Bayesian framework for tumor board decision making

Traditionally, the specific treatment for a cancer patient is decided by a multidisciplinary tumor board, which integrates prior clinical experience, published data, and patient-specific factors to develop a consensus on an optimal therapeutic strategy. However, tumor boards often encounter patients who incompletely match extant data and for whom several treatment options must be evaluated based on imprecise criteria. We propose optimizing treatment outcomes will require a flexible but rigorous mathematical tool that can define the probability of success of given therapies. Here, we propose a Bayesian approach to tumor forecasting using a multi-model framework that can predict response to different targeted therapies within individual patients. By exploiting the integrative power of the Bayesian decision theory, we demonstrate multiple therapeutic options can be simultaneously examined so that the resulting clinical course can be forecasted. From this, we detail a general decisional methodology built upon a robust and well-established mathematical framework that can support the clinical decision process for individual patients within a clinical tumor board.