Dear nmusers,
I have PK data from a compound with a very long half-life (weeks). I have two
issues with this data and would very much like to hear your opinion on this:
· Concentrations were rounded to integers, with LLOQ of 1. This means a
concentration value of 2 could span values from 1.5-2.49, and unfortunately
there is no way to find out what was originally measured, and I have quite a
few measurements around LLOQ. How would you handle this in the coding (also
later when mixing with additional data that was not rounded), and how will this
rounding influence my parameter estimates?
· The sampling times were only documented as dates (except for the day
of dosing). With a long half-life compound, do you consider this to be critical?
Your opinions would be very welcome.
Best regards
Nele
______________________________________________________________
Dr. Nele Käßner
Principal Scientist Modeling and Simulation
Global Pharmacometrics
Experimental Medicine
Takeda Pharmaceuticals International GmbH
Thurgauerstrasse 130
8152 Glattpark-Opfikon (Zürich)
Switzerland
Visitor address:
Alpenstrasse 3
8152 Glattpark-Opfikon (Zürich)
Switzerland
Phone: (+41) 44 / 55 51 404
Mobile: (+41) 79 / 654 33 99
mailto: [email protected]<mailto:[email protected]>
http://www.takeda.com
--------------------------------------------------------------------
The content of this email and of any files transmitted may contain
confidential, proprietary or legally privileged information and is intended
solely for the use of the person/s or entity/ies to whom it is addressed. If
you have received this email in error you have no permission whatsoever to use,
copy, disclose or forward all or any of its contents. Please immediately notify
the sender and thereafter delete this email and any attachments.
rounding issues with PK data
Hi Nele,
For the first point, if you have values that are rounded like that and you will
be mixing rounded and non-rounded values, then the simplest way to handle it
would be to have a different additive error term for measurements that were or
were not subject to the rounding.
A more complex method would be to implement something akin to the M2 method but
using both a lower and upper limit for the range of values that would be mapped
to each value in the file. The higher complexity is unlikely to be worth it.
With the sampling times only being documented as dates, to give the full
answer, we'd need more information: is there a non-terminal phase to the
half-life that extends beyond the first day? Is the terminal half-life > ~5
days? What is there a need for more precision than 1 day resolution to the
half-life and is it possible (i.e. even if you had data down to the second,
would your SE be <1 day for half-life)?
Generally, if the half-life is more than about 5 days, there should be no major
issue with only having dates of collection.
Thanks,
Bill
Quoted reply history
On Jun 10, 2013, at 2:57, "Kaessner, Nele"
<[email protected]<mailto:[email protected]>> wrote:
Dear nmusers,
I have PK data from a compound with a very long half-life (weeks). I have two
issues with this data and would very much like to hear your opinion on this:
· Concentrations were rounded to integers, with LLOQ of 1. This means a
concentration value of 2 could span values from 1.5-2.49, and unfortunately
there is no way to find out what was originally measured, and I have quite a
few measurements around LLOQ. How would you handle this in the coding (also
later when mixing with additional data that was not rounded), and how will this
rounding influence my parameter estimates?
· The sampling times were only documented as dates (except for the day
of dosing). With a long half-life compound, do you consider this to be critical?
Your opinions would be very welcome.
Best regards
Nele
______________________________________________________________
Dr. Nele Käßner
Principal Scientist Modeling and Simulation
Global Pharmacometrics
Experimental Medicine
Takeda Pharmaceuticals International GmbH
Thurgauerstrasse 130
8152 Glattpark-Opfikon (Zürich)
Switzerland
Visitor address:
Alpenstrasse 3
8152 Glattpark-Opfikon (Zürich)
Switzerland
Phone: (+41) 44 / 55 51 404
Mobile: (+41) 79 / 654 33 99
mailto: [email protected]<mailto:[email protected]>
http://www.takeda.com
--------------------------------------------------------------------
The content of this email and of any files transmitted may contain
confidential, proprietary or legally privileged information and is intended
solely for the use of the person/s or entity/ies to whom it is addressed. If
you have received this email in error you have no permission whatsoever to use,
copy, disclose or forward all or any of its contents. Please immediately notify
the sender and thereafter delete this email and any attachments.
