Prednisone and Prednisolone PPK

Dear all, I am working on modelling prednisone and prednisolone PPK in humans given a PO dose of prednisone. Note that prednisone is converted to prednisolone during the first-pass and prednisolone is converted back to prednisone (reversible metabolism). I used a simple 2-cmt PK model (see the code below) that seems to work. However, the model does not appear to be stable and sensitive to initial estimates . Is there an issue of "identifiability" in this model? does anyone has already worked on PPK of this drug? Furthermore, I would like to model the inhibition of conversion of prednisone to prednisolone as function of time and test drug concentration. Any leads much appreciated. Regards, - Chandra ---------------------------------------------------------------------- $INPUT C ID TIME DV AMT CMT EVID MDV $DATA pred2.CSV IGNORE=C $SUBROUTINES ADVAN8 TRANS1 TOL=5 $MODEL NPAR=7 NCOMP=3 COMP=(DEPOT,DEFDOSE) COMP=(PARENT) COMP=(METAB) $PK KA=THETA(1)*EXP(ETA(1)) VP=THETA(2)*EXP(ETA(2)) CLP=THETA(3)*EXP(ETA(3)) VMT=THETA(4)*EXP(ETA(4)) KF=THETA(5)*EXP(ETA(5)) KB=THETA(6)*EXP(ETA(6)) CLM=THETA(7)*EXP(ETA(7)) S2=VP S3=VMT $DES DADT(1)=-KA*A(1) DADT(2)=KA*A(1)-KF*A(2)+KB*A(3)-CLP*A(2)/VP DADT(3)=KF*A(2)-KB*A(3)-CLM*A(3)/VMT $ERROR FLAG=0 IF(AMT.NE.0) FLAG=1 IPRED=LOG(F+FLAG) R1=0 IF (CMT.EQ.2) R1=1 R2=0 IF (CMT.EQ.3) R2=1 Y2=IPRED+ERR(1) Y3=IPRED+ERR(2) Y=R1*Y2+R2*Y3 IRES=EXP(DV)-EXP(IPRED) $THETA ....... $OMEGA ....... $SIGMA ....... $EST SIG=5 METHOD=1 PRINT=1 MAX99 POSTHOC NOABORT

Chandra, You must remember that prednisolone is bound in a saturable manner to plasma proteins resulting in the non-linear PK. Prednisone plasma protein binding however is not satuable. Modeling the unbound prednisolone data or estimating unbound conc from total conc and binding proteins (CBG and albumin) may therefore help the model. Also there is identifiability problem with the prednisolone to prednisone conversion. Dr. Jusko has successfully modeled this conversion in animal models but the animals were given the drug either as prednisolone or prednisone. Hope it helps. Fatemeh Akhlaghi >===== Original Message From "Chandrasekhar Udata" <Udatac >Dear all, > >I am working on modelling prednisone and prednisolone PPK in humans given a PO dose of prednisone. Note that prednisone is converted to prednisolone during the first-pass and prednisolone is converted back to prednisone (reversible metabolism). I used a simple 2-cmt PK model (see the code below) that seems to work. However, the model does not appear to be stable and sensitive to initial estimates . Is there an issue of "identifiability" in this model? does anyone has already worked on PPK of this drug? Furthermore, I would like to model the inhibition of conversion of prednisone to prednisolone as function of time and test drug concentration. Any leads much appreciated. > >Regards, >- Chandra > > > >---------------------------------------------------------------------- > >$INPUT C ID TIME DV AMT CMT EVID MDV >$DATA pred2.CSV IGNORE=C >$SUBROUTINES ADVAN8 TRANS1 TOL=5 >$MODEL NPAR=7 NCOMP=3 > COMP=(DEPOT,DEFDOSE) > COMP=(PARENT) > COMP=(METAB) > >$PK > > KA=THETA(1)*EXP(ETA(1)) > VP=THETA(2)*EXP(ETA(2)) > CLP=THETA(3)*EXP(ETA(3)) > VMT=THETA(4)*EXP(ETA(4)) > KF=THETA(5)*EXP(ETA(5)) > KB=THETA(6)*EXP(ETA(6)) > CLM=THETA(7)*EXP(ETA(7)) > > S2=VP > S3=VMT > >$DES > > DADT(1)=-KA*A(1) > DADT(2)=KA*A(1)-KF*A(2)+KB*A(3)-CLP*A(2)/VP > DADT(3)=KF*A(2)-KB*A(3)-CLM*A(3)/VMT > >$ERROR > FLAG=0 > IF(AMT.NE.0) FLAG=1 > IPRED=LOG(F+FLAG) > R1=0 > IF (CMT.EQ.2) R1=1 > R2=0 > IF (CMT.EQ.3) R2=1 > Y2=IPRED+ERR(1) > Y3=IPRED+ERR(2) > Y=R1*Y2+R2*Y3 > IRES=EXP(DV)-EXP(IPRED) > >$THETA ....... > >$OMEGA ....... > >$SIGMA ....... > >$EST SIG=5 METHOD=1 PRINT=1 MAX=9999 POSTHOC NOABORT Fatemeh Akhlaghi, PharmD, PhD Associate Professor in Pharmacokinetics Biomedical and Pharmaceutical Sciences (BPS) University of Rhode Island 125 Fogarty Hall, 41 Lower College Road Kingston, RI 02881, USA Phone/Fax: (401) 874 9205/(401) 874 2181 Email: fatemeh Laboratory Website: http://www.uri.edu/pharmacy/faculty/aps/akhlaghi

