Table 3 Summary of key findings related to single-cell works revealing primary tumor microenvironment in leukemia

AML

scRNA-seq

The diversity of immunosuppressive CD206 + and CX3CR1 + dendritic cells and different M2 macrophages was defined. Several unique subtypes of TH17-like intermediate population, cytotoxic CD4 + T subset and CD8 + memory-like subset were also identified in AML TME.

Offered a comprehensive AML TME profiling, revealing potential immunotherapy targets.

[81]

AML

scRNA-seq

M2-type macrophage with enhanced oxidative activity and impaired phagocytosis ability in the AML microenvironment. Also, in vitro exposure of leukemic blast to M2 macrophage resulted in the accumulation of CALR-low blast enrichment and the exchange of mitochondria with M2 macrophage. The mechanisms enhanced the survival of AML cells.

Proposed the importance of the interaction of M2 macrophages with AML cells. Revealed potential therapeutic target in terms of metabolism (e.g. FAO/mitochondrial ETC).

[83]

B-ALL

CITE-seq; scRNA-seq

Monocyte abundance is poor prognostic predictive in B-ALL. Non-classical (CD16+) monocyte was attracted by B-ALL and Anti-CSF1R therapy targeting CD16 + monocytes improved the therapeutic outcomes.

Noted that the non-classical monocyte predicts patient survival. Targeting CSF1R of these monocytes together with TKI improved therapeutic outcomes in animal models, revealing a potential therapy combination.

[36]

B-ALL

scRNA-seq

Changes in AP-1-regulated genes were observed in normal pro- and pre-B cells at an early stage of B-ALL. GMP showed tumor suppressor Neat1 downregulation. Monocyte-dendritic precursors (MDP) were continuously active during disease progression. Monocytes increased the interaction with GMP and MDP during progression.

Noted that targeting the MDP, GMP, and monocytes may improve therapeutic outcomes in B-ALL.

[85]

CLL

scRNA-seq

The difference in the number of exhausted CD8 + T cells was significantly larger between the healthy donors (HD) and MBL than between MBL and CLL. Early intervention of ibrutinib can largely reverse the immune dysfunction.

Demonstrated the need for early intervention of CLL by immunotherapy.

[86]

AML

scRNA-seq

Senescent-like CD8 + T-cells were impaired in dealing with AML blasts. Defined a new set of immune effectors

dysfunction (IED) signatures that are associated with the adverse outcome and immunocheckpoint unresponsive TME.

Revealed that senescent-like T cells may also be an underlying treatment target. IED scores helped the AML-risk stratification and facilitated the identification of personal treatment targets.

[87]

B-ALL, AML

scDNA-seq

T-cells acquired the exhaustion/dysfunction signature by chronic immune activation in pediatric leukemia TME, manifesting as the attrition of naïve T cells and TCF1 + stem-like memory T cells, and the terminal differentiation of effector T cells. NK cells also expressed a signature of exhaustion, especially in AML.

Noted that although pediatric leukemia has a shorter natural history of tumor exposure, immune cell exhaustion/dysfunction is still a common event and is negatively correlated to the clinical outcomes.

[90]

CLL

scRNA-seq; scATAC-seq

PD-1int subset that was still functional and PD-1hi subset that was exhausted was identified in CLL TME. IL-10 signaling moderates the PD-1 expression through IL-10R-STAT3 pathway and sustains anti-tumor immunity by preventing excessive exhaustion.

Proposed that combining IL-10 with checkpoint blockade therapy may improve the clinical outcome in CLL patients.

[91]

CLL

scRNA-seq

BCL-2 expression was significantly increased in the T cells of CLL patients and associated with increased regulatory T-cells, exhausted cytotoxic T lymphocytes (CTL) signature, and increased T-cell adhesion.

Showed that BCL2 expression in T-cells is associated with immunosuppressive TME. BCL2 may be an underlying therapeutic target.

[92]

CLL

scRNA-seq

CLL progression mainly occurs in the lymphatic nodes (LN) and is associated with suppressive T-cell states. A small population of activated CLL cells progressed in the lymph nodes. Poor outcome was associated with activated CD4 + memory T cells and M2 macrophages in LN. T-cell inflamed microenvironment was progression inhibitive for the tumor.

Attributed the shorter time-to-first-treatment in CLL patients to increased proportion of activated CLL cells. These cells are potentially more effective in recruiting a tumor-supportive TME, thereby accelerating disease progression.

[95]