Jump to content


Photo

Understanding loading dose in Nonparametric Superposition?

NPS nonparametric superposition loading dose

  • Please log in to reply
3 replies to this topic

#1 simon.hutchings

simon.hutchings

    Member

  • Members
  • PipPip
  • 13 posts

Posted 22 April 2021 - 03:06 PM

Hi Certara PK community!

 

I'm hoping someone has a simple answer for something that has me puzzled regarding NPS.  I'm using NPS to predict PK following 10 days of once-daily dosing of 800 mg based on PK data obtained from 10 days of once-daily dosing of 400 mg.  Please see profiles in attached image.

 

The yellow and purple curves are actual mean data obtained from Days 1 and 10 respectively), and the blue circles are Day 10 concentrations predicted from NPS.

As you will see from the inset text boxes, the only difference between the upper and lower graphs is the loading dose.  The t1/2 from NCA of the observed Day 10 data ranges from 5 to 12h, so based on the usual assumptions by Day 10 we should be at steady state.  On this basis the loading should have no effect on the Day 10 predicted profile, so I am confused as to why we see such a difference between the two graphs above, particularly when a 400 mg loading dose results in what looks like super-proportional increase in Day 10 exposure?!

Any help very gratefully received  :)


Edited by simon.hutchings, 22 April 2021 - 03:07 PM.


#2 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 22 April 2021 - 03:35 PM

Hi Simon - I don't see the attached image, however if you can share the project i maybe there is an issue where your administered dose is out ? (if necessary you can email it confidentially to Support).

Also with NPS be careful you don't have an arterfact by having insufficient samples requested to capture Cmax.

NPS is good for  a quick 'look-see' but I generally prefer to use a simulation if at all possible.  If I do perform NPS I tend to use the individual profiels so I see the range or 'simulated' outcomes and I can always summarise later.

  Simon.



#3 simon.hutchings

simon.hutchings

    Member

  • Members
  • PipPip
  • 13 posts

Posted 23 April 2021 - 11:28 AM

Hi Simon - I don't see the attached image, however if you can share the project i maybe there is an issue where your administered dose is out ? (if necessary you can email it confidentially to Support).

Also with NPS be careful you don't have an arterfact by having insufficient samples requested to capture Cmax.

NPS is good for  a quick 'look-see' but I generally prefer to use a simulation if at all possible.  If I do perform NPS I tend to use the individual profiels so I see the range or 'simulated' outcomes and I can always summarise later.

  Simon.

 

Hmmm - strange... yes, will forward project file to support.

 

Thanks for the swift response



#4 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 26 April 2021 - 09:34 AM

Hi Simon. the issue is because if you don't explicitly state the administered dose (that which generated the initial concentrations), then the first dose is assumed to be that dose.

Hence you were getting a difference between loading at 400 and 800 mg. see my attached plot where admin was set to 400mg for both (purple and dotted black lines) will give identical profiles at steady state.

 

the lesson in my opinion is to always explicitly state the original (administered) dose.  I hope that makes sense ?  Screenshots in the attached Word doc.

 

Simon.

Attached Files







Also tagged with one or more of these keywords: NPS, nonparametric superposition, loading dose

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users