Individualised dosing of amikacin in neonates

Abstract

Sir,We thank van den Anker and Allegaert for their commentson our recent paper (1) regarding individualised dosing ofamikacin in neonates. We are particularly pleased at therecognition given to the pharmacodynamic considerations,because this is where we consider the most mileage can tobe made in improving aminoglycoside drug regimens.We were indeed disappointed in our inability to individu-alisethedosingintervalusingpatient characteristics.Itiswellrecognised that in the newborn population there are wideinter-individual variations in the elimination characteristics ofaminoglycosides. We hypothesised that sepsis, or variousfeaturesofsepsis,mightaccountforthisvariationbutwerenotable to demonstrate a correlation for any of the features thatwe examined. We did not find that the creatinine levelspredicted elimination. With the addition of recent publishedevidence (2), we accept that the enzymatic method forcreatinine analysis is superior to the Jaffe method in theneonatal period. Nevertheless, in the sick neonate, renalclearance varies with changes in patient condition, limitingthe predictive value of an initial creatinine level, unless it isparticularly high. The most useful elimination to measure isthat of the aminoglycoside itself.In choosing the initial aminoglycoside dosing regimen,we considered that the most important aim was to achievean optimal therapeutic peak, and this approach was thebasis for our informed choice of dose. When it comes todosing interval, we accept the evidence that 24-h dosing ispreferable to more frequent dosing, but question the safetyof very extended dosing intervals because at some point thedosing interval will go beyond the postantibiotic effect andbacterial regrowth will occur, leading to treatment failure.Hence, our reluctance to recommend dosing intervalsbeyond 36 h. There are risks involved in complicatedtreatment regimens, and like most recommendations, oursuggestion for dosing intervals is in multiples of 12 h. Thishelps to avoid administration errors. We also recommendedthat a trough level taken 24 h after a dose can be helpful inindividualising the dosing interval for aminoglycosides,given the poor track record of other measures to adequatelypredict elimination. It may be safe to use standard dosingregimens in stable infants without measuring aminoglyco-side levels, but where there is any doubt about safety orefficacy, both peak and trough can be used to optimise thetreatment regimen.While ideally we want to unravel the relationship betweencovariates associated with aminoglycoside pharmacokinetics(PK)inneonates,weneeddosingregimensthatareaccessible