Search the PBPK Model Repository

  Absorption Model

• First-Order (dose and food-dependent fa – saved in different models)

  Volume of Distribution

• Full PBPK (Method 2)
• Notes: Includes a Kp scalar and Kpadipose

  Route of Elimination

• CYP3A4 (80%), CYP2C9 (10%), CYP2C19 (10%)

  Perpetrator DDI

• CYP3A4 Inhibitor

  Validation

• Two clinical studies with fasted and fed groups at varying dose levels describing single and multiple dose exposure of piperaquine were used to verify the PBPK model. All of the simulated studies were within 1.5-fold of the observed values. 
• A clinical DDI study where piperaquine was the victim of a CYP3A4-mediated DDI was accurately recovered using the PBPK model as well as a CYP3A4 perpetrator DDI with the sensitive substrate midazolam.

  Limitations

• Requires separate files for low and high dose due to dose-dependant fa​
• Cmax overprediction, likely due to formulation differences​
• Additional verification for DDIs would be ideal although studies are currently not available in literature

  Updates in V19

• Updated in vitro­ data
• LogP
• Converted model to full PBPK with Vss predicted through Method 2

  Absorption Model

• First-Order, fa and ka predicted by Caco-2 data

  Volume of Distribution

• Minimal (optimized to IV data)

  Route of Elimination

• fm_CYP3A4 = 80, fm_CYP2D6 = 20

  Perpetrator DDI

• Inhibition of CYP2D6

  Validation

• Model performance was verified in healthy volunteers and psychiatric patients. Two clinical studies with IV administration (7 to 10 mg) and ten clinical studies with oral administration (25 to 200 mg) were used for model verification. Simulations for eight of eleven clinical studies with a reported AUC were within 2-fold of the observed value.
• The fm_CYP1A2 was verified through simulations of chlorpromazine in smokers and non-smokers. The fm_CYP2D6­ was verified through simulations of chlorpromazine coadministered with and without quinidine.

  Limitations

• Model was developed using 10 mg IV and 100 mg PO. Use of 7-10 mg IV dosing and oral doses of 100-200 mg were verified, but predictions of oral doses <100 mg PO are overestimated.
• Model does not include P-gp efflux.
• Model is not verified for use in perpetrator DDI simulations with CYP2D6 substrates.

  Absorption Model

First-Order

  Volume of Distribution

• Full PBPK (Method 2)

Note: A Kp scalar (0.5) was used in the model along with optimized partitioning into adipose tissue (Kp,adipose = 0.5) to recover the clinical observed data. 

  Route of Elimination

• CYP2B6 and CYP3A4 (non-linear kinetics); incorporates autoinduction of CYP2B6

  Perpetrator DDI

• Induction of CYP2B6

  Validation

• Two clinical studies describing single dose exposure and two describing multiple dose exposure of artemether were used to verify the PBPK model.  The single dose exposures were within 1.5-fold of observed for both studies. The multiple dose exposures were slightly over-predicted at 2.02 and 2.63-fold for the two studies.  Clinical DDI studies with ketoconazole, rifampicin and efavirenz where artemether was the victim of CYP3A4 (and CYP2B6 for efavirenz)-mediated DDIs were accurately recovered (within 1.25-fold) using the PBPK model.  A clinical DDI study with efavirenz, where artemether was the perpetrator of a CYP2B6-mediated DDI was accurately recovered (within 1.25-fold) using the PBPK model. 

  Limitations

• The tendency towards over-prediction of artemether exposure upon multiple dosing could indicate a greater extent of induction is required. However, any increase in induction potency resulted in under-prediction of single dose exposure, which is of greater importance for the therapeutic effect of artemether.

  Updates in V19

• Updated in vitro­ data
  • fu: 0.083 -> 0.038
  • B:P: 1.7 -> 1.1
• Optimized ka and t_lag
• Converted from minimal PBPK model to full PBPK model
  • Optimized CYP2B6 IndC₅₀