Search the PBPK Model Repository

 Absorption Model

•   First-Order

 Volume of Distribution 

•   Full PBPK (Method 2)

  Note: Kp scalar used

 Route of Elimination

•   CYP2C19, CYP3A4, renal clearance

 Perpetrator DDI

•   CYP2D6 Inhibitor

 Validation

• Proguanil and cycloguanil files were built using in vitro and clinical (Jeppersen et al., 1997) data
• 3 clinical studies describing single and multiple dose exposure of proguanil were used to verify the PBPK model. 66% of studies were within 2-fold, of which 33% were within 1.5-fold. 
• A clinical DDI study where proguanil was the victim of a CYP2C19-mediated DDI was accurately recovered using the PBPK model.  

 Limitations

• Prediction of proguanil exposure was complicated by not knowing the polymorphism classification of subjects in each study, hence the model performance was deemed acceptable using the criteria of being within 2-fold of observed.
• Verification needed for perpetrator DDI assessment as literature data is unavailable at this time
• With a large CL_Rcomponent and chemical relation to metformin, we hypothesise that proguanil may be a substrate for active transport in the kidney. However, owing to a lack of mechanistic information relating to active transport this cannot be built into the model.​

 Updates in V19

• Modification of fm values
• Model converted from minimal to full PBPK distribution model
• Updated CYP2D6 IC50

  Absorption Model

First-Order

  Volume of Distribution

Full PBPK (Method 2)

  Route of Elimination

Formed by CYP2C19, CYP3A4; unknown clearance mechanism

  Perpetrator DDI

• CYP2D6 Inhibitor

  Validation

• Proguanil and cycloguanil files were built using in vitro and clinical (Jeppersen et al., 1997) data
• 5 clinical studies describing single and multiple dose exposure of cycloguanil were used to verify the PBPK model. 60% of studies were within 2-fold, of which 40% were within 1.5-fold.
• A clinical DDI study where proguanil was the victim of a CYP2C19-mediated DDI was accurately recovered using the PBPK model.  

  Limitations

• Prediction of cycloguanil exposure was complicated by not knowing the polymorphism classification of subjects in each study, hence the model performance was deemed acceptable using the criteria of being within 2-fold of observed.
• Verification needed for perpetrator DDI assessment as literature data is unavailable at this time

  Updates in V19

• Recalculated fm values using the corrected dose administered in Jeppesen et al., 1997
• Previous version used paper calculation of total clearance which did not account for the weight of the salt in the 200 mg dose administered
• Model converted from minimal to full PBPK distribution model
• Updated in vitro data
• Updated CYP2D6 IC₅₀

 Absorption Model

First order (different absorption parameters for each formulation)

Minimal PBPK with Vsac and Q (Method 2)

 Route of Elimination

• CYP3A4 = 95%; Additional HLM = 5%

 Perpetrator DDI

• CYP3A4 Mechanism Based Inhibition

 Validation

The exposure of 1000mg BID saquinavir with 100 mg BID ritonavir regimen for hard gel were reasonably well recovered (3/3 within 2-fold). With the exception of the 1000 mg BID saquinavir with 100 mg BID ritonavir regimen for soft gel, the exposures of ritonavir-boosted regimens were well recovered (4/5 within 1.5-fold).

Ten clinical DDI studies where saquinavir (soft gel) was administered with either ritonavir, cimetidine, ketoconazole, rifampin, erythromycin, or rifabutin were used to verify the PBPK model of saquinavir as a victim. In comparison of predicted vs. observed AUC, 80% of the studies were within 2-fold.

Two clinical DDI studies where saquinavir (hard gel) was administered with either ritonavir or nelfinavir were used to verify the PBPK model of saquinavir (hard gel) as a victim. In comparison of predicted vs. observed AUC, 50% of the studies were within 2-fold.

Three clinical DDI studies where saquinavir was administered with either midazolam or rifabutin were used to verify the PBPK model of rifabutin (soft gel) as a perpetrator. In comparison of predicted vs. observed AUC, 100% of the studies were within 2-fold.

 Limitations

• The variability within studies has presented a significant challenge to developing a single model to recover all data.