Including Quantitative Benefit‐Risk Assessment in Seamless Phase 2/3 Designs With Dose Selection
Statistics in Medicine · 2026-11 · 45 (25-27)
DOI: 10.1002/sim.70790 ↗At a glance
Journal article in Statistics in Medicine, published online on 2026-11-16. Written by Marco Ratta and 5 co-authors. Institution: Dipartimento di Scienze Matematiche, Politecnico di Torino, Torino, Italy. Major MEDLINE topics: Clinical Trials, Phase II as Topic, Clinical Trials, Phase III as Topic, Research Design. 51 references. The full text is freely available (open access, CC BY 4.0). Funded by: Medical Research Council, National Institute for Health and Care Research.
Abstract (original language)
Seamless Phase II/III designs are increasingly employed to accelerate drug development, particularly in settings where multiple doses of the same compound need to be evaluated in parallel. In such designs, a restricted subset of doses is selected at the first stage of the trial and subsequently carried forward to the second stage, where efficacy is assessed against control. Dose selection is typically based on a single efficacy endpoint, ensuring consistency between interim decision‐making and final hypothesis testing. However, safety is typically not formally integrated into the selection process during the design stage, which can be problematic, as it may lead to advancing doses that are effective but potentially harmful. In this work, we propose a novel multi‐arm two‐stage design in which the dose with the most favorable benefit‐risk profile at the interim analysis is selected and continued into Phase III. Benefit‐risk assessment is conducted using probabilistic multi‐criteria decision analysis (MCDA), where information from multiple efficacy and safety endpoints is combined into a single aggregated score. For the proposed design, we derive an analytical expression for the decision boundary, ensuring strong control of the Type I error rate, and provide guidance on sample size determination. The operating characteristics are evaluated through extensive simulation studies and compared with existing designs from the literature, demonstrating that the proposed approach is particularly effective in terminating potentially harmful doses at an early stage.
Subjects
MeSH terms (MEDLINE)
Author keywords
Authors and institutions
- Dipartimento di Scienze Matematiche, Politecnico di Torino, Torino, Italy.
- Department of Statistical Methodology, Saryga, Tournus, France.
- Servier National Research Institute (IRIS), Servier, Gif-sur-Yvette, France.
- Astellas Pharma, Boston, Massachussets, USA.
- MRC Biostatistics Unit, University of Cambridge, Cambridge, UK.
ORCID profiles: Marco Ratta, Zhaoyang Teng, Gaëlle Saint-Hilary, Mauro Gasparini, Pavel Mozgunov
Funding
- Medical Research Council
- National Institute for Health and Care Research
Cite
APA
Ratta, M., Skanji, D., Teng, Z., Saint-Hilary, G., Gasparini, M., & Mozgunov, P. (2026). Including Quantitative Benefit‐Risk Assessment in Seamless Phase 2/3 Designs With Dose Selection. Statistics in Medicine, 45(25-27). https://doi.org/10.1002/sim.70790
Vancouver
Ratta M, Skanji D, Teng Z, Saint-Hilary G, Gasparini M, Mozgunov P. Including Quantitative Benefit‐Risk Assessment in Seamless Phase 2/3 Designs With Dose Selection. Statistics in Medicine. 2026;45(25-27). doi:10.1002/sim.70790
BibTeX
@article{Ratta2026,
author = {Ratta, Marco and Skanji, Donia and Teng, Zhaoyang and Saint-Hilary, Gaëlle and Gasparini, Mauro and Mozgunov, Pavel},
title = {Including Quantitative Benefit‐Risk Assessment in Seamless Phase 2/3 Designs With Dose Selection},
journal = {Statistics in Medicine},
year = {2026},
volume = {45},
number = {25-27},
doi = {10.1002/sim.70790}
}
EUVIMED