The Substrate Problem in Cancer Immunotherapy
Immune checkpoint inhibitors (ICIs), such as anti-PD-1 and anti-PD-L1 antibodies, have fundamentally changed oncology by removing the molecular "brakes" that tumors use to suppress adaptive immunity. However, these therapies suffer from a glaring vulnerability: they require a pre-existing population of active, tumor-infiltrating lymphocytes (TILs) to serve as their substrate. If cytotoxic T cells are not already inside the tumor mass, blocking inhibitory checkpoints achieves nothing.
As reviewed by Ming Yi et al. in Exploiting innate immunity for cancer immunotherapy, this limitation leaves the majority of patients with solid tumors non-responsive. This therapeutic failure defines "immunologically cold" tumors, which generally fall into two distinct spatial phenotypes described by Mengke Niu et al. in The enhanced antitumor activity of bispecific antibody targeting PD-1/PD-L1 signaling:
- Immune Deserts: Tumors that completely lack baseline leukocyte presence or recognition.
- Immune-Excluded Tumors: Tumors where the host successfully generates a T-cell response, but the cells are physically or chemically blocked from penetrating the malignant core.
To design therapies that can convert these non-responsive microenvironments into "hot," immuno-responsive zones, researchers must address the specific structural, metabolic, and immunological barriers that drive exclusion.
The Physical Fortress: Stroma and Vascular Chaos
The most direct reason T cells fail to infiltrate a tumor is structural. Tumors actively construct physical walls that isolate themselves from the host immune system.
A primary driver of this immune-excluded phenotype is the extracellular matrix. As detailed by Xinrun Wang et al. in Bispecific antibody targeting TGF-β and PD-L1 for synergistic cancer immunotherapy, a dense stroma of collagen fibers creates a rigid physical shield around the tumor core. This architecture prevents direct, cell-to-cell contact between cytotoxic T cells and malignant cells.
This structural sequestration is highly organized. By combining spatial omics with graph neural networks in 1415 Multimodal spatial signatures of immunotherapy response in hepatocellular carcinoma, Alexandro E. Trevino et al. identified that cancer-associated fibroblasts within immune-rich regions form a distinct "interface niche." This niche acts as a physical wall that traps and isolates immune cell aggregates, preventing them from crossing into the tumor epithelium (Alexandro E. Trevino et al., 478 Spatial omics and deep learning define an immunotherapy resistance niche in hepatocellular carcinoma).
[ Circulating T Cells ]
│
▼ (Chaotic Vasculature Hinders Extravasation)
[ Disorganized Vessels ]
│
▼ (Trapped in Fibroblast/Collagen Wall)
[ Interface Niche / Stroma ] ──X──> [ Tumor Core (Immune Desert) ]
Even if T cells attempt to navigate toward the tumor via the bloodstream, the entry points are often compromised. The authors of Association between radiomics features of DCE-MRI and CD8+ and CD4+ TILs in advanced gastric cancer demonstrated that high levels of asymmetry and disorganization in tumor vascular maps strongly correlate with an absence of cytotoxic T cells. Because these chaotic microvessels lack proper endothelial activation, they fail to facilitate the rolling, adhesion, and extravasation required for leukocytes to exit circulation and enter the tissue.
The Biochemical Shield: Hypoxia and Innate Silence
Beyond physical barriers, cold tumors deploy metabolic and chemical defenses that neutralize immune cells or render the tumor invisible.
The supportive stromal tissue itself drives exclusion through metabolic deprivation. In a study of clear cell ovarian cancer, Aijun Yang et al. found that severe hypoxia localized within the surrounding stroma acts as a functional barrier (CD8 + T cell infiltration is associated with improved survival and negatively correlates with hypoxia in clear cell ovarian cancer). This oxygen-deprived zone suppresses T-cell viability and migration, disabling the cells before they can ever reach the tumor epithelium.
Simultaneously, many tumors maintain an immune-desert state through immunological ignorance. In pediatric and low-progression malignancies, a low tumor mutational burden (TMB) means the cancer expresses very few distinct neoantigens (Christian Hörner et al., CD163+ immune cell infiltrates and presence of CD54+ microvessels are prognostic markers for patients with embryonal rhabdomyosarcoma). Without recognizable targets, the adaptive immune system remains unaware of the malignancy (Nai‐Kong V. Cheung, Immunotherapy and Radioimmunotherapy for Desmoplastic Small Round Cell Tumor).
