From: Weijiang Zhang <ZHANGW@mail.rx.uga.edu>

Subject: Help: oral and ig bioavailability by NONMEM

Date: Tue, 28 Aug 2001 02:46:10 GMT

 

Dear NONMEM users:

When I tried to use NONMEM to analyze my data, there is a

problem I don't know how to do with it. My study is about

giving a compound to six subjects by three different

administration routes: iv, ig and oral. I used subroutines

ADVAN4 TRANS3, that's two compartment with first order

absorption. Everything works out fine except the

bioavailability.

 

Following is my code for absorption rate (KA)(the

absorption is rapid for both routes) and bioavailability:

 

$PK

X1=0

X2=0

IF (WZ.EQ.1) THEN ;WZ is the indicator, WZ=1, oral dose

X1=1

ENDIF

IF(WZ.EQ.2) THEN ;WZ=2, ig dose

X2=1

ENDIF

KA = X1*THETA(5) + X2*THETA(6)

TVF1 = X1*THETA(7) + X2*THETA(8)

F1=TVF1 *(1+X1*ETA(2)+X2*ETA(3)) + (1-X1-X2)*THETA(9)

 

Most parameters are well estimated, however, the F1 for oral

(theta7) was estimated to be 1.52, for ig (theta8) was 0.58,

which were very different from the number (0.95 and 0.70) I

got by noncompartmental analysis using Winnonlin. I've

already double checked my data set to make sure it's

correct.

 

My first question is: Is that right to use F1 here? or F2?

Second: Is my code correct to estimate bioavailability for

oral and ig in this model? What can I do to get the "right"

estimation?

 

Any help will be highly appreciated.

 

Thank you so much for your time and attention.

 

Sincerely,

Weijiang Zhang

UGA

 

******

 

From: "Rik Schoemaker" <RS@chdr.nl>

Subject: Re: Help: oral and ig bioavailability by NONMEM

Date: Tue, 28 Aug 2001 10:09:33 +0200

 

Dear Weijiang,

 

I'm a bit puzzled about your THETA(9) parameter; it seems to be=20

related to "bioavailability" for the IV dose (when both X1 and X2=20

are 0) which however should not operate on compartment 1 because=20

the IV dose is introduced in compartment 2...Why is it there?

Could you try and see if the following syntax produces correct results

(after checking that the IV dose indeed enters comp2):

 

instead of:

TVF1 =3D X1*THETA(7) + X2*THETA(8)

F1=3DTVF1 *(1+X1*ETA(2)+X2*ETA(3)) + (1-X1-X2)*THETA(9)=20

use:

F1=3D X1*THETA(7)*(1+ETA(2)) + X2*THETA(8)*(1+ETA(3))

and drop THETA(9)...

 

Did you try using FOCE estimation with interaction?

 

Good luck,

 

Rik

 

Rik C. Schoemaker, PhD

Biostatistician - pharmacokineticist

 

Centre for Human Drug Research (CHDR)

Zernikedreef 10, 2333CL Leiden, Netherlands

tel +31 (0) 71 5246417; fax +31 (0) 71 5246499

email rs@chdr.nl; http://www.chdr.nl

 

******

 

From: "Piotrovskij, Vladimir [JanBe]" <VPIOTROV@janbe.jnj.com>

Subject: RE: Help: oral and ig bioavailability by NONMEM

Date: Tue, 28 Aug 2001 12:37:24 +0200

 

Weijiang,

 

I agree with Rik: THETA(9) seems to be redundant. Another potential problem:

too few individuals. You cannot estimate so many random effects having only

6 subjects. Fixed effect estimates can also be affected. I think you have to

fit each subject separately. Interoccasion variability should be taken into

account, too.

 

Best regards,

Vladimir

Centre for Human Drug Research (CHDR)

Zernikedreef 10, 2333CL Leiden, Netherlands

tel +31 (0) 71 5246417; fax +31 (0) 71 5246499

email rs@chdr.nl <mailto:rs@chdr.nl> ; http://www.chdr.nl

<http://www.chdr.nl>