Glucocorticoid receptor gene polymorphisms and glucocorticoid resistance in inflammatory bowel disease: a meta-analysis.

BACKGROUND

Studies investigating the associations between glucocorticoid receptor gene polymorphisms and glucocorticoid resistance in inflammatory bowel disease report conflicting results.

AIMS

We conducted a meta-analysis to assess the possible association between the three most commonly investigated glucocorticoid receptor gene (ER22/23EK, N363S, and BclI) polymorphisms and glucocorticoid resistance in inflammatory bowel disease.

METHODS

Articles evaluating the effect of ER22/23EK, N363S, and BclI gene polymorphism on glucocorticoid resistance in inflammatory bowel disease were identified from 1950 to February 2012. After extraction of relevant data, meta-analyses were performed to assess the association between glucocorticoid receptor gene polymorphisms and glucocorticoid resistance in inflammatory bowel disease.

RESULTS

A total of five eligible studies with 942 cases were included. Our analysis showed that ER22/23EK polymorphisms were not associated with glucocorticoid resistance in inflammatory bowel disease [GG versus GA + AA: odds ratio (OR) = 0.58, 95 % confidence interval (CI) 0.16-2.08]. In N363S polymorphisms, AG + GG allele showed no significant effect on glucocorticoid resistance in inflammatory bowel disease compared with AA allele (OR = 1.19, 95 % CI 0.33-4.30). In BclI polymorphisms, there was also no association of CG + GG allele with glucocorticoid resistance (CC versus CG + GG: OR = 1.22, 95 % CI 0.70-2.13). For Crohn's disease (CD) and ulcerative colitis (UC), no statistically significant associations between these three single-nucleotide polymorphisms (SNPs) and glucocorticoid resistance were found. The shape of the funnel plot did not detect publication bias.

CONCLUSIONS

The current meta-analysis found no evidence that glucocorticoid receptor gene polymorphisms (ER22/23EK, N363S, and BclI) are associated with glucocorticoid resistance in inflammatory bowel disease treatment. However, this meta-analysis is underpowered for relatively large effect sizes in some SNPs. More well-designed cohort studies should be conducted to fully characterize such an association.