Skip to main content
  • Consolidating and Functionalizing a Decade of Work – A Retrospective Approach to Classification of Malignant Gliomas Based Solely on Gene Expression Profiling

    Final Number:

    Haya Sarras; Megan Wu MSc; Angela Celebre; Daniele Merico; Jason Karamchandani; Sunit Das MD, PhD

    Study Design:
    Laboratory Investigation

    Subject Category:

    Meeting: Congress of Neurological Surgeons 2015 Annual Meeting

    Introduction: Glioblastoma is the most common and deadly type of brain cancer. Over the past decade, several genetic alternations and actionable targets have been implicated in the initiation, progression and clinical outcome of the disease. As global scientific initiatives continue generating impactful data, there is a pressing need for the development of new tools that will maximize the use of precious clinical samples with limited quantity and/or quality in basic and clinical research

    Methods: We used PrimePCR validated assays to generate a custom real-time PCR screening panel, containing all 74 previously published mRNA targets showing gene expression changes in glioblastoma, and 5 house-keeping genes. A cohort of 21 frozen brain biopsies diagnosed as: 4 normal, 4 oligodendoglioma, 1 diffuse astrocytoma, 1 G-III glioma, and 11 G-IV glioblastoma were used in the analysis. We performed RNA extraction, followed by cDNA synthesis, multiplexed pre-amplification and SYBR-based qPCR to generate expression profiles on all samples.

    Results: We demonstrated that our workflow overcomes some of the key technical challenges currently experienced in glioblastoma research. The workflow showed high tolerance to variation in RNA quality (RIN 8.5-4) and high sensitivity in detection. cDNA input that is equivalent to 3 ng of starting RNA was sufficient to conduct accurate analysis of the entire panel of assays. Using Principal Component Analysis (PCA), we were able to accurately separate G-IV glioblastoma from low grade glioma. We were also able to observe the progression of G-III glioma away from the low grade profile and closer to G-IV clustering, thus confirming the validity of the workflow and analysis method. We also demonstrated the robustness of this analysis method by using the PCA rotation matrix obtained after removing 25% of the samples, which still lead to correct sample clustering.

    Conclusions: This is the first study to consolidate high-throughput data into a single functional panel capable of accurately classifying glioblastoma biopsies based solely on gene expression profiling

    Patient Care: There is a growing need for the development of tools that will maximize the potential use of limited and sub-optimal clinical biopsies to make informative decision at the clinical level.Currently a large fraction of frozen biopsies are generally disqualified from high throughput analysis due to limited quantity and low quality of RNA samples. Such technical challenge limits the implementation of personalized genetic screening at the clinical level. We have shown that data output from high throughput efforts can be consolidated into simples real-time PCR panels for quick laboratory genetic screening to assist with diagnosis and inform treatment decisions from limited and/or sub-optimal RNA samples. The workflow makes use of new Pre-amplification chemistry and validated assays to improve the utility of such samples in clinical research.

    Learning Objectives: To examine if data from retrospective high-throughput genomic analysis over the past decade can be converted to simple real-time pcr panel capable of accurately classifying malignant glioma biopies with limited input or quality.


We use cookies to improve the performance of our site, to analyze the traffic to our site, and to personalize your experience of the site. You can control cookies through your browser settings. Please find more information on the cookies used on our site. Privacy Policy