EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Dual-Foundation-Model Ensemble Predicts Gene Expression in Muscle-Invasive Bladder Cancer for Patient Outcome Stratification

Ingvild Frøberg Mathisen, Florestan J Koll, Robin Sebastian Mayer, Marcel H Schulz, Julia Bein, Felix Chun, Markus Eckstein, Anne Fassl, Jens Köllermann, Mike Wenzel, Nadine Flinner, Peter J Wild

Modern Pathology · 2026-10 · 101092

DOI: 10.1016/j.modpat.2026.101092 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Modern Pathology, 2026-10. Verfasst von Ingvild Frøberg Mathisen und 11 weiteren Autor:innen. Beteiligte Einrichtung: Goethe-University Frankfurt, Dr. Senckenberg Institutes of Pathology and Human Genetics, University Hospital Frankfurt, Frankfurt am Main, Germany. E…. Die Arbeit nennt 74 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Alfons and Gertrud Kassel Trust.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Muscle-invasive bladder cancer (MIBC) exhibits significant heterogeneity in clinical outcomes and treatment responses. While gene expression profiling has provided important biological and prognostic insight, transcriptomic assays are not routinely incorporated into standard clinical practice, in part due to cost. AI models have been demonstrated to predict gene expression from H&E-stain whole-slide images (WSIs) potentially offering a scalable approach for gathering transcriptomic information for patient stratification and biomarker discovery. We developed and tested models to predict expression of 857 genes directly from H&E-stained whole-slide images (WSI). The best performing final model was a dual-ensemble combining predictions from models trained on features extracted with different foundation models. The dual-ensemble reached 776/857 predictable genes with an average correlation of 0.43 (cross-validation), 0.45 (holdout) and 0.40 (external Frankfurt cohort). Predicted expression strongly correlates with expression of clinically relevant markers (like NECTIN4). Estimated hazard ratios of predicted expression correlated strongly with those of measured expression and genes associated with overall survival for both predicted and measured expression in both cohorts were identified. AI-predicted basal vs. luminal subtypes were associated with overall survival and aligned with the ground-truth (accuracies > 0.8). Findings demonstrate that gene expressions inferred from routine staining can support biomarker discovery and patient stratification in MIBC.

Schlagwörter

Stichwörter der Autor:innen

bladder cancercomputational pathologyensemble learningfoundation modelsmachine learningsurvival

Autor:innen und Institutionen

  • Goethe-University Frankfurt, Dr. Senckenberg Institutes of Pathology and Human Genetics, University Hospital Frankfurt, Frankfurt am Main, Germany. Electronic address: Ingvild.FrobergMathisen@ukffm.de.
  • LOEWE-Center Frankfurt Cancer Institute (FCI), Frankfurt am Main, Germany; Goethe-University Frankfurt, Department of Urology, University Hospital Frankfurt, Frankfurt am Main, Germany; Department of Urology, Medical Un…
  • Goethe-University Frankfurt, Dr. Senckenberg Institutes of Pathology and Human Genetics, University Hospital Frankfurt, Frankfurt am Main, Germany; LOEWE-Center Frankfurt Cancer Institute (FCI), Frankfurt am Main, Germa…
  • Goethe-University Frankfurt, Institute for Computational Genomic Medicine, Frankfurt am Main, Germany.
  • Goethe-University Frankfurt, Dr. Senckenberg Institutes of Pathology and Human Genetics, University Hospital Frankfurt, Frankfurt am Main, Germany.
  • LOEWE-Center Frankfurt Cancer Institute (FCI), Frankfurt am Main, Germany; Goethe-University Frankfurt, Department of Urology, University Hospital Frankfurt, Frankfurt am Main, Germany.

Förderung

  • Alfons and Gertrud Kassel Trust

Zitieren

APA

Mathisen, I. F., Koll, F. J., Mayer, R. S., Schulz, M. H., Bein, J., Chun, F., Eckstein, M., Fassl, A., Köllermann, J., Wenzel, M., Flinner, N., & Wild, P. J. (2026). Dual-Foundation-Model Ensemble Predicts Gene Expression in Muscle-Invasive Bladder Cancer for Patient Outcome Stratification. Modern Pathology, 101092. https://doi.org/10.1016/j.modpat.2026.101092

Vancouver

Mathisen IF, Koll FJ, Mayer RS, Schulz MH, Bein J, Chun F, et al. Dual-Foundation-Model Ensemble Predicts Gene Expression in Muscle-Invasive Bladder Cancer for Patient Outcome Stratification. Modern Pathology. 2026:101092. doi:10.1016/j.modpat.2026.101092

BibTeX
@article{Mathisen2026,
  author = {Mathisen, Ingvild Frøberg and Koll, Florestan J and Mayer, Robin Sebastian and Schulz, Marcel H and Bein, Julia and Chun, Felix and Eckstein, Markus and Fassl, Anne and Köllermann, Jens and Wenzel, Mike and Flinner, Nadine and Wild, Peter J},
  title = {Dual-Foundation-Model Ensemble Predicts Gene Expression in Muscle-Invasive Bladder Cancer for Patient Outcome Stratification},
  journal = {Modern Pathology},
  year = {2026},
  pages = {101092},
  doi = {10.1016/j.modpat.2026.101092}
}
Kostenlos starten →