Back to the Top
Are there any formulas to determine how many timepoints should be
sampled for a compound in a Phase I human trial to determine AUC,
t1/2, Cl, Vd, Cmax, and MRT? I have heard that a rule of thumb is
the number of parameters you want to determine plus 3 will give you
the appropriate number of timepoints to sample. When does it become
excessive? Thanks for the time.
Kirk S. Culotta, Pharm.D., B.S., R.Ph.
Clinical Research Pharmacist - Clinical & Translational Research Center
Pharmacology Research - Division of Pharmacy
UT M.D. Anderson Cancer Center
Back to the Top
The following message was posted to: PharmPK
> Are there any formulas to determine how many timepoints
> should be sampled for a compound in a Phase I human trial to
> determine AUC, t1/2, Cl, Vd, Cmax, and MRT? I have heard
> that a rule of thumb is the number of parameters you want to
> determine plus 3 will give you the appropriate number of
> timepoints to sample. When does it become excessive? Thanks
Interesting question. Excessive is the optimal word here. If excessive
= unaffordable then do less samples. If excessive is still affordable -
then why not (afterall this is phase I which is by definition
exploratory and excessive is generally the rule of thumb)?
AUC, Cmax, MRT are not parameters - so should not be considered in this
equation (rule of thumb). If you included Tmax as well - would you
require more samples?
CL, Vd are probably not the only parameters and hence you wouldn't want
to base your design on just these alone otherwise your design will be
degenerate (i.e. you will not be able to do anything of interest with
So, figure out what you really want to know and then design the study
accordingly (my guess is you want to know the same thing everyone wants
to know from phase I). Generally phase I studies have lots of samples.
Phase I studies are only retrospectively inefficient, i.e. only once you
know the PK characteristics can you design a more efficient study. In
contrast phase II/III studies are commonly inefficient (these can be
improved by optimal design techniques).
NB The optimally inefficient phase I study is one that fails.
School of Pharmacy, University of Queensland, Brisbane 4072, Australia
Tel +61 7 3365 8808; Fax +61 7 3365 1688; Email:
POPT (optimal design software):
MCMC PK example: http://www.uq.edu.au/pharmacy/sduffull/MCMC_eg.htm
Back to the Top
Dave, thanks for your response.
I guess what I am trying to say is that say you have a short
continous IV infusion over 3 hours of a compound that historically
has a biphasic distribution that is being given concurrently with
other meds and the purpose of the investigation is to determine the
dispositional characteristics of the drugs in combination. Is it
necessary to sample at 0.5, 1, 1.5, 2, 2.5, and 3 hours post-SOI and
1, 2, 3, 4, 6, 8, 10, 12, 14, 16, 18, and 24hrs post-EOI? Can't any
of these timepoints be removed i.e. 0.5, 1.5, and 2.5hr post-SOI and
2, 4, 8, and 12hr post-SOI without affecting the modeling and
avoiding needless blood sampling in patients? Of course, the more
you sample the better you can characterize the PK, but can't it be
done more efficiently? I think patient sampling should be kept to a
minimum, if possible.
[If the model is biphasic distribution you would have four parameters
(two compartment model) after iv infusion. If you have some idea of
the parameter values optimal sampling techniques will give you the
four best times to sample. For more sample you could 'cluster' around
the optimal times. - db]
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (firstname.lastname@example.org)