Abstract B033: Characterizing the spatial transcriptomic landscape of Osteosarcoma from diagnosis to relapse
Bibliographic record
Abstract
Abstract Introduction This study aims at elucidating mechanisms of immune infiltration in osteosarcoma (OSA) to identify novel therapeutic strategies. Methods Here, joint analysis of spatial transcriptomics (ST, Visium) data from 26 formalin-fixed paraffin embedded OSA samples at diagnosis (n=9) and relapse (n=17, MAPPYACTS NCT02613962 & OS2006 NCT00470223) from 19 donors (including 5 diagnosis/relapse pairs) was performed using the CellsFromSpace unbiased reference-free signal deconvolution and analysis workflow. This methodology enabled the definition of the topology and the molecular signatures of the cancerous, stromal and immune components across all samples, and allowed for the phenotyping of individual tumors based on cancer subset composition. Spatial colocalization analysis was performed using Lee’s bivariate spatial association measure of different cell populations, enabling the characterization of tumor-infiltrating immune cells found across samples. Similar ST analysis of Patient-derived xenograft (PDX) samples was performed on fresh-frozen samples. Subsequent separation of human and mouse reads using the Xenome algorithm allowed for a clear dissection of the cancer vs. microenvironment compartments. Preliminary signature validation on larger cohorts was performed on bulk RNAseq samples from the OS2006 and MAPPYACTS cohorts. Results Our joint spatial transcriptomics analysis of 26 primary and relapse tumors revealed 64 transcriptomic signatures associated with cancer cells, some of which were shared across multiple samples. These signatures could be consolidated into 18 broad phenotypic categories frequently detectable in various samples, showcasing significant intratumoral heterogeneity within the cancerous compartment. By averaging the composition per sample, we found that samples clustered based on the relative intratumoral abundance of cancer phenotypes, indicating the presence of at least two main differentiation archetypes in OSA: “High intratumoral heterogeneity” and “Undifferentiated.” Additionally, spatial niche analysis of immune cells within tumors identified immune populations consistently associated with tumors across samples. This analysis revealed a specific myeloid lineage capable of infiltrating tumors in all samples, likely mediated by chemotactic signals linked to cancer differentiation, as suggested by ligand-receptor analysis. To validate these findings, PDX models of OSA were analyzed, confirming the presence of a similar phenotype of tumor-infiltrating myeloid cells and underscoring the relevance of PDX models in OSA research. Furthermore, preliminary analyses querying the transcriptomic signature of infiltrating myeloid cells in larger bulk RNAseq cohorts revealed promising associations with patient outcomes, detectable even at diagnosis. Conclusions Our research elucidates patterns of intratumoral phenotypic heterogeneity in OSA and identifies the primary immune cell populations specifically infiltrating OSA tumors. This paves the way for developing novel therapies by leveraging these cellular tropism mechanisms. Citation Format: Gaël Moquin-Beaudry, Maria Eugenia Marques Da Costa, Hanane Zair, Corentin Thuilliez, Pierre Khneisser, Nicolas Signolle, Jean-Yves Scoazec, Birgit Geoerger, Nathalie Gaspar, Antonin Marchais. Characterizing the spatial transcriptomic landscape of Osteosarcoma from diagnosis to relapse [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pediatric Cancer Research; 2024 Sep 5-8; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl):Abstract nr B033.
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 | 0.001 |
| 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 | 0.000 |
| 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".