Hello. My name is Ying Yuan.
I'm a professor in Biostatistics
at MD Anderson Cancer Center.
Today I'm going to talk about
how to design early phase drug combination trial.
So here is outline of the talk.
First, I will talk about the challenges
for designing drug combination trials
then I will talk about some specific designs
for drug combination trial,
specifically, I will talk about
how to design a trial to find a single MTD,
and how to design a trial
to find its MTD contour.
Then I will introduce some software
to implement those designs,
and I will finish the talk
with a summary statement.
As we know, the primary objective
of phase I trial
is to find its maximum tolerated dose,
which is defined as a dose
with a specific targeted toxicity rate
So for single-agent trials,
the standard assumption is that
toxicity monotonically increases with the dose.
So in this case, we only find a dose
with a specific target toxicity rate.
And in this monotonic assumption means,
for the single-agent trial
the toxicity order is known.
And this greatly simplifies
the dose escalation and de-escalation.
For example, if we know the current dose
is below the MTD,
we only need to escalate the dose.
If we know the current dose
is above the MTD,
we only needed to de-escalate the dose.