Weekly Digests
‹ Back to February

Bispecific antibody 7370: will AML meet its match?

February 12, 2020

In an effort to address the significant unmet clinical needs in the treatment of acute myeloid leukemia (AML), Yeung and Krishnamoorthy et al. developed a novel bispecific antibody that targets both CD3 on T cells and FLT3 on AML cells. FLT3 was selected as a strong potential target antigen, given its upregulated expression on AML blasts in a high percentage of patients, its role as an oncogenic driver, and its absence from most healthy cell types, besides dendritic cells (DCs) and hematopoietic stem and progenitor cells (HPSCs), which typically express FLT3 at low levels. The researchers optimized and tested their bispecific antibody, deemed 7370, and their results were recently published in Molecular Therapy.

In order to design an optimized bispecific antibody, Yeung and Krishnamoorthy et al. screened a panel of human antibodies and selected representative antibodies targeting different domains of human FLT3 to be incorporated into full-length bispecific antibodies co-targeting CD3. In an in vitro cytotoxicity assay against the AML cell line EOL-1, bispecific antibodies targeting the fourth (the second-most membrane proximal) and fifth (the most membrane proximal) domains were most effective. In NSG mice bearing subcutaneous EOL-1 AML tumors, the bispecific antibody targeting the fourth domain clearly showed greater antitumor efficacy, despite not having the highest affinity, which may reflect a more optimal geometry of the synapse. Several antibody clones targeting domain 4 were then tested, and the top clone was selected to be incorporated into a bispecific antibody. The resulting full-length FLT3/CD3-targeting bispecific antibody, 7370, was able to specifically bind to recombinant FLT3 and CD3 proteins, human T cells, and the AML cell lines EOL-1 (high FLT3 expression) and MV4-11 (low FLT3 expression).

In short-term cytotoxicity assays using healthy donor T cells and EOL-1 or MV4-11 AML cell lines, 7370 induced nearly complete lysis of both EOL-1 or MV4-11 AML cells at an effector-to-target (E:T) ratio of 1:1. Substantial target cell lysis of both EOL-1 or MV4-11 AML cells was observed at an E:T ratio as low as 1:20, though less lysis was induced in MV4-11  cells, suggesting that antigen density affects the efficacy of 7370. In the presence of FLT3+ AML blasts (but not FLT3- cell lines), 7370 potently induced T cells to express activation markers (CD25, CD69, 4-1BB, and PD-1) and produce IFNγ, TNFα, and IL-2, consistent with TCR-driven T cell activation. Evidence of T cell activation was transient, and disappeared (except for CD25) when tumor cells were eliminated.

To test the efficacy of 7370 in vivo, Yeung and Krishnamoorthy et al. engrafted NSG mice with either EOL-1 (high expression of wild-type FLT3), MOLM-13 (medium expression of mutant FLT3), or MV4-11 (low expression of mutant FLT3) cell lines to generate orthotopic xenograft models of AML. Mice were injected with previously activated and expanded human T cells, followed by a single injection of 7370 at various doses. 7370 was able to protect all mice against EOL-1 or MOLM-13 for at least 40 days post-implantation, though a higher dose was required to achieve complete control over MOLM-13 growth. A single dose of 7370 reduced the growth of MV4-11 cells, but was not sufficient to eliminate them entirely, suggesting that low FLT3 expression could be a barrier to antitumor efficacy.

Having shown that 7370 can induce antitumor efficacy in mice, the researchers went on to explore its potential clinical relevance. To this end, they tested PBMCs from 5 patients against their own AML blasts in the presence of 7370 in vitro. The E:T ratio in these samples ranged from 1:25 to 1:64, and expression of FLT3 on AML cells varied. Despite the low E:T ratios, 7370 induced T cell proliferation in a dose-dependent manner, as well as substantial cytotoxic responses from T cells in 4 out of 5 samples. In 2 out of 4 samples, AML cells were over 80% depleted by day 7. The dose of 7370 required was significantly higher than with T cells from healthy individuals, possibly reflecting a lower T cell quality or presence of inhibitory checkpoints in patient-derived T cells.

