Abstract
The pathogenesis of primary CNS lymphoma (PCNSL) is dependent upon constitutive activation of nuclear factor-kappa B (NF-kB) signaling. Recurrent oncogenic mutations involving the adapter protein myeloid differentiation primary response gene 88 (Myd88), as well as the B cell receptor (BCR) transmembrane protein partner, cluster of differentiation 79 (CD79B), are established as key drivers of NF-kB activation.
We previously demonstrated that AZ1495, a CNS penetrating inhibitor of IRAK4, potently antagonized the phosphorylation of the p65 subunit of NF-kB, reduced NF-kB target gene expression in a patient-derived xenograft model of MYD88-L265P mutant intracranial CNS large B-cell lymphoma, and delayed tumor growth and dissemination in an OCI-Ly10 bioluminescence model of CNS lymphoma. (Geng et al., ASH# 622, 2016). Combination AZ1495 plus ibrutinib was synergistic in survival prolongation compared to AZ1495 monotherapy, ibrutinib monotherapy or control (p<0.002 for each comparison) in a MYD88-L265P mutant, clinically refractory patient-derived intracranial model.
The goal of this current investigation is to identify the key targets of IRAK4 and BTK pathways that contribute to growth suppression and potentially the emergence of resistance to pharmacologic antagonists of IRAK4 and BTK in CNS lymphoma. Transcriptional profiling via RNA-Seq of CD19+ patient-derived CNS lymphoma cells isolated from intracranial xenografts in cohorts of mice 6 hours after exposure to AZ1495, Ibrutinib, combination AZ1495 plus Ibrutinib, or control gavage was performed to identify in vivo targets of pharmacologic antagonists of IRAK4 and BTK. Among differentially expressed genes with a FDR <0.05, C-X-C motif chemokine ligand 10 (CXCL10) was significantly reduced by monotherapy with both AZ1495 and ibrutinib and was the most strongly suppressed NF-kB target gene with combination therapy (<3-fold compared to control). By contrast, expression of CXCR3, the receptor for CXCL10, was among the most strongly upregulated NF-kB target genes with combination therapy (>2-fold compared to control).
Using ELISA, we determined that cerebrospinal fluid (CSF) from 33 primary CNS lymphoma (PCNSL) patients at diagnosis contained elevated concentrations of CXCL10 compared to patients with non-malignant, non-inflammatory conditions and that CSF from 30 patients with relapsed CNS lymphoma (PCNSL and SCNSL) contained the highest CXCL10. In both cohorts of CNS lymphoma, patients with higher CXCL10 concentrations in CSF, exceeding 650 ng/ml, experienced significantly shorter overall survival. CXCL10 in CSF was also elevated in patients with neuro-inflammatory conditions such as multiple sclerosis.
CXCL10 expression was confirmed in diagnostic specimens of PCNSL by IHC and in situ hybridization, both by lymphoma and by infiltrating myeloid cells. We demonstrated that polarization of murine microglia with IFN-ϒ potently induced expression of CXCL10, consistent with its role in neuro-inflammation. Furthermore, we demonstrated that CXCL10 significantly enhanced by >40% the in vitro survival of patient-derived high-grade B-cell lymphoma cells isolated from a highly refractory secondary leptomeningeal lymphoma.
Using siRNA, we determined that 50% reduction in expression of CXCL10 and/or CXCR3, demonstrated by qRT-PCR and western blot, significantly reduced lymphoma cell survival by >50%, compared to control siRNA, under low (<2%) serum conditions in vitro in MYD88-L265P mutant OCI-Ly10 lymphoma but not in MYD88 wt Raji lymphoma. Essentially identical results were obtained via reduction in the expression of CXCL10 as well as CXCR3 in multiple clones of OCI-Ly10 cells engineered via CRISPR. Disruption of CXCL10 and/or CXCR3 expression in OCI-LY10 was each associated with > 2-fold increase in expression of the apoptotic marker annexin V by tumor cells at 48 hrs compared to CRISPR control under reduced serum conditions.
These results demonstrate a novel, pro-survival role for the CXCL10-CXCR3 axis in a MYD88-mutant aggressive lymphoma, that is regulated by the NF-kB targeting agents AZ1495 and Ibrutinib, and raise the possibility that paracrine signaling via expression of chemokine CXCL10 by CNS resident myeloid cells may contribute to the CNS tropism of subtypes of aggressive lymphoma. Pharmacologic disruption of the CXCL10-CXCR3 axis may have also potential in treatment and/or prevention of CNS lymphomas.
Supported by the Leukemia and Lymphoma Society
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal