Leveraging QSP Models for MIPD: A Case Study for Warfarin/INR.

Area of Science:

• Pharmacometrics
• Pharmacogenomics
• Systems Pharmacology

Background:

• Warfarin dosing is complex due to high inter-individual variability, often leading to suboptimal patient outcomes.
• Model-informed precision dosing (MIPD) offers a path to individualized warfarin therapy, but model generalizability is a concern.
• Quantitative systems pharmacology (QSP) models offer potential for improved extrapolation but require clinical validation for MIPD.

Purpose of the Study:

• To evaluate the predictive performance of a previously derived mechanistic warfarin/international normalized ratio (INR) model in the context of MIPD.
• To benchmark the QSP-derived model against an established empirical reference model using external clinical data.
• To assess the impact of genetic covariates (CYP2C9, VKORC1) on warfarin dosing predictions.

Main Methods:

• Utilized an external dataset of patient INR data during warfarin initiation.
• Assessed model accuracy and precision, comparing predictions to observed INR values.
• Evaluated covariate contributions and performance in inpatient versus outpatient settings.

Main Results:

• The QSP-derived warfarin/INR model demonstrated comparable predictive performance to the empirical reference model without specific calibration for warfarin initiation.
• Inclusion of CYP2C9 and/or VKORC1 genotypes significantly improved prediction quality, even after incorporating 4 days of INR data.
• Outpatient data showed higher unexplained variability, suggesting potential need for model adjustments during patient transition.

Conclusions:

• QSP-derived mechanistic models hold significant potential for application in MIPD of warfarin.
• These models provide a valuable complementary approach to traditional empirical model development.
• Pharmacogenetic information enhances the predictive accuracy of QSP-based warfarin dosing models.