Dear all
I am a working on a study that measures both short and long term
exposure to drug using plasma and hair drug concentration. What
methods can I use to model hair drug concentration. You can give
references so that I can read.
Thanks once again.
Regards
--
Bernard Ngara
Biometrician
Department of Research Specialist Services
5th Street Extension
Harare
Zimbabwe
+263 772 160 896
Population PKPD using hair drug concentration
Dear Bernard,
This looks like really interesting problem. Based on the idea that it should be a long delay, I would start with the transit compartment model (you can google for the references on this type of models) with the input from the plasma compartment. The last compartment will represent a barber shop. The number of transit compartment can be increased until you get a sufficiently long delay. Observation compartment can be either the last one, or the sum of several, depending on how measurements are done (at a particular hair length, or by grinding the hair together before measurement). Depending on whether hair can eliminate the drug (or it happens only in the barber shop), hair clearance can be assigned to all or only to the last of those transit compartments.
It could be that a simple effect compartment model with a very slow ke0 could describe it as well but you should be able to see it by increasing or decreasing the number of transit compartments.
Regards,
Leonid
--------------------------------------
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web: www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel: (301) 767 5566
Quoted reply history
On 6/12/2015 8:56 AM, Bernard Ngara wrote:
> Dear all
>
> I am a working on a study that measures both short and long term
> exposure to drug using plasma and hair drug concentration. What
> methods can I use to model hair drug concentration. You can give
> references so that I can read.
>
> Thanks once again.
>
> Regards
Dear Leonid and Bernard,
Why not treat hair like a urine compartment that is not emptied until the hair
is pulled out with the root? NONMEM user guides contain all you need to model a
"collection compartment" like urine/hair.
Have fun,
Joachim
Joachim Grevel, PhD
Scientific Director
BAST Inc Limited
Loughborough Innovation Centre
Charnwood Building
Holywell Park, Ashby Road
Loughborough, LE11 3AQ
Tel: +44 (0)1509 222908
Confidentiality Notice: This message is private and may contain confidential
and proprietary information. If you have received this message in error, please
notify us and remove it from your system and note that you must not copy,
distribute or take any action in reliance on it. Any unauthorized use or
disclosure of the contents of this message is not permitted and may be unlawful.
Quoted reply history
-----Original Message-----
From: [email protected] [mailto:[email protected]] On
Behalf Of Leonid Gibiansky
Sent: 12 June 2015 14:55
To: Bernard Ngara; [email protected]
Subject: Re: [NMusers] Population PKPD using hair drug concentration
Dear Bernard,
This looks like really interesting problem. Based on the idea that it should be
a long delay, I would start with the transit compartment model (you can google
for the references on this type of models) with the input from the plasma
compartment. The last compartment will represent a barber shop. The number of
transit compartment can be increased until you get a sufficiently long delay.
Observation compartment can be either the last one, or the sum of several,
depending on how measurements are done (at a particular hair length, or by
grinding the hair together before measurement). Depending on whether hair can
eliminate the drug (or it happens only in the barber shop), hair clearance can
be assigned to all or only to the last of those transit compartments.
It could be that a simple effect compartment model with a very slow ke0 could
describe it as well but you should be able to see it by increasing or
decreasing the number of transit compartments.
Regards,
Leonid
--------------------------------------
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web: www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel: (301) 767 5566
On 6/12/2015 8:56 AM, Bernard Ngara wrote:
> Dear all
>
> I am a working on a study that measures both short and long term
> exposure to drug using plasma and hair drug concentration. What
> methods can I use to model hair drug concentration. You can give
> references so that I can read.
>
> Thanks once again.
>
> Regards
>
Dear Bernard,
I don't know if you only have data collected from the hair on the scalp or
also from other parts of the body. If so, I think that you should try to
distinguish between body hair and scalp hair (maybe using covariates) since
they have different growth rates and so your drug could be eliminated
faster from the hair on the scalp (barber shop) compared to the hair on
your arm, for example.
