Post-dexamethasone cortisol levels in non-functioning adrenal incidentalomas: a pooled individual participant data analysis
TL;DR
Comparing F-1mgDST distributions between individuals with NFAI and population-based controls and to explore whether alternative F-1mgDST thresholds would meaningfully modify patient classification found no F-1mgDST threshold with an accuracy exceeding 60% for distinguishing controls from patients with NFAI.
Abstract
: Non-functioning adrenal incidentalomas (NFAI) are defined by 1 mg-dexamethasone suppression test (F-1mgDST) cortisol levels < 1.8 µg/dL. However, many individuals with NFAI may exhibit subtle hypercortisolism associated with increased cardiometabolic and skeletal risk. The aims of the present study are to compare F-1mgDST distributions between individuals with NFAI and population-based controls and to explore whether alternative F-1mgDST thresholds would meaningfully modify patient classification. This is a pooled individual participant data analysis of 7 studies evaluating cortisol secretion in patients with type 2 diabetes, osteoporosis and/or fragility fractures, hypertension, or adrenal incidentalomas. The study included 1681 participants (915 NFAI, 766 controls) with F-1mgDST < 1.8 µg/dL. Data included age, sex, body mass index (BMI), F-1mgDST, and adrenocorticotroph hormone (ACTH). We assessed the between-group differences in F-1mgDST and ACTH and the accuracy of different F-1mgDST thresholds for identifying NFAI. Compared to controls, patients with NFAI had lower age and ACTH levels (62.1 ± 9.8 and 59.8 ± 12.0 years, 25.1 ± 14.1 pg/mL and 16.3 ± 10.4, respectively, p < 0.001 for all comparisons), but higher BMI and F-1mgDST values (28.1 ± 4.9 and 29.0 ± 5.0 kg/m2, 1.04 ± 0.37 µg/dL and 1.13 ± 0.39 µg/dL, respectively, p < 0.001 for all comparisons) and were more less frequently females (48.0% and 38.6%, respectively, p < 0.001). A propensity score analysis performed on 730 patients and 730 age-, BMI- and gender-matched controls confirmed these findings. We did not identify any F-1mgDST threshold with an accuracy exceeding 60% for distinguishing controls from patients with NFAI. The F-1mgDST threshold the best balance between sensitivity (54.8%) and specificity (57.2%) was set at 1.1 µg/dL (30 nmol/L), although with a low accuracy of 55.8%. Individuals with NFAI have higher F-1mgDST levels than controls. More than 40% overlap exists in F-1mgDST values between NFAI and controls.