Background: Current gentamicin dosing algorithms in adult populations target a high peak concentration (Cmax) assuring efficacy and a drug-free period (concentration <0.5 mg/L) preventing toxicity. In contrast, gentamicin-based regimens in neonatal sepsis often aim for lower peak levels and trough concentrations of 0.5-2.0 mg·L. The latter concentrations are associated with an increased risk of aminoglycoside-related toxicity. Therefore, the primary aim of this study was to assess the target attainment of a simple and practical dosing regimen designed to attain drug-free periods in newborns.
Methods: The study was of prospective observational design. Neonates admitted to a level II neonatal nursery diagnosed with (suspected) early-onset sepsis and commencing intravenous gentamicin therapy of 5 mg·kg every 36 hours were eligible for inclusion. Gentamicin dosing was guided by drug concentration monitoring targeting Cmax values >8 mg·L and estimated trough concentrations <0.5 mg·L. Relationships between body weight (BW), gestational age (GA), postnatal age, and pharmacokinetic parameters were analyzed using the Pearson correlation test, and univariate and multivariate logistic regression analyses were performed to identify covariates predictive of target attainment failure.
Results: A total of 184 patients were included. 90.4% of patients (n = 166) achieved a Cmax value >8 mg·L with a Cmin value <0.5 mg·L. Subsequently, significant correlations were found between GA and Cmax (r = 0.58, P < 0.001) between GA and Cmin (r = 0.44, P < 0.001), between BW and Cmax (r = 0.50, P < 0.001), and between BW and Cmin (r = 0.42, P < 0.001). Correlations between area under the curve (AUC) and GA (r = 0.064, P = 0.4), and between AUC and BW (r = 0.028, P = 0.7) were not significant. During multivariate analysis, only GA (P < 0.001) was retained as an independent predictor of underexposure.
Conclusions: Extended interval dosing of gentamicin resulted in high target attainment rates in neonates admitted to a level II unit. In line with previous reports, low GA and BW were predictive of subtherapeutic peak and toxic trough levels. The AUC, however, was unaffected by the interpatient variation in GA and BW. Since AUC-guided dosing is gaining interest worldwide, the latter finding deserves further exploration in other neonatal cohorts.