Adaptive clinical trial designs to detect interaction between treatment and a dichotomous biomarker

Biomarkers play a crucial role in the design and analysis of clinical trials for personalized medicine. One major goal of these trials is to derive an optimal treatment scheme based on each patient's biomarker level. Although completely randomized trials may be employed, a more efficient design can be attained when patients are adaptively allocated to different treatments throughout the trial using biomarker information. Therefore, we propose a new adaptive allocation method based on using multiple regression models to study treatment–biomarker interactions. We show that this perspective simplifies the derivation of optimal allocations. Moreover, when implemented in real clinical trials, our method can consolidate all the covariates which may not be related to the treatment–biomarker interaction for a joint analysis. Our general idea can be applied to diverse models to derive optimal allocations. Simulation results show that both the optimal allocation and the proposed design can lead to a more efficient trial. The Canadian Journal of Statistics 41: 525–539; 2013 © 2013 Statistical Society of Canada Resume Les biomarqueurs jouent un role essentiel dans la conception et l'analyse d'essais cliniques en medecine personnalisee. L'un des objectifs principaux de ces essais consiste a mettre au point un plan de traitement optimal fonde sur le niveau des biomarqueurs de chaque patient. Bien que des essais entierement randomises soient possibles, un plan plus efficace peut etre obtenu en repartisssant les patients entre les traitements de facon adaptative au cours de l'essai en se basant sur l'information provenant des biomarqueurs. Ainsi, les auteurs proposent une nouvelle methode adaptative de repartition basee sur des modeles de regression multiple evaluant les interactions entre le traitement et les biomarqueurs. Les auteurs montrent que cette technique simplifie les calculs menant a la determination optimale du traitement. De plus, lorsqu'elle est appliquee a de veritables essais cliniques, la methode proposee peut consolider toutes les covariables qui ne sont pas necessairement liees a l'interaction entre le traitement et les biomarqueurs dans le cadre d'une analyse commune. Le concept general des auteurs peut etre applique a divers modeles en vue d’etablir l'assignation optimale. Des simulations montrent que la methode proposee et l'assignation optimale au traitement peuvent mener a des essais plus efficaces. La revue canadienne de statistique 41: 525–539; 2013 © 2013 Societe statistique du Canada

[1]  Li-Xin Zhang,et al.  Asymptotic properties of covariate-adjusted response-adaptive designs , 2006 .

[2]  Uttam Bandyopadhyay,et al.  Adaptive designs for normal responses with prognostic factors , 2001 .

[3]  A. Atkinson.  Optimum biased coin designs for sequential clinical trials with prognostic factors , 1982 .

[4]  S. Pocock,et al.  Sequential treatment assignment with balancing for prognostic factors in the controlled clinical trial. , 1975, Biometrics.

[5]  Feifang Hu,et al.  Asymptotic normality of urn models for clinical trials with delayed response , 2004 .

[6]  Stephane Heritier,et al.  Dynamic balancing randomization in controlled clinical trials , 2005, Statistics in medicine.

[7]  W. Rosenberger,et al.  COVARIATE-ADJUSTED RESPONSE-ADAPTIVE DESIGNS FOR BINARY RESPONSE , 2001, Journal of biopharmaceutical statistics.

[8]  D R Taves,et al.  Minimization: A new method of assigning patients to treatment and control groups , 1974, Clinical pharmacology and therapeutics.

[9]  William F. Rosenberger,et al.  Handling Covariates in the Design of Clinical Trials. , 2008, 1102.3773.

[10]  Orjan Nordle,et al.  A self‐adjusting randomization plan for allocation of patients into two treatment groups , 1977, Clinical pharmacology and therapeutics.

[11]  Feifang Hu,et al.  Balancing continuous covariates based on Kernel densities. , 2013, Contemporary clinical trials.

[12]  Feifang Hu,et al.  Doubly adaptive biased coin designs with delayed responses , 2008 .

[13]  A. Atkinson,et al.  Adaptive biased‐coin designs for skewing the allocation proportion in clinical trials with normal responses , 2005, Statistics in medicine.

[14]  Dongsheng Tu,et al.  K-ras mutations and benefit from cetuximab in advanced colorectal cancer. , 2008, The New England journal of medicine.

[15]  Feifang Hu,et al.  Efficient randomized-adaptive designs , 2009, 0908.3435.

[16]  Val Gebski,et al.  Dynamic balanced randomization for clinical trials , 1993 .

[17]  William F. Rosenberger,et al.  Asymptotic Properties of Adaptive designs for Clinical Trials with delayed Response , 2002 .

[18]  William F. Rosenberger,et al.  Exact properties of Efron’s biased coin randomization procedure , 2010, 1010.0483.

[19]  W. Rosenberger,et al.  The theory of response-adaptive randomization in clinical trials , 2006 .

[20]  M Zelen,et al.  The randomization and stratification of patients to clinical trials. , 1974, Journal of chronic diseases.

[21]  L. J. Wei,et al.  The Adaptive Biased Coin Design for Sequential Experiments , 1978 .

[22]  C. Baker.  Another success for hepatitis A vaccine. , 2007, The New England journal of medicine.

[23]  F. Hu,et al.  Asymptotic properties of covariate-adaptive randomization , 2012, 1210.4666.

[24]  B. Efron.  Forcing a sequential experiment to be balanced , 1971 .

[25]  Lee-Jen Wei,et al.  A Class of Designs for Sequential Clinical Trials , 1977 .

[26]  Uttam Bandyopadhyay,et al.  A covariate adjusted two‐stage allocation design for binary responses in randomized clinical trials , 2007, Statistics in medicine.