I think it's worth mentioning here that Simeoni's model is just as empirical as other tumor growth models. Patrick should also consider Gompertz or logisitic growth models. There is a large body of literature using these latter models which oncologists and cancer pharmacologists are used to and familiar with. Simeoni's model is still relatively new and, while it is couched in terms of 'mechanistic' models, there is still an empirical component to it. I would still use the Gompertz model.
The second part of Patrick's question was how do you use the output from these models to help guide human drug development. So here is what we have used. Typically you get pk in mice and rats from the ADME group. In a separate study by the cancer pharmacology group, you get tumor growth in mice without collecting pk. I then fix the pk and model the tumor growth under the dosing regimen in the xenograft study. The output from the Gompertz model is then an IC50. This is the target you need to acheive in humans, usually assuming trough. You could then allometrically scale the rodent models to get pk in humans and simulate an approximate dosing regimen to acheive a trough concentration or a 24 hour concentration above the IC50. And I don't usually do this work in NONMEM. The mice pk data are usually single time-point studies so a naive pooled approach is reasonable. You could model the xenograft data in NONMEM, but I would expect that your IC50s would be similar regardless of whether you model mean data or individual-level data. Hope this helps pete bonate Peter L. Bonate, PhD, FCP Genzyme Corporation Senior Director Clinical Pharmacology and Pharmacokinetics 4545 Horizon Hill Blvd San Antonio, TX 78229 USA [EMAIL PROTECTED] phone: 210-949-8662 fax: 210-949-8219 crackberry: 210-315-2713 -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, January 07, 2008 1:36 AM To: [EMAIL PROTECTED] Cc: [email protected] Subject: Antwort: [NMusers] PK/PD models to describe anti-ancer drug effect on the tumor volume Dear Patrick, the Simeoni model works well: Simeoni M, Magni P, Cammia C, De Nicolao G, Croci V, Pesenti E, Germani M, Poggesi I, Rocchetti M. Predictive pharmacokinetic-pharmacodynamic modeling of tumor growth kinetics in xenograft models after administration of anticancer agents. Cancer Res. 2004 Feb 1;64(3):1094-101. I've attached a code using this model below. My code used the concentrations in tumor tissue, but you can just as well use the serum concentrations. I also develowed a second model which uses linear tumor growth and considers resistance development over time, which is less pronounced if concentrations of the anti-tumor drug increase. I've attached that code as well, maybe it is useful. Considering the predictiveness of the model, in my case it seemed like the use of physiological life-span seemed to be a good corrector (I got the idea from Monro, Drug Toxicokinetics: Scope and Limitations that Arise from Species Differences in Pharmacodynamic and Carcinogenic Responses, J Pharmacokin Biopharm 22, 41-57, 1994). I.e. if you want your patients to survive 6 months, then use human average age of 75 years and mouse age of 2 years, so a mouse would have to survive 6 months/75*2= 5 days. But this might be different from drug to drug. Best wishes Nele Simeoni: $SUBROUTINES ADVAN6 TOL=3 $MODEL COMP=(GUT) COMP=(CENTRAL) COMP=(TUMOR) COMP=(PD) $PK TVCL=THETA(1) CL=TVCL ; TVV2=THETA(2) V2=TVV2 ; TVKA=THETA(3) KA=TVKA TVF1=THETA(4) F1=TVF1 TVPC=THETA(5) PC=TVPC TVKEO=THETA(6) KEO=TVKEO ; TVL0=THETA(7) L0=TVL0*EXP(ETA(1)) ; TVL1=THETA(8) L1=TVL1 TVK2=THETA(9) K2=TVK2 W0=THETA(10) F4=W0 S2=V2 K20=CL/V2 S4=1 PSI=20 ; $ERROR IPRED=F DEL=0 IF (IPRED.EQ.0) DEL=0.0001 W=F IRES=DV-IPRED IWRES=IRES/(W+DEL) Y=F+SQRT(THETA(12)*THETA(12)+THETA(11)*THETA(11)*F**2)*EPS(1) ; $DES DADT(1)= -KA*A(1) DADT(2)= KA*A(1) -K20*A(2) DADT(3)= KEO*(PC*A(2)/V2 - A(3)) DADT(4)= L0*A(4)/(1+(L0/L1*A(4))**PSI)**(1/PSI)-K2*A(3)*A(4) own code: $SUBROUTINES ADVAN8 TOL=3 $MODEL COMP=(GUT) COMP=(CENTRAL) COMP=(TUMOR) COMP=(PD) COMP=(AUC) $PK TVCL=THETA(1) CL=TVCL*EXP(ETA(1)) ; TVV2=THETA(2) V2=TVV2 ; TVKA=THETA(3) KA=TVKA TVF1=THETA(4) F1=TVF1 TVPC=THETA(5) PC=TVPC TVKEO=THETA(6) KEO=TVKEO TVW0=THETA(7) F4=TVW0 W0=F4 ; dataset: put a dummy dose of 1 into compartment 4, initial tumor weight is then estimated as F4 TVKRES=THETA(8) KRES=TVKRES TVIC50=THETA(9) IC50=TVIC50 TVKIN=THETA(10) KIN=TVKIN TVRED=THETA(13) RED=TVRED S2=V2 K20=CL/V2 S4=1 ; $ERROR IPRED=F DEL=0 IF (IPRED.EQ.0) DEL=0.0001 W=F IRES=DV-IPRED IWRES=IRES/(W+DEL) Y=F+SQRT(THETA(12)*THETA(12)+THETA(11)*THETA(11)*F**2)*EPS(1) ; $DES CAV=A(5)/(T+0.01) RES=IC50*EXP(KRES*T*(1-CAV/(RED+CAV))) DADT(1)= -KA*A(1) DADT(2)= KA*A(1) -K20*A(2) DADT(3)= KEO*(PC*A(2)/V2 - A(3)) DADT(4)= KIN*(1-A(3)/(RES+A(3))) DADT(5)= A(3) _________________________ Dr. Nele Plock Bayer Schering Pharma AG Drug Metabolism & Pharmacokinetics Development Pharmacokinetics Scientific Expert Development Pharmacokinetics D- 13342 Berlin Phone : +49-30-468 15146 Fax: +49-30-468 95146 [EMAIL PROTECTED] http://www.bayerscheringpharma.de Vorstand: Arthur J. Higgins, Vorsitzender | Werner Baumann, Andreas Busch, Ulrich Köstlin, Kemal Malik, Gunnar Riemann Vorsitzender des Aufsichtsrats: Werner Wenning Sitz der Gesellschaft: Berlin | Eintragung: Amtsgericht Charlottenburg 93 HRB 283 Patrick Zhou <[EMAIL PROTECTED]> Gesendet von: [EMAIL PROTECTED] 06.01.2008 20:50 An [email protected] Kopie Thema [NMusers] PK/PD models to describe anti-ancer drug effect on the tumor volume Dear NMusers, Does anyone aware any good published example (or non-public if you are willing to share) of modeling the anti-cancer drug effect on the tumor volume in nude mice model? And how this is normally used in the human dose projection, and any published work of such? Please advice. Thank you very much. Patrick Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now.
