and patient might benefit if updated assessments of the risk of toxicitywere available during the trial. Both of these needs can be addressedwithin a Bayesian framework. In Sections 2.1 through 2.5 we present adescription of selected Bayesian procedures developed for the specificcase where toxicity is assessed on a binary scale (presence or absence ofDLT), only a single agent is under investigation (the levels of any otheragents being fixed) and no relevant pretreatment covariate informationis available to tailor the dosing scheme to individual patient needs.We discuss extensions and modifications of the selected methods inSection 3.2.1.Formulation of the ProblemDose level will be represented by the random variableXwhose realizationis denoted byx. For notational compactness, the same variable will beused for any formulation of dosage deemed appropriate. Thus, forexample,Xmay represent some target drug exposure (e.g., AUC), thephysical amount of agent in appropriate units (e.g., mg/m2), or the
This is the end of the preview.
access the rest of the document.