The Immunotherapy of Acute Myeloid Leukemia: A Clinical Point of View

Abstract

The potential of the immune system to eradicate leukemic cells has been consistently demonstrated by the Graft vs. Leukemia effect occurring after allo-HSCT and in the context of donor leukocyte infusions. Various immunotherapeutic approaches, ranging from the use of antibodies, antibody-drug conjugates, bispecific T-cell engagers, chimeric antigen receptor (CAR) T-cells, and therapeutic infusions of NK cells, are thus currently being tested with promising, yet conflicting, results. This review will concentrate on various types of immunotherapies in preclinical and clinical development, from the point of view of a clinical hematologist. The most promising therapies for clinical translation are the use of bispecific T-cell engagers and CAR-T cells aimed at lineage-restricted antigens, where overall responses (ORR) ranging from 20 to 40% can be achieved in a small series of heavily pretreated patients affected by refractory or relapsing leukemia. Toxicity consists mainly in the occurrence of cytokine-release syndrome, which is mostly manageable with step-up dosing, the early use of cytokine-blocking agents and corticosteroids, and myelosuppression. Various cytokine-enhanced natural killer products are also being tested, mainly as allogeneic off-the-shelf therapies, with a good tolerability profile and promising results (ORR: 20-37.5% in small trials). The in vivo activation of T lymphocytes and NK cells via the inhibition of their immune checkpoints also yielded interesting, yet limited, results (ORR: 33-59%) but with an increased risk of severe Graft vs. Host disease in transplanted patients. Therefore, there are still several hurdles to overcome before the widespread clinical use of these novel compounds.

Keywords: NK cells; T lymphocytes; acute myeloid leukemia; bioengineering; bispecific antibodies; chimeric antigen receptor cells; dual-affinity retargeting antibodies; immune checkpoint inhibitors; immune escape; immunotherapy.

Figure 1

Immunotherapy of acute myeloid leukemia. This figure shows a schematic representation of selected components and agents involved in the immunotherapy of acute myeloid leukemia (AML). Cells exerting immune activity against AML are shown on the left of the figure; opposite to them, immunomodulatory cells and immunosuppressive agents are shown on the right. Potential immunotherapeutic drugs are shown on top of the figure. Besides exposing antigen-derived peptides from MHC to cytotoxic T lymphocytes (CTL, in black) and natural killer cells (NK, in green), AML blasts (in red) actively express immunomodulatory molecules (e.g., indoleamine 2,3 dioxygenase—IDO) and immune checkpoint ligands to evade immune reactions (e.g., PD-L1, CD155/CD112, NKG2DL), as well as maintain sensitivity to potential immunotherapeutic agents (e.g., antibody–drug conjugates—ADCs, bispecific T-cell engagers—BiTEs, and dual-affinity retargeting antibodies—DARTs) by the expression of lineage-restricted antigens (e.g., CD33, CD123). CTL/CAR-T and NK/CAR-NK cells, together with antigen-presenting cells (APCs) are fundamental in activating and eliciting the immune response against AML; opposite to them, T-regulatory cells (T_reg), mesenchymal stromal cells (MSCs), myeloid-derived suppressor cells (MDSCs), and Ttmor-associated macrophages (TAMs) modulate the immune response by means of contact-dependent as well as soluble factors (represented as yellow circles—soluble IDO (sIDO)—and starlets—immunomodulatory cytokines e.g., IL-10, TGF-β, IL-35). Following therapy, immunotoxins and immune checkpoint inhibitors (ICPI) also contribute to the formation of the milieu.

Figure 2

Chimeric antigen receptor T-cells constructs and platforms. (A) Chimeric antigen receptor (CAR) constructs have evolved over time from a basic form that only contained a CD3ζ chain intracellular signal domain (1st generation) to more complex structures that include one (2nd generation) or two (3rd generation) costimulatory domains (most commonly CD28 or 4-1BB). More recent CARs contain interleukin-producing sections (4th generation, e.g., IL-12) or intracellular domains of cytokine receptors (5th generation, e.g., IL-2Rβ). Besides acquiring the CAR, engineered T lymphocytes usually maintain their previous T-cell receptor (TCR), depending on the specific design. Upon recognition by their CAR, CAR-T cells activate against the target cells (e.g., CD33+ AML) by inducing their apoptosis. (B) “Dual CAR” platforms can generate different combinations of CAR-Ts: in the pooled CAR-T, two different clones, each with its specific CAR, are generated; in the compound CAR-T, both CARs, each complete with a costimulatory and an activating domain, are expressed by the same cell; in the split CAR-T, two different CARs (a chimeric antigen receptor and a chimeric costimulatory receptor) are present on the same cell and linked differently to activating or costimulatory domains; finally, in the tandem CAR-T, two antigen-recognizing CARs are both linked to the same costimulatory and activating domains.

Figure 3

Mechanisms of immune escape by acute myeloid leukemia cells. The figure shows a schematic representation of selected possible mechanisms of immune escape by acute myeloid leukemia (AML) cells. All these biological changes by AML cells are ultimately responsible for immune evasion and for inducing an exhaustion phenotype in both T lymphocytes and natural killer (NK) cells (not represented).