Off-the-shelf TCR for graft-versus-leukemia without GVHD

With the recent US Food and Drug Administration approval of chimeric antigen receptor (CAR) T cells for immunotherapy of CD19⁺ B-cell leukemia and lymphoma,^2,3  why pursue such a complex approach for the treatment of hematological malignancies?

The development of adoptive cell therapies for myeloid diseases has not advanced to the same degree as anti-CD19 CAR T-cell therapies. Extensive studies using genomics and proteomics suggest that an ideal surface target may not exist in acute myeloid leukemias (AMLs).⁴  Myeloid lineage–specific molecules are shared with hematopoietic stem cells, a clear limitation for the development of CAR T cells directed against nonpolymorphic cell surface receptors: pancytopenia is not an acceptable toxicity as opposed to B-cell aplasia after anti-CD19 CAR T-cell therapy.

The ability of TCRs to target peptides derived from intracellular proteins in AML and other cancers is a design advantage over CARs. Ideally, the T cell should target mutated proteins that are unique to each individual tumor. Solid tumors induced by carcinogens, for example, melanoma and lung cancer, express numerous mutations that create tumor-specific antigens, which can be targeted by T cells expanded in vivo after neutralizing negative immune checkpoints,⁵  or by adoptive transfer of ex vivo expanded tumor-infiltrating lymphocytes.⁶  Unfortunately, tumor-specific antigens are rare in hematologic malignancies because of the low mutation load, and alternative immunotherapy targets must be found.

A variety of nonpolymorphic tumor-associated antigens are preferentially expressed on myeloid leukemias, including products of oncogenic mutations, such as the Bcr/Abl tyrosine kinase, tissue-specific proteins, such as proteinase-3, and tumor-selective, overexpressed self-proteins, such as WT1. In general, these are weakly expressed antigens, and it remains uncertain whether they can elicit tumor immunity sufficient for a meaningful clinical impact.^7,8 

Preclinical studies have shown that injection of allogeneic T cells primed against a single minor histocompatibility antigen can cure hematologic and solid cancers without causing toxicity to the host. The key advantage of targeting minor histocompatibility antigens, such as HA-1, for leukemia immunotherapy is their high level of expression that is restricted to hematopoietic tissues, without expression in nonhematopoietic tissues, so that the T cells would neither elicit graft-versus-host disease (GVHD) nor cause other undesirable off-target effects.⁹ 

Minor histocompatibility antigens, derived from polymorphic sequences that differ between individuals, are expressed as intracellular proteins, processed, and presented on the leukemia cell surface to TCRs as peptides by HLA molecules. Therefore, the limitations of the proposed approach for cellular immunotherapy of leukemia include the requirement for patient expression of the targeted antigen (HA-1) and the HLA restriction element (HLA-A*02:01) and the need for an allogeneic stem cell transplant from a histocompatible donor that does not express HA-1. The strategy would only work after an allogeneic stem cell transplant from an HA-1–negative donor, because T cells against HA-1 would cause pancytopenia after adoptive transfer into an HA-1–positive recipient. Given these constraints, the approach is translatable to ∼15% to 20% of transplanted patients. If successful, it will provide proof of principle applicable to targeting other minor histocompatibility antigens in almost all patients.¹⁰ 

The authors address several other fundamental problems that could hamper efficacy or safety of the T cells (see table). These include the weak immunogenicity and variable expression of typical intracellular tumor-associated antigens, which could be targeted by TCR gene therapy; the need for both CD8⁺ cytotoxic and CD4⁺ helper T cell anti-tumor specificity; and safety concerns with on-target, off-tumor toxicity to normal myeloid cells. A single lentiviral vector encodes a high affinity TCR against HA-1, a CD8 coreceptor secures the TCR function in CD4 T cells, a selectable marker allows enumeration and enrichment of transfected T cells, and a safety switch could stop the cells in case of toxicity. After transfection, T cells from any donor gain potent antitumor activity.

These elegant molecular biological engineering strategies to target minor histocompatibility antigens combined with allogeneic transplant can overcome specific barriers to successful antitumor immunity for AML.¹  Results of clinical investigation using adoptive transfer of donor T cells engineered with this construct will be eagerly sought. Adoptive cell therapy for AML may now come off the shelf, literally and figuratively.