Frontiers in Oncology (Dec 2024)

Combining dosiomics and machine learning methods for predicting severe cardiac diseases in childhood cancer survivors: the French Childhood Cancer Survivor Study

  • Mahmoud Bentriou,
  • Véronique Letort,
  • Stefania Chounta,
  • Stefania Chounta,
  • Stefania Chounta,
  • Stefania Chounta,
  • Brice Fresneau,
  • Brice Fresneau,
  • Brice Fresneau,
  • Brice Fresneau,
  • Duyen Do,
  • Duyen Do,
  • Duyen Do,
  • Nadia Haddy,
  • Nadia Haddy,
  • Nadia Haddy,
  • Ibrahima Diallo,
  • Ibrahima Diallo,
  • Neige Journy,
  • Neige Journy,
  • Neige Journy,
  • Monia Zidane,
  • Monia Zidane,
  • Monia Zidane,
  • Thibaud Charrier,
  • Thibaud Charrier,
  • Thibaud Charrier,
  • Naila Aba,
  • Naila Aba,
  • Naila Aba,
  • Claire Ducos,
  • Claire Ducos,
  • Claire Ducos,
  • Vincent S. Zossou,
  • Vincent S. Zossou,
  • Vincent S. Zossou,
  • Florent de Vathaire,
  • Florent de Vathaire,
  • Florent de Vathaire,
  • Rodrigue S. Allodji,
  • Rodrigue S. Allodji,
  • Rodrigue S. Allodji,
  • Rodrigue S. Allodji,
  • Sarah Lemler

DOI
https://doi.org/10.3389/fonc.2024.1241221
Journal volume & issue
Vol. 14

Abstract

Read online

BackgroundCardiac disease (CD) is a primary long-term diagnosed pathology among childhood cancer survivors. Dosiomics (radiomics extracted from the dose distribution) have received attention in the past few years to assess better the induced risk of radiotherapy (RT) than standard dosimetric features such as dose-volume indicators. Hence, using the spatial information contained in the dosiomics features with machine learning methods may improve the prediction of CD.MethodsWe considered the 7670 5-year survivors of the French Childhood Cancer Survivors Study (FCCSS). Dose-volume and dosiomics features are extracted from the radiation dose distribution of 3943 patients treated with RT. Survival analysis is performed considering several groups of features and several models [Cox Proportional Hazard with Lasso penalty, Cox with Bootstrap Lasso selection, Random Survival Forests (RSF)]. We establish the performance of dosiomics compared to baseline models by estimating C-index and Integrated Brier Score (IBS) metrics with 5-fold stratified cross-validation and compare their time-dependent error curves.ResultsAn RSF model adjusted on the first-order dosiomics predictors extracted from the whole heart performed best regarding the C-index (0.792 ± 0.049), and an RSF model adjusted on the first-order dosiomics predictors extracted from the heart’s subparts performed best regarding the IBS (0.069 ± 0.05). However, the difference is not statistically significant with the standard models (C-index of Cox PH adjusted on dose-volume indicators: 0.791 ± 0.044; IBS of Cox PH adjusted on the mean dose to the heart: 0.074 ± 0.056).ConclusionIn this study, dosiomics models have slightly better performance metrics but they do not outperform the standard models significantly. Quantiles of the dose distribution may contain enough information to estimate the risk of late radio-induced high-grade CD in childhood cancer survivors.

Keywords