Control Of Acute Myeloid Leukemia Research Articles

AML/T cell interactomics uncover correlates of patient outcomes and the key role of ICAM1 in T cell killing of AML

T cells are important for the control of acute myeloid leukemia (AML), a common and often deadly malignancy. We observed that some AML patient samples are resistant to killing by human-engineered cytotoxic CD4+ T cells. Single-cell RNA-seq of primary AML samples and CD4+ T cells before and after their interaction uncovered transcriptional programs that correlate with AML sensitivity or resistance to CD4+ T cell killing. Resistance-associated AML programs were enriched in AML patients with poor survival, and killing-resistant AML cells did not engage T cells in vitro. Killing-sensitive AML potently activated T cells before being killed, and upregulated ICAM1, a key component of the immune synapse with T cells. Without ICAM1, killing-sensitive AML became resistant to killing by primary ex vivo-isolated CD8+ T cells in vitro, and engineered CD4+ T cells in vitro and in vivo. While AML heterogeneity implies that multiple factors may determine their sensitivity to T cell killing, these data show that ICAM1 acts as an immune trigger, allowing T cell killing, and could play a role in AML patient survival in vivo.

Control of acute myeloid leukemia and generation of immune memory in vivo using AMV564, a bivalent bispecific CD33 x CD3 T cell engager.

Off-the-shelf immunotherapeutics that suppress tumor growth and provide durable protection against relapse could enhance cancer treatment. We report preclinical studies on a CD33 x CD3 bivalent bispecific diabody, AMV564, that not only suppresses tumor growth, but also facilitates memory responses in a mouse model of acute myelogenous leukemia (AML). Mechanistically, a single 5-day treatment with AMV564 seems to reduce tumor burden by redirection of T cells, providing a time window for allogeneic or other T cells that innately recognize tumor antigens to become activated and proliferate. When the concentration of bispecific becomes negligible, the effector: target ratio has also shifted, and these activated T cells mediate long-term tumor control. To test the efficacy of AMV564 in vivo, we generated a CD33+ MOLM13CG bioluminescent human cell line and optimized conditions needed to control these cells for 62 days in vivo in NSG mice. Of note, not only did MOLM13CG become undetectable by bioluminescence imaging in response to infusion of human T cells plus AMV564, but also NSG mice that had cleared the tumor also resisted rechallenge with MOLM13CG in spite of no additional AMV564 treatment. In these mice, we identified effector and effector memory human CD4+ and CD8+ T cells in the peripheral blood immediately prior to rechallenge that expanded significantly during the subsequent 18 days. In addition to the anti-tumor effects of AMV564 on the clearance of MOLM13CG cells in vivo, similar effects were seen when primary CD33+ human AML cells were engrafted in NSG mice even when the human T cells made up only 2% of the peripheral blood cells and AML cells made up 98%. These studies suggest that AMV564 is a novel and effective bispecific diabody for the targeting of CD33+ AML that may provide long-term survival advantages in the clinic.

Oncogenic drivers dictate immune control of acute myeloid leukemia

Acute myeloid leukemia (AML) is a genetically heterogeneous, aggressive hematological malignancy induced by distinct oncogenic driver mutations. The effect of specific AML oncogenes on immune activation or suppression is unclear. Here, we examine immune responses in genetically distinct models of AML and demonstrate that specific AML oncogenes dictate immunogenicity, the quality of immune response and immune escape through immunoediting. Specifically, expression of NrasG12D alone is sufficient to drive a potent anti-leukemia response through increased MHC Class II expression that can be overcome with increased expression of Myc. These data have important implications for the design and implementation of personalized immunotherapies for patients with AML.

Control of acute myeloid leukemia by a trifunctional NKp46-CD16a-NK cell engager targeting CD123

CD123, the alpha chain of the IL-3 receptor, is an attractive target for acute myeloid leukemia (AML) treatment. However, cytotoxic antibodies or T cell engagers targeting CD123 had insufficient efficacy or safety in clinical trials. We show that expression of CD64, the high-affinity receptor for human IgG, on AML blasts confers resistance to anti-CD123 antibody-dependent cell cytotoxicity (ADCC) in vitro. We engineer a trifunctional natural killer cell engager (NKCE) that targets CD123 on AML blasts and NKp46 and CD16a on NK cells (CD123-NKCE). CD123-NKCE has potent antitumor activity against primary AML blasts regardless of CD64 expression and induces NK cell activation and cytokine secretion only in the presence of AML cells. Its antitumor activity in a mouse CD123+ tumor model exceeds that of the benchmark ADCC-enhanced antibody. In nonhuman primates, it had prolonged pharmacodynamic effects, depleting CD123+ cells for more than 10 days with no signs of toxicity and very low inflammatory cytokine induction over a large dose range. These results support clinical development of CD123-NKCE.

