3 replies to this topic

[csheme]

Dear all,

 

I would like to ask your assistance with a project I am working on and will provide a bit of background of the data I have available and what I am attempting to do.

 

I have dose-response data and have modeling and simulated the ED50 values for pharmacological activity in liver following a simple efficacy screen (2 total doses, one on days 0 and one on 6).

 

My question is now on modeling and simulating the (exposure response) in liver. I do not have the liver concentration from this exact study, but I do have liver pk from a separate study which I want to use (please note the dosing regimen from this separate study is weekly for a total of 14 doses).

 

My pk data is limited to liver concentrations (observations for each dose groups i.e. 5, 20, and 60 mg/kg/wk) following the end of treatment.

 

My questions are first how can I build a simple PK model for the liver data alone (I am linking the dataset below). I have never built a PK model based on limited single point tissue data before. 

 

PK DATASET:

 https://www.dropbox....erConc.csv?dl=0

 

Once I have a liver PK model established for the (14 dose regimen), how can I incorporate the pharmacology from the dose response (2 dose regimen) to simulate the exposure response. (I am linking the pharmacology dataset below and the dose response model).

 

SEPARATE PD DATASET:

https://www.dropbox....ology.xlsx?dl=0

 

 

Thank you all tremendously, and wish you a Happy Holiday season!

Edited by csheme, 04 January 2017 - 03:07 PM.

[serge guzy]

Dear colleague

Not enough information to help you efficiently.

In general, using the graphical mode, you can define a plasma compartment  for PK, a link with a compartment you call liver, another link between plasma and one or 2 peripheral compartments to account for all other organs. Then you link the liver compartment to an observation object where your liver observations will be mapped to.

Now if all this is not clear and you were not exposed to Phoenix NLME that way, I strongly suggest you to go one of our courses.

Introductory to get some basic notions, intermediate to be exposed to the kind of problems you just defined and advanced for test cases with lot of coding and complex modeling.

Best Regards

Serge

[Simon Davis]

To add to Serge's comment the link to the training registrations is;

 

http://www.certaraun...e/details&id=27

 

however I think without  prior information for PK parameters you simply don't have enough information to fit or even simulate with any confidence with essentially only one timepoint for each profle, all at the same time.

 Simon.

[csheme]

Thank you Simon, you are confirming my initial thoughts.