Back to the Top
My question is about calculation of T1/2 using winonlin with the following set of data:
Compound A (5MG/KG po)
Time (Hr)
0.00
0.50
1.00
2.00
4.00
7.00
24.00
Plasma Concentration (ng/ml)
0.00
114.33
132.67
150.67
88.20
8.04
0.00
Compound B (50MG/KG PO)
Time (Hr)
0.00
0.50
1.00
2.00
4.00
7.00
24.00
Plasma concentration (mg/ml)
0.00
6253.33
6843.33
6603.33
5270.00
2610.00
10.00
Winnonlin calculated the T1/2 of these compounds as 1.15 and 2.15 hours. My understanding is that winonlin considers the 3 terminal points (not the one which has zero levels) for the calculation of T1/2. The data clearly indicates that the levels of both the compounds remain staedy from 0.5hrs to 4hrs. Going by the classical definition of T1/2 (time it takes for the compound to reach half of its original concentration), I am finding it difficult to understand the values that winnolin has generated. Can someone explain the reason between the discrepancy between what is evident from the data and the winonlin generated T1/2 values.
Thanks
Back to the Top
The following message was posted to: PharmPK
Dear Geeta,
Please review the help file provided with the software, I've extracted the salient points for your questions.
First how WinNonlin Classic (that is versions 5.x and earlier), will determine the data points to include in lambda_z or slope calculations (to get Thalf) if the user chooses NOT to manually select Lz (and does NOT choose "disable curve stripping" under options)
- During the analysis, NCA repeats regressions using the last three points with non-zero concentrations, then the last four points, last five, etc.
- (Points prior to Cmax or prior to the end of infusion are not used unless the user specifically requests that time range.)
- Points with a value of zero for the dependent variable i.e. CONC are excluded.
- For each regression, an adjusted R2 is computed, i.e a 'best fit and then WinNonlin estimates lambda_z using the regression with the largest adjusted R2 and ...
- If the adjusted R2 does not improve, but is within 0.0001 of the largest adjusted R2 value, the regression with the larger number of points is used.
- lambda_z must be positive, and calculated from at least three data points.
> So you get half-lives of 1.15 h and 2.18h for analytes A and B respectively.
In Phoenix WinNonlin, (that is versions 6.0 and later, current version is 6.1), one extra rule is added that Cmax is NOT automatically included so NO half-life is not calculated for your analyte A, (since only two points are seen in the apparent elimination curve.
--
Secondly let's look at "Going by the classical definition of T1/2 (time it takes for the compound to reach half of its original concentration)",
however this definition is talking about the apparent ELIMINATION phase, that is the one you are estimating from e.g. 2 to 7 hours in the examples below. Just looking at the raw data for Analyte B you can see the conc. At 4 hours is 5270, id you half that it would give you 2635, which is conveniently very close to what you observed at 7 hours (2610), so even without regression you might guess a half-life of less than 7-4 i.e. less than 3. In fact the regression below gives a half-life of 2.7 h for Analyte B, (Analyte A was 1.15h)
aaaa 0 0
aaaa 0.5 114.33
aaaa 1 132.67
aaaa 2 * 150.67
aaaa 4 * 88.2
aaaa 7 * 8.04
aaaa 24 0
b 0 0
b 0.5 6253.33
b 1 6843.33
b 2 * 6603.33
b 4 * 5270
b 7 * 2610
b 24 * 10
The statement "the data clearly indicates that the levels of both the compounds remain steady from 0.5hrs to 4hrs." is irrelevant to the calculation since you are not assessing the *apparent elimination* over this time interval.
On another note I am pleased to announce that I am pleased to announce the launch of Pharsight's new product-focused discussion forums for customers. The forums are part of a larger online initiative, the Pharsight Extranet - www.pharsight.com/extranet.
I would stress these are in no way meant to replace the PharmPK discussion list that David diligently maintains, but to give an arena for Pharsight *software* based discussions. In the past I know that many of you have discussed WinNonlin, WinNonMix, Trial Simulator, IVIVC, etc. on this discussion list and thank many of you for your tips and comments on Pharsight's implementations of various routines. I hope you will continue to read and comment on both boards depending on your interest/question.
A one-time registration will give you access to Pharsight's new discussion forums for Phoenix WinNonlin, NLME(tm), & Connect(tm) as well as PKS(tm) and WinNonlin(r) Classic.
Discussion categories for each product have been established and will become home to a growing collection of helpful product information, tricks & tips, and how-to's.
The Extranet also contains a growing library of product and scientific documents available for download. It is our hope that the Pharsight Extranet becomes a valuable online resource where members of the PK/PD analysis and modeling communities can share ideas and information.
Also available are Online support issue tracking and a Online support knowledgebase, both of these have been available for a while but hopefully should be more accessible and integrated as Pharsight adds more functionality in the coming months. Planned enhancements include, online training courses etc.
We are excited about this new resource and look forward to how it will improve interaction within the scientific community as well as the support and interaction between Pharsight and our customers. However I would stress that an actual bug should still be reported to Support as usual to ensure a timely review and response
Reminder - If you are a WinNonlin or WinNonMix user and have not upgraded to Phoenix WinNonlin or Phoenix NLME - it's free! (WinNonlin Enterprise customers also get the Connect module)
Also if you are using 6.0 still please do upgrade to 6.1 as soon as possible as lot of minor, but nice to have, features were added in this release. 6.1 will also be the only version supported by the forthcoming Phoenix Validation Suite.
Best regards,
Simon.
--
Simon.Davis.-a-.certara.com
Senior Scientific Consultant
Pharsight- A Certara(tm) Company
Back to the Top
The following message was posted to: PharmPK
Dear Geeta:
The route of administration for both the compounds is PO. Hence while calculating T1/2, the terminal phase is considered from Cmax (Tmax) till last concentration (if last concentration is zero, don't consider it). Hence your assumption "The data clearly indicates that the levels of both the compounds remain staedy from 0.5hrs to 4hrs" even though virtually true, it is not considered for calculation of T1/2.
Assuming One- compartment for both compounds:
The Tmax for Compound-A is 2 Hrs. Hence for Compound-A, the T1/2 calculation is based on time points 2, 4, 7 Hrs. The elimination rate constant for this data is 0.603 (Hr^-1) and T1/2 is 1.15 Hrs (R-squared value 0.9452).
Similarly the Tmax for Compound-B is 1 Hr. Hence for Compound-B, the T1/2 calculation is based on time points 1, 2, 4, 7, 24 Hrs. The elimination rate constant is 0.295(Hr^-1) and T1/2 is 2.35 Hrs (R-squared value 0.9868). If only three time points are considered (4-24 Hrs), the R-squared value improves to 0.9986 and T1/2 becomes 2.16 Hrs.
Hope this helps.
Thanks & Regards
Sivacharan Kollipara, M. Pharm
Research Scientist
Metabolism & Pharmacokinetics Department
Ranbaxy Laboratories Limited
R&D-III, Plot-20, Sector-18
Gurgaon - 122015
INDIA
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Calculation of t 1/2" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)