Refined adaptive design for D-optimum dose-finding in early phase clinical trials
Dariusz Uciński (University of Zielona Gora)
Friday 9th November, 2012 15:00-16:00 Maths 203
An adaptive procedure for dose finding in clinical trials when both efficacy and
toxicity responses are available is proposed. The adopted approach incorporates both toxicity and efficacy data into the estimation
of the biologically optimal dose of an agent in a phase I trial. Dose selection is
based on all patients treated up until that time point, using a continual reassessment method approach.
Based on analytic formulae for the Fisher information matrix, adaptive D-optimal are then sought subject to the constraint
that the probability of toxicity at a given dose is smaller than a pre-specified value. A novelty here lies
in the fact that patients in each cohort can be treated using different doses, which is in contrast to the common
procedure used in clinical trials. Various aspects of this strategy are investigated, including computational complexity
and the choice of effective algorithms to determine optimal designs.
The search for the optimal solution is performed using the branch-and-bound method in which an extremely simple and efficient technique is employed to produce an upper bound to the maximum objective function. Its idea consists in solving a relaxed problem through the application of a simplicial decomposition algorithm in which the restricted master problem is solved using a multiplicative algorithm for optimal design.
Discussion of extensive simulation results complements the presentation.