Search
  • Dennis Wang

New gene panel to predict prognosis in lung cancer

Updated: Nov 17, 2018

We developed a 865 gene panel measuring mutation burden and neoantigens for identifying subgroups of early-stage lung cancer patients with different prognosis.


Background: Genomic profiling of patient tumors has linked somatic driver mutations to survival outcomes of non-small cell lung cancer (NSCLC) patients, especially for those receiving targeted therapies. However, it remains unclear whether specific non-driver mutations have any prognostic utility. Methods: Whole exomes and transcriptomes were measured from NSCLC xenograft models of patients with diverse clinical outcomes. Penalised regression analysis was performed to identify a set of 865 genes associated with patient survival. The number of somatic copy number aberrations, point mutations and associated expression changes within the 865 genes were used to stratify independent NSCLC patient populations, filtered for chemotherapy naive and early-stage. In-depth genomic analysis and functional testing was conducted on the genomic alterations to understand their effect on improving survival. Results: High burden of somatic alterations are associated with longer disease-free survival (HR=0.153, P=1.48x10-4) in NSCLC patients. When somatic alterations burden was integrated with gene expression, we were able to predict prognosis in three independent patient datasets. Patients with high alteration burden could be further stratified based on the presence of immunogenic mutations, revealing another subgroup of patients with even better prognosis (85% with >5 years survival), and associated with cytotoxic T-cell expression. In addition, 95% of these 865 genes lack documented activity relevant to cancer, but are in pathways regulating cell proliferation, motility and immune response were implicated. Conclusion: Our results demonstrate that non-driver somatic alterations may influence the outcome of cancer patients by increasing beneficial immune response and inhibiting processes associated to tumorigenesis. Read more here.


0 views

The University of Sheffield

 © Dennis Wang, University of Sheffield

genomic medicine | personalized medicine |bioinformatics | computational biology | healthcare AI

  • Twitter Clean Grey
  • LinkedIn Clean Grey