The role of prolactin gradient and normalized ACTH/prolactin ratio in the improvement of sensitivity and specificity of selective blood sampling from inferior petrosal sinuses for differential diagnostics of ACTH-dependent hypercorticism

Abstract

The present study included patients with confirmed diagnosis of ACTH-dependent hypercorticism admitted to the Department of Neuroendocrinology and Osteopathies, Endocrinological Research Centre, between 2008 and 2012. Selective blood sampling from the inferior petrosal sinuses was performed with the stimulation by decompressin administered intravenously at a dose of 8 mcg. The normalized ACTH/prolactin ratio was calculated by dividing the maximum ACTH gradient following decompressin stimulation by the ipsilateral prolactin gradient. The cut off values were calculated from the operating characteristic curves as follows: 1.5 for the prolactin gradient (sensitivity 92.5%, specificity 100%) and 1.18 for the normalized ACTH/prolactin ratio (sensitivity 85.9%; 95% CI 76.8-93.4, specificity 100%; 95% CI 60.9-100). In the study cohort (n=70), the normalized ACTH/prolactin ratio proved to be a more specific but less sensitive parameter than the routine analysis of the ACTH gradient (sensitivity 98.4%, specificity 83.3%). Areas under the operating characteristic curves were on the whole identical. However, comparison of the diagnostic value of selective blood sampling from the inferior petrosal sinuses in combination of the determination of the prolactin gradient and normalized ACTH/prolactin ratio with the results obtained in the patients in whom the catheter position was not controlled (n=47) revealed the advantages of the former approach in terms of both the number of unverified causes of hypercorticism (1 versus 6) and the area under the operating characteristic curves 0.964 (95% CI 0.897-1.032) among 70 patients and 0.910 (95% CI 0.821-0.998) among the first 47 patients (included in the analysis confined to the confirmed cases). It is concluded that the determination of the prolactin gradient and calculation of the normalized ACTH/prolactin ratio allow to increase the diagnostic value of selective blood sampling from the inferior petrosal sinuses under decompressin stimulation. This inference is especially true of doubtful cases.