CSL362: A Monoclonal Antibody to Human Interleukin-3 Receptor (CD123), Optimized for NK Cell-Mediated Cytotoxicity of AML Stem Cells

Abstract 3598

The interleukin-3 receptor alpha chain (IL-3Rα/CD123) is expressed in a variety of hematological malignancies including AML, MDS, B-ALL, Hodgkin's lymphoma, hairy cell leukemia, systemic mastocytosis, plasmacytoid dendritic cell leukemia and CML. In AML, the majority of AML blasts express CD123 and this receptor is selectively over expressed on CD34⁺CD38⁻ leukemic stem cells (LSC) compared to normal hematopoietic stem cells. This difference may provide a biological advantage to the leukemic cells given the survival and proliferation-promoting activities of IL-3, whilst at the same time providing an opportunity to target these malignant cells selectively.

We have shown previously that 7G3, a mouse monoclonal antibody (mAb) which blocks IL-3 binding to CD123, is capable of eliminating human LSC in a mouse model of human AML by a combination of mechanisms, including engagement of the innate immune system via Fc-dependent mechanisms (Jin et al., 2009 Cell Stem Cell, 5:31). We have subsequently humanised and affinity-matured this antibody and, in addition, have engineered the Fc-domain to optimise potential cytotoxicity against AML cells.

The resultant antibody, CSL362, retains the ability to neutralise IL-3 and has enhanced affinity for the FcγRIIIa (CD16) on NK cells. In vitro studies have demonstrated that the increased affinity for CD16 correlates with greater antibody-dependent cell-mediated cytotoxicity (ADCC) against CD123 expressing cell lines compared to CSL360, a non Fc-engineered anti-CD123 mAb. The improved activity was evident as both an increased maximal level of target cell lysis and as a shift in the EC₅₀ of the antibody to lower concentrations. Importantly, both primary AML blasts and CD34⁺CD38⁻CD123⁺LSC were susceptible to CSL362-induced ADCC and this was seen even in samples that were resistant to ADCC by a non Fc-engineered anti-CD123 mAb. In an AML xenograft mouse model, where treatment with the antibody was initiated 4 weeks after engraftment of leukemia cells, CSL362 was more effective in reducing leukemic growth than the non Fc-engineered anti-CD123 mAb.

The evaluation of neutrophils, monocytes, macrophages and NK cells in ADCC assays has revealed that the major effector cell responsible for CSL362-mediated cytotoxicity in human peripheral blood is the NK cell. In clinical samples we have been able to demonstrate autologous depletion ex vivo of target AML blasts (collected at diagnosis and cryopreserved) following incubation with CSL362 and peripheral blood mononuclear cells (taken from the same patient at first remission), indicating that NK cell number and function is sufficiently preserved in such patients for CSL362-directed killing of leukemic target cells.

The pre-clinical data generated thus far strongly support the clinical development of CSL362 for the treatment of AML in patients with adequate NK cell function. A Phase 1 study of CSL362 in patients with CD123 positive AML in remission is underway (Clinical Trials.gov identifier: NCT01632852).