Paulus, Aneel

Abstract Chronic Lymphocytic Leukemia (CLL) is considered a disease of antigen-naive CD5+CD23+CD19+κ/λ+ B-lymphocytes, in which the tumor microenvironment is highly immunosuppressive. Among several human B cell subsets, B10 [CD24hiCD27+CD38hi]/B regulatory cells (Bregs) are functionally discernable in patients with aggressive CLL as they support and regulate immunological tolerance via production of IL-10 and transforming growth factor β (TGF-β). In an ex vivo setting, we noted CD38hi CLL Bregs to promote transformation and expansion of CD4+CD25+FoxP3+Tregs; an effect that was abrogated in the presence of IL10/TGF-β neutralizing antibodies. In patients with CLL, Bregs also prohibit the expansion of cytotoxic T cells (CD8+T cells; Tc), natural killer (NK) cells and other pro-inflammatory lymphocytes (CD4+ helper T cells; Th1 and Th17). We noted that patients with CLL had a significantly higher % of Tregs (11.8±2.0) compared to healthy donors (1.9±0.3) and similar to Bregs, a substantial proportion of CLL Tregs had high CD38 expression (MFI=616.8±36.2). We demonstrated that an anti-CD38 therapeutic antibody was able to eliminate CD38+ Bregs and Tregs from CLL patients by immune effector mechanisms (ADCC, CDC, ADCP) as well as direct mitochondrial/FcγR-mediated apoptosis. Furthermore, in a PDX model of CLL, mice treated with anti-CD38 therapy had decreased % Bregs and Tregs but increased Th17 and CD8+ T-cells (vs. vehicle; p<0.05). These observations carry implications beyond CLL for any Breg/Treg-dependent cancer or disease (i.e. Systemic Lupus Erythematosus, Rheumatoid arthritis, Sjögren’s syndrome), wherein CD38-targeted therapy could potentially rescue/improve an immune tolerant microenvironment by eradication of Bregs.

Abstract Hodgkin’s lymphoma (HL) is characterized by malignant CD30+ Reed-Sternberg (HRS) B-cells surrounded by an ineffective infiltrate of inflammatory immune cells. T-helper and T-regulatory cells presumably provide survival signals for HRS cells; rescuing them from immune-surveillance by CD8+ cytotoxic T lymphocytes and natural killer (NK) cells, which are also functionally anergic. Clinically, ≈10–30% of HL patients resulted in relapse after front line therapy; highlighting the need for development of new treatments. We investigated a novel bispecific CD16axCD30 antibody, AFM13 (Affimed Therapeutics, AG), which engages NK cells and recruits them towards CD30+ tumor cells. Safety of single-agent AFM13 has been demonstrated in a phase I clinical study, however, disease remission was observed in only 23% of patients. We hypothesized that combining AFM13 with another agent whose safety profile is already been established in HL and also activates NK cells (lenalidomide) could potentially increase anti-tumor efficacy of AFM13. We observed increased expression of CD107a, CD69, CD25 and IFNγ (activation markers) in NK cells isolated from HL patients (HL-NK), that were stimulated ex vivo with IL- 15 + lenalidomide. Pre-treatment with IL-2/IL-15 antibody has been shown to activate and restore T/NK cell cytolytic function. In line with this, IL-15+lenalidomide-stimulated HL-NK cells showed functional restoration by inducing specific lysis of human CD30+ lymphoma cells (30.55±1.49%) in a co-culture assay. While, AFM13 is being investigated in combination with anti-PD1 antibody therapy (NCT02665650), our preclinical data provides the rationale for an AFM13 + Len combination regimen for HL and potentially other CD30+ malignancies.

