The BET Inhibitor GS-626510 Converges with Idelalisib through the Inhibition of Tumor-Macrophage Crosstalk and Cytokine Production in Diffuse Large B-Cell Lymphoma

Background: The recent observation that components of the tumor microenvironment, including M2 tumor-associated macrophages, may define a subgroup of diffuse large B-cell lymphoma (DLBCL) patients with poor prognosis after R-CHOP chemotherapy, has spurred debate on the role of the stroma in this disease. This is particularly true for the germinal center subtype (GCB), which is considered to rely rather on the activity of the proto-oncogene c-Myc and its regulation by bromodomain and extra-terminal (BET) proteins, than on exogenous factors. This c-Myc dependency has led to the preclinical evaluation of BET inhibitors (BETis) in GCB-DLBCL. Subsequent studies have suggested a broader role for these molecules in both GCB- and activated B-cell (ABC)-DLBCL subtypes, due to their ability to modulate additional pathways, such as NF-κB and BCL-2 protein signaling, and led to the assessment of different BETi-based combination strategies. Among them, combinatorial inhibition of BET and class I phosphoinositide 3-kinase (PI3K) has proved to be effective in a wide range of preclinical cancer models including B-cell non-Hodgkin lymphoma, where the δ isoform of the PI3K p110 catalytic subunit plays an active role in driving lymphomagenesis. Since about 55% of GCB-DLBCL cases show activation of the PI3K pathway through loss of PTEN, we assessed the hitherto unexplored combination of a PI3Kδi with a BETi in an in vitro GCB-DLBCL co-culture system. Effects on M2 macrophage-DLBCL crosstalk using the first-in-class PI3Kδi idelalisib (IDELA) are reported either alone, or in combination with the novel BETi, GS-626510.

Methods/Results: We observed that IDELA (50-500 nM) induced a 20%‒30% antitumoral effect in a set of 7 GCB-DLBCL cell lines and primary cultures and that this effect was not abrogated by the presence of M2-polarized macrophages, even though M2 presence in cultures enhanced survival of GCB-DLBCL cells. Gene expression profiling (GEP) revealed that M2 macrophages induced transcription of genes in tumor cells related to cell proliferation, macrophage activation, chemokine activity, cytokine-receptor binding, and cytokine-mediated signaling. Globally, we found an enrichment (>5-fold increase) in cytokine-related gene sets ( p <0.00001). Subsequent analysis of 174 cytokines by an ELISA-based assay confirmed an increased secretion (>2-fold) of 36 cytokines by GCB-DLBCL primary cells, when cultured in the presence of M2 macrophages. Although IDELA significantly reduced some of these transcriptional pathways in DLBCL cell lines and biopsies, it caused limited changes in the cytokine secretion profile. A second GEP analysis of GCB-DLBCL cell lines and primary co-cultures treated with IDELA in combination with the BETi (10-500 nM) showed a significant change in transcriptional repression of BCR signaling, NF-κB pathway, mTOR axis, and inflammatory cytokines. Accordingly, in GCB-DLBCL cells, the drug combination led to a significant decrease in 5 cytokines (CCL13, PPBP, IL-1RN, IL-10, and CCL15) both at mRNA and proteins levels, suggesting an improved disruption of M2 macrophage-tumor crosstalk. Combination treatment resulted in increased apoptotic rates, from 60% to 70% at 24 hours and from 80% to 95% at 72 hours, which were associated with an almost complete downregulation of anti-apoptotic MCL-1 (24 hours). This antitumor effect was found to be synergistic, as illustrated by a mean combination index of 0.38 (range: 0.09-0.82) in 4 GCB-DLBCL cell lines.

Conclusion: In summary, we have characterized the impact of the combination of IDELA and GS-626510 on GCB-DLBCL cells and their crosstalk with the M2 macrophage population, an essential component of the DLBCL lymph node microenvironment.