Single-cell and multi-regional biopsy analyses. A, Schematic of scRNA-seq cohort derived from n = 6 CRC primary tumors. Boxplots showing ssGSEA scores for the Hallmark Epithelial Mesenchymal Transition gene set across the various cell types (B) and specifically between epithelial and stromal cells (C; from all six colorectal cancer tumors) in the scRNA-seq dataset (P < 2.2 × 10–16; Wilcoxon test). D, Comparison of ssGSEA scores for the Hallmark Epithelial Mesenchymal Transition gene set between epithelial and stromal cells in each primary CRC (n = 6) in the scRNA-seq dataset (all P < 2.2×10–16; Wilcoxon test). Epithelial cells are shown in green and stromal cells in pink. E, Schematic overview of the BOSS Biopsy cohort consisting of colon cancer resections from 7 patients each with up to n = 5 multi-regional biopsy samples. Heatmaps of ssGSEA scores for the Hallmark gene sets (F) and TF activity scores for the BOSS Biopsy samples (G). Samples are grouped according to patient of origin and the ESTIMATE Stromal Score of each biopsy sample is indicated by the ESTIMATE StromalScore bar at the top of the heatmap. Only the gene sets/TFs significantly and concordantly enriched in stroma or epithelium in both the LCM discovery and LCM validation cohorts are shown (from Fig. 2; Padj < 0.02 – Hallmarks; P < 0.05 – TFs). Gene sets/TFs with names/symbols colored orange were significantly enriched in stroma in the LCM discovery and LCM validation cohorts and gene sets/transcription factors with names/symbols colored blue were significantly enriched in epithelium in the LCM discovery and LCM validation cohorts. H, Scatterplots showing correlation between the ESTIMATE StromalScore and ssGSEA scores for the Hallmark Epithelial Mesenchymal Transition (left; Spearman rho = 0.96, P = 1.7e-08), KRAS Signaling Up (middle; Spearman rho = 0.87, P = 7.9e-07) and MYC Targets V2 (right; Spearman rho = −0.63, P = 0.00037) gene sets. Samples are colored by patient of origin. CRC, colorectal cancer.

SpatialDecon derived immune cell counts aggregated by Region, KM grade and KRAS status with pairwise comparison between groups. Mann-Whitney test used to determine statistical significance between groups

The ConfoundR resource enables stromal influence estimation in cancer tissue. A, Schematic overview of the cohorts and analyses available within the ConfoundR app, accessible via https://confoundr.qub.ac.uk/. B, Expression Boxplots analysis module of ConfoundR enabling the expression of a single gene to be compared between stroma and epithelium samples in each of the ConfoundR datasets. C, Expression Heatmap analysis module of ConfoundR enabling the expression of multiple genes to be visually compared between stroma and epithelium samples in each of the ConfoundR datasets. D, GSEA analysis module of ConfoundR allowing GSEA of existing gene sets from established gene set collections or custom user defined gene sets to be performed comparing stroma with epithelium in each of the ConfoundR datasets.

