The immune system can recognize and attack cancer cells, but tumors often develop ways to evade or suppress these attacks. The collection of immune cells, signaling molecules, and structural components surrounding a tumor is called the tumor immune microenvironment (TIME).
Understanding the TIME in detail is essential for developing better immunotherapies -- treatments that harness the immune system against cancer. Current immunotherapies like immune checkpoint inhibitors (which release the brakes on immune cells) work well for some patients but fail for many others, largely because the TIME is far more complex than a simple on/off switch.
Traditional methods for studying immune cells, such as flow cytometry and bulk RNA sequencing, measure the average activity of thousands or millions of cells at once. This averaging masks the important differences between individual cells -- including rare but functionally critical populations that may drive resistance to treatment.
Single-cell RNA sequencing (scRNA-seq) is a transformative technology that measures gene expression in each individual cell separately, revealing the full diversity of immune cell types, their functional states, and how they interact with tumor cells across different cancer stages.
Tumor cells actively reshape the surrounding immune environment to suppress anti-cancer immunity. One major mechanism involves recruiting regulatory T cells (Tregs) -- immune cells whose normal job is to prevent overactive immune responses. Within tumors, Tregs secrete suppressive signals including IL-10 and IL-35 that disable cytotoxic immune cells, and can even directly kill natural killer (NK) cells and cytotoxic T cells.
Tumor-associated macrophages (TAMs) represent another major immunosuppressive cell type. When influenced by tumor-derived signals, macrophages shift from an anti-cancer (M1) state to a pro-cancer (M2) state that supports tumor blood vessel growth, suppresses immune attacks, and remodels the tissue architecture to favor cancer spread.
Myeloid-derived suppressor cells (MDSCs) are immature immune cells that accumulate in tumors and actively suppress T cell activity through multiple mechanisms including secreting arginase and IL-10. They also contribute directly to tumor expansion by promoting new blood vessel formation and remodeling the extracellular matrix.
Tumor cells also hide from the immune system by reducing the expression of molecules that immune cells use to recognize cancer -- including MHC class I (HLA) antigen presentation molecules. This makes tumor cells effectively invisible to cytotoxic T cells. They also activate specific signaling pathways (JAK/STAT, PI3K/AKT) that prevent cancer cells from dying and drive chronic inflammation that further suppresses immunity.
CD8+ cytotoxic T cells are the immune system's primary cancer killers, but within tumors they can become progressively disabled in a state called T cell exhaustion (TEx). Single-cell sequencing has revealed that exhausted T cells are not a single uniform population but include multiple sub-states -- progenitor exhausted cells (TPE) that can still be reinvigorated, and terminally exhausted cells (TTE) that cannot.
In tumors, exhausted T cells over-express multiple inhibitory immune checkpoints simultaneously -- including PD-1, CTLA-4, and TIGIT. This co-expression of multiple inhibitory signals may explain why blocking only one checkpoint (like PD-1) is insufficient for many patients, suggesting that combination checkpoint therapies may be needed.
CD4+ helper T cells within tumors have a complex dual role. Immunosuppressive Treg subsets and follicular helper T cells (Tfh) are enriched in tumor tissues compared to adjacent normal tissue, while anti-tumor helper T cell subsets are reduced. Single-cell studies have shown that Treg expansion in tumors is associated with worse patient survival across multiple cancer types.
Single-cell analysis has also identified positive prognostic T cell subpopulations -- particularly GZMK+ effector memory T cells and proliferating MKI67+ T cells -- which when present in greater numbers correlate with better treatment outcomes in kidney cancer, liver cancer, and other tumor types.
The traditional classification of tumor-associated macrophages into two simple types -- cancer-killing M1 and cancer-promoting M2 -- has been overturned by single-cell studies. scRNA-seq has revealed that TAMs frequently co-express both M1 and M2 genes simultaneously, and has identified entirely new functionally distinct macrophage subpopulations.
Three newly characterized macrophage subtypes have emerged as particularly important: C1Q+ TAMs with enhanced ability to engulf dead cells; SPP1+ TAMs that promote tumor blood vessel growth and suppress immune function; and FCN1+ TAMs derived from circulating monocytes that are more prevalent in tissue adjacent to tumors. In prostate cancer specifically, TAMs expressing SLC40A1, PLAC8, and FCN1 were associated with worse patient outcomes.
