Unformatted text preview: and spacing of the dose combinations and so define the set of dose combinations to be used in the trial. This approach can easily be generalized to accommodate either nonlinear search regions or combi- nations of more than two agents. For example, the set S of dose levels to be searched for an MTD could be chosen so that given any two distinct dose combinations in S , one will have the levels of all agents higher than the other. As a result, the combinations in S can be unambiguously ordered and the dose-toxicity relationship can be meaningfully modeled as an increasing function of the distance of each permissible dose combination from the ‘‘ minimum ’’ combination. Since this formulation represents each permissible dose combination by a single real number, the design methods described above for single agent trials can be used to select dose combinations in the multiple agent trial....
View Full Document
- Spring '11
- Physics, Bayesian probability, MTD, dose combinations