PGx6951: Analysis of influence of GLCCI1 variant rs37973 on inhaled corticosteroid response in COPD patients treated with fluticasone furoate in HZC112206 and HZC112207

Trial description: Following the 2011 literature report associating polymorphism rs37972 with response to inhaled glucocorticoids in patients with asthma, van den Berge et al. conducted an evaluation of the effect of rs37972, a promoter polymorphism in the gene GLCCI1 (glucocorticoid-induced transcript 1), on response to glucocorticoid treatment in patients with COPD. They studied 63 COPD subjects recruited in the Netherlands who were treated with fluticasone or fluticasone plus salmeterol. Efficacy was measured using change from baseline to 3 months in forced expiratory volume in one second (FEV1). Van den Berge et al. found a significant association between rs37972 and change in FEV1. Specifically, subjects with two copies of the normal allele had a higher response than subjects with either 1 or 2 copies of the variant allele (p-values were 0.006 and 0.02, respectively) and the results were consistent with the variant allele having an additive genetic effect. Van den Berge et al. also saw an association between rs37972 and residual volume (RV). Note that in subjects of European ancestry the genetic marker rs37972 is highly correlated with rs37973 such that, for all practical purposes, the two markers can be analyzed interchangeably.With the development of fluticasone furoate (FF) plus vilanterol for the treatment of COPD, there are a number of opportunities within GSK to investigate whether the association of rs37972 with response to inhaled glucocorticoid treatment can be observed within GSK clinical studies. In particular, HZC112206 and HZC112207 are two studies in which over 530 patients were treated with inhaled FF, a steroid, including approximately 455 self-reported white subjects. This analysis is an attempt to follow up the findings of van den Berge et al., however, HZC112206 and HZC112207 do not include measures of RV. Because reductions in RV result in increases in forced vital capacity (FVC), at least in some COPD patients, the association between rs37973 and FVC will be evaluated. Specifically, the analysis objectives are: 1) Primary: Test for an association between rs37973 and change in FEV1 from baseline to Week 12 of treatment with FF in self-reported non-Hispanic white subjects, adjusting for variables nominally associated with change in FEV1. 2) Secondary: If an association is observed between rs37973 and change in FEV1 from baseline to Week 12 of treatment within subjects on FF treatment, then test for association between rs37973 and treatment effect (steroid response versus placebo) in non-Hispanic white subjects treated with FF. 3) Secondary: Test for an association between rs37973 and change in FVC from baseline to Week 12 of treatment within FF self-reported non-Hispanic white subjects, adjusting for variables nominally associated with change in FVC.The analysis will include an assessment of the comparability of studies HZC112206 and HZC112207 to determine if subject level pooling of data is contraindicated due to heterogeneity. If the two studies cannot be combined, analyses will be performed in each study separately and the results combined using meta-analysis techniques. The effect of rs37973 on change in FEV1 in FF-treated subjects will be evaluated using the variables associated with change in FEV1 as covariates in a linear model. If a significant association is found between rs37973 and change in FEV1 among FF treated subjects, a treatment (that is, placebo versus FF) by genotype interaction will be tested for and box plots for change in FEV1 by genotype within treatment group (placebo or FF) will be generated.