Dexamethasone Is More Potent Than Prednisone in B-Lymphoblastic Leukemia Despite Dose Adjustment

Glucocorticoids are key to combination chemotherapy regimens for B-lymphoblastic leukemia (B-ALL), with response to glucocorticoids alone predicting outcomes. Dexamethasone (dex) and prednisone (pred) are both used for induction remission in children with B-ALL depending on age. Dex is more potent than pred, which also causes more toxicities, and pred is typically given at a higher dose to adjust for this difference in potency. Current Children’s Oncology Group protocols use 6 times more pred to equal dex, but previous studies have suggested that with a pred/dex dose ratio <7, dex is still more effective than pred (Inaba and Pui, Lancet Oncol 2010). To determine how we might improve pred efficacy and subsequently improve outcomes particularly for patients with B-ALL receiving pred, we asked 3 primary questions: 1) How much more potent is dex compared to pred, 2) Why is dex more potent than pred, and 3) How might we increase pred potency specifically in B-ALL without increasing toxicities.

We first determined the pred/dex dose ratio needed to achieve the same effect on cell viability in both cell lines and ex vivo primary patient specimens. We tested 15 B-ALL cell lines and 10 primary specimens with serial dilutions of dex or pred for 72 hours before assessing viability with a resazurin-based reagent (PrestoBlue). The pred EC50 to dex EC50 ratio was greater than 6 for the cell lines (mean 10) and significantly greater than 6 in primary specimens (mean 17, p = 0.02). In primary specimens, this difference was driven by a subset of 3 specimens that were highly sensitive to dexamethasone (EC50 <1 nM, mean pred/dex EC50 ratio 34). However, after removing these specimens, the ratio remained significantly greater than 6 (mean 9, p = 0.04). These data indicate that dex is, on average, 9-10 times more potent than pred, but that the optimal dose ratio is variable between specimens.

Glucocorticoids work primarily by binding the glucocorticoid receptor (GR), a ligand-activated transcription factor, to regulate genes leading to cell death. To better understand why dex is more potent than pred, we measured differences between dex- and pred-induced gene regulation. Using the 3 highly dex-sensitive primary specimens, we compared equal (20 nM) doses after 24-hour treatment using RNA-seq. Although dex and pred regulate similar sets of genes by pathway analysis, dex upregulates and downregulates genes more strongly, and pred only regulates a subset of dex-regulated genes. In addition, dex regulates some effector genes (genes whose regulation drives glucocorticoid-induced cell death) more strongly, including TXNIP and BCL2L11. To determine whether regulation of these genes is equalized by dose adjustment, we performed RNA-seq on a different patient specimen (pred/dex EC50 ratio ~5) treated with 8-fold more pred than dex. Even with dose adjustment, dex still regulated genes more strongly than pred, including the effector genes TXNIP and BCL2.

To understand why dex more strongly regulates genes, we compared DNA binding affinity using Mechanically Induced Trapping of Molecular Interactions (MITOMI 2.0, Fordyce et al, Nat Biotechnol 2010). This demonstrated that dex-bound GR has a higher affinity for GR binding sites than pred-bound GR. Using ChIP-seq, we also observed that GR had a higher enrichment at sites of GR binding in the presence of dex compared to pred, even with dose adjustment.

These studies provide mechanistic insights that can be exploited to improve B-ALL treatment. We show that dex is more potent than pred likely due to increased affinity of dex-bound GR for DNA causing increased genomic occupancy and subsequently enhancing gene regulation. Importantly, dex still regulates genes more strongly even at 8-fold lower concentrations than pred. We hypothesize that enhancing pred-induced gene regulation specifically in B-ALL cells would compensate for this difference in potency while maintaining a milder side effect profile. This can potentially be achieved by inhibiting the leukocyte-restricted PI3Kδ (Zimmerman et al, Cancers 2024; Kruth et al, Blood 2017), which enhances DNA binding by GR, or by inhibiting AURKB (Poulard et al, PNAS 2019), which is overexpressed in relapsed B-ALL and also enhances gene regulation by GR. Another strategy could be to target pathways that are insufficiently regulated by pred compared to dex, such as enhancing redox regulation by TXNIP or promoting apoptosis by BCL2 inhibition with venetoclax.