Abstract LB-454: Genes involved in cell proliferation are down-regulated by dexamethasone in an in vivo model of glioma and predict survival in the TCGA GBM data set

Abstract

Abstract High-grade gliomas are among the most malignant and deadly cancer types overall. Most patients will die within the first year of diagnosis despite aggressive therapy. Conventional glioma post-surgical therapy consists of dexamethasone (DEX) for symptomatic relief and radiation and temozolomide for antineoplastic therapy. DEX has been used for decades to treat brain tumor-associated neurological symptoms with most patients experiencing a dramatic improvement in symptoms. Some investigators have suggested that DEX can inhibit tumor growth, however, there have been no definitive studies that explain the mechanism of how DEX can exert this effect. Today, the accepted role of DEX in glioma is in the management of peri-tumoral edema, presumably through its anti-inflammatory properties. We have used a proneural mouse model of glioma to study the mechanism of how DEX effects tumor cells. We treated glioma bearing mice with 10mg/kg/day of DEX for 2-3 days and employed the Illumina Mouse ref8 array to identify the top differentially expressed genes (DEGs) between DEX treated and control RNA samples isolated from whole tumors. We found that the majority of the DEGs were down-regulated by DEX and were involved in cell cycle progression and proliferation. This anti-proliferative phenotype was validated by PCNA immunhistochemistry, which showed greater staining in controls compared to DEX treated tumors; and staining was primarily localized to tumor cells. More interestingly, when these genes were tested against the TCGA GBM data set, we found that high expression of our gene set predicted a significantly longer survival. To determine whether the anti-proliferative phenotype induced by DEX was a cell autonomous effect, we treated primary mouse glioma cultures with escalating concentrations of DEX in vitro. In agreement with other reports, we observed an increased proliferation in glioma cultures suggesting that the anti-proliferative phenotype observed in vivo is induced indirectly through non-neoplastic stromal cells. In summary, our data confirm previous reports of an in vivo growth inhibitory effect of DEX and support the hypothesis that DEX acts indirectly on tumor cells to exert this effect. The TCGA analysis suggests that inhibiting tumor cell growth through the genes identified in this study may adversely impact survival. A concerning possibility is that if DEX decreases tumor cell proliferation it may also decrease the efficacy of antineoplastic therapy that is most toxic to proliferating cells. These data underscore the importance of understanding the mechanism of how DEX affects tumor and stromal cells in glioma. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-454. doi:1538-7445.AM2012-LB-454