In order to evaluate potential hematological toxicities associated with 7370, Yeung and Krishnamoorthy et al. tested their bispecific antibody in healthy cynomolgus monkeys. After showing that 7370 was cross-reactive to cynomolgus FLT3, and that, like in humans, FLT3 was expressed on DCs in the blood and on CD34+ HSPCs in the bone marrow, four cynomolgus monkeys were treated with two doses of 7370, administered one week apart. One week after the initial dose, FLT3+ DCs (CD1c+ and CD123+) and FLT3+CD34+CD38+ HSPCs were nearly or completely depleted. Two of the monkeys were observed for a two-week recovery period, during which levels of FLT3+ DCs and HSPCs were restored roughly to baseline levels. Other immune cell subsets, including B cells and monocytes, were also depleted following treatment with 7370, but these too were restored over the two-week recovery period. Evaluation of cytokine release patterns showed that 7370 increased expression of IL-6, IL-10, and IFNγ after first, but not the second, dose. Cellularity was minimally to mildly decreased in the bone marrow, consistent with the depletion of FLT3+ cells, and cellularity was minimally to moderately increased in spleens and axillary lymph nodes, consistent with T cell activation. One monkey experienced low-grade adverse events, but overall, 7370 was well-tolerated.

Overall, evidence from mice, patient samples, and monkeys, shows that 7370 can effectively redirect T cells to target and kill FLT3+ target cells, in a dose-dependent and antigen density-dependent manner. 7370 was well-tolerated and its effects on the immune system appeared to be reversible, suggesting that 7370 has the potential to be safe and effective against AML in a clinical setting.

by Lauren Hitchings

Meet the researcher

This week, first author Yik Andy Yeung answered our 3 questions. 

What prompted you to tackle this research question?
We would like to develop a therapy that can broadly treat AML with durable response. AML is a heterogeneous disease. Continual understanding of the underlying disease mechanism did lead to the development of novel targeted therapies, however, such therapies only benefit subset of patients and the disease often relapses. Meanwhile, T cell redirection therapy (Blinatumomab: CD19-CD3 for ALL) showed great promise in delivering long-term responses. However, proper target selection is critical for T cell redirection therapy to work, as broad expression of the target receptor in healthy tissues can severely limit the therapeutic efficacy. Therefore, we sought to identify a target receptor on AML that is critical for their survival and differentially overexpressed as compared to normal cells. That’s how we arrived at FLT3 as a T cell redirection target for AML.

What was the most surprising finding of this study for you?
One of the surprise findings was on the engineering of the candidate molecule. Multiple studies suggested that for T cell redirection therapy, bispecific antibody targeting the membrane proximal region of the target receptor can help bring the T cell closer to the cancer cell, thereby mediating more effective killing. However, when we tested the efficacy of bispecific antibodies targeting different regions of the FLT3 receptor, we found that the most optimal domain to target is not the most membrane proximal domain, rather it is the adjacent domain further away from the membrane. This implies that proper screening of the binding epitope on the target receptor is very important for the engineering of the bispecific antibody.

What was the coolest thing you’ve learned (about) recently outside of work?
Recently, I sat in on my son‘s karate lesson. My son was asked by the sensei (teacher) to demonstrate certain skill sets for the younger apprentice. Watching him performing the forms and steps confidently in the class suddenly reminded me of him at an earlier age, when he could not properly do these steps. However, through the help of sensei and practices, he managed to learn the steps over time. Beside being proud of him, I am very grateful for the sensei’s passion and patience in teaching him. It is such a blessing in life to have a teacher/mentor with such passion, knowledge, and patience.


Yeung Y.A., Krishnamoorthy V., Dettling D., Sommer C., Poulsen K., Ni I., Pham A., Chen W., Liao-Chan S., Lindquist K., Chin S.M., Chunyk A.G., Hu W., Sasu B., Chaparro-Riggers J., Djuretic I. An Optimized Full-Length FLT3/CD3 Bispecific Antibody Demonstrates Potent Anti-leukemia Activity and Reversible Hematological Toxicity. Mol Ther. 2020 Jan 14.

In the Spotlight...

Interleukin-23 engineering improves CAR T cell function in solid tumors

Ma and Xu et al. found that TCR stimulation of memory-like T cells upregulated IL-23R and the IL-23α p19 subunit, but surprisingly not the IL-12β p40 subunit of IL-23. In xenograft and syngeneic cancer models, p40-transduced CAR (CAR.p40) or antigen-specific TCR T cells secreted IL-23 upon CAR or TCR stimulation, showed enhanced proliferation and persistence, upregulated STAT3-responsive genes, and conferred enhanced tumor control and survival. IL-23 produced by CAR.p40 T cells acted in an autocrine fashion, with limited effects on bystander cells. CAR.p40 cells had better efficacy and safety than CAR T cells expressing IL-18 or IL-15.

