Working within NLME to implement a semi-mechanistic PK/PD model to describe the chemotherapy induced myelosuppression of various chemotherapeutics. Interested in using the PK/PD myelosuppression model described by Friberg et al. (J Clin Oncol, 20: 4713, 2002) which includes a proliferating compartment that is sensitive to the drug, 3 transit compartments reflecting cell maturation and a blood cell compartment. Want to know if it is possible to set up this model in NLME. Thanks and kind regards Bob [file name=Friberg_myelsupp_model_2002.pdf size=294601]Certara | Drug Development Solutions (288 KB)
Bob, I’m sorry your question seems to have been overlooked - I am overdue to send out a digest. I don’t see any immediate reason why this should not be possible in NLME, there are several examples with transit compartment models on the forum already that you should be able to find by typing ‘transit’ if that helps you get started. e.g. Certara | Drug Development Solutions Personally I would start with the graphical editor to describe the model on p4715 of the paper you attached and then edit it further in Textual mode. If you have some problems then post your project/code back here and we can try and take it from there. Simon.
Dear Bob and Simon, please find below working code I tested it in simulation mode with dummy one line input id time CObs RATE ProlObs CircObs Aa1 ADDL II 1 0 100 13.88889 4.34 4.94 1000 7 168 it still needs work but it illustrate how e can code transit comparments with a feedback loop. we can integrate the PK later on currently it has a biomarker linked to the ANC model from a recent publication: http://www.nature.com/psp/journal/v2/n12/full/psp201362a.html Samer test(){ sequence{ BM=BMBASE; Circ = BASE; STEM=KEBASE/KP ; Transit1=KEBASE/KP ; Transit2=KEBASE/KP; Transit3=KEBASE/KP } deriv(BM = KIN*(1-EFF) - KOUTBM ) deriv(Circ = -KE Circ + KPTransit3) deriv(STEM = -KP STEM + KPSTEM(1-DRUG)*(BASE/Circ)0.362) deriv(Transit1= -KP Transit1 + KPSTEM ) deriv(Transit2= -KP Transit2 + KPTransit1 ) deriv(Transit3= -KP Transit3 + KPTransit2 ) error(CEps2 = 0.001) observe(CircObs = (Circ)+CEps2) BMCHANGE = ((BM-BMBASE)/BMBASE) DRUG = 0.520(-BMCHANGE)/(1.0099+(-BMCHANGE)) EFF = 1(50/32.819)/(1.0099+(50/32.819)) stparm(KOUT = tvKout ) stparm(KIN = tvKout * BMBASE) stparm(BMBASE =tvBM0 ) fixef(tvBM0 = c(,42554, )) fixef(tvKout = c(,0.002719313, )) FN = (BASE/Circ)*0.362 KP=4/MTT KE=log(2)/kcirc stparm(MTT = tvMTT * exp(nMTT)) stparm(kcirc = tvkcirc * exp(nkcirc)) stparm(BASE = tvCirc0 exp(nBASE) ) fixef(tvCirc0 = c(,4.94, )) fixef(tvMTT = c(, 248, )) fixef(tvkcirc = c(, 7, )) ranef(diag( nMTT,nkcirc,nBASE) = c( 1,1,1)) }
Hi,
Has anybody done Friberg et al.2002 model in NLME Phoenix yet?
Thanks
SK
Hi SK,
not sure if this is the right one, but we have worked out a Friberg model as part of our PML School webinars. Please take a look here:
https://support.certara.com/forums/topic/1097-qa-from-lesson-20-transduction-modeling/
Let us know if there are specific questions.
Bernd
link is broken but yes we have done the Friberg model to model, platelets, neutrophils and more
in phoenix you also have access to real distributed delay models