Dear all, I am working on modelling prednisone and prednisolone PPK in humans given a PO dose of prednisone. Note that prednisone is converted to prednisolone during the first-pass and prednisolone is converted back to prednisone (reversible metabolism). I used a simple 2-cmt PK model (see the code below) that seems to work. However, the model does not appear to be stable and sensitive to initial estimates . Is there an issue of "identifiability" in this model? does anyone has already worked on PPK of this drug? Furthermore, I would like to model the inhibition of conversion of prednisone to prednisolone as function of time and test drug concentration. Any leads much appreciated. Regards, - Chandra ---------------------------------------------------------------------- $INPUT C ID TIME DV AMT CMT EVID MDV $DATA pred2.CSV IGNORE=C $SUBROUTINES ADVAN8 TRANS1 TOL=5 $MODEL NPAR=7 NCOMP=3 COMP=(DEPOT,DEFDOSE) COMP=(PARENT) COMP=(METAB) $PK KA=THETA(1)*EXP(ETA(1)) VP=THETA(2)*EXP(ETA(2)) CLP=THETA(3)*EXP(ETA(3)) VMT=THETA(4)*EXP(ETA(4)) KF=THETA(5)*EXP(ETA(5)) KB=THETA(6)*EXP(ETA(6)) CLM=THETA(7)*EXP(ETA(7)) S2=VP S3=VMT $DES DADT(1)=-KA*A(1) DADT(2)=KA*A(1)-KF*A(2)+KB*A(3)-CLP*A(2)/VP DADT(3)=KF*A(2)-KB*A(3)-CLM*A(3)/VMT $ERROR FLAG=0 IF(AMT.NE.0) FLAG=1 IPRED=LOG(F+FLAG) R1=0 IF (CMT.EQ.2) R1=1 R2=0 IF (CMT.EQ.3) R2=1 Y2=IPRED+ERR(1) Y3=IPRED+ERR(2) Y=R1*Y2+R2*Y3 IRES=EXP(DV)-EXP(IPRED) $THETA ....... $OMEGA ....... $SIGMA ....... $EST SIG=5 METHOD=1 PRINT=1 MAX=9999 POSTHOC NOABORT

Chandra, You must remember that prednisolone is bound in a saturable manner to plasma proteins resulting in the non-linear PK. Prednisone plasma protein binding however is not satuable. Modeling the unbound prednisolone data or estimating unbound conc from total conc and binding proteins (CBG and albumin) may therefore help the model. Also there is identifiability problem with the prednisolone to prednisone conversion. Dr. Jusko has successfully modeled this conversion in animal models but the animals were given the drug either as prednisolone or prednisone. Hope it helps. Fatemeh Akhlaghi >===== Original Message From "Chandrasekhar Udata" <[EMAIL PROTECTED]> ===== >Dear all, > >I am working on modelling prednisone and prednisolone PPK in humans given a PO dose of prednisone. Note that prednisone is converted to prednisolone during the first-pass and prednisolone is converted back to prednisone (reversible metabolism). I used a simple 2-cmt PK model (see the code below) that seems to work. However, the model does not appear to be stable and sensitive to initial estimates . Is there an issue of "identifiability" in this model? does anyone has already worked on PPK of this drug? Furthermore, I would like to model the inhibition of conversion of prednisone to prednisolone as function of time and test drug concentration. Any leads much appreciated. > >Regards, >- Chandra > > > >---------------------------------------------------------------------- > >$INPUT C ID TIME DV AMT CMT EVID MDV >$DATA pred2.CSV IGNORE=C >$SUBROUTINES ADVAN8 TRANS1 TOL=5 >$MODEL NPAR=7 NCOMP=3 > COMP=(DEPOT,DEFDOSE) > COMP=(PARENT) > COMP=(METAB) > >$PK > > KA=THETA(1)*EXP(ETA(1)) > VP=THETA(2)*EXP(ETA(2)) > CLP=THETA(3)*EXP(ETA(3)) > VMT=THETA(4)*EXP(ETA(4)) > KF=THETA(5)*EXP(ETA(5)) > KB=THETA(6)*EXP(ETA(6)) > CLM=THETA(7)*EXP(ETA(7)) > > S2=VP > S3=VMT > >$DES > > DADT(1)=-KA*A(1) > DADT(2)=KA*A(1)-KF*A(2)+KB*A(3)-CLP*A(2)/VP > DADT(3)=KF*A(2)-KB*A(3)-CLM*A(3)/VMT > >$ERROR > FLAG=0 > IF(AMT.NE.0) FLAG=1 > IPRED=LOG(F+FLAG) > R1=0 > IF (CMT.EQ.2) R1=1 > R2=0 > IF (CMT.EQ.3) R2=1 > Y2=IPRED+ERR(1) > Y3=IPRED+ERR(2) > Y=R1*Y2+R2*Y3 > IRES=EXP(DV)-EXP(IPRED) > >$THETA ....... > >$OMEGA ....... > >$SIGMA ....... > >$EST SIG=5 METHOD=1 PRINT=1 MAX=9999 POSTHOC NOABORT Fatemeh Akhlaghi, PharmD, PhD Associate Professor in Pharmacokinetics Biomedical and Pharmaceutical Sciences (BPS) University of Rhode Island 125 Fogarty Hall, 41 Lower College Road Kingston, RI 02881, USA Phone/Fax: (401) 874 9205/(401) 874 2181 Email: [EMAIL PROTECTED] Laboratory Website: http://www.uri.edu/pharmacy/faculty/aps/akhlaghi