Good luck and I am looking forward to the outcome.
Best regards,
Sinziana Cristea
*Research Associate *Simcyp Limited
Quoted reply history
On 12 June 2015 at 14:54, Leonid Gibiansky <[email protected]>
wrote:
> Dear Bernard,
>
> This looks like really interesting problem. Based on the idea that it
> should be a long delay, I would start with the transit compartment model
> (you can google for the references on this type of models) with the input
> from the plasma compartment. The last compartment will represent a barber
> shop. The number of transit compartment can be increased until you get a
> sufficiently long delay. Observation compartment can be either the last
> one, or the sum of several, depending on how measurements are done (at a
> particular hair length, or by grinding the hair together before
> measurement). Depending on whether hair can eliminate the drug (or it
> happens only in the barber shop), hair clearance can be assigned to all or
> only to the last of those transit compartments.
>
> It could be that a simple effect compartment model with a very slow ke0
> could describe it as well but you should be able to see it by increasing or
> decreasing the number of transit compartments.
>
> Regards,
> Leonid
>
>
>
> --------------------------------------
> Leonid Gibiansky, Ph.D.
> President, QuantPharm LLC
> web: www.quantpharm.com
> e-mail: LGibiansky at quantpharm.com
> tel: (301) 767 5566
>
>
>
>
> On 6/12/2015 8:56 AM, Bernard Ngara wrote:
>
>> Dear all
>>
>> I am a working on a study that measures both short and long term
>> exposure to drug using plasma and hair drug concentration. What
>> methods can I use to model hair drug concentration. You can give
>> references so that I can read.
>>
>> Thanks once again.
>>
>> Regards
>>
>>
Dear Bernard,
If you observed a long delay between your measurement and exposure to plasma
drug concentration and you do not want to model the mechanisms underlying
delay, the simples way to start is to use the delayed plasma concentration
C(t-Tdel). For example, the Emax model with delay will be
E(t)=Emax*C(t-Tdel)/(EC50+C(t-Tdel)), etc. You can implement Tdel in NONMEM
using the ALAG feature.
What Leonid suggests might work, but then you need to interpret the transit
compartments and you will need a lot of them. The link model (effect
compartment) makes sense, but might not work if the delay is too long.
My gut feeling is that to model drug concentration in hair you will need more
complex models handling delays than any of these.
Regards,
Wojciech
Quoted reply history
-----Original Message-----
From: [email protected] [mailto:[email protected]] On
Behalf Of Leonid Gibiansky
Sent: Friday, June 12, 2015 9:55 AM
To: Bernard Ngara; [email protected]
Subject: Re: [NMusers] Population PKPD using hair drug concentration
Dear Bernard,
This looks like really interesting problem. Based on the idea that it should be
a long delay, I would start with the transit compartment model (you can google
for the references on this type of models) with the input from the plasma
compartment. The last compartment will represent a barber shop. The number of
transit compartment can be increased until you get a sufficiently long delay.
Observation compartment can be either the last one, or the sum of several,
depending on how measurements are done (at a particular hair length, or by
grinding the hair together before measurement). Depending on whether hair can
eliminate the drug (or it happens only in the barber shop), hair clearance can
be assigned to all or only to the last of those transit compartments.
It could be that a simple effect compartment model with a very slow ke0 could
describe it as well but you should be able to see it by increasing or
decreasing the number of transit compartments.
Regards,
Leonid
--------------------------------------
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web: www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel: (301) 767 5566
On 6/12/2015 8:56 AM, Bernard Ngara wrote:
> Dear all
>
> I am a working on a study that measures both short and long term
> exposure to drug using plasma and hair drug concentration. What
> methods can I use to model hair drug concentration. You can give
> references so that I can read.
>
> Thanks once again.
