Evaluation of the accuracy of inferior petrosal sinus sampling in the differential diagnosis of ACTH-dependent Cushing’s syndrome

Abstract

The differentiation of adrenocorticotropic hormone producing pituitary adenoma (Cushing's disease) from the ectopic ACTH syndrome is always a complex and difficult task, and in rare cases it is not possible to differentiate between the two disorders, even with the use of dynamic endocrine tests and the most advanced imaging techniques. Inferior petrosal sinus sampling (IPSS) with subsequent ACTH measurements became the gold-standard method of the differential diagnostic process. 34 patients with ACTH dependent Cushing's syndrome in whom the source of ACTH secretion couldn't be identified unambiguously with imaging techniques and/or dynamic endocrine tests underwent altogether 41 IPSS between 1999 and 2005. The sensitivity of the method was calculated on the basis of 31 samplings of 25 patients who had definite endocrinological diagnosis confirmed by the recovery from Cushing's syndrome after surgical intervention and/or by histological examinations (22 patients with ACTH-producing pituitary adenoma and 3 patients with ectopic ACTH syndrome). As a result of IPSS, pituitary-dependent Cushing's disease was diagnosed with a baseline central to peripheral ACTH ratio of >2.0 or with a ratio of >3.0 after corticotropin releasing hormone (CRH) administration. IPSS correctly identified ACTH-producing pituitary adenoma in 20 of 28 sampling procedures, with a sensitivity of 71.4%. Three patients had true negative and 8 had false negative results. There was no false positive result. Four of the 8 patients with false negative first sampling had a repeat sampling procedure leading to true positive result in each patient. In patients with Cushing's disease having true positive interventions, the basal and 5 minutes post-CRH ACTH concentrations were diagnostic in 14 and 19 cases, respectively. The sensitivity of IPSS within this series, reported for the first time from Hungary, was lower than it was found in much larger series published in international literature. In addition to technical difficulties, the lower sensitivity can be accounted also for the highly selected nature of the patient group.