the usefulness of its metabolite cortisone (late night salivary cortisone, LNSE) is less well described. The aim of this study was to determine the utility of measuring LNSE in patients with confirmed CS compared with other diagnostic tests and to analyse serial LNSF measurements for evidence of variable hormonogenesis. This was a retrospective observational study including patients with confirmed CS in whom LNSF and LNSE were measured. Twenty-three patients with confirmed CS were included, 21 with Cushing's disease. LNSF had a sensitivity of 92%, LNSE 87% and combined LNSF/LNSE 94% per sample. Four patients had cyclical hormonogenesis, when the definition of one trough and two peaks was applied to LNSF measurements, anda fifth patient fell just outside the criteria. Six patients had evidence of variable hormonogenesis, defined as doubling of LNSF concentration on serial measurements. Sensitivity of 24-h urinary free cortisol (UFC) was 89% per collection. Sixteen patients had simultaneous measurements of LNSF and UFC; in three patients, they provided discordant results. LNSF appears more sensitive than LNSE and UFC in the diagnosis of CS, combining LNSF and LNSE results leads to superior sensitivity. Half of our cohort had evidence of cyclical or variable hormonogenesis. Fluctuations in LNSF did not always correlate with changes in UFC concentration, emphasising the importance of performing more than one screening test, particularly if pretest clinical suspicion is high.