Predicting late recurrence in surgically-treated patients with Cushing's disease.

OBJECTIVE: Cushing's disease (CD) has an uncertain prognosis because patients achieving remission after transsphenoidal pituitary neurosurgery (TSS) may relapse. We aimed to identify factors predicting relapse, focusing on desmopressin (DDAVP) and corticotropin-releasing hormone (CRH) tests after surgery. MATERIALS AND METHODS: 57 patients with CD (mean age 36 yrs) after TSS experienced remission (24 cases); late relapse (LR) (15 cases); or persistent disease (18 cases). RESULTS: The median time to relapse was 40 months. ACTH levels increased after both DDAVP and CRH stimulation, with a significantly higher response in the late recurrence group, showing this to be an indicator of increased risk of relapse. In the logistic regression model, a rise in ACTH > 9 pg/ml after DDAVP and > 36.7 pg/ml after CRH showed a sensitivity of 93% and 73%, respectively, a specificity of 82% and 76% in LR group. The area under the curve was 0.91 for DDAVP, 0.80 for CRH, and 0.95 for DDAVP+CRH test, i.e. the combined tests performed better than each test alone, but not to a statistically significant degree. A response to both tests resulted in a positive predictive value (PPV) of 100%, while no response to either test in a negative predictive value (NPV) of 100%. CONCLUSIONS: ACTH hyper-responsiveness to DDAVP stimulation proved a valuable indicator of relapsing patients with high sensitivity and specificity; in selected cases when a clear high increment of ACTH level is not evident, the CRH test might be used as additional tool to confirm the risk of future relapses.