Background: Secondary adrenal insufficiency is a common complication of transsphenoidal hypophysectomy (TSS) for pituitary adenoma. It is therefore imperative to rapidly and accurately identify patients requiring glucocorticoid replacement, thus minimising risks of adrenal insufficiency or unnecessary glucocorticoid exposure. The gold-standard test of HPA axis reserve, the insulin tolerance test (ITT), cannot be performed safely until after post-operative recovery. Nine am cortisol levels <100 nmol/l in the early post-operative period reliably predict secondary adrenal insufficiency during subsequent ITT. However there is little consensus over the early postoperative 0900 a.m. cortisol concentration which most reliably excludes secondary adrenal insufficiency.
Aim: To assess the performance of early post-TSS 0900 a.m. cortisol measurement to reliably detect and exclude secondary adrenal insufficiency as defined by ITT.
Methods: Data was reviewed from 36 patients (22 males, mean age 44 years) undergoing TSS followed by day 5 postoperative 0900 a.m. cortisol measurement and ITT 6 weeks post-surgery. All patients received postoperative glucocorticoid replacement, which was discontinued if 9am serum cortisol was >300 nmol/l.
Results: Of 20 of 23 patients who failed the ITT (peak cortisol <500 nmol/l), also had a day 5 0900 a.m. serum cortisol <300 nmol/l. Of 9 of 13 patients who passed the ITT, had a day 5 0900 a.m. cortisol >300 nmol/l. The cut-off cortisol level of <300 nmol/l had 86.9% sensitivity, 69.2% specificity and positive predictive value (PPV) of 83.3%, with respect to detection of secondary adrenal insufficiency. Increasing the cut-off to 400 nmol/l or below would result in 100% sensitivity and a PPV of 76.6%, but reduced specificity (46.1%). Decreasing the cut-off to <110 nmol/l would result in 100% specificity and PPV, although sensitivity would be 47.8%.
Conclusion: Day 5 post-TSS 0900 a.m. serum cortisol concentrations <110 nmol/l and >400 nmol/l accurately detect and exclude secondary adrenal insufficiency, respectively. Since levels between 110 and 400 nmol/l are poorly predictive of HPA reserve, glucocorticoid replacement should be continued in such cases until definitive assessment is performed with ITT.