Key Points• In vivo PD-L1 blockade prevents CLL development in the Em-TCL1 adoptive transfer model. • In vivo PD-L1 blockade normalizes T-cell and myeloid cell populations and immune effector functions.Blockade of the programmed cell death 1 (PD-1)/programmed death-ligand 1 (PD-L1) immune checkpoint augments antitumor immunity and induces durable responses in patients with solid cancers, but data on clinical efficacy in leukemias are sparse. Chronic lymphocytic leukemia (CLL) is associated with a tumor-supportive microenvironment and a dysfunctional immune system, as shown by "exhausted" T cells, defective immunologic synapse formation, and immunosuppressive myeloid cells. These defects involve aberrant expression of PD-L1 and are closely mirrored in the Em-TCL1 mouse model for CLL. In this study, we treated mice after adoptive transfer of Em-TCL1 CLL with PD-L1-blocking antibodies, which prevented CLL development and was accompanied by a reactivation of immune effector functions. This included restoration of mature macrophages and major histocompatibility complex class II-expressing dendritic cells and prevention of aberrant and exhaustion-like T-cell phenotypes. In addition, PD-L1 blockade restored CD8 T-cell cytotoxicity and immune synapse formation and normalized T-cell cytokines and proliferation ex vivo and in vivo. Our data demonstrate that early PD-L1 blockade effectively corrects leukemia-induced immune dysfunction and thus prevents CLL development in mice. Targeting PD-L1/PD-1 interactions should therefore be further explored in clinical studies with CLL patients, ideally in combination with novel compounds to help eliminate CLL. (Blood. 2015;126(2):203-211)

Key Points• PD-L1/PD-1-mediated CD8T-cell dysfunction develops with CLL in different organs, and similarities to agingrelated immune defects exist.• PD-1 1 normal T cells have markedly different effector functions than PD-1 1 CLL T cells.T-cell defects, immune suppression, and poor antitumor immune responses are hallmarks of chronic lymphocytic leukemia (CLL), and PD-1/PD-L1 inhibitory signaling has emerged as a major immunosuppressive mechanism. However, the effect of different microenvironments and the confounding influence of aging are poorly understood. The current study uses the Em-TCL1 mouse model, which replicates human T-cell defects, as a preclinical platform to longitudinally examine patterns of T-cell dysfunction alongside developing CLL and in different microenvironments, with a focus on PD-1/PD-L1 interactions. The development of CLL was significantly associated with changes in T-cell phenotype across all organs and function. Although partly mirrored in aging wild-type mice, CLL-specific T-cell changes were identified. Murine CLL cells highly expressed PD-L1 and PD-L2 in all organs, with high PD-L1 expression in the spleen. CD3

Background CLL-induced severe T-cell dysfunction and ineffective anti-tumor immune-responses are hallmarks of the disease, but the specific interactions remain poorly understood. The PD-1/PD-L1 axis is an important mediator of T-cell dysfunction in solid tumors, and we have previously demonstrated that T cells from CLL patients exhibit impaired immunological synapse (IS) formation, predominantly mediated by PD-L1 (CD274) expression on CLL cells. We have also shown that the corresponding T-cell ligand PD-1 (CD279) is also upregulated, probably as a result of chronic antigenic stimulation, and that T cells have similarities to exhausted T cells observed in the context of chronic viral infection. Recent studies demonstrated that ibrutinib has impressive clinical activity in CLL, and mechanisms of action include irreversible binding of essential components of both B-cell- and T-cell-receptor signaling and interactions with the tumor microenvironment. Modulation of PD-1/PD-L1 interactions might therefore be an additional potential mode of action of this drug. Using the well-established Eμ-TCL1 (TCL1) mouse model of CLL, our aims were to demonstrate that (1) altered expression of PD-L1 on CLL cells and PD-1 on T cells and CLL are causally related, (2) the second ligand of PD-1, PD-L2, is also involved in mediating T-cell dysfunction, (3) T-cell effector function and IS formation are directly linked to PD-1 expression and (4) PD-1 associated in vivo T-cell responses can be modulated by treatment with ibrutinib. Methods As we have previously demonstrated that the spleen is the major organ of disease and representative of T-cell changes in peripheral blood and lymph nodes, experiments were performed on spleens from young TCL1 and wild-type (WT) C57Bl/6 mice with established CLL after adoptive transfer (AT) of syngeneic CLL cells (n=10), and on matched litter-mates after AT of healthy mouse B cells (n=10). An additional 12 mice were randomized to treatment with 25 mg/kg/d ibrutinib in 10% HP-β-CD, vehicle control, or sterile water, all administered by gavage, three weeks after AT of syngeneic CLL cells, and sacrificed 20 days later at a pre-defined endpoint. Multicolor flow cytometry was used to characterize T-cell subsets, expression of PD-1, PD-L1 and PD-L2 and T-cell effector function. Entire population CD8 T cells, PD-1+ve and PD-1-ve CD8 T cells were flow-sorted and used in IS formation assays with healthy murine B cells as antigen-presenting cells. Results Our previous studies using aged TCL1 mice and age-matched WT controls indicated that CLL-related PD-1 upregulation on antigen-experienced CD44+ CD8 T cells is masked by aging. However, PD-1 expression could also be induced in young TCL1 and WT mice by AT of CLL cells but not healthy B cells, suggesting a causal relationship with disease. Both PD-L1 and PD-L2 surface expression on CLL B cells were significantly increased compared to healthy B cells. Using TCL1 mice at early stages of CLL development when a healthy CD19+ B-cell population is still present, we were able to confirm that PD-L2 expression is a unique feature of CLL cells, with PD-L2 being virtually absent on healthy B cells. We next compared effector function and the ability to form IS of PD-1+ve and PD-1-ve antigen-experienced CD44+ T-cell subsets in mice with CLL. While proliferation was equally impaired in these subsets, they were both able to degranulate but generally failed to localize granzyme B to the IS. Although subsets produced some IL2/TNFα/IFNγ cytokine responses, PD-1+ve cells had significantly impaired TNFα and slightly impaired IL2 and IFNγ production, and a highly significant impaired ability to form IS compared to PD-1-ve cells. Treatment with ibrutinib reduced PD-1 expression on antigen-experienced CD44+ CD8 T cells and promoted stronger IFNγ production of entire population CD8 T cells, but failed to restore proliferation and granzyme B relocation to the IS. Conclusion Our in vivo data suggests that CLL and PD-1/PD-L1-mediated T-cell dysfunction are causally related, but that phenotypic and functional T-cell changes are not absolute and might be at least partly reversible by ibrutinib treatment. We also show that the second ligand of PD-1, PD-L2, is also a critical mediator of PD-1 associated T-cell dysfunction in CLL. Disclosures: Riches: Celgene: Research Funding. Gribben:Celgene: Research Funding; Pharmacyclics: Honoraria; Roche: Honoraria.