>
> Regards
>
Bernard,
An alternative approach would be consider hair "concentrations" accumulating in an output compartment. Each time the hair is cut you "empty" the compartment and reset it to accumulate until the next hair cut. This is similar to using an output compartment to predict amounts in urine.
NONMEM has a convenient way to obtain a prediction of the amount output from a compartment (such as plasma or the end of transit chain). You just need to estimate an output fraction to explain the actual amount you measure.
Best wishes,
Nick
+--------------------------------------------------------------------+
| |
| OUTPUT FRACTION PARAMETER |
| |
+--------------------------------------------------------------------+
MEANING: Output fraction (F0) parameter for PREDPP
CONTEXT: Additional PK Parameters
USAGE:
$PK
F0= ....
DISCUSSION:
The output fraction parameter is used with PREDPP. It is an optional
additional PK parameter. With NM-TRAN, it are symbolized in the $PK
block by any one of the reserved variables FO, F0, or Fn, where n is
the compartment number of the output compartment.
With any of the kinetic models a (peripheral) output compartment is
always present. Associated with this compartment is a PK parameter,
the output fraction , denoted here by Fo. Of the entire amount, Ao, of
drug introduced into the system by various dosage patterns and then
eliminated from the system during a given time interval, a fraction Fo
of
Ao goes into this output compartment.
If the output compartment is never turned on, the output fraction can
be ignored. If the value of the output fraction is not computed in PK,
it is always understood to be 1
The use of Fo depends on the assumption that the rate of change of
drug amount in the output compartment is linear in the other compart-
ment amounts. Other than this linearity restriction, the system can be
nonlinear.
Quoted reply history
On 12/06/2015 3:54 p.m., Leonid Gibiansky wrote:
> Dear Bernard,
>
> This looks like really interesting problem. Based on the idea that it should be a long delay, I would start with the transit compartment model (you can google for the references on this type of models) with the input from the plasma compartment. The last compartment will represent a barber shop. The number of transit compartment can be increased until you get a sufficiently long delay. Observation compartment can be either the last one, or the sum of several, depending on how measurements are done (at a particular hair length, or by grinding the hair together before measurement). Depending on whether hair can eliminate the drug (or it happens only in the barber shop), hair clearance can be assigned to all or only to the last of those transit compartments.
>
> It could be that a simple effect compartment model with a very slow ke0 could describe it as well but you should be able to see it by increasing or decreasing the number of transit compartments.
>
> Regards,
> Leonid
>
> --------------------------------------
> Leonid Gibiansky, Ph.D.
> President, QuantPharm LLC
> web: www.quantpharm.com
> e-mail: LGibiansky at quantpharm.com
> tel: (301) 767 5566
>
> On 6/12/2015 8:56 AM, Bernard Ngara wrote:
>
> > Dear all
> >
> > I am a working on a study that measures both short and long term
> > exposure to drug using plasma and hair drug concentration. What
> > methods can I use to model hair drug concentration. You can give
> > references so that I can read.
> >
> > Thanks once again.
> >
> > Regards
--
Nick Holford, Professor Clinical Pharmacology
Dept Pharmacology & Clinical Pharmacology, Bldg 503 Room 302A
University of Auckland,85 Park Rd,Private Bag 92019,Auckland,New Zealand
office:+64(9)923-6730 mobile:NZ+64(21)46 23 53 FR+33(7)80 48 55 50
email: [email protected]
http://holford.fmhs.auckland.ac.nz/
Holford SD, Allegaert K, Anderson BJ, Kukanich B, Sousa AB, Steinman A, Pypendop,
B., Mehvar, R., Giorgi, M., Holford,N.H.G. Parent-metabolite pharmacokinetic models
- tests of assumptions and predictions. Journal of Pharmacology & Clinical
Toxicology. 2014;2(2):1023-34.
Holford N. Clinical pharmacology = disease progression + drug action. Br J Clin
Pharmacol. 2015;79(1):18-27.