Bioequivalence and Bioavailability Forum

ramesh_7779 ★ 2011-10-17 08:18 (5357 d 01:07 ago) Posting: # 7501 Views: 9,731	Level C Correlation (IVIVC) [Dissolution / BCS / IVIVC] Post reply
	Dear All, Can anyone pls explain how to do Level C Correlation (IVIVC) by winnonlin with example. Regards Ramesh

Helmut
★★★

Vienna, Austria,
2011-10-17 16:01
(5356 d 17:25 ago)

@ ramesh_7779
Posting: # 7505
Views: 8,410

Level C Correlation (WinNonlin)

Post reply

Dear Ramesh!

❝ Can anyone pls explain how to do Level C Correlation (IVIVC) by winnonlin with example.

Level C is a point-to-point relationship between an in vitro parameter (e.g., t_50%) and an in vivo metric (e.g., t_max, C_max, AUC).

In WinNonlin set up a workbook with two columns. In the first the dissolution (in vitro) parameter and in the second the PK metric (in vivo). You will need at least three formulations to set up a linear Level C and four formulations for a quadratic Level C. Values of the formulations go into the rows of the workbook. Send the data to a linear model; either Model 502 (Linear) or Model 503 (Quadratic). It’s possible to get a negative intercept – that’s an indication of a lag-time.
If you want to predict in vivo response from in vitro data (X) calculate it with the Final Parameters of the results workbook as:

Model 502: INT + SLOPE * X

Model 503: A0 + A1 + A2 * X

For an overview see

Cardot J-M, Beyssac E, and M Alric
In Vitro–In Vivo Correlation: Importance of Dissolution in IVIVC
Dissolution Technologies 14(1), 15–9 (2007)
Online

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

ramesh_7779 ★ 2011-10-18 08:13 (5356 d 01:13 ago) @ Helmut Posting: # 7509 Views: 8,104	Level C Correlation (WinNonlin) Post reply
	Dear Helmut, Thanks for the reply. I am completely ok with Level A correlation, but i didnt get exposure to work with Level C dataset. Since Level A and C are mostly used. If you can share one dataset for Level C correlation it will be useful. Regards Ramesh

Helmut
★★★

Vienna, Austria,
2011-10-18 17:22
(5355 d 16:04 ago)

@ ramesh_7779
Posting: # 7515
Views: 8,194

Level C Correlation (WinNonlin)

Post reply

Dear Ramesh!

❝ I am completely ok with Level A correlation, but i didnt get exposure to work with Level C dataset.

Can you give your example?
‘Didn’t work’ is not helpful (my crystal ball is in the laundry). ;-)

❝ Since Level A and C are mostly used.

Well; Level C is the worst by far. If you don’t succeed with A, try B first.

❝ If you can share one dataset for Level C correlation it will be useful.

A picture tells more than a thousand words.
[image]

Select Model 502 | Linear
In the setup of Data Variables, drag [t50] to [X Variable] and [tmax] to [Y Variable] OK
[Start Modeling]
In the [Linear Workbook] you should get in the [Non-Transposed Final Parameters]-sheet:
Parameter | Estimate | StdError | ----------+-----------+----------+ INT | -0.055556 | 0.299176 | SLOPE | 2.000000 | 0.384900 |

We can predict t_max based on tmax = INT + SLOPE × t50 within the data range we have used in the model. It’s not acceptable to extrapolate!
If we have a new formulation with a t_50% of 0.85 h we get the in vivo t_max as -0.055556 + 2 × 0.85 ≈ 1.64 h.

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

JMCardot
☆

France,
2011-10-24 14:34
(5349 d 18:51 ago)

(edited on 2011-10-24 17:08)
@ Helmut
Posting: # 7536
Views: 8,123

Level C Correlation (WinNonlin)

Post reply

Hello all,

In contradiction to HS I dislike level B ;-)

. In your case the pertinence of level C will depend of the nature of your problem.

To give examples

if you have a class I or II drug and a SR formulation a level C could be sufficient to analyse and solve a Cmax problem.
You could also have multiple level C in which you are going to establish relationships between Cmax and dissolution but also AUC and dissolution etc... I this case the power of IVIVC level C is quite high as you are able to predict all BE parameters.

In all cases you must understand the formulation and process and establish the critical attribute of your formulation and process in order to know how to influence the behavior of your formulation.

Best regards,

JM

Edit: Full quote removed. Please delete anything from the text of the original poster which is not necessary in understanding your answer; see also this post! [Ohlbe]

Helmut
★★★

Vienna, Austria,
2011-10-27 20:27
(5346 d 12:58 ago)

@ JMCardot
Posting: # 7556
Views: 7,976

Level C rulez

Post reply

Dear Jean-Michel!

❝ In contradiction to HS I dislike level B ;-) . In your case the pertinence of level C will depend of the nature of your problem.

❝

❝ To give examples […]

OK, I give in – you are the expert here. ;-)

I still consider MRT / MAT in level B to be useful, but we might be dead if the slightest lag-time comes in to play…

❝ In all cases you must understand the formulation and process and establish the critical attribute of your formulation and process in order to know how to influence the behavior of your formulation.

Yes. Yes!

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Level C Correlation (IVIVC) [Dissolution / BCS / IVIVC]

Level C Correlation (WinNonlin)

Level C Correlation (WinNonlin)

Level C Correlation (WinNonlin)

Level C Correlation (WinNonlin)

Level C rulez