ALBERT

Description: Anti-PD1 therapy in hematologic malignancies has shown clinically inferior effects when compared to solid tumors.1 However, immunotherapy has been part of the standard of care in the treatment of acute leukemia for over 40 years. Recently, loss of MHCII was shown to be a mechanism of Immune escape in patients with acute myeloid leukemia after stem cell transplantation.2 Here we demonstrate, in pre-clinical models, that responses to anti-PD1 can be enhanced by increased antigen presentation through induction of MHCII. We used a Philadelphia chromosome positive (Ph+) Acute Lymphoblastic Leukemia (ALL) syngeneic mouse model and treated 4 cohorts of mice with the following regimens: 1) vehicle 2) single agent Dasatinib; 3) single agent anti-PD1; 4) combination anti-PD1 + Dasatinib (Figure 1a). Single agent Dasatinib, was highly active against this model compared to vehicle. Although anti-PD1 therapy showed little or no activity as a single agent, when combined with sub-therapeutic doses of Dasatinib, we observed significantly enhanced survival of mice compared to single agent anti-PD1 or to single agent Dasatinib. (Figure 1b) In our mouse model, Dasatinib increased the MHCII on the surface of the antigen presenting cells in the tumor microenvironment, most notably in dendritic cells (CD11c+ cells) from mouse tumor-infiltrating lymphocytes collected from the mice bone marrow after treatment (Fig 1c). To prove whether Dasatinib-induced MHCII expression can occur independently of a tumor microenvironment, we treated antigen presenting cell lines (KG1) with Dasatinib and were able to confirm an increased MHCII expression by flow cytometry and by western blotting. (Fig 1d) Since Dasatinib is approved for the treatment of Philadelphia chromosome positive ALL, CML, and has been used in a variety of other settings in the treatment of both malignant hematology and solid tumors, this data has immediate translational potential. In fact, Dasatinib in combination with anti-PD1 therapy is currently being tested in a variety of phase I/II trials in various kinds of malignancies (NCT04284202, NCT03516279). Here we show that induction of MHCII by Dasatinib may serve as a biomarker and could predict the potential benefit from the combination treatment. Additionally, targeting anti-PD1 following Dasatinib administration may increase response rates of patients treated with Dasatinib, providing a rationale for sequential treatment design in patients with persistent minimal residual disease. Figure 1. (a) C57BL/6J mice were injected via tail vein injection with a transplantable, immunocompetent BCR-ABL+ B-ALL model and treated with vehicle/Dasatinib/or anti-PD1 as described in the schema. (b) Overall survival of the mice treated in the previous experiment. (c) MHCII cell surface expression on dendritic cells (CD11+ cells) from the bone marrow of treated mice. (d) relative Cell surface expression of MHCII in the KG1 cell line (APC cell line) treated with Dasatinib vs. vehicle. (e) protein expression of MHCII in the KG1 cell line (APC cell line) treated with Dasatinib vs. vehicle. 1. Masarova L, Kantarjian H, Ravandi F, Sharma P, Garcia-Manero G, Daver N. Update on Immunotherapy in AML and MDS: Monoclonal Antibodies and Checkpoint Inhibitors Paving the Road for Clinical Practice. Adv Exp Med Biol 2018;995:97-116. W2. Christopher MJ, Petti AA, Rettig MP, et al. Immune Escape of Relapsed AML Cells after Allogeneic Transplantation. New England Journal of Medicine 2018;379:2330-41. Figure Disclosures Koller: Jazz Pharmaceuticals, Inc.: Consultancy. Konopleva:Stemline Therapeutics: Consultancy, Research Funding; F. Hoffmann La-Roche: Consultancy, Research Funding; Forty-Seven: Consultancy, Research Funding; Sanofi: Research Funding; Genentech: Consultancy, Research Funding; Ablynx: Research Funding; Cellectis: Research Funding; Agios: Research Funding; AbbVie: Consultancy, Research Funding; Ascentage: Research Funding; Kisoji: Consultancy; Reata Pharmaceutical Inc.;: Patents & Royalties: patents and royalties with patent US 7,795,305 B2 on CDDO-compounds and combination therapies, licensed to Reata Pharmaceutical; AstraZeneca: Research Funding; Amgen: Consultancy; Calithera: Research Funding; Eli Lilly: Research Funding; Rafael Pharmaceutical: Research Funding. OffLabel Disclosure: anti-pd1 therapy in acute lymphoid leukemia