Hi Nele,
That is annoying! These things happen but still....
How I would handle it? I probably would make sure to estimate an additive error
next to a proportional/exponential one for PK. Chances are that you would need
an additive error to start with, but let's assume that in your case the
rounding error would dominate. The additive component may be poorly supported
by the data, in which case you could fix it to 0.25 on the SD scale (~95%
coverage spans 1.5-2.49). It of course deviates from the expected distribution
(uniform) but I would suppose the match is good enough; if you really have a
lot of samples around the LLOQ you may want to do some simulation-estimation
exercises to ensure there is no bias.
Mixing it with other data can be simple: just add a flag for the rounded data,
and differentiate inclusion/estimation of the additive component with the flag.
The dates for sampling time are good enough after the first week from dosing or
so. Generally you would like to make the assumption that the error in IDV is
negligible compared to the error in DV. The only option that you have if you
can't make the assumption I believe is a method described by Jan Freijer to
increase robustness: http://www.page-meeting.org/page/page2005/PAGE2005P76.pdf
Hope this helps,
Jeroen
PS: Bill already answered before I could finish this mail, I am sending anyway
because of some additional considerations
J. Elassaiss-Schaap Senior Principal
Scientist Phone: + 31 412 66 9320
MSD | PK, PD and Drug Metabolism | Clinical PK-PD Mail stop KR
4406 | PO Box 20, 5340 BH Oss, NL
Quoted reply history
From: [email protected]<mailto:[email protected]>
[mailto:[email protected]] On Behalf Of Kaessner, Nele
Sent: Monday, June 10, 2013 8:05
To: [email protected]<mailto:[email protected]>
Subject: [NMusers] rounding issues with PK data
Dear nmusers,
I have PK data from a compound with a very long half-life (weeks). I have two
issues with this data and would very much like to hear your opinion on this:
· Concentrations were rounded to integers, with LLOQ of 1. This means a
concentration value of 2 could span values from 1.5-2.49, and unfortunately
there is no way to find out what was originally measured, and I have quite a
few measurements around LLOQ. How would you handle this in the coding (also
later when mixing with additional data that was not rounded), and how will this
rounding influence my parameter estimates?
· The sampling times were only documented as dates (except for the day
of dosing). With a long half-life compound, do you consider this to be critical?
Your opinions would be very welcome.
Best regards
Nele
______________________________________________________________
Dr. Nele Käßner
Principal Scientist Modeling and Simulation
Global Pharmacometrics
Experimental Medicine
Takeda Pharmaceuticals International GmbH
Thurgauerstrasse 130
8152 Glattpark-Opfikon (Zürich)
Switzerland
Visitor address:
Alpenstrasse 3
8152 Glattpark-Opfikon (Zürich)
Switzerland
Phone: (+41) 44 / 55 51 404
Mobile: (+41) 79 / 654 33 99
mailto: [email protected]<mailto:[email protected]>
http://www.takeda.com
--------------------------------------------------------------------
The content of this email and of any files transmitted may contain
confidential, proprietary or legally privileged information and is intended
solely for the use of the person/s or entity/ies to whom it is addressed. If
you have received this email in error you have no permission whatsoever to use,
copy, disclose or forward all or any of its contents. Please immediately notify
the sender and thereafter delete this email and any attachments.
--------------------------------------------------------------------
Notice: This e-mail message, together with any attachments, contains
information of Merck & Co., Inc. (One Merck Drive, Whitehouse Station,
New Jersey, USA 08889), and/or its affiliates Direct contact information
for affiliates is available at
http://www.merck.com/contact/contacts.html) that may be confidential,
proprietary copyrighted and/or legally privileged. It is intended solely
for the use of the individual or entity named on this message. If you are
not the intended recipient, and have received this message in error,
please notify us immediately by reply e-mail and then delete it from
your system.