[NMusers] RE: Is it reasonable to add covarite to PD parameters kin or kout?

[Hu, Chuanpu [JRDUS]]

Hello -

The following should be helpful:

Approaches to handling pharmacodynamic baseline responses
Chantaratsamon Dansirikul * Hanna E. Silber * Mats O. Karlsson

J/PK/PD 2008

:)
Chuanpu

From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Peiming Ma
Sent: Monday, July 27, 2015 9:31 PM
To: Mark Sale; Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Dear Mark and Li,

I think there is nothing wrong with using baseline as a covariate. Baseline 
observations contain more information than just some pre-treatment observations 
and thus have a lot of explanatory power; this is the reason they appear 
significant more often than not. Comparisons after treatments should try to use 
all information before treatments (thus including baselines).

Statisticians habitually use baselines as covariates in their ANCOVA models for 
obviously good reasons.

Regards,
Peiming

From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Mark Sale
Sent: Tuesday, July 28, 2015 3:34 AM
To: Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Li,
I'm going to go out on a limb (since I haven't seen your data or understand the 
biology) and suggest that the baseline value is not a covariate, but is just 
another observation (presumably with drug concentration = 0).  The baseline 
value is sort of by definition a function of kin.  So, you might look for 
predictors of Kin (age? Disease stage?) rather than using the observed value to 
predict the parameters (which then predict the observed value).  Also note that 
covariates are, by definition, measured without error, and a baseline value is 
likely measure with error (the same error as any other observation?).

Mark



From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Zhao,Li
Sent: Monday, July 27, 2015 2:39 PM
To: nmusers@globomaxnm.com
Subject: [NMusers] Is it reasonable to add covarite to PD parameters kin or 
kout?


Dear NMusers,



Right now I am doing covarite analysis for an indirect response model.



 I tested a few potential covarites and found it's STATISTICALLY significant if 
I add observed baseline value to kin.



But I am not sure if it makes sense to add the baseline value as a covarite to 
kin.



Could you please help me if you have had similar experiences before?



Thank you very much!

[NMusers] RE: Is it reasonable to add covarite to PD parameters kin or kout?

[Mats Karlsson]

Hi,

There are both pros and cons of using baseline observations as a covariate. We 
presented four different options for how to handle baseline and investigated 
estimation properties in the paper below. What we didn't look at, because it is 
a very bad idea, is to use an baseline observation both as a covariate and 
dependent variable.

Approaches to handling pharmacodynamic baseline responses.
Dansirikul C, Silber HE, Karlsson MO.
J Pharmacokinet Pharmacodyn. 2008 Jun;35(3):269-83.

Best regards,
Mats


Mats Karlsson, PhD
Professor of Pharmacometrics

Dept of Pharmaceutical Biosciences
Faculty of Pharmacy
Uppsala University
Box 591
75124 Uppsala

Phone: +46 18 4714105
Fax + 46 18 4714003
www.farmbio.uu.se/research/researchgroups/pharmacometrics/

From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Peiming Ma
Sent: Tuesday, July 28, 2015 3:31 AM
To: Mark Sale; Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Dear Mark and Li,

I think there is nothing wrong with using baseline as a covariate. Baseline 
observations contain more information than just some pre-treatment observations 
and thus have a lot of explanatory power; this is the reason they appear 
significant more often than not. Comparisons after treatments should try to use 
all information before treatments (thus including baselines).

Statisticians habitually use baselines as covariates in their ANCOVA models for 
obviously good reasons.

Regards,
Peiming

From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Mark Sale
Sent: Tuesday, July 28, 2015 3:34 AM
To: Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Li,
I'm going to go out on a limb (since I haven't seen your data or understand the 
biology) and suggest that the baseline value is not a covariate, but is just 
another observation (presumably with drug concentration = 0).  The baseline 
value is sort of by definition a function of kin.  So, you might look for 
predictors of Kin (age? Disease stage?) rather than using the observed value to 
predict the parameters (which then predict the observed value).  Also note that 
covariates are, by definition, measured without error, and a baseline value is 
likely measure with error (the same error as any other observation?).

