Cavernous sinus sampling in selected cases of Cushing's disease.

The classical trias in Cushing's disease, normal or moderately elevated plasma ACTH, significant suppression of cortisol in the high-dose dexamethasone test, and stimulation of ACTH and cortisol in the CRH test, confirms the diagnosis in most cases. However, as a referral centre for complex Cushing's disease cases, we are confronted with problems in the differentiation of pituitary and ectopic Cushing's syndrome and with problems in the localisation of minute ACTH-secreting pituitary adenomas. In this study, cavernous sinus sampling (CSS) was evaluated as a diagnostic tool in complex Cushing's disease cases. Thirty-five patients were transferred to our unit for the treatment of Cushing's disease between January 1999 and August 2000. Of those, 17 patients (including five children) had the combination of equivocal results in endocrinological testing and negative MRI prior to admission. In these cases, CSS was performed preoperatively to confirm the diagnosis and to obtain further information about the localisation of pituitary microadenomas. Twelve of these 17 patients showed the classical trias for Cushing's disease after equivocal tests were repeated. A central-peripheral gradient was also found using CSS. In ten of the 11 patients, where CSS could be successfully performed, the lateralisation of the ACTH-adenoma was correctly predicted (91%). In the eleventh case, a right ACTH-gradient was found in a stalk tumour. In one patient (a child), the catheterisation of the cavernous sinus failed due to anatomical reasons. Eleven of these 12 patients successfully underwent transnasal adenomectomy. In two of the 17 patients, active Cushing's syndrome was excluded. In the remaining three cases, ectopic Cushing's syndrome was suspected based on CSS results. Therefore, these patients did not undergo pituitary exploration. In complex cases, we recommend CSS for diagnostic purposes, especially for the localisation of ACTH-secreting microadenomas within the pituitary. In our experience, CSS has a higher accuracy than inferior petrosal sinus sampling (IPSS) in the localisation of ACTH-adenomas.