Cycle characterisation and clinical complications in patients with cyclic Cushing's syndrome: insights from an international retrospective cohort study

Abstract

Background: Cyclic Cushing's syndrome (cCS) features fluctuating cortisol secretion, often causing diagnostic errors or delays, and possibly poorer outcomes. We aimed to identify unpublished cCS cases to characterise clinical challenges and guide strategies for improving outcomes by characterising cycle patterns, peak frequency, and evaluating complications.

Methods: This was a retrospective observational study at 43 endocrine centres in 21 countries, including patients with confirmed Cushing's syndrome showing two or more hypercortisolaemic peaks and one or more spontaneous eucortisolaemic or hypocortisolaemic trough. Data included both clinical (eg, comorbidities and physical signs of cortisol excess) and biochemical (eg, screening and confirmatory tests) parameters, imaging, treatment, complications, and outcomes.

Findings: Between Dec 1, 2023 and Feb 2, 2025, 116 potentially eligible patients were identified and 110 were included. Most patients were female (84 [76%] of 110 patients), with a median age at diagnosis of 44·0 years (IQR 31·8-58·3). cCS origin was pituitary in 70 (64%), ectopic in 25 (23%), adrenal in three (3%), and occult in 12 (11%). Cyclicity was primarily determined by 24 h urinary free cortisol, with median peaks of 7·40 × ULN (range 0·44-299) and troughs of 0·31 × ULN (0·02-0·98). The median peak count was 3·0 (IQR 2·0-4·0), mostly (55 [86%] of 64 patients) occurring at irregular intervals, and was most frequent and pronounced in ectopic cCS. Symptoms worsened in 87 (81%) of 108 patients during peaks and improved in 79 (74%) of 107 patients during troughs; 31 (28%) of 110 patients had spontaneous adrenal insufficiency. Bilateral inferior petrosal sinus sampling (BIPSS) was performed during troughs in 14 patients (18% of the 78 procedures done). Imaging missed tumours in 35 (32%) of the 110 patients, and nine (8%) underwent unwarranted surgeries at the wrong anatomical site due to misclassification. After 5·8 years (IQR 2·6-10·5) median follow-up, 55 (50%) of 110 patients had complete biochemical surgical remission, seven (6%) had spontaneous remission, 22 (20%) were medically controlled, six (5%) had partial remission, 11 (10%) remained uncontrolled, nine (8%) were lost to follow-up. During the entire observation period, 3% (3/110) died. Delayed diagnosis (45 [41%] of 110 patients) and therapy (47 [43%]) were also observed.

Interpretation: Even in specialised centres, cCS diagnosis and management remain challenging with high rates of spontaneous adrenal insufficiency, inappropriate surgeries, and poor outcomes. Ectopic cCS showed the most frequent and severe peaks. These findings might help to guide imaging localisations or the timing of BIPSS in patients with active occult ACTH-dependent cCS. Hypercortisolism needs to be biochemically confirmed before BIPSS to enable correct tumour localisation. Patients with suspected or proven cCS should be equipped with salivary cortisol collection kits to capture dynamic changes as well as being prescribed glucocorticoids to be used as a precaution.

Funding: None.