Abstract Increased CD38 expression on chronic lymphocytic leukemia (CLL) cells correlates with a negative prognosis, however, its utility as a therapeutic target in this malignancy remains unknown. Our investigations show that the anti-CD38 therapeutic mAb, daratumumab (Dara) induces direct apoptosis of CLL cells and promotes their immune-mediated clearance (through ADCC). Notably, CD38 expression is quite high on a subset of CLL-cells (CD19+/CD38hi/CD24+/IL-10+; B-regulatory-like phenotype), which promote Treg expansion while prohibiting CD8+ cytotoxic T-cells (cTLs). Indeed, immunosuppression is a hallmark of CLL and patients often present with increased number of Tregs and Th2 cells but decreased effector T cells and cTLs. It has recently been reported that CD38hi Tregs are significantly immunosuppressive and when examining this population in peripheral blood mononuclear cells (PBMCs) from CLL patients (n=13), we noted that >50% of CD4+/CD25+/FoxP3+ cells were CD38hi (median MFI ~600). We hypothesized that Dara would kill CD38hi CLL cells à decreased CD38hi Treg formation/direct killing à expansion of cTLs. Indeed, ex-vivo treatment of CLL-patient derived PBMCs showed that Dara was highly lethal to both CD38hi "Breg-like" CLL cells and Tregs; associated with decrease in IL-10, increase in IFN-γ, effector T cells, Th17 subsets and proliferation of cTLs. Altogether this pointed towards recovery of anti-tumor immunity which was further evidenced by enhanced tumor-specific cytolytic activity of cTLs. Our pre-clinical data demonstrates that targeting CD38 in CLL has an anti-tumor effect via 1.) lethal effect on CLL-Bregs and Tregs and 2.) tumor-specific expansion of cTLs, which we will test clinically in an upcoming trial.

Abstract Purpose: IgM monoclonal gammopathy of undetermined significance (MGUS) and Waldenström macroglobulinemia (WM) represent a disease spectrum with highly varied therapeutic management, ranging from observation to chemoimmunotherapy. The current classification relies solely on clinical features and does not explain the heterogeneity that exists within each of these conditions. Further investigation is warranted to shed light on the biology that may account for the clinical differences. Experimental Design: We used bone marrow (BM) clonal CD19+ and/or CD138+ sorted cells, matched BM supernatant, and peripheral blood serum from 32 patients (7 MGUS, 25 WM) to perform the first multi-omics approach including whole-exome sequencing, RNA sequencing, proteomics, metabolomics, and mass cytometry. Results: We identified three clusters with distinct pathway activation, immune content, metabolomic, and clinical features. Cluster 1 included only patients with WM and was characterized by transcriptional silencing of genes involved in cell cycle and immune response, enrichment of mitochondrial metabolism, infiltration of senescent T effector memory cells, and aggressive clinical behavior. Genetic/structural alterations of TNFAIP3 were distinct events of this cluster. Cluster 2 comprised both MGUS and WM patients with upregulation of inflammatory response, senescence and glycolysis signatures, increased activated T follicular helper and T regulatory cells, and indolent clinical behavior. Cluster 3 also included both MGUS and WM patients and exhibited intermediate features, including proliferative and inflammatory signaling, as well as glycolysis and mitochondrial metabolism. Conclusions: We have identified three distinct molecular clusters, suggesting a potential biologic classification that may have therapeutic implications.

Abstract Purpose: Development of B-cell lymphoma 2 (BCL-2)–specific inhibitors poses unique challenges in drug design because of BCL-2 homology domain 3 (BH3) shared homology between BCL-2 family members and the shallow surface of their protein–protein interactions. We report herein discovery and extensive preclinical investigation of lisaftoclax (APG-2575). Experimental Design: Computational modeling was used to design “lead” compounds. Biochemical binding, mitochondrial BH3 profiling, and cell-based viability or apoptosis assays were used to determine the selectivity and potency of BCL-2 inhibitor lisaftoclax. The antitumor effects of lisaftoclax were also evaluated in several xenograft models. Results: Lisaftoclax selectively binds BCL-2 (Ki < 0.1 nmol/L), disrupts BCL-2:BIM complexes, and compromises mitochondrial outer membrane potential, culminating in BAX/BAK-dependent, caspase-mediated apoptosis. Lisaftoclax exerted strong antitumor activity in hematologic cancer cell lines and tumor cells from patients with chronic lymphocytic leukemia, multiple myeloma, or Waldenström macroglobulinemia. After lisaftoclax treatment, prodeath proteins BCL-2‒like protein 11 (BIM) and Noxa increased, and BIM translocated from cytosol to mitochondria. Consistent with these apoptotic activities, lisaftoclax entered malignant cells rapidly, reached plateau in 2 hours, and significantly downregulated mitochondrial respiratory function and ATP production. Furthermore, lisaftoclax inhibited tumor growth in xenograft models, correlating with caspase activation, poly (ADP-ribose) polymerase 1 cleavage, and pharmacokinetics of the compound. Lisaftoclax combined with rituximab or bendamustine/rituximab enhanced antitumor activity in vivo. Conclusions: These findings demonstrate that lisaftoclax is a novel, orally bioavailable BH3 mimetic BCL-2–selective inhibitor with considerable potential for the treatment of certain hematologic malignancies.