Simultaneous PK and Cell Life Span PD modeling

Dear group: Perez-Ruixo and Jusko et al. published NONMEM code for a basic cell lifespan model in 2005 (Population Cell Life Span Models for Effects of Drugs Following Indirect Mechanisms of Action, J Pharmacokinet Pharmacodyn. 2005 Dec;32(5-6):767-93), which has PK parameters modeled first, which are then used for life span PD modeling. I was wondering if anyone in this group has attempted to model PK and the cell life span PD model simultaneously. Please see followiing part of my control streams, which yielded the error message: "167 $ MODEL: MORE THAN ONE DEFAULT COMPARTMENT FOR DOSE OR OBSERVATION". Are there any work-arounds for this? Many thanks! Wei-jian $MODEL COMPARTMENT=(DEPOT) ; [CMT 1] COMPARTMENT=(CENTROL, DEFOBS) ; [CMT 2] COMPARTMENT=(PERIPH) ; [CMT 3] COMPARTMENT=(RESPONSE) ; [CMT 4] COMPARTMENT=(DDEPOT) ; [CMT 5; delay] COMPARTMENT=(DCENTROL, DEFOBS) ; [CMT 6; delay] COMPARTMENT=(DPERIPH) ; [CMT 7; delay] $PK CALLFL=-2 ;Call with every event record and at additional and lagged dose times ;DEFINE PK PARAMETERS TVKA = THETA(1) KA = TVKA*EXP(ETA(1)) TVCL = THETA(2) CL = TVCL*EXP(ETA(2)) TVV2 = THETA(3) V2 = TVV2*EXP(ETA(3)) TVQ = THETA(4) Q = TVQ*EXP(ETA(4)) TVV3 = THETA(5) V3 = TVV3*EXP(ETA(5)) TVF1 = THETA(6) F1 = TVF1*EXP(ETA(6)) S2 = V2 S3 = V3 ;DEFINE PD PARAMETERS KIN = THETA(7)*EXP(ETA(7)) ALAG6 = THETA(8)*EXP(ETA(8)) ALAG5 = ALAG6 ALAG7 = ALAG6 F4 = KIN*ALAG6 EMAX = THETA(9)*EXP(ETA(9)) IC50 = THETA(10)*EXP(ETA(10)) GAMA = THETA(11)*EXP(ETA(11)) $DES C2 = A(2)/V2 E1 = EMAX*(C2**GAMA)/((IC50**GAMA)+(C2**GAMA)) C6 = A(6)/V2 E2 = EMAX*(C6**GAMA)/((IC50**GAMA)+(C6**GAMA)) DADT(1)=-KA*A(1) DADT(2) = KA*A(1)-CL*A(2)/V2-Q*A(2)/V2+Q*A(3)/V3 DADT(3) = Q*A(2)/V2-Q*A(3)/V3 DADT(5) = -KA*A(5) DADT(6) = KA*A(5)-CL*A(6)/V2-Q*A(6)/V2+Q*A(7)/V3 DADT(7) = Q*A(6)/V2-Q*A(7)/V3 DADT(4) = KIN*(1-E1) - KIN*(1-E2) $ERROR CP = A(2)/V2 CONC = CP*EXP(EPS(1)) EFF = A(4)*EXP(EPS(2)) Y = CONC*(1-TYPE) + EFF*TYPE ____________________________________________________________________________________ Boardwalk for $500? In 2007? Ha! Play Monopoly Here and Now (it's updated for today's economy) at Yahoo! Games. http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow

Wei-jian The error message means exactly what it says, you have, in the $MODEL record more than one compartment labeled as DEFOBS (compartments 2 and 6). Probably you don't want/need any compartment as the default dose or observation, but rather should be explicitly saying which compartment any dose/observation is associated with (likely you'll need initialization "doses" for the cells, along with drug doses, although in NM6 there are other ways to initialize the cell compartments, but if you want to do pk/cells simultaneously you'll need to make observations in two compartments). So, you'll need to add a CMT data item for doses, although in your code you do explicitly state which compartment is to be used to calculate the DV. Mark Sale MD Next Level Solutions, LLC www.NextLevelSolns.com

Hi Wei-jian, I let others help you with the NONMEM coding. However, I would like to ask what type of data set you have. It seems that you have such a rich data set for both PK and PD observations that allows you to estimate a total of 14 different parameters, including 11 associated with variability of every single structural model parameter. Is it really reasonable to, e.g., expect your plasma concentration data to provide reliable estimates of both structural model parameters of a 2-comp model with an absorption component as well as the variability terms associated with every single parameter of the model? I would suggest that you start with the simplest model you believe is reasonable in reflecting the underlying mechanism of the plasma concentration and effect data and include a single ETA to start with. Once you have the structural model in place, you could explore the possibility of estimating the variability terms with the model parameters. This approach should allow you to get the PK/PD model in place early on, with much shortened computing time. Toufigh

Mark: Thanks for your suggestions! I have removed "DEFOBS" from CMT6, and it is now running. Toufigh: Yes, I do have dense PK and PD data for this modeling exercise. I presented the simultaneous PK/PD modeling results at last month's AAPS NBC as a poster. I am now attempting to use the cell life span model to see if there any improvements in model fitting. Note that this is an mAb which reduces the NK cell count in the peripheral blood. Thanks! Wei-jian ____________________________________________________________________________________ We won't tell. Get more on shows you hate to love (and love to hate): Yahoo! TV's Guilty Pleasures list. http://tv.yahoo.com/collections/265

