Spatial Omics Solutions for Pathology Research
Histology can show where a lesion begins, how tissue architecture is disrupted, and which regions appear morphologically distinct—but morphology alone may not explain the molecular programs operating within those regions. Spatial omics connects pathology annotations with RNA and protein states to characterize lesions, invasive fronts, stromal boundaries, inflammatory regions, and apparently normal adjacent tissue without losing their spatial context. CD Genomics supports pathology-guided studies using FFPE or fresh-frozen specimens, from whole-section discovery and region-focused profiling to targeted in situ validation. We help align tissue selection, pathology review, spatial resolution, assay choice, and downstream analysis so that archived cohorts and newly collected samples can produce interpretable molecular evidence while preserving scarce material for follow-up research.
Morphology Linked to Molecular States
Lesion, Boundary, and Compartment Analysis
FFPE and Fresh-Frozen Study Options
Why Morphology and Non-Spatial Molecular Data Leave Different Evidence Gaps
Pathology images preserve architecture, while bulk and dissociated-cell assays provide molecular depth. The missing step is a coordinate system that lets a research team test whether a molecular program occupies the lesion, compartment, interface, or microenvironment visible on the slide.
1
Similar morphology can contain different molecular states
Two regions may look alike on H&E yet differ in gene programs, cell composition, or protein phenotype. Morphology can define where to look, but it does not determine the molecular state by itself.
2
Molecular averages erase lesion and boundary context
Bulk extraction blends regions, while dissociation removes the original coordinates. A signal may be real but impossible to attribute to a lesion core, adjacent stroma, infiltrating front, vessel-associated niche, or unaffected compartment.
3
FFPE and image artifacts can be mistaken for biology
Fixation history, decalcification, necrosis, folds, tears, low cellularity, and registration error can alter signal or interpretation. These regions need explicit QC and annotation rather than being treated as ordinary biological variation.
Decision rule: If the question is only whether an analyte changes across a specimen, a bulk assay may be sufficient. Spatial profiling becomes decision-relevant when the answer depends on where the change occurs, which tissue compartment carries it, or whether a molecular boundary agrees with the pathology image.
Study Decisions for Pathology-Informed Spatial Profiling
A pathology research project should start by deciding whether the unknown is distributed across the whole section, restricted to pre-defined regions, or ready for targeted validation across a larger cohort. That choice determines the assay, annotation plan, section allocation, and analysis unit.
Survey
Whole-Section Molecular Mapping
Use broad spatial profiling when the important compartments, gradients, or molecular domains are not known in advance.
- Question: Which molecular programs organize the tissue?
- Image role: Register molecular domains to H&E or IF and review unexpected boundaries.
- Benefit: Discover heterogeneity without restricting the first analysis to a pre-selected ROI.
Regions
Pathology-Guided Region Profiling
Use pre-specified ROIs or compartments when a pathologist can identify the lesions, interfaces, or structures that must be compared.
- Question: How do defined regions differ in RNA or protein state?
- Image role: Lock inclusion, exclusion, and segmentation rules before molecular comparisons.
- Benefit: Concentrate measurement on the regions that change the research decision.
Cohort
FFPE Cohort and TMA Validation
Use targeted RNA or protein panels when a shortlist is ready to be examined across archived blocks or a spatial-ready tissue microarray.
- Question: Does the spatial pattern recur across independent specimens?
- Image role: Standardize core selection, tissue class, and compartment annotation.
- Benefit: Move from a discovery slide to a replicate-aware cohort test.
How to Match Spatial Technologies to the Pathology Question
Coverage, spatial resolution, tissue area, sample preservation, and the role of the pathology image create real trade-offs. Use the smallest assay combination that can connect the region of interest to the molecular evidence needed for the study.
