Reliability of the diagnostic tests for Cushing’s syndrome performed in a tertiary referral center Mutlu Gu ¨nes • Ozlem Celik • Pinar Kadioglu Published online: 1 April 2012 Ó Springer Science+Business Media, LLC 2012 Abstract The study aimed to retrospectively evaluate the reliability of the diagnostic and location tests in Cushing’s Syndrome (CS). Eighty-seven patients diagnosed with CS between 1995 and 2007 by Endocrinology Metabolism Department of Cerrahpasa Medical School were included in the study. The control group consisted of 91 patients who presented to the outpatient clinic because of obesity. The diagnostic tests were as follows: 1 mg dexamethasone suppression test (DST), 24-h urinary free cortisol (UFC), midnight cortisol level (MCL), ACTH level and overnight 8 mg DST. The sensitivity and specificity of UFC were 81 and 66 % respectively for the cut-off point of 50 lg/day, whereas they were 64 and 76 % respectively for the cut-off point of 100 lg/day. For the cut-off value of 1.8/lg/dL for MCL and 1 mDST, the sensitivity rates were 100 and 98 %, while the specificity rates were 88 and 33 %, respectively. Among the location tests, the sensitivity and specificity of ACTH under 10 pg/mL for adrenal CS were 92 and 94 % respectively. The sensitivity and specificity of ACTH higher than 30 pg/mL for ACTH-dependent CS were 69 and 100 % respectively. The sensitivity rates of 8 mg DST for 50 and 60 % suppressions were 83 and 79 % respectively, whereas the specificity rates were 75 and 88 % respectively. 1 mg DST (cut-off \ 1.8 lg/dL) and UFC (50 lg/24 h) are appropriate tests for screening CS. Overnight 8 mg DST with 60 % suppression for Cushing’s Disease (CD) and ACTH levels \ 10 pg/mL for adrenal CS, ACTH levels [ 30 pg/mL for ACTH dependency were identified as the best tests for the differential diagnosis of the subtypes. Keywords Cushing’s disease Á Diagnostic tests Á Midnight cortisol Á Urinary free cortisol Á ACTH Introduction Cushing’s syndrome (CS) develops due to endogenous excess cortisol production or more commonly, as a result of long-term glucocorticoid therapy [1]. The annual incidence of endogenous CS among general population is 0.7–2.4 per million [1, 2]. Insulin resistance, central obesity, dyslipi- demia, hypertension, impaired glucose tolerance, diabetes mellitus, atherosclerosis and early cardiovascular disease are associated with CS. It has been shown that standard mortality rate was increased by 3.8–5 folds in the patients whose moderate hypercortisolemia persisted despite ther- apy, as compared to the general population [3–5]. Symp- toms and signs which were characteristic for CS such as easy bruising, plethora, strias, weight gain, edema, dorso- cervical fat pad, proximal myopathy and depression are also seen in the pseudo-Cushing states [3]. The incidence of CS is increased since the screening tests are performed not only in those with the symptoms of CS but also in those with obesity, metabolic syndrome, polycystic ovary syn- drome and/or osteoporosis [6–8]. The incidence of CS in uncontrolled diabetic patients has increased up to 2–3 % [9]. Considering that they overlap in terms of clinical char- acteristics and they concern a large patient population, the test that would be used for the differentiation of pseudo- M. Gu ¨nes Á O. Celik Á P. Kadioglu Division of Endocrinology and Metabolism, Department of Internal Medicine, Cerrahpasa Medical School, University of Istanbul, Istanbul, Turkey P. Kadioglu (&) Cerrahpasa Tip Faku ¨ltesi, Ic Hastalıkları Anabilim Dali, Endokrinoloji-Metabolizma ve Diyabet Bilim Dali, 34303 Cerrahpasa, Istanbul, Turkey e-mail: kadioglup@yahoo.com 123 Pituitary (2013) 16:139–145 DOI 10.1007/s11102-012-0387-7