Evolutionary dynamics of metastatic cancers

Cancer is a genetic disease characterized by the uncontrolled growth of abnormal cells. Some of the acquired genetic and epigenetic alterations increase the proliferation rate of cells and can lead to the formation of a primary tumor followed by the dissemination of cells to form metastases. Metastases are responsible for 90% of cancer-related deaths, yet the mechanisms by which cancer cells disseminate and eventually colonize foreign organs have remained poorly understood. For not yet fully understood reasons, the probability, the timing, and the target organs of metastases greatly differ across cancer types. Therefore, we propose to study the evolution of metastasis, the final biological stage of cancer, by two complementary approaches: (i) mathematical modeling and (ii) phylogenomics. Various mathematical models have been developed to study the number and size of metastases. However, the timing of metastasis as well as the spread among metastases has not yet been considered in a comprehensive model of cancer evolution. We aim to study metastatic progression within a multi-step model of cancer and will investigate the overall evolutionary dynamics of the primary tumor and its metastases. We will identify the parameter regimes leading to early and late metastasis, calculate the probabilities of metastasis-to-metastasis spread and tumor reseeding, and explore the expected effects on tumor progression and patient survival. Phylogenomic methods utilize the heterogeneity of the acquired genetic alterations among cancer cells to reconstruct the evolutionary history of a patients cancer. Despite the tremendous progress in DNA sequencing of cancer cells to detect these acquired genetic alterations, classical phylogenetic methods struggle with the noisy sequencing data. We propose to develop a new phylogenomic method, Netomics, that allows us to accurately reconstruct metastatic seeding patterns of a cancer with commonly available sequencing technologies. We will apply this new algorithm on new data of sixteen pancreatic cancer patients and four melanoma patients where between 4 and 53 samples of different regions of the primary tumor and distinct metastases have been obtained per patient. Last, we will utilize the developed mathematical models to generate in silico sequencing data of metastases for a variously evolving cancers. These data will enable us to perform an objective benchmarking across existing phylogenomic tools as well as our new tool Netomics. Effective treatment of cancer requires a comprehensive understanding of the evolutionary rules governing metastatic spread. The reconstructed phylogenies will shed new light on the evolution of metastasis and provide the necessary input to develop a general model to analyze and predict metastatic progression. Understanding the different mechanisms of metastatic progression operating in cancers may have significant implications for clinical decision-making and will likely provide predictive value for a patients prognosis as well as for the response to therapy.