Description: The inferior cure rate of T-cell acute lymphoblastic leukemia (T-ALL) is associated with inherent drug resistance. The activating NOTCH1 gene mutations have been reported to cause chemoresistance at the stem cell level1. Direct NOTCH1 inhibition has failed in clinical trials due to a narrow therapeutic window but targeting key oncogenic and metabolic pathways downstream of mutated NOTCH1 may offer novel approaches. We previously reported that rapid transformation of thymocytes at the DN3 differentiation stage into preleukemic stem cells (pre-LSC) requires elevated Notch1 in addition to the presence of Scl/Lmo11. Notably, we showed that cellular metabolism of NOTCH1-mutated T-ALLs depends on Oxidative Phosphorylation (OxPhos) and that OxPhos inhibition using the complex I inhibitor IACS-010759 (OxPhos-i) is efficacious in NOTCH1-mutated T-ALL patient derived xenografts (PDXs)2. Here, we investigated the link between NOTCH1-mutated chemoresistance and OxPhos in pre-leukemic and leukemic cells, utilizing comprehensive molecular and functional assays. We hypothesized that chemotherapy aided by OxPhos-i overcomes chemoresistance, depletes LSCs and combats T-ALL. First, we analyzed the role of OxPhos in downstream Notch1 targets at the pre- and leukemic stage considering four stages of thymocyte differentiation (D1-D4), in a mouse model of human T-ALL1. Gene set enrichment analysis (GSEA) implicated increased expression of Notch1 target genes starting at DN1, and OxPhos target genes were the highest-ranked gene set at DN3. Next, activation of Notch1 by its ligand DL4 and inhibition of OxPhos reduced viability of pre-LSCs, indicating that ligand-dependent activation of Notch1 signaling upregulates the OxPhos pathway and sensitizes pre-LSCs to OxPhos-i. To clarify the role of Notch1 signaling, we examined the effect of IACS-010759 on pre-leukemic thymocytes harboring LMO1, SCL-LMO1, NOTCH1, LMO1-NOTCH1 and SCL-LMO1-NOTCH1 with and without DL4 stimulation. We found that in the absence of DL4, only thymocytes harboring the Notch1 oncogene responded to OxPhos-i, whereas all DL4-stimulated thymocytes responded regardless of Notch1 status (Fig. 1a). In addition, at the leukemic stage, we found elevation of the OxPhos pathway driven by oncogenic Notch1 when we compared transcriptomes of SCL-LMO1 induced T-ALL in the presence or absence of the NOTCH1 oncogene. In line with the murine T-ALL NOTCH1 model, we performed transcriptome analysis of two independent T-ALL patient cohorts prior to chemotherapy, COG TARGET ALL (n=263) and AALL1231 (n=75), comparing transcriptomes of NOTCH1-mutated vs NOTCH1-wt T-ALLs. We found co-segregation of NOTCH1 mutations with significant upregulation of OxPhos and TCA cycle genes and downregulation of apoptosis signaling. Aiming to reverse the NOTCH1-controlled anti-apoptotic program and chemoresistance, we next tested the combination of Vincristine, Dexamethasone and L-Asparaginase (VXL) with IACS-010759. When compared to vehicle, OxPhos-i or VXL alone, only the VXL-OxPhos-i treatment caused an energetic crisis indicated by decreased OCR and ECAR (Seahorse), which translated to a profound reduction of viability (CTG, flow cytometry) in T-ALL cell lines (n=9) and primary T-ALL samples (n=5). Additionally, the IACS-VXL combination in vivo resulted in pan-metabolic blockade, which caused metabolic shut-down and triggered early induction of apoptosis in leukemic cells in peripheral blood, spleen and bone marrow (Fig. 1b). Single cell Proteomic analysis (CyTOF) of spleen showed reduced expression of cell proliferation marker -ki67, c-myc, ERK and p38 proteins, and reduction in number of leukemic cells. Finally, this combination therapy resulted in reduced leukemia burden and extension of overall survival across all three aggressive NOTCH1-mutated T-ALL PDX models (p