Back to the Top
I have following two questions. I appreciate if you give me some comments on
In the case of multiple dose study, we compare AUCinf on day 1 to AUCtau at
steady-state, usually on day 7. Do you know the method to estimate
intra-subject variation between day 1 and the steady-state ? With using
population pharmacokinetics, intra-subject variation can be estimated in
addition to inter-subject variation. However, I have no idea for estimation of
intra-subject variation among the different administered day. For example, if
day 7's AUC is smaller than that on day 1, how can we conclude that it is due
to intra-subject variation but not drug effect and so on ? Could you tell me
pharmacokinetic method to estimate this intra-subject variation ?
Regarding to a drug which acts on cell cycle like 5-FU, time to maintain
effective concentration would be good indicator for the effect. In the case
that the minimum effective concentration in vivo is not known, AUC can be used
as a surrogate indicator. I have an experience that AUC is a good indicator
for the effect but not Cmax in the case of cell cycle acting drug. However,
textbook said Cmax is correlated to AUC. In my case, Cmax is also well
correlated to AUC, but AUC is much better indicator than Cmax regarding to the
effect. Do you have any answer for this paradox ?
I am looking forward to hearing from you.
Back to the Top
The major objective of comparing AUC(0-inf) after single dose and AUC(tau) at
steady state is to assess whether kinetics of the drug are altered on multiple
dosing. It would be wrong to compare AUC (0-inf) after single dose and AUC
(tau) to assess intrasubject variability.
One possible approach to assess intrasubject variability is to determine
AUC(tau) on two occasions at steady state. One would therefore have to
determine three AUC values. Also following administration of the last multiple
dose one could take blood samples beyond tau (usually 3 half lives). I use
such a study design extensively in drug development. It is a very powerful
design and provides key information on a drug including:
1) Is there an unexpected increase (induction) or decrease (capacity limited
kinetics) in the clearance on multiple dosing and can the steady state drug
levels predictable from the single dose data?
2) Mean residence time and multiple dosing half life of the drug following
single and multiple dosing.
3) Inter dose variability in the plasma concentrations at steady state. This
is not exactly the same as the intrasubject variability but in my opinion it
is a very useful measure of variability in the actual clinical use of the
4) Half life of the drug following single and multiple dosing. Note that the
terminal phase half life (t1/2 lambda_z) is not the same as multiple dosing
t1/2 (derived from mean residence time). It is the latter half life which is
the most accurate determination of dosing interval.
Above are just a few of the important pharmacokinetic parameters derived from
the suggested study design.
Back to the Top
There are two aspects to the above statement that are wrong:
1) The concentration that needs to be measured is what gets to the tumor, not
what is present in the blood. As a colleague of mine put it many years ago,
measuring drug levels in blood is like a gum-shoe observing who goes in an
out of a house. It gives you absolutely no clue as to what happens in a given
room. And as we (and others) have shown, kinetics of drugs at their target
(effector) sites can (and probably, must) be measured using noninvasive
2) In the particular case of 5-FU, whose tumoral targeting is primarily
determined by first passage processes, we have shown that one of the
effective indicators of its effect is given by its tumoral pharmacokinetics.
Those patients whose tumors trap 5-FU (tumoral t1/2 of free 5-FU > 20 min)
have a high likelihood of responding; and that patients who do not trap 5-FU
will not respond (p<.00001). We have now identified further conditions that
are simultaneously (and independently) necessary for correlation with
response to chemotherapy (Ann. Oncol. 8:203-4,1997).
=========================================================================| Professor Walter Wolf, Ph.D. E-Mail: wwolfw.-at-.hsc.usc.edu |
| Director, Pharmacokinetic Imaging Program |
| Department of Pharmaceutical Sciences Telephone: 213-342-1405 |
| University of Southern California Fax: 213-342-9804 |
| 1985 Zonal Ave., Los Angeles, CA 90033 |
| Center for Noninvasive Pharmacology, Los Angeles Oncologic Institute |
| MRI at St. Vincent Medical Center Telephone: 213-484-7235 |
| 2131 Third St., Los Angeles, CA 90057 Fax: 213-484-7447 |
Date: Mon, 23 Jun 1997 12:13:47 -0500
From: "Gabrielsson, Johan"
Subject: New PK/PD book edition
I just want to inform you about a new, updated, and expanded release of
Pharmacokinetic and Pharmacodynamic Data Analysis: Concepts and
Applications, 2nd edition by J. Gabrielsson and Daniel Weiner, has been
"A must to the library of anyone involved in
pharmacokinetic-pharmacodynamic data analysis and interpretation"
This new and expanded edition differs substantially from the first
edition in that pharmacokinetic and pharmacodynamic theory as well as
modeling concepts are expanded. Detailed sections on NCA, non-linear
kinetics (capacity, time, flow, and binding), turnover concepts,
pulsative functions, push-and-pull and precursor models, binding- and
displacement models, toxicokinetic models and determination of initial
parameter estimates for complex models are included. PART ONE of the
text is a basic introduction to pharmacokinetic and pharmacodynamic
modeling principles and can be understood by scientists with very little
formal training in this area. Complex mathematics are minimized.
Extensive graphical presentations accompany the theory. This enables the
scientist to conceptualize the underlying principles, such as clearance,
potency, tolerance weighting, and search algorithms. Issues on program
validation are also discussed. PART TWO is aimed to illustrate the
application of the concepts to specific modeling situations commonly
encountered at the drug industry and in research labs. Several new
exciting data sets on tolerance-rebound, dose-response-time models,
bacterial growth model, turnover, non-linear kinetics, oscillating input
functions, and toxicokinetics are also solved and discussed in this
section. The book bridges the gap between standard texts in
pharmacokinetics, pharmacology and numerical analysis, and provides
unique insight into the field of pharmacometrics in an easy to
* Substantially expanded concepts sections (kinetics, dynamics,
* Increased graphical presentation of concepts and data
* State-of-the-art applications section
The book + all example files on a diskette is available from the Swedish
Pharmaceutical Press (FAX +46-(0)8-14 95 80) or D. Weiner, SCI, Inc.
(FAX 919-859-6871, phone: 919-859-6868 or E-mail:
75450.3171.-at-.compuserve.com) for USD 160.
Back to the Top
We find interesting your questions regarding intra-subject variability
of drugs when we have determined the induividual PK on two different
Assuming the clearance of the drug do not change whithin the two days,
then AUCinf = AUCtau.
If the AUC are different, as is the case, it can be due to a) inter-dose
associated variability or also b) drug clearance change. To dilucidate
this issue, you can quantify inter- and residual-variability and
inter-occassion variability using NoN lineal Mixed Effects Model (NONMEM
program) or other similar statistical computer resource.
Interoccassional variability is a concept included in the NONMEM program
that may be useful for your problems.
Hospital Dr. Peset
Servicio de Farmacia
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (email@example.com)