gms | German Medical Science

50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds)
12. Jahrestagung der Deutschen Arbeitsgemeinschaft für Epidemiologie (dae)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie
Deutsche Arbeitsgemeinschaft für Epidemiologie

12. bis 15.09.2005, Freiburg im Breisgau

The design of phase III trials investigating simultaneously a set of targeted therapies with different targets

Meeting Abstract

Search Medline for

  • Werner Vach - Syddansk Universitet, Odense M
  • Rene dePont Christensen - Syddansk Universitet, Odense M

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. Deutsche Arbeitsgemeinschaft für Epidemiologie. 50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 12. Jahrestagung der Deutschen Arbeitsgemeinschaft für Epidemiologie. Freiburg im Breisgau, 12.-15.09.2005. Düsseldorf, Köln: German Medical Science; 2005. Doc05gmds190

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/meetings/gmds2005/05gmds257.shtml

Published: September 8, 2005

© 2005 Vach et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Background

Targeted treatments are a recent, promising development in cancer treatment research. Such treatments try to take individual characteristics into account, typically the presence of a genetic or molecular marker, which we can measure with more or less accuracy by some biochemical method. Hence for each targeted treatment we have a corresponding, empirical condition, where presence of this condition suggests to apply the specific, targeted therapy. However, as any other innovative treatment, targeted therapies have to demonstrate their superiority to the conventional standard treatment in clinical trials, before they can be introduced as regular treatments.

The design of such trials is a new challenge, as they differ substantially from conventional trials. The basic principle will be of course, that patients satisfying the condition for a targeted therapy are randomised to the targeted therapy or to the standard treatment. However, how should we handle the patients satisfying several conditions? This is the basic question we will consider in our contribution, and will show how we can make efficient use of these patients. We will take the viewpoint that we are mainly interested in evaluating the single therapies, and not in evaluating combinations of therapies, although we will partially allow that several targeted therapies are administered in one patient. We further take the viewpoint that we would like to evaluate a targeted therapy by a simple comparison of subjects treated with the targeted therapy and subjects treated with the standard therapyamong subjects satisfying the corresponding condition.

The fundamental problem we have to handle is that the conditions can be associate with the prognoses of the patient. Hence for a fair comparison of a targeted therapy with the standard therapy we have to ensure, that the presence of the conditions corresponding to the other targeted therapies are balanced in the two groups. If we further allow the administration of multiple targeted therapies in a patient, we have also to ensure that the alternative targeted therapies are balanced between the two groups. And of course any combination of conditions and/or therapies has to be balanced, too.

Results

The most simple idea to use a patient satisfying several conditions is to make first a random decision, for which therapy the patient should contribute in the evaluation, and then to randomise the patient between the standard therapy and the therapy selected. This way we ensure that each patient makes a contribution. However, each patient contributes to only one evaluation.

One can easly show that one can improve the efficiency by randomising a patient satisfying several conditions with equal probability to any of the therapies corresponding to his/her conditions or to the standard therapy, as we can use now the patients receiving standard therapy as a control group for each evaluation.

We will show, that a further, very substantial improvement is possible, if one allows to randomise patients to two therapies, and discuss the magnitude of this improvement at hand of some examples.

We further dicuss, how one can evaluate the overall effect of all therapies using the designs suggested.