Abstract 4476: PD-1 and TIGIT co-expression identifies a circulating CD8 T cell population predictive of response to anti-PD-1 therapy in melanoma and Merkel-cell carcinoma patients

Immunotherapies targeting the PD-1 pathway have profoundly transformed the clinical care of cancer patients for a growing variety of cancer types. However, most patients do not experience durable clinical benefit. The definition of robust and convenient biomarkers of PD-1 therapy efficacy to stratify patients beforehand or early after initiation of the therapy that could guide therapeutic management is still lacking while being a very active research field. Biomarkers described to date include tumor burden, neoantigen load, presence and number of PD-1+ CD8+ at the tumor margin, T-cell inflamed tumor microenvironment and PD-L1 expression by the tumor cells or other immune cells and composition of the gut microbiota. Most of these parameters are closely related/influenced by the presence, activation status and functional capacities of CD8+ T cells infiltrating the tumor site demonstrating their pivotal role for anti-PD-1 mediated anti-tumor efficacy. A population of PD-1high CD8 TILs was consequently described as predictive of PD-1 blockade in NSCLC. The exact contribution for clinical efficacy of TILs versus distinct CD8+ T cells from peripheral origins recirculating to the tumor site remains to be elucidated. Notably, immunological responses to PD-1 blockade at the periphery were described within the very first days following the first therapy dose. Therefore, describing circulating cellular population predictive of PD-1 inhibitor efficacy could represent a convenient, non-invasive and rapid method to assess anti-tumor benefits. Original findings reported in this study identified a circulating CD8 T cell population delineated by the co-expression of TIGIT and PD-1 inhibitory receptors as an early immune marker of anti-PD-1 efficacy in three independent cohorts of cancer patients (two melanoma patient's cohorts and one Merkel-cell carcinoma patient's cohort). The frequency of this double positive (DPOS) population even appeared predictive of PD-1 inhibitor therapy efficacy at baseline in the MCC cohort. Furthermore, to understand the mechanistical relevance of this subset for PD-1 blockade efficacy, we thoroughly described this DPOS T cell subset by flow cytometry, gene expression analysis, anti-tumor reactivity assay and TCR repertoire analysis, and compared it to its double negative (DNEG), PD-1 and TIGIT single positive counterparts. This DPOS subset was enriched in activated and proliferative T cells, retained expression of co-stimulatory molecules and