Incidence of Cushing's syndrome in patients with significant hypercortisoluria

Abstract

Objective: To investigate the incidence of Cushing's syndrome (CS) in patients with significant hypercortisoluria and the performance of urinary free cortisol (UFC) screening. Design: Retrospective file review. Methods: The computerized database of a publicly funded health maintenance organization (HMO) in Israel was searched for all patients who underwent 24-h UFC testing in 2005-2014 with a result of more than twice the upper limit of normal (ULN). The patients' medical files were reviewed for a subsequent diagnosis of CS by an expert endocrinologist. Findings were evaluated for patterns in CS diagnosis and UFC testing over time. Results: Of 41 183 individuals tested, 510 (1.2%) had UFC >2× ULN (214 >3× ULN). Eighty-five (16.7%) individuals were diagnosed with CS (63 female and mean age 47.2 ± 15.1 years), mainly Cushing's disease (55.3%) or adrenal Cushing's syndrome (37.6%). The number of UFC tests increased steadily, from 1804 in 2005 to 6464 in 2014; yet, the resultant detection rate of CS remained generally stable. The calculated incidence of CS in the general HMO-insured population based only on the patients identified in the present cohort was 4.5 new cases/million/year (median 4.9/million/year, range 1.7-5.9/million/year), which was also relatively stable. The most common reason for referral for UFC screening was obesity. Of the 148 patients before bariatric surgery with UFC >2× ULN, 2 were diagnosed with CS. Conclusions: The incidence of CS is higher than previously suggested. The consistently increasing number of UFC tests being performed has not been accompanied by a similar increase in CS detection rate. The expected yield of routine UFC testing before bariatric surgery is low.