Open-Systems-Pharmacology / PK-Sim

PK-Sim® is a comprehensive software tool for whole-body physiologically based pharmacokinetic modeling
Other
107 stars 50 forks source link

Longstanding issue, unsure if intentional: DDI AUCR and CmaxR are inflated for compounds undergoing EHC #2303

Open prvmalik opened 2 years ago

prvmalik commented 2 years ago

When the victim drug undergoes biliary excretion and EHC.

prvmalik commented 2 years ago

Compare the AUCinf ratio by PK Analysis NCA from two separate simulations vs. the AUCR calculated in the tool.

sfrechen commented 2 years ago

@Yuri05 : how is AUCR / CmaxR calculated in PK-Sim? What is done in background?

Yuri05 commented 2 years ago

how is AUCR / CmaxR calculated in PK-Sim? What is done in background?

A second simulation is created (a copy of the original simulation) and all "DDI-Parameters" are then changed in the way that the corresponding DDI processes are disabled:

Code disabling DDI parameters https://github.com/Open-Systems-Pharmacology/PK-Sim/blob/ec53f4f8e519ac4ffc9ca064d4435620eca55fd0/src/PKSim.Core/Model/SimulationFactory.cs#L226-L241

Then the ratio is calculated based on the PK parameter in the Peripheral Venous Blood / Plasma:

prvmalik commented 2 years ago

Imatinib.zip Example file here (ITZ DDI), and also affects digoxin.

StephanSchaller commented 2 years ago

Yes, this fails if we have more than 1on1 inhibition, right?

sfrechen commented 2 years ago

A second simulation is created (a copy of the original simulation) and all "DDI-Parameters" are then changed in the way that the corresponding DDI processes are disabled:

  • EC50 = 1
  • k_kinact_half = 1
  • kinact = 0
  • Emax = 0
  • ki = ♾️
  • ki_u = ♾️
  • ki_c = ♾️

That sounds not really like the best implementation as interactions may still play a role at other levels. The best and most clean solution would actually be to put the dose(s) (and its correspoding volume of administration if applicable) of the perpetrator(s) to 0. In that sense, I would define perpetrators as all other compounds that are not metabolite of the drug for which the DDI ratio should be calculated.

Yuri05 commented 2 years ago

That sounds not really like the best implementation as interactions may still play a role at other levels. The best and most clean solution would actually be to put the dose(s) (and its correspoding volume of administration if applicable) of the perpetrator(s) to 0. In that sense, I would define perpetrators as all other compounds that are not metabolite of the drug for which the DDI ratio should be calculated.

As far as I remember this was the first implementation - but this resulted in problems. @msevestre Do you remember the details?

Yuri05 commented 2 years ago

s. also https://github.com/Open-Systems-Pharmacology/PK-Sim/issues/175

Old YT discussion ![grafik](https://user-images.githubusercontent.com/25061876/184341708-62f7dbff-46b3-4f98-a19d-57ee17199d01.png)
sfrechen commented 2 years ago

Here is a proposal:

  1. Check which compounds are administered (let's call them "parents")
  2. Only offer AUCR / CMAXR calculation button for these compounds
  3. If button is clicked, disable all administrations except those of the parent of which the botton is clicked
  4. Calculate AUC and CMAX for all compounds (not only parents) in the new simulation with the disabled administrations.
  5. Exclude compounds with an AUC of 0 and CMX of 0 in this new simulation for AUCR and CMAXR calculation.
  6. Calculate AUCR and CMAXR for remaining compounds (after step 5 exclusion)