Abstract:Purpose: Predicting the efficacy of radiotherapy in individual patients has drawn widespread attention, but the limited sample size remains a bottleneck for utilizing high-dimensional multi-omics data to guide personalized radiotherapy. We hypothesize the recently developed meta-learning framework could address this limitation. Methods and materials: By combining gene expression, DNA methylation, and clinical data of 806 patients who had received radiotherapy from The Cancer Genome Atlas (TCGA), we applied the Model-Agnostic Meta-Learning (MAML) framework to tasks consisting of pan-cancer data, to obtain the best initial parameters of a neural network for a specific cancer with smaller number of samples. The performance of meta-learning framework was compared with four traditional machine learning methods based on two training schemes, and tested on Cancer Cell Line Encyclopedia (CCLE) and Chinese Glioma Genome Atlas (CGGA) datasets. Moreover, biological significance of the models was investigated by survival analysis and feature interpretation. Results: The mean AUC (Area under the ROC Curve) [95% confidence interval] of our models across nine cancer types was 0.702 [0.691-0.713], which improved by 0.166 on average over other the four machine learning methods on two training schemes. Our models performed significantly better (p < 0.05) in seven cancer types and performed comparable to the other predictors in the rest of two cancer types. The more pan-cancer samples were used to transfer meta-knowledge, the greater the performance improved (p < 0.05). The predicted response scores that our models generated were negatively correlated with cell radiosensitivity index in four cancer types (p < 0.05), while not statistically significant in the other three cancer types. Moreover, the predicted response scores were shown to be prognostic factors in seven cancer types and eight potential radiosensitivity-related genes were identified. Conclusions: For the first time, we established the meta-learning approach to improving individual radiation response prediction by transferring common knowledge from pan-cancer data with MAML framework. The results demonstrated the superiority, generalizability, and biological significance of our approach.

Machine Learning & Molecular Radiation Tumor Biomarkers

A Radiomic Approach to Access Tumor Immune Status by CD8<SUP>+</SUP>TRMs on Surgically Resected Non-Small-Cell Lung Cancer

Machine learning applications in radiation oncology

Genomics models in radiotherapy: from mechanistic to machine learning

Machine Learning Applications In Radiation Therapy

Machine Learning in Lung Cancer Radiomics

Exploring non-invasive precision treatment in non-small cell lung cancer patients through deep learning radiomics across imaging features and molecular phenotypes

Machine and deep learning methods for radiomics

Machine learning in image‐based outcome prediction after radiotherapy: A review

Application of machine learning for high-throughput tumor marker screening

Machine learning and glioma imaging biomarkers

Applications and limitations of machine learning in radiation oncology

Radiomics and Machine Learning in Brain Tumors and Their Habitat: A Systematic Review

From patterns to patients: Advances in clinical machine learning for cancer diagnosis, prognosis, and treatment

Towards computational fluorescence microscopy: Machine learning-based integrated prediction of morphological and molecular tumor profiles

Machine Learning-Assisted Decision Making in Orthopaedic Oncology

Machine Learning for the Prediction of Organ Damage from Cancer Radiotherapy

Machine learning analysis reveals tumor stiffness and hypoperfusion as biomarkers predictive of cancer treatment efficacy

Detecting Blood Methylation Signatures in Response to Childhood Cancer Radiotherapy via Machine Learning Methods

A meta-learning approach to improving radiation response prediction in cancers

The Applications of Radiomics in Precision Diagnosis and Treatment of Oncology: Opportunities and Challenges