Dear NONMEM users,
I have a data set from a bioavailablity study (2×2 study design). 18
people took a single dose of a marketed drug (marked as R), then after a
wash-out period, took a single dose the corresponding generic drug (marked T).
I want to estimate the PK parameters of T and R seperately. One approach we
handle the dataset is to consider the 18 people as 36 people. Is it right?
I still want to handle them as 18 people. I read something useful in the NONMEM
guides on EVID, but still comfused about the data file structure and the NONMEM
codes.
Thank you very much ahead for any comments.
Xipei Wang
Beijing, China
Part of my codes and data are as below:
data set:
ID TIME AMT DV EVID DRUG
1 0 100 . 1 1
1 0.3 . 0.7 0 1
1 0.5 . 2 0 1
....
1 48 . 0.6 0 1
1 0 . . 3 .
1 0 100 . 1 2
1 0.3 . 0.3 0 2
1 0.5 . 1.6 0 2
....
NONMEM CODES：
$INPUT ID TIME AMT DV EVID DRUG
$DATA sample.csv
$SUBROUTINES ADVAN2 TRANS2
$PK
KA=THETA(1)*EXP(ETA(1))
CL=THETA(2)*EXP(ETA(2))
V=THETA(3)*EXP(ETA(3))
IF (DRUG.EQ.2) THEN
KA=THETA(4)*EXP(ETA(4))
CL=THETA(5)*EXP(ETA(5))
V=THETA(6)*EXP(ETA(6))
ENDIF
S2=V/1000
$ERROR
Y=F+ERR(1)
$THETA
(0, 2) ;KA
(0, 8) ;CL
(0, 4) ;V
(0, 2) ;
(0, 8) ;
(0, 4) ;
$OMEGA
0.04
0.04
0.04
0.04
0.04
0.04
$SIGMA
0.04
$ESTIMATION MAXEVAL=9999 METHOD=1
NONMEM codes about PK analysis from a cross-over study design
Dear Xipei
I recommend you use EVID=4 at time=0 in the 2nd period on the dataset. Also you
consider the relative BA(F1 parameter).
ID TIME AMT DV EVID DRUG
1 0 100 . 1 1
1 0.3 . 0.7 0 1
1 0.5 . 2 0 1
....
1 48 . 0.6 0 1
1 0 100 . 4 2
1 0.3 . 0.3 0 2
1 0.5 . 1.6 0 2
$PK
KA=THETA(1)*EXP(ETA(1))
CL=THETA(2)*EXP(ETA(2))
V=THETA(3)*EXP(ETA(3))
F1=1
IF (DRUG.EQ.2) THEN
KA=THETA(4)*EXP(ETA(4))
CL=THETA(5)*EXP(ETA(5))
V=THETA(6)*EXP(ETA(6))
F1=THETA(7) *EXP(ETA(7)); Relative BA parameter
ENDIF
Thanks!
Yoshi
Quoted reply history
From: [email protected] [mailto:[email protected]] On
Behalf Of wangxipei
Sent: Wednesday, March 16, 2011 2:51 PM
To: nmusers
Subject: [NMusers] NONMEM codes about PK analysis from a cross-over study design
Dear NONMEM users,
I have a data set from a bioavailablity study (2×2 study design). 18
people took a single dose of a marketed drug (marked as R), then after a
wash-out period, took a single dose the corresponding generic drug (marked T).
I want to estimate the PK parameters of T and R seperately. One approach we
handle the dataset is to consider the 18 people as 36 people. Is it right?
I still want to handle them as 18 people. I read something useful in the NONMEM
guides on EVID, but still comfused about the data file structure and the NONMEM
codes.
Thank you very much ahead for any comments.
Xipei Wang
Beijing, China
Part of my codes and data are as below:
data set:
ID TIME AMT DV EVID DRUG
1 0 100 . 1 1
1 0.3 . 0.7 0 1
1 0.5 . 2 0 1
....
1 48 . 0.6 0 1
1 0 . . 3 .
1 0 100 . 1 2
1 0.3 . 0.3 0 2
1 0.5 . 1.6 0 2
....
