Toward predicting CYP2D6-mediated variable drug response from CYP2D6 gene sequencing data

Area of Science:

• Pharmacogenomics
• Personalized Medicine
• Enzyme Kinetics

Background:

• Pharmacogenomics aids personalized medicine by tailoring drug treatment based on genetic profiles.
• Current clinical practice uses categorical CYP450 allele assignments to predict enzyme activity, but this leaves significant drug response variability unexplained.

Purpose of the Study:

• To develop and validate a novel approach for predicting CYP2D6 enzyme activity using continuous genetic data.
• To improve the accuracy of predicting interindividual variability in drug response compared to conventional methods.

Main Methods:

• Full CYP2D6 gene sequences were obtained using long-read amplicon-based sequencing.
• A neural network was trained using patient data (n=561) on tamoxifen metabolism to predict CYP2D6 activity.
• In vitro experiments using human embryonic kidney cells confirmed the impact of genetic variants on CYP2D6 substrate metabolism.

Main Results:

• The continuous-scale model explained 79% of CYP2D6 activity variability, surpassing the 54% explained by the conventional *-allele approach.
• The model accurately assigned activities to known alleles and predicted effects of uncharacterized variant combinations.
• Results were replicated in independent cohorts for tamoxifen and venlafaxine, showing improved prediction accuracy (R² adjusted 0.66 vs. 0.35 for tamoxifen; 0.64 vs. 0.55 for venlafaxine).

Conclusions:

• A continuous-scale genotype assignment significantly enhances the prediction of CYP2D6 enzyme activity.
• This approach offers a more accurate prediction of individual drug response, advancing personalized medicine.
• The findings support the clinical utility of comprehensive genetic profiling for optimizing drug therapy.