Predicting relapse after transsphenoidal surgery for Cushing's disease.

Although selective transsphenoidal surgery is an effective treatment for pituitary-dependent Cushing's syndrome the definition of cure as distinct from improvement is unclear. Complete tumor removal should be associated with very low serum cortisol levels because of long-term suppression of normal corticotrophs but the optimum timing of this investigation after surgery has not been established. Eleven consecutive patients with surgical and histological confirmation of a corticotroph adenoma removed at transsphenoidal surgery for proven Cushing's disease were studied with 0900 h serum cortisol levels at 5-14 days and 6-12 weeks postoperatively. Patients were maintained on hydrocortisone 10 mg three times daily (final dose at 1800 h) pending recovery of the hypothalamic-pituitary axis which was assessed by periodic short tetracosactrin tests and continued remission of the condition was confirmed by low dose dexamethasone suppression testing. All patients achieved clinical resolution of their disease but four subsequently developed biochemical evidence of recurrence with incomplete suppression on low dose dexamethasone testing at 2-48 months after surgery. These patients had 0900 h serum cortisol levels of 124, 95, 186, and 265 nmol/L at 5-14 days and 334, 52, 130, and 240 nmol/L at 6-12 weeks postoperatively. The remaining seven patients, who are in remission after 8-83 (median 24) months of follow-up, demonstrated 0900 h serum cortisol levels of 30-75 (median 31) nmol/L at 5-14 days but lower levels at 6-12 weeks (< 20 nmol/L in three patients and 22, 30, 30, and 33 nmol/L in the remainder). In this series, serum cortisol measurements at 6-12 weeks after transsphenoidal surgery for Cushing's disease are lower than levels obtained within 2 weeks of surgery and appear to give better discrimination of continuing remission; levels less than 35 nmol/L suggest a favorable long-term outcome.