Adipokine profile in glucocorticoid-treated patients: baseline plasma leptin level predicts occurrence of lipodystrophy

Context Glucocorticoid therapy may result in adipose tissue redistribution of unknown pathophysiology. Objectives To evaluate the effects of glucocorticoids on adipokine levels and adipose tissue inflammation. To compare the results in patients with or without glucocorticoid-induced lipodystrophy (GIL) after 3 months of therapy. Design and setting Prospective monocentric study. Patients Adult patients initiating systemic, high-dose prednisone therapy for at least 3 months. Blood samples and subcutaneous abdominal adipose tissue biopsies were collected at baseline and month 3. The presence of GIL after 3 months of therapy was assessed using standardized photography. Results Thirty-two patients were enrolled. Blood samples and subcutaneous abdominal adipose tissue were available at baseline and month 3 for 30 patients [median age: 61 (38-79) years, 77% women]. Among those 30 patients, 15 were classified as GIL+ and 15 were GIL- at month 3. Between baseline and month 3, adiponectin and leptin levels increased in the overall population while the level of resistin remained unchanged. At baseline, leptin level was higher [19.3 (8.3-31.1) vs 4.5 (2.4-11.3) mu g/l, P = 0.006] and resistin level lower [7.1 (6.3-12.4) vs 10.4 (8.0-21.7) mu g/l, P = 0.05] in GIL+ than in GIL- patients. Baseline leptin level was predictive of GIL occurrence. Receiver operating characteristic curve analysis demonstrated that the best diagnostic accuracy was obtained with a baseline leptin cut-off of 5.9 mu g/l (sensitivity: 93%, specificity: 60%). At month 3, leptin and adiponectin levels increased more in the GIL+ than in the GIL- group, as did the number of anti-inflammatory M2 macrophages in subcutaneous abdominal fat. Conclusion Glucocorticoid-induced lipodystrophy is associated with a different adipokine profile both before and after glucocorticoid therapy. Serum leptin level prior to glucocorticoid therapy is highly predictive of GIL occurrence.