Predictors of remission of hyperprolactinaemia after long-term withdrawal of cabergoline therapy

Abstract

Background: Remission rates of 76, 69.5 and 64.3% have been reported in patients with nontumoural hyperprolactinaemia (NTH), microprolactinoma and macroprolactinoma, respectively, 2-5 years after cabergoline (CAB) withdrawal.

Objective: To report the estimated recurrence rate at 24-96 months after CAB withdrawal and indicate predictors of disease remission.

Design: Observational, analytical, prospective.

Patients: Of 381 previously untreated de novo patients with hyperprolactinaemia, 221 (58%) (173 women, 48 men; 27 with NTH, 115 with micro-, and 79 with macroprolactinoma) were studied.

Measurements: Using multiple regression analysis the diagnostic accuracy of nadir PRL levels (t = 7.6, P < 0.0001) and nadir maximal tumour diameter at CAB withdrawal (t = 3.9, P < 0.001) was analysed using receiver operating characteristic (ROC) curves.

Results: The recurrence of hyperprolactinaemia was 25.9, 33.9 and 53.1% in patients with NTH, micro- or macroprolactinoma, respectively. To predict the last PRL level after withdrawal, the optimum cut-off of nadir PRL levels at withdrawal was 162 mU/l (5.4 microg/l) [sensitivity (95% CI) 76% (67-84%), specificity 65% (51-77%)] and that of nadir maximal tumour diameter was 3.1 mm [sensitivity 52% (41-63%), specificity 86% (79-91%)]. The patients achieving both nadir PRL levels </= 162 mU/l and maximal tumour diameter </= 3.1 mm (n = 111) at CAB withdrawal had a significantly lower Kaplan-Meier estimate of recurrence of hyperprolactinaemia (20%) at 24-96 months than those who did not fulfil any of these criteria [(n = 38) 90%; P < 0.0001]. Patients achieving nadir PRL levels </= 162 mU/l (n = 26) or maximal tumour diameter </= 3.1 mm during CAB treatment (n = 46) had an estimated recurrence rate of hyperprolactinaemia of 50 and 56%, respectively.

Conclusion: Persistent remission of hyperprolactinaemia without any evidence of tumour re-growth after 24-96 months of CAB withdrawal occurred in the majority of patients with NTH and microprolactinoma and in about half of those with macroprolactinoma. Nadir PRL levels and maximal tumour diameter at CAB withdrawal of </= 162 mU/l and </= 3.1 mm predicted remission of hyperprolactinaemia in 80% of patients.