Spatial Omics Reframes Cancer as a Tissue-Scale Disease
Cancer biology has moved far beyond the idea of a uniform tumor mass. Malignant cells, immune infiltrates, stromal populations, blood vessels, extracellular matrix and metabolites all occupy distinct tissue neighborhoods, and those neighborhoods shape how disease behaves at the bedside. In practice, the question is no longer only which genes are active, but where they are active and which neighboring cells are close enough to matter.
Bulk RNA sequencing captured broad molecular programs, yet it blended together signals from different compartments and erased the tissue architecture that governs invasion, immune escape and drug failure. Single-cell RNA sequencing solved part of that problem by separating cellular states, but it still required tissue dissociation, which strips away the spatial relationships that define tumor ecology. Spatial omics fills that gap by anchoring molecular readouts to histology, coordinates and cellular neighborhoods.
Why Location Changes the Biology
A tumor is not just a collection of cells; it is an organized ecosystem. Spatial organization influences whether malignant cells can migrate, whether immune cells can enter productive contact with their targets, and whether stromal barriers wall off therapy-sensitive regions. Even when two lesions look similar under routine pathology, the cellular geography inside them may be radically different, and those differences can help explain divergence in progression and treatment response.
The appeal of spatial omics lies in that simple but consequential principle. RNA, protein and metabolite signals gain biological meaning when they are interpreted in place. A transcript that appears modest in bulk analysis may be highly concentrated in an invasive edge, a hypoxic niche or a perivascular zone, where its functional significance is amplified by proximity to other cell types.
Mapping Tumor Microenvironments Across Multiple Layers
The review brings together several spatial technologies rather than treating one assay as sufficient on its own. Spatial transcriptomics, spatial proteomics, multiplexed imaging and spatial metabolomics each answer different questions. Transcriptomic mapping identifies local gene-expression programs. Proteomic platforms can capture signaling states closer to functional phenotype. Multiplexed imaging preserves morphology while showing how cells are arranged. Spatial metabolomics adds a layer that is especially relevant in cancer, because nutrients, intermediates and drug exposure are all unevenly distributed within tissue.
This multimodal approach is particularly valuable for dissecting the tumor microenvironment. Immune cells do not act in isolation; their effect depends on whether they are excluded, clustered, exhausted, activated or organized into specialized structures. Fibroblasts and endothelial cells can either support immune entry or reinforce physical and biochemical barriers. The same tumor can contain both permissive and suppressive niches, and spatial methods are built to identify those contrasts rather than average them away.
Immune Niches, Tertiary Lymphoid Structures and Clonal Evolution
Spatial analysis is especially useful for studying tertiary lymphoid structures and other immune-rich compartments. These organized aggregates may represent local immune activity within tumors, but their significance depends on context: which cells are present, how they are arranged, and which molecular programs dominate at the interface between immune and malignant tissue. Spatial tools can also track how clonal populations evolve across the tumor core, invasive front and metastatic routes, offering a more anatomically grounded view of selection pressure.
One sentence captures the practical consequence: tumor evolution is not evenly distributed across space.
By preserving location, spatial approaches can reveal how different clones occupy distinct ecological niches and how those niches may protect resistant populations. In a tissue section, an apparently homogeneous lesion may actually be a mosaic of competing microenvironments with different proliferative, inflammatory and metabolic rules.
Metabolic Reprogramming Is Spatial, Not Uniform
The metabolic emphasis in this review is clinically important. Cancer metabolism is often discussed as a tumor-wide feature, but metabolic dependencies are frequently spatially restricted. Oxygen gradients, nutrient depletion, acidification and competition for amino acids all vary by tissue location. Spatial metabolomics makes those gradients visible and allows investigators to connect them to cell state, stromal composition and immune suppression.
Particular attention is placed on lipid metabolism, amino acid competition and ferroptosis-associated states. Each of these processes can be intensified or blunted by local tissue conditions. A region with restricted nutrient access may force malignant cells into alternate fuel usage, while nearby immune cells may be starved of the substrates they need for effector function. When therapy arrives, those same gradients can influence whether a drug reaches its target at an effective concentration.
Therapy Resistance Emerges From Tissue Geography
Resistance is rarely just a molecular mutation problem. Spatial omics frames resistance as a relationship between the tumor and its environment. Drugs do not distribute evenly through every compartment, and their effects are shaped by vascular access, stromal density, extracellular matrix architecture and local metabolic state. Spatial pharmacokinetics and spatial pharmacodynamics, as discussed in the review, may therefore become critical tools for understanding why one tumor region responds while another persists.
This has direct implications for combination therapy. If a resistant niche is protected by metabolism, immune exclusion or poor drug penetration, then treatment planning may need to account for the physical placement of that niche as much as its gene-expression profile. The same lesion may require different strategies in different regions, not because the disease changed in time, but because it was already heterogeneous in space.
From Descriptive Atlases to Mechanism-Driven Oncology
The field is moving from tissue mapping toward clinically interpretable spatial oncology. That shift depends on more than better imaging resolution. It requires robust multimodal integration, careful tissue workflows and functional validation that connects spatial patterns to biological mechanism. Without those safeguards, spatial datasets risk becoming elaborate atlases without therapeutic consequence.
Pathology will be central to that transition. So will bioinformatics, pharmacology and engineering. The review makes a strong case that spatial data should be interpreted alongside histology rather than as a separate technical exercise. In that framework, the microscope becomes a molecular coordinate system, not just a diagnostic tool.
A second practical reality follows from that logic: spatial biomarkers will only matter if they can influence patient stratification and treatment design.
What the Field Still Needs Before Routine Translation
Several hurdles remain before spatial omics can move from research platforms into routine cancer care. Standardized tissue handling is one of them, because spatial readouts are highly sensitive to how specimens are collected, processed and aligned with downstream assays. Multimodal integration is another, since RNA, protein, imaging and metabolite data must be merged without losing biological interpretability. Functional validation remains essential as well, because a spatial association alone does not prove causation.
The review closes with a translational agenda rather than a claim of immediate clinical readiness. Its central message is clear: cancer behaves as a spatially organized disease, and the methods used to study it must preserve that organization. For oncology teams, the next challenge is not whether spatial heterogeneity exists, but how to convert spatial information into decisions that change treatment selection, combination design and resistance monitoring.