Prediction Of Serum Digoxin Concentration Using Estimated Glomerular Filtration Rate In Thai Population

Area of Science:

• Pharmacokinetics and Pharmacodynamics
• Nephrology
• Cardiology

Background:

• Serum digoxin concentration (SDC) monitoring is essential for optimizing digoxin therapy, but may not be feasible in all clinical settings.
• Renal function significantly influences digoxin elimination, making accurate prediction of SDC vital for managing efficacy and toxicity.
• Various equations exist to estimate renal function, but their performance in predicting SDC using different pharmacokinetic models needs evaluation.

Purpose of the Study:

• To compare measured SDC with predicted SDC using creatinine clearance (CrCl) from the Cockcroft-Gault (CG) equation and estimated glomerular filtration rate (eGFR) from multiple equations (CKD-EPI, Re-MDRD4, Thai-MDRD4, Thai-eGFR).
• To assess the accuracy of these predictions within two established pharmacokinetic models: Sheiner's and Konishi's.
• To identify the most reliable methods for predicting SDC based on renal function parameters.

Main Methods:

• A retrospective study involving 124 patients with cardiovascular disease and a steady-state SDC between 0.5-2.0 mcg/L.
• Utilized CrCl (CG equation) and various BSA-adjusted eGFRs (CKD-EPI, Re-MDRD4, Thai-MDRD4, Thai-eGFR) as inputs for Sheiner's and Konishi's pharmacokinetic models.
• Analyzed discrepancies between measured and predicted SDC to evaluate prediction accuracy.

Main Results:

• In Sheiner's model, predicted SDC using CG and BSA-adjusted CKD-EPI equations showed no significant difference from measured SDC (p=0.669 and p=0.374, respectively).
• The CG, CKD-EPI, and Re-MDRD4 equations demonstrated better prediction accuracy with minimal errors for patients with serum creatinine ≥0.9 mg/dL.
• In Konishi's model, predicted SDC using CG and the studied eGFR equations were significantly lower than measured SDC (p<0.05).

Conclusions:

• The CG and BSA-adjusted CKD-EPI equations are recommended for predicting SDC within Sheiner's model, especially for patients with serum creatinine ≥0.9 mg/dL.
• Other studied eGFR equations tended to underestimate SDC in both Sheiner's and Konishi's models.
• Accurate prediction of SDC using appropriate renal function estimates is crucial for therapeutic drug monitoring when direct measurement is unavailable.