Predicting Pharmacokinetics of Drugs in Patients with Heart Failure and Optimizing Their Dosing Strategies Using a Physiologically Based Pharmacokinetic Model.

Background: Heart failure (HF), as the end stage of various cardiac diseases, alters blood flow to key organs responsible for drug clearance. This can lead to unpredictable and often suboptimal drug exposure, creating a critical need for quantitative tools to guide precise dosing in this vulnerable population. Methods: This study aimed to establish a whole-body physiologically based pharmacokinetic (PBPK) model for characterizing drug pharmacokinetics in both healthy subjects and patients across the HF severity spectrum. Eight commonly used drugs (digoxin, furosemide, bumetanide, torasemide, captopril, valsartan, felodipine and midazolam) for treating HF and its comorbidities were selected. Following successful validation against clinical data from healthy subjects, the PBPK model was extrapolated to HF patients. Pharmacokinetics of the eight drugs in 1000 virtual HF patients were simulated by replacing tissue blood flows and compared using clinical observations. Results: Most of the observed concentrations were encompassed within the 5th-95th percentiles of simulated values from 1000 virtual HF patients. Predicted area under the concentration-time curve and maximum plasma concentration fell within the 0.5~2.0-fold range relative to clinical observations. Sensitivity analysis demonstrated that intrinsic renal clearance, unbound fraction in blood, muscular blood flow, and effective permeability coefficient significantly impact plasma exposure of digoxin at a steady state. Oral digoxin dosing regimens for HF patients were optimized via the validated PBPK model to ensure that steady-state plasma concentrations in all HF patients remain below the toxicity threshold (2.0 ng/mL). Conclusions: A PBPK model was successfully developed to predict the plasma concentration-time profiles of the eight tested drugs in both healthy subjects and HF patients. Furthermore, this model may also be applied to guide digoxin dose optimization for HF patients.