| Technology or Service | Analytical Role | Useful Output | When to Choose | Boundary |
|---|---|---|---|---|
| Whole-transcriptome spatial profiling Visium / Visium HD |
Survey broad gene programs and spatial domains across an intact section | Registered expression map, spatial domains, region contrasts, pathway localization | When heterogeneity is unknown or whole-section context must be preserved | Resolution and cell assignment vary; FFPE feasibility depends on tissue and analyte quality |
| Targeted in situ RNA Xenium / CosMx SMI |
Localize selected transcripts at cell or subcellular coordinates | Transcript map, cell segmentation, phenotype map, boundary or neighborhood analysis | When candidate genes are known and precise localization is more important than broad discovery | Panel-limited; segmentation and transcript assignment require QC |
| Morphology-guided regional profiling GeoMx DSP |
Measure RNA or protein in pathologist-defined ROIs, AOIs, or marker-defined compartments | ROI image set, region-by-feature matrix, compartment contrasts, spatial signatures | When lesions or tissue structures can be defined before measurement, especially in FFPE cohorts | Region-level measurements may combine multiple cells; ROI selection can introduce bias |
| Multiplex spatial protein imaging mIHC |
Test selected protein phenotypes and cell relationships in tissue | Multiplex image, cell phenotype map, density and proximity statistics | When RNA-defined findings need protein-level confirmation or cohort-scale targeted imaging | Panel and antibody specificity constrain the evidence; protein abundance does not prove function |
| Spatial-ready TMA construction | Standardize selected tissue regions across many FFPE blocks | Core map, block metadata, section-allocation record, spatial-assay-ready slides | When a validated marker set or spatial phenotype must be examined across a cohort | Small cores may miss whole-tissue heterogeneity; core number and placement need pathology review |
| Tissue microregion retrieval | Recover a defined region for follow-up molecular assays | Retrieved microregion, coordinates, area documentation, downstream assay material | When the question requires physical isolation of a pathology-defined compartment | Sampling is destructive and material may be limited; adjacent-section differences remain relevant |
Integration logic: A pathology image defines tissue structures and QC exclusions; spatial RNA or protein data test the molecular state within those coordinates; registration links the evidence layers; and expert review determines whether molecular domains support, refine, or challenge the initial annotation. Disagreement is a result to investigate, not an automatic assay failure.
If FFPE and fresh-frozen material are both available, review the fresh-frozen versus FFPE workflow guide before allocating blocks or sections.
Configurable Study Routes for Spatial Pathology Research
These routes represent different research decisions, not fixed commercial packages. Final sample numbers, platforms, sections, staining, annotation rules, and analyses are set after feasibility and pathology review.
Whole-Section Discovery Route
For tissues in which the relevant molecular domains or boundaries are not known in advance
Comparison Logic
Spatial domains within a section, disease vs reference tissue, or matched morphological classes across biological replicates
Evidence Layers
H&E or IF image + broad spatial RNA profiling; optional single-cell reference or targeted validation
Key Deliverables
Registered spatial map, molecular domains, region-level differential analysis, image–omics concordance review, and validation shortlist
Evidence Boundary
Diagnostic classification, causal proof, or universal disease markers
Pathology-Guided
Region and Boundary Route
For defined lesions, compartments, interfaces, or structures that require molecular comparison
Comparison Logic
Lesion vs adjacent tissue, core vs edge, epithelium vs stroma, inflamed vs non-inflamed, or other pre-specified regions
Evidence Layers
Expert annotation + ROI/AOI RNA or protein profiling, or targeted in situ RNA with cell segmentation
Key Deliverables
Annotation file, region-by-feature matrix, compartment contrasts, boundary gradients, QC exclusions, and follow-up priorities
Evidence Boundary
Post hoc region selection without a documented rule or direct proof of cell–cell interaction
FFPE Cohort Validation Route
For a defined signature or phenotype ready to be tested across archived specimens or a TMA
Comparison Logic
Pre-specified tissue classes, cohorts, experimental groups, or matched regions across independent blocks
Evidence Layers
Spatial-ready TMA or selected sections + targeted RNA and/or multiplex protein imaging
Key Deliverables
Cohort annotation matrix, spatial phenotype frequencies, replicate-aware contrasts, missingness/QC summary, and evidence status
Evidence Boundary
Clinical validation, prognosis, treatment selection, or individual health assessment
What to prepare for feasibility review: tissue type and preservation; species; block or slide availability; fixation, decalcification, and storage history; section area and thickness; H&E or IF images; biological replicate structure; intended tissue classes; existing molecular data; and the decision the spatial result should support.
Replication boundary: Spots, cells, ROIs, or cores taken from one specimen are not independent biological replicates. Statistical inference should use the true biological unit—such as donor, animal, independent model, or block—while accounting for paired and nested designs.
Outputs That Connect Pathology Annotations to Molecular Evidence
The output package is organized for multidisciplinary review. It shows where a result was measured, how it was aligned to the image, which regions were excluded, what was computationally inferred, and what question should move to validation.
Registered Tissue Map
High-resolution H&E or IF image aligned with spots, cells, ROIs, or AOIs, including transformation metadata and visible registration landmarks. This lets reviewers inspect the molecular result in the same tissue coordinate system.
Annotation and QC Layer
Pathology-defined regions, molecular domains, section boundaries, and technical exclusions are delivered as inspectable overlays and tables rather than hidden analysis assumptions.
