Jump to content


Photo

Transdermal PK profile modeling


  • Please log in to reply
11 replies to this topic

#1 Seokhyun Hong

Seokhyun Hong

    Newbie

  • Members
  • Pip
  • 5 posts

Posted 22 September 2014 - 09:22 AM

Hello.

I'd like to fit single dose transdermal PK profile using Phoenix model.

Basically, I used "Transdermal example of nicotine patch" text model with some modification

 

test(){

    deriv(A1 = Frate + Srate - Ke * A1 - K12 * A1 + K21 * A2)

    deriv(A2 = K12 * A1 - K21 * A2)

  Dose = 10000

  Frate = (t <= Ftau ? ((Dose-SDose)/Ftau) : 0)

  Srate = (t <= Stau ? (SDose/Stau) : 0)

    C = A1 / V

    error(CEps = 1)

    observe(CObs = C * (1 + CEps))

  stparm(Ftau = tvFtau * exp(nFtau))

  stparm(Stau = tvStau * exp(nStau))

  stparm(SDose = tvSDose * exp(nSDose))

    stparm(V = tvV * exp(nV))

    stparm(Ke = tvKe * exp(nKe))

    stparm(K12 = tvK12 * exp(nK12))

    stparm(K21 = tvK21 * exp(nK21))

    stparm(Tlag = tvTlag * exp(nTlag))

    fixef(tvV = c(0, 384, ))

    fixef(tvKe = c(0, 0.15, ))

    fixef(tvK12 = c(0, 0.01, ))

    fixef(tvK21 = c(0, 0.9, ))

    fixef(tvTlag = c(0, 0.75, ))

  fixef(tvFtau = c(8, 12, 16))

  fixef(tvStau = c(16, 24, 28))

  fixef(tvSDose = c(0, 8000, ))

    ranef(diag(nV, nKe, nK12, nK21, nTlag, nFtau, nStau, nSDose) = c(1, 1, 1, 1, 1, 1, 1, 1))

}

 

 

But, Our PK profile has "Tlag" and I'd like to enter the Tlag as a parameter in above model.

How can I put the "Tlag" in the model?

Can you modify this model to Phoenix graphical model?

and How can I simulate multiple dose profile (Tau = 24hr) using this model?

Please give me advices

Thanks [file name=TDS_modeling.phxproj size=365000]http://www.pharsight.com/extranet/media/kunena/attachments/legacy/files/TDS_modeling.phxproj[/file]

Attached Files



#2 Samer Mouksassi

Samer Mouksassi

    Advanced Member

  • Members
  • PipPipPip
  • 90 posts

Posted 22 September 2014 - 03:50 PM

# one way is to take advantage of the dosepoint satement and its argument tlag

but then you need to map dosing information

or you can code this into your equation Frate equation

I prefer the dospoint style see below

 

 

 

test(){

    deriv(A1 = - Ke * A1 - K12 * A1 + K21 * A2)

    deriv(A2 = K12 * A1 - K21 * A2)

 

    dosepoint(A1,duration=tvFtau,tlag=tvTlag,bioavail=FREL)

    dosepoint2(A1,duration=tvStau,bioavail=1-FREL)

 

 

 # Dose = 10000

  #Frate = (t <= Ftau ? ((Dose-SDose)/Ftau) : 0)

  #Srate = (t <= Stau ? (SDose/Stau) : 0)

    C = A1 / V

    error(CEps = 1)

    observe(CObs = C * (1 + CEps))

  stparm(Ftau = tvFtau * exp(nFtau))

  stparm(Stau = tvStau * exp(nStau))

 

    stparm(V = tvV * exp(nV))

    stparm(Ke = tvKe * exp(nKe))

    stparm(K12 = tvK12 * exp(nK12))

    stparm(K21 = tvK21 * exp(nK21))

    stparm(Tlag = tvTlag * exp(nTlag))

    fixef(tvV = c(0, 384, ))

    fixef(tvKe = c(0, 0.15, ))

    fixef(tvK12 = c(0, 0.01, ))

    fixef(tvK21 = c(0, 0.9, ))

    fixef(tvTlag = c(0, 0.75, ))

  fixef(tvFtau = c(8, 12, 16))

  fixef(tvStau = c(16, 24, 28))

  fixef(FREL = c(0, 0.8,1 ))

    ranef(diag(nV, nKe, nK12, nK21, nTlag, nFtau, nStau, nSDose) = c(1, 1, 1, 1, 1, 1, 1, 1))

}



#3 Seokhyun Hong

Seokhyun Hong

    Newbie

  • Members
  • Pip
  • 5 posts

Posted 22 September 2014 - 04:41 PM

Dear smouksassi

I appreciate your kind reply.
I have an another question about dosepoint.
Follwing your advice,
How can I map dosing information? (A1 1, A1 1 rate, A1 2, A1 2 rate, time)
In case of total transdermal patch dose is 10 mg, How can I split the dose?

I'll be waiting for your reply.

