Author of

• 11273

  +

  MO05 - Prognostic and Predictive Biomarkers II (ID 95)

  • Event: WCLC 2013
  • Type: Mini Oral Abstract Session
  • Track: Medical Oncology
  • Presentations: 1
  • Moderators:J. Hu, S. O'Toole
  • Coordinates: 10/28/2013, 16:15 - 17:45, Parkside Auditorium, Level 1

  • 11274

    +

    MO05.01 - Validation of gene expression biomarker panels in non-small cell lung cancer (ID 2928)

    16:15 - 17:45  |  Author(s): M. Huebner
    • Abstract
    • Presentation
    • Slides

    Background
    Many studies in the literature have suggested that gene expression biomarkers may guide patient classification and clinical management in NSCLC. Despite minimal external validation and no clinical trial evidence, gene expression biomarker panels have been proposed as tools for making treatment decisions. Recent controversy surrounding the validity of such data and its potential applicability to clinical practice led us to perform an external validation study of published gene expression biomarker panels.

    Methods
    We performed gene expression profiling for a total of 209 patients with both Affymetrix whole transcriptome U133Plus2 arrays in addition to a NSCLC-specific array constructed by our group for assessment of mRNA expression in frozen tumor specimens of NSCLC. Clinical outcome data were collected and analyzed for correlations of gene expression with disease-free and overall survival. Cox proportional hazard models were used to test significance of individual genes and for gene sets defined by each panel. Panels tested included those previously published from Michigan, Mayo Clinic, Taiwan, Toronto, and UCSF.

    Results
    Expression profiling data were generated for a total of 209 patients with NSCLC. This included U133Plus2 arrays of 242 tumor samples and 105 matched surrounding normal lung tissue, as well as 111 tumor profiles using the NSCLC-specific array. There were 98 women and 111 men in the study cohort, with 120 patients having Stage I NSCLC (57.4%), 38 with Stage II (18.2%), 50 with Stage III (23.9%), and one patient with Stage IV disease (0.5%). Mean follow-up time after surgical resection was 62.4 ± 48 months. Seventy-four patients (35.1%) developed post-resection recurrence after a mean of 53.3 ± 49.3 months, of which 62 patients died (83.8%). Known clinical predictors such as TNM stage, histology, and tumor grade were predictive of survival. Although many genes within the published biomarker panels were significantly correlated with disease-free and overall survival, none provided additive prognostic value beyond standard clinical predictors.

    Conclusion
    Although a number of individual gene expression biomarkers have prognostic significance in univariate models, published biomarker panels perform poorly in external validation studies such as this. The additive prognostic value beyond standard, known clinical predictors in the TNM staging system casts doubt as to whether such information will be useful in clinical practice. Despite the success of gene expression biomarkers for molecular subtyping in other cancers, our data suggests that this information has a low likelihood of clinical translation in NSCLC for unselected patients.

    Only 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, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

    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.