Background According to The American Cancer Society, ovarian cancer (OC) will cause ~13,000 deaths in the US in 2024. Despite advances in therapy, clinical outcomes remain poor and there is a great unmet need for newer therapies for OC patients. Bispecific T cell engagers (BTEs) consist of two single chain variable fragments (scFvs) linked via a peptide linker. Several BTEs targeting hematological cancers have been approved by the FDA, however the impact of BTEs for solid tumor therapy remains limited. 1^st-generation BTEs have a short half-life, reducing availability and compromising their ability to control solid tumors. Solid tumors are heterogeneous making it difficult to get complete tumor control with a single BTE. Tumors downregulate TAA under immune pressure, which further necessitates simultaneously targeting multiple TAAs to improve anti-tumor effect. BTEs have complex manufacturing processes and stringent cold chain requirements which increase the cost of therapy. Development of novel BTEs with improved half-lives and in vivo delivery strategies which allow rapid customization of BTEs is important.

Methods We designed two DNA encoded BTEs targeting FSHR and Her2 using the knob-into-hole design and studied their ability to be assembled in vivo into functional therapeutics. The scFvs targeting FSHR and Her2 were linked to hFc with mutations to create the knob. The scFv targeting CD3 was linked with hFc to create the hole. After establishing potency and specificity of KIH_BTEs in vitro, we studied their pK and activity in vivo.

Results FSHRxCD3_KIH and Her2xCD3_KIH BTEs demonstrated potent (femtomolar levels) and specific killing in vitro across several OC cell lines derived from tumors with different mutation profiles and resistance to many first-line OC therapies. Both KIH_BTEs demonstrated superior pK and significantly improved tumor control compared to the respective 1^st-generation BTEs. Finally, mice treated with combination of both KIH_BTEs had improved tumor control compared to those treated with single KIH_BTE.

Conclusions We demonstrate that using DNA delivery, complex Knob-into-hole designs can be formed in vivo. We observed that FSHRxCD3_KIH and Her2xCD3_KIH are extremely potent in vitro and can significantly extend the half-lives of BTEs in vivo. KIH_BTEs are also better at controlling tumors in vivo compared to 1^st-generation BTE. We found that there is no cost in expression/activity when we co-deliver two KIH_BTEs simultaneously. Finally, we demonstrate that antigen heterogeneity and antigen downregulation can be overcome by co-delivering KIH_BTEs targeting two different TAAs. Further development of KIH_BTEs for treatment of OC is warranted.