Improving early phase oncology clinical trial design: A case study

Authors: Phillips, A.J. and Clark, T.P.

Journal: Pharmaceutical Statistics

Volume: 21

Issue: 6

Pages: 1370-1375

eISSN: 1539-1612

ISSN: 1539-1604

DOI: 10.1002/pst.2252

Abstract:

This short communication presents a first in human Bayesian Optimal Interval design case study. The study design and associated operating characteristics are discussed, together with study amendments proposed whilst the study was ongoing. Simulations investigating the impact of the amendments on the operating characteristics of the study design are presented. Lessons learnt from the case study, including providing practical advice when designing smarter early phase oncology trials to identify the maximum tolerate dose are also summarised. It is argued that model-assisted designs are simple to implement, flexible and perform significantly better than the commonly used “3 + 3” design, and thus should become the go to design for statisticians when limited information is known about the dose toxicity curve.

Source: Scopus

Improving early phase oncology clinical trial design: A case study.

Authors: Phillips, A.J. and Clark, T.P.

Journal: Pharm Stat

Volume: 21

Issue: 6

Pages: 1370-1375

eISSN: 1539-1612

DOI: 10.1002/pst.2252

Abstract:

This short communication presents a first in human Bayesian Optimal Interval design case study. The study design and associated operating characteristics are discussed, together with study amendments proposed whilst the study was ongoing. Simulations investigating the impact of the amendments on the operating characteristics of the study design are presented. Lessons learnt from the case study, including providing practical advice when designing smarter early phase oncology trials to identify the maximum tolerate dose are also summarised. It is argued that model-assisted designs are simple to implement, flexible and perform significantly better than the commonly used "3 + 3" design, and thus should become the go to design for statisticians when limited information is known about the dose toxicity curve.

Source: PubMed

Improving early phase oncology clinical trial design: A case study

Authors: Phillips, A.J. and Clark, T.P.

Journal: PHARMACEUTICAL STATISTICS

Volume: 21

Issue: 6

Pages: 1370-1375

eISSN: 1539-1612

ISSN: 1539-1604

DOI: 10.1002/pst.2252

Source: Web of Science (Lite)

Improving early phase oncology clinical trial design: A case study.

Authors: Phillips, A.J. and Clark, T.P.

Journal: Pharmaceutical statistics

Volume: 21

Issue: 6

Pages: 1370-1375

eISSN: 1539-1612

ISSN: 1539-1604

DOI: 10.1002/pst.2252

Abstract:

This short communication presents a first in human Bayesian Optimal Interval design case study. The study design and associated operating characteristics are discussed, together with study amendments proposed whilst the study was ongoing. Simulations investigating the impact of the amendments on the operating characteristics of the study design are presented. Lessons learnt from the case study, including providing practical advice when designing smarter early phase oncology trials to identify the maximum tolerate dose are also summarised. It is argued that model-assisted designs are simple to implement, flexible and perform significantly better than the commonly used "3 + 3" design, and thus should become the go to design for statisticians when limited information is known about the dose toxicity curve.

Source: Europe PubMed Central