Which patients benefit from provocative adrenal testing after transsphenoidal pituitary surgery?

OBJECTIVE Recent data suggest that recovery of anterior pituitary function promptly follows surgical decompression and that hypothalamic-pituitary-adrenal axis assessment need not be delayed following transsphenoidal pituitary surgery. We hypothesized that one protocol for both glucocorticoid supplementation and axis investigation prior to discharge may be applied to all transsphenoidal pituitary surgery patients, The protocol examined the merits of preoperative testing and of basal and hypoglycaemia-stimulated cortisol and ACTH measurements in post-operative axis evaluation. DESIGN Rapid tetracosactrin stimulation testing classified patients according to preoperative adrenal integrity. All patients received tapered doses of hydrocortisone beginning on the morning of surgery and discontinued after 48 hours. PATIENTS Of 28 consecutive patients with various pituitary tumours, 19 completed all aspects of the protocol, All evaluable information from the other 9 was incorporated into the final conclusions and recommendations. MEASUREMENTS Morning serum cortisol was measured 24 hours after the last hydrocortisone dose. Plasma ACTH and serum cortisol were measured during insulin tolerance testing within 8 days after surgery. Patients received clinical evaluations and repeat testing as clinically indicated during 6-30 months of follow-up. RESULTS Both peak serum cortisol > 550 nmol/l and peak plasma ACTH of > 4.4 pmol/l during insulin tolerance testing were 100% sensitive and specific in predicting sustained hypothalamic-pituitary-adrenal axis integrity after surgery. For patients entering surgery with normal tetracosactrin tests, an initial morning serum cortisol > 270 nmol/l provided 100% specificity for preserved axis integrity, but a low cortisol did not indicate axis dysfunction. For patients entering surgery with cortisol deficiency, an initial morning cortisol < 60 nmol/l indicated sustained axis failure, but higher values proved inconclusive. However, the basal cortisol, but not ACTH, on the day of insulin tolerance testing conclusively defined axis status in 18 of 19 study patients (95%). CONCLUSIONS We conclude that (1) a 48-hour perioperative hydrocortisone reducing regimen may be used in all uncomplicated transsphenoidal pituitary surgery cases regardless of pituitary-adrenal axis status before surgery; (2) preoperative adrenal testing aids interpretation of the initial morning serum cortisol and may be used to direct further testing; (3) a single morning serum cortisol drawn 24 hours after glucocorticoid withdrawal suffices for pituitary adrenal axis investigation if results suggest no change in axis function, as occurred in most study patients; (4) while insulin tolerance testing 5-8 days after surgery is 100% accurate in determining the need for sustained glucocorticoid replacement due to clinically significant hypopituitarism, repeat morning cortisol measurement obviates provocative testing in 95% of cases; and (5) basal and stimulated plasma ACTH values provide no information additional to serum cortisol measurements in post-operative axis evaluation.