AT1412, a Patient-Derived Antibody in Development for the Treatment of CD9 Positive Precursor B-Acute Lymphoblastic Leukemia

Despite rapid advances in immunotherapeutic options for precursor B-acute lymphoblastic leukemia (ALL), outcomes remain poor especially for adult ALL and relapsed pediatric ALL. With conventional chemotherapy, remission percentages in adult ALL range from 75 to 90%, but relapse rates are high and long-term leukemia-free survival ranges between 35-70% depending on age and risk group. The introduction of CD19 targeting immunotherapy has significantly improved patient outcomes in (relapsed) B-ALL. However, tumor escape via downregulation of CD19 occurs in a significant number of patients. Therefore an ongoing urgency remains for the identification of additional or alternative immunotherapeutic targets for the treatment of ALL.

AT1412 is an antibody that was identified from the peripheral blood memory B cell pool of a patient cured of metastatic melanoma after adoptive T-cell therapy, using a B cell immortalization technology (AIMSelect) with ectopic Bcl-6 and Bcl-xL expression as described previously [Kwakkenbos et al. Nat. Med. 2010]. The antibody was selected based on differential binding to melanoma cells as compared to healthy melanocytes and was shown to be successful in killing melanoma cells in vitro and in vivo [manuscript submitted]. In addition to melanoma, AT1412 binds other tumor types including B-ALL, gastric, colon- and pancreatic cancer. The target of AT1412 is the tetraspanin CD9, which is expressed by more than half of all B-ALL. Expression of CD9 has been correlated with adverse prognosis [Liang et al. Cancer Biomark. 2018].

We assessed binding of this human CD9 antibody to a panel of ALL cell lines using flow cytometry. Binding of AT1412 to the B-ALL cell lines SUP-B15, MHH-CALL-2 and CCRF-SB varied as expected based on the CD9 levels that we detected using a commercial CD9 antibody. AT1412 induced antibody dependent cellular cytotoxicity (ADCC) on these cells, in line with the level of AT1412 binding. No binding was seen to the T-ALL cell line Jurkat.

Importantly, these findings were confirmed in primary ALL samples, obtained prospectively at diagnosis from a cohort of patients with T- or B-ALL (n=30). AT1412 showed binding to 61% of B-ALL samples but not to T-ALL samples. The potential of AT1412 to induce ADCC was tested on patient samples from the same panel. Remarkably, AT1412 induced ADCC of all B-ALL samples it bound to (8 out of 14) and of none of the T-ALL samples. Cytotoxicity significantly correlated with the level of AT1412 binding. These findings were supported by the observation that AT1412 induced B-ALL cell death when a freshly drawn whole bone marrow sample from a patient with newly diagnosed B-ALL was cocultured with AT1412. AT1412-induced cell death of B-ALL blasts occurred without affecting the monocytic, granulocytic and lymphocytic populations. This cell death was not observed when this patient's ALL blasts were incubated with AML-targeting antibodies. Remarkably, AT1412 induced cell death in the absence of added effector cells or other (chemo)therapeutic agents, while the bone marrow sample contained over 80% blasts and as little as 3% lymphocytes. We are currently investigating the in vivo efficacy of the antibody in a humanized immune system mouse model with human B-ALL.

Taken together, the majority of precursor B-ALL blasts express CD9 and expression of CD9 is associated with a dismal outcome. Our data demonstrate that CD9 can be successfully targeted by the human CD9 antibody AT1412, suggesting that AT1412 has the potential to be developed as a therapeutic antibody for B-ALL. AT1412 is currently being advanced through preclinical development.