| Summary: | Background: Lung cancer is the world's leading cause of cancer death, with a 5 year
survival rate of ~16%. Several factors contribute to this poor prognosis: the limited detection of disease at treatable stages, the high metastatic potential of any primary tumour, and the variable effectiveness of chemotherapy. We applied high resolution whole genome profiling technologies to uncover genes associated with specific lung cancer phenotypes and to delineate clonal relationships between tumours.
Hypotheses: (i) Shared genetic features in tumours from the same patient are evidence
of a common progenitor. (ii) Continuing clonal evolution facilitates selection for
resistance genes during drug exposure. (iii) Drug response can be predicted for pretreatment lung cancer by evaluating specific gene changes.
Materials/Methods: DNA alteration data from non-small cell lung cancers (NSCLC) were
integrated with mRNA/protein expression data to identify genes contributing to
tumourigenesis. Fine-mapped DNA alteration boundaries were used to evaluate clonality, discriminating multiple primary tumours from intrapulmonary metastasis. Subsequently, this approach was applied to define chemoresistance gene candidates in cells grown under drug selection. Genome alteration data for early stage lung tumours
were also analyzed to define gene changes driving post-treatment recurrence in patients.
Results: We optimized collection of and genomic analysis for clinical lung cancer,
identifying novel oncogene candidates (including genes contributing to tumour
invasion). In addition, we successfully used DNA alteration boundaries to discriminate
clonally-related tumours and define ongoing clonal evolution in both tumours and cancer
cell lines, providing evidence in support of our first two hypotheses. We also identified
dysregulated genes and gene pathways associated with post-treatment recurrence for
clinical lung cancer. These last data suggest that chemoresistance may be an intrinsic process for the majority of cells in a pre-treatment tumour, lending support to our third hypothesis. Significantly, we also detected distinct recurrence-associated gene changes within tumour histology subgroups, indicating that NSCLC may not be treatable as a single disease entity.
Conclusions: Global analysis of DNA alterations is an effective means for defining clonal relationships between tumours. Further, tumour phenotypes such as chemoresistance are governed by complex activation of a variety of gene networks.
|
|---|
