Influence of short-term glucocorticoid therapy on regulatory T cells in vivo.

Background: Pre- and early clinical studies on patients with autoimmune diseases suggested that induction of regulatory T(T(reg)) cells may contribute to the immunosuppressive effects of glucocorticoids (GCs).
Objective: We readdressed the influence of GC therapy on T(reg) cells in immunocompetent human subjects and naïve mice.
Methods: Mice were treated with increasing doses of intravenous dexamethasone followed by oral taper, and T(reg) cells in spleen and blood were analyzed by FACS. Sixteen patients with sudden hearing loss but without an inflammatory disease received high-dose intravenous prednisolone followed by stepwise dose reduction to low oral prednisolone. Peripheral blood T(reg) cells were analyzed prior and after a 14 day GC therapy based on different markers.
Results: Repeated GC administration to mice for three days dose-dependently decreased the absolute numbers of T(reg) cells in blood (100 mg dexamethasone/kg body weight: 2.8±1.8×10(4) cells/ml vs. 33±11×10(4) in control mice) and spleen (dexamethasone: 2.8±1.9×10(5)/spleen vs. 95±22×10(5)/spleen in control mice), which slowly recovered after 14 days taper in spleen but not in blood. The relative frequency of FOXP3(+) T(reg) cells amongst the CD4(+) T cells also decreased in a dose dependent manner with the effect being more pronounced in blood than in spleen. The suppressive capacity of T(reg) cells was unaltered by GC treatment in vitro. In immunocompetent humans, GCs induced mild T cell lymphocytosis. However, it did not change the relative frequency of circulating T(reg) cells in a relevant manner, although there was some variation depending on the definition of the T(reg) cells (FOXP3(+): 4.0±1.5% vs 3.4±1.5%*; AITR(+): 0.6±0.4 vs 0.5±0.3%, CD127(low): 4.0±1.3 vs 5.0±3.0%* and CTLA4+: 13.8±11.5 vs 15.6±12.5%; * p<0.05).
Conclusion: Short-term GC therapy does not induce the hitherto supposed increase in circulating T(reg) cell frequency, neither in immunocompetent humans nor in mice. Thus, it is questionable that the clinical efficacy of GCs is achieved by modulating T(reg) cell numbers.