Limited sampling approach for model-informed precision dosing of daptomycin to rapidly achieving the target area under the concentration-time curve: A simulation study

Abstract

Daptomycin, an anti-methicillin-resistant Staphylococcus aureus drug, causes exposure-dependent muscle toxicity and eosinophilic pneumonia. Although the area under the concentration-time curve (AUC)-guided dosing is crucial, an optimal blood sampling strategy is lacking. This study aimed to identify an optimal limited sampling strategy using Bayesian forecasting to rapidly achieve the target AUC. Two validated population pharmacokinetic models generated a virtual population of 1000 individuals (models 1 and 2 represent diverse patients and kidney transplant recipients, respectively). The AUC for each blood sample was assessed using the probability of achieving the estimated/reference AUC ratio on the second day (AUC_24–48) and at the steady state (AUC_ss). In Model 1, Bayesian posterior probabilities for AUC_24–48 increased from 50.7% (a priori) to 59.4% and for AUC_ss from 48.9% (a priori) to 61.9%, with one-point C_trough sampling at 24 h. With two-point sampling at 7 and 24 h, the probabilities increased to 73.8% for AUC_24–48 and 69.7% for AUC_ss. In Model 2, the probabilities for both AUC_24–48 and AUC_ss with one-point C_trough or two-point sampling incorporating C_trough sampling increased compared to a priori probabilities. These results suggest that two-point sampling incorporating C_trough during initial dosing enhanced achieving the target AUC_24–48 and AUC_ss rapidly.