Region and Boundary Comparisons
Region-by-feature matrices, differential expression, pathway activity, cell composition, or protein phenotype results compare the pre-specified compartments and transition zones.
Evidence and Follow-Up Summary
Each central finding is labeled as measured, computationally inferred, or unresolved, with an appropriate next step such as targeted RNA localization, protein staining, expanded cohort testing, or functional validation.
| Analysis | Concrete Deliverable | Typical Format |
|---|---|---|
| Image registration | Registered pathology image with molecular coordinates and transformation record | TIFF/PNG + coordinate table |
| Tissue annotation | Region polygons, tissue-class labels, exclusion masks, and reviewer notes | GeoJSON/CSV + overlay image |
| Spatial-domain analysis | Molecular domain map with marker genes or proteins and pathology concordance table | Interactive HTML + CSV + PNG |
| Region comparison | Region-by-feature matrix, differential results, pathway scores, and effect-size plots | CSV + PDF/PNG |
| Cell or neighborhood analysis | Cell-phenotype map, composition table, proximity matrix, and neighborhood graph | Interactive HTML + CSV + PNG |
| Cohort review | Specimen-level QC, missingness, spatial phenotype frequency, and replicate-aware comparison | CSV + PDF summary |
Morphology–Molecular Registration Dashboard
A coordinated view of the tissue image, expert annotations, molecular domains, alignment landmarks, and excluded regions for review before downstream interpretation.
Compartment and Boundary Evidence View
Pre-defined regions are compared with molecular gradients, cell composition, or protein phenotypes while keeping non-informative tissue separate from biological classes.
Research Interpretation Note: A molecular domain that aligns with a pathology region strengthens the spatial association but does not prove diagnosis or causality. A domain that does not align may reflect biological heterogeneity, image-registration limits, section-to-section differences, assay resolution, or tissue quality and should be reviewed across evidence layers.
Why Choose CD Genomics for Spatial Pathology Research
FFPE-Oriented Feasibility Review
Platform selection can account for fixation and storage history, tissue area, section quality, morphology, analyte condition, target abundance, and cohort scale. This helps protect scarce archival material before the final section-allocation plan is locked.
Image-to-Omics Registration
Pathology images, tissue annotations, and molecular coordinates can be organized in a shared analysis framework with QC landmarks and exclusions retained. Reviewers can trace a result back to its location rather than accepting a detached molecular label.
Multiple Spatial Evidence Routes
Whole-transcriptome discovery, targeted in situ RNA, morphology-guided RNA/protein profiling, multiplex protein imaging, tissue microarrays, and microregion retrieval can be matched to different pathology questions. A project can remain focused on the modality that changes the decision.
Research-Question-Led Bioinformatics
Analysis can be structured around tissue domains, annotated compartments, boundaries, cell composition, RNA–protein evidence, cohort recurrence, and technical exclusions. Measured results and computational inference remain distinguishable in the delivered outputs.
For morphology-centered tumor questions, review Spatial Omics Solutions for Tumor Morphology. For broader tumor ecosystems, see the Cancer and Oncology solution and the Tumor Microenvironment solution.
Published Research Examples
These independent studies demonstrate complementary pathology workflows: integrating FFPE morphology with whole-transcriptome and targeted in situ data, and using histopathologic interpretation to organize spatial molecular patterns in kidney tissue. The illustrations are original conceptual summaries, not reproductions of published figures.
Example 1: Connecting FFPE Morphology to Whole-Transcriptome and Cell-Level RNA Maps
Janesick A, Shelansky R, Gottscho AD, et al. Nature Communications. 2023;14:8353. doi:10.1038/s41467-023-43458-x
Research Question
Can complementary single-cell, whole-transcriptome spatial, and targeted in situ assays resolve molecular heterogeneity and thin tissue boundaries in FFPE breast-cancer specimens?
Study Design
The researchers analyzed FFPE breast-cancer blocks using Chromium Single Cell Gene Expression Flex, Visium CytAssist, Xenium In Situ, and H&E or immunofluorescence on adjacent or matched sections. Image registration connected pathology annotations with spatial expression data.
Key Findings
- Whole-transcriptome spatial profiling located molecularly distinct in situ and invasive tumor domains in the tissue.
- Targeted in situ mapping refined cell assignments and revealed thin myoepithelial boundaries and rare boundary-cell populations.
- H&E registration enabled direct comparison between molecular regions and pathologist annotations.
