Telomere Length Dynamics and Chromosomal Instability for Predicting Individual Radiosensitivity and Risk via Machine...

Area of Science:

• Oncology
• Genetics
• Radiotherapy Research

Background:

• Predicting patient response to radiotherapy and risk of late adverse effects is essential for personalized cancer treatment.
• Telomeres are key biomarkers for radiosensitivity, as radiation exposure can cause telomere pathologies linked to radiation-induced late effects.

Purpose of the Study:

• To assess telomere length longitudinally in prostate cancer patients undergoing Intensity Modulated Radiation Therapy (IMRT).
• To evaluate genome instability using chromosome aberrations for predicting secondary malignancy risk.
• To implement telomere length data in a machine learning model to predict radiotherapy outcomes and individual radiosensitivity.

Main Methods:

• Longitudinal assessment of telomere length in 15 prostate cancer patients using Telomere Fluorescence in situ Hybridization (Telo-FISH).
• Assessment of chromosome aberrations via directional Genomic Hybridization (dGH) for inversion detection.
• Development of an XGBoost machine learning model using baseline and in vitro irradiated telomere length data.

Main Results:

• Telomere length and chromosomal instability were measured to predict post-radiotherapy telomeric outcomes.
• The study presents the first machine learning model integrating individual telomere length for radiosensitivity prediction.
• Combined telomere length and chromosomal instability data offer insights into individual radiosensitivity and late effects risk.

Conclusions:

• Individual telomere length and chromosomal instability are valuable biomarkers for predicting radiotherapy response and late effects.
• Machine learning models incorporating telomere data can enhance personalized radiotherapy treatment strategies.
• This approach aids in assessing individual patient risk for adverse health outcomes and secondary malignancies.