Abstract B128: Spatial proteomics extend tumour subtype and microenvironment classifications across genotype-phenotype axes of pancreatic cancer
Bibliographic record
Abstract
Abstract Human pancreatic ductal adenocarcinomas (PDAC) are composed of sparse, heterogeneous tumour cell populations embedded within a dense, desmoplastic stroma and are highly resistant to existing therapies. Guided by underlying genomic aberrations, cancer cell states and phenotypes interact with their microenvironment to shape disease trajectory and therapeutic response. To quantify the cellular phenotypic heterogeneity of these complex tumour-microenvironment interactions we designed custom multiplexed histopathology imaging panels against single cell RNA sequencing-defined tumour, immune, and stromal cell compartments. Using imaging mass cytometry (IMC), we profiled the co-localization and organization of 83 cell types and their functional states in 221 resected pancreatic tumours. This identified extensive inter- and intra-patient heterogeneity including the presence of both classical and basal cell types in most patients and a gradient of pancreatic epithelial identity confirmed by scRNAseq. This single cell content defined an expanded spectrum of classifiable intermediate tumour subtypes between classical and basal which have specific associations to neighboring immune and stroma cellular content, ploidy, established broad transcriptional subtypes, heterogeneity, and patient outcome. Eight reproducible cancer microenvironments were quantified, empowering cross-tumour comparisons. Matched whole genome sequencing (WGS) identified aberrations associated with specific tumour phenotypes and microenvironments. These results support asynchronous tumour and microenvironment genotype-phenotype axes which subdivide basal to classical tumour polarization and categorize co-existing microenvironments from stiff, ECM-rich, immune suppressed to immune infiltrated regions alongside fibrovascularized CD105+ stroma. This is based upon spatially heterogeneous co-occurring microenvironment niches and correlated tumour phenotypes which are associated with KRAS or MYC amplification and CDKN2A or LATS2 deletion on one axis, mutation of epigenetic modifiers in intermediate and hybrid subtypes, and RNF43 deletion on the other axis. To deeply profile the signalling pathways and enriched biological processes characteristic of each tissue state, we performed deep proteomics measurements of each microenvironment and tumor phenotype through whole-slide IMC informed laser capture microdissection and mass spectrometry of 273 micro-regions from a subset of patients. Finally, to identify a robust, minimized set of outcome-relevant features, we used a multi-modal technology optimized machine learning model trained to predict overall survival. We compare the predictive potential of WGS and spatial proteomics and show that a combination of genomic and cellular content outperforms clinical features and platform specific models thereby demonstrating the synergistic benefit of integrated multi-modal data. Our findings classify the spatial organization of human PDAC to identify an intertwined tumour/microenvironment axis that is influenced by specific genomic aberrations. Citation Format: Noor Shakfa, Ferris Nowlan, Tiak Ju Tan, Sibyl Drissler, Beth Sunnucks, Jennifer Gorman, Chengxin Yu, Michael Geuenich, Sheng Ben Liang, Barbara Gruenwald, Ayelet Borgida, Cassandra Wong, Brendon Seale, Zhen Yuan Lin, Edward Chen, Golnaz Abazari, Miralem Mrkjonic, Julie Wilson, Kieran Campbell, Anne-Claude Gingras, Rob Grant, Grainne O'Kane, Faiyez Notta, Steve Gallinger, Hartland Jackson. Spatial proteomics extend tumour subtype and microenvironment classifications across genotype-phenotype axes of pancreatic cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research—Emerging Science Driving Transformative Solutions; Boston, MA; 2025 Sep 28-Oct 1; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_3):Abstract nr B128.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".