Immune Checkpoint Inhibitors Reverse the Mesenchymal Stromal Cell-Mediated Immunosuppression in Acute Myeloid Leukemia Microenvironment

Introduction: Acute myeloid leukemia (AML) cells can evade the immune system attack through the support of the bone marrow microenvironment and various cellular and molecular mechanisms. Within the bone marrow microenvironment, mesenchymal stromal cells (MSCs) were identified as a key contributor in promoting AML survival. Furthermore, MSCs can suppress the activity and proliferation of T cells through the expression and secretion of the immune checkpoint ligand, PD-L1. In addition, overexpression of inhibitory checkpoint receptors on CD8⁺ T cells of AML patients further hints at the critical involvement of the immune checkpoint pathways in AML immune escape. This evidence supports the clinical use of immune checkpoint inhibitors (CPI) in reversing MSC-mediated T cell suppression. However, the role of immune checkpoint inhibitors in AML, specifically for MSC-mediated immune evasion, has not yet been firmly established. We hypothesize that 1) bone marrow MSCs assist AML cells evading the T cell attack through the overexpression of inhibitory immune checkpoints, and 2) MSC-mediated T cell suppression could be reversed using CPI.

Methods: In this study, we established a 3-dimensional (3D) biomimetic model of decellularized Wharton's jelly matrix and primary AML patient samples to simulate the bone marrow microenvironment and the interplay between AML cells, MSCs, and T cells. The MSCs were isolated and expanded from the bone marrow mononuclear cells of each AML patient (n=4) to be used for the 3D model development. The AML cells and PBMCs for each of the corresponding patient were seeded in the presence or absence of MSCs and incubated for 1 day to assess the level of AML killing and the level of T cell activation using a far red-labeled AML cell-based killing assay and flow cytometry of intracellular cytokines (IFNγ, TNFα), respectively. An identical experimental set up was incubated for 3 days for immunophenotypic characterization of the surface markers (AML cells: CD33, CD34, HLA-ABC, PD-L1; CD8⁺ T cells: CD3, CD4, CD8, PD-1, LAG-3, TIM-3) using flow cytometry. With the established 3D model, the effect of CPI (bispecific anti-PD-1 x anti-LAG-3 DART®molecule, MGD013/tebotelimab; anti-LAG-3 IgG4 antibody; and anti-PD-1 IgG4 antibody, nivolumab) on AML killing and T-cell activity were assessed and compared to a palivizumab-based IgG4 antibody as a control. The DART molecule and antibodies, except for nivolumab, were provided by MacroGenics, Inc., MD, USA.

Results: The patient-derived bone marrow MSCs were immunophenotypically characterized by flow cytometry to be ≥95% positive in CD73, CD90, and CD105 expression; and <2% positive in CD45 expression. Immunofluorescence confocal microscopy of the cryostat serial sectioned 3D matrix scaffolds showed the physical contacts and the potential co-localization between the three cell types within the 3D matrix scaffold. To identify the impact of the MSCs on CD8⁺ T cell killing of AML cells within the 3D model, we observed a lowered level of T cell-based AML killing (n=4; P=0.0273), accompanied with a lowered intracellular IFNγ and TNFα levels of CD8⁺ T cells (n=4; P=0.0288 and P=0.0433, respectively). A trend of higher exhaustion marker LAG-3 expression was also observed in the presence of MSCs (n=4; P=0.0602) with a statistically significant higher expression of TIM-3 (n=4; P=0.0416). Assessing the influence of MSCs on AML immune evasion, a trend of lowered MHC-I expression (n=4; P=0.0684) and a higher level of PD-L1 expression (n=4; P=0.0349) on AML cells was observed in the presence of MSCs. Comparing the effect of the immune CPI relative to the control antibody in reversing the MSC-mediated immunosuppression, the LAG-3 inhibitor-treated condition was shown to have a higher level of AML killing (n=4; P=0.0178), compared to nivolumab and MGD013 treatments. The elevated AML killing of the LAG-3 inhibitor treatment condition is accompanied with a higher level of intracellular IFNγ expression of CD8⁺ T cells (n=4; P=0.0141).

Conclusion: MSCs play a major role in influencing CD8⁺ T cell-mediated killing of AML cells through the suppression of CD8⁺ T cell activity, potentially involving the LAG-3 immune checkpoint mechanism. LAG-3 immune checkpoint inhibitor showed effective response in reversing T-cell suppression within the AML microenvironment and warrants further investigation as an immunotherapeutic in AML.

Lin:Aptevo; Bio-Path Holdings; Ciclomed; Cleave; Jazz; Jazz Pharmaceuticals; Leukemia & Lymphoma Society; Kura Oncology; Trovagene: Research Funding; Jazz Pharmaceuticals; Servier: Consultancy. McGuirk:CRISPR therapeutics: Consultancy; Envision: Consultancy; Sana technologies: Consultancy; Allo Vir: Consultancy; Autolus: Consultancy; NEKTAR therapeutics: Consultancy; Caribou bio: Consultancy; Novartis: Consultancy; Legend biotech: Consultancy; Kite: Consultancy; BMS: Consultancy. Godwin:Sinochips Diagnostics: Honoraria, Other: Co-founder; Leidos Biomedical Research, Inc.: Research Funding; Clara Biotech, Inc.: Research Funding; VITRAC Therapeutics: Research Funding; Biovica, Inc.: Honoraria; Predicine: Research Funding; DetectOn Diagnostics, Inc.: Consultancy; Exokeryx, Inc.: Consultancy; BioFluidica, Inc.: Research Funding; Amprion, LLC: Consultancy; Biological Dynamics, Inc.: Consultancy. Abdelhakim:Iovance Biotherapeutics: Research Funding.