Compelling Evidence for Targeted Immunotherapy As Treatment for AML

Despite numerous advances in the treatment of acute myeloid leukemia (AML) in recent years, including the development of multiple molecularly targeted agents, relapse remains a major problem, even after allogeneic stem cell transplantation. While immunologic therapies, particularly the bispecific T-cell engager blinatumomab, antibody-drug conjugate (ADC) inotuzumab, and CAR T-cells (CAR-T), have been remarkably successful in relapsed/refractory B-cell acute lymphoblastic leukemia (B-ALL),^1-3  similar approaches have yet to be successfully implemented in AML. A major roadblock limiting the success of immunotherapies in AML is the paucity of AML-specific antigens; in their absence, immunotherapies target antigens on both AML and normal hematopoietic progenitors, leading to unacceptable toxicity.

In this study, Dr. Florence Borot and colleagues present a novel mechanism using CD33-directed immunotherapy in AML that overcomes the hindrance posed by the lack of AML-specific antigens. Their approach combines targeting myeloid lineage–specific antigen (LSA), CD33, a transmembrane protein expressed on more than 90 percent of AML blasts, with the transplantation of genetically engineered stem cells lacking CD33. CD33 knockout mice develop normally without apparent hematologic defects.⁴  This group shows that hematopoietic stem/progenitor cells (HSPCs) that have undergone genetic ablation of CD33 do function normally, providing further evidence that CD33 may be dispensable. They provide additional proof that ablating a myeloid LSA such as CD33 in HSPCs enables the use of anti-CD33 CAR-T or antibody therapy, with concomitant engraftment and growth of CD33-deleted hematopoietic cells transplanted into the same host.

The group used CRISPR/ CRISPR-associated protein 9 (Cas9) to ablate expression of CD33. The Cas9 endonuclease can locate, bind, and cleave double-stranded DNA targets complementary to guide CRISPR RNAs (crRNAs).⁵  In this study, guides were designed to target exon 3 of CD33, since this sequence is common to all CD33 transcripts and has minimal similarity with the family of pseudogenes of which CD33 is a member. There was high deletion efficiency; on average, CD33 expression was reduced from 85 percent to 10 percent of CD34⁺ cells. They then tested the ability of CD33 gene-edited stem cells (CD33^del) to engraft and contribute to hematopoiesis. Mice were injected with CD33^del HPSCs, and peripheral blood analysis post-transplantation revealed the presence of mature cells of myeloid and lymphoid lineages, all CD33-negative. Importantly, there was no functional difference in the CD34⁺CD33^del and CD34⁺CD33^WT myeloid subsets, and whole-genome sequencing of HSPCs edited with Cas9/RNA complexes did not identify any insertions or deletions at off-target loci.

The CAR-T designed in this study contains a single-chain variable region of anti-CD33, along with a costimulatory, transmembrane, and intracellular domain. The CAR cDNA was cloned into a lentiviral vector. Peripheral blood from healthy donors was transduced with lentiviral particles containing the CAR construct. This demonstrated robust expression of the CAR as well as binding with CD33. CAR-T cytotoxicity was proportional to the expression level of CD33 on target cells, with the highest cytotoxicity seen in CD33⁺ myeloid leukemia cells. Irradiated mice were injected simultaneously with both CD33⁺ HL-60 leukemia cells and CD34⁺cd33^del cells, to mimic relapsed AML in the model. One week later, the mice were treated with allogeneic CD33-targeted CAR-T, control T cells, gemtuzumab ozogamicin (GO), or a combination of CAR-T and GO. No leukemia cells were found in bone marrow aspirates at 3.5 weeks or eight weeks in the CAR-T group, GO group, or combined CAR-T/GO group. Furthermore, engraftment of CD34⁺CD33^del cells was demonstrated in the therapy model by the presence of CD45⁺, CD33 negative cells in the bone marrow aspirate of all groups.