Dose‐escalation strategies using subgroup information

News

  • Author: Amy Cotterill and Thomas Jaki
  • Date: 17 May 2019
  • Copyright: Image copyright of Patrick Rhodes

Dose‐escalation trials commonly assume a homogeneous trial population to identify a single recommended dose of the experimental treatment for use in future trials. Wrongly assuming a homogeneous population can lead to a diluted treatment effect. Equally, exclusion of a subgroup that could in fact benefit from the treatment can cause a beneficial treatment effect to be missed. Accounting for a potential subgroup effect (ie, difference in reaction to the treatment between subgroups) in dose‐escalation can increase the chance of finding the treatment to be efficacious in a larger patient population.

In a paper published in Pharmaceutical Statistics which is available via Open Access, a standard Bayesian model‐based method of dose‐escalation is extended to account for a subgroup effect by including covariates for subgroup membership in the dose‐toxicity model.

The paper is available via the link here and the authors explain their findings in further detail below:

Dose-escalation strategies which use subgroup information

Amy Cotterill and Thomas Jaki

Pharmaceutical Statistics, Volume 17, Issue 5, September/October 2018, pages 414-436

thumbnail image: Dose‐escalation strategies using subgroup information

Dose-escalation trials commonly assume that all patients eligible to receive an experimental treatment will have the same reaction to it in terms of beneficial response and toxic side-effects. This is the assumption of a homogeneous trial population. The aim of a dose-escalation trial in a homogeneous trial population is therefore to identify a single optimum dose of an experimental treatment for use in future trials.

Wrongly assuming a homogeneous population can lead to a diluted treatment effect. Equally, exclusion of a subgroup that could in fact benefit from the treatment can cause a beneficial treatment effect to be missed. Accounting for a potential difference in reaction to the experimental treatment between subgroups (i.e. subgroup effect) in dose-escalation can increase the chance of finding the treatment to be safe and effective in a larger patient population.

Subgroups which may react differently to treatment can be identified prior to the trial using results of previous similar studies. For example, a study of a similar drug in the same application, or of the same drug in a different disease, may highlight a potential subgroup effect. In this paper we consider the case of two pre-determined subgroups where a characteristic such as ethnicity, pre-treatment or presence of a specific genetic mutation is used to decide subgroup membership. Patients with the characteristic are classed as biomarker positive, and those without are classed as biomarker negative.

A simple design which accounts for subgroup membership would be to run separate dose-escalation trials in each subgroup. This would result in an optimum dose being identified in each subgroup. Although this design performs well, it uses available data inefficiently and makes no conclusions about the presence of a subgroup effect. A hypothesis test could potentially rectify this problem but the small sample number of patients usually used in dose-escalation trials would result in a low powered test.

We propose an alternative design which continually assesses the presence of a subgroup effect. This enables efficient use of the available trial data throughout escalation and in identifying the optimum dose(s). A simulation study, based on real trial data, was carried out and this design was found to be both promising and feasible.

Related Topics

Related Publications

Related Content

Site Footer

Address:

This website is provided by John Wiley & Sons Limited, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ (Company No: 00641132, VAT No: 376766987)

Published features on StatisticsViews.com are checked for statistical accuracy by a panel from the European Network for Business and Industrial Statistics (ENBIS)   to whom Wiley and StatisticsViews.com express their gratitude. This panel are: Ron Kenett, David Steinberg, Shirley Coleman, Irena Ograjenšek, Fabrizio Ruggeri, Rainer Göb, Philippe Castagliola, Xavier Tort-Martorell, Bart De Ketelaere, Antonio Pievatolo, Martina Vandebroek, Lance Mitchell, Gilbert Saporta, Helmut Waldl and Stelios Psarakis.