Abstract:
It could be argued that design and implementation of MTD - seeking Phase I cancer trial should be trivial. The sense that this is so can come from the fact that the mathematics involved is elementary (seeking a maximum or minimum of a function), statistics is basic (estimation of a probability or of a dose) and the number of parameters or dimensions is minimal (mostly 1-2). The non-triviality comes from the difficulty in optimizing a simultaneous delivery of many factors: mathematical rigor, expression of prior beliefs, concern for safety, clarity in day to day implementation of the trial, decisions of the investigators, possible errors in dose assignment and the need for 'coherence' in escalation (YK Cheung, Biometrika (2005)), to name but a few. We will discuss "a+b", CRM and TITE-CRM methods for escalation.
Lunch is $30 for CCASA members, $35 for nonmembers. Nonmembers, join the chapter for a year for only $15 and get the discount plus all the other benefits of membership!
Please register for the luncheon by Friday October 17, 2008.