mittyri
★★  

Russia,
2019-09-02 22:15

Posting: # 20531
Views: 328
 

 PK model parameters IR vs MR [PK / PD]

Dear All!

recently I got a question from my colleague and I didn't find good answer. Maybe you can help.
Suppose there's a data from MD study for IR formulation and the model built is good enough for the data description.
Now you get another dataset (SD) with the same analyte but MR formulation. Trying to build the model you figure out that the parameters differ significantly (both Cl and V).
What could be a reason of that observation?
We know that the kidneys do not care about lineage of entity.

Kind regards,
Mittyri
jag009
★★★

NJ,
2019-09-03 16:07

@ mittyri
Posting: # 20532
Views: 268
 

 PK model parameters IR vs MR

Hi,

1) Is the MR equivalent to IR with respect to AUC?
2) How different is the Cl and V (V in particular)?
3) how good is the model fit?

J
mittyri
★★  

Russia,
2019-09-03 16:54

@ jag009
Posting: # 20534
Views: 265
 

 PK model parameters IR vs MR

Hi John,

» 1) Is the MR equivalent to IR with respect to AUC?

No, MR AUC < IR AUC

» 2) How different is the Cl and V (V in particular)?

more than 10 times

» 3) how good is the model fit?

Internally validated with another piece of data. CV is less than 20% for all parameters
Tried to use IR params for MR dataset, during convergence they are going away. The problem is not in the model fit by itself

Kind regards,
Mittyri
ElMaestro
★★★

Belgium?,
2019-09-03 17:43

@ mittyri
Posting: # 20535
Views: 262
 

 PK model parameters IR vs MR

Hi mittyri,

» recently I got a question from my colleague and I didn't find good answer. Maybe you can help.
» Suppose there's a data from MD study for IR formulation and the model built is good enough for the data description.
» Now you get another dataset (SD) with the same analyte but MR formulation. Trying to build the model you figure out that the parameters differ significantly (both Cl and V).
» What could be a reason of that observation?

You have a difference in the absorbed fraction from the two formulations which screws everything up.


» We know that the kidneys do not care about lineage of entity.

What is meant with this term??

I could be wrong, but...
Best regards,
ElMaestro
mittyri
★★  

Russia,
2019-09-03 18:05

@ ElMaestro
Posting: # 20536
Views: 260
 

 PK model parameters IR vs MR

Hi ElMaestro,

» You have a difference in the absorbed fraction from the two formulations which screws everything up.

could you please explain what is happening here? Does it mean that changes in absorbed fraction did significantly modified PK of the drug? Are there any examples for another drugs where it was already observed?

» » We know that the kidneys do not care about lineage of entity.
» What is meant with this term??

Well, nothing special. I mean clearance does not depend on the route of administration for particular drug (or type of formulation).

But it seems to me that due to low absorption rate by time unit the PK properties of the analyte could be closer to some low dose IR (where non-linear kinetics rules). Is that correct?

Kind regards,
Mittyri
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2019-09-07 13:03

@ mittyri
Posting: # 20552
Views: 197
 

 Parameter-identifiability?

Hi mittyri,

» […] data from MD study for IR formulation […] another dataset (SD) […] but MR formulation. Trying to build the model you figure out that the parameters differ significantly (both Cl and V).
» What could be a reason of that observation?

Flip-flop PK? Rather the rule than an exception with MR formulations. You might run into parameter-identifiability issues. Unless you have data from an IV dose you don’t know what the true elimination is.
An idea: Parameterize models with rate-constants (instead of CL) and compare kel of the IR formulation with ka of the MR formulation. Are they similar?

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
mittyri
★★  

Russia,
2019-09-07 22:25

@ Helmut
Posting: # 20553
Views: 185
 

 transit compartments...

Hi Helmut,

» An idea: Parameterize models with rate-constants (instead of CL) and compare kel of the IR formulation with ka of the MR formulation. Are they similar?

as always - thanks for support and suggestion!
You know MR is a nightmare for ka modelling (that's why we go into one stage IVIVC sometimes).
For that particular case absorption was modelled with a transit compartments chain. So there's no simple Ka to compare...

Kind regards,
Mittyri
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2019-09-08 10:56

@ mittyri
Posting: # 20554
Views: 166
 

 No organs in (Pop)PK

Hi mittyri,

» You know MR is a nightmare for ka modelling […]. For that particular case absorption was modelled with a transit compartments chain. So there's no simple Ka to compare...

OK, clearer now. Keep in in mind that CL, V, rate constants belong to the – purely empiric – (Pop)PK concept.
I think that when you wrote
» » » » » We know that the kidneys do not care about lineage of entity.
you fell into this trap.

There are no organs associated with any of the PK parameters.* See also this post and read the heated debates of the two churches at David Bourne’s PKPD-list.
You have models describing IR and MR. Don’t worry about ‘differences’ in CL/V. Since CL/V in both models are purely empiric and have no connection to physiology at all, be happy and go ahead. If you are interested to explore what’s going on, IMHO, the only way out would be PBPK – which is not another league but another sport.

Remember that all models are wrong;
the practical question is
how wrong do they have to be to not be useful.
   George E.P. Box


» » » […] due to low absorption rate by time unit the PK properties […] could be closer to some low dose IR (where non-linear kinetics rules). Is that correct?

I don’t think so. Generally it is the other way ’round. Nonlinear PK at high concentrations (due to saturation) gets linear at low ones. Think about C2H5OH. ;-)


  • Oral diclofenac follows a simple one-compartment model with a lag-time. When I had pain in my right knee, I tried hard to convince my synovia, ligaments, and cartilage that my body is a well - stirred beaker. Unfortunately the message didn’t get across.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
Activity
 Thread view
Bioequivalence and Bioavailability Forum |  Admin contact
19,789 posts in 4,196 threads, 1,358 registered users;
online 10 (1 registered, 9 guests [including 9 identified bots]).
Forum time (Europe/Vienna): 15:07 CEST

100% of all disasters are failures of design, not analysis.    Ronald G. Marks

The BIOEQUIVALENCE / BIOAVAILABILITY FORUM is hosted by
BEBAC Ing. Helmut Schütz
HTML5