Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Evaluation of gene expression-based predictors of lymph node metastasis in bladder cancer

Johansson, Hafdís Birta ; Liedberg, Fredrik LU ; Bernardo, Carina LU orcid ; Zadoroznyj, Aymeric LU ; Mattsson, Carl-Adam ; Höglund, Mattias LU ; Eriksson, Pontus LU and Sjödahl, Gottfrid LU orcid (2025) In Bladder Cancer 11(3).
Abstract

Background: The presence of cancer in pelvic lymph nodes removed during radical surgery for muscle-invasive bladder cancer (MIBC) is a key determinant of patient outcome. It would be beneficial to predict node status preoperatively to tailor the use of neoadjuvant chemotherapy and extent of lymph node dissection. Of 12 published node status predictors based on tumor RNA expression signatures, none have been successfully validated in subsequent reports. Objective: We aimed to validate all published node status predictors and evaluate new prediction models in MIBC. Methods: Gene expression data and node status from two MIBC cohorts were used to test 12 published node-predictive signatures. The overlap in differential expression was... (More)

Background: The presence of cancer in pelvic lymph nodes removed during radical surgery for muscle-invasive bladder cancer (MIBC) is a key determinant of patient outcome. It would be beneficial to predict node status preoperatively to tailor the use of neoadjuvant chemotherapy and extent of lymph node dissection. Of 12 published node status predictors based on tumor RNA expression signatures, none have been successfully validated in subsequent reports. Objective: We aimed to validate all published node status predictors and evaluate new prediction models in MIBC. Methods: Gene expression data and node status from two MIBC cohorts were used to test 12 published node-predictive signatures. The overlap in differential expression was examined across the two datasets, and new prediction models were tested in cross-validation and by application to the independent cohort. Results: Published node status predictors performed either no better, or only slightly better than chance in the two independent validation datasets (maximum AUC 0.59 and 0.65, and maximum balanced accuracy 0.54 and 0.57). Among very few genes and signatures differentially expressed in the same direction in both data sets we identified upregulation of interferon-response signatures in node negative cases. Transcriptomic predictors trained in one dataset performed poorly when applied to the independent dataset (AUC 0.60-0.62). Conclusions: In this systematic evaluation, neither the 12 published signatures nor our own models reached an adequate performance for clinical node status prediction in independent data. This indicates that the biological determinants of nodal spread are poorly captured by bulk tumor RNA expression profiles.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
bladder cancer, node status, lymph-node, signatures
in
Bladder Cancer
volume
11
issue
3
pages
12 pages
publisher
SAGE Publications
external identifiers
  • pmid:40861318
ISSN
2352-3727
DOI
10.1177/23523735251370895
language
English
LU publication?
yes
additional info
© The Author(s) 2025.
id
89ac9979-8d80-4fb0-be9d-e132a0a073cb
date added to LUP
2026-03-30 09:40:03
date last changed
2026-03-30 10:57:41
@article{89ac9979-8d80-4fb0-be9d-e132a0a073cb,
  abstract     = {{<p> Background: The presence of cancer in pelvic lymph nodes removed during radical surgery for muscle-invasive bladder cancer (MIBC) is a key determinant of patient outcome. It would be beneficial to predict node status preoperatively to tailor the use of neoadjuvant chemotherapy and extent of lymph node dissection. Of 12 published node status predictors based on tumor RNA expression signatures, none have been successfully validated in subsequent reports.  Objective: We aimed to validate all published node status predictors and evaluate new prediction models in MIBC.  Methods: Gene expression data and node status from two MIBC cohorts were used to test 12 published node-predictive signatures. The overlap in differential expression was examined across the two datasets, and new prediction models were tested in cross-validation and by application to the independent cohort.  Results: Published node status predictors performed either no better, or only slightly better than chance in the two independent validation datasets (maximum AUC 0.59 and 0.65, and maximum balanced accuracy 0.54 and 0.57). Among very few genes and signatures differentially expressed in the same direction in both data sets we identified upregulation of interferon-response signatures in node negative cases. Transcriptomic predictors trained in one dataset performed poorly when applied to the independent dataset (AUC 0.60-0.62). Conclusions: In this systematic evaluation, neither the 12 published signatures nor our own models reached an adequate performance for clinical node status prediction in independent data. This indicates that the biological determinants of nodal spread are poorly captured by bulk tumor RNA expression profiles. </p>}},
  author       = {{Johansson, Hafdís Birta and Liedberg, Fredrik and Bernardo, Carina and Zadoroznyj, Aymeric and Mattsson, Carl-Adam and Höglund, Mattias and Eriksson, Pontus and Sjödahl, Gottfrid}},
  issn         = {{2352-3727}},
  keywords     = {{bladder cancer; node status; lymph-node; signatures}},
  language     = {{eng}},
  month        = {{08}},
  number       = {{3}},
  publisher    = {{SAGE Publications}},
  series       = {{Bladder Cancer}},
  title        = {{Evaluation of gene expression-based predictors of lymph node metastasis in bladder cancer}},
  url          = {{http://dx.doi.org/10.1177/23523735251370895}},
  doi          = {{10.1177/23523735251370895}},
  volume       = {{11}},
  year         = {{2025}},
}