Background Our lab has identified a new PD-L1 antibody (clone H1A) with a unique mechanism of action. Unlike current FDA-approved therapies that block PD-L1’s interaction with PD-1, H1A destabilizes PD-L1 at the cell surface and induces its degradation. Recent research using H1A have demonstrated its effect on human PBMCs, including increased effector T cell populations and cytotoxicity, compared to PD-L1 blocking antibodies. While this data is promising, there is a need to evaluate this H1A antibody in an immunocompetent host.

Methods Our lab has developed a humanized PD-1/PD-L1 chimeric mouse strain (huPD-H1), in which the murine PD-1/PD-L1 have been replaced with human PD-1/PD-L1 versions. Additionally, we engineered tumor cell lines to replace murine PD-L1 with human PD-L1 (B16F10-huPD-L1, MC38-huPD-L1, and E0771-huPD-L1). Human PD-L1 expression in the tumor cells and huPD-H1 mice were validated by flow cytometry. H1A and Atezolizumab binding to PD-L1 expressing tumor cells and huPD-H1 PBMCs were confirmed by flow cytometry. In vivo studies used a subcutaneous tumor injection, followed by antibody treatments (200ug, every 2 days for 5 total doses) once the tumors were palpable. For rechallenge studies, complete responder mice received a second subcutaneous injection of the same tumor type and another subcutaneous injection of a different tumor type on the opposing flank.

Results The tumor cell lines and PBMCs of huPD-H1 mice expressed human PD-L1. H1A and Atezolizumab showed similar binding to PD-L1 on tumor cells and huPD-H1 PBMCs. H1A and Atezolizumab did not have a response in the B16F10-huPD-L1 model, but both had a moderate response in the MC38-huPD-L1 model. Once CD8 T cells were depleted in the MC38-huPD-L1 model, the responses were abrogated. H1A and Atezolizumab had strong responses in the E0771-huPD-L1 model. Interestingly, H1A resulted in a higher percentage of complete responders compared to Atezolizumab. When rechallenged, H1A-CR’s and Atezolizumab-CR’s rejected the E0771-huPD-L1 tumor but failed to reject the B16F10-huPD-L1 tumor.

Conclusions Our preclinical in vivo studies demonstrated that while H1A and Atezolizumab treatments can have similar response rates, H1A can achieve better responses in certain tumor models. Additionally, our studies demonstrate that H1A promotes an antitumor response that relies on CD8 T cell responses and can elicit a specific memory immune response. Our work evaluates a new PD-L1 targeting antibody with a unique mechanism of action. Because of the novelty of this antibody, it can potentially increase the response rates of patients with lethal cancers.