Primary colorectal cancer (CRC) is manageable with surgical and non-surgical therapies. However, high affinity for distant spread to the liver (CRLM) and lack of subsequent therapeutic options has maintained CRC as one of the most lethal malignant diseases worldwide. Tumors consist of malignant cells as well as a complex micro-environment (TME) of stromal and immune components, all of which have been implicated in disease severity and treatment resistance. Spatial transcriptomics (ST) is novel technology which allows molecular profiling of tissue while preserving tissue architecture. By studying matched CRC and CRLM with modern ST tools which allow scrutiny of the malignant cells and TME we aim to learn more about mechanisms of metastasis and why CRLM are resistant to treatment. 41 patients (29 untreated) underwent synchronous resection of CRC and CRLM within our institution between April 2002 and June 2010 (64% 5 year mortality). Paired CRC and CRLM from 4 patients underwent single-cell ST assessment on the Nanostring CosMx Spatial Molecular Imager (23 fields of view (FOV)). Paired CRC and CRLM from 4 patients underwent regional ST assessment (PanCk+, PanCk-, aSMA+ comprtments) on the Nanostring GeoMx Digital Spatial Profiler (116 FOV). Paired CRC and CRLM from 2 patients underwent regional ST assessment on the Visium 10X Genomics platform (4 FOV). Results/Conclusions: GeoMx offers entire slide visualization with bespoke and convenient FOV selection with reliable segmentation of TME components with whole transcriptome available. Visium 10X offers high resolution ST assessment of large pieces of tissue but slide preparation is difficult without specific tools. CosMx offers single-cell in-situ allowing the highest resolution with neighborhood and cellular interaction analysis possible but at lower plex and increased complexity. All technologies demonstrated differences between CRLM and their matched CRC. CRLM were generally more collagenous and chemokine rich. Furthermore, transcriptomic and morphological differences exist between CRLM from different patients. Specifically, CosMx demonstrated that epithelial cells and fibroblasts in CRC express MZTA compared to COL9A2, NEAT1 and IGF2 in CRLM. Neutrophils in CRLM under-express CXCL8. Desmoplastic CRLM were chemokine high whereas replacement CRLM over-expressed IGF2, DUSP5 and COL9A2. GeoMx identified MET and IL17 as potential drug targets in CRLM. These data suggest that CRLM evolve once metastasis is established and the immune response of the host to the metastasis is pivotal. Future therapeutic goals should emphasize personalized medicine and molecular staging of CRLM with specific targeted therapies. Colin Wood, Luke McNickle, Andrew Cameron, Ritupam Sarma, Vaidehi Pandya, Joao Da Silva Filho, Colin steele, John Cole, Joanne Edwards, Paul Horgan, Campbell Roxburgh. Multi-scale multi-omic assessment of matched synchronous colorectal cancer liver metastases using multiple spatial transcriptomic tools [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4586.

Supplementary details regarding methods and materials

Topographic immune cell deconvolution primary colorectal cancer and CRLM. A, Images (from Fig. 4A) of primary colorectal cancer (bottom) and CRLM (top) with 48 ROIs superimposed. Abundance estimates as determined from transcriptome by SpatialDecon for 14 cell populations illustrated for each ROI with color coding detailing the annotated tumor region. Radius is proportional to the estimated cell counts within the ROI. The immune cell count per region was extracted and the square root of the ratio to the mean immune cell count per region (41.37) of all immune cells was calculated and is displayed. The square root of the ratio was calculated to minimize the skew caused by variance of highly expressed cell types. B, Box plots demonstrating the median and interquartile range for each cell type analyzed organized by cell type and topographic region and grouped by KM grade. All ROIs taken from primary colorectal cancer except TLR are grouped as primary. Dendritic cells were removed due to insignificant counts. The Mann–Whitney test was used to assess for statistical significance; *, P < 0.05.

Mutational characterization of primary colorectal cancer and CRLM. A, Venn diagram demonstrating primary colorectal cancer and CRLM that underwent genomic analysis. Co-Barplot illustrating most frequently mutated genes across 13 matched primary colorectal cancer and CRLM, including mutation type. Genes are ordered by mutational frequency. Sections were sequenced using GPOL (Glasgow Precision Oncology Laboratory) mutational panel. B, Oncoplot demonstrating concurrent mutations in the 13 matched lesions. Patients are ranked according to co-mutational burden on the y-axis and ranked according to KM grade on the x-axis. Blue, gene mutated in primary only; red, gene mutated in metastasis only; purple, mutated in primary and metastasis. The right-hand three columns denote the percentage of total patients with each mutation type. C, Correlation matrix demonstrating co-occurrence of mutations, with left of the blue demarcation line representing primary colorectal cancer, right of the blue demarcation line representing CRLM (pair-wise Fisher exact test; *, P < 0.05). Gene names are displayed along the x- and y-axes ordered by mutational frequency. Dark green boxes represent significant co-occurrence. D, Box plot illustrating mutational burden in primary colorectal cancer and CRLM according to KM grade using the Mann–Whitney test to assess for statistically significant difference between KM groups.

