Tamara van Donge

Chapter 4 70 from robust in vivo interventional trials to support such higher targets for improved patient outcomes are lacking. The potential benefit of continuous infusions is being discussed for many beta-lactams, especially for critically ill patients. 24 Given the lack of data to support a positive clinical impact of this approach and since the target pathogens with highest MIC for amoxicillin (e.g. enterococci), account for a relatively small proportion of culture confirmed neonatal sepsis episodes, it remains unclear that routine use of continuous amoxicillin infusion at the empiric treatment stage is justifiable. 14 Practically speaking, intravenous access in this population is often precarious, central venous lines are avoided whenever possible. Presumably, the rationale for higher PKPD targets used in clinical practice than supported by in vitro data is that beta-lactams are considered extremely safe because of their wide therapeutic window. Data linking high levels of beta-lactam exposure to neurotoxic events in vulnerable populations, however, are accumulating. 25-28 Shaffer and colleagues described ampicillin-induced neurotoxicity in very low birth weight neonates. 22 Of note, the investigation of neurotoxicity in this population is challenging because acute symptoms may remain unnoticed due to their subtle nature. Furthermore, both short and long-term neurotoxic effects of amoxicillin can be hard to distinguish from effects of the disease itself. 28 We were unable to identify any additional robust data on potential neurotoxic effects of aminopenicillins when administered to neonates, including which parameter is relevant for toxic effects (Cmax, area under the curve, time above concentration). The fact that poor neurodevelopmental outcomes are seen not only in infants with culture-proven sepsis, but also in culture-negative infants exposed to antibiotic treatment when compared to unexposed infants, underscores the importance of considering the contribution of antibiotics, including commonly used beta-lactams. 29,30 We therefore hypothesize that clinically used amoxicillin doses may result in toxicity in some exposed neonates, representing an upper limit of its therapeutic window. Limitations of this study should be recognized. Regarding empirical treatment when the pathogen is unknown, combination therapy with an aminoglycoside is the first line treatment of choice with the aim to broaden coverage of the spectrum of the pathogens. Aminoglycosides target mainly Gram-negative bacteria and staphylococci, and are not appropriate for other Gram-positive bacteria. Evidence on synergistic antibacterial evidence is lacking, and only in a few situations (e.g. endocarditis) are robust data available to support this effect. Only few population PK models for amoxicillin in preterm and term neonates exist. 5,6,8 The model of Tang et al. was preferred since it was developed with prospectively collected data originating from a multi-centre study and extensively