Background/Purpose: Steroids are widely used in clinical practice for treating systemic lupus erythematosus (SLE), and can have a strong effect on immune cells in Whole Blood (WB), especially B cells and T-cells. Steroids are often used as background medication during clinical trials of SLE drugs and therefore can easily confound efficacy end points associated with therapies targeting B cells and T cells in the trials. Moreover, response to steroids might vary from individual to individual and might not completely agree with protocol steroid usage. Our goal is to create a gene signature as an objective measure of steroid response that can be used in SLE trials to aid in the interpretation of potential confounding steroid effects.

Methods: We conducted 5 in vitro experiments where PBMCs from normal healthy volunteers were treated with prednisolone. Differential gene expression analysis identified genes that were up regulated by prednisolone treatment, and 64 common genes were identified as a potential prednisolone gene signature. Composite scores of the prednisolone signature were calculated using the Single-sample Gene Set Enrichment Analysis (ssGSEA) algorithm (Barbie et al., 2009) for WB samples treated in vitro with prednisolone. The same was done for whole blood samples from normal healthy subjects who had in vivo administration of either a selective glucocorticoid receptor agonist (SEGRA) or prednisolone in a phase I study. The prednisolone signature was further investigated in baseline WB samples of SLE patients from a multicenter, exploratory, phase IIb randomized, double-blind, placebo controlled trial, and signature scores were compared to neutrophil counts and protocol steroid usage.

Results: Both the in vitro and in vivo studies showed the high specificity of the prednisolone signature. When applied to WB gene expression data from SLE patients, the prednisolone signature scores were positively correlated with neutrophil counts (CBC) in WB, and negatively correlated with in vitro derived T cell and B cell signature scores. Both of these observations can be explained by the steroid mechanism of action. Steroid signature scores did not completely correlate with protocol steroid usage.

Conclusion: We identified a highly specific steroid gene signature that can be used to account for effects of steroids in global clinical trials where PAX gene tubes are routinely collected.

Figure 1: Prednisolone signature scores increased in a dose responsive manner in GR compound trial.

Figure 2: Prednisolone signature scores significantly correlate with netrophil counts in WB from baseline SLE patients