Gene Set Enrichment Analysis results obtained by interrogating ranked list of differentially expressed KM high vs KM low genes against the REACTOME curated gene set database using ClusterProfiler package

Clinicopathological, morphological, and treatment characteristics for synchronously resected primary colorectal cancer and paired CRLM.

IHC characterization of matched primary colorectal cancer and CRLM with integration of morphological and mutational features. A, Representative images of CD3 and CD66b immunohistochemical staining (Patient B). Whole section demonstrated at ×0.5 magnifications; ROIs corresponding to tumor center (TC) and invasive edge (IE) of primary and CRLM are shown at ×6 and ×10 (black box). Scale bar, 100 μm. B, Intrapatient comparison between primary and metastasis of CD3 and CD66b cell counts at tumor center and invasive edge. P values calculated using the Mann–Whitney test. C, Kaplan–Meier survival plots (log-rank test, P values displayed) demonstrating the prognostic impact of CD3 and CD66b cell density at CRLM IE identified by IHC for IE of CRLM. High and low values determined according to median expression. D, Comparison of CD3 and CD66b cell density at (i) primary IE, (ii) CRLM IE, and (iii) CD3 primary IE and metastatic TC. Spearman Rho analysis. E, Box plots illustrating the relationship between KM grade and CD3 and CD66b cell counts at the IE and TC of primary colorectal cancer and CRLM. The P value was calculated using the Mann–Whitney test. F, Box plot representing relationship between mutational features (APC, TP53, KRAS, Serrated) and CD3 and CD66b cell density at TC and IE of primary colorectal cancer and CRLM. The P value was calculated using the Mann–Whitney test. Lesions of serrated origin were defined as APCwild + KRAS/BRAFmutation.

Spatial transcriptomic confirms the confounding effects of the stroma. A, Whole slide image of colon cancer case selected for spatial transcriptomic analysis. The tissue was stained with PanCK and CD45 with PanCK+ regions (green) identifying epithelium and CD45+ regions (purple) identifying immune components. Small circles indicate the ROIs selected for spatial transcriptomic analysis; ROI 4: high epithelial content, ROI 11: mixed epithelial content, ROI 10 demonstrates a ROI with low epithelial content, ROI 6: no epithelial content. B, Scatterplot showing the correlation between ssGSEA scores for the full Hallmark Epithelial Mesenchymal Transition gene set (n = 200 genes) and the corresponding reduced GeoMx Epithelial Mesenchymal Transition gene set (n = 81 genes) in the FOCUS clinical trial cohort (Spearman rho = 0.95). Samples colored by CMS calls (CMS1: n = 62; CMS2: n = 155; CMS3: n = 29; CMS4: n = 66; UNK: n = 44). C, Heatmap of ssGSEA scores for the Hallmark gene sets for the PanCK+ (epithelium; n = 8) and PanCK− (stroma; n = 11) areas within the regions of interest. Only the Hallmark gene sets identified as significantly and concordantly enriched in stroma or epithelium in both the LCM discovery and LCM validation cohorts are shown (the GeoMx versions of these Hallmark gene sets were used). D, GSEA comparing PanCK− areas (stroma; n = 11) to PanCK+ areas (epithelium; n = 8) for the Hallmark Epithelial Mesenchymal Transition gene set (GeoMx version). CMS, consensus molecular subtypes; LCM, laser capture microdissection; NES, normalized enrichment scale; ROI, region of interest.