Best regards.

shhong Posted Image

Attached Thumbnails

  • Dose.jpg


#4 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 22 September 2014 - 05:04 PM

Dear Shong

You can split the dose using this tyoe of code. I added a tlag to see the 2 splitted doses.

best

Serge

 

 

test(){

    deriv(A1 = - (Cl * C))

    urinecpt(A0 = (Cl * C))

    C = A1 / V

    dosepoint(A1, bioavail = (1-f), idosevar = A1Dose, infdosevar = A1InfDose, infratevar = A1InfRate)

 

# first dose fraction 1-f

    dosepoint2(A1, tlag = (tlag), bioavail = (f))

# second dose fraction f after tlag

    error(CEps = 1)

    observe(CObs = C + CEps)

    stparm(V = tvV)

    stparm(Cl = tvCl)

    stparm(f = ilogit(tvf))

    stparm(tlag = tvtlag * exp(ntlag))

    fixef(tvV = c(, 50, ))

    fixef(tvCl = c(, 5, ))

    fixef(tvf = c(, 0, ))

    fixef(tvtlag = c(, 6, ))

    ranef(diag(ntlag) = c(0.0001))

}



#5 Samer Mouksassi

Samer Mouksassi

    Advanced Member

  • Members
  • PipPipPip
  • 90 posts

Posted 22 September 2014 - 05:49 PM

as per the code Serge and I sent it is the FREL that is dividing the total dose of

10 000 to be split between the slow and fast rates.

 

in the mapping you just input time =0 and A1 1 = total dose = 10 000 and A1 2 = 10 000

and you leave rates empty.

 

now the total dose that will end up in the system =

(10000 * F) + (10000 *(1- F) ) = 10 000



#6 Seokhyun Hong

Seokhyun Hong

    Newbie

  • Members
  • Pip
  • 5 posts

Posted 23 September 2014 - 04:41 AM

Thanks for your reply.

 

I fitted our profile well.

then, I'd like to simulate multiple dose profile (tau = 24) using ADDL input option.

In case of split dose (infusion), How do I set the dosing option?

I attached project file.

Please give me advice again.

 

Best regards [file name=TDS_modeling-20140922.phxproj size=729271]http://www.pharsight.com/extranet/media/kunena/attachments/legacy/files/TDS_modeling-20140922.phxproj[/file]

Attached Files



#7 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 23 September 2014 - 06:04 AM

You got me on that one because both use the same dosepoint A1 and A11 and A12 are not dosepoints.

The simplest way is just to write all the doses in the dosing tab

 

A1 1 A1 1 Rate A1 2 A1 2 Rate Time

13300000 13300000 0

13300000 13300000 24

13300000 13300000 48

13300000 13300000 72

 

etc...

 

This will work for sure.

best

Serge



#8 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 23 September 2014 - 06:51 AM

I found a way to deal with ADDL if you just put extravascular input with infinite Ka, then it will work because the input dose is outside the central compartment.

 

Look at the project and input options and let me know if it is clear.

best

Serge [file name=split_dose_2.phxproj size=98671]http://www.pharsight.com/extranet/media/kunena/attachments/legacy/files/split_dose_2.phxproj[/file]

Attached Files



#9 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 23 September 2014 - 11:11 AM

I am still working on it. I will come back after the holidays we have in Israel right now.

What I sent you does not seem 100% correct but not sure yet.

best

Serge



#10 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 28 September 2014 - 10:43 AM

Dear Shhong

I made 3 versions of the model and all gave me the same simulation results. Please download the new Phoenix version that just came out.

The project with the new version is attached. If you still need the previous version, let me know and I will make the translation to have it working on the previous version too.

best

Serge [file name=TDS_modeling_simulation_new_version.phxproj size=1479226]http://www.pharsight.com/extranet/media/kunena/attachments/legacy/files/TDS_modeling_simulation_new_version.phxproj[/file]

Attached Files



#11 serge guzy

serge guzy

    Advanced Member

  • Members
  • PipPipPip
  • 485 posts

Posted 29 September 2014 - 10:14 AM

Dear Shhong I translated the 3 models to have them working on 1.3. It is attached. The only version we cannot make is the ADDL with split dose in the central compt. The 3 models are as follows: 1: ka infinite, 2 extravascular compts. ADDL works 2: split dose into central, all doses are explicitly shown in the data set. It works 3: Since the kinetic is linear, the principle of superposition applies and you can model the 2 input separately and add the two response to get the same profile as before. It works too. best Regards Serge [file name=split_dose_phoenix1point3.phxproj size=1104117]http://www.pharsight.com/extranet/media/kunena/attachments/legacy/files/split_dose_phoenix1point3.phxproj[/file]

Attached Files


Edited by Simon Davis, 18 February 2016 - 12:58 PM.


#12 Seokhyun Hong

Seokhyun Hong

    Newbie

  • Members
  • Pip
  • 5 posts

Posted 30 September 2014 - 01:32 AM

Dear Serge

 

I'm very thankful for your kind answers.

It was very helpful for me

 

Best regards.

 

Hong






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users