Epigenetic patterns and methylation-based models for robust outcome prediction in osteosarcoma
Abstract
Osteosarcoma (OSA) is an aggressive bone malignancy primarily affecting children and young adults. Long-term survival has remained unchanged for at least three decades, and no robust molecular prognostic factors have emerged that might help improve this outcome. The complex genomic architecture and genetic heterogeneity of OSA have constrained the identification and validation of such molecular predictors. Recent work has shown that DNA methylation is associated with outcome, yet clinically applicable models building on these findings have not been developed. To address this unmet need, we developed and validated region-based methylation models for individualized risk prognostication and treatment response prediction at the time of diagnosis. These findings were further contextualized through an unsupervised exploration of methylation features, showing an association between widespread global hypermethylation and unfavorable outcomes. The prognostic effect of methylation was independent of known confounding factors. Additionally, an examination of methylation-derived epigenetic age provided orthogonal information with significant association to outcome. Finally, clustering phenotypes were compared between methylation and several other genomic layers (mRNA, miRNA, and copy number variation) and showed little overlap with one another. Together, these findings establish a foundation for unique methylation-based outcome stratification at the time of diagnosis and underscore the need for continued investigation toward clinical application.
Related
- RNF152 promoter methylation and gene expression in oral premalignant lesions and oral squamous cell carcinoma: implications for epigenetic biomarker development
- Fusion of Deep Learning Radiomics and Clinical Features for Individualized Prediction of Neoadjuvant Chemotherapy Response in Osteosarcoma
- Methylation profiling reveals epigenetic programs associated with chr18 loss in small intestinal neuroendocrine tumors
- Ensemble‐Based Machine Learning Models for Classifying Intervention‐Related Psychological Outcome Patterns in Menopausal Women Treated With Pimpinella anisum
- COPD/OSA phenotypes and inpatient outcome patterns among primary total hip arthroplasty hospitalizations: a national inpatient sample study, 2016–2022
BibTeX
@misc{bowers2026epigenetic,
title = {{Epigenetic patterns and methylation-based models for robust outcome prediction in osteosarcoma}},
author = {Joshua J. Bowers and Nikhil Ramavenkat and Christopher E. Lietz and Miya Hugaboom and Aditya Madupur and Edwin Choy and Gregory M. Cote and Hang Lee and Martin J. Aryee and Dimitrios Spentzos},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.1101/2025.11.19.25340108},
url = {https://doi.org/10.1101/2025.11.19.25340108},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Epigenetic patterns and methylation-based models for robust outcome prediction in osteosarcoma
AU - Joshua J. Bowers
AU - Nikhil Ramavenkat
AU - Christopher E. Lietz
AU - Miya Hugaboom
AU - Aditya Madupur
AU - Edwin Choy
AU - Gregory M. Cote
AU - Hang Lee
AU - Martin J. Aryee
AU - Dimitrios Spentzos
PY - 2026
DA - 2026/10/14
PB - medRxiv
DO - 10.1101/2025.11.19.25340108
UR - https://doi.org/10.1101/2025.11.19.25340108
AB - Osteosarcoma (OSA) is an aggressive bone malignancy primarily affecting children and young adults. Long-term survival has remained unchanged for at least three decades, and no robust molecular prognostic factors have emerged that might help improve this outcome. The complex genomic architecture and genetic heterogeneity of OSA have constrained the identification and validation of such molecular predictors. Recent work has shown that DNA methylation is associated with outcome, yet clinically applicable models building on these findings have not been developed. To address this unmet need, we developed and validated region-based methylation models for individualized risk prognostication and treatment response prediction at the time of diagnosis. These findings were further contextualized through an unsupervised exploration of methylation features, showing an association between widespread global hypermethylation and unfavorable outcomes. The prognostic effect of methylation was independent of known confounding factors. Additionally, an examination of methylation-derived epigenetic age provided orthogonal information with significant association to outcome. Finally, clustering phenotypes were compared between methylation and several other genomic layers (mRNA, miRNA, and copy number variation) and showed little overlap with one another. Together, these findings establish a foundation for unique methylation-based outcome stratification at the time of diagnosis and underscore the need for continued investigation toward clinical application.
ER -