While cancer immunotherapy is a promising strategy for improving patients’ survival and quality of life, currently only a fraction of cancer patients (~25-35%) respond to checkpoint inhibitor therapies (CIT). Patients with inflamed tumors, where CD8 T cells infiltrate the tumor, respond better to CIT compared to those with immune excluded tumors, where CD8 T cells accumulate at the tumor-stroma boundary. 

The excluded immune phenotype is found in more than 50% of all human tumors and is characterized by elevated TGFβ signaling¹. Preclinical studies demonstrated that combining CIT with TGFβ inhibition converts immune excluded tumors to inflamed and increases anti-tumor responses². Unfortunately, this increased anti-tumor efficacy has not been observed in the clinic so far. The lack of efficacy has been accompanied by on target toxicities³. This disconnect between what has been observed in animal models and the clinic warrants additional investigations into the mechanism of action of the CIT/anti-TGFβ combination, with the hope that having a deeper understanding of the biology can help guide future clinical trials. 

Despite the fact that CD8 T cells are the key component of the anti-tumor response, the direct or indirect impact of TGFβ inhibition combined with CIT on CD8 T cells origin and differentiation remains incompletely understood. With this in mind, we aimed to perform an in-depth analysis of the effects of anti-PD-L1/TGFβ combination therapy in a mouse model of T cell exclusion:  the EMT6 breast cancer model. We used a multi-omics approach to analyze how anti-PD-L1/TGFβ combination therapy affects tumor infiltrating CD8 T cells and their interaction with the other main cell types in the tumor microenvironment (TME): tumor cells, myeloid cells and fibroblasts. We show that in immune excluded tumors CD8 T cells are present with three major phenotypes: (1) progenitor-like CD8 T cells (T_SCL) with a diverse T cell repertoire, (2) clonally expanded T progenitor exhausted cells (T_PEX) characterized by high levels of TGFβ signaling and expression of exhaustion markers like TOX, and (3) different flavors of effector T cells (T_EFF) characterized by either high expression of granzymes or IFNγ. Using bioinformatic inferences based both on single-cell RNA sequencing (scRNAseq) and T cell receptor sequencing (scTCRseq) analyses, we show that T_SCL are the progenitor cells that can give rise to the different CD8 T cell phenotypes present in the TME. Anti-PD-L1/TGFβ combination therapy expands T_SCL within the tumor and pushes CD8 T cell differentiation towards the IFNγ^hi phenotype, effectively fueling an IFNγ dependent anti-tumor response. IFNγ neutralization or IFNGR1 deletion in EMT6 tumor cells impaired anti-tumor response demonstrating the central role of IFNγ response induced by anti-PD-L1/TGFβ combination therapy. 

Our analysis in an immune excluded mouse model describes how anti-PD-L1 and anti-TGFβ induces a shift towards a more effective CD8 T cell population in the TME and transforms myeloid, stromal, and tumor niches to yield an immune-supportive ecosystem. The data published in this paper improves our understanding of TGFβ biology and might help guide future clinical trials, where the expansion of T_SCL could be monitored as a measure of patient’s response to the combination treatment.

1. Hegde, P. S. & Chen, D. S. Top 10 Challenges in Cancer Immunotherapy. Immunity 52, 17–35 (2020).
2. Mariathasan, S. et al. TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature 554, 544–548 (2018).
3. Robbrecht, D. et al. Safety and efficacy results from the expansion phase of the first-in-human study evaluating TGFβ inhibitor SAR439459 alone and combined with cemiplimab in adults with advanced solid tumors. J. Clin. Oncol. 40, 2524–2524 (2022).