Clonal Kinetics and Single Cell Transcriptional Profiling of Adoptively Transferred CD19 CAR-T Cells

INTRODUCTION:

When introduced into polyclonal T cells, chimeric antigen receptors (CAR) redirect specificity of the engineered T cells to an antigen recognized by the CAR. We conducted a phase I/II clinical trial of treatment of relapsed and refractory CD19-positive B cell malignancies using a defined formulation of CD4⁺ and CD8⁺ CD19-specific CAR-T cells (NCT01865617). Little is known about the transcriptional heterogeneity of CAR-T cells in the infused product and their clonal kinetics after adoptive transfer.

METHODS:

To understand the factors that impact clonal CAR-T cell behavior in vivo, we performed TCRBV sequencing and single cell transcriptional profiling (10X Genomics) on CD8⁺ CAR-T cells isolated from infused products and the blood of treated patients. TCRBV sequencing was performed on 0.8 to 1.5 million cells from the infused product and 700-65,000 CAR-T cells from blood after CAR-T infusion. For single-cell RNA sequencing (scRNAseq), we obtained paired 5' gene expression and V(D)J data from individual CAR-T cells isolated from infused products and from the blood at the peak of in vivo expansion, after contraction, and at a late time point.

RESULTS:

High-throughput sequencing of the TCRBV genes revealed that CAR-T cells were polyclonal in the infused products, and during in vivo expansion and contraction, and at late times (≥ 3 months) after adoptive transfer. We evaluated the diversity of the TCRBV repertoire using the Shannon entropy index, and found that clonal diversity was highest in the infused product and declined at later time points after adoptive transfer. Loss of diversity after adoptive transfer was due to both expansion of higher frequency CAR-T cell clones and loss of low-frequency clones. We identified distinct CAR-T cell clones in the infused product and in blood at multiple time points after infusion that exhibited different kinetics of expansion and contraction. To examine the transcriptional programs that regulate the fate of CAR-T cells after infusion, we performed scRNAseq on CD8⁺ CAR-T cells, and found transcriptional heterogeneity in the infused products, which declined in CD8⁺ CAR-T cells isolated from patient blood after adoptive transfer. Gene set enrichment analysis showed that the infused products expressed higher levels of genes associated with hypoxia, glycolysis, and proliferation, and lower levels of genes associated with cytotoxicity compared to CAR-T cells isolated after adoptive transfer. In the infused product, genes associated with cytotoxicity were expressed at higher levels in CAR-T cells harboring clonotypes that were subsequently represented at relatively higher levels in vivo after adoptive transfer.

CONCLUSIONS:

There is transcriptional heterogeneity in the infused product and distinct CAR-T cell clones exhibit different kinetics of expansion and contraction after infusion. A better understanding of the kinetics of clonal expansion of CAR-T cells after adoptive transfer may provide insight into strategies to improve CAR-T cell immunotherapy.