Fully Myeloablative Antibody-Drug Conjugates Condition for Hematopoietic Stem Cell Transplantation and Provide Durable Antileukemia Benefit

INTRODUCTION: Allogeneic hematopoietic stem cell transplantation (allo-HSCT) holds the potential for cure of numerous blood diseases and provides the best chance for durable control of acute myeloid leukemia (AML). Successful allo-HSCT requires that host hematopoietic stem cells (HSCs) are depleted to make bone marrow (BM) space for the incoming donor HSCs, and that host T and NK cell mediated immunity is suppressed to prevent donor HSC rejection. To meet these requirements, conditioning regimens comprised of chemotherapy and/or irradiation are used, which carry risks of severe toxicities that may prevent patients from undergoing this potentially curative therapy. Thus, there is an unmet clinical need for novel conditioning regimens that minimize toxicity without sacrificing therapeutic efficacy. Previously, we showed in mouse models that saporin-based antibody-drug conjugates (ADCs) targeting CD45 or cKit combined with Janus kinase 1/2 (JAK1/2) inhibitors enabled stable, high-level engraftment in fully mismatched HSCT with ≥99% donor myeloid lineage chimerism. However, saporin ADCs are not fully myeloablative, which limited their antitumor efficacy in prior studies and would likely limit the ability of these conditioning agents to provide protection against AML relapse. We hypothesized that alternative toxic payloads to saporin would enable generation of fully myeloablative ADCs capable of enhanced antileukemia benefit. METHODS: We conjugated anti-mouse CD45.2 (clone 104), anti-mouse cKit (clone 2B8), and their respective isotype control antibodies with pyrrolobenzodiazepine (PBD) dimers via maleimide-thiol chemistry. PBD-conjugated ADCs were tested for cytotoxicity in vitro against the murine YAC-1 (CD45.2+) or P815 (cKit+) cell lines, as well as the human Jurkat (CD45.2-cKit-) cell line using XTT viability assays. Cytotoxicity against BM HSCs from B6 mice and AML1 primary leukemia cells (B6-derived, Dnmt3aR878H/+ and FLT3-ITD; provided by Dr. Timothy Ley) was tested using methylcellulose assays. HSC depletion in vivo was assessed 7 days after treatment with a single dose of CD45- or cKit-PBD. Syngeneic and fully mismatched HSCT studies with CD45-PBD were done by conditioning B6 mice with 50 µg ADC 7 days before transplant (d-7), then infusing 10x106 B6-GFP (syngeneic) or 20x106 BALB/c (allogeneic) BM cells on d0. For in vivo leukemia killing studies, B6 mice were infused with 5x105 AML1 cells on d-6, treated with 50 mg CD45-PBD or 75 mg CD45-saporin on d-2, then transplanted on d0 with 10x106 T cell depleted BM cells from 129/Sv mice. In allo-HSCT studies, mice received oral ruxolitinib (2 mg/g chow) starting on d-3. RESULTS: CD45-PBD targeted murine YAC-1 cells and cKit-PBD targeted P815 cells in vitro with picomolar-range potencies; neither ADC was effective against human Jurkat cells. Target-specific killing was confirmed with blocking studies using excess unconjugated anti-CD45.2 or anti-cKit. While CD45-PBD and cKit-PBD showed similarly high efficacy and picomolar-range potencies at targeting AML1 leukemia cells in vitro, cKit-PBD was approximately 40-fold more potent at depleting B6 HSCs in vitro than CD45-PBD (IC50 values: 243 fM for cKit-PBD versus 10 pM for CD45-PBD). In vivo, a single 50 µg dose of CD45-PBD or 25 µg dose of cKit-PBD completely ablated the HSC niche in B6 mice. Initial transplantation studies done with CD45-PBD confirm that this ADC enables both syngeneic and allogeneic HSCT with near complete donor chimerism in all recipient mice (Fig. 1). Finally, a single CD45-PBD dose provided durable protection against AML1 leukemia, whereas the nonmyeloablative CD45-saporin was only minimally protective (Fig. 2). CONCLUSIONS: We successfully generated fully myeloablative CD45- and cKit-ADCs using PBD as the toxic payload. Notably, CD45-PBD enabled HSCT with near total conversion to donor hematopoiesis and was highly protective against the aggressive AML1 primary leukemia. Studies thus far with cKit-PBD have shown superior killing of murine HSCs in vitro and in vivo compared to CD45-PBD; testing of cKit-PBD in HSCT models and in vivo AML1 leukemia killing studies are in progress. Given the high potency of the PBD payload, we are conducting extensive histopathologic studies and serum chemistry testing to evaluate the acute and chronic organ toxicities of conditioning with PBD-based conjugates. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