Background: Clinical studies have demonstrated that targeted immunotherapy using PD-1/PD-L1 antibodies induces tumor regression and prolongs disease stabilization in advanced solid cancers. Data on the clinical efficacy in hematological malignancies is largely missing, even though PD-L1/PD-1 interactions have been described as major mediators of immune dysfunction in several leukemias and lymphomas. They are therefore ideal to study if PD-L1/PD-1 blockade has the potential to control disease by restoring anti-tumor immune responses. Several groups showed that chronic lymphocytic leukemia (CLL) provokes immune evasion via PD-L1/PD-1 inhibitory signaling, and that this is very closely mirrored in the Eµ-TCL1 (TCL1) murine model for CLL. Our recent data suggest that in this model, aberrant PD-L1 expression in myeloid cells contributes to the immune defect in CLL. We further demonstrated that the T cell and myeloid cell immune defects in ageing leukemic mice can be induced in young wild-type (WT) mice by adoptive transfer (AT) of murine CLL. In the current study, we used the AT model to test if in vivoPD-L1 blockade corrects leukemia-induced cellular immune dysfunction in myeloid and T cells and enhances anti-tumor immunity. Methods: WT mice transplanted with 4x107 TCL1 splenocytes were randomized to treatment with 10 mg/kg α-murine-PD-L1 (n=15) or isotype antibody (n=10), which was administered i.p. every 3 days starting 1 day after AT, and sacrificed 31 days later. Matched non-transplanted WT mice (n=6) served as additional controls. Immune cell subsets, expression of immune checkpoint markers and T cell effector functions were analyzed by multicolor flow cytometry using cells isolated form spleen, peripheral blood (PB), bone marrow (BM) and peritoneal cavity (PC). Cell proliferation was measured by EdU incorporation in vivo. Immune synapse (IS) formation was assessed by confocal microscopy. Serum cytokines were quantified by multiplex bead arrays. Results: We first confirmed successful engraftment and presence of disease by immunohistochemistry. Compared to isotype controls, α-PD-L1 treated mice had significantly lower spleen weights (median 0.2 g vs 0.9 g, p<0.0001) and a highly significant lower relative frequency of CD19+CD5+ CLL lymphocytes in spleen (1.55% vs 71.69%), PB (10.5% vs 63.53%) and BM (0.26% vs 2.74%) demonstrating very effective tumor control. Compared to non-transplanted animals, α-PD-L1 treated mice showed alterations in almost all phenotypical and functional immune cell parameters, especially in regards to immune cell activation, indicating encounter with and immunological challenge by CLL cells. Along with disease control, α-PD-L1 treated mice had improved immune status as multiple inflammatory cytokines in the serum, including IL-10, TNF-α, CCL2 and GM-CSF were decreased and splenic infiltration of monocytes was reduced. While CLL development skewed monocytes towards Ly6Clow patrolling monocytes, α-PD-L1 treatment restored the presence of Ly6Chi inflammatory monocytes and decreased the expression of adhesion molecules ICAM-1 and PECAM-1. These monocytes regained their differentiation capacity as shown by increased numbers of macrophages and mature MHC-IIhi dendritic cells in the spleens of treated mice. In the T cell compartment, in vivo PD-L1 blockade prevented the CLL-induced CD4/CD8 ratio inversion, the loss of naïve CD8 T cells and the shift towards antigen-experienced and terminally differentiated T cells in spleen, BM and PB. Aberrant expression of immune checkpoint markers PD-1, KLRG-1, LAG-3, and 2B4 was also significantly reduced. The CLL-associated loss of intracellular IL-2 and the increased secretion of IL-4 and IFN-γ in CD4 T cells were prevented in α-PD-L1 treated mice. Respective cytokine patterns were observed in the serum. Functionally, PD-L1 blockade restored CD8 degranulation and IS formation to the level of healthy T cells, and significantly improved both ex vivo and in vivoT cell proliferation. Conclusion: Our in vivodata demonstrate that early PD-L1 blockade very effectively controls CLL development and enables complex effector function of myeloid and T cells, thus restoring anti-tumor immune responses. Targeting PD-L1/PD-1 interactions should therefore be further explored in clinical studies, potentially in combination with novel substances. BH/FM and MS/JGG contributed equally to first and last authorship. Disclosures No relevant conflicts of interest to declare.

No abstract