Hi
I have collected data on absolute neutrophil counts
(ANC) from children (n=30) with acute leukaemia (from
the maintenance phase of therapy, lasting up to 3
years). I also have the 6-Mercaptopurine dosing data,
but no plasma concentration data. Each child attended
clinic every 2 weeks to measure ANC and adjustment of
6-MP dose. All children received 75mg/m2. Doses
started at 100%, but were reduced to 75% or 50%
depending on ANC count. On most occasions when the ANC
level dropped below 1.0, the 6-MP was withheld until
the ANC level returned back to >1.
I have tried to implement a K-PD model as described by
P. Jacqmin et al (JPP;34(1) 2007) without success. I
provide the code and snippet of the data file (The
dose is actual dose administered and I have initiated
the PD CMP=3 by putting the initial amount in the AMT
column). I have not had any success with trying to
develop a base model; the model spectactularly fails
to describe the observed data. The population PRED
estimate is virtually the same (a constant) for all
time points for all patients (even though ANC sampling
times and doses varied between patients). The IPRED is
also the same as PRED. I have tried many different
initial estimates. What am I doing wrong? Can anyone
please diagnose the problem?
$MODEL
NPAR=6 NCOMP=3
COMP=(DOSE)
COMP=(BIOPHASE)
COMP=(ANCLEVEL)
$PK
KIN=THETA(1)*EXP(ETA(1))
KOUT=THETA(2)*EXP(ETA(2))
KDE=THETA(3)*EXP(ETA(3))
EDK50=THETA(4)*EXP(ETA(4))
GAMMA=THETA(5)*EXP(ETA(5))
KA=THETA(6)*EXP(ETA(6))
S2=1
;volume of biophase fixed
;KDE IS THE ELIMINATION RATE CONSTANT FROM BIOPHASE
;IRG IS THE (VIRTUAL)INFUSION RATE IN THE INHIBITION
FUNCTION
;EDK50 IS THE IR THAT LEADS TO 50% NHIBTION OF KIN
$DES
DADT(1)=-KA*A(1)
DADT(2)=KA*A(1)-KDE*A(2)
VIR=A(2)*KDE
IRG=VIR**GAMMA
COEF=1-IRG/(EDK50**GAMMA+IRG)
DADT(3)=KIN*COEF-KOUT*A(3)
$ERROR
Y = F + F*ERR(1)+ERR(2)
IPRED=F
$EST METHOD=0 POSTHOC PRINT=5 MAX=9999 SIG=3
MSFO=6mp2.msf NOABORT
$THETA
(0.0001,0.01,10) ;[KIN]
(0.0001,0.001,10) ;[KOUT]
(0.01,1,100) ;[KDE]
(0.1,20,100);[EDK50]
(0.001,1,5) ;[GAMMA]
(0.01,0.1,10)
ID TIME AMT DV CMT MDV EVID SS II
1 1 0.5 . 3 1 1 . .
1 1 . 0.5 3 0 0 . .
1 8 . 2.4 3 0 0 . .
1 8 20 . 1 1 1 . .
1 14 20 . 1 1 1 1 1
1 15 . 0.4 3 0 0 . .
1 22 . 0.6 3 0 0 . .
1 29 . 0.8 3 0 0 . .
1 29 21 . 1 1 1 . .
1 35 21 . 1 1 1 1 1
1 36 . 1.4 3 0 0 . .
1 36 31 . 1 1 1 . .
1 42 31 . 1 1 1 1 1
1 43 . 0.9 3 0 0 . .
1 43 21 . 1 1 1 . .
1 49 21 . 1 1 1 1 1
1 50 . 1.7 3 0 0 . .
1 50 31 . 1 1 1 . .
1 63 31 . 1 1 1 1 1
1 64 . 0.1 3 0 0 . .
1 71 . 0.4 3 0 0 . .
1 78 . 0.6 3 0 0 . .
thanks
M Sarasia
____________________________________________________________________________________
Tonight's top picks. What will you watch tonight? Preview the hottest shows on
Yahoo! TV.
http://tv.yahoo.com/
KPD Model
Dear,
You may want to test a different structural model. See the attached
reference:
http://www.page-meeting.org/pdf_assets/8835-PAGE%202007.pdf
Best regards,
Juan Jose Perez Ruixo.
