Dynamic phenotype monitoring to prevent genotype–phenotype discrepancies in pharmacogenetic-guided drug therapy

Abstract

Background One of the limiting factors for the clinical application of pharmacogenomics (PGx) is phenoconversion, i.e., the dynamic discrepancy between genotype-predicted and actual drug-metabolizing phenotypes. Systemic inflammation, polypharmacy, transporter dysfunction, redox and mitochondrial stress, and epigenetic modification can rapidly alter cytochrome P450 (CYP) enzyme and transporter activity. This suppression can lead to unexpected poor metabolizing phenotypes, increased drug accumulation, and even an increased risk of drug-induced liver injury. Therefore, static PGx genotyping should be complemented by real-time functional biomarker monitoring to achieve more accurate, effective, and safe drug delivery. Methods This structured narrative review integrates a systematic literature search with expert-guided thematic synthesis. Mechanistic insights are supported by selectively included preclinical data. Candidate biomarkers—including miR-122, 4β-hydroxycholesterol, and GLDH—were identified through an iterative, criteria-based selection process that prioritizes mechanistic plausibility, clinical relevance, and favorable kinetics. Results Mechanistic analyses have implicated cytokine-mediated signaling pathways (IL-6/STAT3, NF-κB), nuclear receptor repression (PXR/CAR), proteasomal CYP degradation, miRNA-driven mRNA destabilization, and enzyme inactivation as causes of phenoconversion. Dysfunction of the transporters OATP, BSEP, and MRP2, particularly in SLCO and ABCC genetic variants, creates a dual intrahepatic bottleneck that exacerbates drug and metabolite accumulation. The biomarker matrix, which collectively considers 4β-hydroxycholesterol, miR-122, GLDH, M30, sCD163, and acute phase reactants (CRP/IL-6), provides a theoretical framework to explore early hepatocellular stress secondary to inflammation-mediated CYP suppression and phenoconversion, thereby serving as an investigative tool to model genotype–phenotype discordance and anticipate potential variations in drug exposure tolerance. Conclusion Within this hypothesis-generating framework, an integrated analysis of routine (CRP, transaminases), functional, and molecular biomarkers offers a novel strategy for interpreting clinically significant genotype–phenotype discordance. This approach shows the functional consequences of altered CYP activity rather than genetic predictions. Combining these multi-level parameters systematically reflects the main mechanistic domains: systemic inflammation-induced phenoconversion (CRP, IL-6), mitochondrial and oxidative stress (AST/ALT ratio, GLDH), early hepatocyte stress and apoptosis (miR-122, M30), and actual CYP3A4 metabolic capacity (4β-OHC). A PGx panel integrated with a dynamic biomarker could lead to safer, more effective, and adaptive drug dosing and reduced liver injury for high-risk patients.