Contributed by Anna Scherer

REVIEW: Neoadjuvant checkpoint blockade for cancer immunotherapy

Topalian et al. reviewed results of neoadjuvant (presurgical) PD-1 blockade in earlier stage cancer, focusing on emerging biological mechanisms, trial designs, response evaluation, and immune-related tumor pathology. Neoadjuvant PD-1 inhibition enhanced antitumor T cell immunity by direct DC priming of T cell responses, intratumorally and in tumor-draining LNs, and expanding additional sub-dominant tumor-specific T cell clones. In neoadjuvant trials of several different tumor types, histologically measuring immune-mediated tumor regression, coupled with the rich correlatives from the resected tumor tissue, may provide early surrogate markers.

Contributed by Katherine Turner

Antigen Experienced T Cells from Peripheral Blood Recognize p53 Neoantigens

Malekzadeh et al. show that peripheral blood lymphocytes (PBL) contain a small fraction of mutated TP53 neoantigen-reactive T cells, which was consistent with tumor-infiltrating lymphocyte response to mutant TP53. In vitro stimulation (IVS) using either p53 neoantigen long peptide or tandem minigenes, with multiple hotspot TP53 mutations, followed by 4-1BB/OX40 enrichment, increased the frequency of PBL-derived reactive T cells to as high as 70%. IVS increased the TCRB clonality in PBL-derived T cells, and IVS T cells recognized naturally presented, HLA-restricted neoantigens.

Contributed by Shishir Pant

Control of metastases via myeloid CD39 and NK cell effector function

Yan and Li et al. demonstrated that an antibody blocking CD39 ATPase enzymatic activity reduced metastases in several murine tumor models. Metastasis control was dependent on NK cells, IFNγ, and expression of CD39 on myeloid cells, and was independent of Fc, CD8+ T cells, or CD39 expression on NK cells. The mechanism required expression of the ATP receptor P2X7 on myeloid cells, as well as functional downstream inflammasome and IL-18 production, which triggered NK cell effector responses. The antitumor effect was further improved when anti-CD39 was combined with other immunotherapies (anti-PD-1, IL-2, IL-15, or A2AR inhibitor).

Contributed by Anna Scherer

Human pDCs Are Superior to cDC2s in Attracting Cytolytic Lymphocytes in Melanoma Patients Receiving DC Vaccination

Analysis of chemokine gene expression/secretion showed that activation of pDCs and cDC2s led to biased expression of CXCR3/CCR5 (Th1)- and CXCR1/R2 (Th2/17)- associated ligands, respectively. CD8+αβ, CD56+, and γδ T and NK cells were recruited by activated pDC+/-cDC2-conditioned media in vitro, were detected by IHC in sections of, and by FACS in cultures of skin biopsies of melanoma patients vaccinated intradermally (ID) with activated pDCs and cDC2s, and were shown by spectral imaging clustered near DCs in the deep reticular dermis of melanoma patients injected ID with activated pDCs. cDC2s recruited CD4+ Th2/17, but not Th1, cells in vitro.

Contributed by Paula Hochman

Microenvironment-Dependent Gradient of CTL Exhaustion in the AE17sOVA Murine Mesothelioma Tumor Model

Hope and Spantidea et al. showed that in mice bearing an OVA-secreting, asbestos-induced mesothelioma cell line, adoptively transferred naive OT-I CD8+ T cells initially expanded and produced high levels of IFNγ and TNFα within the tumor at day 15, but by day 22, intratumoral donor OT-I T cells showed signs of exhaustion, including significant contraction, loss of polyfunctionality, and increased expression of multiple inhibitory receptors. On day 22, donor OT-I CD8+ T cells were most exhausted in the tumor, less exhausted in the tumor-draining lymph nodes, and least exhausted in the spleen.

Contributed by Anna Scherer

Everything New this Week In...

Close Modal

Small change for you. Big change for us!

This Thanksgiving season, show your support for cancer research by donating your change.

In less than a minute, link your credit card with our partner RoundUp App.

Every purchase you make with that card will be rounded up and the change will be donated to ACIR.

All transactions are securely made through Stripe.