Spatially resolved transcriptomic analysis using Nanostring Cancer Transcriptome Atlas gene sets. A, Representative images of mIF staining of four matched primary colorectal cancer and CRLM. DAPI, blue; Pancytokeratin, green; CD45, pink; αSMA, yellow. Topographic regions are annotated and each box represents hand-selected area of tumor. Eight regions were taken from CRLM and four from primary colorectal cancer per patient. Patient A: KMhigh, KRAS-wt, good prognosis. Patient B: KMhigh, KRAS-mt, good prognosis. Patient C: KMlow, KRAS/TP53 co-mutation, poor prognosis. Patient D: KM-low, KRAS-wt, BRAF-mt high-mutational burden lesion. B, PCA plot of all ROIs selected. The patient from whom the lesion originated is represented by shape. A red border indicates region arises from CRLM and white border represents primary colorectal cancer. The topographical region within the lesion is illustrated by the innermost color of the shape. KM high metastatic edges, green circle; KMlow metastatic edges, red circle; dashed blue line, epithelial regions of primary colorectal cancer and CRLM. C, Heatmap demonstrating single sample GSEA for every ROI, ordered on the x-axis by patient and ROI. Key presented to aid patient identification. The y-axis represents annotated gene sets from Cancer Transcriptome Atlas ordered by and clustered within modules of Immune Response, Adaptive Immune, Innate Immune, Signaling Pathways. Cell Function, Metabolism. Each cell represents the NES scaled by pathway. D, Heatmap demonstrating GSEA providing interpatient comparison of selected areas between KMhigh and KMlow patients and intrapatient comparison between primary and metastatic sites and intralesional comparison between tumor center and immune edge. Subset of regions filtered before GSEA is demonstrated in subgroup. Subsequent groups compared in GSEA identified as groups A and B. Group A annotated at top of diagram and represented by green. Group B annotated at bottom of diagram and represented by orange. Cells of heatmap represent −log10(Padj) for comparison; cell is tinted green if pathway is upregulated in group A and orange if upregulated in group B.

Initial characterization of tumor epithelium and stromal datasets. A, Schematic of the segregation strategies in the discovery and validation cohorts, drawn using BioRender. B, Heatmap of MCP-counter scores for the LCM discovery cohort, according to epithelium and stromal regions. C, Heatmap of MCP-counter scores for the LCM validation cohort, according to epithelium and stromal regions. D, Heatmap of MCP-counter scores for the FACS validation cohort. E, CMS classifications (using CMSclassifier) for the matched epithelium and stroma samples in the LCM discovery cohort. F, CMS calls (using CMSclassifier) for the matched epithelium and stroma samples in the laser capture microdissected validation cohort. G, CMS calls (using CMSclassifier) for the four lineages in the FACS validation cohort. CMS, consensus molecular subtypes; FSC, forward light scatter; LCM, laser capture microdissection; MCP, microenvironment cell population; SSC, side light scatter.

Application of findings to bulk colorectal cancer tumor data. A, Schematic summary of the clinical validation dataset from the FOCUS clinical trial. B, Scatterplot showing correlation between desmoplastic stroma percentage (DS%) determined from H&E assessment and ESTIMATE Stromal Score determined by transcriptomic data in the FOCUS clinical trial samples (Spearman rho = 0.73, P < 2.2e-16), colored by CMS calls (CMS1: n = 62; CMS2: n = 155; CMS3: n = 29; CMS4: n = 66; UNK: n = 44). C, Heatmap of ssGSEA scores for the Hallmark gene sets (identified in Fig. 2 as significantly enriched in the stroma/epithelium in the LCM discovery and validation cohorts) for the FOCUS clinical trial samples. Samples ranked in order of DS% from lowest (left) to highest (right). Gene sets with names colored orange were significantly enriched in stroma in the LCM discovery and LCM validation cohorts and gene sets with names colored blue were significantly enriched in epithelium in the LCM discovery and LCM validation cohorts. D, Heatmap of activity scores for TFs (identified as significantly enriched in the stroma/epithelium in the LCM discovery and validation cohorts) for the FOCUS clinical trial samples. Samples are arranged in order of DS% from lowest (left) to highest (right). Gene sets with names colored orange were significantly enriched in stroma in the LCM discovery and LCM validation cohorts and gene sets with names colored blue were significantly enriched in epithelium in the LCM discovery and LCM validation cohorts. E, Scatterplots showing the correlation between DS% determined from H&E and ssGSEA scores for the Epithelial Mesenchymal Transition (left; Spearman rho = 0.69, P < 2.2e-16), KRAS Signaling Up (middle; Spearman rho = 0.48, P < 2.2e-16) and MYC Targets V2 (right; Spearman rho = -0.41, P < 2.2e-16) Hallmark gene sets. We identified two cases representative of low and high DS% in each of these analyses (red circles). F, H&E along with HALO mark-up for the representative low and high DS% samples identified in (E). CRC, colorectal cancer; LCM, laser capture microdissection.

