Author of

• 34780

  +

  P37 - Pathology - Biomarker Testing (ID 107)

  • Event: WCLC 2020
  • Type: Posters
  • Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
  • Presentations: 1
  • Moderators:
  • Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall

  • 34794

    +

    P37.12 - Concordance of Next-Generation Sequencing Between Tissue and Liquid Biopsies in Non-Small Cell Lung Cancer (ID 1560)

    00:00 - 00:00  |  Presenting Author(s): Fahmin Basher
    • Abstract

    Introduction
    

    While genetic profiling has become standard of care for patients diagnosed with non-small cell lung cancer (NSCLC), next-generation sequencing (NGS) provides a wealth of information about targetable mutations. Advances in genetic testing have led to sequencing platforms that utilize tissue itself or extracellular circulating tumor DNA in the blood, known as a “liquid biopsy.”

    Methods
    

    We identified 55 patients with NSCLC who had undergone both tissue and liquid biopsy, using Foundation One and Guardant 360 at the University of Miami / Sylvester Comprehensive Cancer Center between January 2016 and December 2018, and performed retrospective analysis to determine patient characteristics as well concordance between different NGS platforms.

    Results
    

    In our patient population, 34% of patients had never smoked prior to diagnosis, while 22% had more than a 30 pack-year smoking history. 64% of patients had no treatment prior to initial NGS. 40% of patients had both testing done essentially simultaneously, while 60% of patients had one test done after disease progression. Of these patients, therapy was changed as a result in 73%. Median number of days between tests was 21 days, with 56% of testing done within 90 days of the previous testing. Nine patients had an additional Foundation One tissue NGS performed. Concordance across all genes tested in both platforms was 98 ± 0.2%. Concordance with consideration of genetic alterations detected in both assays was 24.5 ± 3.0%. The median number of gene alterations determined by Foundation One testing was 4 (range 1-9), while the median for gene alterations detected by Guardant 360 was 3 (range 1-13). The median number of variants of unknown significance (VUS) was 10 (range 5-25). We also calculated sensitivity and specificity across 8 genes, using tissue-based NGS as the gold standard, as shown in Table 1.

    Table 1. Sensitivity, sensitivity, and diagnostic accuracy across 8 genes in NSCLC.

    Gene	Sensitivity	Specificity	Positive Predictive Value	Negative Predictive Value
    EGFR	78.3%	97.1%	94.7%	86.8%
    TP53	59.4%	84.6%	82.6%	62.%
    ALK	100%	98.1%	66.7%	100%
    KRAS	58.3%	95.5%	77.8%	89.4%
    BRAF	50%	94.2%	40%	96.1%
    ROS	n/a	98.2%	n/a	100%
    RET	100%	98.1%	50%	100%
    MET	100%	92.6%	20%	100%

    Conclusion
    

    Our analysis indicates a role for both tissue-based and circulating tumor DNA-based NGS for determination of targetable mutations and thus appropriate treatment regimens. Low levels of concordance are potentially related to post-treatment changes in the tumor genetic profile as well as evolution in the testing itself.