Tissue-resident memory T cells play a key role in the efficacy of cancer vaccines

ABSTRACT

Resident memory CD8⁺T cells (T_RM) usually defined by the CD103 marker represent a new subset of long-lived memory T cells that remain in the tissues. We directly demonstrate their specific role in cancer vaccine-induced tumor regression. In human, they also seem to play a major role in tumor immunosurveillance.

KEYWORDS:

• resident memory T cell
• cancer vaccine
• tumor microenvironment
• route of immunization
• immunosurveillance

Cancer immunotherapy based on immunomodulators (anti-CTLA-4, anti-PD-1/PD-L1) has recently demonstrated clinical efficacy for the treatment of cancer patients, but up until now cancer vaccines have failed to show any clinical impact in these same patients. Preclinical models have emphasized the role of CD8⁺T cells in tumor protection mediated by cancer vaccines. However, in humans, no clear correlation has been observed between the detection of anti-tumor CD8⁺T cells in the blood and their clinical benefit after cancer vaccine administration.¹

Resident memory CD8⁺T cells (T_RM) usually defined by the CD103 marker represent a new subset of long-lived memory T cells that remain in the tissues, do not recirculate and therefore cannot be monitored by whole blood analysis. T_RM appear as a specific T cell lineage as the transcription factor Hobit together with Blimp1 mediated the development of these cells from T cell precursor²

In a recent paper, we address the role of this population in the efficacy of cancer vaccines. After mucosal (intranasal) but not systemic (intramuscular) administration of a cancer vaccine targeting the E7 protein from HPV16, we induced anti-E7-T_RM in the head and neck and lung tissue which correlated with the inhibition of the orthotopic tumor growth.³ We then selected various strategies to specifically define the role of these cells, not biased by anti-E7 circulating CD8⁺T cells also induced by the vaccine: i) the use of FTY720, which blocks the recruitment of circulating T cells in the tissues; ii) co-administration of the vaccine with anti-TGFβ, which preferentially elicits circulating effector T cells over T_RM; iii) parabiosis experiments, in which the blood circulations of 2 mice are connected, but not the non-circulating T cells (T_RM). In all these experiments, we demonstrated that T_RM are required for the efficacy of the cancer vaccine to inhibit tumor growth and that the presence of circulating CD8⁺T cells alone was not sufficient to control tumor growth.³ In the past, the role of T_RM has already been demonstrated after vaccination against pathogens.¹ In preclinical model of cancer, mucosal immunization was more efficient than the conventional systemic route (intramuscular, subcutaneous) to elicit T_RM at the mucosal tumor site.⁴ A correlation was observed between the ability to elicit these cells at the tumor site and control of tumor growth.^1,4-5 When dendritic cells were reprogrammed via treatment with the β-glucan, curdlan, a ligand of dectin-1, to induce mucosal antitumor CD8⁺T cells expressing CD103, these cells with a T_RM phenotype were found to be more efficient than control CD8⁺T cells to inhibit tumor growth.⁶ The location of T_RM at the site of tumor, the enrichment of T_RM in the tumor antigen specific fraction of CD8⁺T cells, their constitutive expression of deployment-ready mRNAs encoding pro-inflammatory cytokines and cytotoxic mediators and the potentiation of their lytic function after the CD103-E cadherin (expressed by epithelial tumor cells) interaction may explain the particular efficacy of T_RM in tumor immunosurveillance⁷ (Fig. 1). In addition, they express multiple chemokines which could recruit and amplify secondary local recruitement of effector immune cells (Fig. 1).

Figure 1. Properties of resident memory CD8⁺T cells explaining their specific role in tumor immunosurveillance. TRM is located in the tumor nest via its specific interaction between CD103 and E-cadherin expressed by epithelial cells. This interaction and the preformed mRNA for cytotoxic molecules (IFNγ, TNF, Granzyme B) increase their cytotoxic activitiy toward tumor cells. They are enriched in the tumor specific fraction of CD8⁺T cells. They produce chemokines and IFNγ also favors chemokine production by epithelial cells, which results to a secondary recruitment of anti-tumor effector cells.

Mucosal priming seems to be required for the induction of T_RM in lung tissue, as after nasal but not systemic vaccination, a co-culture of DC collected from the lung with CD8⁺T cells could promote a T_RM phenotype .⁸ In human, a lung-tissue-resident CD1c⁺DCs were able to drive CD103 expression on CD8⁺T cells.

To extrapolate these results to the role of T_RM in human cancers, we set up an in situ multiparametric immunofluorescence study and clearly demonstrated the presence of T_RM in human non-small cell lung tumor tissue. In multivariate analysis, the presence of T_RM in a series of human lung cancers was correlated with better overall survival than other infiltrating immune cells. Interestingly, Ganesan et al recently confirmed these results by identifying the prognostic value of T_RM in human lung cancer independently of the CD8⁺T cell density.⁹

In melanoma patients vaccinated with a mixture of Melan-A peptide combined with CpG and Montanide, the ability to elicit anti-Melan A CD8⁺T cells expressing VLA1, a surrogate marker of T_RM, was correlated with better survival of vaccinated patients.¹⁰

Studies from our group and others therefore strongly argues in favor of the induction of T_RM as a new goal to develop clinically successful therapeutic cancer vaccines. We also propose that the mucosal route of administration should be preferred to the systemic route (intramuscular, subcutaneous route) to preferentially elicit these cells. Our result could explain why monitoring of anti-tumor CD8⁺T cells often leads to controversial results in terms of correlation with the efficacy of cancer vaccines. T_RM may constitute a new surrogate biomarker of spontaneous or vaccine-induced cancer immunosurveillance.