A Pharmacokinetic/Pharmacodynamic Model to Determine Optimal Dosing Targets for Amikacin in Neonatal Sepsis

Objectives: To develop a PK/PD model to explore the determinants of treatment failure (TF) with the use of amikacin in neonates with suspected or proven sepsis. Methods: A retrospective chart review was performed including all neonates treated with amikacin at Dunedin Hospital from Oct 2003 to Jan 2007. In these neonates, amikacin was commenced for treatment of suspected late onset neonatal sepsis. For those neonates with culture positive sepsis, TF was defined when an infant was diagnosised with a repeat infection caused by the same bacteria or a treatment switch to vancomycin due to no apparent clinical improvement. Amikacin E-tests were performed on the bacteria cultured from septic infants to determined amikacin minimum inhibitory concentrations (MIC). The population PK analysis included all the study subjects and was performed using NONMEM, version 5. The PD analysis included all the neonates with culture proven sepsis and used the posthoc estimates from the PK model, the MIC data and clinical characteristics in a logistic regression model using Stata, version 8. Results: The PK model analysed 358 amikacin concentrations from 80 preterm and term neonates. A one-compartment first order elimination model was utilised. The median postmenstrual age (PMA) for the infants was 30.29 weeks, ranging (24.71 to 44.14). Median (range) current weight (CWT) was 1.06 kg (0.45 to 4.43). The final covariate model estimated clearance (CL) = 0.23 • 0.0746 (PMA) • 0.691 (CWT) and volume of distribution (V) = 0.957 •0.89(CWT). The PD model included 35 confirmed septic episodes from 26 infants. Fourteen episodes met the TF criteria. Median (range) amikacin peak concentration was 27.7 (17.1 to 36.8) mg/L, MIC was 3 (\u3c0.5 to 12) mg/L and peak/MIC was 10.8 (2.3 to 71). In the PD model, two infants with peak/MIC \u3c6 had TF, relative risk (RR) 2.86 (1.57–5.19). Three out of four patients with peak/MIC \u3c8 had TF, RR 2.25 (0.95–5.34). There was no influence from area under the curve (AUC), gender, postnatal age (PNA) or peak and MIC as independent variables. Hence, other than peak/MIC \u3c8, there were no independent predictors of TF. Conclusions: The present study confirms the need for adequate amikacin peak concentrations to reduce the risk of TF. Therefore there is a need to develop a dosing regimen that reliably delivers a peak/MIC \u3e8 for septic neonates