Objectives: Milrinone is an inotropic agent with vasodilating properties used in the treatment of ventricular dysfunction. Milrinone is predominantly eliminated by the kidneys and accumulates in the setting of acute kidney injury (AKI). The purpose of this study was to evaluate milrinone pharmacokinetics in children with AKI with or without continuous renal replacement therapy (CRRT).
Methods: Retrospective collection of milrinone therapeutic drug monitoring data in patients with AKI, including those requiring CRRT, through chart review from January 2008 to March 2014. Pharmacokinetic (PK) data were analyzed by Bayesian estimation using a pediatric population PK model (MW/Pharm). Clearance estimates were allometrically scaled to body weight.
Results: Data on 11 patients were available for analysis. Three patients required CRRT. Milrinone concentrations during continuous infusion varied 30-fold and ranged from 44 to 1343 ng/mL. Of the 33 samples obtained in 11 patients, 24 were outside the target range (72.7%), with 16 (48.5%) above and 8 (24.2%) below. Patients with AKI had significantly lower milrinone clearance (4.72 ± 2.26 L/h per 70 kg) compared with published data in patients without AKI. There was large between-patient variability in milrinone clearance (range: 2.91-13.6 L/h per 70 kg). Clearance in patients on CRRT ranged from 2.8 to 7.19 L/h per 70 kg. A significant correlation between milrinone clearance and estimated creatinine clearance was observed (r = 0.70, P = 0.0097). Allometrically scaled milrinone clearance was lower in the youngest patients (younger than 2 years), suggestive of ongoing renal maturation and existing AKI.
Conclusions: Pediatric patients with AKI have significantly lower milrinone clearance compared with published data in patients without AKI. Large variability was noted in milrinone concentrations, and they were frequently outside the target range. The large between-patient variability in milrinone concentrations suggests that dosing regimens should be individualized in this population of critically ill patients. Evaluation of PK model-based milrinone dose optimization and the use of biomarkers as predictors of changes in clearance warrant further study.