Tumor-associated neutrophils (TANs), long overlooked in cancer immunology, show considerable heterogeneity revealed by single-cell analysis. Beyond the simple anti-tumor (TAN-1) vs. pro-tumor (TAN-2) division, at least four functional TAN subpopulations have now been identified in pancreatic cancer, with distinct metabolic programs and recruitment pathways.
Dendritic cells (DCs) -- which are responsible for alerting T cells to cancer -- also show important heterogeneity. A newly identified LAMP3+ DC subpopulation can migrate from tumors to lymph nodes to trigger immune responses, but in some contexts can also recruit immunosuppressive Tregs. This dual role highlights the complexity of dendritic cell biology in cancer.
The tumor immune microenvironment does not remain static -- it evolves dramatically as cancer progresses from early stages to late and metastatic disease. Single-cell studies across multiple cancer types consistently show that immune suppression intensifies with disease progression: immunosuppressive cells increase in proportion while anti-tumor immune populations decrease.
In breast cancer metastases, the total number of immune cells decreases but the proportion of immunosuppressive cells increases. Single-cell analysis revealed that fatty acid metabolism shifts immune cells toward a pro-tumor phenotype in metastatic lesions, suggesting that targeting specific metabolic pathways could delay cancer spread.
For brain metastases -- which occur in many cancer types including melanoma and breast cancer -- single-cell studies found that the immune environment is particularly immunosuppressive, with disease-specific differences. Melanoma brain metastases show abundant T cells (which may explain why immunotherapy has some effect in melanoma brain disease), while breast cancer brain metastases are dominated by neutrophils with fewer T cells.
In colorectal cancer liver metastasis, SPP1+ macrophages are specifically enriched, and immune cell metabolism shifts in ways that promote tumor survival. These organ-specific immune landscape differences explain why the same cancer type can respond very differently to immunotherapy when it metastasizes to different organs.
Beyond the well-known PD-1/PD-L1 and CTLA-4 checkpoint pathways, single-cell studies have revealed additional immune checkpoints that are co-expressed on exhausted T cells and may represent important new targets. These include LAG-3, TIM-3, and TIGIT -- all of which have entered clinical trials.
The LAG-3 antibody relatlimab combined with the PD-1 antibody nivolumab (the RELATIVITY-047 trial) showed significantly improved progression-free survival compared to nivolumab alone in untreated metastatic melanoma, demonstrating that dual checkpoint blockade targeting these new pathways can translate into real clinical benefit.
TIGIT blockade has shown promise in cervical cancer and osteosarcoma models, and is being evaluated in combination with PD-1/PD-L1 inhibitors in Phase I and II trials. For TIM-3, the anti-TIM-3 antibody TSR-022 combined with a PD-1 antibody showed dose-dependent activity in lung cancer patients who had failed PD-1 therapy alone -- suggesting TIM-3 inhibition may help overcome resistance to existing immunotherapies.
Single-cell analysis of cytokines and chemokines has also identified molecular signals that recruit immunosuppressive cells into tumors. The CCL2/CCR2 axis that recruits immunosuppressive macrophages and the CXCL5/CXCR2 axis for MDSC recruitment are both potential targets -- though a clinical trial of the CCL2 blocking antibody carlumab in prostate cancer showed it was well-tolerated but lacked significant single-agent activity, highlighting the need for combination strategies.
Single-cell RNA sequencing has transformed cancer immunology from a field that studied average behaviors of bulk cell populations to one that can map the full diversity of individual immune cells within a tumor at a given moment. This resolution has revealed the extraordinary complexity of the TIME and identified dozens of new potential therapeutic targets.
The field is moving toward single-cell atlases -- comprehensive maps of immune cell types and states across many cancer types, stages, and patients. These public databases will enable researchers to compare findings across studies and identify universal vs. cancer-specific immune features, accelerating the development of broadly applicable immunotherapy strategies.
A key insight is that effective immunotherapy will likely require multi-target combination approaches rather than single-agent treatments. The simultaneous presence of multiple suppressive mechanisms (Tregs, TAMs, MDSCs, multiple checkpoint proteins) means that blocking any single pathway is insufficient for durable cancer control in many patients.
Despite its power, scRNA-seq has important limitations: it provides a snapshot of one moment in time, the cell isolation process can alter cell states, and sample sizes remain limited compared to clinical trials. Future integration of scRNA-seq with spatial transcriptomics (which preserves tissue location information) and longitudinal sampling will be needed to fully understand the dynamic evolution of the TIME during cancer progression and treatment.