Gene Signature In Radiation Oncology

Aims: In the era of precision medicine,prognostic markers and therapy-predictive markers are important for improving clinical outcome in cancer.As a high-throughput technology,gene signatures have been used to predict prognosis and treatment response in many cancer types.In our study,we proposed that the gene signature may serve as a prognostic biomarker for cancer and a predictive biomarker for estimating the survival or sides effects of radiotherapy-treated patients.Methods: 1)Lnc RNA expression data of breast cancer patients derived from Gene Expression Omnibus(GEO)dataset was used to analyze the association between lnc RNA signature and clinical survival in the training set.The validation for the association was performed in another three independent testing sets.2)We have tested a 31-gene signature in two previously published glioblastoma multiforme(GBM)datasets from Gene Expression Omnibus(GEO)and The Cancer Genome Atlas(TCGA).Patients were stratified into radiosensitive(RS)and radioresistant(RR)groups according to the above-mentioned 31-gene signature.The Kaplan-Meier method was used to estimate survival time and differences in survival times were then compared in each set.3)From June to September 2014,19 patients with T1c-T3 a,histologically proven prostate cancer patients were enrolled in the clinical trial of carbon ion radiation therapy(CIRT).RNA-seq was performed using peripheral blood lymphocytes-derived RNA obtained before and after CIRT.The end point was defined as the incidence of acute and late adverse events that were evaluated based on the Common Terminology Criteria for Adverse Events(CTCAE)version 4.03.We performed Boruta algorithm to identify feature genes related to carbon ion induced toxicity.Based on the selected gene sets,multiple classification models were applied for predicting a patient's likelihood of developing adverse effects.Results: 1)A set of four long non-coding RNA(lnc RNA)genes have been identified to predict breast cancer patients' survival.2)Radiotherapy-treated GBM patients assigned to RS group had an improved overall survival compared with RR group.Gene set enrichment analysis(GSEA)revealed that epithelial mesenchymal transition(EMT)pathway was enriched in the radioresistant subgroup.3)A set of 7 genes was identified capable of differentiating between patients with and without radiation induced adverse effects.Predictive capabilities were obtained using different models for predicting new sample toxicity in training and testing sets.Based on the model evaluation,a 7-gene signature was established with two optimal classifiers(support vector machine and random forest).Conclusions: These results suggest that gene signature is a predictive biomarker for cancer patients' prognosis and radiotherapy-treated patients' prognosis or side effects.