RE: PBPK structural model in NONMEM!
Dear Ka
To me one of main issues here is the identifiability of partition
coefficients because you don't have any observations to uniquely estimate
these parameters and there is not an observable relationship between renal
excretion and partition coefficients either. In such cases incorporation of
prior knowledge may be helpful, see:
Langdon G, Gueorguieva I, Aarons L and Karlsson M (2007) Linking preclinical
and clinical whole-body physiologically based pharmacokinetic models with
prior distributions in NONMEM. Eur J Clin Pharmacol 63:485-498.
( http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=C
itation&list_uids=17345074 &db=PubMed&dopt=Citation&list_uids=17345074).
I noticed you have normalised weight to 25, are you subjects children?
Please note that if this is the case then body composition and blood flow to
organs change by age and weight; however you haven't considered eta for
partition coefficients. Renal function is a big player for your drug and you
may also need considering maturation of renal function in children:
Rhodin MM, Anderson BJ, Peters AM, Coulthard MG, Wilkins B, Cole M, Chatelut
E, Grubb A, Veal GJ, Keir MJ and Holford NH (2009) Human renal function
maturation: a quantitative description using weight and postmenstrual age.
Pediatr Nephrol 24:67-76.
( http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=C
itation&list_uids=18846389 &db=PubMed&dopt=Citation&list_uids=18846389).
In addition to incorporation of physiological knowledge into your model if
you have any physicochemical or in vitro data you can still use such
information. Your case as well as Paul Hutson case (Duration of Absorption
Time From Depot (Gut) as Covariate posted on Friday) are good examples of
joining bottom-up and top-down approaches, see:
http://www.pkuk.org.uk/ContentImages/MasoudJamei.pdf.
Regards
Masoud
Quoted reply history
From: [email protected] [mailto:[email protected]] On
Behalf Of ke fang
Sent: 13 December 2009 06:28
To: nonmem usersgroup
Subject: [NMusers] PBPK structural model in NONMEM!
Dear all,
I'm now developing a simple PBPK structural model in NONMEM. The input to
the PBPK model was a i.m. injection, and the only output was the renal
excretion.. Because the drug was a long acting formulation, so I try to
model the absorption into a fast phase and a slow phase. Now, here comes the
problem. I don't know how to get DOSE into the model. Here is the control
stream file:
$PROB WBPBPK POPULATION MODEL
$INPUT ID TIME CONC=DV AMT WT EVID CMT
$DATA ..\PBPK.TXT IGNORE=#
$SUBROUTINES ADVAN6 TOL=3
$MODEL
COMP = (LIV) ; 1 - LIVER COMPARTMENT
COMP = (KID) ; 2 - KIDNEY COMPARTMENT
COMP = (MUS) ; 3 - MUSCLE COMPARTMENT
COMP = (INJ) ; 4 - INJECTIONSITE COMPARTMENT RECEIVING DOSE
COMP = (VEN) ; 5 - VENOUS COMPARTMENT
COMP = (OUT) ; 6 - KIDNEY EXCRETION COMPARTMENT
$PK
; ------------------------ TWO ABSORBTION PHASE-------------------- ;
PHSF = 0.714 ; FAST PHASE ABSORPTION FRACTION
PHSS = 0.286 ; SLOW PHASE ABSORPTION FRACTION
KF = 0.032334 ; FAST PHASE ABSORPTION ABSORPTION RATE
KS = 0.00835 ; SLOW PHASE ABSORPTION ABSORPTION RATE
; ------------------------- BLOOD FLOWS (Q,L/HR) -------------------- ;
QVEN = 2.049*WT ;L/H/KG;
QLIV =QVEN*0.3053*(WT/25) ;L/H/KG;
QKID =QVEN*0.1398*(WT/25) ;L/H/KG;
QMUS =QVEN*0.2524*(WT/25) ;L/H/KG;
QINJ = QVEN*0.2524*(0.5/WT/25) ;L/H/KG;
;----------------------------TISSUE VOLUMES (V,L)----------------------;
VLIV =0.0294*WT ;KG;
VKID =0.004*WT ;KG;
VMUS =0.4007*WT ;KG;
VINJ =0.5 ;KG;
VVEN =0.06*WT ;KG;
;----------------------------PARTITION COEFFICIENTS-------------------;
PLIV = EXP(THETA(2))
PKID = EXP(THETA(3))
PMUS = EXP(THETA(4))
;---------------------KIDNEY EXCRETION-------------------;
TVEXCR=EXP(THETA(1))
EXCR=TVEXCR*EXP(ETA(1))
$DES
;--------------------COMPARTMENT CONCENTRATIONS-------------------;
C1 = A(1)/VLIV
C2 = A(2)/VKID
C3 = A(3)/VMUS
C4 = A(4)/VINJ
C5 = A(5)/VVEN
DADT(1)=QLIV*(C5-C1/PLIV)
DADT(2)=(QKID*(C5-C2/PKID)-C2*EXCR)
DADT(3)=QMUS*(C5-C3/PMUS)
DADT(4)=(QINJ*(C5-C4/PMUS))
DADT(5)=(QLIV*C1/PLIV+QKID*C2/PKID+QMUS*C3/PMUS+QINJ*C4/PMUS-QVEN*C5)
DADT(6)=EXCR
$ERROR
Y=A(3)/VMUS+EPS(1)
;------------------------------INITIAL
ESTIMATES--------------------------------;
$THETA(0,0.5) ; 1-EXCR
$THETA(0,1.82,10) ; 2-PLIV
$THETA(0,6.46,10) ; 3-PKID
$THETA(0,0.486) ; 4-PMUS
$OMEGA(0.5) ; 1-EXCR
$SIGMA(1) ; 1-ERROR
$EST PRINT=5 MAX=9990 POSTHOC SIGDIG=3 METHOD=1
$COV MATRIX=R
$TABLE ID TIME CONC WT AMT NOPRINT ONEHEADER FILE=PBPK.FIT
Can anyone give some hint?
Thanks in advance!
Ke, Fang