Back to the Top
Hi,
I am running a 3-compartmental PK model with a non-linear elimation
(Michealis-Menten) from the central compartment. The data were from
3 trials with total patients of 224. The drug was given as 20-min iv
infusion weekly or every three weeks at different dose leves.
When it was run with first order (FO) estimation, the model got
converged successfully and produced very reasonable PK parameters
(THETAs), inter-individual variabilities (ETAs), and relative
standard errors of estimations. The fitting looks pretty good.
However, when it runs with first order conditional estimation (FOCE),
convergence becomes almost impossible. Interation becomes extremely
slow, and after several iterations, there is always an error message
like "...Occurs during search for ETA at a nonzero value of ETA.
Numerical difficulties with integration routine. Maximum No. of
evaluaitons of differential equations exceeded 100000".
Does anyone have some good ideas to solve this problem? I appreciate!
Also, is it acceptable to report parameters produced by FO? I am
wondering if there are some guidelines or references for the choice
of FO and FOCE.
Thanks you in advance for any help!
Sincerely,
Jing
Back to the Top
The following message was posted to: PharmPK
Hello Jing
The difficulties with FOCE can often be resolved with a simpler
structure
for your random effects in your model. It is a common issue that is
observed
as you (error message)use methods that are more complex such as FOCE or
FOCE+Interaction in comparison to FO. Results from FO method are often
published in literature. It also depends on the purpose of your
modeling of
the data. If the analysis would result in critical recommendations
then it
might be useful to run your models with methods such as FOCE etc.
Venkatesh Atul Bhattaram
Pharmacometrics
DPE-1, OCPB
CDER, FDA.
Back to the Top
The following message was posted to: PharmPK
Jing,
What you are seeing is a fairly common experience. You can try a
number of
different things, none of which will really speed up the run times.
They are
what they are. First, try using a transform-both-sides approach
using the
Ln-transformation. Modeling on a log-scale always seems to help.
Second,
you might try reparameterizing your Michaelis-menten model to a more
stable
form. There are two possibilities here:
Cp = A(1)/V1
CL = Vmax/(exp(km) + Cp)
or
Cp = A(1)/V1
CL = CL_linear/(1 + alpha*Cp)
You can also reduce the number of variance components, especially
near zero
ones.
I wish I had a solution to the Maxevals problem short of changing the
NONMEM
code.
Good luck,
Pete Bonate
Peter L. Bonate, PhD, FCP
Director, Pharmacokinetics
Genzyme Corporation
4545 Horizon Hill Blvd
San Antonio, TX 78229
phone: 210-949-8662
fax: 210-949-8219
email: peter.bonate.-a-.genzyme.com
Back to the Top
The following message was posted to: PharmPK
Dear Jing,
You may want to browse the nmusers list archives for more tips, but the
solution to your problem is described in
http://huxley.phor.com/nonmem/nm/98oct071998.html
For guidelines regarding the use of conditional estimation methods,
see the conditional estimations guide (part VII of the nonmem users
guide).
Best regards,
Jeroen
J. Elassaiss-Schaap
Scientist PK/PD
Organon NV
PO Box 20, 5340 BH Oss, Netherlands
Phone: + 31 412 66 9320
Fax: + 31 412 66 2506
e-mail: jeroen.elassaiss.aaa.organon.com
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)