Relevance to This Solution
The study shows why broad discovery and targeted localization are complementary in pathology research. A whole-section map can nominate domains, while a targeted in situ assay can resolve selected genes and cells at a finer spatial scale.
Boundary
The work examined selected breast-cancer blocks, and some assays used serial sections. Section-to-section variation, panel selection, segmentation, and registration should be considered before generalizing the molecular patterns.
Example 2: Histopathologic Interpretation Organizes Human Kidney Spatial Transcriptomics
Isnard P, Li D, Xuanyuan Q, et al. American Journal of Pathology. 2025;195(1):69–88. doi:10.1016/j.ajpath.2024.06.011
Research Question
Can histopathologic features provide an interpretable framework for connecting human kidney morphology with spatial transcriptomic patterns?
Study Design
The researchers developed a morphology-based approach for interpreting human kidney spatial transcriptomics, integrating histopathologic review with molecular clustering and projecting pathology-informed patterns across additional tissue sections.
Key Findings
- Histopathologic features helped organize molecularly distinct kidney tissue regions.
- Pathology-informed groupings supported interpretation of spatial transcriptomic clusters across sections.
- The workflow demonstrated how morphology can guide molecular review without treating image appearance as a substitute for transcriptomic evidence.
Relevance to This Solution
The study illustrates how pathology review can provide a biological coordinate system for spatial transcriptomics in a non-oncology tissue. This helps researchers compare molecular domains with recognizable tissue structures and prioritize discordant regions for closer review.
Boundary
Pathology-informed clusters remain research interpretations and depend on tissue quality, annotation consistency, spatial resolution, and the studied kidney specimens. They do not establish diagnosis or clinical classification.
Frequently Asked Questions
Yes, several whole-transcriptome spatial, targeted in situ RNA, morphology-guided RNA/protein, and multiplex protein workflows support FFPE tissue. Feasibility still depends on fixation and storage history, decalcification, tissue area, section integrity, morphology, analyte condition, target abundance, and platform-specific QC. Review the block and available slides before committing scarce material.
Use whole-section discovery when the important domains or boundaries are not known. Use pathology-guided ROI profiling when lesions or compartments can be defined before measurement and the decision depends on comparing them. A hybrid design can survey representative sections first and then validate selected regions or markers in additional specimens.
Images can support tissue selection, region annotation, registration, cell or compartment segmentation, QC exclusion, spatial-domain review, and interpretation. The image is not merely decorative: its coordinate system should remain linked to the molecular data, and any transformation or serial-section alignment should be documented.
Mismatch can be biologically informative, but technical explanations must be reviewed first. Registration error, section-to-section differences, tissue damage, assay resolution, segmentation, low signal, or annotation uncertainty may contribute. The appropriate response is to inspect the image, QC layer, molecular markers, and replicate evidence together rather than forcing agreement.
No. These services are designed for research. A spatial association can support mechanistic hypotheses and research biomarker evaluation, but it does not by itself establish clinical performance, causality, prognosis, or an individual treatment decision. Those claims require separately designed and validated studies.
Define the biological tissue classes, inclusion rules, exclusion rules, and comparison unit before examining the final molecular contrasts whenever possible. Necrosis, folds, tears, low cellularity, detachment, and assay-QC failure should be labeled separately from biological reference regions. Consensus or blinded annotation may be useful when region classification is subjective.
There is no universal number. The design depends on the biological unit, expected heterogeneity, paired or nested structure, tissue availability, effect size, platform capacity, and intended statistical comparison. Many spots, cells, ROIs, or TMA cores from one specimen do not replace independent biological replicates.
Provide tissue type, species, preservation, fixation and decalcification history, storage information, block or slide inventory, representative H&E or IF images, section area and thickness, biological replicate structure, tissue classes of interest, existing molecular data, and the research decision the result should support. The sample submission guidelines can help prepare this information.
Related Services, Solutions, and Planning Resources
Continue from your pathology question to an appropriate FFPE, whole-section, region-based, targeted, or cohort workflow.
References
Janesick A, Shelansky R, Gottscho AD, et al. High resolution mapping of the tumor microenvironment using integrated single-cell, spatial and in situ analysis. Nature Communications. 2023;14:8353. https://doi.org/10.1038/s41467-023-43458-x
Isnard P, Li D, Xuanyuan Q, et al. Histopathologic Analysis of Human Kidney Spatial Transcriptomics Data: Toward Precision Pathology. American Journal of Pathology. 2025;195(1):69–88. https://doi.org/10.1016/j.ajpath.2024.06.011
For research purposes only. Not intended for clinical diagnosis, treatment, or individual health assessment.
Online Inquiry