CLEC12A and CD33 coexpression as a preferential target for pediatric AML combinatorial immunotherapy

AbstractEmerging immunotherapies such as chimeric antigen receptor T cells have advanced the treatment of acute lymphoblastic leukemia. In contrast, long-term control of acute myeloid leukemia (AML) cannot be achieved by single lineage-specific targeting while sparing benign hematopoiesis. In addition, heterogeneity of AML warrants combinatorial targeting, and several suitable immunotargets (HAVCR2/CD33 and HAVCR2/CLEC12A) have been identified in adult AML. However, clinical and biologic characteristics of AML differ between children and the elderly. Here, we analyzed 36 bone marrow (BM) samples of pediatric AML patients and 13 age-matched healthy donors using whole RNA sequencing of sorted CD45dim and CD34+CD38−CD45dim BM populations and flow cytometry for surface expression of putative target antigens. Pediatric AML clusters apart from healthy myeloid BM precursors in principal-component analysis. Known immunotargets of adult AML, such as IL3RA, were not overexpressed in pediatric AML compared with healthy precursors by RNA sequencing. CD33 and CLEC12A were the most upregulated immunotargets on the RNA level and showed the highest surface expression on AML detected by flow cytometry. KMT2A-mutated infant AML clusters separately by RNA sequencing and overexpresses FLT3, and hence, CD33/FLT3 cotargeting is an additional specific option for this subgroup. CLEC12A and CD33/CLEC12Adouble-positive expression was absent in CD34+CD38−CD45RA−CD90+ hematopoietic stem cells (HSCs) and nonhematopoietic tissue, while CD33 and FLT3 are expressed on HSCs. In summary, we show that expression of immunotargets in pediatric AML differs from known expression profiles in adult AML. We identify CLEC12A and CD33 as preferential generic combinatorial immunotargets in pediatric AML and CD33 and FLT3 as immunotargets specific for KMT2A-mutated infant AML.

TCR and DNAM Dependent Lysis of Acute Monocytic Leukemia (AMoL) by Vg9Vd2 T Lymphocytes.

