ISSN 1470-3947 (print) | ISSN 1479-6848 (online)

Endocrine Abstracts (2011) 25 P309

Utility of basal DHEAS measurement in the detection of subclinical autonomous glucocorticoid hypersecretion in adrenal incidentaloma

Anand K Annamalai1,2, Narayanan Kandasamy1,2, Natalie Freeman1,2, Kuhan Venugopal1,2, Jonathan Chia1,2, Ashley Shaw1,3, Helen L Simpson1,2, David Halsall1,4 & Mark Gurnell1,2

1Addenbrookes Hospital, Cambridge, UK; 2Wolfson Diabetes and Endocrine Clinic, Cambridge, UK; 3Department of Radiology, Cambridge, UK; 4Department of Biochemistry, Cambridge, UK.

Background: Adrenal incidentalomas (AI) are identified in 4–7% of patients >40 years undergoing abdominal CT/MRI. Evidence of subclinical autonomous glucocorticoid hypersecretion (SAGH) is found in 5–10% of cases depending on the diagnostic criteria/thresholds adopted.

Aim: To examine the utility of basal DHEAS measurement in the detection of SAGH in a cohort of patients with AI.

Methods: Ninety-six consecutive subjects (49 females, 47 males: mean age 62 years, range 29–84 years) were referred to our clinic over a 3-year period. All patients were assessed by two clinicians (AA and MG), and underwent repeat dedicated adrenal imaging and biochemical profiling, including overnight dexamethasone suppression testing (normal cortisol suppression: <50 nmol/l). Cases in whom there was no clinical or biochemical evidence of endocrine hyperfunction, and with appropriate radiological features, were categorised as non-functioning adrenal adenomas (NFAA). SAGH was confirmed by the finding of low ACTH levels, abnormal circadian rhythm, and failure of cortisol to suppress during a 48 h low dose dexamethasone suppression test (normal: <50 nmol/l). DHEAS levels were classified as undetectable, suppressed (but detectable) and normal.

Results: Amongst NFAA (n=41), DHEAS levels were normal in 38, suppressed in one, and undetectable in two subjects. Amongst SAGH (n=18) subjects, nine had undetectable DHEAS, three had suppressed, and six normal levels. Receiver operating characteristic analysis using DHEAS concentration to diagnose SAGH returned an area under the curve of 0.85 (0.75–0.96). At the optimum cut-off of 0.9 nmol/l, sensitivity and specificity were 78% (52–94%) and 74% (58–87%) respectively. At a cut-point of 0.5 nmol/l the test is specific (95%) with a sensitivity for the detection of SAGH of 50%.

Conclusion: The finding of a subnormal DHEAS level in a patient with an AI is an important pointer to the potential presence of SAGH and merits further investigation.