Mining single-cell transcriptomic data reveals distinct T-cell population in pediatric B-ALL and AML at diagnosis

Immunotherapy represents a promising strategy to improve outcomes in pediatric leukemia; however, its efficacy remains considerably lower in acute myeloid leukemia (AML) compared with B-cell acute lymphoblastic leukemia (B-ALL). To characterize T-cell subsets in AML and B-ALL, we took advantage of existing data sets and performed an integrated single-cell RNA sequencing analysis of T cells from pediatric patients with B-ALL (n = 89) and AML (n = 26), and healthy donors (n = 9), mostly from bone marrow at diagnosis. In total, 47 610 T cells were analyzed, revealing 17 transcriptionally distinct subsets. Comparative analysis identified T-cell subsets distinguishing B-ALL from AML, such as proliferative, naïve CD4, and progenitor exhausted, and naïve or resting CD4 regulatory T cells. We identified a rare T-cell subset expressing hemoglobin genes, which was enriched in B-ALL and characterized by upregulation of heme metabolism and chronic hypoxia-associated pathways, and its abundance was associated with better outcomes. Collectively, our findings delineate the transcriptional and functional heterogeneity of T cells in pediatric B-ALL and AML and provide insights that may inform future T-cell-based immunotherapeutic strategies.