Author of

• 25967

  +

  P3.13 - Targeted Therapy (Not CME Accredited Session) (ID 979)

  • Event: WCLC 2018
  • Type: Poster Viewing in the Exhibit Hall
  • Track:
  • Presentations: 1
  • Moderators:
  • Coordinates: 9/26/2018, 12:00 - 13:30, Exhibit Hall

  • 27689

    +

    P3.13-25 - Development of a Comprehensive Genomic Profiling System to Detect Actionable Genetic Alterations and Tumor Mutation Burden   (ID 14000)

    12:00 - 13:30  |  Author(s): John Simmons
    • Abstract
    • Slides

    Background
    

    Current non-small-cell lung cancer (NSCLC) treatment includes targeted therapies for EGFR, ROS1, BRAF and ALK. Patients treated with immunotherapies have shown both high and durable response rates. Tumor mutational burden (TMB) has emerged as a promising biomarker, with multiple clinical studies correlating high TMB with better response to immunotherapy. However, in order for widespread adoption to be successful, the community needs more clarity and guidance around TMB analytical performance before implementation into routine diagnostic practice. First, analytical accuracy across a broad dynamic range of mutation burden is critical. Second, comparison to a common reference and standardization of reporting TMB is also required for clinicians to translate results across different technologies and methods. Here, we describe the development and analytical validation of a decentralized comprehensive gene profiling kit to determine TMB and other clinically relevant somatic variants from a 507-gene next-generation sequencing panel.

    a9ded1e5ce5d75814730bb4caaf49419 Method
    

    Identification of somatic mutations with sufficient accuracy is critical for inferring mutational load from a targeted gene panel. We developed a machine learning algorithm to optimize sensitivity and specificity for somatic mutation detection across the entire coding sequence of a 507-gene panel. We compared the overall accuracy of this approach using simulated and experimentally validated whole-exome and targeted gene analyses. We then assessed the accuracy of the 507 gene panel for prediction of TMB in silico across a wide dynamic range using several cohorts of publicly available NSCLC whole exome sequencing datasets. Then, cases from an independent NSCLC cohort were experimentally evaluated and compared to whole exome sequencing results to demonstrate the analytical accuracy across a spectrum of tumor mutational loads. Finally, sensitivity and specificity of clinically relevant genetic alterations in NSCLC were determined using both contrived and clinical samples compared to orthogonal methods.

    4c3880bb027f159e801041b1021e88e8 Result
    

    Our results indicate that the 507-gene system, combined with proprietary software leveraging machine learning algorithms, accurately identifies somatic mutations, and achieves high analytical accuracy for determining TMB across a broad dynamic range. In addition, this system has the capability to detect somatic genetic variants that are important for therapeutic stratification in NSCLC.

    8eea62084ca7e541d918e823422bd82e Conclusion
    

    This 507-gene decentralized system provides accurate results for the determination of TMB across a broad range of mutational burden, and detects clinically relevant genetic alterations important in NSCLC.

    6f8b794f3246b0c1e1780bb4d4d5dc53

    Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.