Mark



From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Zhao,Li
Sent: Monday, July 27, 2015 2:39 PM
To: nmusers@globomaxnm.com
Subject: [NMusers] Is it reasonable to add covarite to PD parameters kin or 
kout?


Dear NMusers,



Right now I am doing covarite analysis for an indirect response model.



 I tested a few potential covarites and found it's STATISTICALLY significant if 
I add observed baseline value to kin.



But I am not sure if it makes sense to add the baseline value as a covarite to 
kin.



Could you please help me if you have had similar experiences before?



Thank you very much!

[NMusers] Implementation of the OMEGA covariance matrix

[Ana Miranda Bastos]

Dear NMusers,

I wonder if someone could help me to implement the OMEGA covariance matrix 
below.

$OMEGA  
 0.09  ;   PPV_VMAX
 0.00 0.09  ; PPV_V1
 0.00 0.00 0.09  ; PPV_V2
 0.00 0.00 0.00 0.09  ; PPV_KD
 0.00 0.01 0.00 0.00 0.09  ;   PPV_BMAX
 0.00 0.01 0.00 0.00 0.01 0.09  ; PPV KM


I don't want to fix any value to zero, but I don't know how to better 
illustrate that I would like to consider a covariance between:
. V1 and Bmax 
. Bmax and Km
. V1 and Km


Any advice is highly appreciated. Thank you in advance.

Kind regards,

Ana



--
Ana Bastos, Pharm, MSc, PhD student
Pharmacologie cellulaire et moléculaire
(Cellular and Molecular Pharmacology Unit)
Louvain Drug Research Institute
Université catholique de Louvain (Catholic University of Louvain)
UCL 7370 avenue E. Mounier 73
1200 Bruxelles, Belgique
--
http://www.facm.ucl.ac.be

Re: AW: [NMusers] Time 'T' in $Error

[sbihorel]

Hi,

From a purely technical point of view, you can use TIME in your $ERROR 
record if you want to implement your drug effect as a direct effect on 
top of the circadian rhythm (this type of model may or may not 
appropriate, I leave you judge of that).


SinfS = ( 1 + Amplitude * sin ( (2*PI/Period) * (TIME-Phase) ) )
SBP   = (SBas + SinfS) * (1-EffS)

In a direct effect model, you only need to evaluate your PD endpoint at 
the value of TIME that were observed/included in your dataset, because 
you only need the value of your drug concentration at a given time to 
calculate the value of the effect at this time.


If you want to implement a circadian rhythm in an indirect response 
model, you will have to derive your sine/consine function of time and 
use it in $DES. There are multiple papers from Dr. Jusko's team on the 
topic, eg:


http://www.ncbi.nlm.nih.gov/pubmed/10672436

Sebastien

On 7/28/2015 2:07 AM, Katrin Volz wrote:


Dear Bill, dear all

Thanks so much for your answer.

More concrete I would like to model blood pressure including a 
circadian rhythm. Therefore I would like to use a sine or cosine 
function with ‘T’ instead of ‘Time’.


SinfS = ( 1 + Amplitude * sin ( (2*PI/Period) * (T-Phase) ) )  ; sine 
or cosine function


SBP   = (SBas + SinfS) * (1-EffS) 
 ; SBP = systolic 
blood pressure, SBas = Baseline systolic blood pressure, (1-EffS = effect)


I could also model this effect using a turnover model and avoid this 
problem, but I would like to evaluate this kind of PD model (if there 
is any meaningful way to code).


I also had the idea to code this sine function as a function of time 
in a compartment and rename this compartment in $Error, but I am not 
sure if there is any way to do this.


Thanks again and best regards.

Katrin

*Von:*Denney, William S. [mailto:william.s.den...@pfizer.com]
*Gesendet:* 27 July 2015 17:34
*An:* Katrin Volz ; 
nmusers@globomaxnm.com

*Betreff:* RE: [NMusers] Time 'T' in $Error

Hi Katrin,

$ERROR is executed once per data row.  The time when $ERROR is run is 
the TIME value (the discrete times of the measurement). For this 
specific example, you can just use TIME.  You will need to code your 
$DES block so that what you’re wanting to integrate as a function of 
time is in a compartment amount (e.g. A(1)).  With your example, it’s 
hard to see how T is used in the $DES, but perhaps you could share the 
real example to give a bit more concrete advice on how to integrate 
the answer into your code.


