Simulation study for evaluating an adaptive-randomisation Bayesian hybrid trial design with enrichment.
Details
State: Public
Version: Final published version
License: CC BY-NC-ND 4.0
Serval ID
serval:BIB_EB6E8007B5B0
Type
Article: article from journal or magazin.
Collection
Publications
Institution
Title
Simulation study for evaluating an adaptive-randomisation Bayesian hybrid trial design with enrichment.
Journal
Contemporary clinical trials communications
ISSN
2451-8654 (Electronic)
ISSN-L
2451-8654
Publication state
Published
Issued date
06/2023
Peer-reviewed
Oui
Volume
33
Pages
101141
Language
english
Notes
Publication types: Journal Article
Publication Status: epublish
Publication Status: epublish
Abstract
As we enter the era of precision medicine, the role of adaptive designs, such as response-adaptive randomisation or enrichment designs in drug discovery and development, has become increasingly important to identify the treatment given to a patient based on one or more biomarkers. Tailoring the ventilation supply technique according to the responsiveness of patients to positive end-expiratory pressure is a suitable setting for such a design.
In the setting of marker-strategy design, we propose a Bayesian response-adaptive randomisation with enrichment design based on group sequential analyses. This design combines the elements of enrichment design and response-adaptive randomisation. Concerning the enrichment strategy, Bayesian treatment-by-subset interaction measures were used to adaptively enrich the patients most likely to benefit from an experimental treatment while controlling the false-positive rate.The operating characteristics of the design were assessed by simulation and compared to those of alternate designs.
The results obtained allowed the detection of the superiority of one treatment over another and the presence of a treatment-by-subgroup interaction while keeping the false-positive rate at approximately 5\% and reducing the average number of included patients. In addition, simulation studies identified that the number of interim analyses and the burn-in period may have an impact on the performance of the scheme.
The proposed design highlights important objectives of precision medicine, such as determining whether the experimental treatment is superior to another and identifying wheter such an efficacy could depend on patient profile.
In the setting of marker-strategy design, we propose a Bayesian response-adaptive randomisation with enrichment design based on group sequential analyses. This design combines the elements of enrichment design and response-adaptive randomisation. Concerning the enrichment strategy, Bayesian treatment-by-subset interaction measures were used to adaptively enrich the patients most likely to benefit from an experimental treatment while controlling the false-positive rate.The operating characteristics of the design were assessed by simulation and compared to those of alternate designs.
The results obtained allowed the detection of the superiority of one treatment over another and the presence of a treatment-by-subgroup interaction while keeping the false-positive rate at approximately 5\% and reducing the average number of included patients. In addition, simulation studies identified that the number of interim analyses and the burn-in period may have an impact on the performance of the scheme.
The proposed design highlights important objectives of precision medicine, such as determining whether the experimental treatment is superior to another and identifying wheter such an efficacy could depend on patient profile.
Keywords
Bayesian designs, Marker-based strategy designs, Precision medicine, Response-adaptive randomisation
Pubmed
Web of science
Open Access
Yes
Create date
07/07/2023 9:46
Last modification date
09/08/2024 16:08