Identification of predictive drug response signatures by whole genome profiling of lung tumors

Non-small cell lung cancer (NSCLC) accounts for an estimated 80% of known lung cancer cases. In many instances, these tumors are inoperable by the time of diagnosis, leaving chemotherapy as the main option for treatment. Unfortunately, tumor response to chemotherapy can vary and the presence of even a few drug-resistant cells within a tumor may result in disease recurrence as these cells expand to form a new tumor mass. Using archived NSCLC tumour samples, Timon Buys is working to identify genomic “signatures” that might predict how well or poorly a drug will act to destroy tumour tissues. First, he is using a technology called array comparative genomic hybridization, which allows researchers to assess changes in gene copy number throughout the whole human genome. Second, isolated patient cancer tissue will be grown and treated in mice to preview how a specific case will behave after treatment with chemotherapy drugs. Information from these studies could point to ways to effectively eliminate tumor recurrence due to drug resistance, helping to improve the prospects of patient survival.