Thanks,

Bill

*From:*owner-nmus...@globomaxnm.com 
 
[mailto:owner-nmus...@globomaxnm.com] *On Behalf Of *Katrin Volz

*Sent:* Monday, July 27, 2015 10:46 AM
*To:* nmusers@globomaxnm.com 
*Subject:* [NMusers] Time 'T' in $Error

Dear NMuser,

I would like to model a PD effect using a function which includes 
continuous time ‘ T ’.


The PD Effect is coded in $Error.

In general like this:

*$ERROR*

MyFunction = xxx + T

PDEffect = Basline * MyFunction

But ‘T’ cannot be used in $Error.

I tried to code the function with ‘T’ in $DES and rename it to use it 
in $ERROR:


*$DES*

MyFunction = xxx + T

*$ERROR*

MyNewFunction = MyFunction

PDEffect = Baseline * MyNewFunction

but it didn’t worked (to rename was not permitted by NONMEM).

Does anyone of you has an idea how to solve this problem?

Thanks and best regards.

Katrin

Anke-Katrin Volz

PhD Student

Saarland University

Clinical Pharmacy

Campus C2 2 | Room 0.31

D- 66123 Saarbrücken

mail ankekatrin.v...@uni-saarland.de 



fon +49 [0]681 302 22 84

fax +49 [0]681 302 70 258

www.clinicalpharmacy.me 

Re: AW: [NMusers] Time 'T' in $Error

[Jeroen Elassaiss-Schaap]

Dear Katrin,


You still could code your model in $ERROR using TIME, as your diurnal 
rhythm is proportional to baseline.


If you would like to explore indirect models, you could try the harmonic 
approach as described in Chakraborty et al. 1999 - 
http://europepmc.org/abstract/med/10533696; the beauty is that you can 
model the baseline in $PRED and use these estimates right into $DES for 
your drug effect on kin.


For blood pressure rhythms you might also want to try the more flexible 
model laid out by Sallstrom et al. 2005,

http://europepmc.org/abstract/med/16328099.

Hope this helps,
Jeroen


--

http://pd-value.com
jer...@pd-value.com
@PD_value
+31 6 23118438
-- More value out of your data!


Op 28-07-15 om 08:07 schreef Katrin Volz:


Dear Bill, dear all

Thanks so much for your answer.

More concrete I would like to model blood pressure including a 
circadian rhythm. Therefore I would like to use a sine or cosine 
function with ‘T’ instead of ‘Time’.


SinfS = ( 1 + Amplitude * sin ( (2*PI/Period) * (T-Phase) ) )  ; sine 
or cosine function


SBP   = (SBas + SinfS) * (1-EffS) 
 ; SBP = systolic 
blood pressure, SBas = Baseline systolic blood pressure, (1-EffS = effect)


I could also model this effect using a turnover model and avoid this 
problem, but I would like to evaluate this kind of PD model (if there 
is any meaningful way to code).


I also had the idea to code this sine function as a function of time 
in a compartment and rename this compartment in $Error, but I am not 
sure if there is any way to do this.


Thanks again and best regards.

Katrin

*Von:*Denney, William S. [mailto:william.s.den...@pfizer.com]
*Gesendet:* 27 July 2015 17:34
*An:* Katrin Volz ; 
nmusers@globomaxnm.com

*Betreff:* RE: [NMusers] Time 'T' in $Error

Hi Katrin,

$ERROR is executed once per data row.  The time when $ERROR is run is 
the TIME value (the discrete times of the measurement). For this 
specific example, you can just use TIME.  You will need to code your 
$DES block so that what you’re wanting to integrate as a function of 
time is in a compartment amount (e.g. A(1)).  With your example, it’s 
hard to see how T is used in the $DES, but perhaps you could share the 
real example to give a bit more concrete advice on how to integrate 
the answer into your code.


