ISHL12 Abstract 129

Multi-effector cell targeting with half-life extended bispecific scFv in Hodgkin lymphoma

Immunotherapy has revolutionized the treatment of many hematological cancers. One form of cancer immunotherapy is exploiting immune effector cells, for example, T-cells or NK-cells, by retargeting them to cancer-associated antigens which have been identified and validated for most hematological cancers. However, most strategies to do so employed to date suffer from disadvantages such as unfavorable pharmacokinetics and unspecific immune system activation with current bispecific antibodies, high complexity and cost with CAR T-cells or lack of efficacy with many simple monoclonal antibodies targeting tumor-associated antigens. Thus, it is highly desirable to develop new technologies that can overcome these limitations. Here, we introduce our strategy to retarget various immune effector cells simultaneously with half-life extended bispecific scFv constructs to Hodgkin lymphoma-associated antigens e.g., CD30. In doing so, we target CD3, CD16 and CD28 on immune effector cells. Half-life extension is achieved by fusing the scFv constructs to a repetitive amino acid sequence without secondary structure and immunogenicity developed by us. After optimizing the production of our constructs in mammalian protein production systems we performed functional in vitro validation. Our data shows that the designed bispecific scFv constructs can bind their targets specifically and simultaneously and facilitate the formation of an artificial immunological synapse with immune cells secreting toxic proteins such as perforin and granzymes leading to the apoptosis of the lymphoma cells. We show direct cytotoxicity towards tumor cells in vitro and a target-cell specific release of cytokines such as TNFα and IFNγ by effector cells. We are now moving some constructs into cell line xenograft mouse models for further in vivo validation.

Authors

  • Zhiyuan Shi
  • Verena Kern
  • Max Freihammer
  • Jennifer Schitz
  • Sven Borchmann