CSF1R and BTK Inhibitions As Novel Strategies to Disrupt the Dialogue between Mantle Cell Lymphoma and Macrophages

The aggressive clinical behavior of mantle cell lymphoma (MCL) and its short-term response to treatments highlight a need for novel options. The microenvironment strongly controls MCL cell survival, proliferation and chemoresistance, nevertheless little is known regarding the interactions that occur in the tumor niches. Here, we studied the interplay between primary MCL cells (n=55) and macrophages and identified mechanism-based targeted strategies to disrupt this dialogue.

Using ex vivo co-cultures of peripheral blood MCL cells and monocytes (healthy donors), we showed that monocytes greatly improved MCL cell survival, whereas MCL cells poorly survived when cultured alone (co-culture, 85.9%; alone, 6.5%; D7; n=17; p<0.001). We also demonstrated that monocytes supported the proliferation of conventional (n=7) and blastoid (n=4), but not leukemic non-nodal MCL subtypes (n=5) (p<0.05).

During co-culture, primary MCL cells polarized monocytes into MCL-associated-macrophages (MϕMCL). To define the nature of MϕMCL, we studied their phenotype and function with 12 markers differentially expressed in M1 and M2 macrophages (CD14, CD68, CD163, CD11b, CD86, HLA-DR, CD16, CD23, IL10, IL6, IGF1, BAFF). The analysis showed that even though MϕMCL distinctly clustered from both M1 and M2, they shared more similarities with M2 (CD163^high, CD11b^low, IL10⁺, IGF1⁺).

To determine how MCL cells polarized monocytes into CD163^high MϕMCL, we analyzed the expression of known M2-polarizing factors and observed that CSF1 and IL10 were overexpressed in MCL cells compared to normal B cells (GEP, p<0.01). CSF1 and IL-10 were detected at the protein level in co-culture (ELISA), and CSF1, but not IL-10, neutralization significantly blocked MϕMCL polarization ex vivo. Previous studies reported modulations of the MCL secretome upon BTK inhibition. Here, we demonstrated that both CSF1 and IL-10 were inhibited upon ibrutinib treatment (0.5 µM) in sensitive MCL cells, whereas no modulation was observed in ibrutinib-resistant cells. Reduction of the CSF1 level by ibrutinib consequently inhibited M2-like polarization and resulted in the inhibition of MϕMCL-dependent tumoral survival (n=8/14) and proliferation (n=4/4). Nevertheless, 6 MCL samples cocultured with MϕMCL appeared resistant to ibrutinib. To test whether targeting the CSF1R could bypass this resistance, we treated MCL/MϕMCL coculture with the inhibitor GW2580. We showed that CSF1R inhibition efficiently reduced the viability of ibrutinib-resistant MCL cells. Moreover, the combination in low concentrations of ibrutinib/GW2580 (125nM) induced a supra-additive effect in ibrutinib-sensitive samples.

In vivo, we showed higher levels of CSF1 and IL-10 in the plasma of MCL patients (n=28) compared to match-aged HD (p<0.01). We also showed that CD163 was overexpressed at the surface of circulating monocytes in MCL compared to HD (n=32; p < 0.05), which was consistent with the CD163-inducing properties of CSF1 and IL-10. Taking advantage of a local clinical trial (NCT02558816), we evaluated modulations of CSF1 and IL-10 plasma concentrations as well as CD163 expression on monocytes from 8 patients treated with ibrutinib. We observed a reduction of CSF1 and IL-10 as well as inhibition of CD163 expression in 7/8 patients after 8 days (D8) of treatment (median CD163+ reduction, - 58%). We performed longitudinal follow-up of 4 patients treated with ibrutinib, 3 were characterized by a dramatic reduction in CD163 at D8 and achieved a durable complete response. In contrast, 1 patient who displayed an increase in CD163 at D8, progressed upon treatment.

In conclusion, through secretion of IL-10 and CSF1, MCL cells polarized monocytes into M2-like macrophages (MϕMCL), which in turn support tumor survival and proliferation. Ibrutinib counteracts the MCL/MϕMCL dialogue through inhibition of CSF1 production and, consequently, impairs the MϕMCL-dependent expansion of ibrutinib-sensitive MCL cells. Of note, our in vivo retrospective analysis highlights that CSF1, IL-10 and CD163 modulations might be early markers of ibrutinib response. A larger cohort of MCL patients treated with ibrutinib is now necessary to confirm the strength of this soluble and cellular signature. Lastly, targeting the CSF1R is an alternative to disrupt the MCL/MϕMCL pro-tumoral dialogue, especially for ibrutinib-refractory patients for which no therapeutic alternative is available.