Deciphering and predicting the evolution of cancer cell populations

ended

The fundamental evolutionary nature of cancer is well recognized but an understanding of the dynamic evolutionary changes occurring throughout a tumour’s lifetime and their clinical implications is in its infancy. Current approaches to reveal cancer evolution by sequencing of multiple biopsies remain of limited use in the clinic due to sample access problems in multi-metastatic disease. Circulating tumour DNA (ctDNA) is thought to comprehensively sample subclones across metastatic sites. However, available technologies either have high sensitivity but are restricted to the analysis of small gene panels or they allow sequencing of large target regions such as exomes but with too limited sensitivity to detect rare subclones. We developed a novel error corrected sequencing technology that will be applied to perform deep exome sequencing on longitudinal ctDNA samples from highly heterogeneous metastatic gastro-oesophageal carcinomas. This will track the evolution of the entire cancer population over the lifetime of these tumours, from metastatic disease over drug therapy to end-stage disease and enable ground breaking insights into cancer population evolution rules and mechanisms. Specifically, we will: 1. Define the genomic landscape and drivers of metastatic and end stage disease. 2. Understand the rules of cancer evolutionary dynamics of entire cancer cell populations. 3. Predict cancer evolution and define the limits of predictability. 4. Rapidly identify drug resistance mechanisms to chemo- and immunotherapy based on signals of Darwinian selection such as parallel and convergent evolution. Our sequencing technology and analysis framework will also transform the way cancer evolution metrics can be accessed and interpreted in the clinic which will have major impacts, ranging from better biomarkers to predict cancer evolution to the identification of drug targets that drive disease progression and therapy resistance.