1. Abstract 1210: Periostin expression in glioma correlates with genes related to mesenchymal transition and survival
- Author
-
Andrew D. Trister, Jason K. Rockhill, Andrei M. Mikheev, Stephen H. Friend, and Robert C. Rostomily
- Subjects
Oncology ,Cancer Research ,medicine.medical_specialty ,Pathology ,business.industry ,Matricellular protein ,Astrocytoma ,Periostin ,Ductal carcinoma ,medicine.disease ,Breast cancer ,Internal medicine ,Glioma ,Gene expression ,medicine ,business ,Survival analysis - Abstract
Background: Gliomas are the most common primary brain tumors and their prognosis is related to WHO histopathological grade. Glioblastoma (grade IV astrocytoma or GBM) is characterized by microscopic invasion into surrounding brain and universally poor prognosis despite treatment with surgery, radiation and chemotherapy . Recently, the matricellular protein periostin (POSTN) has been shown to be associated with increased parenchymal invasion (measured by edema on MRI) and poor prognosis in GBM. The specific aim of the present analysis was to determine the mechanistic impact of POSTN on glioma outcome. Methods: We used the gene expression of POSTN from 559 GBM patients (pts) included in the Cancer Genome Atlas (TCGA) to build a gene expression model of the expression of 12184 genes to predict POSTN expression using elastic net. This model was used to predict the POSTN expression in a testing set of 419 pts with gliomas included in Repository for Brain Neoplasia Data (REMBRANDT) (99 grade II, 71 grade III and 125 grade IV and 124 with no grade). Receiver-operating characteristic (ROC) and survival analysis were performed to measure the performance of the model, and gene-set enrichment analysis (GSEA) was used to reveal network topology perturbations. Results: The gene expression model discovered 721 genes highly correlated to POSTN expression that predict POSTN in the testing set with an area under the curve (AUC) of 0.96 on ROC curve. GSEA reveals genes involved in “mesenchymal transition,” transition of invasive ductal carcinoma (IDC) from ductal carcinoma in situ (DCIS) in breast cancer and stem cell signatures. Survival analysis of the TCGA GBM pts showed that the cohort with “high POSTN” had worse survival (median 12 months versus 15 months, log-rank p=0.0002). When applied to all pts in REMBRANDT, the model predicted classes also had significantly different survival (median 13.4 versus 38 months, log-rank p Conclusions: We have developed a gene expression model related to POSTN, a gene linked to poor prognosis in GBM, to investigate the role correlated genes may play in the aggressive phenotype. Some of the genes found to be highly correlated to POSTN are related to mesenchymal transition, invasive behavior in breast cancer and stemness. We verify that high POSTN expression is a strong prognostic indicator for poor outcome in GBM, and reveal for the first time that pts with grade II and III glioma with a “high” POSTN signature have significantly worse survival. Given that low grade glioma pts often have less aggressive treatment at time of diagnosis, we propose studying the role of early chemoradiation in the subset of pts with poor POSTN signature to potentially improve their outcome. Citation Format: Andrew D. Trister, Andrei M. Mikheev, Jason K. Rockhill, Stephen H. Friend, Robert C. Rostomily. Periostin expression in glioma correlates with genes related to mesenchymal transition and survival. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 1210. doi:10.1158/1538-7445.AM2013-1210
- Published
- 2013
- Full Text
- View/download PDF