Early morning cortisol levels as predictors of short-term and long-term adrenal function after endonasal transsphenoidal surgery for pituitary adenomas and Rathke's cleft cysts.

OBJECTIVE Patients undergoing pituitary adenoma or Rathke cleft cyst (RCC) removal are often administered perioperative glucocorticoids regardless of lesion size and preoperative adrenocorticotropic hormone/cortisol levels. To minimize unnecessary glucocorticoid therapy, we describe a protocol in which patients with normal preoperative serum cortisol and adrenocorticotropic hormone levels are given glucocorticoids only if postoperative day 1 or 2 (POD1 or POD2) cortisol levels decrease below normal. METHODS A total of 207 consecutive patients undergoing endonasal surgery for an adenoma or RCC were considered for study. Of these, 68 patients with preoperative adrenal insufficiency or Cushing disease were excluded. Glucocorticoids were withheld unless POD1 or POD2 morning cortisol values were below normal (≤4 μg/dL). Subsequent adrenal status was assessed through follow-up biochemical and clinical evaluations. RESULTS The 139 patients included 119 with macroadenomas, 14 microadenomas, and 6 RCCs (follow-up, 3-41 months; median, 10 months). Nine patients (6.5%), all with macroadenomas (mean diameter, 26 ± 10 mm) had low POD1 or POD2 cortisol values and received glucocorticoids; of these, five patients were weaned off within 3-28 weeks of surgery. Overall, 12 of 139 patients (8.6%) were treated for early (n = 9) or delayed (n = 3) adrenal insufficiency but only 5 patients (3.6%) remain on glucocorticoid replacement. No patient experienced an adrenal crisis. Using morning POD1 or POD2 cortisol values >4 μg/dL as a measure of adequate hypothalamic-pituitary-adrenal axis function, yields a sensitivity of 96%, a specificity of 57%, and a positive predictive value of 98%. CONCLUSIONS In patients with normal preoperative cortisol levels undergoing endonasal removal of a pituitary adenoma or RCC, normal morning cortisol values on POD1 and POD2 reliably predicts adequate and safe adrenal function in 98% of patients. This simple protocol of withholding postoperative glucocorticoids avoids unnecessary steroid exposure and poses minimal risk to the well-informed closely monitored patient.