4 replies to this topic

[fang ke]

Dear Simon,

Recently, i have take look at the IVIVC case study located at http://pharsight.com /events/eventsonline_archive.php#replay_015.

I have a question with one data in this presentation.

In case study 1, 3 oral formulation were used to develop the IVIVC model. But what dose this VAL formulation mean?

Do you have this example project data for sharing?

Thanks in advance.

 

Coco

[Simon Davis]

Hi Coco,

  Sorry I don't have a copy of that project - the presentation was made by our consulting division so it may have been based on customer data that was anonomised for the presentation.

 

  However I did quickly download the PDF and if I'm looking at the same slide 11 "case Study 1" as you then the "Val" is a formulation for which both Dissolution and In Vivo data is available; it has been 'held back' to test the goodness of the model by a process known as external validation. Since it was not used to build the model you can test it by seeing how closely it's actual in vivo profiles compare to what the model would predict.

 

 Simon.

[Dr abdus samad]

Hi Simon

 

There are two types of validations are used one is internal validation which is used for the IVIVC and one is external validation.

 

In most of the cases the internal validation is of three batches (slow, medium and fast release profiles).

For the external validation we have to formulate one more batch with little bit similar release profile. So my query is that since we don’t know the exact profile of the batch which is used for the external validation then how it can be used for the validation.

As per my previous concept the validation is something like a standard for which you know all the profile and characteristics in advance.

So please let me know if any criteria for selection of the internal and external validation.

Hope for your positive reply.

[Simon Davis]

Dear Dr Samad,

 

 The external validation is an optional, but very useful, step in building your IVIVC. Imagine you have 4 formulations and both dissolution and in vivo data for all 4.

 

 Build your IVIVC model with only 3 of the sets and then use the 4th to 'externally validate' your model; i.e.as you already know the in vivo results for this set, the prediction from the model can be compared to see if it is a good predictor.

 

   If you do NOT have in vivo data from a formulation it is mapped to the PREDICTION DATA input and the results from this compared to your TARGET formulation.

 

 Simon.

[Dr abdus samad]

Dear Simon

Thank you very much for your esteemed and quick reply.

I have few more queries that

1. If the data of the fourth formulation is not available with me then can I take this data from the published literatures for the validation purpose or I should not have to take any other data and I should only rely on the internal validation data?

2. Secondly for the target formulation should I need both the invitro or invivo data or only invitro data is enough to get the invivo data prediction

3. Thirdly can I get the complete profile of invivo data from the IVIVC prediction? Like the data time vs. concentration.

4. Can you please provide me the complete case studies which are performed from IVIVC toolkit?

Hope for the positive reply.