Abstract 1041Poster Board I-63 BACKGROUNd:Compelling evidences have demonstrated the role of the immune system in the control of acute myeloid leukemia (AML). So far, T cells and natural killer (NK) cells are the major immune effectors shown to be involved in AML control. The graft-versus-leukemia (GVL) effect following allogenic stem cell transplantation as well as donor lymphocyte infusions indicate that T lymphocytes can control and eliminate AML cells. Leukemia-specific antigenic peptides have been characterized (proteinase-3 and Wilms tumor 1 protein) and serve as targets for peptide-based vaccine trials in AML. Allogenic NK cells have anti-leukemic activity as shown by killer cell inhibitory receptor (KIR)-mismatched haplo-identical stem cell transplantation. Less is known regarding the role of gd T cells in the control of AML. Recently the reconstitution of Vd1 T lymphocytes post transplantation has been shown to correlate with a better prognosis. In the present study, we have analyzed gd T cells in patients with AML and in a mouse model of human AML and focused on (Vg9) Vd2 T cells, the main subset of circulating gd T cells with anti-neoplastic activity. Human Vg9Vd2 T lymphocytes can be activated by nonpeptidic antigens such as the mevalonate pathway-derived isopentenyl pyrophosphate or synthetic phosphoantigen such as bromohydrin pyrophosphate (BrHPP). This population may be suitable for the adoptive immunotherapy of acute myeloid leukemia (AML). However little is known about the frequency, the function and the mechanisms underlying Vg9Vd2 T-cell recognition of AML. We have focused this study on AMoL which are targets of NK and ab T cells. OBJECTIVE OF THE STUDYto describe Vg9Vd2 T cells in patients with AML and investigate their ability to induce an effective cytotoxic response against autologous AML blast in vitro and in vivo. EXPERIMENTAL PROCEDURe:We compared the phenotype and the absolute circulating Vg9Vd2 T cell levels in the blood and the bone marrow (BM) in 12 patients with AMoL (FAB AML-M4 and -M5) and 12 healthy volunteers (HV) using multi parametric flow cytometry. All patients and volunteers gave written informed consent. Vg9Vd2 T cells of AML patient were expanded ex vivo using BrHPP or Zoledronic acid plus IL2. The functions of expanded Vg9Vd2 T cells were assessed in vitro by their cytotoxicity against leukemic blasts (CD107a staining, 51Cr assay) and in vivo in immunodeficient mice transplanted with human AML cell line (U937). In these experiments, the ability of adoptively transferred Vg9Vd2 T cells to migrate into BM and improve mice survival was assessed after i.v. infusion of U937 cells into healthy female NOD-SCID, common _-chain knockout mice (NOG mice). Mice were then treated twice i.v. with 40.106 Vg9Vd2 T cells. RESULTs:Vg9Vd2 T lymphocytes are present in the blood as well as BM of AMoL patients at a lower frequency as compared to HV (median 2.07/μl vs 34/μL respectively P<0.001). Vg9Vd2 T lymphocytes from AML patients are endowed with in vitro proliferation in response to BrHPP or Zoledronic acid plus IL2 but lower than HV (fold increase median 33 versus 69, P=0.051). Expanded Vg9Vd2 express activation markers (CD69 and CCR5) and exhibit an effector/memory phenotype (CD45RA- CD27-). Their lytic potential toward autologous AML blast was equivalent to those of HV by 51Cr experiments and CD107a staining and involves the perforin-granzyme pathway. Their activity depends on both TCRVd2 and DNAX accessory molecule-1 (DNAM-1) as demonstrated by antibody blockade. In vivo data show that, upon sacrifice, Vg9Vd2 were detected in BM, spleen and blood of mice. Preliminary Kaplan-Meier analysis of pooled cohorts of Vg9Vd2-treated and untreated mice reveals that mice receiving Vg9Vd2 T cells displayed superior survival compared with untreated controls (P=0.0047). CONCLUSIOn:Altogether, our data indicate that Vg9Vd2 T cells are decreased in AML patients and have a more limited expansion potential. However, they are able to kill autologous AML blast upon stimulation in a TCRVd2 as well as the DNAM-1 receptor dependent manner. These results provide a rationale for the clinical evaluation of adoptive transfer of ex vivo expanded allogenic Vg9Vd2T cells or direct activation of Vg9Vd2T cells with IL2 + phosphoantigens in patients with AML. Disclosures:No relevant conflicts of interest to declare.

Analysis of NK Cells Receptors and NK Cells Ligands in Acute Myeloid Leukemia: Comparisons Between the Blood and the Bone Marrow.

