Identifying Novel Biomarkers And Predicting Response Of Rectal Cancer To Neoadjuvant Chemoradiotherapy By Microarray Gene Expression Profiling

Identifying novel biomarkers and predicting response of rectal cancer to neoadjuvant chemoradiotherapy by microarray gene expression profilingPurpose:Neoadjuvant chemoradiotherapy was widely used to improve local control and sphincter preservation in patients with locally advanced rectal cancer.But the spectrum of tumor responsiveness of rectal cancer to neoadjuvant chemoradiotherapy included complete response to complete resistance.The aim of study here was to explore the gene expression of locally advanced middle/lower rectal cancer and identify novel biomarkers that can be used for characterization and prediction of response to neoadjuvant chemoradiotherapy using microarray gene expression profiling.Methods:Pretherapeutic tumorous tissues were obtained from 29 patients with locally advanced middle/lower rectal cancer.All patients received preoperative radiation accompanied by Oxaliplatin and Capecitabine.Standardized surgery with total mesorectal excision was performed after an interval of approximately 6 weeks.The histologic work-up of surgical specimens was carried out according to International Union Against Cancer(UICC) standards.Responsiveness of patients was evaluated as T-level downstaging of the primary tumor by comparing the clinical T category (determined by rectal Magnetic Resonance Imaging) with histologic T category.Gene expression profiling of rectal cancer were obtained by Human Genome GeneChip Plus U133 2.0 Array.By using Significance Analysis of Microarrays(SAM),a set of discriminating genes was identified between responders and non-responders. Hierachical Clustering Analysis was applied to predict the response of the rectal cancer.Gene Ontology and DAVID 2008 Functional Annotation Bioinformatics Microarray Analysis tools were used to gene functional category clustering analysis. As interested novel biomarkers,the CFLAR,SERPINB5,EML2,MUC5,UBR2 was performed further analysis. Results:In 29 patients,responders and non-responders showed significant different expression levels for 146 genes(False Discovery Rate=0.05) by Significance Analysis of Microarrays(SAM).Using this gene set,we were able to establish a new model to predict response to neoadjuvant chemoradiotherapy in rectal cancer with accuracy of 96.55%.Sensitivity(correct prediction of response) was 100%,and specificity (correct prediction of non-response) was 93.8%,with positive and negative predictive value of 92.9%and 100%,respectively.The list of discriminating genes included cellular processes which comprise cell proliferation and differentiaton,metabolism, membrane proteins,nuclear acid interaction,signal transduction.In further study, CFLAR,SERPINB5,EML2,MUC5 and UBR2 showed significant higher expression in non-responders(p＜0.05).Conclusions:The present study suggested the possibility that microarray gene expression profiling may be useful in predicting response of locally advanced rectal cancer to neoadjuvant chemoradiotherapy and helpful in individualized therapy.Global expression profiles of responders and non-responders may provides insight into molecular biologic rules of rectal cancer and explores new therapeutic targets.