Linezolid Population Pharmacokinetics to Improve Dosing in Cardiosurgical Patients: Factoring a New Drug–Drug Interaction Pathway
Manjunath P. Pai, Pier Giorgio Cojutti, Valentina Gerussi, Paola Della Siega, Carlo Tascini, Federico Pea
Abstract
BACKGROUND: Linezolid-induced myelosuppression limits optimal therapy in cardiosurgical patients with deep-seated infections at current doses. METHODS: Adult patients who received a cardiac surgery intervention and linezolid for a documented or presumed serious gram-positive infection were evaluated. Therapeutic monitoring data, dosing, concomitant medications, and other pertinent laboratory data were collected retrospectively. A population pharmacokinetic model was constructed to identify covariates and test potential drug-drug interactions that may account for interpatient variability. Simulations from the final model identified doses that achieve a target therapeutic trough concentration of 2-8 mg/L. RESULTS: This study included 150 patients (79.3% male) with sepsis and hospital-acquired pneumonia in 71.7% as the primary indication. The population had a median (minimum-maximum) age, body weight, and estimated glomerular filtration rate (eGFR) of 66 (30-85) years, 76 (45-130) kg, and 46.8 (4.9-153.7) mL/minute, respectively. The standard linezolid dosage regimen achieved the therapeutic range in only 54.7% of patients. Lower-than-standard doses were necessary in the majority of patients (77%). A 2-compartment Michaelis-Menten clearance model with weight, kidney function, and the number of interacting drugs identified as covariates that best fit the concentration-time data was used. Cyclosporine had the greatest effect on lowering the maximum elimination rate (Vmax) of linezolid. Empiric linezolid doses of 300-450 mg every 12 hours based on eGFR and the number of interacting medications are suggested by this analysis. CONCLUSIONS: Lower empiric linezolid doses in cardiosurgical patients may avoid toxicities. Confirmatory studies are necessary to verify these potential drug interactions.