PD1 blockade augments adoptive T cell therapy via endogenous T cells rather than direct enhancement of transferred T cells

Abstract Adoptive T cell therapy (ACT) has yielded promising clinical results but improvements are needed. Cell therapy efficacy may be improved by blockade of the inhibitory receptor PD1. Independent T cell models demonstrated cooperation between ACT and α-PD1. The cooperative effect of α-PD1 was abolished in three separate hosts devoid of polyclonal T cells (Rag1-KO, Trac-KO and irrelevant TCRtg hosts). Moreover, endogenous CD4 and CD8 T cells in the tumor increased and transferred T cells decreased in number following α-PD1 treatment. Efficacy of ACT was not improved when transferred cells were engineered with PD1-targeting miRNA or over-expression of PD1 without the cytoplasmic signaling domain. However, ACT with PD1-KO T cells was improved with the combination of α-PD1. Additionally, transferred T cells did not respond to α-PD1 when they were repeatedly stimulated until high co-expression of PD1+LAG3+TIM3+ (85% vs 34%) prior to transfer, when they were allowed prolonged in vivo tumor engagement prior to α-PD1 administration, or when they were targeting tumors with enforced expression of PDL1. Lastly, α-PD1 therapy showed no efficacy during the treatment of less immunogenic tumors, engineered with ubiquitin-tagged antigen constructs rather than RFP-antigen constructs. This finding established that PD1-sensitive endogenous T cells targeted immunogenic antigens within RFP. The role of endogenous T cells was further corroborated by TCR clonotype skewing when α-PD1 was administered. These data demonstrated that blocking the PD1 axis did not directly benefit adoptively transferred T cells. Rather, the benefit of α-PD1 was dependent on the presence of endogenous T cells specific for immunogenic tumor antigens.