Introduction: T cell exhaustion is a hallmark of CTCL and alterations in mRNA profiles correlate with immune checkpoint expression, with potential clinical relevance (Querfeld et al. 2018). There is no immunophenotypic marker that can distinguish malignant CD4+ T cells from benign CD4+ T cells in the infiltrate and intratumoral heterogeneity poses a major challenge to treatments and long-term remissions. The microenvironment in CTCL harbors multiple immune cells that may contribute to the development of resistance to drug treatments; however, the genomic and molecular determinants of response to therapeutic agents remain incompletely understood. The aim of our study was to distinguish malignant from non-malignant T cells based on TCR α/β repertoires and to understand the transcriptional landscapes of malignant and non-malignant cells in the TME while on anti-PD-L1 therapy.
Methods: Migrated cells from skin explants were harvested and subsequently analyzed by our paired single-cell RNA and T cell receptor (TCR; alpha/beta) sequencing on ~3000-4000 cells from skin lesions of 6 patients with mycosis fungoides at baseline and cycle 1 day 15 with anti-PD-L1 + lenalidomide.
Results: We identified 14 gene clusters. Differential expression (DE) of genes in each of the unique clusters were identified by comparing gene expression from cells in each cluster to that of all other cells in the dataset, using a cut-off of P < 0.05 and further requiring expression of the gene in >25% of cells in the cluster. Thus, DE-identified genes are expressed either uniquely or by a large proportion of cells within each cluster compared to all other clusters. TCR clones in these cells were also characterized. Through this combined analysis, we demonstrated differences in the diversity, clonal expansion and T cell phenotypes that differentiated expanded malignant T cell populations (cluster 0-3) from non-malignant T cells including tumor infiltrating lymphocytes (TILs), regulatory T cells (Tregs), NK/T cells, and from immune cells such as B cells, antigen presenting cells (dendritic cells, macrophages) and other cells (stromal, epithelial cells) (cluster 4-13). Comparing baseline to C1D15 we were able to identify microenvironmental changes that occurred during treatment, specifically characterized the expression and significance of PD1, LAG3, CTLA4, TIM3 and ICOS in malignant and non-malignant T cell clusters, which demonstrated differential expression of these targets in malignant T cells (clusters 0-4). Non-malignant T cell phenotyping revealed an enriched tumor-infiltrating CD8+ T cell population at baseline with upregulation of LAG3 gene expression, and FOXP3+ CD4+ regulatory T cell population with high expression of CTLA4 and ICOS consistent with inducible Tregs (iTregs) in all, but one baseline sample that did not resolve during treatment (C1D15).
Conclusions: Paired scRNA and TCRseq revealed distinctive functional composition of T cells and other immune cells. Combined scRNA expression and scTCR analysis identified malignant from non-malignant T cell subsets. Malignant T cell clones diminished in responders during treatment, while shifted or emerged in non-responders. Clonal enrichment of iTregs and exhausted CD4 and CD8 T cells were identified that did not resolve during treatment. suggesting that potential targeting of ICOS, CTLA4 and/or LAG3 will reverse T cell dysfunction in TILS and iTregs, respectively and increase clinical benefit of anti-PD-L1 blockade.
Querfeld:Stemline: Consultancy; MiRagen: Consultancy; Kyowa Kirin: Consultancy; Bioniz: Consultancy; Helsinn: Consultancy; Trillium: Consultancy; Celgene: Research Funding. Rosen:Novartis: Consultancy; Pebromene: Consultancy; Aileron Therapeutics: Consultancy; Celgene: Speakers Bureau; paradigm Medical Communications: Speakers Bureau; Abbvie: Speakers Bureau; Seattle Genetics: Consultancy; NeoGenomics: Consultancy.
Asterisk with author names denotes non-ASH members.
This icon denotes a clinically relevant abstract