Immune cell spatial deconvolution. A, Representative images from 1 CRLM (Patient A, Fig. 4A) showing cell detection from the Qupath package used on a CD3 and CD66b IHC-stained liver metastasis to count number of CD3 and CD66b-positive cells from 21 regions from a total of three CRLM. This count was compared with the SpatialDecon-derived count that uses the transcriptomic data from the corresponding ROI in the GeoMx mIF-stained matched sample (See Supplementary Fig. S5). B, Bland Altman plot comparing transcriptome SpatialDecon-derived cell count versus the IHC-derived cell count. C, Correlation plot comparing transcriptome SpatialDecon-derived cell count versus the IHC-derived cell count.

BULK IO360 transcriptomic characterization of matched primary colorectal cancer and CRLM. A, PCA plot demonstrating two principal components of minimal variance for all samples. Primary colorectal cancer samples are demonstrated by circles and yellow outline. CRLM are demonstrated by triangular points and brown outline. KMhigh, KMlow Stromalow and KMlow Stromahigh samples are depicted by color. B, Unsupervised analysis using gene expression correlation matrix for all samples. Patient, site, KM grade, and TSP are depicted by key. Spearman correlation of all expressed genes performed between each sample sequenced and plotted on the heatmap. k-means clustering of heatmap to demonstrate correlated samples. Red, strong correlation. Blue, negative correlation. C, Volcano plots demonstrating differential gene expression results and clustered heatmap of significant genes for (i) all primary colorectal cancer versus CRLM; (ii) KM grade: KM high versus KM low primary colorectal cancer; (iii) KM grade: KM high versus KM low CRLM. The x-axis of volcano plot demonstrates log2-fold change; y-axis demonstrates –log10P. Colored points demonstrate significant changes in gene expression between groups (P < 0.05 and logFC > 1.5). Volcano plots demonstrate top 20 differentially expressed genes for each group. D, Heatmap demonstrating GSEA results comparing the different tumors grouped according with KM grade and TSP using io360-curated gene sets annotated on the right of the diagram. Heatmap squares represent log10-adjusted P value. Green, upregulation in group 1; orange, upregulation in group 2. The heatmap is clustered by y-axis only to demonstrate frequently upregulated gene sets. E, Box plot comparisons of immune cell populations and selected cell:cell ratios between KM grade and TSP segregated groups using deconvolution software included in the nCounter package. Annotated subgroups are: Kmhigh, Kmlow Stromalow, Kmlow Stromahigh. The y-axis represents log10 of estimated cell count. Primary colorectal cancer is represented in i and CRLM in ii.

