Impact of POR and cytochrome b5 variability on the genotype-phenotype discordance of CYP2A6 and CYP2D6: A quantitative analysis in Chinese human livers

) derived via a bias-corrected in vitro-in vivo extrapolation approach. The results revealed significant phenotypic overlap, leading to poor prediction accuracy when relying solely on genotyping. For CYP2A6, the concordance rates were only 50.0%-58.3% for the ∗4 allele and 0% for the ∗9 allele at the microsomal level. Similarly, for CYP2D6, the 100TT genotype exhibited low prediction accuracy (2.4%-14.3%), and the 1661CC genotype showed a 100% discrepancy. Correlation analysis demonstrated that these discrepancies were significantly driven by variations in microsomal enzyme content and cytochrome P450 oxidoreductase activity (for CYP2A6), as well as cytochrome b5 (Cytb5) content (specifically for CYP2D6). In conclusion, genotyping alone is insufficient for precise clearance prediction. Integrating physiological covariates-specifically POR activity and Cytb5 content-into in vitro-in vivo extrapolation models is essential to resolve genotype-phenotype discordance and optimize CYP2A6- and CYP2D6-mediated drug dosing. SIGNIFICANCE STATEMENT: This study addresses a critical limitation in CYP2A6/CYP2D6 genotype-guided therapy: it quantifies their extensive phenotypic overlap (genotyping alone: 0%-58.3% reliable dosing). Its novel multilevel analysis finds cytochrome P450 oxidoreductase/Cytb5 as drivers, aiding precision dosing and genetic-only risks mitigation.