Subscribe to RSS
DOI: 10.1055/s-0034-1372153
Adrenal function after adrenalectomy for subclinical and overt Cushing's syndrome: A systematic review of the literature
Introduction: The aims of this review are to analyze the prevalence of postoperative adrenal insufficiency (pAI) and the time to recovery a normal adrenal function in patients with subclinical (SCS) and overt (CS) Cushing syndrome.
Methods: Using the PubMed database, a systematic review of the literature was conducted selecting human studies published from 1980.
Results: Of the 1522 screened papers, 28 were selected (13 retrospective, 14 prospective and 1 randomized controlled trial). Of the 268 SCS patients, 248 (93%) underwent unilateral adrenalectomy. The prevalence of pAI was 69% in SCS patients and 99.7% in CS subjects. No deaths were reported in SCS patients, whereas 2 CS subjects died due to acute AI. Considering the heterogeneity in the diagnosis of SCS among the studies, SCS patients were re-classified in 3 groups: patients diagnosed only by the dexamethasone test (DST), patients classified with DST + 1 additional test, and DST + 2 tests (dex+2). The prevalence of pAI in the 3 groups was 51%, 62%, and 96%, respectively, and it was lower than that of CS patients for the first 2 groups (P < 0.001). No difference was found between dex+2 and CS groups (P = 0.106). The prevalence of recovery a normal adrenocortical function was 98% in SCS and 94% in CS patients. The mean time to recovery was lower in SCS patients (6.4 months) with respect to CS subjects (10.8 months, P = 0.010). The midnight serum cortisol showed the best accuracy in predicting the pAI (sensitivity 64%, specificity 81%). Patients with at least 2 impaired tests between DST, urinary cortisol, midnight serum cortisol, and ACTH had the highest risk of developing pAI. Cortisol after CRH test < 120 nmol/L has also been associated with pAI.
Conclusion: The pAI should be considered in the postoperative management of SCS patients. Prevalence and time to recovery a normal adrenocortical function are tightly related to the degree of hypercortisolism and to the diagnostic criteria to define SCS.