3 replies to this topic

[Helmut Schütz]

Hi!

 

Any ideas to set up the following model: 1comp, but the input rate is not a constant (zero order), but decreases with time following a simple exponential (like Dose/time=A·exp(-B·t). Background: I try to model formulations (like patches, implants), where the release rate decreases with time. Some of these formulations deliver at a constant rate and therefore can be modeled like an infusiuon, but others not...

[Simon Davis]

I would suggest to consider this as a first-order model with extravascular input.

 

in your formula, the B that's being estimated is Ka; the A is bioavailability.

 

  If there is also a "bolus" component, that can go directly into the central compartment.

 

I hope that makes sense, Simon.

[Helmut Schütz]

Hi Simon!

I would suggest to consider this as a first-order model with extravascular input.
in your formula, the B that's being estimated is Ka; the A is bioavailability.

Nice to call what I posted a 'formula' ;-)

I was thinking (in conventional WNL code) of something like:

f = kin*exp(-kdeg*x)/(V*k10)*(1-exp(-k10*x))

which would reduce to

f = kin/(V*k10)*(1-exp(-k10*x))

if kdeg approaches zero.

[Simon Davis]

The suggestion is not to use a closed form solution but to write derivs and choose non-stiff solver to start. e.g.

Derivs for compartments, same as whatever regular model you'd choose.

Stparm(ka = ka0*exp(-kae*t))

Please let us know how that works out for your problem. DO you perhaps have a snippet of data you could share?

Simon