A New Adaptive Simon-Based Design Focusing on Subpopulation Heterogeneity
|Jacques Medioni1,2*, Caroline Tournoux-Facon3and Yann De4|
|1Department of Medical Oncology, Hôpital Européen Georges Pompidou, AP-HP, 20 rue Leblanc, 75015, Paris, France|
|2Paris Descartes University, Paris, France|
|3University Hospital-Pôle Régional de Cancérologie-CIC P802, 86000 Poitiers, France|
|4Epidemiology and Clinical Research Department, Hopital Bichat, 46 Rue Henri Huchard, 75018 Paris, France|
|*Corresponding Author :||Jacques Medioni
Département d’Oncologie Médicale
Hôpital Européen Georges Pompidou
20, rue Leblanc, 75015 Paris, France
Tel: (33) 1 56 09 27 81
Fax: (33) 1 56 09 548
E-mail: [email protected]
|Received November 21, 2015; Accepted March 04, 2016; Published March 11, 2016|
|Citation:Medioni J, Tournoux-Facon C, De Y (2016) A New Adaptive Simon- Based Design Focusing on Subpopulation Heterogeneity. Drug Des 5:128. doi: 10.4172/2169-0138.1000128|
|Copyright: © 2016 Medioni J, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.|
Background: Phase III trials can fail, leading to termination of drug development. This can result from heterogeneous subpopulations such as a drug-sensitive and a drug-insensitive subpopulation of patients or biological subtypes.
Methods: Traditional phase II methods do not detect heterogeneity of subpopulations. We proposed a new adaptive design for phase II trials, adapted from Simon’s “Minimax” design, where heterogeneity between two subpopulations could be highlighted.
Results: Drug inefficacy in one or two subpopulations could be determined at stage one and two and drug efficacy at stage two. Single Simon design and two independent Simon designs were compared to the adaptive design using calculated type I and II errors, expected and maximum sample size, and the probability of detecting drug-insensitive subpopulations. For the adaptive design, the type I and II errors calculated were similar to those of a single Simon design, sample size was smaller than with the two independent Simon designs (between 25 and 40% fewer patients) and the probability of detecting drug-insensitive subpopulations remained at 40%. An example with real data is presented.
Conclusion. In the event of different subpopulations of patients or biological subtypes, our adaptive design can help select the drug sensitive subpopulation in one single trial.