Differential predictive value of resident memory CD8 ⁺ T cell subpopulations in non-small-cell lung cancer patients treated by immunotherapy

A high density of resident memory T cells (T _RM ) in tumors correlates with improved clinical outcomes in immunotherapy-treated patients. However, in preclinical models, only some subpopulations of T _RM are associated with cancer vaccine efficacy.

We identified two main T _RM subpopulations in tumor-infiltrating lymphocytes derived from non-small cell lung cancer (NSCLC) patients: one co-expressing CD103 and CD49a (DP), and the other expressing only CD49a (MP); both exhibiting additional T _RM surface markers like CD69. DP T _RM exhibited greater functionality compared to MP T _RM . Analysis of T-cell receptor (TCR) repertoire and of the stemness marker TCF-1 revealed shared TCRs between populations, with the MP subset appearing more progenitor-like phenotype. In two NSCLC patient cohorts, only DP T _RM predicted PD-1 blockade response. Multivariate analysis, including various biomarkers (CD8, TCF1 ⁺ CD8 ⁺ T cells, and PD-L1) associated with responses to anti-PD(L)1, showed that only intra-tumoral infiltration by DP T _RM remained significant. This study highlights the non-equivalence of T _RM populations and emphasizes the importance of distinguishing between them to better define their role in antitumor immunity and as a biomarker of response to immunotherapy.