ECE2017 Eposter Presentations: Adrenal and Neuroendocrine Tumours Adrenal cortex (to include Cushing's) (86 abstracts)
1Department of Medical Endocrinology, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark; 2Center of Rheumatology and Joint Diseases, Copenhagen University Hospital, Rigshospitalet Blegdamsvej, Copenhagen, Denmark; 3Center of Rheumatology and Joint Diseases, Copenhagen University Hospital, Frederiksberg Hospital, Frederiksberg, Denmark; 4Center of Rheumatology and Joint Diseases, Copenhagen University Hospital, Gentofte Hospital, Gentofte, Denmark; 5Center of Rheumatology and Joint Diseases, Copenhagen University Hospital, Rigshospitalet Glostrup, Glostrup, Denmark; 6Department of Clinical Biochemistry, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark; 7Department of Clinical Biochemistry, Copenhagen University Hospital, Nordsjællands Hospital Hillerød, Hillerød, Denmark.
Introduction: Evaluation of glucocorticoid production generally requires a dynamic test. Cut-off levels for baseline cortisol concentrations predicting the outcome of a Synacthen test have been proposed for different cortisol assays. With introduction of the new Roche Elecsys Cortisol II assay, P-cortisol concentrations are expected to decrease by 20%. We have investigated cut-off levels for baseline P-cortisol concentrations measured with the Roche Elecsys Cortisol II assay that could predict the response to a Synacthen test in patients at risk of glucocorticoid-induced adrenal insufficiency.
Methods: In a cross-sectional study, 110 prednisolone treated rheumatologic patients had a 250 μg Synacthen test performed, fasting, in the morning, starting between 0800 and 1030 h, 3648 h after the last prednisolone dose. P-cortisol was measured before and 30 min after Synacthen injection. The locally validated assay specific cut-off for normal adrenal function was 30 min P-cortisol ≥420 nmol/l.
Results: Forty-six patients (42%) had an insufficient response to the Synacthen test. Baseline and 30 min P-cortisol correlated positively (P<0.0001, r=0.86). Receiver-operator curve analysis showed an area under the curve of 0.94 (95% CI: 0.900.98). All patients with baseline P-cortisol <139 nmol/l also failed the Synacthen test (positive predictive value=100%). All patients with baseline P-cortisol >310 nmol/l had a normal response to the Synacthen test (negative predictive value =100%). Applying these cut-off values baseline P-cortisol measurements predicted the response to the Synacthen test in 58/110 (53%) of cases; P-cortisol >310 nmol/l (33%); P-cortisol <139 nmol/l (20%).
Conclusion: We have presented assay specific baseline P-cortisol concentrations that could predict the response to a Synacthen test in half of patients at risk of glucocorticoid-induced adrenal insufficiency and potentially reduce the number of needed Synacthen tests in the worldwide many glucocorticoid treated patients. Baseline morning cortisol measurements might become a valid diagnostic screening tool in the future.