Posterior Probability of Adverse Events (AE) Categories for Phase I Studies

• Activity Number: 342
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Kyounghwa Bae and John W. Seaman III, PhD and Mark VonTress*+
• Companies: Alcon Laboratories and Alcon Laboratories and Alcon Laboratories
• Keywords: Phase I ; Bayesian ; Advers Events

Abstract

Phase I trials are often utilized to find the maximum tolerated dose (MTD) of a treatment with a dose-limiting toxicity (DLT) no more than some pre-determined rate. In the Bayesian setting, the posterior distribution for the probabilities of a DLT is used to choose a dose for the next patient cohort using information that has been collected in the trial. Neuenschwander et al. (2008) used a Bayesian approach to summarize and categorize the posterior probability of a DLT after each patient cohort which allows for the incorporation of uncertainty, rather than using a point estimate. We use this concept in a logistic regression, as its parameters can be understood intuitively by a non-statistician. This modified method calculates a posterior probability of adverse event categories and also determines the MTD using the Escalation With Over dose Control method (2010). A user friendly R program for the modified method has been developed.