Background: Correct identification of patients with HPA axis dysfunction following trans-sphenoidal pituitary surgery is important. Our centre measures post-operative day 5 0900 h serum cortisol (24 h after the last dose of glucocorticoid) to identify the need for glucocorticoid replacement. This is given if post-operative day 5 0900 h cortisol concentration is <300 nmol/l, until dynamic pituitary function assessment occurs.
Methods: Data were reviewed for 51 patients undergoing an insulin tolerance test (ITT) ~6 weeks after pituitary surgery (20102015) using a new cortisol assay. The peak cortisol concentration during an ITT was compared to post-operative day 5 0900 h serum cortisol.
Results: Of the 27 patients who passed the ITT (peak cortisol >450 nmol/l) 18 had post-operative day 5 0900 h cortisol concentration <300 nmol/l. Of the 24 patients who failed the ITT, seven had post-operative day 5 0900 h cortisol concentrations of >300 nmol/l. The cortisol concentration cut-off point of 300 nmol/l had 70.8% (95% CI: 48.9187.38%) diagnostic sensitivity and 33.3% (95% CI: 16.5254.0%) diagnostic specificity. 100% (85.75100.00%) sensitivity was achieved by increasing the cut-off to 633 nmol/l. Decreasing the cut-off to 101 nmol/l resulted in 100% (87.23100.00%) specificity. The ROC curve investigating the predictive ability of post-operative day 5 0900 h cortisol had an AUC of 0.60 (95% CI 0.440.76).
Conclusions: Following pituitary surgery, post-operative day 5 0900 h cortisol measurement is effective at identifying patients who will subsequently fail an ITT, and hence, require glucocorticoid replacement. However, the current cut-off of 300 nmol/l falsely reassures some patients they do not need glucocorticoid replacement post-operatively.