Abstract

Abstract Glioblastoma (GB) is the most common malignant primary brain tumor in adults and disproportionately contributes to cancer mortality and morbidity. It has been repeatedly hypothesized that GB patients with varying response to treatment and hence prognosis have differential gene or protein expression profiles. To evaluate proteomic differences between long term and short term survivors with glioblastoma, an unbiased, label-free, unfractionated LC-MS/MS methodology was used on snap-frozen tumor samples from 10 patients with short term survival and 6 patients with long term survival from University Hospitals Case Medical Center. Short term survival was defined as 18 months after diagnosis. Significant protein intensity differences between long and short term survivors were tested using a mixed model. For each protein, the survival group was set as a fixed effect; while the peptide was set as a random effect which allowed for the within-protein correlation of the peptides to be accounted for in the model. In total 1490 proteins were detected in the glioblastoma sample set. One hundred eighty-three (183) proteins were found to be significant between survival groups using an adaptive False Discovery Rate p-value <0.05. Isocitrate Dehydrogenase 1 (IDH1), a mitochondrial energy-production enzyme, which has been implicated in differential response to temozolomide, was among these significant proteins. The top canonical pathways using these top 183 proteins, as generated by Ingenuity Pathway Analysis (IPA), were oxidative phosphorylation and mitochondrial dysfunction. These two pathways were at least twice as statistically significant as compared to any other discovered pathways. Examination of the individual proteins revealed a significant down regulation of oxidative phosphorylation, including subunits of cytochrome-oxidase and the BC1 complex in short term survivors. In addition, matched formalin-fixed paraffin embedded (FFPE) tumor tissue was used in order to verify the protein targets. Using 2 slides of 6 microns each, sufficient protein over 1 microgram of protein was extracted in order to perform an unbiased label-free LC-MS/MS analysis and 930 total proteins were detected. One hundred twenty-six (126) of the 183 significant proteins were successfully identified in the matched FFPE samples; thus FFPE can be used to verify many of the protein targets predicted to be changing. Further validation of the proteomics signature is ongoing in an expanded, independent cohort of glioblastoma patients with targeted validation. Progress in treatment for glioblastoma patients has been slowed by the absence of biomarkers to define treatment response and/or stratify patient sub-groups and help guide physician treatment. Proteomics analysis provides the most direct readout of the functional state of the tumor and thus promises improved functional annotation of patient status. 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 1279. doi:1538-7445.AM2012-1279

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call