From: "Ritu Lal" Ritu.Lal@XenoPort.com
Subject: [NMusers] FW:  question for absorption modeling
Date: Fri, 4 Nov 2005 08:41:54 -0800

Dear NM users,

I am trying the following models for modeling the absorption phase of my
compound. This is data for a drug administered orally. Currently, I am using
ADVAN2 (one compt. oral absorption model with first order absorption and lag
time), but there is some model misspecification in the absorption phase. 

These are the 2 other models I would like to try:
1. sequential first and zero-order abn with lag time, one compt model
2. only zero order absorption with lag time, one cpmpt model


How do I set up my data file (how do I enter RATE for the zero-order absorption?)
and control file for these 2 situations? Can I continue using the library models
or do I need to switch to user written subroutines?

Thanks,
Ritu 
Ritu Lal, Ph.D. 
Sr. Director, Pharmacokinetics and Drug Metabolism  
Xenoport
3410 Central Expressway 
Santa Clara, CA 95051 
Phone (408)616-7199 
Cell (408)483-3741 
Fax (408)616-7212 
e-mail ritu.lal@xenoport.com 

_______________________________________________________

From:     Justin Wilkins justin.wilkins@farmbio.uu.se
Subject:     Re: [NMusers] question for absorption modeling
Date:     Mon, 07 Nov 2005 10:42:58 +0100

Hi there,

I don't think there'll be any problem using the library routines for this. Specifying RATE=-2
indicates that the durations (as opposed to rates) of zero-order bolus doses will be calculated
by PK. For ADVAN2/TRANS2 you might use a data file structure similar to the following example:

#ID     RATE    TIME    DV      MDV     AMT     EVID
      1      -2       0       0       1     600    1
      1       0    1.55    3.98       0       0    0
      1       0    3.13    4.32       0       0    0
      1       0    4.78    3.22       0       0    0
      2      -2       0       0       1     600    1
      2       0    3.18    0.55       0       0    0
      2       0    4.48    3.62       0       0    0

... and so on.

You would specify the D1 (duration of zero order infusion) parameter in the model
specification file, similarly to the below (while retaining KA):

TVCL  = THETA(1)          ; CL
CL    = TVCL*EXP(ETA(1))  ; CL

TVV   = THETA(2)          ; V2
V     = TVV*EXP(ETA(2))   ; V2

TVKA  = THETA(3)          ; KA
KA    = TVKA*EXP(ETA(3))  ; KA

TVD1  = THETA(4)          ; D1
D1    = TVD1*EXP(ETA(4))  ; D1

Another option to consider when dealing with awkward absorption and associated lag times
might be the transit compartment model - it's still in manuscript but a broad outline
was presented at PAGE 2004:

PAGE 13 (2004) Abstr 513 [www.page-meeting.org/?abstract=513]

Best regards
Justin 

_______________________________________________________