Bioequivalence and Bioavailability Forum

Dear John,

let’s see where you get a quicker answer – over here or to your cross-post at Pharsight’s Extranet.

❝ Can I compute macro constant for a one-compartment model?

Yes.*

❝ I used winnonlin to fit my IR profile and the best fit was a one compartment model with lagtime. […] I know how to figure out the macro constants from micro constants for a multicompartment model. Winnonlin's 2nd parameter output (from the modelling model) doesn't give out the macro constants for a one compartment model.

I know – that’s weird. Consider filing a request. Alternatively instead of using the built-in (precompiled) model modify #4 in PK2.lib (within seco … end) or define the secondary parameters in Phoenix.

Here’s the way to go (WinNonlin’s notation):

• C₀ = D/V_F
  
• A₁ = C₀[k₀₁/(k₀₁-k₁₀)]e^{k₁₀·t_lag} and α₁ = k₀₁
  
• A₂ = C₀[k₀₁/(k₀₁-k₁₀)]e^{k₀₁·t_lag} and α₂ = k₁₀
  
• t₀ = [ln(A₂) – ln(A₁)]/(k₀₁ – k₁₀); check: t₀ = t_lag?
  
• Check whether you get identical concentrations from
  WinNonlin: C = D/V_F·k₀₁/(k₀₁-k₁₀)·[e^{-k₁₀·(t-t_lag)}-e^{-k₀₁·(t-t_lag)}] and
  Macro-constants: C = Ae^-α₁t - Be^-α₂t

Use A₁ and α₁ (the part of the equation describing elimination) for deconvolution.

---

• JG Wagner
  Pharmacokinetics for the Pharmaceutical Scientist
  Technometrics Publishing, Lancaster 1993, pp 4–5 [sic]
  ISBN 1-56676-032-1