Modeling Neoplastic Progression in Barrett's Esophagus - Renewal -2

Patients with Barrett's Esophagus are at increased risk of developing esophageal adenocarcinoma, but there is currently no reliable way to determine who is at high versus low risk. We propose to quantify the evolutionary dynamics in Barrett's and determine if those measures predict cancer risk. Neoplastic progression is fundamentally a stochastic evolutionary process at the cell level. However, there are often many different genetic and epigenetic alterations that can contribute to the development of a cancer, and the ones that evolve in any given cancer are random. This makes it difficult to assess the risk that a pre-cancerous tissue will evolve into cancer based simply on what mutations have occurred previously. We propose a different approach: Measure the dynamics of cell level evolution rather than the products of that evolution. We hypothesize that the rate of cellular evolution in a tissue should predict how likely and how quickly that tissue will evolve into a cancer. We will test this hypothesis in Barrett’s esophagus. We propose to fit both population genetic and phylogenetic models to data from Barrett’s esophagus biopsies spread over space and time in order to estimate parameters of the rate of cellular evolution and test if they are predictive of progression to esophageal adenocarcinoma. Five parameters determine the rate of evolution: 1) stem cell population size, 2) stem cell generation time, 3) mutation rate, 4) the strength of selection, and 5) heritability of the alterations. We have recently shown that we can measure all those parameters in clinical biopsies. We will measure all five parameters in Barrett’s esophagus based on CpG methylation and copy number alterations using methylation arrays, as well as single nucleotide variants identified in whole genome sequencing, in a prospective cohort of Barrett’s patients having either stable, benign disease or validated cancer outcomes. We will use those measures of the rate of evolution in Barrett’s esophagus to derive multivariable risk models of progression to esophageal adenocarcinoma as well as a classification system for the types of evolution present, that are clinically useful for the management of Barrett’s esophagus. If we are successful, we will reduce morbidity in patients likely to progress through earlier intervention, reduce suffering and stress in low-risk patients, and focus resources and interventions on those high-risk patients who will benefit from them. Our proof of principle would also justify using our tools to measure evolution in other pre-cancers to predict progression. In full-blown cancers, our measures may also predict their response to therapy and to select therapies that are matched to the evolutionary dynamics of a patient’s cancer.