Transsphenoidal pituitary surgery in Cushing's disease: can we predict outcome?

OBJECTIVE To assess the results of transsphenoidal pituitary surgery in patients with Cushing's disease over a period of 18 years, and to determine if there are factors which will predict the outcome. PATIENTS Sixty-nine sequential patients treated surgically by a single surgeon in Newcastle upon Tyne between 1980 and 1997 were identified and data from 61 of these have been analysed. DESIGN Retrospective analysis of outcome measures. MAIN OUTCOME MEASURES Patients were divided into three groups (remission, failure and relapse) depending on the late outcome of their treatment as determined at the time of analysis, i.e. 88 months (median) years after surgery. Remission is defined as biochemical reversal of hypercortisolism with re-emergence of diurnal circadian rhythm, resolution of clinical features and adequate suppression on low-dose dexamethasone testing. Failure is defined as the absence of any of these features. Relapse is defined as the re-emergence of Cushing's disease more than one year after operation. Clinical features such as weight, sex, hypertension, associated endocrine disorders and smoking, biochemical studies including preoperative and postoperative serum cortisol, urine free cortisol, serum ACTH, radiological, histological and surgical findings were assessed in relation to these three groups to determine whether any factors could reliably predict failure or relapse after treatment. RESULTS Of the 61 patients included in this study, 48 (78.7%) achieved initial remission and 13 (21.3%) failed treatment. Seven patients suffered subsequent relapse (range 22-158 months) in their condition after apparent remission, leaving a final group of 41 patients (67.2%) in the remission group. Tumour was identified at surgery in 52 patients, of whom 38 achieved remission. In comparison, only 3 of 9 patients in whom no tumour was identified achieved remission. This difference was significant (P = 0.048). When both radiological and histological findings were positive, the likelihood of achieving remission was significantly higher than if both modalities were negative (P = 0.038). There were significant differences between remission and failure groups when 2- and 6-week postoperative serum cortisol levels (P = 0.002 and 0.001, respectively) and 6-week postoperative urine free cortisol levels (P = 0.026) were compared. This allowed identification of patients who failed surgical treatment in the early postoperative period. Complications of surgery included transitory DI in 13, transitory CSF leak in 8 and transitory nasal discharge and cacosmia in 3. Twelve of 41 patients required some form of hormonal replacement therapy despite achieving long-term remission. Thirteen patients underwent a second operation, of whom 5 achieved remission. CONCLUSIONS Transsphenoidal pituitary surgery is a safe method of treatment in patients with Cushing's disease. Operative findings, radiological and histological findings, together with early postoperative serum cortisol and urine free cortisol estimates may identify failures in treatment. Alternative treatment might then be required for these patients. Because of the risk of late relapse, patients require life-long follow-up.