Exploratory machine-learning evaluation of clinical and dosimetric features in late prostate radiotherapy toxicity.
5 affiliations
- 1School of Physics, Clinical & Optometric Sciences, Central Quad, Technological University Dublin, City Campus, Grangegorman, Ireland.
- 2Centre for Radiation and Environmental Science, Physical to Life Sciences Research Hub, Technological University, Dublin, Ireland.
- 3St. Vincent's Private Hospital, Merrion Road, Dublin 4, D04 N2E0, Ireland.
- 4Division of Genetics and Genome Biology, Leicester Cancer Research Centre, University of Leicester, University Road, Leicester, LE1 7RH, United Kingdom.
- 5ADAPT Research Centre, School of Computer Science and Statistics, Trinity College Dublin, Ireland.
Abstract
Prediction of late radiotherapy toxicity remains challenging, as routinely available clinical and dosimetric variables may incompletely capture individual radiosensitivity. This study evaluated interpretable machine-learning models for predicting baseline- corrected late gastrointestinal (GI) and genitourinary (GU) toxicity following prostate radiotherapy and examined model-derived feature contributions. Data from 229 men with prostate cancer enrolled in the REQUITE cohort were retrospectively analysed. Late toxicity was defined as an increase of ≥1 CTCAE grade from baseline occurring ≥12 months following radiotherapy. Prespecified clinical and dosimetric predictors were evaluated separately for GI and GU toxicity, with correlated dose-volume parameters represented using organ-specific equivalent uniform dose (EUD). L1-regularised logistic regression, Elastic Net and XGBoost were evaluated using repeated nested stratified cross-validation. Performance was assessed using out-of-fold predictions, bootstrap confidence intervals and outcome-label permutation testing. SHAP analysis assessed feature contributions and stability across repeats. GI and GU toxicity worsening occurred in 107 (46.7%) and 125 (54.6%) patients, respectively. GI ROC-AUCs were 0.573, 0.571 and 0.587 for L1, Elastic Net and XGBoost, respectively, with permutation testing indicating above-chance discrimination (p = 0.036, 0.047 and 0.022). Rectal EUD was the most consistently influential GI predictor. GU discrimination was below chance, with no evidence of discrimination beyond the null distribution. GU feature importance was more heterogeneous and model dependent. Routine clinical and dosimetric variables provided limited predictive information for late GI toxicity but did not reliably discriminate GU toxicity. The findings support investigation of additional dosimetric and biological markers for toxicity prediction.
Related
- Comparative Dosimetric Evaluation of Tomotherapy with a 1-cm Field Width versus Volumetric-modulated Arc Therapy for Whole-brain Radiotherapy with Hippocampal Avoidance
- Hypofractionated and conventionally fractionated radiotherapy for synchronous bilateral breast cancer: clinical and dosimetric outcomes
- Microbeam radiotherapy - clinical and dosimetric investigation of a first clinical application.
- Dosimetric Comparison of Intensity-Modulated Radiotherapy and Four-Field Three-Dimensional Conformal Radiotherapy in the Management of Prostate Cancer: A Retrospective Dual-Centre Analysis
- Dosimetric and Treatment Delivery Efficiency Comparison of Two Ring-gantry Linear Accelerator Platforms for Liver Stereotactic Body Radiotherapy
BibTeX
@article{jordan2026exploratory,
title = {{Exploratory machine-learning evaluation of clinical and dosimetric features in late prostate radiotherapy toxicity.}},
author = {Karl Jordan and Christopher J Talbot and John D Kelleher and Fiona Lyng and Aidan D Meade},
journal = {Clinical and translational radiation oncology},
year = {2026},
month = oct,
doi = {10.1016/j.ctro.2026.101305},
url = {https://europepmc.org/article/MED/42999685},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Exploratory machine-learning evaluation of clinical and dosimetric features in late prostate radiotherapy toxicity.
AU - Karl Jordan
AU - Christopher J Talbot
AU - John D Kelleher
AU - Fiona Lyng
AU - Aidan D Meade
PY - 2026
DA - 2026/10/15
JO - Clinical and translational radiation oncology
DO - 10.1016/j.ctro.2026.101305
UR - https://europepmc.org/article/MED/42999685
AB - Prediction of late radiotherapy toxicity remains challenging, as routinely available clinical and dosimetric variables may incompletely capture individual radiosensitivity. This study evaluated interpretable machine-learning models for predicting baseline- corrected late gastrointestinal (GI) and genitourinary (GU) toxicity following prostate radiotherapy and examined model-derived feature contributions. Data from 229 men with prostate cancer enrolled in the REQUITE cohort were retrospectively analysed. Late toxicity was defined as an increase of ≥1 CTCAE grade from baseline occurring ≥12 months following radiotherapy. Prespecified clinical and dosimetric predictors were evaluated separately for GI and GU toxicity, with correlated dose-volume parameters represented using organ-specific equivalent uniform dose (EUD). L1-regularised logistic regression, Elastic Net and XGBoost were evaluated using repeated nested stratified cross-validation. Performance was assessed using out-of-fold predictions, bootstrap confidence intervals and outcome-label permutation testing. SHAP analysis assessed feature contributions and stability across repeats. GI and GU toxicity worsening occurred in 107 (46.7%) and 125 (54.6%) patients, respectively. GI ROC-AUCs were 0.573, 0.571 and 0.587 for L1, Elastic Net and XGBoost, respectively, with permutation testing indicating above-chance discrimination (p = 0.036, 0.047 and 0.022). Rectal EUD was the most consistently influential GI predictor. GU discrimination was below chance, with no evidence of discrimination beyond the null distribution. GU feature importance was more heterogeneous and model dependent. Routine clinical and dosimetric variables provided limited predictive information for late GI toxicity but did not reliably discriminate GU toxicity. The findings support investigation of additional dosimetric and biological markers for toxicity prediction.
SN - 2405-6308
ER -