Quoted reply history
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of hussain mulla
Sent: Thursday, September 27, 2007 3:14 AM
To: [email protected]
Subject: [NMusers] KPD Model
Hi
I have collected data on absolute neutrophil counts
(ANC) from children (n=30) with acute leukaemia (from
the maintenance phase of therapy, lasting up to 3
years). I also have the 6-Mercaptopurine dosing data,
but no plasma concentration data. Each child attended
clinic every 2 weeks to measure ANC and adjustment of
6-MP dose. All children received 75mg/m2. Doses
started at 100%, but were reduced to 75% or 50%
depending on ANC count. On most occasions when the ANC
level dropped below 1.0, the 6-MP was withheld until
the ANC level returned back to >1.
I have tried to implement a K-PD model as described by
P. Jacqmin et al (JPP;34(1) 2007) without success. I
provide the code and snippet of the data file (The
dose is actual dose administered and I have initiated
the PD CMP=3 by putting the initial amount in the AMT
column). I have not had any success with trying to
develop a base model; the model spectactularly fails
to describe the observed data. The population PRED
estimate is virtually the same (a constant) for all
time points for all patients (even though ANC sampling
times and doses varied between patients). The IPRED is
also the same as PRED. I have tried many different
initial estimates. What am I doing wrong? Can anyone
please diagnose the problem?
$MODEL
NPAR=6 NCOMP=3
COMP=(DOSE)
COMP=(BIOPHASE)
COMP=(ANCLEVEL)
$PK
KIN=THETA(1)*EXP(ETA(1))
KOUT=THETA(2)*EXP(ETA(2))
KDE=THETA(3)*EXP(ETA(3))
EDK50=THETA(4)*EXP(ETA(4))
GAMMA=THETA(5)*EXP(ETA(5))
KA=THETA(6)*EXP(ETA(6))
S2=1
;volume of biophase fixed
;KDE IS THE ELIMINATION RATE CONSTANT FROM BIOPHASE
;IRG IS THE (VIRTUAL)INFUSION RATE IN THE INHIBITION
FUNCTION
;EDK50 IS THE IR THAT LEADS TO 50% NHIBTION OF KIN
$DES
DADT(1)=-KA*A(1)
DADT(2)=KA*A(1)-KDE*A(2)
VIR=A(2)*KDE
IRG=VIR**GAMMA
COEF=1-IRG/(EDK50**GAMMA+IRG)
DADT(3)=KIN*COEF-KOUT*A(3)
$ERROR
Y = F + F*ERR(1)+ERR(2)
IPRED=F
$EST METHOD=0 POSTHOC PRINT=5 MAX=9999 SIG=3
MSFO=6mp2.msf NOABORT
$THETA
(0.0001,0.01,10) ;[KIN]
(0.0001,0.001,10) ;[KOUT]
(0.01,1,100) ;[KDE]
(0.1,20,100);[EDK50]
(0.001,1,5) ;[GAMMA]
(0.01,0.1,10)
ID TIME AMT DV CMT MDV EVID SS II
1 1 0.5 . 3 1 1 . .
1 1 . 0.5 3 0 0 . .
1 8 . 2.4 3 0 0 . .
1 8 20 . 1 1 1 . .
1 14 20 . 1 1 1 1 1
1 15 . 0.4 3 0 0 . .
1 22 . 0.6 3 0 0 . .
1 29 . 0.8 3 0 0 . .
1 29 21 . 1 1 1 . .
1 35 21 . 1 1 1 1 1
1 36 . 1.4 3 0 0 . .
1 36 31 . 1 1 1 . .
1 42 31 . 1 1 1 1 1
1 43 . 0.9 3 0 0 . .
1 43 21 . 1 1 1 . .
1 49 21 . 1 1 1 1 1
1 50 . 1.7 3 0 0 . .
1 50 31 . 1 1 1 . .
1 63 31 . 1 1 1 1 1
1 64 . 0.1 3 0 0 . .
1 71 . 0.4 3 0 0 . .
1 78 . 0.6 3 0 0 . .
thanks
M Sarasia
____________________________________________________________________________
________
Tonight's top picks. What will you watch tonight? Preview the hottest shows
on Yahoo! TV.
http://tv.yahoo.com/