Back to the Top
Dear all,
Can anybody help us in finding out absorption rate constants from IV
and oral profile using deconvolution analysis in winnonlin
(in detail)
Regards,
Raghav.
Back to the Top
The following message was posted to: PharmPK
Dear Dr Chowdray,
I think you'll find a fairly detailed example from page 130 of
the WinNonlin Examples Guide 5.2.1.pdf you can find in the folder
C:\Program Files\Pharsight\WinNonlin\User Docs. The user manual and
online help also explain a little more about this, if you still have a
specific technical problem you can contact Support or perhaps consider
attending one of our training courses - please see below;
Best regards,
Simon.
--
Simon Davis
Senior Scientific Consultant
Pharsight- A Certara(tm) Company
Back to the Top
The following message was posted to: PharmPK
Dear Raghav,
First, there is no such thing as an absorption rate constant - there
is an
absorption rate _coefficient_, which is time-dependent. Mathematical
shortcuts such as you describe can sometimes be used successfully, but
they
are approximations based on simplifying assumptions. Unless you know
ahead
of time that your drug is absorbed rapidly, you are taking a chance by
assuming a constant Ka. In doing so, you will have to adjust other PK
parameters to fit Cp-time, and you may even get a nice plot, but both
the Ka
and the PK parameters will be wrong to balance the errors - two wrongs
to
try to make a right.
A more accurate way to "deconvolute" the time-dependent value of the
absorption rate coefficient is provided in GastroPlus. That involves a
mechanistic model to calculate absorption rate based on local
conditions in
the gastrointestinal tract. As the remaining unabsorbed drug transits
through regions of different pH, different permeability, and different
fluid
volume, the concentration gradient of drug in solution between the
lumen and
the enterocytes changes, which changes the effective value of Ka in each
region at any point in time. In other words, drug will be getting
absorbed
in multiple regions at the same time, each with its own time-dependent
Ka.
If transporters are involved, it gets more complex - but state-of-the-
art
simulations can deal with that situation as well.
The "net Ka" at any point in time can be deduced (deconvoluted) from the
combination of all of the local absorption rates and the total amount of
drug in solution at that point in time. If you model absorption rate
as Ka
times mass of drug in solution then:
d(abs)/dt = ka * M(solution)
If you solve for Ka you get:
Ka = [d(abs)/dt] / [M(solution)]
If you look at the plots of net time-dependent Ka vs time, you'll
typically
see something akin to a Cp-time curve, starting at 0 at time 0, then
rising
to a peak, then decreasing, sometimes increasing again as the drug moves
into caecum and colon.
Of course, you wouldn't actually apply the Ka-vs-time curve to calculate
absorption because in order to find it, you will have already modeled
the
absorption phenomena in greater detail, so the value of knowing a
time-dependent Ka is more of a curiosity.
Walt Woltosz
Chairman & CEO
Simulations Plus, Inc. (NASDAQ: SLP)
42505 10th Street West
Lancaster, CA 93534-7059
U.S.A.
http://www.simulations-plus.com
E-mail: walt.-at-.simulations-plus.com
Back to the Top
The following message was posted to: PharmPK
Hi Walt,
The GastroPlus method of charactering the absorption profile from
physiological mechanisms is to be recommended. However, "black box"
deconvolution methods (like that in WinNonlin) can, of course, also
provide
a non-mechanistic absorption profile (without assuming first-order
absorption).
Charlie
Back to the Top
The following message was posted to: PharmPK
Charlie,
Actually, the "black box" deconvolution methods to not provide an
absorption
profile. They can and do provide a non-mechanistic systemic availability
profile - provided, of course, that the PK parameters are correct. If
they
are not, then the systemic availability will be adjusted to compensate
for
the errors in the PK (two wrongs <> right). The danger here is getting a
nice-looking plot with wrong PK and systemic availability and then
trying to
predict a different formulation or dose with the wrong values.
Remember that the FDA-accepted definition of absorption means only
leaving
the intestinal lumen and entering the enterocytes. First pass
extraction in
the gut and liver often results in a very different amount of drug that
reaches the systemic circulation from that which entered the
enterocytes.
The older references that use the word absorption (Wagner-Nelson,
Loo-Riegelman, etc.) are really referring to systemic availability.
That's
unfortunate, because this is an important distinction, and I believe we
should encourage everyone to use these terms with care.
Best regards,
Walt Woltosz
Chairman & CEO
Simulations Plus, Inc. (NASDAQ: SLP)
42505 10th Street West
Lancaster, CA 93534-7059
U.S.A.
http://www.simulations-plus.com
E-mail: walt.-at-.simulations-plus.com
[Walt do you have a FDA reference for your definition of absorption? A
recent search still seems to describe absorption in terms of systemic
availability - db]
Back to the Top
Dear Members,
I have been estimating the human oral pharmacokinetic parmeters for a
new drug.
To simulte the oral pharmacokinetics, needed the average absorption
rate constants (ka) from
animals (rat ,dog and monkey) and these constants are obtained from
each individual species using deconvolution analysis of the IV and
oral data, and i have a problem related to input in winnolin and
using deconvolution for this data.
I used estimated parameters (A and alpha ) using Model 7 from IV data
and used these parameters for deconvoluting oral data.
Please suggest !
Raghav
Back to the Top
The following message was posted to: PharmPK
Dear Raghav,
The most effective way to predict human PK from preclinical data is
with a
PBPK mode with a mechanistic absorption simulation (NOT a "constant
Ka" -
there is no such thing).
If you would like some help, feel free to e-mail me directly.
Best regards,
Walt Woltosz
Chairman & CEO
Simulations Plus, Inc. (NASDAQ: SLP)
42505 10th Street West
Lancaster, CA 93534-7059
U.S.A.
http://www.simulations-plus.com
E-mail: walt.at.simulations-plus.com
Back to the Top
The following message was posted to: PharmPK
Dear Raghav,
The most effective way to predict human PK from preclinical data is
with a
PBPK mode with a mechanistic absorption simulation (NOT a "constant
Ka" -
there is no such thing).
If you would like some help, feel free to e-mail me directly.
Best regards,
Walt Woltosz
Chairman & CEO
Simulations Plus, Inc. (NASDAQ: SLP)
42505 10th Street West
Lancaster, CA 93534-7059
U.S.A.
http://www.simulations-plus.com
E-mail: walt.at.simulations-plus.com
PharmPK Discussion List Archive Index page
Copyright 1995-2009 David W. A. Bourne (david@boomer.org)