NONMEM CODES：
$INPUT ID TIME AMT DV EVID DRUG
$DATA sample.csv
$SUBROUTINES ADVAN2 TRANS2
$PK
KA=THETA(1)*EXP(ETA(1))
CL=THETA(2)*EXP(ETA(2))
V=THETA(3)*EXP(ETA(3))
IF (DRUG.EQ.2) THEN
KA=THETA(4)*EXP(ETA(4))
CL=THETA(5)*EXP(ETA(5))
V=THETA(6)*EXP(ETA(6))
ENDIF
S2=V/1000
$ERROR
Y=F+ERR(1)
$THETA
(0, 2) ;KA
(0, 8) ;CL
(0, 4) ;V
(0, 2) ;
(0, 8) ;
(0, 4) ;
$OMEGA
0.04
0.04
0.04
0.04
0.04
0.04
$SIGMA
0.04
$ESTIMATION MAXEVAL=9999 METHOD=1
With Yoshi’s suggestion, you may try FORM as a covariate (e.g. if
formulation is R, FORM=0 and T, FORM=1) and put it in the code as following
But I doubt all of theta 4,5,6,7 are necessary at this situation. If you
want to consider any change in PK parameters by periods, use of IOV in the
ETA terms may be better answer.
Good Luck
Dong-Seok Yim
$PK
KA=(THETA(1) + FORM*THETA(4))*EXP(ETA(1))
CL=(THETA(2)+ FORM*THETA(5))*EXP(ETA(2))
V=(THETA(3) + FORM*THETA(6))*EXP(ETA(3))
F1=1 + FORM*THETA(7)
Quoted reply history
From: [email protected] [mailto:[email protected]] On
Behalf Of [email protected]
Sent: Wednesday, March 16, 2011 3:19 PM
To: [email protected]; [email protected]
Subject: RE: [NMusers] NONMEM codes about PK analysis from a cross-over
study design
Dear Xipei
I recommend you use EVID=4 at time=0 in the 2nd period on the dataset. Also
you consider the relative BA(F1 parameter).
ID TIME AMT DV EVID DRUG
1 0 100 . 1 1
1 0.3 . 0.7 0 1
1 0.5 . 2 0 1
....
1 48 . 0.6 0 1
1 0 100 . 4 2
1 0.3 . 0.3 0 2
1 0.5 . 1.6 0 2
$PK
KA=THETA(1)*EXP(ETA(1))
CL=THETA(2)*EXP(ETA(2))
V=THETA(3)*EXP(ETA(3))
F1=1
IF (DRUG.EQ.2) THEN
KA=THETA(4)*EXP(ETA(4))
CL=THETA(5)*EXP(ETA(5))
V=THETA(6)*EXP(ETA(6))
F1=THETA(7) *EXP(ETA(7)); Relative BA parameter
ENDIF
Thanks!
Yoshi
From: [email protected] [mailto:[email protected]] On
Behalf Of wangxipei
Sent: Wednesday, March 16, 2011 2:51 PM
To: nmusers
Subject: [NMusers] NONMEM codes about PK analysis from a cross-over study
design
Dear NONMEM users,
I have a data set from a bioavailablity study (2×2 study design). 18
people took a single dose of a marketed drug (marked as R), then after a
wash-out period, took a single dose the corresponding generic drug (marked
T). I want to estimate the PK parameters of T and R seperately. One
approach we handle the dataset is to consider the 18 people as 36 people.
Is it right?
I still want to handle them as 18 people. I read something useful in the
NONMEM guides on EVID, but still comfused about the data file structure and
the NONMEM codes.
Thank you very much ahead for any comments.
Xipei Wang
Beijing, China
Part of my codes and data are as below:
data set:
ID TIME AMT DV EVID DRUG
1 0 100 . 1 1
1 0.3 . 0.7 0 1
1 0.5 . 2 0 1
....
1 48 . 0.6 0 1
1 0 . . 3 .
1 0 100 . 1 2
1 0.3 . 0.3 0 2
1 0.5 . 1.6 0 2
....
NONMEM CODES：
$INPUT ID TIME AMT DV EVID DRUG
$DATA sample.csv
$SUBROUTINES ADVAN2 TRANS2
$PK
KA=THETA(1)*EXP(ETA(1))
CL=THETA(2)*EXP(ETA(2))
V=THETA(3)*EXP(ETA(3))
IF (DRUG.EQ.2) THEN
KA=THETA(4)*EXP(ETA(4))
CL=THETA(5)*EXP(ETA(5))
V=THETA(6)*EXP(ETA(6))
ENDIF
S2=V/1000
$ERROR
Y=F+ERR(1)
$THETA
(0, 2) ;KA
(0, 8) ;CL
(0, 4) ;V
(0, 2) ;
(0, 8) ;
(0, 4) ;
$OMEGA
0.04
0.04
0.04
0.04
0.04
0.04
$SIGMA
0.04
$ESTIMATION MAXEVAL=9999 METHOD=1