Endocrine Abstracts

Introduction: The use of Inferior Petrosal Sinus Sampling (IPSS) to differentiate between central and ectopic ACTH dependent Cushing’s, although requires expertise, is widely available. The value of adding prolactin measurement to that of ACTH to improve IPSS diagnostic accuracy remains controversial. We evaluated the impact of adding prolactin measurement to IPSS procedures at a large academic center.

Methods: Leftover samples from patients who were investigated by IPSS procedure for differentiation of ACTH source were stored at -70°C until analysis. ACTH had been measured and results reported. Retrospective measurement of prolactin was performed on the saved sample aliquots using Cobas 4000 (Roche Diagnostics, IA, USA). Prolactin normalized ACTH IPSS peripheral ratios >0.8 indicated central Cushing’s and <0.6 indicated ectopic source of ACTH. Previously reported IPSS studies were reevaluated following prolactin normalization of ACTH.

Results: Patients (n = 10) with ACTH dependent Cushing’s and who had undergone IPSS were entered into the study. Prolactin levels in IPSS samples ranged from 38 mIU/l to 12,978 mIU/l. Although left and right inferior petrosal ACTH values were significantly different (P = 0.03) suggesting lateralization, prolactin normalized ACTH levels were not (P = 0.37). Patients (n = 8) were reported as having central ACTH production. One patient had ectopic ACTH production, whereas one patient was difficult to cannulate and thus to assess. In this patient, although ACTH data suggested a questionable pituitary source for ACTH, the patient was classified as having ectopic ACTH syndrome following prolactin normalization of ACTH. Subsequent clinical findings were consistent with ectopic ACTH syndrome.

Conclusions: Although the addition of prolactin to the analysis of IPSS ACTH provided a reassurance for ACTH ratio calculations as well as an aid in difficult catheterization studies, tumour lateralization became less obvious when using prolactin normalized ACTH levels. A larger study is required to confirm those findings.