Population Pharmacokinetic-pharmacodynamic Model Analysis of Dapagliflozin for HbA1c-lowering Effects in Japanese Patients with Type 2 Diabetes Mellitus using Long-term Real-world Data.

Objectives: Dapagliflozin, a sodium-glucose cotransporter 2 inhibitor, has demonstrated population-level benefits in patients with various metabolic, cardiovascular, and renal comorbidities. However, significant inter-individual differences exist in plasma exposure and response to dapagliflozin. This study aimed to identify factors influencing the HbA1c-lowering effects of dapagliflozin using long-term real-world data and a population pharmacokinetic-pharmacodynamic (PK-PD) modeling approach. Methods: A PK-PD model was applied to analyze 415 plasma dapagliflozin concentrations and 508 HbA1c measurements from 85 patients with type 2 diabetes mellitus (T2DM) treated with dapagliflozin for one year. The long-term real-world data enabled the evaluation of treatment variability over time. Inter-individual variability in PK-PD parameters was assessed, and covariate analysis was performed to identify patient-specific factors affecting drug response. Results: HbA1c time profiles were well described using the PK-PD turnover model with an E_max function. Body weight significantly influenced the apparent clearance of dapagliflozin, though its clinical impact on systemic exposure was minimal. Long-term real-world data analysis revealed substantial inter-individual variability in HbA1c response. Conclusion: By integrating pharmacometric modeling with long-term real-world data, this study provided unique insights into the determinants of dapagliflozin efficacy in routine clinical practice. These findings highlight factors that may not be captured in short-term clinical trials. These findings emphasize the importance of individualized treatment strategies and suggest that future research should incorporate additional covariates, such as variations in glycemic response dynamics, to further refine dose optimization and personalized diabetes management.