Hi, Then let me ask the readers about their experience with estimating ETAs on every single structural model parameter. I have seen many examples of NM control streams, where in addition to a complex structural model also several variability terms are implemented. In my limited experience, it is rather unlikely to be able to get meaningful estimates of all these parameters based on measurements of a single compartment in the model. It would be interesting to hear other people's comments. Toufigh

Toufigh, When I first started this some very smart people told me that you should put an ETA on everything, because, biologically, there is variability on everything. Since, then, I've decided that this isn't a feasible approach. The same argument would be used to justify physiologically based pk models - even though they have two many parameters to be feasible. At the other end, if you talk to some very traditional statistician, they are entirely "data driven", don't put anything in that the data don't tell you to, i.e., ignore your knowledge of biology entirely. So, we all decide where we are on this empirical/theoretical spectrum. Personally, I like to think I'm somewhere in the middle - biologically based, but really need the data to tell me what should go into a model. The good news, I think, is that Bayesian statistics lets us do this. Could be done formally, but you'd have to specify, up front, how confident yo! u were that each feature is in the model. If you have lots of reason to believe that volume is a linear function of weight, then (IMHO) you are justified putting it in even if the current data in hand don't really tell you that. On the other hand, I'd be hesitant to put clearance as a function of astrological sign, even if the current data support that (unless the support is very, very strong) - because my prior is so low. People more knowledgeable about Bayesian methods could put this in a more formal structure. But, to your question, my answer is NO, you might start with a prior that there is an ETA on any given parameter, but if the current data don't support it (by which I mean either in terms of log-likelihood test [I know, you can't formally test whether and ETA is significant] or failure to converge) I don't hesitate to remove it. I don't see a reason to treat variance parameters any different from structural model parameters or co! variate parameters in this regard.Mark Mark Sale MD Next Level Solutions, LLC www.NextLevelSolns.com

Toufigh, Mark Sale said: > When I first started this some very smart people told me that you should > put an ETA on everything, because, biologically, there is variability on > everything. This is still correct. There is variability in all biological parameters. Many of them are biologically connected and thus apparently random differences are statistically correlated. So it makes sense to estimate the variance-covariance matrix for the parameter random effects (OMEGA with full BLOCK() in NONMEM terms). The problem with trying to estimate the variability is caused by 1) sub-optimal designs 2) approximate estimation methods. The first can be improved by using an optimal design program (see France Mentre's excellent review at PAGE 2007 http://www.page-meeting.org/?abstract=1179 ). The second can be improved by using a better estimation method (see see France Mentre's excellent review at PAGE 2005 http://www.page-meeting.org/default.asp?id=26&keuze=abstract-view&goto=abstracts&orderby=author&abstract_id=833 } So if you have been given data from a typical sub-optimal design and are using NONMEM (none of the better algorithm implementations are as flexible (except possibly S-Adapt MCPEM http://www.page-meeting.org/?abstract=1111 ) ) then you have to be pragmatic and let the data tell you which parameters have distinguishable between subject variability. NONMEM gives some indirect help with this by issuing a warning about parameters being at a constraint boundary. If you haven't turned off this semi-helpful feature and your THETAs are well clear of their boundaries then you should check which OMEGAs are close to zero and consider fixing them to zero. Nick Mark Sale - Next Level Solutions wrote: > > Toufigh, > When I first started this some very smart people told me that you should > put an ETA on everything, because, > biologically, there is variability on everything. Since, then, I've decided > that this isn't a feasible approach. The > same argument would be used to justify physiologically based pk models - even > though they have two many > parameters to be feasible. At the other end, if you talk to some very > traditional statistician, they are entirely > "data driven", don't put anything in that the data don't tell you to, i.e., > ignore your knowledge of biology entirely. > So, we all decide where we are on this empirical/theoretical spectrum. > Personally, I like to think I'm somewhere > in the middle - biologically based, but really need the data to tell me what > should go into a model. The good news, > I think, is that Bayesian statistics lets us do this. Could be done > formally, but you'd have to specify, up front, > how confident yo! > u were that each feature is in the model. If you have lots of reason to > believe that volume is a > linear function of weight, then (IMHO) you are justified putting it in even > if the current data in hand don't really > tell you that. On the other hand, I'd be hesitant to put clearance as a > function of astrological sign, even if the > current data support that (unless the support is very, very strong) - because > my prior is so low. People more > knowledgeable about Bayesian methods could put this in a more formal > structure. > But, to your question, my answer is NO, you might start with a prior that > there is an ETA on any given > parameter, but if the current data don't support it (by which I mean either > in terms of log-likelihood test [I know, > you can't formally test whether and ETA is significant] or failure to > converge) I don't hesitate to remove it. I don't > see a reason to treat variance parameters any different from structural model > parameters or co! > variate parameters > in this regard. > Mark > > Mark Sale MD > Next Level Solutions, LLC > www.NextLevelSolns.com >