A Bayesian three-parameter logistic model for early- and late-onset DLTs in oncology Phase I studies.
J Biopharm Stat
; 26(2): 339-51, 2016.
Article
em En
| MEDLINE
| ID: mdl-25629564
ABSTRACT
We introduce a three-parameter logistic model to analyze the dose limiting toxicity (DLT) as a time-to-event endpoint in oncology Phase I trials. In the proposed model, patients are allowed to stay on trial without the constraint of a maximum follow-up time. Our model accommodates late-onset DLT as well as early-onset DLT, by both of which the dose recommendation is informed. A Bayesian approach is used to incorporate prior knowledge of the test treatment into dose recommendation. Simulation examples show that our proposed model has good operating characteristics in assessing the maximum tolerated dose (MTD).
Palavras-chave
Texto completo:
1
Base de dados:
MEDLINE
Assunto principal:
Modelos Logísticos
/
Teorema de Bayes
/
Ensaios Clínicos Fase I como Assunto
/
Oncologia
/
Antineoplásicos
Tipo de estudo:
Guideline
/
Prognostic_studies
/
Risk_factors_studies
Limite:
Humans
Idioma:
En
Ano de publicação:
2016
Tipo de documento:
Article