Serum pharmacodynamic biomarkers for chronic corticosteroid treatment of children

Abstract

Corticosteroids are extensively used in pediatrics, yet the burden of side effects is significant. Availability of a simple, fast, and reliable biochemical read out of steroidal drug pharmacodynamics could enable a rapid and objective assessment of safety and efficacy of corticosteroids and aid development of corticosteroid replacement drugs. To identify potential corticosteroid responsive biomarkers we performed proteome profiling of serum samples from DMD and IBD patients with and without corticosteroid treatment using SOMAscan aptamer panel testing 1,129 proteins in <0.1 cc of sera. Ten pro-inflammatory proteins were elevated in untreated patients and suppressed by corticosteroids (MMP12, IL22RA2, CCL22, IGFBP2, FCER2, LY9, ITGa1/b1, LTa1/b2, ANGPT2 and FGG). These are candidate biomarkers for anti-inflammatory efficacy of corticosteroids. Known safety concerns were validated, including elevated non-fasting insulin (insulin resistance), and elevated angiotensinogen (salt retention). These were extended by new candidates for metabolism disturbances (leptin, afamin), stunting of growth (growth hormone binding protein), and connective tissue remodeling (MMP3). Significant suppression of multiple adrenal steroid hormones was also seen in treated children (reductions of 17-hydroxyprogesterone, corticosterone, 11-deoxycortisol and testosterone). A panel of new pharmacodynamic biomarkers for corticosteroids in children was defined. Future studies will need to bridge specific biomarkers to mechanism of drug action, and specific clinical outcomes.

Figure 1. Corticosteroid-responsive serum protein biomarkers in children with DMD.

Shown are SOMAscan analyses of cross-sectional groups of corticosteroid-naïve children (DMD; n = 9), corticosteroid-treated children (DMD + CS; n = 5), and healthy children (CT; n = 4). ITGa1/b1: Integrin alpha-I/beta-1 complex. IL22RA2: interleukin-22 receptor subunit alpha2; MMP12: matrix metalloproteinase 12 also known as macrophage metalloelastase; MMP3: matrix metalloproteinase 3 also known as stromelysin 1. Values for each protein are plotted as RFU (relative fluorescent units) and represent the average mean and standard error in each group.

Figure 2. Correlation of SOMAscan and ELISA assays for corticosteroid-responsive proteins.

Serum samples from DMD boys were tested in parallel by the two methods, and values plotted. The number of samples tested was different from protein to protein due the availability of serum samples.

Figure 3. Adrenal steroidal hormones are suppressed by chronic corticosteroid treatment in DMD and IBD children.

(a) Longitudinal analysis of DMD children (n = 9) and age-matched controls (n = 4). (b) Longitudinal analysis of IBD children (n = 11). Values for steroids are plotted in ng/mL and represent the average mean and standard error in each group.