7 replies to this topic

[lizao007]

Hi,

 

I have a naive question here.

 

We want to do an NCA analysis with multiple dosing at Day 0, 7, and 14 during a 21-day Cycle 1. I know a dosing record is not mandatory when calculating AUC. However, my PK parameters (AUClast) turned out to be different when I left dosing empty comparing when I input a multi-dosing record (calculation results in the attachment section). Meanwhile, R gave me the same results with the one without dosing record in Pheonix WNL.

 

I am wondering why this can happen and I still want to somehow find a way to include dosing since a clearance is required for the analyses.

 

Thank you guys!

 

 

 

 

 

Attached Thumbnails

• 
• 

[Simon Davis]

I don't really follow what you've done, are you comapring like with like? Can you send your project?

Do you want AUC last or AUC to a common appropriate timepoint? e.g. AUC0_24?

 

Did you use sort keys to identify each profile the same way?

 Simon.

[lizao007]

I don't really follow what you've done, are you comapring like with like? Can you send your project?

Do you want AUC last or AUC to a common appropriate timepoint? e.g. AUC0_24?

 

Did you use sort keys to identify each profile the same way?

 Simon.

 

Hi Simon,

 

Sorry for the confuse.

 

I am comparing 1. AUClast results after NCA with multi-dosing record and

                          2. AUClast results after NCA with no dosing record

They are weirdly different under the two scenarios. I am attaching my mock Phoenix project in the attachment for clarification.

 

This is my first time dealing with multi-dosing and sorry if i made stupid mistakes:)

Thank you!

Attached Files

•  trial.phxproj   160.52KB   432 downloads

[Helmut Schütz]

Hi Liza,

 

looking at your project I don’t understand what you want to achieve.

1. Assessing whether the drug follows linear PK? That would mean that the superposition principle is applicable – which could be done by comparing AUC_0–τ in steady state with AUC_0–∞ after the first dose. However, this approach is only possible with equal doses.
2. You had escalating doses which is not informative. Before you know whether the drug follows linear PK you can’t even play around with dose-normalization. OK, I did it anyway: dose-normalized AUC_0–7 0.405, AUC_7–14 0.481, AUC_14–21 0.422. Now what?
   Are you trying to kill two birds with one stone?
3. NCA is valid only either after a single dose or in (pseudo-) steady state; not in between. Even modeling is problematic.
4. A hint: Always use the same units for dosing and sampling.

[lizao007]

 

Hi Liza,

 

looking at your project I don’t understand what you want to achieve.

1. Assessing whether the drug follows linear PK? That would mean that the superposition principle is applicable – which could be done by comparing AUC_0–τ in steady state with AUC_0–∞ after the first dose. However, this approach is only possible with equal doses.
2. You had escalating doses which is not informative. Before you know whether the drug follows linear PK you can’t even play around with dose-normalization. OK, I did it anyway: dose-normalized AUC_0–7 0.405, AUC_7–14 0.481, AUC_14–21 0.422. Now what?
   Are you trying to kill two birds with one stone?
3. NCA is valid only either after a single dose or in (pseudo-) steady state; not in between. Even modeling is problematic.
4. A hint: Always use the same units for dosing and sampling.

 

 

Thank you so much for the help.

I simply want to calculate PK parameters for report purpose. Based on your bullet point 3, can I say clearance and Vss cannot be acquired via Phoenix NCA if multiple doses are applied? 

[Helmut Schütz]

Hi Liza,

 

I simply want to calculate PK parameters for report purpose. Based on your bullet point 3, can I say clearance and Vss cannot be acquired via Phoenix NCA if multiple doses are applied? 

 

That’s not software-related but basic PK. You are neither in steady state nor did you administer equal doses. Whatever you will do, CL and V calculated for the last dosing interval will be wrong since you are violating the underlying assumptions. Sorry.

[simon.hutchings]

I came across this topic while searching for an answer to what I think is a related problem.

 

We have PK data following two 'single' oral doses administered at time 0 h and 4 h.  There is no further dosing, and both doses were the same (i.e. 100 mg per dose). T1/2 is approx 3 h.

 

Am I correct that it is inappropriate to estimate Cl/F or Vd/F in this situation?

[Simon Davis]

Correct. The NCA formulas will not support two doses and analysing the first dose profile i.e. truncating data at 4 h is very unlikely to be a good estimate either.

 

You could try modelling the data with either the WNL classic or Phoenix models.

 

 SImon.