Abstract 1643Poster Board I-669 BackgroundThe significant role of NK cells in the control of acute myeloid leukemia (AML) has been demonstrated in the setting of allogeneic stem cell transplantation. However, the implication of NK cells against autologous leukemic cells needs to be clarified.We have previously described deficient expression of NK activatory receptors in AML at diagnosis, in particular the natural cytotoxicity receptors (NCR) namely NKp30 and NKp46. So, defective cytotoxicity of AML cells can be explained by abnormalities of activating NK-receptor expression allowing immune escape from NK cells. However, immune escape can be also due to defective activating NK receptor-ligand interactions due to abnormal expression of their ligands on blasts cells. These defects have been observed in particular on NK cells or blasts cells isolated from the blood. Few studies have analysed the bone marrow component although blasts cells concentrate here. We postulated that abnormalities of NK cells receptors or ligands expression are more severe in bone marrow, in near contact with the blasts, compared with blood. We sought to identify disparities between deficient expression of NK or ligands in the bone marrow in comparison with the blood. MethodsWe realized a phenotypic analysis of NK cells and blasts cells at the diagnosis of AML. The level of activatory NK receptors (NKRa) knew to mediate NK cell recognition and lysis of AML blasts cells (NCR (NKp30 and NKp46) and DNAM-1) was investigated by flow cytometry. The expression of NKG2D ligands (MICA/B and ULBP1-3) and DNAM-1 ligands (Nectin-2 and PVR) receptors were also analysed. These analyses were realised with coupled specimens obtained in the same patient at diagnosis of AML (n=19), peripheral blood and bone marrow samples in order to detect discrepancies between these two sites. A control group (age-matched; n=15) for blood samples was included for this study. All biological samples were obtained from patients and healthy volunteers after informed consent. ResultsA total of 19 patients were included in this study. We included 6 cases of AML 5, 4 cases of AML 4, 4 cases of AML 2, 4 cases of AML 1 and one case of AML 0. Flow cytometry data for NKRa were only available for 11 patients. We confirmed the deficient expression of NKp30 and NKp46 receptors (as determined by MFI) on NK cells from blood of AML patients. In AML patient, the ratio MFI (MFI receptor/MFI control isotype) of NKp30 (4.27 +/- 2.97; p<0.0001) and of NKp46 (5.96 +/- 5.67; p<0.0001) significantly differed from healthy volunteers (NKp30 26.65 +/- 6.12; NKp46 39.73 +/- 9.66). Moreover, the deficient expression of these receptors was also observed on NK cells from the bone marrow (NKp30 3.66 +/- 2.22, p<0.0001; NKp46 6.71+/- 6.42, p<0.0001). However, we can not demonstrated significant differences between the NKRa expression on NK cells from blood versus from bone marrow (NKp30 p= 0.8438; NKp46 p= 0.9476 and DNAM-1 p= 0.3579).The expression of the ligands for NKRa was analysed to compare the expression on blasts cells isolated from the blood compared to blasts cells isolated from the bone marrow. Flow cytometry data for ligands were only available for 17 patients. We observed a strong expression of HLA class I molecules on blasts cells that was equivalent in the blood and in bone marrow. DNAM-1 ligands (PVR, Nectin 2) were expressed on blasts cells (see figure). NKG2D ligands were also expressed but to a lesser extent with predominant ULBP1 expression. However, we can not observed significant differences in the expression of ligands between the blood and the bone marrow. ConclusionsThe deficiency of activating NK cells receptors expression at AML diagnosis is significant, present in a majority of patients, and consistent across the 2 components, ie blood and bone marrow. These defects are one component of the immune escape from NK cells. We have speculated that these abnormalities were more pronounced in bone marrow, near blasts cells, because these abnormalities are in part induced by blasts cells. However, we can not demonstrated significant differences in the expression of activating receptors or ligands between blood and bone marrow. We are accumulating more data in order to detect differences between sub-groups of AML. DisclosuresNo relevant conflicts of interest to declare.

Improved detection of minimal acute myeloid leukemia cells by the use of the combined parameters of XE‐2100 hematology analyzer

For the diagnosis and therapy of acute leukemia, it is important to detect a small number of leukemic cells precisely. Although several automated hematology analyzers that carry blast-detecting programs have been developed, they do not exert sufficient detection sensitivity to exceed the sensitivity of manual eye counting method. We constructed a new blast-detecting program by combining the numerical information acquired from five cytometric parameters presented by XE-2100. The sensitivity and specificity of this blast multi-scoring program were assessed in comparison with the Blasts flag program equipped originally in XE-2100. The blast-detecting sensitivity was found to be highly improved in the blast multi-scoring program as compared with the Blasts flag program without much decreasing the specificity. A small number of leukemic myeloblasts was detected at the better sensitivity than the eye counting method in the clinical course of the patients with acute myeloid leukemia. The daily practical use of this blast multi-scoring program will surely contribute to sensitive, objective, and real-time evaluation of the control of acute myeloid leukemia with a low cost.