A neonatal amikacin covariate model can be used to predict ontogeny of other drugs eliminated through glomerular filtration in neonates

Abstract

Purpose: Recently, a covariate model characterizing developmental changes in clearance of amikacin in neonates has been developed using birth bodyweight and postnatal age. The aim of this study was to evaluate whether this covariate model can be used to predict maturation in clearance of other renally excreted drugs.

Methods: Five different neonatal datasets were available on netilmicin, vancomycin, tobramycin and gentamicin. The extensively validated covariate model for amikacin clearance was used to predict clearance of these drugs. In addition, independent reference models were developed based on a systematic covariate analysis.

Results: The descriptive and predictive properties of the models developed using the amikacin covariate model were good, and fairly similar to the independent reference models (goodness-of-fit plots, NPDE). Moreover, similar clearance values were obtained for both approaches. Finally, the same covariates as in the covariate model of amikacin, i.e. birth bodyweight and postnatal age, were identified on clearance in the independent reference models.

Conclusions: This study shows that pediatric covariate models may contain physiological information since information derived from one drug can be used to describe other drugs. This semi-physiological approach may be used to optimize sparse data analysis and to derive individualized dosing algorithms for drugs in children.

Fig. 1

Covariate model of amikacin (16) which was applied to the other renally excreted drugs. The figure illustrates the model-based predicted amikacin clearance (CL) values versus birth bodyweight (bBW) for post-natal age of 0, 14 and 28 days with (grey) and without (black) co-administration of ibuprofen. Birth bodyweight reflects the antenatal maturation of the kidney, postnatal age is reflecting the postnatal maturation. Reproduced from [De Cock RF, Allegaert K, Schreuder MF, et al. Maturation of the glomerular filtration rate in neonates, as reflected by amikacin clearance. Clin Pharmacokinet 2012 Feb 1;51 (2): 105–17] with permission from Adis (© Springer International Publishing AG [2012]. All rights reserved.).

Fig. 2

Observed versus population predicted concentrations for the models using the amikacin covariate model (top) and the independent reference models (bottom) of netilmicin, vancomycin, tobramycin, gentamicin A and gentamicin B.

Fig. 3

Individual post hoc (black) and population predicted (grey) clearance values (l/h) versus the most predictive covariate, birth bodyweight (g), for the models using the amikacin covariate model.

Fig. 4

Results of the NPDE analysis for the models using the amikacin covariate model of (a) netilmicin, (b) tobramycin, (c) vancomycin, (d) gentamicin A and (e) gentamicin B. Left panel: Histograms of the NPDE distribution with the solid line representing a normal distribution as a reference, Middle panel: NPDE versus time (hours); Right panel: NPDE versus observed concentrations (mg/L).

Fig. 5

Individual and population predicted clearances for the models using the amikacin covariate model versus the independent reference models for netilmicin, vancomycin, tobramycin, gentamicin dataset A and gentamicin dataset B.

Fig. 6

Population clearance values for the models using the amikacin covariate model (black) and the independent reference models (dotted line) versus birth bodyweight for PNA 1 (above), 14 (middle) and 28 days (below) for netilmicin, vancomycin, tobramycin, gentamicin A and gentamicin B. For the clearance values obtained with the amikacin covariate model, the full study range of the amikacin dataset is used while for the independent reference models, the study range available for that particular dataset is used causing the differences seen in the length of both lines illustrating the population clearance values using both approaches. Moreover for tobramycin, no clearance values for to the independent reference model could be illustrated for PNA 14 and 28 days since no data were available.