Biochemical predictors of outcome of pituitary surgery for Cushing's disease.

OBJECTIVE Transsphenoidal surgery (TS) is the primary therapy for Cushing's disease (CD). The aims of this retrospective study were twofold: (i) investigate early and late results of TS for CD, and (ii) evaluate various postoperative tests in order to predict the outcome of TS. METHODS We reviewed the long-term outcome in 79 patients with CD who underwent TS (median follow-up 84 months, range 6-197). Within 2 weeks after surgery, morning serum cortisol concentrations were obtained (n = 78) and corticotropin-releasing hormone (CRH) (n = 53) and metyrapone tests (n = 72) were performed. Three groups of outcome were identified: sustained remission, early failure (persistent CD), and late relapse. RESULTS Immediate postoperative remission was achieved in 51 patients (65%), whereas 28 patients (35%) had persistent CD after TS. Ten patients developed recurrent CD after initial remission (20%). Morning cortisol: all relapses but one recorded serum cortisol >50 nmol/l. A cortisol threshold value of 200 nmol/l has a positive predictive value of 79% for immediate surgical failure (negative predictive failure [NPV] 97%). CRH test: CRH-stimulated peak cortisol > or =600 nmol/l predicted early failure in 78% (NPV 100%). All relapses recorded CRH-stimulated peak cortisol >or =485 nmol/l. Metyrapone test: 11-deoxycortisol >or =345 nmol/l predicted an early failure in 86% of cases (NPV 94%). CONCLUSION Predictive factors of surgical failure are morning cortisol >or =200 nmol/l, 11-deoxycortisol >or =345 nmol/l after metyrapone and CRH-stimulated cortisol >or =600 nmol/l. CRH and/or metyrapone testing are not superior to morning cortisol concentration in the prediction of outcome of TS. Careful long-term follow-up remains necessary independent of the outcome of biochemical testing.