Thanks,

Bill

*From:*owner-nmus...@globomaxnm.com 
 
[mailto:owner-nmus...@globomaxnm.com] *On Behalf Of *Katrin Volz

*Sent:* Monday, July 27, 2015 10:46 AM
*To:* nmusers@globomaxnm.com 
*Subject:* [NMusers] Time 'T' in $Error

Dear NMuser,

I would like to model a PD effect using a function which includes 
continuous time ‘ T ’.


The PD Effect is coded in $Error.

In general like this:

*$ERROR*

MyFunction = xxx + T

PDEffect = Basline * MyFunction

But ‘T’ cannot be used in $Error.

I tried to code the function with ‘T’ in $DES and rename it to use it 
in $ERROR:


*$DES*

MyFunction = xxx + T

*$ERROR*

MyNewFunction = MyFunction

PDEffect = Baseline * MyNewFunction

but it didn’t worked (to rename was not permitted by NONMEM).

Does anyone of you has an idea how to solve this problem?

Thanks and best regards.

Katrin

Anke-Katrin Volz

PhD Student

Saarland University

Clinical Pharmacy

Campus C2 2 | Room 0.31

D- 66123 Saarbrücken

mail ankekatrin.v...@uni-saarland.de 



fon +49 [0]681 302 22 84

fax +49 [0]681 302 70 258

www.clinicalpharmacy.me 

RE: [NMusers] Implementation of the OMEGA covariance matrix

[Rudy Gunawan]

Hi Ana,

What you can do is rearrange your ETA order, ie, the full block and the rest as 
with a separate '$OMEGA'
$OMEGA BLOCK(3)
0.09; V1
0.01 0.09   ;BMAX
0.01 0.01 0.09  ;KM 
$OMEGA ...  

Hope this helps,

Rudy

-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Ana Miranda Bastos
Sent: Tuesday, July 28, 2015 5:52 AM
To: nmusers@globomaxnm.com
Subject: [NMusers] Implementation of the OMEGA covariance matrix

Dear NMusers,

I wonder if someone could help me to implement the OMEGA covariance matrix 
below.

$OMEGA  
 0.09  ;   PPV_VMAX
 0.00 0.09  ; PPV_V1
 0.00 0.00 0.09  ; PPV_V2
 0.00 0.00 0.00 0.09  ; PPV_KD
 0.00 0.01 0.00 0.00 0.09  ;   PPV_BMAX
 0.00 0.01 0.00 0.00 0.01 0.09  ; PPV KM


I don't want to fix any value to zero, but I don't know how to better 
illustrate that I would like to consider a covariance between:
. V1 and Bmax 
. Bmax and Km
. V1 and Km


Any advice is highly appreciated. Thank you in advance.

Kind regards,

Ana



--
Ana Bastos, Pharm, MSc, PhD student
Pharmacologie cellulaire et moléculaire
(Cellular and Molecular Pharmacology Unit)
Louvain Drug Research Institute
Université catholique de Louvain (Catholic University of Louvain)
UCL 7370 avenue E. Mounier 73
1200 Bruxelles, Belgique
--
http://www.facm.ucl.ac.be

RE: [NMusers] Implementation of the OMEGA covariance matrix

[Krekels, E.H.J.]

Dear Ana,

If you only want to consider covariance between 3 parameters, I suggest you put 
these three in an $OMEGA BLOCK(3) and list the remaining OMEGAs separately 
outside the block.

Best,
Elke

-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Ana Miranda Bastos
Sent: Tuesday, July 28, 2015 2:52 PM
To: nmusers@globomaxnm.com
Subject: [NMusers] Implementation of the OMEGA covariance matrix

Dear NMusers,

I wonder if someone could help me to implement the OMEGA covariance matrix 
below.

$OMEGA  
 0.09  ;   PPV_VMAX
 0.00 0.09  ; PPV_V1
 0.00 0.00 0.09  ; PPV_V2
 0.00 0.00 0.00 0.09  ; PPV_KD
 0.00 0.01 0.00 0.00 0.09  ;   PPV_BMAX
 0.00 0.01 0.00 0.00 0.01 0.09  ; PPV KM


I don't want to fix any value to zero, but I don't know how to better 
illustrate that I would like to consider a covariance between:
. V1 and Bmax 
. Bmax and Km
. V1 and Km


Any advice is highly appreciated. Thank you in advance.

