Endocrine Abstracts

JOINT2972

Background: Adrenal insufficiency (AI) under glucocorticoid replacement therapy, adrenocortical tumors with mild autonomous cortisol secretion (MACS), and Cushing syndrome (CS), are characterized by different patterns of dysregulation of cortisol circadian rhythm. Despite the relationship between cortisol and energy metabolism is established, little is known about circadian-specific metabolic derangements in any of these conditions.

Aim: To characterize circadian fluctuations of acylcarnitines, sphingomyelins, glycerophospholipids, amino acids, and biogenic amines in healthy subjects (HS), and in patients affected by AI, MACS, and CS in dried blood spots (DBS).

Methods: We enrolled patients with AI (n=8) under dual-release hydrocortisone, and subjects with MACS (n=12) and CS (n=9) at diagnosis. HS (n=10) were drawn from the general population. All subjects underwent a 7-days standardized isocaloric Mediterranean diet. On day 7th, subjects collected DBS samples 30 minutes before and 2 hours after each meal (breakfast, lunch, and dinner), and at 11 pm. We measured 21 amino acids, 21 biogenic amines, 40 acylcarnitines, 15 sphingomyelins, and 90 glycerophospholipids by a targeted metabolomics LC–MS/MS method previously validated for DBS. Data were explored by a two-step approach. Firstly, a BORUTA algorithm trained on disease groups separation was performed. According to an all-features-matter approach, permutations of features were tested in a random forest classifier, pruning out uninformative features and keeping features describing differences among groups. Secondly, Factor Analysis (FA) was applied to reduce data dimensionality. To characterize the circadian fluctuation of interesting molecular patterns, the linear combination of selected molecule concentrations and FA-derived coefficients was evaluated along the 7 time-points.

Results: Two molecular patterns were highlighted. In pattern A, phosphatidylcholine (PC)-ae-C34:2, PC-ae-C34:3, PC-aa-C34:2, PC-aa-C36:2 and PC-aa-C36:0 differentiated HS from patients with AI, MACS, and CS. Overall, higher concentrations were detected in HS with respect to disease groups, with maximum difference at 2 hours after lunch (P<0.05). In pattern B, PC-aa-C38:4, PC-aa-C36:4, lysoPC-a-C20:4, PC-aa-C36:5 and Met differentiated HS and patients with AI from those with MACS and CS. In particular, patients with CS and MACS showed higher concentrations than HS and patients with AI, with maximum differences at 2 hours after lunch and 30 minutes before dinner (P<0.05 for both). At the same time-points, patients with AI had lower levels of compounds of pattern B than HS (P<0.05).

Conclusions: The phosphatidylcholine system is predominantly affected in different states of glucocorticoid replacement and cortisol excess. Dysregulation is mostly evident in the afternoon. Phosphatidylcholines appear relevant biomarkers of cortisol-related metabolic alterations.