Dear NMusers,
I'm working on a PK-PD model to describe paclitaxel intra-tumoural PK and PD
post intraperitoneal administration.
Hereto, we collected a.o. tumour tissue at different time-points post dosing.
After tissue collection we divided the tumour tissue specimens in different
portions according to depth from tumour surface.
As part of the PK model I'm trying to model the concentration-decay over time
as well as the concentration decay over depth.
However, I'm currently facing some problems.
At the moment, my model deals with the concentration-time profile under $DES
and then corrects for depth under $ERROR.
$DES
...
DADT(5)=A(1)*VM/(KM+A(1))-K50*A(5)
$ERROR
...
INT=-7
IF(F.GT.0) INT = LOG(F)
IF(CMT.EQ.5) THEN
IPRED=INT+SLOPE*DEPTH
Y = IPRED+EPS(2)
ENDIF
...
This works out fine.
However, since I'm correcting for depth in a post-hoc fashion, I'm wondering
which information NM is using during the integration step.
Is it using the average of the DVs sharing the same TIME value (without taking
into account the DEPTH variable), is it only using the first DV value from the
ones sharing a TIME value?
Secondly, I was wondering whether it is possible to apply the Depth correction
within the $DES statement? Or would this require the use of partial
differential equations rather than ODE's?
My dataset looks like this:
ID
TIME
DV
CMT
AM0
DOSE
RAT0
EV0
MV0
DEPTH
SIZE
17
0
0
1
30000
30000
0
1
1
0
0
17
0.75
0.912189
5
0
30000
0
0
0
1.25
7.5
17
0.75
1.150486
5
0
30000
0
0
0
1.25
7.5
17
0.75
0.202403
5
0
30000
0
0
0
3.75
7.5
17
0.75
2.187764
6
0
30000
0
0
0
1.25
7.5
17
0.75
1.641103
6
0
30000
0
0
0
1.25
7.5
17
0.75
1.495206
6
0
30000
0
0
0
3.75
7.5
Kind regards,
Pieter Colin, Pharmacist
Ph.D. Student (Pre-) Clinical PK/PD Modelling & Simulation
Laboratory of Medical Biochemistry and Clinical Analysis
Faculty of Pharmaceutical Sciences
Ghent University
Harelbekestraat 72
B-9000 Gent
Belgium
Tel.: +32-9-264-81-14
Fax: +32-9-264-81-97
E-mail: [email protected]<mailto:[email protected]>