Kind regards,

Ana



--
Ana Bastos, Pharm, MSc, PhD student
Pharmacologie cellulaire et moléculaire
(Cellular and Molecular Pharmacology Unit)
Louvain Drug Research Institute
Université catholique de Louvain (Catholic University of Louvain)
UCL 7370 avenue E. Mounier 73
1200 Bruxelles, Belgique
--
http://www.facm.ucl.ac.be

[NMusers] RE: Is it reasonable to add covarite to PD parameters kin or kout?

[Dodds, Mike]

Li,

It's worth pointing out that the typical IDR formulation assumes proportional 
effect on response.  That is, if the drug effect is inhibition of synthesis 
(e.g Imax=0.5), a subject with baseline 100 will go to 50 and a subject with 
baseline 1000 will go to 500.  If your data is suggesting that a subject with 
baseline 100 goes to 50 and a subject with baseline 1000 goes to 950, the 
proportional response formulation is probably a mismatch.

Introduction of baseline as a covariate (in this case, with a large negative 
coefficient?) would align the model with data, but you may instead look at 
parameterizing the drug effect component of the model differently.

Warm Regards,
Mike

From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Mats Karlsson
Sent: Tuesday, July 28, 2015 3:15 AM
To: Peiming Ma; Mark Sale; Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Hi,

There are both pros and cons of using baseline observations as a covariate. We 
presented four different options for how to handle baseline and investigated 
estimation properties in the paper below. What we didn't look at, because it is 
a very bad idea, is to use an baseline observation both as a covariate and 
dependent variable.

Approaches to handling pharmacodynamic baseline responses.
Dansirikul C, Silber HE, Karlsson MO.
J Pharmacokinet Pharmacodyn. 2008 Jun;35(3):269-83.

Best regards,
Mats


Mats Karlsson, PhD
Professor of Pharmacometrics

Dept of Pharmaceutical Biosciences
Faculty of Pharmacy
Uppsala University
Box 591
75124 Uppsala

Phone: +46 18 4714105
Fax + 46 18 4714003
www.farmbio.uu.se/research/researchgroups/pharmacometrics/

From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Peiming Ma
Sent: Tuesday, July 28, 2015 3:31 AM
To: Mark Sale; Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Dear Mark and Li,

I think there is nothing wrong with using baseline as a covariate. Baseline 
observations contain more information than just some pre-treatment observations 
and thus have a lot of explanatory power; this is the reason they appear 
significant more often than not. Comparisons after treatments should try to use 
all information before treatments (thus including baselines).

Statisticians habitually use baselines as covariates in their ANCOVA models for 
obviously good reasons.

Regards,
Peiming

From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Mark Sale
Sent: Tuesday, July 28, 2015 3:34 AM
To: Zhao,Li; nmusers@globomaxnm.com
Subject: [NMusers] RE: Is it reasonable to add covarite to PD parameters kin or 
kout?

Li,
I'm going to go out on a limb (since I haven't seen your data or understand the 
biology) and suggest that the baseline value is not a covariate, but is just 
another observation (presumably with drug concentration = 0).  The baseline 
value is sort of by definition a function of kin.  So, you might look for 
predictors of Kin (age? Disease stage?) rather than using the observed value to 
predict the parameters (which then predict the observed value).  Also note that 
covariates are, by definition, measured without error, and a baseline value is 
likely measure with error (the same error as any other observation?).

Mark



From: owner-nmus...@globomaxnm.com 
[mailto:owner-nmus...@globomaxnm.com] On Behalf Of Zhao,Li
Sent: Monday, July 27, 2015 2:39 PM
To: nmusers@globomaxnm.com
Subject: [NMusers] Is it reasonable to add covarite to PD parameters kin or 
kout?


Dear NMusers,



Right now I am doing covarite analysis for an indirect response model.



 I tested a few potential covarites and found it's STATISTICALLY significant if 
I add observed baseline value to kin.



But I am not sure if it makes sense to add the baseline value as a covarite to 
kin.



Could you please help me if you have had similar experiences before?



Thank you very much!