[ Low Mutational Burden ] ──> No Neoantigens ──> No T-Cell Recognition
[ ENPP1 Activation ] ──> Degrades cGAMP ──> STING Pathway Silenced ──> No Chemokines (CXCL11/16)
Even when abnormal antigens are present, the innate immune system—the primary alarm system required to recruit adaptive cells—is frequently silenced. For instance, Songnan Wang et al. discovered that the enzyme ENPP1 acts as an innate immune checkpoint by rapidly degrading extracellular cGAMP (ENPP1 is an innate immune checkpoint of the anticancer cGAMP–STING pathway in breast cancer). By destroying cGAMP, the tumor prevents the activation of the STING pathway. Without this innate trigger, the microenvironment cannot produce the critical pro-inflammatory cytokines and chemokines (such as CXCL11 or CXCL16) needed to recruit specialized type-I conventional dendritic cells and cytotoxic T cells (X. Li et al., XCL1-secreting CEA CAR-T cells enhance endogenous CD8+ T cell responses to tumor neoantigens to confer a long-term antitumor immunity).
Strategies to Melt the Cold Front
To overcome these multi-layered defenses, modern therapeutic strategies focus on three distinct mechanisms: dismantling the physical stroma, forcing innate recruitment, and engineering localized immune hubs.
Dismantling the Stroma
To breach the physical matrix, researchers are engineering multi-targeting bispecific molecules. The strategy evaluated by Xinrun Wang et al. combines a PD-L1 inhibitor with a TGF-β neutralizing entity (Bispecific antibody targeting TGF-β and PD-L1 for synergistic cancer immunotherapy). Because TGF-β is the primary cytokine driving fibroblast activation and collagen deposition, neutralizing it physically dismantles the stromal barrier, opening up pathways for T cells to infiltrate the tumor core while simultaneously blocking adaptive checkpoints.
Sounding the Innate Alarm
Other approaches bypass the tumor's cloaking mechanisms by forcing an artificial inflammatory response. Weijiang Zhang et al. demonstrated that small-molecule STING agonists can inflame the cervical cancer microenvironment, providing the missing innate trigger needed to recruit T cells and overcome anti-PD-1 resistance (STING agonist inflames the cervical cancer immune microenvironment and overcomes anti-PD-1 therapy resistance).
Taking this further, Ming Yi et al. showed that an antibody conjugate that simultaneously targets STING, TGF-β, and PD-L1 successfully activates the CXCL16–CXCR6 signaling axis, actively pulling highly cytotoxic T-cell populations across the microenvironmental border (Triple targeting of STING, TGF-β, and PD-L1 boosts CXCL16–CXCR6 signaling for potent antitumor response). Similarly, intralesional oncolytic viruses like MEM-288 turn the tumor into its own "vaccine factory" by delivering type 1 interferon and CD40 signaling molecules directly into the core, forcing local vascular and immune activation (Andreas Saltos et al., 1132 Impact of intralesional oncolytic viral therapy targeting in situ activation of CD40 and type 1 interferon signaling pathways on the TME and systemic T cell immunity in murine models and cancer patients).
Bypassing Ignorance via In-Situ Engineering
When a tumor completely lacks recognizable neoantigens, researchers can engineer an immune response from scratch. Juanita Lopez demonstrated that individualized mRNA vaccines targeting patient-specific neoantigens can successfully prime de novo T-cell populations, forcing them to traffic into previously cold solid tumors (Autogene cevumeran with or without atezolizumab in advanced solid tumors: a phase 1 trial).
Finally, to sustain these cells without relying on distant, suppressed lymph nodes, Liang Zhang et al. developed light-activated frameworks capable of inducing the formation of Tertiary Lymphoid Structures (TLS) directly within the tumor microenvironment (Trigger inducible tertiary lymphoid structure formation using covalent organic frameworks for cancer immunotherapy). By building synthetic immune niches inside the tumor's perimeter, these therapies ensure that once the cold front is melted, the immune system maintains a localized, self-sustaining factory to prevent tumor recurrence.