Back to the Top
Dear all,
I would like to know, how to calculate the Pharmacodynemic Parameters
and PK/PD correlation? Is it possible to calculate through WinNonLin
software ?
The following is the study design.
A double-blind, randomized, two-way crossover study to compare the
pharmacokinetics and Pharmacodynamics of single doses of Test Product
with Reference Product using euglycemic glucose clamp technique in
healthy, adult, male, human subjects.
Pharmacokinetic Parameters: INS-AUC(0-24h), INS-AUC(0-Yen), Cmax,
Tmax, t1/2
Pharmacodynemic Parameters : GIR-AUC(0-24h), GIRmax, TGIRmax.
Thank you very much in advance for your reply.
Thanks with Regards,
Jay Parikh
Biostatistician
Veeda Clinical Research Pvt. Ltd.
Ahmedabad, India
Web: www.veedacr.com
Back to the Top
The following message was posted to: PharmPK
Jay,
> I would like to know, how to calculate the Pharmacodynemic Parameters
> and PK/PD correlation? Is it possible to calculate through WinNonLin
> software ?
> The following is the study design.
>
> A double-blind, randomized, two-way crossover study to compare the
> pharmacokinetics and Pharmacodynamics of single doses of Test Product
> with Reference Product using euglycemic glucose clamp technique in
> healthy, adult, male, human subjects.
>
> Pharmacokinetic Parameters: INS-AUC(0-24h), INS-AUC(0-Yen), Cmax,
> Tmax, t1/2
>
> Pharmacodynemic Parameters : GIR-AUC(0-24h), GIRmax, TGIRmax.
There is a simple answer to your question and a complicated answer.
The answer depends on why you want to know the correlation -- and
most likely you dont want to know it!
Simple Answer
As you appear to be a statistician then I assume you know how to
compute a correlation coefficient. The simple answer is to compute
the correlation of INS-AUC with GIR-AUC. This is probably without
much meaning but may be enough if you are simply trying to fill in
the gaps in a statistical analysis plan.
Complicated Answer
People with limited background in PKPD sometimes speak of PKPD
correlation when what they mean is a linked model connecting the time
course of concentration (PK) to the effects (PD). Correlation in the
statistical sense of describing the way two random variables covary
is not the issue.
If you have true parameters of a model for the glucose-insulin system
(e.g. basal insulin secretion rate, basal glucose clearance -- not
simple statistics like Cmax, Tmax, AUC) then you might be interested
in the correlation of the between subject differences in the
parameters. But this is not of much value except for Monte Carlo
simulation.
It is possible to build plausible PKPD models for the glucose-insulin
system and to estimate the parameters using mixed effect non-linear
models e.g.
http://www.page-meeting.org/default.asp?abstract=1142
If you build such a model then you can do many interesting things to
understand the glucose-insulin system and the pharmacological
properties of the drugs you are testing.
I suggest you talk to the people who paid for the study and found out
what their real objective is for doing the study. You can then decide
if you can achieve the objective with a simple correlation
coefficient or with a complex model of physiology and pharmacology.
Nick
--
Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New
Zealand
n.holford.-a-.auckland.ac.nz
www.health.auckland.ac.nz/pharmacology/staff/nholford
Back to the Top
The following message was posted to: PharmPK
Jay,
There may be another factor to consider in your data, accounting for
endogenous insulin. A euglycaemic clamp does not account for
variability due to endogenous insulin. Alternative approaches include
suppression of endogenous insulin by hypoglycaemia (hypoglycaemic clamp)
or somatostatin, studying subjects with type 1 diabetes, or using
labelled insulin.
Matt
Matt Doogue
Clinical Pharmacologist / Endocrinologist
Flinders Medical Centre
Back to the Top
The following message was posted to: PharmPK
Which version of WinNonlin are you using? Note that in addition to the
many compartmental library models supplied with WinNonlin, (as well
those you write yourself), we also have Model 220 for an NCA assessment
of PD data. Please take a look at the Examples guide to see how you
should enter your data to obtain the parameters you are seeking, as well
as how to build PK-PD Link models.
Best regards,
Simon.
--
Simon Davis
Senior Scientific Consultant
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "PD and PK-PD correlation calculation" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)