Fitting Volume as a dose-dependent parameter

Phoenix
1

My drug volume changes with dose due to TMDD (small molecule with slow off-rate) and I am looking for help how to set up the extravascular PK model that will fit a different volume for each dose level. My dose range covers 100X so I think it’s a good data set. I read i may need to set up volume as a covariate but am struggling with that. Thanks for any tips.

1 Like
2

Hi Chuck,
Here are a couple of suggestions from a colleague. Hope these help, please let us know how you get on, Simon.

If you prefer a single trend, you can drive V with a simple power law of the administered dose:

```

# In addition to your dose mapping, carry the administered dose as a covariate

fcovariate(DoseAmt) # say 10, 100, 1000

fcovariate(RefDose) # say 10

stparm(V = tvV * (DoseAmt/RefDose)^(dVdose) * exp(nV))

fixef(dVdose = c(, 0, ))          # start near 0; positive means V rises with dose

```

If you prefer more complex trends (i.e. for each dose level - bin), let V use indicator terms for each bin:

```

\# --- categorical covariate carrying the dose level (per dose) ---

\# Put the integer "DoseBin" (1..K) in the data, you may collect different not distinguishable dose levels at one group

fcovariate(DoseBin())

\# V gets K-1 indicator shifts; bin 1 is the reference

group(dV = exp(dV2 * (DoseBin==2)

                            + dV3 \* (DoseBin==3)

                            + dV4 \* (DoseBin==4)   # add/remove as needed)

  ))

stparm(V  = tvV \* dV \* exp(nV))

```

Set DoseBin at each dose record (e.g., 1 for 10 mg, 2 for 30 mg, 3 for 100 mg, etc.).