Comprehensive Review of Factors Influencing Intrasubject Variability in Bioequivalence Studies

Intrasubject variability (ISV) in bioequivalence (BE) studies remains a critical challenge in generic drug development; however, the mechanistic determinants of this variability are not yet fully characterized. This narrative review synthesizes evidence across physicochemical, pharmacokinetic, physiological, and study design domains using a unified mechanistic framework. Our analysis identified a mechanistic hierarchy. Low solubility (BCS Class II/IV) demonstrated a 4.1-fold higher incidence of highly variable drug (HVD) compared with high-solubility classes. First-pass metabolism was observed in 83% of HVDs but only 21% of non-HVDs, making it the strongest pharmacokinetic predictor of HVD status. Subject recruitment stratified by genotyping of CYP2D6 and CYP3A5 polymorphisms reduced ISV by up to 54% in drugs primarily metabolized by these enzymes. Replicate study designs, enhanced measurement precision, and fed conditions reduced intrasubject variability (ISV) in pharmacokinetic parameters; specifically, fed conditions showed a reduction in ISV in 79.4% of comparisons, whereas multiple-dose studies did not consistently reduce ISV. A cross-population comparison revealed that ISV in drugs metabolized via CYP enzymes was consistently higher in the Mexican population than in the South Korean population. ISV in Cmax exceeded that of AUC0-t in over 80% of products, indicating that variability in absorption rate is the primary contributing factor. A four-stage mechanistic framework, termed the Intrasubject Variability (ISV) Cascade Model, was developed to describe ISV: absorption variability generation (Stage 1), amplification through first-pass metabolism (Stage 2), modulation by study design (Stage 3), and translation into bioequivalence study design requirements (Stage 4). A proposed decision tree provides structured guidance for ISV risk assessment and optimization of BE study design.