Circumvention of glucocorticoid resistance in childhood leukemia.

In this study, we determined if in vitro resistance to prednisolone and dexamethasone could be circumvented by cortivazol or methylprednisolone, or reversed by meta-iodobenzylguanidine in pediatric lymphoblastic and myeloid leukemia. As there were strong correlations between the LC50 values (drug concentration inducing 50% leukemic cell kill, LCK) of the different glucocorticoids and median prednisolone/methylprednisolone, prednisolone/dexamethasone and prednisolone/cortivazol LC50 ratios did not differ between the leukemia subtypes, we conclude that none of the glucocorticoids had preferential anti-leukemic activity. Meta-iodobenzylguanidine however, partially reversed glucocorticoid resistance in 19% of the lymphoblastic leukemia samples.

Glucocorticoid receptor alpha, beta and gamma expression vs in vitro glucocorticod resistance in childhood leukemia.

Alternative splicing of the primary glucocorticoid receptor (GR) transcript, resulting in glucocorticoid receptor alpha GRalpha, glucocorticoid receptor beta GRbeta and glucocorticoid receptor gamma GRgamma, may influence glucocorticoid (GC) resistance in childhood leukemia. To test this hypothesis, we determined GRalpha/beta protein and GRalpha/beta/gamma mRNA expression levels in 43 initial acute lymphoblastic leukemia (iALL), 10 initial myeloid leukemia (iAML), 11 relapsed ALL (rALL) samples and one rAML sample. The results were correlated with in vitro GC resistance. GRalpha mRNA correlated with protein expression (rho=0.39-0.56, P<0.05), but the protein to mRNA ratio was median 2.2-fold lower in rALL than in iALL (P<0.05). GRbeta mRNA was median 137-fold lower than GRalpha mRNA and correlated with GRalpha mRNA expression (rho=0.71, P<0.0001). GRbeta could not be detected at the protein level. GRgamma accounted for a median of 2.8% (range 0.95-7.4%) of all GR transcripts. GRalpha (protein and mRNA) and GRbeta (mRNA) expressions or GRalpha/GRbeta ratios did not correlate with in vitro GC resistance in iALL, but GRgamma (mRNA) did (rho=0.52, P=0.007). These results suggest that GRbeta is not involved in GC resistance in childhood leukemia. The association between GRgamma expression and in vitro GC resistance in iALL and the decreased protein/mRNA ratio in rALL, a subgroup resistant to GCs, warrants further exploration.