Objective: Adrenal insufficiency (AI) is a potentially life-threatening condition. It is known that high doses of inhaled corticosteroids (ICS) can induce systemic adverse effects. Currently, there are no data on the prevalence of AI associated with the use of ICS. This study aimed to investigate the prevalence and clinical presentation of AI (associated or not associated with exogenous Cushing's syndrome) in patients who were prescribed ICS by French physicians during the period 2000-5.
Methods: All metropolitan French paediatricians, endocrinologists, pulmonologists and intensive care physicians (n = 11 783) were mailed questionnaires requesting information regarding cases of AI associated or not associated with exogenous Cushing's syndrome between 2000 and 2005. Data collected included patient demographics, oral corticosteroid or ICS used during the year preceding the diagnosis of AI, underlying condition(s), concomitant treatment(s), presenting clinical signs and symptoms, results of laboratory investigations and outcome. The French pharmacovigilance database was screened for spontaneous reports to determine the frequency of AI associated with the use of ICS, using the capture-recapture method.
Results: Forty-six cases of AI were identified. Twenty-three cases presented with clinical symptoms of AI alone and 23 with exogenous Cushing's syndrome. ICS prescribed were fluticasone propionate (n = 24), budesonide (n = 12) and beclometasone dipropionate (n = 5). In 82% (n = 32) of cases for which data were available, ICS were prescribed at high doses. A potential drug interaction was found in 12 cases. Thirteen cases of AI were identified in the French pharmacovigilance database, one of which was common with the questionnaire survey. The capture-recapture method provided an estimation of 598 (95% CI 551, 648) cases of AI associated with the use of ICS for the 2000-5 period in France.
Conclusion: The results of this study confirm the occurrence of adrenal insufficiency in patients treated with ICS. Although the prevalence of ICS-induced AI reported in this study is low, the likelihood of under-diagnosis underlines the need to consider this risk in patients when prescribing these drugs.