From: "Hyeong-Seok Lim"
Subject: [NMusers] Modeling with insufficient data
Date: Tue, October 26, 2004 6:28 am
Dear all,
I am trying to model a drug and its 3 metabolites using NONMEM.
After oral administration, the drug is converted to 2 first stage
metabolites, which are subsequently converted to a second stage metabolite.
However, I don't have a concentration data of one out of the
2 first stage metabolites.
Could anyone advise on how to model all the 4 compounds (the parent
drug and its 3 metabolites) with concentration data of only 3 compounds?
The only relevant information is that the amount ratio of the 2 first
stage metabolites in urine is 4:3.
Thanks in advance.
HS Lim
National Cancer Center, Republic of Korea
E-mail: mdlhs@ncc.re.kr
Modeling with insufficient data
From:"Bhattaram, Atul" BhattaramA@cder.fda.gov
Subject: RE: [NMusers] Modeling with insufficient data
Date: Tue, October 26, 2004 6:48 am
Hello Lim
I dont think you need to model the drug and three metabolites. You could do with the
drug and two metabolites (It depends on the time course of your metabolites relative
to the parent drug). You can fix the volume of distribution of the metabolites to
either 1 or some factor of the volume of distribution of the parent drug. You would
have to worry about identifiability issues if you have insufficient data.
Venkatesh Atul Bhattaram
Pharmacometrics
DPE-1, OCPB
CDER, FDA.
From: "Diane R Mould" drmould@attglobal.net
Subject: RE: [NMusers] Modeling with insufficient data
Date: Tue, October 26, 2004 6:59 am
Hi Hyeong-Seok
how are things going?
that's going to be some awful run time, but its quite do-able. what
I would suggest you do before modeling them simultaneously is to model
them separately to get estimates of the volumes and clearances of each.
then begin adding things piecewise. so model parent and first stage
metabolites (possibly one at a time given the likely run times)
set them up as each having its own compartment. so parent might be
compartments one and two (if its oral and 1 comp pk) then metabolite
1 is compartment 3 and metabolite 2 is compartment 4
there is no need to initialize these compartments as there would be
for an indirect type model
so something like this
$SUBS ADVAN6 TOL=3
$MODEL
COMP=(gut)
COMP=(cent defdose)
COMP=(met1)
COMP=(met2)
COMP=(TRA)
COMP=(met3)
$PK
"FIRST
"COMMON/PRCOMG/IDUM1,IDUM2,IMAX,IDUM4,IDUM5
"INTEGER IDUM1,IDUM2,IMAX,IDUM4,IDUM5
"IMAX=50000000
then define the parameters and des
$DES
DADT(1)=-ka*a(1)
dadt(2)=ka*a(1)-A(2)*K10-k23*a(2)-k24*a(2)
dadt(3)=k23*a(2)-k20*a(3)
dadt(4)=...
when you are missing a piece, that's ok, but as Atul pointed out,
you may have some difficulty identifying parameters. you may have
to fix something. I hope that this is enough to get you started
Diane
_______________________________________________________