Stromal influence on widely used transcriptional signatures. A, GSEA of Hallmark gene sets in LCM discovery and validation cohorts. Only gene sets significantly and concordantly enriched in stroma or epithelium in both the discovery and validation cohorts are shown (Padj < 0.02). B, Heatmap of ssGSEA scores for the Hallmark gene sets in the FACS validation cohort samples. Only the gene sets significantly and concordantly enriched in stroma or epithelium in both the LCM discovery and validation cohorts are shown (Padj < 0.02). C, TFs whose activity was significantly and concordantly enriched in stroma or epithelium in both the LCM discovery and validation cohorts (P < 0.05). D, Heatmap of the inferred activity scores for the same TFs in the FACS validation cohort. For all panels in Fig. 2, gene sets/transcription factors with names/symbols colored orange were significantly and consistently enriched in stroma in the LCM discovery and validation cohorts, whereas gene sets/TFs with names/symbols colored blue were consistently and significantly enriched in epithelium in the LCM discovery and validation cohorts (gene sets: Padj < 0.02; TFs: P < 0.05). LCM, laser capture microdissection; NES, normalized enrichment scale; TF, transcription factor.

All supplementary figures and supplementary tables 1 and 2 from the manuscript

BACKGROUND:Pancreatic Ductal Adenocarcinoma (PDAC) patients exhibit varied responses to multimodal therapy. RNA gene sequencing has unravelled distinct tumour biology subtypes, forming the focus of this review exploring its impact on survival outcomes. METHODS:A systematic search across PubMed, Medline, Embase, and CINAHL databases targeted studies assessing long-term overall and disease-free survival in PDAC patients with molecular subtyping. RESULTS:Fifteen studies including 2731 patients were identified. Molecular subtyping was performed by RNA sequencing and Immunohistochemistry in 14 studies and by Mass Spectrometry in 1 study. Two main tumour subtypes were identified (classical and basal-like or squamous) with basal like associated with poorer outcomes. Further subtypes were identified in individual studies. Superior survival was seen with classical subtype in all other analyses that compared the classical and basal subtypes. High risk stromal subtypes were identified on further analysis of the stroma and were associated with a worse survival independent of the tumour subtype. CONCLUSION:Molecular subtyping of PDAC specimens can identify patients with high-risk tumour biology and poor survival outcomes. Routine subtyping is limited by the cost of RNA sequencing and the volume of raw data generated which has made its translation into routine clinical practice difficult.

BACKGROUND:Predictors of long-term survival after resection of adenocarcinoma arising from intraductal papillary mucinous neoplasms are unknown. This study determines predictors of long-term (>5 years) disease-free survival and recurrence in adenocarcinoma arising from intraductal papillary mucinous neoplasms and derives a prognostic model for disease-free survival. METHODS:Consecutive patients who underwent pancreatic resection for adenocarcinoma arising from intraductal papillary mucinous neoplasms in 18 academic pancreatic centers in Europe and Asia between 2010 to 2017 with at least 5-year follow-up were identified. Factors associated with disease-free survival were determined using Cox proportional hazards model. Internal validation was performed, and discrimination and calibration indices were assessed. RESULTS:In the study, 288 patients (median age, 70 years; 52% male) were identified; 140 (48%) patients developed recurrence after a median follow-up of 98 months (interquartile range, 78.4-123), 57 patients (19.8%) developed locoregional recurrence, and 109 patients (37.8%) systemic recurrence. At 5 years after resection, the overall and disease-free survival was 46.5% (134/288) and 35.0% (101/288), respectively. On Cox proportional hazards model analysis, multivisceral resection (hazard ratio, 2.20; 95% confidence interval, 1.06-4.60), pancreatic tail location (hazard ratio, 2.34; 95% confidence interval, 1.22-4.50), poor tumor differentiation (hazard ratio, 2.48; 95% confidence interval, 1.10-5.30), lymphovascular invasion (hazard ratio, 1.74; 95% confidence interval, 1.06-2.88), and perineural invasion (hazard ratio, 1.83; 95% confidence interval, 1.09-3.10) were negatively associated with long-term disease-free survival. The final predictive model incorporated 8 predictors and demonstrated good predictive ability for disease-free survival (C-index, 0.74; calibration, slope 1.00). CONCLUSION:A third of patients achieve long-term disease-free survival (>5 years) after pancreatic resection for adenocarcinoma arising from intraductal papillary mucinous neoplasms. The predictive model developed in the current study can be used to estimate the probability of long-term disease-free survival.