Hi, Chandra, 1: You may need to use free steroid concentrations to model. (in vitro protein binding information is available in Dr. Jusko's papers (early 90s')) 2: You will have identifiability problem for your model if only prednisolone or prednione is given (even you measured both steroid concentrations.) You can fix a few parameters in order to get your model stable. We published a paper last year for prednisolone and prednisone PK/PD (not POP approach), and you may get some information from it. ( http://www.ncbi.nlm.nih.gov/pubmed/17318442) Good luck, Jian

setting a nonlinear CL from prednisone to prednisolone is probably a reasonable fix, a hill equation instead of just estimating a Cl could work, although you'd probably need to fix one of the terms in the hill equation since data is limited Alice. Alice I Nichols, PhD Sr Director Early Development and Clinical Pharmacology Wyeth Research 500 Arcola Rd Collegeville, PA 19426 tel: 484-865-8741/ fax: 484-865-9075 [EMAIL PROTECTED] >>> "Chandrasekhar Udata" <[EMAIL PROTECTED]> 12/4/2008 7:57:16 PM >>> Dear all, I am working on modelling prednisone and prednisolone PPK in humans given a PO dose of prednisone. Note that prednisone is converted to prednisolone during the first-pass and prednisolone is converted back to prednisone (reversible metabolism). I used a simple 2-cmt PK model (see the code below) that seems to work. However, the model does not appear to be stable and sensitive to initial estimates . Is there an issue of "identifiability" in this model? does anyone has already worked on PPK of this drug? Furthermore, I would like to model the inhibition of conversion of prednisone to prednisolone as function of time and test drug concentration. Any leads much appreciated. Regards, - Chandra ---------------------------------------------------------------------- $INPUT C ID TIME DV AMT CMT EVID MDV $DATA pred2.CSV IGNORE=C $SUBROUTINES ADVAN8 TRANS1 TOL=5 $MODEL NPAR=7 NCOMP=3 COMP=(DEPOT,DEFDOSE) COMP=(PARENT) COMP=(METAB) $PK KA=THETA(1)*EXP(ETA(1)) VP=THETA(2)*EXP(ETA(2)) CLP=THETA(3)*EXP(ETA(3)) VMT=THETA(4)*EXP(ETA(4)) KF=THETA(5)*EXP(ETA(5)) KB=THETA(6)*EXP(ETA(6)) CLM=THETA(7)*EXP(ETA(7)) S2=VP S3=VMT $DES DADT(1)=-KA*A(1) DADT(2)=KA*A(1)-KF*A(2)+KB*A(3)-CLP*A(2)/VP DADT(3)=KF*A(2)-KB*A(3)-CLM*A(3)/VMT $ERROR FLAG=0 IF(AMT.NE.0) FLAG=1 IPRED=LOG(F+FLAG) R1=0 IF (CMT.EQ.2) R1=1 R2=0 IF (CMT.EQ.3) R2=1 Y2=IPRED+ERR(1) Y3=IPRED+ERR(2) Y=R1*Y2+R2*Y3 IRES=EXP(DV)-EXP(IPRED) $THETA ....... $OMEGA ....... $SIGMA ....... $EST SIG=5 METHOD=1 PRINT=1 MAX=9999 POSTHOC NOABORT Alice Nichols.vcf Description: Binary data